Professional Documents
Culture Documents
Blood Banking and Transfusion Medicine - Basic Principles and Practice 2nd Ed - C. Hillyer, Et Al., (Churchill-Livingstone, 2007) BBS PDF
Blood Banking and Transfusion Medicine - Basic Principles and Practice 2nd Ed - C. Hillyer, Et Al., (Churchill-Livingstone, 2007) BBS PDF
Ste 1800
Philadelphia, PA 19103-2899
All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means,
electronic or mechanical, including photocopying, recording, or any information storage and retrieval system,
without permission in writing from the publisher.
Permissions may be sought directly from Elsevier’s Health Sciences Rights Department in Philadelphia, PA, USA:
phone: (+1) 215 239 3804, fax: (+1) 215 239 3805, e-mail: healthpermissions@elsevier.com. You may also com-
plete your request on-line via the Elsevier homepage (http://www.elsevier.com), by selecting ‘Customer Support’
and then ‘Obtaining Permissions’.
Notice
Knowledge and best practice in this field are constantly changing. As new research and experience broaden
our knowledge, changes in practice, treatment and drug therapy may become necessary or appropriate.
Readers are advised to check the most current information provided (i) on procedures featured or (ii) by the
manufacturer of each product to be administered, to verify the recommended dose or formula, the method
and duration of administration, and contraindications. It is the responsibility of the practitioner, relying on
their own experience and knowledge of the patient, to make diagnoses, to determine dosages and the best
treatment for each individual patient, and to take all appropriate safety precautions. To the fullest extent of
the law, neither the Publisher nor the Editors assume any liability for any injury and/or damage to persons or
property arising out or related to any use of the material contained in this book.
The Publisher
Dr. Hillyer is a tenured professor in the Departments of Pathology and Pediatrics, as well as
the Division of Hematology/Oncology, Winship Cancer Institute, Emory University School of
Medicine. He serves as director of the Transfusion Medicine Program at Emory and oversees the
Emory University Hospital Blood Bank, the blood and tissue banks of Children’s Healthcare of
Atlanta, and the Emory Center for the Advancement of International Transfusion Safety. He is
an editor of three textbooks on transfusion medicine and an author of more than 120 articles
and chapters pertaining to transfusion, HIV, cytokines, and herpesviruses (most notably CMV).
Nationally recognized as an expert in hematology and blood transfusion, Dr. Hillyer is President
of AABB (2006–2007) and is a Trustee of the National Blood Foundation (NBF). He has been
awarded research funding from the NIH, CDC, NBF, and other agencies. He currently serves as
principal investigator of a program project grant, several R-series awards, the Emory site of the
NHLBI’s Transfusion Medicine/Hemostasis Clinical Trial Network, and REDS-II. He also is a
co-principal investigator of AABB’s contract with HHS to provide technical assistance to six
developing nations under the President’s Emergency Plan for AIDS Relief (PEPFAR). Dr. Hillyer
is an associate editor of Transfusion and part-time medical director of the American Red Cross
Southern Region. Dr. Hillyer is board certified in Transfusion Medicine, Hematology, Medical
Oncology, and Internal Medicine. He received his BS from Trinity College and his MD from
the University of Rochester School of Medicine, with postgraduate training and fellowships in
hematology-oncology, transfusion medicine, and bone marrow transplantation at Tufts–New
England Medical Center in Boston.
v
Dr. Silberstein is a tenured professor in the Department of Pathology, Harvard Medical School,
and a Senior Investigator at the CBR Institute for Blood Research. He serves as director of the Joint
Program in Transfusion Medicine, with responsibility for the blood and tissue programs at Boston
Children’s Hospital, the Brigham and Women’s Hospitals, and the Dana-Farber Cancer Institute.
Dr. Silberstein has recently created the Center for Human Cell Therapy at Harvard Medical School.
The goal of this innovative center related to transfusion medicine is to facilitate the translation
of proof-of-principle discoveries to clinical applications. Dr. Silberstein is editor of several texts,
including Hematology and the Handbook of Transfusion Medicine. He is a member of the edito-
rial boards of Blood and Transfusion. Dr. Silberstein is a highly respected physician-scientist well
known for his mentorship; he has trained more than 45 fellows with PhD and MD backgrounds
in transfusion medicine-related research. A leader and expert in transfusion medicine and hema-
tology, Dr. Silberstein’s research has focused on the immunology of B-cells and hematopoiesis,
leading to the publication of more than 75 papers and numerous book chapters and reviews.
Dr. Silberstein is board certified in Transfusion Medicine, Hematology, and Internal Medicine.
He received his Baccalaureate and MD degrees from the University of Leiden, the Netherlands,
and accomplished postgraduate training in Hematology/Oncology and Transfusion Medicine at
Tufts–New England Medical Center in Boston.
ABOUT THE EDITORS Dr. Ness is director of the Transfusion Medicine Division at The Johns Hopkins Hospital and pro-
fessor of Pathology, Medicine, and Oncology at The Johns Hopkins University School of Medicine.
For many years he also acted as CEO and medical director of the Greater Chesapeake and Potomac
Region of the American Red Cross Blood Services. Dr. Ness has served the AABB for a number
of years and was President in 1999. He served on the editorial board of Transfusion until named
Editor in 2003. Dr. Ness has been a member of the American Society of Clinical Pathologists
Board of Registry Blood Bank examination committee, and the FDA Blood Products Advisory
Committee, and he consults for many commercial and nonprofit organizations. He is the editor
of several textbooks on transfusion medicine and has published more than 150 articles. Dr. Ness’
research focuses on transfusion-related complications and has been funded by the NIH and
CDC. He was involved in the initial REDS program and now acts as consultant to REDS-II.
He serves as principal investigator for the Johns Hopkins site of the Transfusion Medicine/
Hemostasis Clinical Trial Network, funded by NHLBI. Dr. Ness is co-principal investigator of
a project funded by the REDS-II program to study donor virus epidemiology issues in China.
He has worked extensively in international blood safety initiatives in China, Thailand, Vietnam,
Botswana, and Nigeria. Dr. Ness received his undergraduate education at the Massachusetts
Institute of Technology and his MD degree from the State University of New York at Buffalo. His
postgraduate work includes residency in internal medicine at Johns Hopkins, fellowship train-
ing in hematology-oncology at the University of California, San Francisco, and a transfusion
medicine fellowship at Irwin Memorial Blood Bank in San Francisco.
Dr. Anderson is the Kraft Family Professor of Medicine at Harvard Medical School and serves as
chief of the Division of Hematologic Neoplasia, director of the Jerome Lipper Multiple Myeloma
Center, and vice chair of the Joint Program in Transfusion Medicine at Dana-Farber Cancer
Institute. Currently, Dr. Anderson is chair of the NCCN Multiple Myeloma Clinical Practice
Guidelines Committee, is a Cancer and Leukemia Group B Principal Investigator, and is on the
Board of Scientific Advisors of the International Myeloma Foundation. He has published more
than 300 original articles and 200 book chapters, and has edited multiple textbooks on mul-
tiple myeloma and transfusion medicine. He is a Doris Duke Distinguished Clinical Research
Scientist and has had long-term RO1, PO1, and SPORE funding from the NIH and other agencies.
Dr. Anderson has received numerous awards, including the 2001 Charles C. Lund Award of the
American Red Cross Blood Services, the 2003 Waldenstrom’s award for research in plasma cell
dyscrasias, the 2004 Johnson & Johnson Focused Giving Award for Setting New Directions in
Science and Technology, and the 2005 Robert A. Kyle Lifetime Achievement Award. Dr. Anderson
graduated from Johns Hopkins Medical School, trained in internal medicine at Johns Hopkins
Hospital, and completed hematology, medical oncology, and tumor immunology fellowships at
vi the Dana-Farber Cancer Institute.
Dr. Roback is a tenured associate professor in the Department of Pathology and Laboratory
Medicine at Emory University, associate director of the Emory Transfusion Medicine Program, and
co-director of the Emory University Hospital Blood Bank and Stem Cell Processing Laboratory.
Dr. Roback’s research focuses on human and animal models of CMV infection, emphasizing
approaches to accelerate and improve the antiviral immune response following hematopoietic
stem cell transplantation. He also is inventor or co-inventor of a number of novel devices and
methodologies for rapid pretransfusion blood testing. Dr. Roback’s investigations have been
funded by the NIH, CDC, NBF, and DOD. He is a co-principal investigator of the Emory site for
REDS-II. Dr. Roback has authored 40 peer-reviewed publications and invited reviews, as well as
16 book chapters. He teaches medical, residency, and graduate school courses and was recognized
for excellent clinical pathology teaching with the Golden Apple Award. An active member of a
number of AABB committees, Dr. Roback is editor-in-chief of the 16th edition of the AABB’s
Technical Manual, member of the editorial board for the journal Transfusion, and co-chair
of the NHLBI’s Global Blood Safety and Availability task force on future transfusion medicine
research initiatives. He is a Diplomate of the American Board of Pathology in Clinical Pathology
and Blood Banking and Transfusion Medicine. Dr. Roback received his Baccalaureate degree from
Johns Hopkins University and was awarded a PhD in experimental pathology and an MD from
the University of Chicago. He completed a postdoctoral research fellowship and anatomic pathol-
ogy residency training at Albert Einstein College of Medicine and subsequently completed clinical
pathology and transfusion medicine training at Emory University.
Contributors
Sharon Adams, MT, CHS (ABHI) Richard J. Benjamin, MS, MBChB, PhD
Supervisor, HLA Laboratory Chief Medical Officer
Department of Transfusion Medicine American Red Cross Biomedical Services
Warren G. Magnuson Clinical Center National Headquarters, Washington, D.C.
National Institutes of Health Assistant Professor of Pathology
Bethesda, Maryland, USA Joint Program in Transfusion Medicine
Harvard Medical School
Boston, Massachusetts, USA
Barbara Alving, MD, MACP
Professor of Medicine
Uniformed Services University of the Health Sciences Howard Benn, MD
Bethesda, Maryland, USA Chief Fellow, Department of Hematology/
Oncology
Seton Hall University School of Graduate Medical
Education
Kenneth C. Anderson, MD South Orange, New Jersey, USA
Chief, Division Hematologic Neoplasia
Director, Jerome Lipper Multiple Myeloma Center
Dana-Farber Cancer Institute
Kraft Family Professor of Medicine Ginine M. Beyer, MD
Harvard Medical School Department of Pathology
Boston, Massachusetts University of Maryland School of Medicine
Baltimore, Maryland, USA vii
James P. AuBuchon, MD
E. Elizabeth French Professor and Chair of Pathology Morris A. Blajchman, MD, FRCP
Dartmouth-Hitchcock Medical Center Professor, Pathology and Molecular Medicine
Lebanon, New Hampshire, USA Head, Transfusion Medicine Services, Hamilton
Regional Laboratory
Medical Director, Canadian Blood Services
Hamilton, Ontario, Canada
Nicholas Bandarenko, MD
Associate Professor of Pathology and Laboratory
Medicine
Transfusion Medicine Service Neil Blumberg, MD
University of North Carolina, Chapel Hill Director, Clinical Laboratories
Chapel Hill, North Carolina, USA Director, Transfusion Medicine
Professor of Pathology and Laboratory Medicine,
University of Rochester School of Medicine
Rochester, New York, USA
Jon Barrett, MBBCh, FRCOG, MD,
FRCSC
Associate Professor, Department of Obstetrics and
Gynecology, University of Toronto Mark E. Brecher, MD
Senior Investigator Maternal and Infant Research Professor, Department of Pathology and
Unit of Center for Research in Women’s Health Laboratory Medicine
Chief of Maternal Fetal Medicine, Sunnybrook and Director, Clinical Pathology
Women’s College Health Sciences Center University of North Carolina
Toronto, Ontario, Canada Chapel Hill, North Carolina, USA
CONTRIBUTORS
Hal E. Broxmeyer, PhD Chief Medical Officer and Vice President, Research
Distinguished Professor, Chairman and Mary and Medical Affairs
Margaret Walther Professor of Microbiology Cerus Corporation
and Immunology Concord, California, USA
Professor of Medicine
Scientific Director of the Walther Oncology Center
Indiana University School of Medicine Robert L. Crookes, MBChB
Indianapolis, Indiana, USA Medical Director
South African National Blood Service
Johannesburg, South Africa
Michael P. Busch, MD, PhD
Vice President, Research, Blood Systems, Inc.,
Scottsdale, Arizona Elizabeth E. Culler, MD
Director, Blood Systems Research Institute, Medical Director
San Francisco, California Blood Assurance, Inc.
Adjunct Professor, Department of Laboratory Chattanooga, Tennessee, USA
Medicine
University of California,
San Francisco, California, USA
Melody J. Cunningham, MD
Assistant Professor of Pediatrics
Harvard Medical School
Jeannie L. Callum, BA, MD, FRCPC Children’s Hospital Boston
Assistant Professor, Department of Laboratory Boston, Massachusetts, USA
Medicine and Pathobiology
University of Toronto
Director, Blood and Tissue Banks
Sunnybrook and Women’s College Health Richard J. Davey, MD
Sciences Center Director, Transfusion Service
Toronto, Ontario, Canada The Methodist Hospital
Houston, Texas, USA
Sally A. Campbell-Lee, MD
Assistant Professor, Department of Pathology Dana V. Devine, PhD
Associate Medical Director, Division of Transfusion Professor of Pathology and Laboratory Medicine
viii
Medicine Centre for Blood Research
Medical Director, Johns Hopkins Bayview University of British Columbia
Transfusion Medicine Service Executive Director, Research and Development
Baltimore, Maryland, USA Canadian Blood Services
Vancouver, British Columbia, Canada
Jeffrey L. Carson, MD
Richard C. Reynolds Professor of Medicine Roger Y. Dodd, PhD
Chief, Division of General Internal Medicine Vice President, Research and Development
University of Medicine and Dentistry, New Jersey Director, Holland Laboratory
Robert Wood Johnson Medical School American Red Cross, Biomedical Services
New Brunswick, New Jersey, USA Rockville, Maryland, USA
Laurence Corash, MD
Professor, Laboratory Medicine, University of Walter H. Dzik, MD
California, San Francisco Co-Director, Blood Transfusion Service
Attending Physician, Laboratory Medicine and Massachusetts General Hospital
Medicine-Hematology Division Associate Professor of Pathology
The Medical Center at the University of California, Harvard Medical School
San Francisco Boston, Massachusetts, USA
CONTRIBUTORS
James R. Eckman, MD Joanna M. Heal, MRCP, MBBS
Director, Georgia Comprehensive Sickle Cell Center Associate Medical Director, American Red Cross
Grady Health System Blood Services, New York-Penn Region
Professor of Hematology/Oncology and Medicine Associate Clinical Professor of Medicine,
Winship Cancer Institute Hematology-Oncology Unit
Emory University School of Medicine University of Rochester School of Medicine
Atlanta, Georgia, USA Rochester, New York, USA
Krista L. Hillyer, MD
Chief Medical Officer, American Red Cross Blood
Mindy Goldman, MD, FRCP(C) Services, Southern Region
Executive Medical Director Assistant Professor, Department of Pathology
Donor and Transplantation Services and Laboratory Medicine
Canadian Blood Services Emory University School of Medicine
Ottawa, Ontario, Canada Atlanta, Georgia, USA
Cassandra D. Josephson, MD
Assistant Professor, Departments of Pathology Karen Shoos Lipton, JD
and Pediatrics Chief Executive Officer
Emory University School of Medicine AABB
Assistant Director, Blood Banks and Transfusion Bethesda, Maryland, USA
Services
Attending Pediatric Hematologist/Oncologist,
Department of Pediatrics
Children’s Healthcare of Atlanta Lennart E. Lögdberg, MD, PhD
Atlanta, Georgia, USA Associate Professor, Department of Pathology
and Laboratory Medicine
Director, Crawford W. Long Hospital Transfusion
Richard M. Kaufman, MD Services
Medical Director, Adult Transfusion Service, Emory University School of Medicine
Brigham and Women’s Hospital Atlanta, Georgia, USA
Assistant Professor of Pathology, Harvard Medical
School
Boston, Massachusetts, USA
x
Naomi L. C. Luban, MD
Interim Executive Director, Center for Cancer
and Blood Disorders
Thomas S. Kickler, MD Chair, Laboratory Medicine and Pathology
Professor of Medicine and Pathology Director, Transfusion Medicine/The Edward J. Miller
Johns Hopkins University School of Medicine Donor Center
Baltimore, Maryland, USA Vice Chair for Academic Affairs, Department
of Pediatrics
Children’s National Medical Center
Professor, Pediatrics and Pathology
Diane Killion, JD The George Washington University Medical Center
Staff Counsel Washington, D.C., USA
AABB
Bethesda, Maryland, USA
Catherine S. Manno, MD
Karen E. King, MD Professor and Associate Chair for Clinical Affairs
Associate Medical Director Department of Pediatrics
Transfusion Medicine Division Children’s Hospital of Philadelphia
Johns Hopkins University School of Medicine University of Pennsylvania School of Medicine
Baltimore, Maryland, USA Philadelphia, Pennsylvania, USA
CONTRIBUTORS
Francesco M. Marincola, MD
Director, HLA and Immunogenetics Research Professor, Laboratory Medicine
Laboratory University of Connecticut Health Sciences Center
Department of Transfusion Medicine Director, Blood Bank, John Dempsey Hospital
Warren G. Magnuson Clinical Center Farmington, Connecticut, USA
National Institutes of Health
Bethesda, Maryland, USA
Thomas H. Price, MD
Executive Vice-President, Medical Division
Bruce C. McLeod, MD Medical Director, Puget Sound Blood Center
Professor of Medicine and Pathology Professor of Medicine, University of Washington
Director, Blood Center Seattle, Washington, USA
Rush University Medical Center
Chicago, Illinois, USA
Jayashree Ramasethu, MD
Associate Professor of Clinical Pediatrics
Jay E. Menitove, MD Division of Neonatology
Clinical Professor, Internal Medicine Department of Pediatrics
University of Kansas School of Medicine Georgetown University Hospital
Kansas City, Kansas; Executive Director and Washington, D.C., USA
Medical Director, Community Blood Center
of Greater Kansas City
Kansas City, Missouri, USA Sandra M. Ramirez-Arcos, MSc, PhD
Associate Scientist, Canadian Blood Services
Adjunct Professor, University of Ottawa
Peter A. Millward, MD Research and Development, Infectious Diseases
Assistant Professor, Clinical Pathology Ottawa, Ontario, Canada
Milton S. Hershey Medical Center
Pennsylvania State University
Hershey, Pennsylvania, USA
William Reed, MD
Assistant Medical Director, Research
Blood Systems Research Institute
Edward L. Murphy, MD, MPH Clinical Associate Professor
Professor, Laboratory Medicine and Epidemiology/ xi
Department of Laboratory Medicine
Biostatistics
Medical Director, Human Islet and Cellular
University of California, San Francisco
Therapy Laboratory
San Francisco, California, USA
University of California
San Francisco, California, USA
Paul M. Ness, MD
Director, Transfusion Medicine Division
Johns Hopkins Medical Institutions Marion E. Reid, PhD
Professor, Pathology, Medicine, and Oncology Director, Immunohematology
Johns Hopkins University School of Medicine New York Blood Center
Baltimore, Maryland, USA New York, New York, USA
The editors are pleased to introduce the Second Edition text. We have made a concerted effort to ensure that each
of Blood Banking and Transfusion Medicine. Substantial chapter includes the most up-to-date scientific underpin-
modifications and additions have been made to the text, nings of transfusion biology as well as detailed informa-
reflecting advancements in a number of areas, including tion that can be applied to clinical transfusion practice. It is
cellular therapy, component preparation, infectious disease our goal that this textbook remain the definitive source of
testing, and the underlying biology of transfusion therapy. blood banking and transfusion medicine biology, technol-
In addition, we have continued to integrate elements of ogy, and practice for physicians, technologists, nurses, and
Anderson and Ness’s excellent textbook The Scientific Basis administrative personnel, and we sincerely welcome read-
of Transfusion Medicine, which can be noted by the reader ers’ observations, criticisms, and suggestions so that we can
as a number of new chapters entitled “Principles of . . . .” We continue to work to improve this book. Finally, we thank
are grateful for the many suggestions offered by readers of you for your support of this text, the field of transfusion
the First Edition that led to additional improvements in the medicine, and the patients we serve.
C. D. Hillyer
L. E. Silberstein
P. M. Ness
K. C. Anderson
J. D. Roback
xv
Acknowledgments
We, the editors, would like to acknowledge the outstanding tech- Jackson, and James Hillyer; the family and friends of Les
nical and professional support of Sue Rollins and the expertise, Silberstein; Barbara, Jennie, Steven, and Molly Ness; Cynthia,
guidance, and friendship of Dolores Meloni. We would also like Emily, David, and Peter Anderson; and Linda, Evan, and Ethan
to thank our friends and families for their unconditional love Roback. Finally, we would like to acknowledge and thank the
and support, without which this edition could not have come many mentors, physicians, and patients who have served as
to fruition. We thank especially Krista, Whitney, Peter, Margot, inspiration, colleagues, and friends.
xvii
Chapter 1
A Brief History of Blood Transfusion
Kim A. Janatpour ● Paul V. Holland
EARLY HISTORY with a contumacious and violent fever.” The boy had been
treated with multiple bleeds, following which “his wit
Since the beginning of human history, blood has been rec- seemed wholly sunk, his memory perfectly soft, and his
ognized as a vital force, the essence of life. Prehistoric man body so heavy and drowsie that he was not fit for any thing.”
created cave drawings showing individuals bleeding from Denis attributed these symptoms to the bloodletting he had
traumatic wounds. In the Bible, Leviticus states “the life of received. As treatment, Denis exchanged 3 ounces of the
the flesh is in the blood.” The Chinese Huang Di Nei Ching boy’s blood for 9 ounces of lamb’s blood. Denis chose ani-
(770–221 bc) held that blood contained the soul. Blood mal blood because he believed it purer than that of humans
played a central theme in ancient rituals. Egyptians and due to man’s “debauchery and irregularities in eating and
Romans took blood baths for physical and spiritual resto- drinking” and reasoned that if man could use animal milk
ration,1 and Romans even drank the blood of fallen gladia- as nutrient, animal blood would be safe. Following the infu-
tors in the belief that the blood could transmit the gladiator’s sion of lamb’s blood, the patient complained about “a great
vitality. Precolumbian North American Indians bled the body heat along his arm,” but otherwise suffered no ill effects.
“of its greatest power” as self-punishment. In the Middle Denis subsequently performed such transfusions on three
Ages, the drinking of blood was advocated as a tonic for more patients, the last of which resulted in the first mal-
rejuvenation and for the treatment of various diseases.2 Pope practice suit for blood transfusion.4 Antoine Mauroy was
Innocent VIII drank the blood from three young boys in a 34-year-old madman who was brought to Denis after he
1492. Unfortunately, the boys and the Pope died.2 The idea was found wandering the streets of Paris in the winter of
that infusion of blood could be beneficial did not emerge 1667. Mauroy had suffered for years from severe “phren-
until the 17th century. sies,” during which he would beat his wife, strip off his 1
From the time of Hippocrates (c. 450 bc), disease clothes, and run through the streets, setting house fires. At
was believed to be caused by an imbalance of the four this time, blood was believed to affect one’s temperament 3
humours—blood, phlegm, yellow bile, and black bile. Of and character; therefore, it was reasoned that blood transfu-
these, blood was the most important (Galen [130–201 ad] sion could be used to treat mental ailments. Denis’s patron,
really advanced the humoural theory). The most popu- Monsieur de Montmort, proposed transfusing Mauroy to
lar treatment for most ailments, even as late as the 18th allay the “heat of his blood.”5 Denis transfused Mauroy with
century, was blood letting (Fig. 1–1). Without the correct calf ’s blood, hoping that the calf ’s docile nature would be
understanding of blood circulation, intravenous blood imparted to Mauroy. Although the patient complained of
infusion could not even be imagined. This changed in 1628 heat moving up his arm, he tolerated the transfusion well. A
with William Harvey’s description of the circulatory system. few days later, a second, larger transfusion was performed.
Harvey’s identification of separate yet connected arterial This time, however, the patient complained “of great pains
and venous systems in his De Motu Cordis paved the way for in his kidneys, and that he was not well in his stomack,
an entirely new arena of blood investigation.3 that he was ready to [choak] unless they gave him his lib-
In 1656, Christopher Wren used a quill with an attached erty.”6 The transfusion was quickly discontinued, after
bladder to demonstrate that the intravenous injection of which the patient vomited and passed urine “black as soot.”
substances into animals had systemic effects.1,2 In 1666, Miraculously, the patient not only survived this hemolytic
Richard Lower successfully transfused blood from one dog transfusion reaction, but also appeared to be cured, show-
to another, which led Samuel Pepys to speculate on the ing “a surprising calmness, and a great presence of mind …
potential benefits of human transfusion, stating that “bad and a general lassitude in all his limbs.” In fact, upon seeing
blood” might be mended by “borrowing” blood “from a his wife a few days later, Mauroy greeted her tenderly, relat-
better body.”3 ing “with great presence of mind all that had befallen him.”
Denis was astonished—the man who “used to do nothing
but swear and beat his wife” had dramatically, almost magi-
THE FIRST ANIMAL-TO-HUMAN cally, been cured.7
TRANSFUSIONS Also, later in 1667, Richard Lower successfully transfused
a Cambridge University student described as “cracked a little
The first published animal-to-human transfusion was per- in the head” with sheep’s blood.3,4 A bitter debate followed
formed June 15, 1667, by Jean Baptiste Denis, a physician to between Denis and Lower as to who could claim to have dis-
Louis XIV, on a 16-year-old boy who had been “tormented covered blood transfusion.4
HISTORY Figure 1–1 A collection of bloodletting instru-
ments. (From Star D. Blood. An epic history of
medicine and commerce. New York, HarperCollins
Publishers, 2002. With permission.)
Although a select group of scientists was excited about FIRST HUMAN-TO-HUMAN TRANSFUSION
the concept of transfusion, others were adamantly opposed
to the practice. Denis, in particular, suffered harsh criticism After being banned for more than 150 years, the use of blood
from his peers. With this intense debate and criticism as the transfusion was revived during the late 18th century. A foot-
backdrop, Mauroy suffered a relapse; his wife begged Denis note in an American journal indicates that the first human-
to transfuse her husband again. Denis found the patient to-human transfusion had been performed by Philip Syng
to be very ill, so was hesitant to perform the transfusion, Physick, the “Father of American Surgery,” in 1795, although
but reluctantly agreed. Before the transfusion began, how- this has never been confirmed.5,9 In 1816, John Henry
ever, Mauroy died and his widow refused to allow Denis Leacock, a Barbados physician, presented his dissertation “On
to examine the body. The widow had been offered money the Transfusion of Blood in Extreme Cases of Haemorrhage.”
from Denis’s rivals to charge him with murder; she offered Leacock subsequently performed and published a set of ani-
to drop the matter if Denis would agree to support her mal experiments that proved that the donor and recipient
financially. Denis refused, and the case went to court. must be of the same species.10
Denis was exonerated when it was discovered that Mauroy Although Leacock apparently went no further with the
had been poisoned with arsenic by his wife. Nonetheless, experiments, his work inspired James Blundell, an obstetri-
although Denis was acquitted of malpractice, the general cian and physiologist at Guy’s Hospital in London, to carry
opposition to transfusion ultimately led the French and out additional investigations. At the time, obstetricians could
English courts, and much of the rest of Europe, to ban all only stand by and watch helplessly as patients exsanguinated
human transfusions.1,4,5,7,8 postpartum. Blundell was convinced that blood transfusion
A BRIEF HISTORY OF BLOOD TRANSFUSION
could save patients’ lives. His extensive experimentation con- The discovery of blood groups led Ludvig Hektoen of
firmed Leacock’s findings that blood could be used to treat Chicago to advocate selecting donors by blood group and
hemorrhagic shock, but only blood from the same species crossmatching.8 In 1913, Reuben Ottenberg conclusively
could be used. Recognizing the potentially serious risks of demonstrated the importance of compatibility testing in his
transfusion, Blundell began attempting human-to-human report of 128 cases of transfusion.15 However, even as recently
transfusion in cases that were otherwise hopeless. Over a as 1937, some suggested that crossmatching was unnecessary
decade, he performed 10 such transfusions, all without suc- if the selection of donors was restricted to individuals of the
cess. However, in August 1825, Blundell successfully trans- same blood group.5
fused a woman dying from postpartum hemorrhage with The inheritance pattern of blood groups was finally
blood from her husband. Other successes followed, includ- proved by Felix Bernstein in 1924.3 Sadly, differences in race
ing three cases of postpartum hemorrhage, and a young boy distribution of blood groups were manipulated and misused
who was hypovolemic following amputation of his leg.11 in Germany during World War I (WWI) and World War
Subsequently, other reports of transfusion followed from II (WWII), during which time blood group B was deemed
Europe and then the United States, where it was reported a marker for Slavic or Jewish race, and blood group A was
that transfusion was used by the Union Army during the considered associated with intelligence and industry. In the
American Civil War.5,9 1950s in Louisiana, it was a misdemeanor for a physician to
Significant progress in understanding the basis for the give blood from a black donor to a white person without
incompatibility between species was made by Emil Ponfick consent. In the United States, segregation of blood by race
and Leonard Landois in the late 1800s.8 The first revela- existed until the 1960s.3
tion came from Ponfick, who observed red cell lysis in the
blood of a woman who died after receiving a transfusion
of sheep blood. From animal experiments, Ponfick found DISCOVERY OF RH BLOOD GROUPS
that incompatible transfusions were associated with hem-
orrhage and “congestion” of the kidneys, lungs, and liver. Although a major discovery in transfusion medicine, ABO
He also recognized that the red urine that transfused ani- blood group typing was not sufficient to prevent many fatal
mals excreted was caused by hemoglobinuria, not hema- hemolytic transfusion reactions. In 1939 Philip Levine pub-
turia. Landois’s observation that human red cells would lished a case report of post-transfusion hemolysis in a blood
lyse when mixed in vitro with the sera of other animals set group O patient who received blood from her blood group
the stage for the study of the immunologic basis of blood O husband. Levine found that incubation of the patient’s
incompatibility.8 serum with her husband’s red cells resulted in agglutination.
Additionally, the woman’s serum was found to agglutinate 80
of 104 other samples of ABO-compatible blood. The name
DISCOVERY OF ABO BLOOD GROUPS of the offending antibody came from parallel experiments
conducted by Landsteiner and Alex Wiener in which anti- 1
Before 1901, the prevailing belief was that all human bodies produced by immunization of rabbits and guinea pigs
blood was the same. However, this changed in 1901 with with blood from rhesus monkeys caused red cell agglutina- 5
Karl Landsteiner’s landmark discovery of ABO blood tion of 85% of humans tested. Those individuals whose red
groups.12 Landsteiner, an Austrian immunologist, noticed cells were agglutinated by these antibodies were classified as
that human blood mixed in test tubes with other speci- rhesus (Rh) positive.3 Levine was able to show that Rh anti-
mens of human blood sometimes resulted in agglutination. bodies were the main cause of serious hemolytic disease of
By incubating red cells from some individuals with serum the newborn (erythroblastosis fetalis).16 Later, it was appre-
from others, he identified agglutination patterns, leading to ciated that the Rh system is composed of numerous alleles.
the initial identification of three blood groups, A, B, and C The current system of nomenclature—c, C, d, D, e, E—was
(C was later renamed O).3,13 In 1902, Alfred Decastello and proposed in 1944 by Cambridge geneticist Sir Ronald Fisher.
Adriano Sturli, two of Landsteiner’s former students, found Subsequent development of Rh immune globulin (RhIG) for
the fourth blood group, AB.3 Landsteiner also contributed to prevention of hemolytic disease of the newborn was a major
forensic science by developing a method for blood typing of advance. The use of the antiglobulin test, first described by
dried blood specimens.14 Carlo Moreschi in 1908 and rediscovered in 1945 by Robin
Interestingly, the importance of the blood groups was Coombs, Rob Race, and Arthur Mourant, allowed the iden-
not immediately recognized; blood group typing did not tification of many other blood group antigens in the decades
become part of routine practice for several years. Richard that followed.3,17
Weil, a pathologist at the German Hospital in New York,
was the first to perform ABO typing and began compat-
ibility testing in 1907; he was also the first to suggest inheri- BLOOD COAGULATION, PRESERVATION,
tance of ABO types.5 Also in 1907 and 1910, respectively, AND STORAGE
Jan Jansky of Czechoslovakia and Moss of the United
States independently identified four human blood groups.3 Despite some successes by Blundell and contemporaries,
However, the Roman numeral systems that Jansky and transfusions often failed to save lives, and remained a rarity
Moss each used for designating the four blood groups were until the early 20th century. Clotting remained a significant
completely reversed. Tremendous confusion ensued with problem. A variety of devices, involving valves, syringes, and
the three different nomenclatures. Finally, in 1927, the tubing, were invented to facilitate the collection and infu-
American Association of Immunologists adopted a new sion of blood from one individual to another, including two
classification scheme proposed by Landsteiner, the current invented by Blundell—the “Gravitator” and the “Impellor.”
ABO terminology.3 The impellor consisted of a double-walled funnel in which
HISTORY Figure 1–2 A. Blundell’s “impellor.” (Modified from
Jones HW, Mackmull G. The influence of James Blundell
on the development of blood transfusion. Ann Med
Hist 1928;10:242.) B, Sketch of Blundell’s gravitator.
(Modified from Blundell J. Observations on transfusion of
blood. Lancet 1828;2:321.)
the outer compartment was filled with warm water. The CITRATE ANTICOAGULATION
donor blood flowed into the funnel, was sucked into a
syringe, and was forced along tubing into a cannula inserted A chemical approach to anticoagulation was first attempted
into the patient’s vein by means of two oppositely acting by Braxton Hicks, a 19th-century obstetrician, who experi-
spring valves below the funnel8 (Fig. 1–2). Gesellius used an mented with phosphate of soda. Unfortunately, none of the
equally complex device, in which the donor’s back was lanced four patients in whom it was used survived.8 Other sub-
multiple times and capillary blood extracted using suction stances used in anticoagulation attempts included sodium
cups5,8 (Fig. 1–3). James Aveling used a simpler method for bicarbonate, ammonium oxalate, arsphenamine, sodium
direct blood transfusion from a donor using two silver can- iodide, sodium sulfate, and hirudin.8 Surprisingly, these
nulae, inserted into the recipient and donor, and connected initial attempts did not include sodium citrate, which had
by rubber tubing with a compressible bulb in the middle to long been used in laboratories as an anticoagulant.8 The
promote and sustain flow.11 The Aveling device is featured 1% concentration of citrate commonly used in the labora-
in the first known photograph of an actual blood transfu- tory, however, was toxic to humans.3 Nonetheless, in 1914
sion, taken at Bellevue Hospital in New York City in the Albert Hustin reported the first human transfusion using
1870s9 (Fig. 1–4). In 1908, Alexis Carrel, a French researcher citrated blood.8 In 1915, Richard Lewisohn of the Mount
working at the Rockefeller Institute for Medical Research in Sinai Hospital in New York proved that a 0.2% sodium
New York, perfected a surgical technique for the direct anas- citrate solution was effective as an anticoagulant for blood,
tomosis of donor artery to recipient vein.3 Although highly while having no toxicity even when as much as 2500 mL of
effective at providing blood to the patient without clotting, citrated blood were transfused.3 Also in 1915, Richard Weil,
performance of this technique required tremendous skill. an American pathologist, found that citrated blood could
Further, it required donors willing to undergo the painful be refrigerated for several days before use.18 Lewinsohn and
procedure. It was also impossible to accurately estimate the Weil, as well as Rous and Turner, found that addition of dex-
amount of blood passed from donor to recipient; donors trose to citrate would preserve blood for up to 2 weeks1,5
often became hypotensive or recipients developed circulatory (Fig. 1–5). This permitted the first transfusion of stored blood
overload.3 in WWI by an American army physician, Oswald Robertson,
A BRIEF HISTORY OF BLOOD TRANSFUSION
Figure 1–3 Collection and transfusion of capillary blood by the
method of Gesellius. (Modified from Gesellius F: Die Transfusion des
Blutes. Leipzig, E. Hoppe, 1873.)
•
system of his own design, Duran-Iorda collected the blood
into glass bottles containing a citrate and glucose solution.i-"
Blood was then transported to front-line hospitals in vehicles
fitted with refrigerators." At the height of fighting, Duran-
Iorda's blood center in Barcelona was processing up to 75
blood donations per hour? When it became evident that the
Nationalists would win the war, Duran-Iorda left Spain for
England, where he assisted Janet Vaughan in establishing a
blood bank at Hammersmith Hospital in 1938. Because war
with Germany was imminent, the Medical Research Council
supported Vaughan's proposal to establish four blood depots
in London. In 1938, the War Office also created the army
blood supply depot under the control of Lionel Whitby.' The
Army's policy, to supply blood group 0 red cells at the bat-
tlefronts with blood that had been collected centrally rather
than collected at the front from troops, proved to be highly
successful.'
TRANSFUSION IN WWII
•
WWII were evident, but like whole blood, plasma had its
limitations. The protein-rich solution was highly prone to
bacterial contamination. Freeze-dried plasma circumvented BLOOD COMPONENTS
this problem, but was cumbersome and awkward to use on
the battlefield. In 1940, Edwin Cohn, a Harvard chemist, Another advance in transfusion was development of the first
isolated various fractions of plasma. Fraction V was found cellseparator in 1951by Edwin Cohn; the cellseparator allowed
to be composed of albumin, which, in limited clinical stud- blood to be separated into red cells, white cells, platelets, and
ies of volunteers and accident victims, was found to restore plasma. Although in principle it was possible to harvest any
circulatory collapse. Professor 1. S. Ravdin at the University particular component of the blood, in practice, satisfactory
of Pennsylvania established albumin's efficacy following yields of only red cells or plasma could be obtained."
the bombing of Pearl Harbor in 1942, where albumin was Cohn's cell separator paved the way for component ther-
used to treat injuries of 87 patients? Most of these patients apy, but the development of plastic containers made modern
showed some clinical improvement, and only four suffered component therapy possible. Up until the 1950s, blood was
minor reactions.' Based on this success, the u.s. military collected through steel needles and rubber tubing into glass,
added albumin production to efforts to produce plasma on rubber-stoppered bottles, which were reused following wash-
a large scale? Albumin had a number of advantages over ing and sterilization. 1 Pyrogenic reactions and air embolism
plasma. Because of the method of production, cold ethanol were known risks to blood collected using these materials. In
precipitation/fractionation, albumin is free of bacteria. In 1952, Carl Walter, a researcher under Harvey Cushing, and
addition, because albumin is highly concentrated, it could be William Murphy described a system in which the blood was
transported in small vials easily on the battlefield. However, collected into a collapsible bag of polyvinyl resin.l-" Plastic
production of a single unit of albumin required pooling of had the flexibility to permit the removal of plasma following
multiple blood donations, and, like plasma, production of sedimentation or centrifugation, techniques that became the
albumin was difficult. During WWII, both plasma and albu- foundations for component production. 1,37
min were produced in vast amounts. By the end of 1943,
over 2.5 million packages of dried plasma and nearly 125,000
units of albumin had been sent to the u.s. military? FACTOR VIII CONCENTRATES
Cohn's fractionation of plasma yielded numerous benefits
beyond the purification of albumin. Fraction I was found to The ability to separate blood into components resulted in
contain fibrinogen; fractions II and III contained immuno- major advances for the treatment of hemophilia. Initially,
globulins that proved effective in the temporary prevention hemophiliacs were treated with fresh frozen plasma; however,
HISTORY massive volumes were required to replenish the deficient fac- and West Nile virus. Some type of pathogen inactivation
tor VIII in these patients. A more concentrated form of the applied to blood components might obviate the need for yet
required factor VIII was found in the cryoprecipitated por- more testing of known and unknown pathogens that might
tion of plasma by Judith Pool in 1965.38 In 1968 Brinkhous be transfusion-transmissible.
and Shanbrom produced concentrated factor VIII by pooling
hundreds to thousands of units of plasma. These factor con-
centrates could be carried by the patient and administered NONINFECTIOUS COMPLICATIONS
by self-injection at the earliest sign of bleeding.7,39 Factor OF TRANSFUSION
VIII concentrates were a major advance in the treatment of
hemophilia A, but they came with a very high risk of infec- In addition to infectious diseases, other risks of transfusion
tious disease transmission, first with hepatitis and later with became apparent as transfusions increased. Transfused leu-
human immunodeficiency virus (HIV). kocytes were found to have a number of undesirable effects
(e.g., graft-versus-host disease [GVHD] and febrile reac-
tions).8 In 1970, Graw and colleagues demonstrated that
INFECTIOUS DISEASE TRANSMISSION GVHD could be prevented by the irradiation of blood com-
ponents.44 In 1962, the first-generation leukocyte filter was
The first hint that hepatitis could be caused by blood transfu- shown to be effective in the prevention of febrile transfu-
sion came during WWII following administration of yellow sion reactions.45 The additional benefits of leukoreduction,
fever vaccines produced using human serum as a stabilizer.1 including reduction of leukotropic viruses, such as HTLV 1/2
This was followed by a report in 1943 of a series of seven and CMV, and minimization of the risk of human leukocyte
cases of jaundice following transfusion of whole blood or antigen (HLA) sensitization, have resulted in the widespread
plasma.40 At the time, serum hepatitis (or hepatitis B) was adoption of leukoreduction of cellular blood components.
assumed to be the major cause of transfusion-associated
hepatitis. This began the awareness that blood transfusion
could transmit potentially deadly viral diseases. In 1962, the THE MODERN ERA
connection between paying for units and an increased risk
of post-transfusion hepatitis was made by J. Garrett Allen, a Transfusion medicine continues to evolve in the modern era.
Stanford surgeon41; however, it wasn’t until a decade later that New, problematic infectious agents emerge, such as West
the National Blood Policy mandated a voluntary (unpaid) Nile virus, which is usually transmitted to humans by mos-
donation system in the United States.1 Subsequently, in 1975 quitoes. West Nile virus was first reported to be transmitted
transfusion-associated hepatitis was shown to be primarily by transfusion and organ transplantation in 2002.
due to non-A, non-B hepatitis, or what became identified As detection and prevention of transfusion-transmitted
later as hepatitis C.42 Despite the risk of hepatitis transmis- viral infections has improved, bacterial contamination has
I sion, blood utilization continued to increase. In the United evolved into one of the most significant causes of transfu-
States, for example, blood utilization doubled between 1971 sion-transmitted infectious disease. Bacterial detection systems
10 and 1980.1 However, the emergence of acquired immunode- are being used on platelet components with some success.
ficiency syndrome (AIDS) changed this trend. Fortunately, new methodologies designed to inactivate
The first case of AIDS was reported in 1981. This myste- bacteria have been developed and are undergoing evalua-
rious disease initially occurred only in gay men. However, tion.46
within a few years, it became clear that AIDS was caused by Alternatives to human blood for transfusion, so called
a blood-borne virus. The first reported case of transfusion- “blood substitutes,” continue to be actively investigated but
associated AIDS occurred in a 20-month-old infant who had have thus far had limited success. Current “blood substi-
received multiple transfusions for hemolytic disease of the tutes” are limited to substances designed to carry oxygen.
newborn.43 Hemophiliacs, dependent on lifelong infusions Unfortunately, these substances have limited application
of factor VIII concentrates, were particularly devastated by and are associated with a unique set of risks and potential
the disease. In 1985, the first serologic test to detect HIV was complications.47
implemented by blood banks to protect the blood supply. The history of transfusion medicine parallels mankind’s
However, the possibility that other, as yet unknown, patho- understanding of physiology, immunology, chemistry, infec-
gens could also be transmitted by blood transfusion resulted tious diseases, and advances in technology. What began
in a new awareness that blood should be used judiciously. as a belief that blood carries important healing properties
Multiple serologic tests are now routinely employed on every has been validated by science. However, despite numerous
blood component for a variety of infectious agents, including advances over the centuries, blood remains an indispensable,
hepatitis B, hepatitis C, HIV, cytomegalovirus (CMV), syphi- life-giving force. Interestingly, some of the more impor-
lis, and human T-lymphotropic virus (HTLV). Despite the tant advances in blood banking/transfusion medicine have
improvement in blood safety that serologic testing provides, occurred as the result of wars.
the risk of infectious disease transmission is not completely
eradicated from the blood supply due to the window period
in which donors are infectious, but have not yet developed SUMMARY REMARKS
detectable antibody. The development of nucleic acid ampli-
fication technology (NAT) has limited the risk of infectious Blood transfusions today are an indispensable part of many
disease transmission even more by allowing the direct detec- medical and surgical therapies. The use of blood and its
tion of even small quantities of pathogen. Currently, NAT components temporarily replaces what may be lost or not
testing is widely employed for HIV and hepatitis C detection, produced before, during, or after a disease process and/or
and is used, in certain areas, for detection of hepatitis B virus its treatment. The benefits of transfusion today far outweigh
A BRIEF HISTORY OF BLOOD TRANSFUSION
their minute (yet real) risks with all the current safeguards to 23. Moore SB. A brief history of the early years of blood transfusion at the
select donors, test blood, and ensure that compatible blood Mayo Clinic: the first blood bank in the United States (1935). Transfus
Med Rev 2005;19:241–245.
is transfused to the correct patient. Transfusion medicine has 24. Telischi M. Evolution of Cook County Hospital blood bank. Transfu-
come a long way due to multiple pathfinders, adventurous sion 1974;14:623–628.
physicians, and courageous donors and patients, especially 25. Jorda FD. The Barcelona blood transfusion service. Lancet 1939;1:773–
in the last half century. We owe much to these pioneers. 776.
26. www.pbs.org/wnet/redgold
27. www.cdrewu.edu
Acknowledgments 28. Schmidt PJ. Charles Drew, a legend in our time. Transfusion
1997;37:234–236.
The authors wish to acknowledge Drs. Leo McCarthy and 29. Stern K, Goodman HS, Berger M. Experimental isoimmunization to
hemoantigens in man. J Immunol 1961;87:189–198.
Paul Schmidt for their generous contributions of photos, 30. Clarke CA, Donohoe WT, McConnell RB, et al. Further experimental
information, and review of this chapter. studies on the prevention of Rh haemolytic disease. BMJ 1963;5336:979–
984.
31. Clarke CA, McConnell RB. Prevention of Rh-hemolytic disease. Spring-
REFERENCES field, Ill., Charles C Thomas, 1972.
32. Freda VJ, Gorman JG, Pollack W. Successful prevention of experimen-
1. Rossi E, Simon T, Moss G (eds). Principles of Transfusion Medicine. tal Rh sensitization in man with an anti-Rh gamma2globulin antibody
Baltimore, Williams & Wilkins, 1991. preparation: A preliminary report. Transfusion 1964;77:26–32.
2. Zmijewski CM. Immunohematology, 3rd ed. New York, Appleton-Cen- 33. Freda VJ, Gorman JG, Pollack W. Rh factor; prevention of immuniza-
tury-Crofts, 1978. tion and clinical trial on mothers. Science 1966;151:828–830.
3. Giangrande PL. The history of blood transfusion. Br J Haematol 34. [No authors listed]. Prevention of Rh-haemolytic disease: results of the
2000;110:758–767. clinical trial. A combined study from centres in England and Baltimore.
4. Myhre BA. The first recorded blood transfusions: 1656 to 1668. Transfu- BMJ 1966;2:907–914.
sion 1990;30:358–362. 35. Pollack W, Gorman JG, Freda VJ, et al: Results of clinical trials of Rho-
5. Petz L, Swisher S, Kleinman S, Spence R (eds). Clinical Practice of GAM in women. Transfusion 1968;8:151–153.
Transfusion Medicine, 3rd ed. New York, Churchill Livingstone, 1996. 36. Walter CW, Murphy WP. A closed gravity technique for the preserva-
6. Denis J. A letter concerning a new way of curing sundry diseases by tion of whole blood in ACD solution utilizing plastic equipment. Surg
transfusion of blood. Philos Trans R Soc Lond [Biol] 1667;2:489. Gynecol Obstet 1952;95:113–119.
7. Star D. Blood. An epic history of medicine and commerce. New York, 37. Sack T, Gibson JG, Buckley ES. The preservation of whole ACD blood
HarperCollins Publishers, 2002. collected stored and transfused in plastic equipment. Surg Gynecol
8. Greenwalt TJ. A short history of transfusion medicine. Transfusion Obstet 1952;95:113–119.
1997;37:550–563. 38. Pool JG, Shannon AE. Production of high-potency concentrates
9. Schmidt PJ. The first photograph of blood transfusion. Transfusion of antihemophilic globulin in a closed-bag system. N Engl J Med
2001;41:968–969. 1965;273:1443–1447.
10. Schmidt PJ, Leacock AG. Forgotten transfusion history: John Leacock of 39. Brinkhous KM, Shanbrom E, Roberts HR, et al. A new high-potency
Barbados. BMJ 2002;325:1485–1487. glycine-precipitated antihemophilic factor (AHF) concentrate. Treat-
11. Baskett TF. James Blundell: the first transfusion of human blood. Resus- ment of classical hemophilia and hemophilia with inhibitors. JAMA
citation 2002;52:229–233. 1968;205:613–617.
12. Landsteiner K. Uber Agglunationserscheinungen normalen menschli- 40. Beeson PB. Jaundice occurring one to four months after transfusion of
1
chen Blutes. Wein Klin Wschr 1901;14:1132–1134. blood or plasma. JAMA 1943;121:1332–1334.
13. Watkins WM. The ABO blood group system: historical background. 41. Allen JG, Sayman WA. Serum hepatitis from transfusion of blood. 11
Transfus Med 2001;11:243–265. JAMA 1962;180:1079–1085.
14. Levine P. A review of Landsteiner’s contributions to human blood 42. Feinstone SM, Kapikian AZ, Purcell RH, et al. Transfusion-asso-
groups. Transfusion 1961;1:45–52. ciated hepatitis not due to viral hepatitis type A or B. N Engl J Med
15. Ottenberg R, Kaliski DJ. Accidents in transfusion: their prevention by 1975;292:767–770.
preliminary blood examination: based on an experience of 128 transfu- 43. Ammann AJ, Cowan MJ, Wara DW, et al. Acquired immunodeficiency
sions. JAMA 1913;61:2138–2140. in an infant: possible transmission by means of blood products. Lancet
16. Mollison PL. Blood Transfusion in Clinical Medicine, 7th ed. Oxford, 1983;1:956–958.
Blackwell Scientific Publications, 1983. 44. Graw RG Jr, Buckner CD, Whang-Peng J, et al. Complication of
17. Moreschi C. Neue Tatsachen Uber die Blutkorperchen-agglutination. bone-marrow transplantation. Graft-versus-host disease resulting
Zentralbl Bakteriol Parasitenkd Infektkr 1908;1 Originale 46:49–51. from chronic-myelogenous-leukaemia leucocyte transfusions. Lancet
18. Weil R. Sodium citrate in the transfusion of blood. JAMA 1915;64: 1970;2:338–341.
425–426. 45. Greenwalt TJ, Gajewski M, McKenna JL. A new method for preparing
19. Hess JR, Schmidt PJ. The first blood banker: Oswald Hope Robertson. buffy coat-poor blood. Transfusion 1962;2:221–229.
Transfusion 2000;40:110–113. 46. Lin L, Hanson CV, Alter HJ, et al. Inactivation of viruses in platelet con-
20. Loutit JF, Mollison PL. Advantages of a disodium-citrate-glucose mix- centrates by photochemical treatment with amotosalen and long-wave-
ture as a blood preservative. BMJ 1943;2:744–745. length ultraviolet light. Transfusion 2005;45:580–590.
21. Gibson JG, Gregory CB, Button LN. Citrate-phosphate-dextrose solu- 47. Klein HG. Blood substitutes: how close to a solution? Dev Biol (Basel)
tion for preservation of human blood: a further report. Transfusion 2005;120:45–52.
1961;1:280–287.
22. Smith AU. Prevention of haemolysis during freezing and thawing of red
blood cells. Lancet 1950;2:910–911.
A. Immunohematology
i. Basic Principles
Chapter 2
Principles of the Immune System
Central to Transfusion Medicine
Terrence L. Geiger
II
Table 2–1 Recognition of Pathogen-Specific Motifs by Toll-like Receptors
16
Receptor Recognition Motif Pathogen
Cytokine Function
Figure 2–1 Antigen receptor gene rearrangement. The germline genetic sequence consists of multiple variable (V), diversity (D), and junctional
(J) segments. Rearrangement of a D and J segment with excision of intervening DNA sequence is followed by that of the DJ sequence to a V
region. Completion of rearrangement juxtaposes sequences necessary for RNA transcription. The RNA transcript is spliced to create an mRNA
that is translated into an antigen receptor chain. The diagram is illustrative and does not indicate the full complexity of the T-cell receptor or B-cell
receptor loci.
\CJ
Z Thus antigen receptors are most variable at critical antigen- interaction between rare antigen-specific T cells and B cells
;;;;; binding sites. Further, antigen receptors are heterodimers and APCs bearing their cognate antigen (Fig. 2-2).
Z
«co composed of two polypeptides, a V-D-J-C chain, and a V-D-J The cell carrying antigen to lymphoid tissue may trans-
o chain. Each of these receptor components uses an indepen- fer that antigen to professional APCs within that tissue."
o
o---' dent set of V, D, J, and C fragments. The chains recombine Alternatively the courier cell, often a resident tissue den-
co independently but recognize antigen together. Thus billions dritic cell, may serve as the APC within the lymphoid tis-
of distinct antigen receptors may form from a relatively small sue, directly presenting antigen to lymphocytes. Conversion
number of gene fragments. of such courier cells to effective professional APCs requires
Antigen receptors are expressed on only two classesof cells, their maturation under the influence of cytokines or Toll-
B lymphocytes and T lymphocytes. On B cells they form the like receptor signals that are present in the context of inflam-
B-cell receptor (BCR), which is a cell surface form of immu- mation and innate immune activation. These stimuli induce
noglobulin. On T cells they form the T-cell receptor (TCR). the formation of signaling molecules on the surface of the
Each T cell or B cell expresses only a single (or in some rare APC that can bind and activate lymphocyte receptors. They
casestwo) immune receptors on their cellsurface." Therefore, also induce the secretion of cytokines and other BRMs by
each T cell or B cell is essentially a clone recognizing a distinct the APCs.29,3o Such supplemental, or costimulatory, sig-
piece of an antigen, called an epitope. Because of the diver- nals can both modify and augment the signal received by a
sity of immune receptors, the frequency of T cells or B cells lymphocyte through its antigen receptor. In the presence of
able to respond to any single antigenic protein is extremely costimulatory signaling, a lymphocyte may become more
low, typically ranging from 1110,000 to 11100,000 cells." fully activated, permitting its expansion and differentiation
Receptor engagement activates a lymphocyte and may lead into an effector cell. In contrast, the absence of costimulatory
to the selection and rapid outgrowth of rare antigen-specific signals indicates to a lymphocyte that antigen is not being
clones. Cell activation by antigen may also induce differentia- presented in the context of inflammation or host danger.
tion into effector lymphocyte forms designed to participate in Signals through antigen receptor in this circumstance lead
distinct types of immune responses, such as allergic responses to incomplete stimulation that may promote the develop-
or delayed type hypersensitivity responses. ment of immune tolerance, sometimes through the death
of the responding antigen-specific lymphocyte.v" As self-
derived antigens will tend not to be presented in the context
The Trail of Antigen
of inflammation, this form of tolerance is one route through
Foreign antigens can enter the body through different routes. which the immune system rids itself of self-reactive lympho-
In the case of infection, entry is generally through the skin or cytes that may cause autoimmunity.
mucosal tissues. In the case of transfusion, direct intravenous
inoculation occurs. In both situations, antigens make their
Antigen Presentation to T Cells
way to lymphoid tissue. Antigens in the skin are picked up
by specialized resident skin dendritic cells called Langerhans The TCR is a complex that includes six different polypeptides,
•
cells and are then carried to draining lymph nodes." Antigens most commonly ex, ~,y, 8, E, and ~ (Fig. 2_3).33,34 The ex and
in the blood are picked up by cells in the spleen. Antigens in ~ chains, which include recombined VJC and VDJC regions,
the gut are taken up by cells in the gut-associated lymphoid respectively, are directly involved in antigen recognition. The
tissue." They are then presented to T lymphocytes and B remaining polypeptides do not recognize antigen but trans-
lymphocytes by specialized antigen-presenting cells (APCs), mit signals into T cells after antigen engagement occurs. The
also called professional APes, in the lymphoid tissues. The y, 8, and E signaling chains are collectively referred to as the
lymphoid tissues are specifically designed to facilitate the CD3 complex.
A B
Figure 2-2 Schematic depiction of the structure of a lymph node (A) and an area of the splenic white pulp (B). Lymphocytes enter through the
high endothelial venule (HEV) or central arteriole and then migrate to T- or B-rich areas where they may interact with antigen-presenting cells. T
cells and B cells may also engage each other, often at the interface between the B- and T-cell zones of the spleen or lymph node. (From Mondino
A, Khoruts A, Jenkins MK. The anatomy of T cell activation. Proc Natl Acad Sci USA 1996;93:2246.)
PRINCIPLES OF THE IMMUNE SYSTEM CENTRAL TO TRANSFUSION MEDICINE
α1 α2
N
N
C C
β2m
α3
A B
Figure 2–3 The TCR–CD3 complex consists of eight polypeptide
Figure 2–4 Structure of a class I MHC molecule. A, Three-dimen-
chains. Specificity is determined by the αβ chains or, in some T cells,
sional ribbon model of the extracellular portion of the molecule. The
by analogous γδ TCR chains. Expression of the CD3 γ, δ, and ε chains,
peptide-binding groove lies between the helices at the top of the mol-
as well as ζ or its splice variants, are necessary for surface expression
ecule. B, Top view. (From Bjorkman PJ et al Structure of the human class
and signaling.
I histocompatability antigen, HLA-A2. Nature 1987;329:506.)
T cells do not see antigen alone through their TCR. They It could be imagined that a pathogen may readily evade the
see it only in the context of major histocompatibility complex, immune system by simply designing proteins that do not con-
or MHC, proteins. The significance of the MHC, also called tain MHC-binding motifs. However, this is not possible for two
the human leukocyte antigen (HLA) complex in humans, was reasons. First, we express several different class I and class II
first recognized through the work of Peter Gorer and George MHC molecules, three of each for the major, or classical, MHC
Snell in the 1930s.2 Gorer and Snell were studying tumor graft forms on each chromosome. These are called HLA-A, -B, and
rejection and determined that this genetic locus was absolutely -C for the human class I locus, and HLA-DR, -DQ, and -DP
critical in determining whether grafted cells were rejected. We for the class II locus. Therefore, an individual heterozygous
now know that the MHC serves as a window through which for each locus may have a total of six different class I mole-
T cells see the world around them. T lymphocytes recognize cules, and, because the α and β chains of the class II molecule
antigen only in the context of MHC molecules. Recognition encoded on each chromosome may heterodimerize either in
is remarkably specific. Even subtle allelic variations in a single cis or in trans, an even larger number of distinct class II MHC 2
MHC molecule are sufficient to abrogate T-cell recognition.35 molecules. If a pathogen’s proteins failed to bind one HLA
MHC proteins fall into two categories, class I and class II protein, it would likely bind one of the others. More impor- 19
molecules.36 Class I MHC molecules are heterodimers con- tantly, the MHC genes are remarkable among all genes for their
sisting of a larger α or “heavy” chain, which is membrane polymorphism, or allelic variability. Because of this variability,
bound, and a smaller soluble protein called β2-microglobu- most of us have different sets of MHC genes. Indeed, no other
lin. Class II MHC molecules are heterodimers of two mem- human genetic locus is as polymorphic as the MHC.42 Sequence
brane-bound polypeptides, an α chain and a β chain of variation of the allelic variants is particularly prominent at the
roughly equal size. How the MHC presents antigen to T cells peptide-binding groove, and different alleles of even a single
was not clear until the late 1980s when x-ray crystallographic MHC gene therefore often have distinct peptide-binding and
structures of MHC molecules were solved (Fig. 2–4).37–39 antigen-presentation properties. MHC polymorphism ensures
Although the amino acid sequences of class I and class II that a pathogen that attempted to evade the immune system of
molecules are distinct, they form similar three-dimensional one individual would not be able to do so for others, thereby
structures. The upper surface of the MHC molecule forms protecting the population as a whole.
a platter, called a β sheet, on top of which lie two parallel The polymorphism and polygenicity of the MHC also
tubular coils of peptide, called α helices. Between the α heli- presents a problem in transfusion and transplantation.
ces lies a groove into which various protein fragments, or T lymphocytes respond particularly vigorously against allo-
peptides, may bind. These peptides, presented together with geneic (allelically distinct) MHC molecules. Each of these
their restricting MHC molecules, are the antigenic structures allogeneic MHC molecules can bind thousands of differ-
recognized by T lymphocytes. ent peptides. Peptides derived from self-antigens, to which
Class I MHC molecules bind peptides approximately 8 to a person’s immune system is normally tolerant, when com-
10 amino acids in length, whereas class II MHC molecules bined with allogeneic MHC molecules, form complexes
bind longer peptides, approximately 13 to 24 amino acids. that are foreign to the host immune system.43 It is estimated
Only a few interactions between peptide amino acids and that roughly 1% to 10% of a person’s T cells will respond to
the MHC are needed to stabilize binding.40,41 This allows for another person’s allogeneic MHC combined with the array
the formation of large numbers of different peptide–MHC of self-peptides that may bind in its groove. For stem cell
complexes. A typical antigenic protein will have one or more transplant recipients, this recognition may be manifested
peptides able to bind any specific MHC molecule. The MHC by the development of graft-versus-host disease (GVHD) or
therefore serves as a platform that displays the antigenic uni- graft rejection. In the case of transfusion, it may be made
verse to T cells. The T cells view these peptides through their apparent by the development of immune responses resulting
TCR, which recognize and bind the peptide–MHC complex. in transfusion-associated GVHD or platelet refractoriness.
BLOOD BANKING
response, most notably interferon-γ.45 As class I MHC mol-
Specialized Functions of Class I and Class II
ecules are synthesized, they are extruded into a membranous
MHC Molecules
compartment within the cell called the endoplasmic reticu-
Despite their similar structures, class I and class II MHC lum. Appropriately sized, proteasomally derived peptide
molecules have distinct roles. Key to this distinction is the fragments are continuously transported into the endoplas-
repertoire of peptides that each MHC molecule binds. Class mic reticulum by a special peptide transporter, called trans-
I MHC binds peptides derived from within a cell and thus porter of antigenic peptides, or TAP1/2.46,47 The transported
displays a cell’s intracellular milieu. Class II MHC displays peptides can then bind to the nascent class I MHC molecules
antigens acquired from the extracellular space. Thus, each (Fig. 2–5).48 MHC molecules bound to peptide are then
set of MHC proteins is telling T cells a different tale. In the transported to the cell surface.
event of a viral infection, presentation of viral antigens on Class II MHC molecules take a different route to the cell
a cell’s class I MHC would indicate to a T cell that the cell surface. They are also extruded into the endoplasmic reticu-
is infected. Presentation on class II MHC molecules would lum. However, unlike class I MHC molecules, they bind to
instead indicate that virus is present in the environment, but a protein called the invariant chain as they are synthesized
not necessarily within that cell. (Fig. 2–6). A peptide segment of the invariant chain folds
The distinct presentation capabilities of class I and class into the peptide-binding groove of the class II molecule,
II MHC results from the different mechanisms through blocking any other peptides from binding. Rather than going
which each protein acquires its peptides. Within the cyto- directly to the cell surface like class I MHC molecules, the
plasm of the cell proteins are continuously being degraded class II molecules are first diverted to specialized vesicles, or
by a large proteolytic complex called the proteasome.44 The membrane-contained compartments, within the cell where
peptide fragments formed by the proteasome are perfectly they meet up with extracellularly derived antigens.49–51
sized for binding class I MHC molecules. The production Cells may take up external antigens through a variety of
of antigenic peptides may be enhanced and the sites of pro- processes, including phagocytosis, or wholesale consumption
teolysis modified by cytokines produced during an immune of cells or particles; pinocytosis, or internalization of small
␣ ER membrane
II
 ATP
20
 PA28
␣
Peptide
Peptide
A B
2m
TAP1 TAP2
C
Figure 2–5 A, Schematic drawing of a cut-open 20S proteasome generating peptide fragments from an unfolded polypeptide chain threaded into
the narrow entry formed by an outer ring of β-type subunits. Cleavage occurs in the central cavity formed by two seven-membered rings of β-type
subunits. A PA28 regulator complex, inducible by IFN-γ is bound at the bottom of the cylinder. B, Hypothetical cross-section of TAP1/TAP2 heterodi-
mer embedded into the endoplasmic reticulum membrane. The adenosine triphosphate (ATP)-binding domain, which extends into the cytoplasm,
is shown. ATP cleavage provides energy for peptide translocation. C, Maturation of class I molecules in the endoplasmic reticulum. Nascent class I
heavy chains initially associate with calnexin. On binding of β2-microglobulin, the heterodimer dissociates from calnexin and forms a complex with
calreticulin, tapasin, and TAP1/2. After peptide binding, the heavy-chain β2-microglobulin complexes are released and exit to the cell surface. (From
Koopman JO, Hammerling GJ, Momburg F. Generation, intracellular transport, and loading of peptides associated with MHC class I molecules. Curr
Opin Immunol 1997;9:81.)
PRINCIPLES OF THE IMMUNE SYSTEM CENTRAL TO TRANSFUSION MEDICINE
Figure 2–6 Intracellular pathways of class I and class II MHC molecules. Class I MHC molecules acquire antigenic peptides, generated by the
proteasome in the cytosol, that are translocated into the endoplasmic reticulum by the TAP molecules (bottom of figure). The class I MHC–peptide
complex is transported through the Golgi complex directly to the cell surface for presentation to CD8 T cells (A). In contrast, class II MHC molecules
acquire antigenic peptides derived from antigens that are internalized in the endocytic pathways (B). Class II MHC heterodimers associate in the
endoplasmic reticulum (bottom of figure) with invariant chains to form nonameric αβ–Ii complexes. At the trans-Golgi network (TGN), these com- 2
plexes are targeted to class II MHC compartments (MIIC) in the endocytic pathway as a result of targeting signals within the Ii cytoplasmic tail (not
shown). There the class II MHC-associated Ii is degraded in distinct steps, at least partially by cathepsin proteases (C), leaving a peptide from Ii in the
class II MHC–peptide-binding groove. This peptide is exchanged for antigenic peptides in a process catalyzed by HLA-DM molecules. Peptide-loaded
21
class II MHC complexes are then transported to the plasma membrane for presentation to CD4 T cells (D). (From Pieters J. MHC class II restricted
antigen presentation. Curr Opin Immunol 1997;9:91.)
amounts of the extracellular fluid; and receptor-mediated the binding affinity, or interaction strength, of antibodies for
internalization, which may occur through bound immu- their target antigens.55,56 For most TCRs, this interaction is
noglobulin or sugar-binding receptors such as the man- too weak and too transient to generate a signal capable of
nose receptor.52–54 This material is broken down by enzymes fully activating a T cell. The TCR is able to signal because
within the endocytic, or internalization, pathway. Vesicles it also receives assistance from co-receptors called CD4 and
containing endocytosed and degraded material merge with CD8. CD4 binds to class II MHC molecules and CD8 binds
the class II MHC containing vesicles. The invariant chain is to class I MHC molecules. When the TCR engages cognate
degraded by proteases at this point, and vesicular peptides peptide–MHC, CD4 or CD8 co-associates.
can then bind within the class II MHC-binding groove. The CD4 and CD8 are transmembrane proteins that per-
class II MHC, with externally derived bound peptide, then form two tasks.57 First, because they both associate with
makes its way to the cell surface. the TCR and bind MHC at a site different from where the
TCR binds peptide–MHC, they increase the binding affinity
T-Cell Co-receptors and the Recognition of of the TCR complex for peptide–MHC. Second, and more
Class I and Class II MHC Molecules importantly, they carry bound to their intracellular domain
MHC molecules present the antigenic universe to T lympho- an enzyme of the src family, called lck (Fig. 2–7).58 Lck is a
cytes. Considering the different types of antigens presented type of enzyme called a tyrosine kinase. It adds phosphates
by class I and class II MHC, intracellular versus extracellular, to certain tyrosine residues present in proteins. Tyrosine
it is not surprising that these molecules will engage differ- phosphorylation of the γ, δ, ε, and ζ chains of the TCR is
ent types of T cells and that this engagement will result in the first step in generating a signal once a TCR recognizes
different types of responses. To understand how this occurs antigen. The complex of CD4 or CD8 co-receptor with the
requires some background in how T cells recognize antigen. TCR and peptide–MHC aligns the lck kinase with specific
The variable α and β chains of the TCR recognize pep- conserved tyrosines present within the signaling chains of
tide–MHC complexes. However, their affinity for peptide– the TCR called immunoreceptor tyrosine-based activation
MHC is poor, approximately 100- to 1000-fold lower than motifs, or ITAMs.59 Once phosphorylated, these ITAMs
BLOOD BANKING
Ag/MHC
Ag/MHC Ag/MHC
TCR TCR
TCR CD4
CD4
CD4
PO4 PO4
PO4 SH2
Lck SH2
PO4 Lck
Lck
PO4
SH2
SH2
SH2
SH2
ZAP–70
ZAP–70
Cellular substrates
Figure 2–7 Model of TCR, Lck and ζ-associated protein (ZAP)-70 interactions during antigen recognition by a CD4 T cell. The phosphorylation
of tyrosine residues of the immunoreceptor tyrosine-based activation motifs within the ζ and CD3 chains has been simplified. (From Weiss A. T-cell
antigen receptor signal transduction: A tale of tails and cytoplasmic protein kinases. Cell 1993;73:211.)
serve as docking sites for additional proteins, particularly majority of nucleated cells indeed express class I MHC con-
ZAP-70, another tyrosine kinase that is in turn phosphory- stitutively, though class I MHC expression can be enhanced
lated and activated by lck. This triggers a cascade of signal- by cytokines, such as in the setting of inflammation.62
ing inside the cell as additional proteins are recruited to these Not all cells need to express class II MHC, which presents
phosphorylated molecules. In a manner analogous to how external antigens. Class II MHC expression is limited to spe-
the coagulation cascade may be triggered by single event, cialized cells that provide T cells with environmental infor-
such as exposure to tissue factor, a low amplitude signal that mation. These include professional APCs, dendritic cells,
initially results from TCR–MHC engagement is enzymati- macrophages, B cells, and endothelial cells, though other cell
cally amplified within a T cell. The impact of CD4 and CD8 types may be induced to express class II MHC in the context
is dramatic, and it has been estimated that these molecules of inflammation.
enhance signaling through the TCR by a factor of 100.57 Of the professional APCs that present antigen to T cells
Through its sophisticated signaling apparatus, T cells may on class II MHC, dendritic cells are particularly efficient.
II be discriminately stimulated by 100 or fewer antigen–MHC They reside in tissues in a quiescent state. When activated
complexes on the surface of an APC.60 by any of a large variety of inflammatory stimuli, such as
22 Mature T cells express either CD4 or CD8 and do not cytokines or Toll-like receptor signals, they sample antigens
co-express both of these. This contrasts with their progeni- from their environment, which they present on cell surface
tors that develop in the thymus, which co-express CD4 and class II MHC molecules.30 After accumulating large quanti-
CD8. During development, these CD4+D8+ thymocytes are ties of antigen, the dendritic cells cease internalizing antigen
tested.61 Cells that interact strongly with self-antigen–MHC and migrate to lymphoid tissues, particularly lymph nodes,
complexes in the thymus are deleted in a process termed where they present antigen to lymphocytes.
negative selection. Negative selection is a major mechanism There are several types of dendritic cells.63 Some, the fol-
through which the adaptive immune repertoire is depleted licular dendritic cells, are particularly adept at presenting
of self-antigen–specific T cells. Cells that do not interact with whole antigen to B lymphocytes.64 This is important because
MHC present in the thymus do not receive necessary sur- B cells recognize whole antigens or large pieces of antigens
vival signals and undergo death through neglect. In contrast, through their immunoglobulin receptors. In contrast, most
in a process termed positive selection, cells with receptors that dendritic cells, including myeloid, lymphoid, and plasma-
interact loosely with class I MHC molecules receive survival cytoid dendritic cells, digest antigens into small fragments
signals, downmodulate CD4, and are left expressing only and present these to T lymphocytes. Because of the scarcity
CD8. Thymocytes that interact with class II MHC down- of lymphocytes specific for any single antigen, presentation
modulate CD8 and become CD4-expressing T cells. Thus, in specialized lymphoid tissue is essential. The professional
T cells can be divided into two major groups based on their APCs structure themselves into a reticular network through
interaction with either class I or class II MHC molecules, and which lymphocytes continuously and rapidly percolate.65
these groups can be identified by the alternative expression After a dendritic cell is stimulated through its Toll-like recep-
of CD4 or CD8. Further, CD4 and CD8 define classes of T tor or other receptors, it may secrete chemokines that will
cells with distinct functional potentials. attract T lymphocytes to it and costimulatory molecules that
can activate them. The lymphocytes scan these dendritic
Cells That Present Antigen cells with their antigen receptor for the presence of cognate
Because class I MHC presents antigens derived from intra- antigen. With recognition through their antigen and costim-
cellular pathogens, like viruses, or that result from intrin- ulatory receptors, the lymphocytes pause, alter their metabo-
sic abnormalities, such as malignant transformation, it is lism, and modulate gene expression. They may then expand
important that all cells capable of self-propagation express and differentiate into effector forms.
class I MHC. CD8+ T cells recognizing class I MHC-derived Other professional APCs are macrophages and B lympho-
antigens may eliminate these aberrant or infected cells. The cytes. Macrophages are potent phagocytic cells when activated
sitivity (DTH) reactions.72,73 Th2 cells, which secrete IL-4, IL-
Increased avidity is particularly important for IgM. IgM cell types and continuous intercellular dialogue. This dialogue
is formed early in the course of an immune response, and occurs even when we fail to clinically observe an immuno-
because it will not have undergone somatic hypermutation, is logic consequence of transfusion. It even occurs within stored
typically of low affinity. This is compensated for by its high blood components, which respond to their storage conditions
valency and thereby enhanced avidity. by producing cytokines and other BRMs that are then infused
As described, naive B cells secrete IgM when stimu- into patients.136,137 The complexity of the immune response
lated to produce antibody. A minimal requirement for B leads to unpredictability in the effects of a transfusion. We still
cells to switch isotypes is the engagement of cell surface cannot predict who will or will not have an adverse outcome
CD40 by CD40L on T cells. The isotype the B cell converts to a specific transfusion event. Nevertheless, a continuous
to depends on the inflammatory environment at the time conscientiousness toward the nature of the immune response,
of isotype switching. Cytokines are particularly important. how the immune response may perturb the intended out-
IFN-γ will promote formation of subclasses of IgG that effi- comes of our transfusions, and how transfusion may perturb
ciently bind complement. IL-4 will promote IgE formation. the immune system itself, can only help inform the clinical
TGF-β promotes the formation of IgA.133–135 These differ- decisions we make.
ent antibody isotypes and even subclasses within isotypes
have distinct functional properties, including complement
fixation, Fc-receptor binding, and transport across epithe- REFERENCES
lial barriers or the placenta. Some of these properties are
listed in Table 2–3. 1. Silverstein AM. Autoimmunity versus horror autotoxicus: the struggle 2
for recognition. Nat Immunol 2001;2:279–281.
All antibodies are not equivalent, and the properties an 2. Silverstein AM. The History of immunology. In Paul WE (ed). Funda-
antibody possesses will define its functional capabilities. mental Immunology. New York, Raven Press, 1989, pp 21–38. 27
These properties include the antibody isotype and valency, 3. Owen R. Karl Landsteiner and the first human marker locus. Genetics
affinity for antigen, temperature dependence in binding, 2000;155:995–998.
4. Ohashi PS, Oehen S, Buerki K, et al. Ablation of tolerance and induc-
and titer. For instance, IgM cold agglutinins generally bind tion of diabetes by virus infection in viral antigen transgenic mice. Cell
poorly at physiologic temperatures and require high titers 1991;65:305–317.
for clinically significant effects. Some antibody isotypes are 5. Soldavila G, Geiger T, Flavell R. Breaking immunologic ignorance to an
most efficient at fixing complement, and anti-RBC alloanti- antigenic peptide of simian virus 40 large T antigen. J Immunol 1995;
bodies of these isotypes are of most concern for intravascular 155:5590–5600.
6. Krutzik SR, Modlin RL. The role of Toll-like receptors in combating
hemolytic reactions. Only IgG isotype anti-RBC antibodies mycobacteria. Semin Immunol 2004;16:35–41.
pass the placenta and cause hemolytic disease of the new- 7. Selsted ME, Ouellette AJ. Mammalian defensins in the antimicrobial
born. IgE antibodies bind to specific IgE receptors on mast immune response. Nat Immunol 2005;6:551–557.
cells and when crosslinked activate the mast cells to secrete 8. Janeway CA Jr, Medzhitov R. Innate immune recognition. Annu Rev
Immunol 2002;20:197–216.
histamine and other BRMs that generate allergic symptoms. 9. Miller SI, Ernst RK, Bader MW. LPS, TLR4 and infectious disease diver-
Sometimes we can take advantage of the properties of an sity. Annu Rev Immunol 2005;3:36–46.
antibody. For example, the rate of IgM diffusion into the 10. Iwasaki A, Medzhitov R. Toll-like receptor control of the adaptive
extravascular space is low compared with IgG. This facilitates immune responses. Nat Immunol 2004;5:987–995.
IgM removal by plasmapheresis. 11. Murphy PM. The molecular biology of leukocyte chemoattractant
receptors. Annu Rev Immunol 1994;12:593–633.
12. Hedrick JA, Zlotnik A. Chemokines and lymphocyte biology. Curr Opin
Immunol 1996;8:343–347.
CONCLUSION 13. Munoz-Fernandez MA, Fernandez MA, Fresno M. Synergism between
tumor necrosis factor-α and interferon-γ on macrophage activation for
the killing of intracellular Trypanosoma cruzi through a nitric oxide-
The immune system is a dynamic web of interacting cells and dependent mechanism. Eur J Immunol 1992;22:301–307.
molecules that protect our bodies from harm. In the practice 14. Trinchieri G. Interleukin-12 and the regulation of innate resistance and
of transfusion medicine, we may only become aware of the adaptive immunity. Nat Rev Immunol 2003;3:133–146.
immune system when our patients manifest symptoms, such 15. Takeda K, Akira S. Toll-like receptors in innate immunity. Int Immunol
as when they become alloimmunized or develop allergic reac- 2005;17:1–14.
16. Lecellier CH, Dunoyer P, Arar K, et al. A cellular microRNA mediates
tions. However, underlying these responses are not only end- antiviral defense in human cells. Science 2005;308:557–560.
effector molecules, cytokines, or antibodies, but also complex 17. Gao LH, Zheng SS. Cytomegalovirus and chronic allograft rejection in
pathways of innate and adaptive immunity involving many liver transplantation. World J Gastroenterol 2004;10:1857–1861.
BLOOD BANKING 18. Cainelli F, Vento S. Infections and solid organ transplant rejection: 49. Bakke O, Doberstein B. MHC class II associated invariant chain contains
a cause-and-effect relationship? Lancet Infect Dis 2002;2:539–549. a sorting signal for endosomal compartments. Cell 1990;63:707–716.
19. Falcone M, Bloom BR. A T helper cell 2 (Th2) immune response against 50. Amigorena S, Drake J, Webster P, Mellman I. Transient accumulation of
non-self antigens modifies the cytokine profile of autoimmune T cells new MHC molecules in a novel endocytic compartment in B lympho-
and protects against experimental allergic encephalomyelitis. J Exp Med cytes. Nature 1994;369:113–120.
1997;185:901–907. 51. Castellino F, Germain R. Extensive trafficking of MHC class II invariant
20. Stohlman SA, Pei L, Cua DJ, et al. Activation of regulatory cells sup- chain complexes in the endocytic pathway and appearance of peptide
presses experimental allergic encephalomyelitis via secretion of IL-10. loaded class II in multiple compartments. Immunity 1995;2:73–88.
J Immunol 1999;163:6338–6344. 52. Sallusto F, Sella M, Danieli C, Lanzavecchia A. Dendritic cells use
21. Aharoni R, Teitelbaum D, Sela M, Arnon R. Bystander suppression macropinocytosis and the mannose receptor to concentrate macro-
of experimental autoimmune encephalomyelitis by T cell lines and molecules in the major histocompatibility complex class II compart-
clones of the Th2 type induced by copolymer 1. J Neuroimmunol ment: downregulation by cytokines and bacterial products. J Exp Med
1998;91:135–146. 1995;182:389–400.
22. Marciani DJ. Vaccine adjuvants: Role and mechanisms of action in vac- 53. Bonnerot C, Lankar D, Hanau D, et al. Role of B cell receptor Ig α and
cine immunogenicity. Drug Discov Today 2003;8:934–943. β subunits in MHC class II-restricted antigen presentation. Immunity
23. Toyonaga B, Mak T. Genes of the T-cell antigen receptor in normal and 1995;3:335–347.
malignant T cells. Annu Rev Immunol 1987;5:585–620. 54. Lanzavecchia A. Mechanisms of antigen uptake for presentation. Curr
24. Chothia C, Boswell DR, Lesk AM. The outline structure of the T cell αβ Opin Immunol 1996;8:348–354.
receptor. EMBO 1988;7:3745–3755. 55. Sykulev Y, Brunmark A, Jackson M, et al. Kinetics and affinity of reac-
25. Malissen M, Trucy J, Jouvin-Marche E, et al. Regulation of TCR alpha tions between an antigen-specific TCR and peptide–MHC complexes.
and beta gene allelic exclusion during T-cell development. Immunol Immunity 1994;1:15–22.
Today 1992;13:315–322. 56. Matsui K, Boniface J, Steffner P, et al. Kinetics of T cell receptor binding
26. Tse H, Schwarz R, Paul W. Cell–cell interactions in the T cell prolifera- to peptide/IEk complexes: correlation of the dissociation rate with T-
tive response. J Immunol 1980;125:401. cell responsiveness. Proc Natl Acad Sci USA 1994;91:12862–12866.
27. Carbone FR, Belz GT, Heath WR. Transfer of antigen between migrating 57. Janeway CA Jr. The T cell receptor as a multicomponent signalling
and lymph node-resident DCs in peripheral T-cell tolerance and immu- machine: CD4/CD8 coreceptors and CD45 in T cell activation. Annu
nity. Trends Immunol 2004;25:655–658. Rev Immunol 1992;10:645–674.
28. Man AL, Prieto-Garcia ME, Nicoletti C. Improving M cell mediated 58. Li QJ, Dinner AR, Qi S, et al. CD4 enhances T cell sensitivity to antigen
transport across mucosal barriers: do certain bacteria hold the keys? by coordinating Lck accumulation at the immunological synapse. Nat
Immunology 2004;113:15–22. Immunol 2004;5:791–799.
29. Frauwirth KA, Thompson CB. Activation and inhibition of lympho- 59. Weiss A, Littman D. Signal transduction by lymphocyte antigen recep-
cytes by costimulation. J Clin Invest 2002;109:295–299. tors. Cell 1994;76:263–274.
30. Guermonprez P, Valladeau J, Zitvogel L, et al. Antigen presentation and 60. Demotz S, Grey HM, Sette A. The minimal number of class II MHC–
T cell stimulation by dendritic cells. Annu Rev Immunol 2002;20:621–667. antigen complexes needed for T cell activation. Science 1990;249:
31. Schwartz RH. T cell anergy. Annu Rev Immunol 2003;21:305–334. 1028–1030.
32. Heath WR, Carbone FR. Cross-presentation, dendritic cells, tolerance 61. Starr TK, Jameson SC, Hogquist KA. Positive and negative selection of
and immunity. Annu Rev Immunol 2001;19:47–64. T cells. Annu Rev Immunol 2003;21:139–176.
33. Malissen B, Ardouin L, Lin SY, et al. Function of the CD3 subunits of the 62. Boehm U, Klamp T, Groot M, Howard JC. Cellular responses to inter-
pre-TCR and TCR complexes during T cell development. Adv Immunol feron-γ. Annu Rev Immunol 1997;15:749–795.
1999;72:103–148. 63. Wilson HL, O’Neill HC. Murine dendritic cell development: difficulties
34. Davis MM, Bjorkman PJ. T-cell antigen receptor genes and T-cell recog- associated with subset analysis. Immunol Cell Biol 2003;81:239–246.
II nition. Nature 1988;334:395–402. 64. Park CS, Choi YS. How do follicular dendritic cells interact intimately
35. Sprent J. T lymphocytes and the thymus. In Paul W (ed). Fundamental with B cells in the germinal centre? Immunology 2005;114:2–10.
28 Immunology. New York, Raven Press, 1989, pp 69–93. 65. Huang AY, Qi H, Germain RN. Illuminating the landscape of in vivo
36. Campbell R, Trowsdale J. Map of the human MHC. Immunol Today immunity: insights from dynamic in situ imaging of secondary lym-
1993;14:349–352. phoid tissues. Immunity 2004;21:331–339.
37. Bjorkman PJ, Saper MA, Samraoui B, et al. The foreign antigen binding 66. Kabelitz D, Holtmeier W. γδ T cells link innate and adaptive immune
site and T cell recognition regions of class I histocompatibility antigens. responses. Chem Immunol Allergy 2005;86:151–183.
Nature 1987;329:512–518. 67. Shin S, El-Diwany R, Schaffert S, et al. Antigen recognition determi-
38. Bjorkman PJ, Saper MA, Samraoui B, et al. Structure of the human class nants of γδ T cell receptors. Science 2005;308:252–255.
I histocompatibility antigen, HLA-A2. Nature 1987;329:506–512. 68. Brandes M, Willimann K, Moser B. Professional antigen-presentation
39. Fremont D, Hendrikson W, Marrack P, Kappler J. Structures of an function by human γδT cells. Science 2005;309:264–268.
MHC class II molecule with covalently bound single peptide. Science 69. Mackay CR. Migration pathways and immunologic memory among
1996;272:1001–1004. T lymphocytes. Semin Immunol 1992;4:51–8.
40. Falk K, Rotzsche O, Stevanovic S, Jung G, Rammensee H. Allele specific 70. Mossman T, Coffman R. Th1 and Th2 cells: different patterns of lym-
motifs revealed by sequencing of self peptides eluted from MHC phokine secretion lead to different functional properties. Annu Rev
molecules. Nature 1991;351:290–296. Immunol 1989;7:145–173.
41. Hunt D, Henderson R, Shabanowitz J, et al. Characterization of peptides 71. Abbas A, Murphy M, Sher A. Functional diversity of helper T lympho-
bound to the class I molecule HLA-A2.1 by mass spectrometry. Science cytes. Nature 1996;383:787–793.
1992;255:1261–1263. 72. Romagnani S. Type 1 T helper and type 2 T helper cells: functions,
42. Parham P, Ohta T. Population biology of antigen presentation by MHC regulation and role in protection and disease. Int J Clin Lab Res
class I molecules. Science 1996;272:67–74. 1991;21:152–158.
43. Sherman LA, Chattopadhyay S. The molecular basis of allorecognition. 73. Cher D, Mossman T. Two types of murine helper T cell clone. II.
Annu Rev Immunol 1993;11:385–402. Delayed type hypersensitivity is mediated by Th1 clones. J Immunol
44. Groettrup M, Soza A, Kuckelkorn U, Kloetzel P. Peptide antigen produc- 1987;138:3688–3694.
tion by the proteasome: complexity provides efficiency. Immunol Today 74. Coffman R, Seymour B, Lebman D, et al. The role of helper T cell prod-
1996;17:429–435. ucts in mouse B cell differentiation and isotype regulation. J Immunol
45. Groettrup M, Khan S, Schwarz K, Schmidtke G. Interferon-gamma 1988;102:5–28.
inducible exchanges of 20S proteasome active site subunits: why? 75. Bernhagen J, Bacher M, Calandra T, et al. An essential role for macro-
Biochimie 2001;83:367–372. phage migration inhibitory factor in the tuberculin delayed-type hyper-
46. Lehner P, Cresswell P. Processing and delivery of peptides presented by sensitivity reaction. J Exp Med 1996;183:277–282.
MHC class I molecules. Curr Opin Immunol 1996;8:59–67. 76. Grabbe S, Schwarz T. Immunoregulatory mechanisms involved in
47. Koopmann J, Post M, Neefjes J, et al. Translocation of long peptides by elicitation of allergic contact hypersensitivity. Immunol Today 1998;
transporters associated with antigen processing (TAP). Eur J Immunol 19:37–44.
1996;26:1720–1728. 77. Siraganian RP. Mast cell signal transduction from the high-affinity IgE
48. Sadasivan B, Lehner P, Ortmann B, et al. Roles for calreticulin and a receptor. Curr Opin Immunol 2003;15:639–46.
novel glycoprotein, tapasin, in the interaction of class I MHC molecules 78. Capron A, Dessaint JP. Immunologic aspects of schistosomiasis. Annu
with TAP. Immunity 1996;5:103–114. Rev Med 1992;43:209–218.
109. Wraith DC, Nicolson KS, Whitley NT. Regulatory CD4+ T cells
The existence of the complement system has great impact BASIC BIOCHEMISTRY OF THE
on the practice of transfusion medicine. In its normal COMPLEMENT SYSTEM
immune activities, complement functions to kill patho-
gens, mediate inflammation, maintain the solubility of Classical Pathway Activation
immune complexes, promote a normal adaptive immune
response, and opsonize particles for phagocytosis. Activation of the complement system occurs via two path-
However, complement can also mediate pathogenic pro- ways; with the activation of C3 (the third component of
cesses, including anaphylaxis, intravascular hemolysis of complement), these pathways join to form a common path-
transfused blood cells, leukocyte mobilization in transfu- way that completes the cascade (Fig. 3–1). The primary func-
sion-related acute lung injury (TRALI), and activation of tion of both pathways is the generation of enzyme complexes
platelets. that activate C3 by cleaving it to C3b. The antibody-mediated
The group of proteins known to constitute the comple- activation of complement occurs by the classical pathway, so
ment system was first recognized in the 1880s as the labile called because it was the first pathway recognized. Activators
bactericidal activity in serum.1 Paul Ehrlich coined the term of the classical pathway include not only antibody molecules,
for the phenomenon by proposing a model of antibody- but also several nonimmunoglobulin proteins (Table 3–1).
mediated cytotoxicity in which a serum factor complements Only immunoglobulins (Ig) of the M and G isotypes activate
the bactericidal activity of antibody. A detailed understand- complement by the classical pathway. In humans, IgG3 and
II ing of the biochemistry of the complement system required IgG1 are strong complement activators, but IgG2 is a poor
the development of techniques that would permit the iso- activator, and IgG4 does not activate complement. These dif-
30 lation of individual complement proteins; this technology ferences result fro m variation in the ability of the different
finally appeared in the 1960s. With the explosion of research IgG subclasses to bind the first component of complement,
activity in the complement field in the 1970s came the rec- C1. The ability of an antibody to activate complement, with
ognition that this system was much more complex than had the accompanying opsonization and perhaps lysis of the cell,
been imagined. This complexity is amply demonstrated by parallels the opsonic potential of the IgGs themselves. The
the fact that at least 25 complement proteins are involved varying risk of phagocytic destruction by crystallizable frag-
in the activation and regulation of that activity known to ment (Fc) receptor-mediated recognition of IgG is an impor-
Ehrlich simply as complement. tant feature in distinguishing clinically significant antibodies
The understanding of the complement system is made from those with less destructive potential.
easier by setting a proper context. Like coagulation, the C1 is a multisubunit complex that contains the initial
complement system is an activated enzyme cascade. In such antibody-binding subunits, C1q, as well as two types of
cascades, proteins normally circulate in an inactive form, the zymogen subunits, C1r and C1s, that acquire serine prote-
zymogen. When the pathway is initiated, the first protein in ase activity on activation of the complex. Each molecule of
the sequence is converted from a zymogen to an activated C1 contains six C1q subunits, and two each of the C1r and
enzyme, which acts on the next protein zymogen in the cas- C1s subunits. The fixation of a C1 molecule to the surface
cade. Such pathways are amplifying, because each enzyme of the cell by the C1q subunits requires a minimum of one
molecule generated can act on multiple substrate molecules. molecule of IgM or at least two molecules of IgG for efficient
Activated enzyme pathways are also characterized by the activation. C1q itself contains six identical subunits com-
presence of regulatory proteins, both humoral and cellular, posed of a triple helical region with homology to collagen
that prevent the activated enzymes from converting all avail- and a globular domain at the distal end. The proximal end
able substrate. of C1q is associated with the other subunits of the C1 com-
The nuances of the complement system have long struck plex, C1r and C1s, in a calcium-dependent manner. Two of
fear into the hearts of basic scientist and clinician alike. the six C1q subunits must be bound by antibody to effect
Clinical aspects of complement biology in the pathophysi- activation. Although a single molecule of IgM is capable of
ology of disease have been reviewed by Morgan2 and more activating complement, it must be bound to antigen, where
recently by others.3,4 This chapter describes the central role it assumes a staple-shaped conformation. Fluid-phase or
of complement in many physiologic processes, including planar IgM does not activate complement, inasmuch as the
those associated with the use of blood components. C1q-binding sites are exposed only in the staple form.5
Figure 3–1 The activation pathway of complement.
31
Once two subunits of C1q are bound to antibody, the activated C1 complex is the initiation complex of the classi-
molecular conformation changes, in that the angle between cal pathway.
the subunits is greatly reduced; the resulting stress on the Although the majority of classical pathway activation is
molecule facilitates the autocatalysis of the C1r subunits.6 On antibody mediated, antibody-independent activation of the
autoactivation, C1r acquires serine protease activity, which classical pathway has been described in several situations (see
is directed against the C1s subunits. Once C1s is cleaved by Table 3–1). Such activation may involve the direct binding
activated C1r, it also acquires serine protease activity. This of C1q to a surface, or it may be mediated by other plasma
Surfaces containing IgG1 and IgG3; IgG2 weakly; IgM Some dialysis membranes, especially cuprophane
N-acetylglucosamine Some negatively charged surfaces Desialated erythrocytes
and mannose Crystalline cholesterol Surfaces that promote the binding of factor B
IgA complexes
Protein aggregates
Microbial pathogens
Tumor cells
BLOOD BANKING proteins, such as C-reactive protein (CRP) or mannose- alternative arises from the fact that this mechanism of C3
binding lectin. Activated C1s has the next two proteins in cleavage was discovered many decades after the classical
the pathway, C4 and C2, as its substrates. The C4 molecule is pathway. It should in no way be considered a secondary path-
cleaved by C1s to C4a and C4b. Some of the C4b molecules way of complement activation, because it is the only pathway
bind covalently to the cell surface through a reactive thiol of complement that can respond to microorganisms in the
ester bond that is exposed when the C4a fragment is removed; absence of specific antibody. It is, therefore, a front-line host
most C4b molecules are inactivated by hydrolysis and never defense. Activators of the alternative pathway include a broad
bind to a cell surface. In a magnesium-dependent reaction, a spectrum of substances, from renal dialysis membranes to
C2 molecule binds to a molecule of C4b; the C2 is then also immune complexes to microorganisms (Table 3–2). The ini-
cleaved by C1s. This cleavage results in the generation of the tial step in the activation of the alternative pathway is the
active serine protease, C2a, which remains associated with generation of a partially activated molecule of C3, which has
C4b. Clearly, the cleavage of both C4 and C2 by the same mol- been described as C3b-like.13 This molecule is presumably
ecule of C1s is sensitive to the local geometry; deposition of generated by the low-grade, spontaneous hydrolysis of C3
the C4b molecule far from the activated C1 molecule results that occurs in the body and is not the result of the presence
in the termination of the activation pathway. of the activator substance per se. C3(H2O) has the important
The bimolecular complex, C4b2a, is the C3-convert- characteristic of being able to interact in the fluid phase with
ing complex of the classical pathway. The cleavage of C3 by the complement protein factor B in a magnesium-depen-
C4b2a results in the formation of two fragments, C3a and dent manner. Once factor B has associated with C3(H2O),
C3b. C3b, like C4b, contains a reactive thiol ester bond that it is cleaved by factor D, a circulating serine protease that
enables it to bind covalently to the cell surface. The intramo- has specificity for factor B bound to C3(H2O). This cleavage
lecular thiol ester present within C3 is formed by a transac- results in the formation of two fragments, Ba and Bb. The
ylation reaction between the thiol group of 988Cys and the Bb fragment has serine protease activity and remains associ-
gamma amide group of 991Gln.7 The exposure of this reac- ated with C3(H2O). The bimolecular complex C3(H2O)Bb
tive center results in the interaction with cell surface moieties has C3 as its substrate. The C3(H2O)Bb complex cleaves C3
through the formation of an ester or an amide bond. in the same way as C4b2a, and the resulting C3b molecules
As indicated earlier, the activation of the classical pathway can bind covalently to the cell surface. If C3b binds to an
does not require immunoglobulin. The most recent develop- activator surface, factor B associates with it and is cleaved
ments in complement biochemistry have included an appre- by factor D. This results in the formation of an alternative
ciation for the role of molecules of the pentraxin and lectin pathway C3-converting complex on the surface of the activa-
families in the activation of complement. Pentraxins, a fam- tor. Because this complex has C3 as its substrate, it produces
ily of molecules named for their cyclic pentameric subunit a feedback amplification loop for the deposition of C3b onto
structure, include two proteins that have been implicated in the activator.
complement activation: CRP and serum amyloid P.8–10 Both The binding of a C3b molecule in the immediate vicinity
II proteins bind C1q, thereby initiating the classical pathway of a C4b2a or C3bBb enzyme complex produces a trimolecu-
without a requirement for antibody. lar complex that is capable of cleaving C5. The C3b molecule
32 has a binding site for C5; if the geometry is appropriate, C5
is presented to the enzyme complex, where it is cleaved by
Lectin Pathway Activation
C2a or Bb serine proteases. The cleavage produces the frag-
The carbohydrate-binding properties of plant-derived lectins ments C5a and C5b. Although there is a great deal of homol-
are well known in the field of blood banking, in which they ogy among C3, C4, and C5, there is no reactive thiol ester
are useful phenotyping reagents. Lectins or lectin-like pro- in the C5 molecule. Therefore, the cleavage of C5 does not
teins are also found in mammals. One of the C-type animal generate a fragment capable of binding covalently to a cell
lectins, mannose-binding lectin (MBL), belongs to a group surface. The C5b molecule remains briefly associated with
of soluble pattern recognition receptors. These molecules C3b before its association with the terminal proteins of the
and their membrane-bound cousins the Toll-like receptors, complement pathway.
recognize pathogen-associated molecular patterns located on
infectious organisms. MBL recognizes mannose-containing
Formation of the Membrane
carbohydrates and also binds N-acetylglucosamine. MBL
binding to its substrates has been reported to mediate com- Attack Complex
plement activation.11 The hexameric form of MBL is structur- Membrane attack complex of complement refers to the associ-
ally similar to C1q and will bind the C1r2s2 complex. MBL ation of the complement proteins C5, C6, C7, C8, and C9 to
is found in association with one of three forms of a serine form a potentially cytolytic complex. When C5 is activated in
protease, MBL-associated protease (MASP1, MASP2, or either the classical or alternative pathway, the resulting C5b
MASP3). Complement activation may be mediated by either molecule contains binding sites for the next components in
C1r2s2 or by the direct action of MASPs. The relative biologi- the pathway. This part of the complement pathway is still
cal significance of the pentraxin and lectin pathways is not a cascade, but the activation of a component results in the
fully understood.12 However, both CRP and MBL are acute- exposure of binding sites for other terminal complement
phase reactants and may reach significant serum concentra- proteins rather than in the acquisition of enzymatic activ-
tion during infection or inflammation. ity. C5b, but not C5, contains a binding site for C6, which
becomes bound while the C5b molecule is associated with
Alternative Pathway Activation its C3b tether. The C5b6 complex may be released from C3b,
or it may remain anchored until it binds a molecule of C7.
A different C3-converting complex is generated by the acti- Once C7 is attached, the trimolecular complex undergoes a
vation of the alternative pathway of complement. The term transition in which the normal hydrophilic character of the
PRINCIPLES OF THE COMPLEMENT SYSTEM CENTRAL TO TRANSFUSION MEDICINE
Table 3–2 Characteristics of the Proteins of Complement
Average Serum
Concentration
Pathway Protein Molecular Weight (kDa) Structure (mg/mL)
individual members of the complex is lost and a transient ment of the proteins is the simplest of these, inasmuch
hydrophobic character is acquired. At this time, C5b-7 may as several of the key enzymes function only when surface
associate with a membrane through the hydrophobic region; associated. The continuing activation of the complement
if it fails to interact with a membrane, the complex inacti- pathway is highly dependent on the spatial arrangement 3
vates by self-aggregation or by interaction with the inhibitors of the enzyme and substrate molecules. For example, a C2
described later. molecule associated with a C4b molecule that has been 33
Once C5b-7 is membrane associated, a binding site for deposited more than 60 nm from the activated C1 will
C8 is exposed in C5b. The binding of C8 causes the com- not be activated. Also, the enzymatic proteins of comple-
plex to insert more deeply into the membrane. C9 binds ment all occur in multisubunit or multimolecular com-
to C8 and undergoes significant conformational changes. plexes, each of which has a divalent cation requirement.
Not only does it acquire considerable hydrophobic charac- The reduction of divalent cation concentration, as well as
ter, pushing the membrane attack complex deeper into the the inherent dissociation constant of the protein–protein
membrane, but it also gains affinity for other molecules of interaction, means that the complement enzyme complexes
C9, which polymerize into the complex. Electron micro- simply fall apart.
scopic studies suggest that the C5b-9(n) complex resembles An important way in which the body controls comple-
a hollow cylinder. This cylinder is formed by the polymer- ment activation is through the action of specific protein
ized C9 molecules, which number between 12 and 18 in inhibitors of the complement proteins. These proteins are of
each complex. The C5b-8 is thought to play little active role two sorts: plasma proteins and cellular membrane proteins
in the structure of the cylindrical pore; it is, however, the (Table 3–3). Cells that are constantly exposed to the plasma
essential catalyst for the pore’s formation. The biochemi- milieu have evolved their own defense mechanisms against
cal characteristics of the complement proteins are given in complement activation. Selective pressure to evolve defense
Table 3–2. proteins is present because the complement system does not
distinguish between “good” and “bad” targets. A variety of
proteins that are involved in complement regulation on cell
The Regulation of Complement Activation surfaces have been described.14
The plasma proteins that inhibit complement work at all
Plasma Proteins steps of complement activation. The only known inhibitor
As discussed previously, the activation of the complement of activated C1 is the C1 inhibitor protein, which is also an
system is expressed as the activation of several enzymes of inhibitor of kallikrein, plasmin, factor XIa, and factor XIIa.
the serine protease group. As with any zymogen-to-enzyme C1 inhibitor inactivates C1 by binding to the active sites of
conversion, there must be a way of regulating the activity of the C1r and C1s subunits with very high affinity; this binding
the enzyme, or else it will cleave all available substrate mol- has been suggested to be covalent. Because the C1 complex is
ecules. With the complement proteins, this regulation is composed of two subunits each of C1r and C1s, the inactiva-
achieved by multiple mechanisms. The geometric arrange- tion of the complex requires the binding of two molecules
BLOOD BANKING
Table 3–3 Characteristics of the Complement Regulatory Proteins
Average Serum
Molecular Weight Concentration
Pathway Protein (kDa) Structure (mg/mL)
Proteins Homology
Figure 3–2 The domain structure of the terminal complex proteins. This figure illustrates the relative homologies among the proteins of the mem-
brane attack complex. The highly homologous regions are found within cysteine-rich domains types I, II, and III (according to Morgan2). In addition,
C6 and C7 contain cysteine-rich short consensus repeats (SCRs). (From Anderson KC, Ness PM. Scientific Basis of Transfusion Medicine, 2nd ed.
Philadelphia, Saunders, 2000.)
BLOOD BANKING
Table 3–5 Chromosomal Assignments of the Complement Proteins
DAF, delay accelerating factor; MCP, membrane cofactor protein; MHC, major histocompatibility complex; RCA, regulators of complement
activation.
enzyme, carboxypeptidase R, has been identified as the pri- some types of filters appear to remove C3a that is generated
mary inactivator of kinin as well as anaphylatoxin peptides.43 during processing and storage.45
Unlike carboxypeptidase N, carboxypeptidase R is itself rap- The membrane attack complex itself causes significant
idly inactivated under normal purification conditions. After activation in a wide variety of cell types. Because most stud-
removal of the arginine, the peptide acquires the designation ies of the C5b-9 complex have focused on its lytic effect on
“des arg.” C3a and C4a are rapidly and completely inacti- erythrocytes, the functional effect on nucleated cells and
vated by the two carboxypeptidases; C5a is somewhat more platelets has lately gained appreciation. Although there are
resistant to inactivation. In addition, neutrophils can still some reports in the literature that C5b-7 has chemotactic
respond to C5a des arg, albeit with some 103 weaker affinity; activity for neutrophils, it is the fully formed C5b-9 that
the removal of arginine from C3a and C4a results in com- has the greatest effect on these cells. In comparison with
plete loss of biological activity. erythrocytes, nucleated cells are very resistant to C5b-9
The relative resistance to carboxypeptidases and retention lysis. In response to C5b-9, nucleated cells and platelets
of bioactivity make C5a the most physiologically important coalesce the C5b-9 complexes and bud them off in vesicles
of the anaphylactic complement peptides. Generation of of plasma membrane. In nucleated cells, especially neutro-
II C5a causes many of the effects seen in inflammation that phils, this process is accompanied by the activation of the
are mediated by neutrophils. After activation by the bind- cell and the release of enzymes, leukotrienes, prostaglan-
38 ing of C5a into specific receptors, neutrophils bind to the dins, thromboxanes, and reactive oxygen metabolites from
capillary endothelium and migrate through the vessel wall, the cell.46
after the concentration gradient of the C5a. Once they are in Platelets interact with the activated proteins of the com-
contact with the higher concentrations of C5a present at the plement system at several levels.47 Specific platelet receptors
site of complement activation, neutrophils release granule for C1q have been identified.48–50 Although the precise role
contents and reactive metabolites, including lysozyme, reac- of C1q receptors remains to be determined, one such recep-
tive oxygen species, and eicosanoids. Although this is part of tor has been reported to play a role in phagocytosis in other
the normal mechanism of response to tissue injury or infec- cell types.51 The exposure of human platelets to C3a alters the
tion, the generation of large amounts of C5a or its presence response to physiologic agonists but does not induce plate-
in inappropriate locations can cause significant damage to let aggregation. The effect of C5a on human platelets has not
uninvolved tissues. been investigated. Membrane attack complex can also affect
platelet function.52 C5b-9 can induce the formation of plate-
Other Complement Activation Peptides let membrane vesicles (or microparticles), increase the pro-
coagulant activity of the platelet, and cause some degree of
Other peptides generated during the course of complement arachidonic acid generation.53 Studies of the effects of C5b-9
activation have been identified as inducing cell activation or on platelets have been carried out in purified protein systems;
chemotaxis. Factor Ba, the peptide cleaved from factor B by the significance of these observations for the whole plasma
factor D, is a weak chemotactic factor. The cleavage prod- system remains to be determined. The platelet may modify
ucts of factor B—Ba and Bb—have been reported to inhibit the effects of activated complement fragments through the
and stimulate B-cell proliferation, respectively. Peptides with action of its surface regulatory proteins, DAF, MCP, C8-bind-
functions similar to those of C3a and C5a can be generated ing protein, and CD59. Platelets also contain internal pools of
by the action of noncomplement enzymes, particularly plas- vitronectin, which may contribute to the local modulation of
min, on C3 and C5. These C3a-like and C5a-like peptides complement. The significant effects of activated complement
may play a role in the activation of both platelets and white fragments on the physiologic processes of platelets must
cells under existing blood storage protocols. The use of nega- be considered in the setting of platelet concentrate storage.
tively charged leukoreduction filters may significantly affect Complement is activated under storage conditions,54 with or
the levels of various contact activation peptides in blood without prestorage leukoreduction. Activated complement
products.44 In general, negatively charged artificial mem- fragments as well as C5b-9 may contribute to the platelet
brane surfaces promote complement activation; however, storage lesion.55
PRINCIPLES OF THE COMPLEMENT SYSTEM CENTRAL TO TRANSFUSION MEDICINE
EFFECTS OF COMPLEMENT ACTIVATION system. If a macrophage is already activated, it will bind and
ON CELL SURVIVAL ingest cells that bear only C3b. Normal, resting macrophages
require that IgG also be present on the erythrocyte surface.
This experimental result is supported by the in vivo observa-
Cytotoxic Effects
tion that the erythrocytes of patients with cold agglutinin dis-
In the transfusion medicine setting, the effects of comple- ease circulate through the spleen bearing C3b but no IgG and
ment activation are graphically illustrated by the acute intra- are not sequestered. That is not to say that C3 has no in vivo
vascular transfusion reaction. The generation of C3a and role in clearance. Bacteria coated with C3 only are efficiently
C5a produces bronchospasm and hypotension. C5a can also phagocytized by macrophages, resting or activated, even in
stimulate the production of interleukin-1 from macrophages, the absence of IgG. Macrophages in the liver (Kupffer cells)
thereby causing fever.56 The recruitment of large numbers of are capable of clearing erythrocytes coated with C3b only.
neutrophils to the lung by C5a generation produces ventila- In addition, persons who are genetically deficient in one of
tion–perfusion abnormalities. This effect may have devastat- the components of the early classical pathway are unable to
ing consequences for the recipients of transfusion products bind C3 to anti-D-coated erythrocytes; those target cells are
who develop TRALI subsequent to the fixation of comple- cleared from the circulation much more slowly than in per-
ment by anti-HLA antibodies. Two other hallmarks of the sons with intact complement systems.60
acute intravascular transfusion reaction, hemoglobinemia The opsonic and cytotoxic effects of complement are
and hemoglobinuria, result from the activation of comple- important in the ex vivo setting of blood storage. Many
ment on the red cell surface in sufficient amounts to over- types of bacteria that have been implicated in transfusion-
whelm the cellular and plasma control proteins, thereby mediated sepsis are either lysed or opsonized by the com-
lysing the cell. The generation of activated complement pro- plement proteins in the blood unit. These bacteria are then
teins affects more than red cell survival. The activation of engulfed by phagocytes also present in the bag. The removal
cells with the concomitant release of enzymes contributes to of phagocytes by leukodepletion within 24 hours of collec-
the activation of the coagulation system, as does the action of tion results in the removal of contaminating bacteria as well.
complement proteins on coagulation substrates and endo- Leukodepletion more than 24 hours after collection may fail
thelial cells. In vitro studies suggest that these processes may to sterilize the unit, because white cell breakdown may result
trigger the disseminated intravascular coagulation seen in in the release of viable bacteria from phagolysosomes.
severe cases of hemolytic transfusion reaction. Evidence for
the direct cytolysis of platelets or granulocytes during trans- The Role of Complement Receptors in
fusion is not abundant, perhaps because of the difficulty in Cell Clearance and in Maintenance of the
constructing an adequate study design. However, comple- Immune Response
ment activation may be directly linked to platelet destruc-
tion in the setting of paroxysmal nocturnal hemoglobinuria, The removal of complement-coated target cells is mediated
sepsis, and thrombotic thrombocytopenic purpura/hemo- by specific receptors for C3 and its activation and degrada- 3
lytic uremia syndrome. Each of these conditions is associ- tion fragments. As previously discussed, the activation of C3
ated with complement activation and platelet dysfunction. produces two fragments, C3a, the anaphylatoxin, and C3b, 39
The complement-mediated destruction of antibody-coated which is covalently bound to the cell surface. In the pres-
platelets can be experimentally induced. Antibody to the ence of factor H, MCP, or the complement receptor CR1,
human platelet antigen P1A1 (HPA-1) fixes sufficient comple- C3b is rapidly cleaved by factor I to an inactivated form,
ment to lyse target platelets.57 In vitro platelet lysis can also iC3b (Fig. 3–3). This cleavage is estimated to occur in vivo
be induced by cold agglutinin anti-I antibodies58; antibodies within 5 minutes of the generation of C3b. Inactivated C3b
of this specificity have been proposed to mediate the throm- undergoes another, slower interaction with factor I that
bocytopenia associated with Epstein-Barr virus infection. results in a second cleavage on the opposite side of the thiol
ester bond from the initial cleavage. This second cleavage
occurs within 30 minutes of generation of C3b and results
Opsonic Effects
in the generation of the C3c fragment, which is no longer
The opsonic effects of complement activation result in the tethered to the cell, and the C3dg fragment, which contains
accelerated clearance of particles or cells that bear C3 as the thiol ester bond and remains bound to the cell surface.
well as IgG. Although complement activation is the primary Under laboratory conditions, the C3dg fragment can be fur-
mechanism of cell destruction in the acute hemolytic trans- ther cleaved by trypsin to leave the C3d fragment on the cell
fusion reaction, the extravascular destruction of erythrocytes surface; the frequency of generation of this fragment in vivo
does not have the absolute requirement of complement acti- is uncertain.
vation. The primary clearance mechanism is through the Several complement receptors have been identified to
phagocyte Fc receptor with recognition of the IgG present date. They have overlapping cell distribution (Table 3–6).
on the cell surface.59 Opsonization of cells by complement, The specificity of these receptors for individual fragments of
rather than through lysis, means that the cell has been able to C3 is reflected in the biological response to receptor occu-
regulate complement effectively to prevent the assembly of pation. The central function of the complement receptor
cytolytic membrane attack complexes. This regulation may CR1 is clearance of immune complexes from the circulation
reflect the titer, avidity, affinity, or thermal amplitude of the through its interaction with C3b contained in the complex.61
antibody in its interaction with the cell target. The bulk of the total CR1 in the circulation is present on
In addition, as described previously, only IgG1 and IgG3 red cells, because they provide the greatest mass of cells in
are efficient complement activators. The presence of C3b (or the peripheral blood. The erythrocyte is therefore essential
its degradation products) on the cell target in addition to in transporting immune complexes from the plasma to the
antibody accelerates the clearance by the reticuloendothelial resident macrophages in the spleen.
BLOOD BANKING The further processing of C3b to iC3b enables the com-
plement fragment to interact with the complement receptor
CR3, also known as CD11b/CD18. This receptor, with its dis-
tribution on monocytes, macrophages, and neutrophils, plays
a major role in immune-mediated phagocytosis. CR3 belongs
to a family of related receptor molecules that have the β chain
CD18. The pairing of the CD11a α chain to CD18 defines
the leukocyte function-associated antigen 1, which is impor-
tant in mediating killing by T lymphocytes. CD18 may also
pair with the α chain CD11c, forming the complex known as
p150,95. This molecule has been unofficially designated CR4
in recognition of its binding affinity for iC3b.
The importance of the complement receptors CR3 and
CR4 is indicated by the severity of the deficiency state, in
which patients have defective phagocytic function and are
susceptible to recurrent infections.62 The degradation of
iC3b by factor I produces C3dg, the principal ligand for
the receptor CR2. The expression of CR2 is restricted to
B lymphocytes, on which the binding of C3dg triggers B-
cell activation and proliferation. In addition to its role as a
C3dg-binding protein, CR2 is the cellular receptor through
which the Epstein-Barr virus gains entry to B lymphocytes.
The identification and characterization of the complement
receptors has led to the development of new therapeutic
modalities. For example, a recombinant CR1 has been engi-
neered with the transmembrane region missing. This mol-
ecule retains the complement-regulatory ability of CR1 but
is soluble in plasma. In animal models, this protein, sCR1,
has proved efficacious in reducing the complement activa-
tion seen during thrombolytic therapy.63
Fragments of C3 are important in maintaining the
immune response.64 Animals with an experimental deple-
tion of C3 fail to mount a normal IgG response to secondary
II immunization. This observation suggests that C3 is impor-
tant in the development of immunologic memory, but such
40 a role for C3 has not been confirmed in humans. It has also
Figure 3–3 Degradation fragments of C3. Native C3 is cleaved by been established that the complement receptors CD21 and
C4b2a or C3bBb, resulting in the exposure of a reactive thiol ester
in the α chain of C3 and the generation of two fragments, C3a and CD35 are important in the regulation of B-cell immunity
C3b. C3b is inactivated by factor I in the presence of factor H or CR1 while the complement regulatory proteins CD46 and CD55
to iC3b, which remains bound to the surface by the thiol ester. A fur- have an additional role in T-cell function through the regula-
ther cleavage by factor I on the other side of the covalent bond from tion of cytokine production.65,66
the first cleavage results in the release of the C3c fragment; the C3dg
fragment remains surface bound. In vitro, the C3dg fragment may be
further degraded to C3d, C3e, and C3g. The C3d fragment, a terminal Therapeutic Complement Inhibition
breakdown product of C3, remains covalently bound to the surface.
(From Anderson KC, Ness PM. Scientific Basis of Transfusion Medicine, The great advances in our understanding of both comple-
2nd ed. Philadelphia, Saunders, 2000.) ment system biochemistry and the role of complement in
*
CR designation is unofficial.
PRINCIPLES OF THE COMPLEMENT SYSTEM CENTRAL TO TRANSFUSION MEDICINE
pathophysiology have led to the creation of strategies to tion peptides of either the classical (C4d) or alternative (fac-
control complement activation in order to minimize its del- tor Bb) pathway or both (C3a, C5a, iC3b, and C5b-9) are
eterious effects.67 Several promising compounds are in pre- currently available. In addition, descriptions of many other
clinical development; one monoclonal antibody against C5 monoclonal antibody-based activation peptide assays can
has been used successfully in a clinical trial in the treatment now be found in the literature. These activation-dependent
of PNH to reduce ongoing hemolysis.68 Although comple- assays enable definite differentiation between patients with
ment inhibitors have been proposed for use in acute hemo- complement activation and those with decreased production
lytic transfusion reactions, no clinical studies have yet been of complement proteins.
performed.69 Until recently, the assessment of patients with suspected
complement deficiencies has relied on somewhat cumber-
some gel methods. However, a new ELISA-based procedure
LABORATORY ANALYSIS OF has been developed for detection of complement deficien-
COMPLEMENT cies, including those due to a loss of proteins in the lectin
pathway.72 The continuing improvement of testing methods
Measurement of Cell-Associated should make these assays more accessible to nonspecialists.
In summary, the burgeoning of research activity on the
Complement
complement system has produced a clearer understanding
The methods available for the measurement of cell-associ- of the biochemistry of these important proteins. This knowl-
ated immunoglobulins are readily adaptable to the detec- edge has led to the development of better diagnostic tools, to
tion of cell-bound complement. Traditionally, the presence an increased clarity around the role of complement in dis-
of cell-associated C3b and its cleavage products is detected ease pathophysiology, as well as to the invention of therapeu-
by means of an aggregating anti-C3d antibody (C3 Coombs’ tic modalities to control unwanted complement activation.
test). The presence of small amounts of C4d antigen on the
erythrocyte is identified in the blood bank as the Chido-
Rodgers blood group antigen system. The antigenic differ- REFERENCES
ence is actually determined by the C4A and C4B alleles, with 1. Ross GD. Immunobiology of the Complement System. Orlando, Fla.,
Rodgers specificity found in the C4A isotypes and Chido Academic Press, 1986, pp 1–19.
specificity found in the C4B isotypes. Techniques devel- 2. Morgan BP. Complement: Clinical Aspects and Relevance to Disease.
oped to measure cell-associated immunoglobulin can all be New York, Academic Press, 1990.
adapted to measure cell-bound complement, especially C3. 3. Sjoholm AG, Jonsson G, Braconier JH, et al. Complement deficiency
and diseases: an update. Mol Immunol 2006;43:78–85.
These methods include fluorescence flow cytometry, radio- 4. Thiel S, Frederiksen PD, Jensenius JC. Clinical manifestations of man-
immunoassay, and enzyme-linked immunosorbent assay nan-binding lectin deficiency. Mol Immunol 2006;43:86–96.
(ELISA).70,71 In addition, the development of monoclonal 5. Sim RB, Reid KBM. C1: Molecular interactions with activating systems.
antibodies that can distinguish among the various fragments Immunol Today 1991;12:307–311. 3
6. Perkins SJ, Nealis AS. Solution structure of human and mouse immu-
of C3 may prove useful in laboratory diagnosis. noglobulin M by synchrotron X-ray scattering and molecular graphics
modelling. J Mol Biol 1991;221:1345–1366. 41
7. Levine RP, Dodds AW. The thiolester bond of C3. Curr Top Microbiol
Measurement of Complement Activation Immunol 1989;153:73–82.
or Deficiency States 8. Szalai AJ, Briles DE, Volanakis JE. Role of complement in C-reactive
protein-mediated protection of mice from Streptococcus pneumoniae.
Until the mid-1980s, there were no readily accessible tests of Infect Immunol 1996;64:4850–4853.
complement activation. The activation of complement was 9. Bristow CL, Boackle RJ. Evidence for the binding of human serum amy-
implied by a decrease in the level of an individual compo- loid P component to C1q and Fab. Mol Immunol 1986;23:1045–1052.
nent, most often C4, in a clinical setting in which activation 10. Pepys MB, Blatz ML. Acute phase proteins with special reference to C-
reactive protein and related proteins (pentraxins) and serum amyloid
was suspected. Although this approach may be of some util- A protein. Adv Immunol 1983;34:141–212.
ity in the outpatient setting, it is of little use in the investi- 11. Lu JH, Thiel S, Wiedemann H, et al. Binding of the pentamer/hexamer
gation of very sick patients, especially those receiving blood forms of mannan-binding protein to zymosan activates the proenzyme
products. The problem with this diagnostic approach is two- C1rC1s2 complex of the classical pathway of complement, without
involvement of C1q. J Immunol 1990;144:2287–2294.
fold. First, reduction in the C4 level can occur from decreased 12. Fujita T, Matushita M, Endo Y. The lectin-complement pathway—its role
synthesis of the protein in the liver rather than from con- in innate immunity and evolution. Immunol Rev 2004;198:195–202.
sumption. Second, because of the presence of null genes for 13. Lachmann PJ, Hughes-Jones NC. Initiation of complement activation.
C4, the normal range for C4 is very wide; a person with a Springer Semin Immunopathol 1984;7:143–162.
C4 level that usually is at the high end of the normal range 14. Devine DV. The regulation of complement on cell surfaces. Trans Med
Rev 1991;5:123–131.
may activate more than 50% of C4 before the measured level 15. Donaldson VH, Bissler JJ. C1 inhibitors and their genes: an update.
exceeds the normal range. Only laboratories with specialized J Clin Lab Med 1992;119:330–222.
complement expertise had methods permitting the detec- 16. Hillarp A, Dahlback B. The protein S-binding site localized to the cen-
tion of activation peptides of complement, generally by gel tral core of C4b-binding protein. J Biol Chem 1987;262:11300–11307.
17. Hammer CH, Jacobs RM, Frank MM. Isolation and characterization
electrophoresis techniques or by radioimmunoassay of small of a novel plasma protein which binds to activated C4 of the classical
complement fragments remaining in solution after precipi- complement pathway. J Biol Chem 1989;264:2283–2291.
tation of native protein and large fragments. 18. Farries TC, Lachmann PJ, Harrison RA. Analysis of the interaction
With the advent of monoclonal antibody technology, between properdin and factor B, components of the alternative pathway
antibodies have been made that are capable of recognizing C3 convertase of complement. Biochem J 1988;253:667–675.
19. Murphy BF, Kirszbaum L, Walker ID, et al. SP-40,40, a newly identified
neoantigens expressed in the activation peptides of comple- normal human serum protein found in the SC5b-9 complex of comple-
ment. These antibodies have been used in the formulation ment and in the immune deposits in glomerulonephritis. J Clin Invest
of commercial ELISA kits. ELISA tests that detect the activa- 1988;81:1858–1864.
BLOOD BANKING 20. Choi NH, Mazda T, Tomita M. A serum protein Sp-40,40 modulates cell-removal filter and measurement of venous blood bradykinin level in
the formation of membrane attack complex of complement on eryth- patients who received filtered platelets. Transfusion 1997;37:457–462.
rocytes. Mol Immunol 1989;26:835–840. 45. Snyder EL, Mechanic S, Baril L, et al. Removal of soluble biologic
21. Jenne DE, Lowin B, Peitsch MC, et al. Clusterin (complement lysis response modifiers (complement and chemokines) by a bedside white
inhibitor) forms a high density lipoprotein complex with apolipopro- cell-reduction filter. Transfusion 1996;36:707–713.
tein A-I in human plasma. J Biol Chem 1991;266:11030–11036. 46. Morgan BP. Complement membrane attack on nucleated cells: Resis-
22. Nicholson-Weller A, Burge J, Fearon DT, et al. Isolation of a human eryth- tance, recovery and non-lethal effects. Biochem J 1989;264:1–14.
rocyte membrane glycoprotein with decay accelerating activity for C3 47. Devine DV. The effects of complement activation on platelets. Curr Top
convertases of the complement system. J Immunol 1982;129:184–189. Microbiol Immunol 1992;178:101–113.
23. Medof ME, Kinoshita T, Nussenzweig V. Inhibition of complement acti- 48. Peerschke EIB, Ghebrehiwet B. Human blood platelets possess specific
vation on the surface of cells after incorporation of decay-accelerating binding sites for C1q. J Immunol 1987;138:1537–1541.
factor (DAF) into their membranes. J Exp Med 1984;160:1558–1578. 49. Nepomuceno RR, Henschen-Edman AH, Burgess WH, et al. cDNA clon-
24. Pangburn MK. Differences between the binding sites of the comple- ing and primary structure analysis of C1qR(P), the human C1q/MBL/
ment regulatory proteins DAF, CR1, and factor H on C3 convertases. SPA receptor that mediates enhanced phagocytosis in vitro. Immunity
J Immunol 1986;136:2216–2221. 1997;6:119–129.
25. Kinoshita T, Medof ME, Nussenzweig V. Endogenous association of 50. Nepomuceno RR, Tenner AJ. C1qRP, the C1q receptor that enhances
decay-accelerating factor (DAF) with C4b and C3b on cell membranes. phagocytosis, is detected specifically in human cells of myeloid lineage,
J Immunol 1986;136:3390–3395. endothelial cells, and platelets. J Immunol 1998;160:1929–1935.
26. Berger M, Medof ME. Increased expression of complement decay accel- 51. Butko P, Nicholson-Weller A, Wessels MR. Role of complement and
erating factor during activation of human neutrophils. J Clin Invest complement receptor C1qR in the antibody-independent killing of
1987;79:214–220. group B streptococcus. Adv Exp Med Biol 1997;418:941–943.
27. Seya T, Atkinson JP. Functional properties of membrane cofactor pro- 52. Sims PJ, Wiedmer T. The response of human platelets to activated com-
tein of complement. Biochem J 1989;264:581–588. ponents of the complement system. Immunol Today 1991;12:338–342.
28. Schoermark S, Rauterberg, EW, Shin ML, et al. Homologous species 53. Wiedmer T, Esmon CT, Sims PJ. Complement proteins C5b-9 stimulate
restriction in lysis of human erythrocytes: a membrane-derived pro- procoagulant activity through platelet prothrombinase. Blood 1986;
tein with C8-binding capacity functions as an inhibitor. J Immunol 68:875–880.
1986;136:1772–1776. 54. Bode AP, Miller DT, Newman SL, et al. Plasmin activity and comple-
29. Zalman LS, Wood LM, Muller-Eberhard JH. Isolation of a human eryth- ment activation during storage of citrated platelet concentrates. J Lab
rocyte membrane protein capable of inhibiting expression of homolo- Clin Med 1989;113:94–102.
gous complement transmembrane channels. Proc Nat Acad Sci USA 55. Gyongyossy-Issa MIC, McLeod E, Devine DV. Complement activation
1986;83:6975–6979. in platelet concentrates is surface-dependent and modulated by the
30. Holguin MH, Frederick LR, Bernshaw NJ, et al. Isolation and charac- platelets. J Lab Clin Med 1994;123:859–868.
terization of a membrane protein from normal human erythrocytes 56. Dalmasso AP. Complement in the pathophysiology and diagnosis of
that inhibits reactive lysis of the erythrocytes of paroxysmal nocturnal human disease. CRC Crit Rev Clin Lab Sci 1986;24:123–283.
hemoglobinuria. J Clin Invest 1989;84:7–17. 57. Cines DB, Schreiber RD. Effect of anti-P1A1 antibody on human plate-
31. Sugita Y, Nakamo Y, Tomita M. Isolation from human erythrocytes of lets: I. The role of complement. Blood 1979;53:567–577.
a new membrane protein which inhibits the formation of complement 58. Dixon RH, Rosse WF. Mechanisms of complement-mediated activa-
transmembrane channels. J Biochem (Tokyo) 1988;104:633–637. tion of human blood platelets in vitro: comparison of normal and
32. Okada N, Harada R, Fujita T, et al. A novel membrane glycoprotein paroxysmal nocturnal hemoglobinuria platelets. J Clin Invest 1977;59:
capable of inhibiting membrane attach by homologous complement. 360–368.
Int Immunol 1989;1:205–208. 59. Lutz HU. Innate immune and non-immune mediators of erythrocyte
II 33. Stefanova I, Hilgert I, Kirstofova H, et al. Characterization of a broadly clearance. Cell Mol Biol 2004;50:107–116.
expressed human leukocyte antigen MEM-43 anchored in membrane 60. Schreiber AD. An experimental model of immune hemolytic anemia.
42 through phosphatidylinositol. Mol Immunol 1989;26:153–161. Ann Intern Med 1977;87:211–217.
34. Lockert DH, Kaufman KM, Chang CP, et al. Identity of the segment of 61. Schifferli JA, Ng YC, Peters DK. The role of complement and its receptor in
human complement C8 recognized by complement regulatory protein the elimination of immune complexes. N Engl J Med 1986;315:488–495.
CD59. J Biol Chem 1995;270:19723–19728. 62. Anderson DC, Springer TA. Leukocyte adhesion deficiency: an inher-
35. Husler T, Lockert DH, Kaufman KM, et al. Chimeras of human comple- ited defect in the MAC-1, LFA-1, and p150,95 glycoproteins. Annu Rev
ment C9 reveal the site recognized by complement regulatory protein Med 1987;38:175–194.
CD59. J Biol Chem 1995;270:3483–3486. 63. Weisman HF, Bartow T, Leppo MK, et al. Soluble complement recep-
36. Parker CJ, Omine M, Richards S, et al. Diagnosis and management of tor type 1: in vivo inhibitor of complement suppressing post-ischemic
paroxysmal nocturnal hemoglobinuria. Blood 2005;106:3699–3709. myocardial inflammation and necrosis. Science 1990;249:146–151.
37. Telen MJ, Green AM. The Inab phenotype: characterization of the 64. Erdei A, Fust G, Gergely J. The role of C3 in the immune response.
membrane protein and complement regulatory defect. Blood 1989;74: Immunol Today 1991;12:332–337.
437–441. 65. Liu J, Miwa T, Hilliard B, et al. The complement inhibitory protein
38. Merry AH, Rawlinson VI, Uchikawa M, et al. Studies on the sensitiv- DAF (CD55) suppresses T cell immunity in vivo. J Exp Med 2005;201:
ity to complement-mediated lysis of erythrocytes (Inab phenotype) 567–577.
with a deficiency of DAF (decay accelerating factor). Br J Haematol 66. Wagner C, Hansch GM. Receptors for complement C3 on T-lym-
1989;73:248–253. phocytes: relics of evolution or functional molecules? Mol Immunol
39. Yamashina M, Ueda E, Kinoshita T, et al. Inherited complete deficiency of 2006;43:22–30.
20-kilodalton homologous restriction factor (CD59) as a cause of parox- 67. Mollnes TE, Kirschfink M. Strategies of therapeutic complement inhibi-
ysmal nocturnal hemoglobinuria. N Engl J Med 1990;323:1184–1189. tion. Mol Immunol 2006;43:107–121.
40. Farries TC, Atkinson JP. Evolution of the complement system. Immunol 68. Hill A, Hillmen P, Richards SJ, et al. Sustained response and long-term
Today 1991;12:295–300. safety of eculizumab in paroxysmal nocturnal hemoglobinuria. Blood
41. Marcus D, Alper CA. Methods for allotyping complement proteins. In 2005;106:2559–2265.
Rose N, Manual of Clinical Immunology. Washington, D.C., American 69. Yazdanbakhsh K, Kang S, Tamasauskas D, et al. Complement receptor
Society for Microbiology, 1986, pp 185–196. 1 inhibitors for prevention of immune-mediated red cell destruction:
42. Winkelstein JA, Colten HR. Genetically determined disorders of the Ppotential use in transfusion therapy. Blood 2003;15:5046–5052.
complement system. In The Metabolic Basis of Disease. St. Louis, 70. Garratty G. The significance of IgG on the red cell surface. Trans Med
McGraw-Hill, 1987, pp 2711–2737. Rev 1987;1:47–57.
43. Campbell W, Okada H. An arginine specific carboxypeptidase gener- 71. Schwartz KA. Platelet antibody: review of detection methods. Am J
ated in blood during coagulation or inflammation which is unrelated Hematol 1988;29:106–14.
to carboxypeptidase N or its subunits. Biochem Biophys Res Comm 72. Seelen MA, Roos A, Wieslander J, et al. Functional analysis of the
1989;162:933–939. classical, alternative and MBL pathways of the complement system:
44. Shiba M, Tadokoro K, Sawanobori M, et al. Activation of the contact sys- standardization and validation of a simple ELISA. J Immunol Meth
tem by filtration of platelet concentrates with a negatively charged white 2005;296:187–198.
Chapter 4
Principles of Red Blood Cell Allo- and
Autoantibody Formation and Function
James C. Zimring
Erythrocyte blood group antigens are polymorphic, inher- requiring transfusion has an unexpected blood group anti-
ited carbohydrate or protein structures located on the outside body. Although the laboratory can characterize an antibody
surface of the red blood cell (RBC) membrane. Our ability in terms of specificity, immunoglobulin class, and in vitro
to detect and identify blood group antigens and antibodies characteristics, it cannot always predict the antibody’s clini-
has contributed significantly to the current safe, supportive cal significance. Therefore, when a blood group antibody is
blood transfusion practice and to the appropriate manage- detected, an important, although often overlooked, first step
ment of pregnancies at risk for hemolytic disease of the fetus is to review all available pertinent patient information (Table
and newborn (HDN). Exposure to erythrocytes carrying an 5–1). One can then assess the potential for adverse effects by
antigen lacking on the RBCs of the recipient can elicit an correlating the serologic information with the patient history
immune response in some individuals. Thus, blood group and also with historical clinical experience in other patients.
antigens are clinically important in allogeneic blood transfu- A patient who has not been exposed to RBCs by transfusion 5
sions, maternofetal blood group incompatibility, and organ or pregnancy is unlikely to have a clinically significant allo-
transplantation. antibody. 53
By virtue of their relative ease of detection and generally
straightforward mode of inheritance, blood group antigens
have been used in genetic, forensic, and anthropologic inves- ERYTHROCYTE MEMBRANE
tigations. The polymorphisms of blood groups have been
exploited as a tool to monitor in vivo survival of transfused The erythrocyte membrane consists of lipids, proteins, and
RBCs, to monitor engraftment of bone marrow transplants, carbohydrates, which interact to form a dynamic and fluid
and to monitor for blood doping by allogeneic transfusion in structure. By dry weight, the ratio of protein-to-lipid-to-
sports and athletics. Antigen profiles have been used to pre- carbohydrate in the RBC membrane is 49:43:8. The RBC
dict inheritance of diseases encoded by a gene in close prox- membrane behaves as a semisolid, with elastic and viscous
imity to the gene encoding the blood group antigen (e.g., the properties that are not observed with simple lipid vesicles.
association of gene deletions causing chronic granuloma- These properties are critical for the RBC to survive in the cir-
tous disease with the loss of expression of Kell). Blood group culation for approximately 120 days during numerous cycles
antigens have also contributed to our understanding of cell (approximately 75,000) and passages through narrow veins
membrane structure. The lack of Kell/Kx expression is asso- and sinusoids in the spleen. The RBC accomplishes this goal
ciated with acanthocytes, Rh-deficient RBCs are stomato- without intracellular machinery to repair damage. The mul-
cytic, and mild elliptocytosis occurs in RBCs lacking Gerbich tiple connections between the membrane skeleton and the
glycophorin C (GPC) and/or glycophorin D (GPD). lipid bilayer cause the bilayer to follow the contours of the
More recently, in the postgenomic era, knowledge about membrane skeleton. Together, the membrane skeleton and
the molecular basis associated with blood group antigens lipid bilayer give the erythrocyte shape and resilience.1,2
and phenotypes is being applied to the detection of single
nucleotide polymorphisms (SNPs) associated with blood Lipids
group antigens. Microarray and microchip technologies hold
promise for transfusion medicine testing. The lipids in the RBC membrane form a bilayer, with the
This chapter reviews the scientific basis of molecules that hydrophobic tails on the inside and the hydrophilic polar
express RBC blood group antigens and their relevance and head groups to either the outside (extracellular) or the inside
application to the practice of transfusion medicine. One of (cytoplasmic) surface (Fig. 5–1). The following three types
the most important clinical concerns in contemporary trans- of lipids occur in the RBC membrane: phospholipid (50%),
fusion medicine occurs with the discovery that a patient cholesterol (40%), and glycolipids (10%). The arrangement
BLOOD BANKING
Table 5–1 Information for Problem-Solving in Immunohematology
Patient demographics Diagnosis, age, sex, ethnicity, transfusion and/or pregnancy history, drugs, intravenous
fluids (lactated Ringer’s solution, IVIgG, antilymphocyte globulin), infections, malignancies,
hemoglobinopathies, stem cell transplantation
Initial serologic results ABO, Rh, direct antiglobulin test, phenotype, antibody detection, autologous control,
crossmatch
Hematology/chemistry values Hemoglobin, hematocrit, bilirubin, lactate dehydrogenase, reticulocyte count, haptoglobin,
hemoglobinuria, albumin:globulin ratio, red blood cell (RBC) morphology
Sample characteristics Site and technique of collection, age of sample, anticoagulant, hemolysis, lipemic, color of
serum/plasma, agglutinates/aggregates in the sample
Other Check records in current and previous institutions for antibodies to blood group antigens
Antibody identification Autologous control, phase of reactivity, potentiator (saline, albumin, low-ionic-strength
solution, polyethylene glycol), reaction strength, effect of chemicals on antigen
(proteases, thiol reagents), pattern of reactivity (single antibody or mixture of antibodies),
characteristics of reactivity (mixed field, rouleaux), hemolysis, preservatives/antibiotics in
reagents, use of washed RBCs
of phospholipids in the bilayer is asymmetrical. The outer fluid and flexible. The RBC membrane skeleton is associ-
leaflet predominantly contains the neutral phospholipids ated with the lipid bilayer through specific interactions with
(phosphatidylcholine and sphingomyelin), and the inner transmembrane proteins.
leaflet predominantly contains the aminophospholip- Specific protein components of the RBC membrane skel-
ids (phosphatidylethanolamine and phosphatidylserine). eton, which is associated with the inner leaflet of the lipid
The presence of phosphatidylserine, which is negatively bilayer, interact with the cytoplasmic domains of some anti-
charged, on the inner monolayer results in a significant gen-carrying transmembrane proteins.3–5 Two major and
difference in charge between the two sides of the bilayer. well-defined interactions are ankyrin, which binds to spectrin
The lipid molecules can diffuse rapidly within their own in the membrane skeleton and the cytoplasmic domain of the
monolayer, but they rarely “flip-flop,” maintaining mem- multipass transmembrane protein, band 3 (anion exchanger,
brane “sidedness.”1 AE1),6–14 and protein 4.1, which provides a link between spec-
trin, actin, and p55 in the membrane skeleton to the single-pass
transmembrane proteins GPC and GPD.15–22 Some integral
Proteins
transmembrane proteins interact with other transmembrane
II Peripheral proteins form a meshwork under the lipid bilayer proteins, forming either small or large macromolecular com-
that is called the membrane skeleton. This name implies a plexes (e.g., glycophorin A (GPA) with glycophorin B (GPB)2;
54 relatively rigid structure; however, the meshwork is actually band 3 with GPA23–27; Kell with Kx28,29; RhD, RhCE, RhAG,
ABO
H
P Le NH2
GLOB Ii
Extracellular
Cromer
MNSs Lutheran
Rh* (12) Diego (14) Yt
Gerbich LW
Kell Duffy (7) Kx* (10) Colton (7) Dombrock
Indian Xg
Kidd (7) JMH
Knops Ok
EMM
Cytoplasmic
COOH NH2 COOH NH2 COOH
Figure 5–1 Diagram of a cross-section of the red blood cell membrane lipid bilayer and various membrane components that carry blood group
antigens. GPI, glycosylphosphatidylinositol.
LW (ICAM4), CD47, and GPB).4,30 The recent discovery that
Carbohydrates Terminology
Carbohydrates are essentially restricted to the extracellular A working party on terminology for RBC surface antigens,
surface of the RBC membrane, where they collectively form sanctioned by the International Society for Blood Transfusion
a negatively charged environment that is largely responsible (ISBT), has categorized the blood group antigens into four
for keeping the RBCs from adhering to one another and to classifications34: the genetically discrete blood group sys-
the endothelium. The majority of carbohydrates are attached tems (Table 5–3 summarizes the name, ISBT system number,
to lipids on ceramide and to proteins by attachment to aspar- chromosome location, gene name, associated antigens, com-
agine (N-linked) or to serine or threonine (O-linked) during ponent name, and possible functions for each system); blood
passage through the lumen of the Golgi.32 Some blood group group collections that consist of serologically, biochemically,
antigens are determined by the terminal carbohydrate resi- or genetically related antigens; 700 series of low-incidence
due (e.g., A and B) whereas others require the presence of a antigens; and 901 series of high-incidence antigens (Table
chain of carbohydrate residues (e.g., Leb and I). 5–4 contains the ISBT name, number, and associated anti-
The carbohydrates form the glycocalyx, a negatively gens. The chromosome location for these antigens has not
charged barrier approximately 10 Å thick around the out- yet been determined).
side of the RBC membrane. This barrier can keep immuno- As RBC antigens were discovered, notations were devised
globulin G (IgG) antibodies, particularly those recognizing to describe them. The terminology used is inconsistent: a
antigens that reside close to the lipid bilayer, from readily single letter (e.g., A, D, K), a symbol with a superscript (e.g.,
interacting with the corresponding antigen. Thus, the gly- Fya, Jkb, Lua), or a numerical notation (e.g., Fy3, Lu4, K12)
cocalyx affects the ability of an IgG antibody to cause direct is used. Even within the same blood group system, antigens
agglutination. have been named with different schemes, resulting in a cum- 5
bersome terminology for describing phenotypes. The use of
the same symbol with a different superscript letter (e.g., Fya 55
RBC BLOOD GROUP ANTIGENS and Fyb) indicates products of alleles (antithetical antigens).34
Some confusing terminology persists; for example, the P1
Figure 5–1 depicts the membrane components that are antigen is the sole antigen in the P blood group system, and
known to carry blood group antigens. Some antigens are car- the P antigen is the only antigen in the GLOB system. There
bohydrates attached to lipids or to proteins, some are protein, is also a GLOB collection, with Pk and LKE antigens. In clin-
and some require both protein and carbohydrates. Although ical practice, the traditional terminology is still extensively
the blood group antigens do not themselves have a function, used; this is the terminology that will be used throughout
the molecules on which they are carried do. The function of this chapter for the discussion of red cell antigens.
proteins carrying blood group antigens has been determined Collectively, there are over 260 antigens recognized by the
through observation of the morphology of RBCs that lack ISBT. We restrict discussion to the more commonly encoun-
the protein (Table 5–2), direct experimentation, or compari- tered antibodies and antigens.
son of the predicted protein sequence with protein databases
to identify similar proteins whose function is known, even if Inheritance
in other tissues or organisms.
Recognition of a blood group antigen begins with dis- Most blood group antigens are encoded by genes on auto-
covery of an antibody. When an individual whose RBCs somes.35 Most are codominant (e.g., M/N, E/e, K/k, Fya/Fyb),
B-CAM, cell adhesion molecule; GP, glycophorin; ISBT, International Society of Blood Transfusion; ISGN, International Society for Gene
Nomenclature.
Table 5–4 Other Blood Groups Whose Chromosome Locations Have Not Yet Been Determined
International Society of
Name Blood Transfusion Number Associated Antigens
Collections
Cost 205 Csa, Csb
i 207 i
Er 208 Era, Erb
Globoside 209 P, Pk, LKE
Unnamed 210 Lec, Led
Series
Low-incidence antigens 700 By, Chra, Bi, Bxa, Rd, Pta + 15 more
High-incidence antigens 901 Vel, Lan, Ata, Jra, JMH, Emm, AnWj, Sda, Duclos, PEL, ABTI, MAM
MEMBRANE BLOOD GROUP ANTIGENS AND ANTIBODIES
but some appear to be dominant if an antibody to the pre- able inferences about protein function. The understanding of
sumed antithetical antigen has not been discovered (e.g., Ula, blood groups has benefited from this revolution.
Cra). Some rare phenotypes appear to be inherited in a dom-
inant manner, for example, dominant type Lu(a−b−), or in a Natural Knockouts
recessive manner, for example, recessive type Lu(a−b−). Xga
and Kx are encoded by genes on the X chromosome and are The detection of an alloantibody to a high prevalence antigen
inherited in a classic X-linked manner. See Table 5–3 for the during compatibility testing or prenatal testing has led to the
chromosomal location of genes encoding blood groups. discovery of RBCs with null phenotypes. Such RBCs lack the
specific carbohydrate or carrier protein and therefore, all blood
group antigens within that system. Thus, these RBCs serve
Expression
as natural knockout models and these null phenotypes have
Some blood group antigens are also found on nonerythroid provided insights into the function of the carrier proteins.
cells. Some examples are A, B, H, Kna (CD35), Ina (CD44),
Oka (CD147), and Cromer-related antigens (CD55), which Function
have a wide tissue distribution.36,37
The ability to culture stem cells and sort erythroid lineages In general, the polymorphisms that we recognize as blood
and the availability of monoclonal antibodies have enabled the group antigens and that have significance in transfusion
estimation of the timing of expression of blood group anti- medicine do not appear to alter the function of the carrier
gens during in vitro erythroid maturation. The blood group molecule. The predicted functions are often based on the
antigens appear in the following order: GPC, Kell, RhAG, LW, function of closely related proteins in other tissues; however,
RhCE, GPA, band 3, RhD, Lutheran, and Duffy.38,39 their role in the mature RBC may not be the same as in other
cells, altered forms may function as recognition signals in
Maturation senescent RBCs, or they may have an important role during
earlier stages of erythroid development.5
Several blood group antigens are not expressed or are only The possible functions of the various components car-
weakly expressed on fetal RBCs and do not reach adult levels rying blood group antigens can be divided into the follow-
until a person is approximately age 2. Cord RBCs do not express ing broad categories: contributors to membrane structural
Lea, Sda, Ch, Rg, or AnWj antigens. Antibodies to these antigens integrity, transport proteins, receptors for extracellular
are unlikely to cause HDN, because RBC expression is a prereq- ligands, adhesion proteins, extracellular enzymes, comple-
uisite for HDN. Expression of A, B, H, P1, I, Leb, Lua, Lub, Yta, ment regulatory proteins, and maintainers of surface charge
Vel, Doa, Dob, Gya, Hy, Joa, Xga, Kn, and Bg is greatly reduced on in the glycocalyx (see Table 5–3).4,45,46,53–55 Specific details
RBCs from cord blood compared to adult RBCs. In contrast, the can be found in the chapters that follow.
i and LW antigens are more strongly expressed on RBCs from
cord blood than on RBCs from adults. Furthermore, although 5
adults express more LW on D-positive RBCs than on D-nega- BLOOD GROUP ANTIBODIES
tive RBCs, cord RBCs express LW antigens equally regardless of 57
D type. This information makes the testing of cord RBCs useful An antigen can be defined as a substance that will induce the
in antibody identification studies. production of antibodies, although there are certain prop-
erties of a molecule, such as foreignness, size, and chemical
complexity, that are associated with increased immunoge-
Molecular Genetic Basis
nicity. Proteins usually induce the most vigorous immune
Of the 29 genes (or gene families) that encode blood group anti- responses, followed by carbohydrates, whereas lipids and
gens, 28 have been cloned and sequenced (only the gene encod- nucleic acids are usually not strong immunogens, although
ing P remains to be clarified),40–44 and the genetic basis of many clinically significant antibodies specific for these types of
blood group antigens has been determined.45–48 This knowledge molecules do exist (e.g., antiphospholipid antibodies and
is being applied to transfusion medicine and is discussed later in anti-DNA antibodies found in some autoimmune diseases).
the chapter. Relevant details about blood groups and molecu- Most antigens require helper activity, usually in the form
lar knowledge are given in Chapters 6 through 8. Information of cytokines, from helper T cells to induce strong antibody
concerning the genetic basis of blood group antigens can also responses, and this helper activity is required for production
be obtained from the Blood Group Antigen Mutation Database: of antibody classes other than IgM. Some antigens, usually
www.bioc.aecom.yu.edu/bgmut/index.php. carbohydrate in nature, can induce antibodies in the absence
of helper T-cell activity, but these responses are primarily
IgM, with little, if any, antibodies of other classes produced.
Evolution
Antibodies are produced by B lymphocytes, also known as B
It has been known for many years from agglutination with cells. The basic antibody molecule consists of four polypeptide
human sera that blood group antigens have homologues in chains—two heavy chains of 50 to 70 kilodaltons, depending on
anthropoid apes.49 The availability of monoclonal antibodies the antibody class, and two light chains of approximately 25 kilo-
and molecular analysis of the gene homologues from nonhu- daltons. The two heavy chains produced by a B cell are identical,
man primates have contributed to defining the epitopes of as are the two light chains. In each polypeptide chain, whether
human blood group antigens.26,49–52 Additionally, because of it is a heavy chain or a light chain, approximately the first 100
the conserved nature of proteins, protein sequence comparisons amino acids are known as the variable region, and the remainder
are contributing to predictions about the function of human of the polypeptide chain constitutes the constant region, which is
proteins (see Table 5–3). The sequencing of the genomes of identical in all antibodies of the same class. The antigen-binding
many model organisms has allowed researchers to make test- site is formed by association of the variable regions of one heavy
BLOOD BANKING chain and one light chain, which means that each four-chain other clinically significant antibodies occur with an incidence
unit has two identical antigen-binding sites. However, in most of less than 1% of immunized patients.57–59 Antibodies that are
immune responses, a large number of B cells are stimulated by considered clinically insignificant unless the antibody reacts
an antigen, and each antigen-specific B cell will produce a unique in tests performed strictly at 37°C are anti-P1, anti-M, anti-
antibody, resulting in a heterogeneous response consisting of N, anti-Lua, anti-Lea, anti-Leb, and anti-Sda. Other clinically
many different antibody molecules directed toward multiple insignificant antibodies that react at 37°C in the indirect anti-
sites on the antigen. globulin test (IAT) are those of the Knops and Chido-Rodgers
systems and anti-JMH (Table 5–6).
The incidence of a blood group antibody depends on both
Immunogenicity
the prevalence in the population and the immunogenicity of
Several factors influence the ability to stimulate antibody the antigen. Immunized patients frequently produce mul-
production, including antigen size, complexity, and dose as tiple antibodies, and the more antibodies present, the more
well as host human leukocyte antigen genotype and other, as difficult they are to identify.
yet unidentified, susceptibility factors. Most carbohydrate-
based RBC antigens are T independent and therefore tend Detection and Identification
to elicit an IgM response. The protein-based antigens usu-
ally are T dependent and induce an IgM primary response Compatibility testing (testing patient’s serum against donor’s
that progresses to IgG.56 Antigen exposure usually occurs by RBCs) still uses techniques that were described 100 years ago
transfusion of products containing RBCs or during preg- for direct agglutination and 50 years ago for indirect aggluti-
nancy (immune antibodies) or by exposure to microbes nation. Even today, with our detailed understanding of blood
(apparently, naturally occurring antibodies). Table 5–5 sum- group antigens, we have no single technical procedure able to
marizes the usual type of immunoglobulin response and the detect all known blood group antibodies. The hemagglutina-
potential clinical significance, in transfusion or in HDN, of tion technique is simple and inexpensive, does not require
selected blood group antibodies, which are listed in order of sophisticated equipment, and when done correctly is sensi-
clinical significance.36 tive and specific in terms of clinical relevance. Agglutination
should be graded according to the strength of reaction, and
Clinical Significance an evaluation of the reaction strength can aid in identifica-
tion of antibodies, especially when multiple antibodies are
Antibodies recognizing antigens in the ABO blood group sys- present in a serum.33
tem are by far the most clinically significant. This is because The first blood group antigens to be identified were
they occur naturally in people whose RBCs lack the corre- those that could be agglutinated by the alloantibodies when
sponding antigen. Other clinically significant antibodies occur antigen-positive RBCs were suspended in a saline medium
in the following order, from the most commonly to the least (direct agglutination). This direct agglutination reflects the
II commonly encountered in transfusion practice: anti-D, anti- fact that these antibodies are usually IgM and detect carbo-
K, anti-E, anti-c, anti-Fya, anti-C, anti-Jka, anti-S, anti-Jkb. All hydrate antigens (ABO, P1, Le, and H antigens). Although
58
Clinical
Antibody Specificity IgM (Direct) IgG (Indirect) Transfusion Reaction HDN
Table 5–7 Reactivity of Antigen-Positive Red Blood Cells after Treatment with Ficin/Papain or DTT
†
variable with ficin/papain.
BLOOD BANKING
Patient has clinically significant antibody?
No Yes
Yes No
Screen inventory
at donor center
Test siblings/family
Use autologous
II
60 the first source to investigate for antigen-negative units. Most It is important not to delay surgery or transfusion unnec-
national blood donor centers can often assess national and essarily by attempting to obtain antigen-negative RBCs for
international Rare Donor Registries and can help provide patients with clinically insignificant antibodies. Also, in
the appropriate antigen-negative blood. hemolytic anemia due to warm-reactive autoantibodies,
compatibility may be difficult to demonstrate. The impor-
tant issue is to be sure that there are no underlying, clinically
COMPATIBILITY PROCEDURES significant alloantibodies. This fact can be determined with
autologous or homologous autoabsorptions,33 which require
Most laboratories follow recommendations made by extra time and, possibly, the services of an immunohema-
the American Association of Blood Banks,61 and all must tology reference laboratory. It is helpful to remember that
comply with regulations of state and federal agencies. patients without a history of immunization (transfusion or
Routine approaches to compatibility testing involve testing pregnancy) are unlikely to have underlying alloantibodies.
a blood sample from a prospective recipient for ABO and Unfortunately, a patient’s transfusion history is not always
D blood groups and for the presence of blood group anti- reliable, because many patients are unaware or forget that
bodies. In most cases, no unexpected antibodies are detected, they have received transfusions.62
and donor RBCs of appropriate ABO and D blood type
are selected for transfusion. A sample of the donor’s RBCs
may be tested (crossmatched) with the patient’s serum by CHOICE OF ANTIGEN-NEGATIVE BLOOD
means of either an immediate spin procedure or the IAT. FOR DISEASES REQUIRING LONG-TERM
Alternatively, the blood can be issued without direct cross- TRANSFUSION THERAPY
testing if a computer check is performed.33 Any blood typ-
ing problems should be resolved, and when antibodies are Transfusion management of patients who require long-term
detected, they should be identified. Knowing the specificity of transfusion therapy, in particular patients with sickle cell
an antibody helps establish whether it is likely to be clinically anemia, has been the subject of controversy and debate.63,64
significant and determine the approach necessary to provide There is still no consensus as to the best and most practi-
an adequate supply of compatible blood. Once an antibody cal approach, although the goal is to provide blood with
has been identified, antigen-negative RBC products selected maximal survival. In addition to the traditional practice
for transfusion should be tested with the patient’s serum by of providing antigen-negative blood only after the patient
the IAT to ensure compatibility. has made an antibody, approaches that have been adopted
MEMBRANE BLOOD GROUP ANTIGENS AND ANTIBODIES
approach. Indeed, in some regions, it is a challenge to provide
Table 5–8 Antigen-Negative Incidence—
antigen-negative blood to patients who have already made the
Common Polymorphic Antigens*
antibodies. Extensive screening for antigen-negative donors may
Incidence of Antigen Negativity be realized with the advent of microarray technology.
Although direct and indirect hemagglutination tests can give rise to the null phenotype, for example, for Rh,30,73
have served the transfusion community well for more than Kell,100 and Jk systems,101 and the p phenotype.102
100 and more than 50 years, respectively, in some aspects,
hemagglutination has limitations. For example, it gives
Antigen Identification by DNA Testing
only an indirect measure of the potential complications in
an at-risk pregnancy, it cannot indicate RHD zygosity in D- Although there are many molecular events that give rise to
positive people, it cannot be relied on to type some recently blood group antigens and phenotypes, the majority of geneti-
II transfused patients, and it requires the availability of specific cally defined blood group antigens are the consequence of an
reliable antisera. The characterization of genes and determi- SNP. Thus, simple DNA-based assays can be used to identify
62 nation of the molecular basis of antigens and phenotypes has defined SNPs in genes encoding blood groups. Innumerable
made it possible to use the polymerase chain reaction80,81 to DNA-based assays have been described to detect specific
amplify regions of DNA of interest to detect alleles encoding blood group SNPs. They include polymerase chain reaction
blood groups. The knowledge can also be applied to express (PCR)-restriction fragment length polymorphism (RFLP),
antigens in heterologous systems as a first step to detect and allele-specific (AS)-PCR, sequence-specific (SS)-PCR as
identify blood group antibodies in a single, objective, auto- single or multiplex assays, real-time quantitative PCR (Q-
mated assay. PCR; RQ-PCR), a single nucleotide dye terminator extension
method, and high throughput microarray technology. The
DNA Basis of Blood Group Antigens PCR-RFLP, AS-PCR, and SS-PCR assays can be visualized
in a gel or by electropherogram printouts. Semi-automated
Once a gene is shown to encode a protein that carries a blood methods that use a specific tag attached to the primers, which
group, focused analysis of people with known serologically can then be analyzed in a digital readable format by a mass-
defined antigen profiles is used to determine the molecular spectrometer or pyrosequencer, have tremendous potential
basis of variant forms of the gene. This approach has been in the clinical laboratory. Another exciting opportunity is
extremely powerful because antisera-based definitions of the use of microchip technology for the purpose of antigen
blood groups readily distinguish variants within each blood identification.103,104 The process uses a “chip,” which is com-
group system. The molecular basis associated with many posed of spots of DNA from many genes attached to a solid
blood group antigens has been determined over a relatively surface in a gridlike array. The fluorescent-labeled cDNA or
short period.30,46,54,73,82–99 The available wealth of serologi- DNA with a sequence that matches the sequence of one of
cally defined variants has contributed to the rapid rate with the gene fragments adheres and can be detected. The system
which the genetic diversity of blood group genes has been is most frequently used for studying genes involved in regula-
revealed. Initially, molecular information associated with tion and expression.105 However, it is now also being used to
each variant was obtained from only a small number of analyze specific blood group genes. Some clinical applications
samples and applied to analysis with the hopeful assumption of DNA analysis for blood groups are listed in Table 5–11.
that the molecular analysis would always correlate with RBC
antigen typing. With the gathering of more information it In the Transfusion Setting
became obvious that several molecular events can result in
FOR TRANSFUSION-DEPENDENT PATIENTS
discrepant genotype and RBC phenotype (some are listed in
Table 5–10). Furthermore, analyses of the null phenotypes Certain medical conditions, such as sickle cell disease,
have demonstrated that multiple, diverse genetic events thalassemia, autoimmune hemolytic anemia, and aplastic
MEMBRANE BLOOD GROUP ANTIGENS AND ANTIBODIES
Table 5–10 Examples of Events That Confound Analysis of Genotype and Phenotype
anemia, often require chronic blood transfusion. In this situ- transfusions.108–111 Determination of a patient’s blood type
ation or when a patient receives a massive transfusion, the by analysis of DNA is particularly useful when a patient who
presence of donor RBCs in the patient’s peripheral blood is transfusion-dependent has produced alloantibodies. This
makes RBC phenotyping by conventional hemagglutina- is because identification of the patient’s probable phenotype
tion techniques complex, time-consuming, and possibly allows the laboratory to determine to which antigens the
inaccurate. Indeed the interpretation of RBC typing results patient can and cannot respond to make alloantibodies.
of multitransfused patients, based on such parameters as
For Patients Whose RBCs Have a Positive DAT. DNA-based
number of units transfused, length of time between trans-
antigen typing of patients with autoimmune hemolytic anemia,
fusion and sample collection, and size of patient (the “best
whose RBCs are coated with immunoglobulin, is valuable when
guess”), is often incorrect.106 It is desirable to determine the
available diagnostic antibodies require the IAT and IgG removal
blood type of a patient as part of the antibody identifica-
techniques (e.g., EDTA-acid-glycine, chloroquine diphosphate)
tion process, and molecular DNA approaches are being used 5
are not effective at removing bound immunoglobulin, or when
to type patients to overcome this limitation of hemaggluti-
these techniques destroy the antigen of interest.
nation. Interference or false results from transfused donor 63
WBCs surviving in the recipient’s circulation depends on the For Donors. DNA-based assays can be used to antigen-type
design of the assay used.107 PCR assays designed for transfu- donor blood both for transfusion and for antibody identifica-
sion medicine testing do not detect post-transfusion DNA tion reagent panels. This is particularly useful when antibod-
chimerism, and blood group determinations can reliably be ies are not available or are weakly reactive. A good example
made using DNA prepared from a blood sample collected is the Dombrock blood group polymorphism, where DNA-
after the patient has received recent chronic or even massive based assays112–114 are used to type patients and donors for
Doa and Dob to overcome the problem of not having reliable
typing reagents. Furthermore, the newer DNA technologies
have the potential for screening pools of DNA for rare blood
Table 5–11 Clinical Applications of DNA types and thereby increasing the number of donors that can
Analysis for Blood Group Antigens be tested. As automated procedures attain higher and faster
throughput at lower cost, typing of blood donors by DNA-
To type patients who have been recently transfused based assays is likely to become more widespread, revolution-
To type patients whose RBCs are coated with
immunoglobulin izing the provision of antigen-negative blood to patients.
To identify a fetus at risk for hemolytic disease of the With donor DNA-based typing, the presence of a grossly
newborn normal gene whose product is not expressed on the RBC
To determine which phenotypically antigen-negative surface would lead to the donor being falsely typed as anti-
patients can receive antigen-positive RBCs
To type donors for antibody identification panels
gen-positive, and although this would mean loss of an
To type patients who have an antigen that is expressed antigen-negative donor, it would not jeopardize the safety of
weakly on RBCs the patient in blood transfusion practice.
To determine RHD zygosity
To mass screen for antigen-negative donors IN THE PRENATAL SETTING
To resolve blood group A, B, and D discrepancies
To determine the origin of engrafted leukocytes in a stem Hemagglutination titers give only an indirect indication of
cell recipient the risk and severity in HDN. Thus, antigen typing by DNA-
To determine the origin of lymphocytes in a patient with based assays has particular value in the prenatal setting to
graft-versus-host disease determine whether the fetus has inherited the paternal antigen
For tissue typing
For paternity and immigration testing
of clinical relevance. If the fetus is predicted to be antigen-
For forensic testing negative, the fetus is not at risk for HDN and the mother
need not be aggressively monitored.
BLOOD BANKING Specific criteria should be met before initiating fetal DNA the appropriate assay to detect a mutation that silences a gene
testing. The mother’s serum contains an IgG antibody of should be part of the DNA-based testing (e.g., GATA box
potential clinical significance and the father is heterozygous analysis with FY typing,137 presence of RHD pseudogene with
for the gene encoding the antigen of interest (or paternity RHD typing,133 and exon 5 analysis with GYPB(S) typing).139
is in doubt). It is helpful to know the ethnic origin and to Also, it is important to obtain an accurate medical history for
genotype both mother and father concurrently with the fetus the patient because with certain medical treatments, such as
to confirm the inheritance and to identify potential variants stem cell transplantation and kidney transplants, typing results
that could influence interpretation of the test results. in tests using DNA from different sources (such as WBCs,
Fetal DNA can be isolated from cells obtained by inva- buccal smears, or urine sediment) may differ. Thus, thought
sive procedures such as amniocentesis or chorionic villus should be given as to the purpose of the testing when estab-
sampling and by noninvasive procedures from trophoblasts lishing standard operating procedures for use in the clinical
collected by transcervical sampling115–117 and from fetal laboratory.
erythroblasts isolated from the maternal circulation.118 Before interpreting the results of DNA testing, it is impor-
Remarkably, cell-free fetal-derived DNA can be extracted tant to obtain an accurate medical history and to establish
from maternal serum or plasma119–122 and RHD typing is if the study subject is a surrogate mother, if she has been
possible.120,123–129 Use of cell-free fetal DNA from maternal impregnated with nonspousal sperm, or if she has received
plasma has overcome concerns that fetal lymphocytes from a stem cell transplant. This information, although critical,
previous pregnancies can persist in maternal blood130 or is often not provided. For prenatal diagnosis of a fetus at
skin131 for years. This is not a problem with cell-free fetal risk of HDN, the approach to genotyping should err on the
DNA because it clears the maternal circulation within 3 days side of caution.
of termination of the pregnancy.120,121 Fetal DNA obtained When performing any blood group DNA analysis in the
from maternal plasma appears to be less stable and reports prenatal setting, determining the RHD status of the fetus, in
vary regarding optimal sample storage.120,129 Large-scale tri- addition to the test being ordered, can be important. If the
als and standardization of protocols are still needed; how- fetus has a normal RHD, there is no need to provide Rh-
ever, it is likely that determination of fetal RHD using this negative blood for intrauterine transfusions. This is espe-
noninvasive procedure will become routine clinical practice. cially true if the mother has anti-c and fetal DNA is being
RHD testing for D expression is the prime target because, typed for RHc.
at least in the majority of white individuals, the D-negative
mother has a complete deletion of RHD, thereby permit- Antibody Identification
ting detection of the paternally derived fetal RHD gene.
For analysis of the inheritance of blood group antigens In addition to the value of typing donors for antibody iden-
determined by SNPs (e.g., K/k) rather than by the pres- tification panels when appropriate antibodies are not avail-
ence or absence of the gene, a sample source that contains able, DNA analyses are being used to aid in the identification
II primarily fetal DNA (e.g., amniocytes) is preferred. When of Dombrock (anti-Hy, anti-Joa), Knops (anti-Kna, anti-Sla,
isolating fetal DNA from maternal plasma, proper positive and anti-McCa), and Rh (anti-hrS, anti-hrB) antibodies.
64 controls that test for the presence of fetal DNA are critical. Determination of the phenotype of the patient allows more
Y chromosome markers are useful when the fetus is a male, accurate classification in Dombrock and Knops blood group
but polymorphic paternal markers are needed in the case systems.
of a female. DNA can be transfected into cells and grown in tissue
DNA analysis for the prediction of fetal D phenotype is culture to express blood group antigens. Indeed, single-pass
based on detecting the presence or absence of portions of (Kell) and multipass (Duffy) proteins have been expressed
RHD. A prerequisite to the accurate interpretation of DNA in high levels in mouse erythroleukemic cells or 293T cells
analysis in clinical applications is the extensive molecu- and detected by human polyclonal antibodies.140 Similar
lar characterization of apparently identical phenotypes in experiments have been performed on Lutheran antigens.141
mother and fetus. The Rh blood group system is complex, Thus, it is theoretically possible to produce a panel of cell
and many hybrid genes have been described. For example, in lines expressing individual proteins for development of an
Europeans, the molecular basis of the D-negative phenotype automated, objective, single-step antibody detection and
was, with few exceptions, associated with deletion of the identification procedure. Such an approach would eliminate
entire RHD, until a study of serologically D-negative sam- the need for antigen-matched, short-dated, potentially bio-
ples revealed nine novel RHD gene-positive haplotypes.95,132 hazardous RBC screening and panel products derived from
Approximately one third of Japanese D-negative people humans. However, although promising, some major hurdles
have an intact but inactive RHD. The majority of D-negative are yet to be overcome; for example, antigens from all blood
Africans and approximately one quarter of D-negative African group systems must be expressed at levels that are at least
Americans have an inactive RHD pseudogene (RHDΨ), and equivalent to those on RBCs and the detection system should
many others have hybrid RHD-CE-D genes.133 These hybrid have low background levels of reactivity. The highly clini-
genes complicate DNA testing and can lead to false results if cally significant Rh antigens are proving difficult to express
they are not understood by the investigator.30,74 in adequate levels.
When recommendations for clinical practice are based on Recombinant cells expressing blood group antigens can
molecular analyses, it is important to remember that, in rare be used for adsorption of specific antibodies as part of anti-
situations, a genotype determination will not correlate with body detection and identification procedures, or prior to
antigen expression on the RBC (see Table 5–10).48,95–97,133–138 crossmatching if the antibody is clinically insignificant. In
If a patient has a grossly normal gene that is not expressed addition, genes can be engineered to express soluble forms
(usually a null phenotype), he or she could produce an anti- of antigens for use in antibody inhibition, again as part of
body if transfused with antigen-positive blood. When feasible, antibody detection and identification procedures, or prior
to crossmatching.142–144 Concentrated forms of recombinant
II
68
Chapter 6
ABO and Related Antigens and Antibodies
Connie M. Westhoff ● Marion E. Reid
This chapter summarizes current knowledge of the blood chain. The terminal sugar determines antigen specificity:
groups composed of terminal carbohydrate moieties, the an N-acetylgalactosamine residue results in expression of
ABO, H, Lewis, Ii, and P systems (International Society of the A antigen, and a terminal galactose residue is responsible
Blood Transfusion [ISBT] system or collection numbers for the B antigen. These structures are similar, differing only
given in parentheses). The antigens are the products of the in that A antigen has a substituted amino group on carbon
action of glycosyltransferase enzymes, and they share bio- 2 (Fig. 6–1). The precursor substrate for A and B antigens is
chemical synthesis pathways and precursor framework oli- the H antigen (Fig. 6–2). The terminal fucose in α(1,2) link-
gosaccharide molecules. The antigens are carried on large age to galactose is responsible for H antigen specificity, and
oligosaccharide chains covalently linked to proteins (glyco- large amounts of H are present on group O RBCs, because
proteins), lipids (glycolipids), or both on the red blood cell H is not converted to A or B. Some H antigen precursor also
(RBC). In addition, some are expressed on tissues, and soluble remains on A and B RBCs, listed as follows in descending
forms can be found in various secretions and excretions. order of frequency: A2, B, A2B, A1, A1B.
RBC membrane proteins carry well over 2 × 106 A, B, and
H antigens (ABH) combined per RBC, and most (80%) are
ABO AND Hh SYSTEMS (ISBT SYSTEMS located on the major integral membrane protein, band 3
001 AND 018) (anion exchanger 1, or AE1), which is the RBC Cl−HCO3−
anion exchanger. In addition, the glucose transporter, Rh-
History associated glycoprotein (RhAG), and aquaporin-1 (Colton)
also carry A, B, and H, but in smaller amounts (reviewed
ABO was the first blood group system to be discovered. In in Lowe5). A, B, and H antigens are also found in lesser 6
1900, Landsteiner mixed sera and RBCs from his colleagues amounts as lipid-linked glycoconjugates associated with the
and observed agglutination.1 On the basis of the agglutina- RBC membrane via a ceramide moiety. 69
tion pattern, he named the first two blood group antigens After their discovery on RBCs, the ABH antigens were
A and B, using the first two letters of the alphabet. RBCs also found on many tissues. Large amounts are expressed on
not agglutinated by either sera were first called type C but endothelial and epithelial cells of the lung and gut and on the
became known as “ohne A” and “ohne B” (ohne is German epithelial cells of the urinary and reproductive tracts; hence,
for “without”) and finally O. Landsteiner received the Nobel they are called histo-blood group antigens. ABH antigens are
Prize in 1930, 30 years after the discovery of the ABO blood also found in secretions (particularly saliva) and fluids (milk
groups. and urine) of 80% of the population who have the secretor
Unfortunately, administering transfusions based on agglu- (Se) phenotype.6
tination reactions was largely ignored, and surgeons contin- ABH antigens are also found on platelets. The amount
ued performing direct donor-to-patient blood transfusions. of A or B antigen on platelets varies between individu-
This practice provided many opportunities to observe symp- als, with approximately 8% of blood group A and 10% of
toms of severe and fatal hemolytic transfusion reactions. Not B donors reported to have “high” level expression.7 Thus,
until methods were developed to store and preserve blood ABO incompatibility can compromise the outcome in plate-
(1914–1917) did blood transfusion move to the blood bank let transfusions. Recent studies have refocused attention on
environment, and World War I saw the first large-scale the interesting observation that platelets from donors with
transfusions based on serologic ABO selection of donors.2 an A2 phenotype lack both A and H antigens.8,9 The bio-
chemical basis for the lack of A antigen is not yet known, but
Antigens this observation has implications for transfusion practice in
that approximately 20% of group A platelets would be from
The ABO antigens were biochemically characterized in the A2 donors (see incidence of A2 phenotype). These platelets
1950s and 1960s (reviewed in Yamamoto3) as carbohydrate would be appropriate for “universal” use, and platelets from
structures on glycoproteins and glycolipids. The ABO A2 donors may also be a superior product for patients under-
system consists of four antigens with ISBT numbers—A, B, going A/O major mismatch allogeneic progenitor cell trans-
A,B, and A1—but there are additional subgroups (e.g., A2, plantation.9
A3, AX, Ael, B3), and Hh is a separate system.4 The antigens As tissue antigens, ABH are important in solid organ
are synthesized in a stepwise fashion by glycosyltransferase transplantation. Recipient antibodies react with antigens on
enzymes that sequentially add specific monosaccharide sugars the transplanted organ, and complement activation at the
in specific linkages to a growing oligosaccharide precursor surface of endothelial cells results in rapid destruction and
BLOOD BANKING 6CH
2OH
6CH OH
2
A2 has an average incidence of 20% of group A patients.15
5 O 5 O Subgroup A2 is rare in Asians.
HO H OH HO H OH Inherited and Acquired ABH Antigen Variants
4 1 4 1
H H H H Subgroups of A and B have weaker expression of the respec-
OH H OH H
tive antigens. The difference between A1 and A2 is quanti-
3 2 3 2
tative, because the number of A antigens is reduced on A2.
H NHCOCH3 H OH The difference is also qualitative, because there are structural
N-acetyl-galactosamine galactose differences in the branching of the oligosaccharide chains.
(A antigen) (B antigen) The structural difference explains why A-subgroup individu-
Figure 6–1 Terminal carbohydrates that define the A and B antigens.
als often make anti-A1. The reagent Dolichos biflorus lectin
The terminal galactose residues differ only in that the A antigen has a distinguishes A1 from A2 and other A subgroups.
substituted amino-acetyl group on carbon number 2. Acquired B antigen results from the action of bacte-
rial deacetylase, an enzyme that can remove an acetyl
group from the A-terminal sugar, N-acetylgalactosamine.
acute rejection.10 However, successful transplantation across Galactosamine is similar to galactose, the B-specific ter-
ABO barriers is possible, particularly with blood group A2 minal residue, and anti-B reagents can cross-react with
to O, and combined with current immunosuppressive and the deacetylated structure.16 Acquired B usually occurs
pretreatment regimens (reviewed by Rydberg11). As observed in individuals suffering from colon or rectal carcinoma,
for platelets, the absence of A antigen on tissues of group A2 intestinal obstruction, or infections involving gram-nega-
indicates significant lack of A2-transferase activity in non- tive bacteria from the gut. It is transient and, because the
erythroid tissues.12,13 acquired B develops at the expense of A antigen, the phe-
The incidence of ABO blood groups differs in popula- notype is found only in group A patients. There is a report
tions (Table 6–1).14 Group B is found twice as frequently in of a transfusion fatality in a group A patient mistyped
African Americans and Asians as in white persons. Group A as AB because of acquired B who was transfused with
subgroups are more common than group B subgroups, and group AB blood.17
Fuc
Fuc
Fuc Fuc
A Gene
A Gal GlcNAc R A
Gal
β1→4
GlcNAc R
GlcNAc
β1→4
Gal GlcNAc β1→3
β1→3 Gal
GalNAc β1→4 GalNAc β1→3
α1→3
α1→3
α1→2
α1→2
Fuc
Gal = Galactose
Fuc Fuc = Fucose
GalNAc = N-acetylgalactosamine
B Gene GlcNAc = N-acetylglucosamine
B Gal GlcNAc R B
Gal
β1→4
GlcNAc R
GlcNAc
β1→4
Gal GlcNAc β1→3
β1→3 Gal
Gal α1→3 β1→4 β1→3
Gal α1→3
α1→2 α1→2
Fuc
Fuc
Figure 6–2 Synthesis of A, B, and H, and Lewis antigens. Oligosaccharide precursor core type 1 and type 2 structures differ only in the linkage
between the terminal galactose (Gal) and the N-acetylglucosamine (GlcNAc), shown underscored. Terminal carbohydrates that define the antigens
are shown in black.
ABO AND RELATED ANTIGENS AND ANTIBODIES
Table 6–1 ABO Blood Groups and Incidence A1 A2 B O(1)
156 266268
Incidence (%)
Catalytic 354 235 116
Domain COOH
Phenotype White African American Asian COOH
COOH 87
A1 34 19 27 COOH 176
A2 10 8 Rare
B 9 19 25
A1B 3 3 5
Golgi Membrane
A2B 1 1 Rare
O 44 49 43
Cytosol
NH2 NH2 NH2 NH2
Leu
A2 NH2 COOH
156
Asn
A3 NH2 COOH
291
O1 NH2 COOH
86 116
Leu Ala
II
cisAB NH2 COOH
72 156 268
H Antigen
Common
Secretor H+ Yes HH, Hh, SeSe, Sese
Nonsecretor H+ No HH, Hh, sese
H-Deficient
Bombay H- No hh, sese Anti-H
Para-Bombay H weak No (H), sese Anti-H
Para-Bombay H weak Yes (H), SeSe, Sese Anti-HI
BLOOD BANKING Hemolytic disease of the newborn caused by ABO antibod-
Table 6–3 Lewis Blood Group Phenotypes
ies is usually mild, for the following reasons: placental transfer
and Incidence
is limited to the fraction of IgG anti-A and anti-B found in
maternal serum, fetal ABO antigens are not fully developed,54 Incidence (%)
and ABO tissue antigens provide additional targets for the anti-
bodies. ABO-HDN is most often seen in non-group O infants Phenotype White African Asian
of group O mothers, because anti-A, anti-B, and anti-A,B of
Le(a–b +.) 72 55 72
group O mothers often has a significant IgG component. Le(a+b−) 22 23 22
Potent anti-H (along with anti-A and anti-B) found in Oh Le(a−b−) 6 22 6
(Bombay) or para-Bombay nonsecretors destroys transfused Le( a. +b.+) Rare Rare 3
RBCs of any ABO group, so these individuals must be trans-
fused only with blood of the Bombay phenotype.55 In con-
trast, anti-H in non-Bombay individuals is usually IgM and
clinically insignificant. Anti-IH is not uncommonly found in chains, which are found in secretions but not in RBCs (see
patient sera and is usually IgM; compatible blood is easily Fig. 6–2). The nature and substrate specificity of the enzymes
found among donors of identical ABO type.15,56 have been elucidated only recently. The α1,4-fucosyltrans-
ferase encoded by LE (FUT3) catalyzes the addition of a
Enzyme-Converted O Cells (ECO) fucose to carbon 4 of the subterminal N-acetylglucosamine
(GlcNAc) residue of the type 1 precursor chains, creating
Blood group O is considered the universal donor because it can the Lea structure (see Fig. 6–2). (Note that this transferase
be transfused to patients of all ABO types. Therefore, enzymes cannot act similarly on the type 2 chains found on RBCs,
that remove terminal carbohydrates from the nonreducing end because they already have Ga1 on carbon 4 of the subtermi-
of carbohydrate chains could be used to remove terminal A and nal GlcNAc, which blocks the acceptor site; this fact explains
B sugars to convert the blood supply to all universal group O why the Lewis antigens are not synthesized in RBCs.)
units. An enzyme from coffee beans, α-galactosidase, has been The Lea structure remains unchanged, resulting in the
the most successful at removing galactose to convert blood Le(a+b−) phenotype, unless the individual is a secretor (Se-
group B to group O. RBCs treated in this manner have nor- FUT2). In the presence of the α1,2-fucosyltransferase encoded
mal survival when transfused to group B, A, or O recipients.57 by the secretor locus, the Lea structure is converted to Leb by
Removal of N-acetylgalactosamine to convert group A to the addition of a fucose residue to carbon 2 of the terminal
group O has been much more problematic, owing to the inac- galactose residue on the same chain. Leb antigen is synthe-
cessibility of the carbohydrates on internal branching chains, sized at the expense of Lea antigen, resulting in the Le (a−b+)
especially those found on A1 cells. The procedures required to phenotype. This finding explains early observations that indi-
convert B to O, which include exposure to low pH followed viduals with RBCs that typed as Le(b+) were secretors of ABH
II by numerous washings, make them impractical for general use. substance, those with Le(a+) RBCs were nonsecretors, and
This is an active area of research and alternative enzymes and individuals with Le(a−b−) RBCs could be either secretors or
improved methodologies are under investigation.58,59 nonsecretors of ABH. Ninety percent of white persons inherit
74
normal LE (FUT3), and 80% carry a functional Se (FUT2),
THE LEWIS SYSTEM (ISBT SYSTEM 007) accounting for the prevalence of the Le(a−b+) phenotype in
the white population (see Table 6–3).5,50
History Le(a−b−) arises from homozygous defects in LE (FUT3),
regardless of the Se (FUT2). Le(a+b+) individuals have weak
The Lewis system, first reported in 1946 by Mourant,60 was expression of both Lea and Leb and are sometimes called
named after the first patient to make the antibody. What was partial secretors. They have significantly reduced α1,2-fucos-
thought at the time to be the antithetical antigen was found yltransferase activity, and this uncommon phenotype is
in 1948, and the designations Lea and Leb were applied at a principally found in Taiwanese.
later time.19 We now know that these antigens are not anti-
thetical, because they are not products of alternative forms of
a single gene. Rather, they result from the sequential action Genes
of two fucosyltransferases encoded at independent loci (LE- The gene responsible for Lewis carbohydrates, LE (FUT3),
FUT3 and Se-FUT2). located on chromosome 19 (19p13.3), encodes a 361-amino
acid, type II membrane-bound enzyme.63 It is one of the
Antigens series of genes located on chromosome 19 that encode fucos-
Lewis antigens, unlike antigens of all other blood group sys- yltransferases. Chromosome 19 is also the location for both
tems except Chido-Rodgers, are not intrinsic to the RBC but the secretor gene Se (FUT2)—which not only determines
are synthesized in the intestinal epithelial cells. Lewis anti- ABH secretor status but also interacts with LE (FUT3) to
gens circulate in plasma while bound to glycosphingolipids synthesize Leb antigens—and the H gene (FUT1)—which is
and are passively adsorbed onto RBCs.61,62 responsible for H antigen on RBCs.
There are four Lewis phenotypes—Le(a+b−), Le(a−b+), Le(a−b+) individuals have a least one functional LE (FUT3)
Le(a−b−), and the rare Le(a+b+). African Americans have a and Se (FUT2), and are secretors of ABH antigens. Le(a+b−)
higher incidence of the Le(a−b−) phenotype, and Le(a+b+) individuals have at least one functional LE (FUT3), but this
is very rare in white persons and African Americans but not phenotype identifies the 20% of individuals with defective Se
uncommon in Asians (Table 6–3). (FUT2) (see discussion of the ABO system). Le(a+b+), found
Lea and Leb are synthesized in a stepwise manner by two in relatively high prevalence in Taiwan, results from inheritance
transferase enzymes that add fucose residues only to type 1 of Se(w), which encodes an amino acid change in the catalytic
ABO AND RELATED ANTIGENS AND ANTIBODIES
domain resulting in reduced activity of the enzyme encoded hemolysis, or cold agglutinin disease, due to anti-I. They used
by Se (FUT2). Le(a−b−) individuals have point mutations in the symbol I to emphasize the high degree of Individuality
the LE (FUT3) gene (le/le). Many different defective alleles of RBCs failing to react with the patient’s serum at room
(le1-le6) have been reported, and mutations at nucleotides 508, temperature. Because the patient’s serum was only weakly
1067, and 202 severely reduce or inactivate the fucosyltrans- reactive in vitro with bovine RBCs, she was transfused with
ferase.39 A mutation at nucleotide 59 results in an amino acid a small volume of bovine RBCs, but an anaphylactic-type
change in the transmembrane domain, which does not affect reaction discouraged Wiener and colleagues70 from further
the enzyme activity but may affect the Golgi localization. This attempts to transfuse bovine RBCs.
mutation is responsible for the paradoxical Le(a−b−) RBC
phenotype in people with Lewis antigens in their saliva.64
Antigens
I and i antigens on RBCs are subterminal portions of the
Expression
same carbohydrate chains that carry ABH antigens.18 I and
Lewis antigens are not expressed on cord RBCs. Lewis anti- i are not allelic; rather, they differ in their branching struc-
gen levels on RBCs are often diminished during pregnancy, ture. The i antigen, found predominantly on fetal and infant
possibly due to pregnancy-associated changes in plasma RBCs, is characterized by disaccharide units (Gal-GlcNAc)
lipoproteins.65 Lewis antigens are also found on lymphocytes linked in a straight chain.71 During the first 2 years of life,
and platelets (secondarily adsorbed from the plasma); on many of these linear chains are modified into branched
other tissues, including pancreas, stomach, intestine, skeletal chains,41 resulting in the appearance of I antigens found on
muscle, renal cortex, and adrenal glands; and in soluble form adult RBCs. I specificity develops at the expense of i antigens
in saliva as glycoproteins. Lewis antigens may be of some con- when the branched structures appear.
sequence in renal allografts. Graft survival has been reported
to be reduced in patients who lack Lewis antigens, but this Gene
issue is controversial.18 The Leb antigen may be a receptor for
Helicobacter pylori, a feature that could explain the associa- The gene responsible for I antigen synthesis on RBCs and
tion of gastric ulcers and secretor status.66,67 in tissues, called IGnT (GCNT2), consists of three exons
and is located on chromosome 6p24.72 The gene encodes
β1,6-N-acetylglucosaminyltransferase, a type II membrane
Antibodies
protein similar in structure to other glycosyltransferases.73
Lewis antibodies are primarily IgM and are usually not clini- This enzyme is responsible for the branching synthesis of I
cally significant. Lewis antibodies often complicate antibody antigen and so is probably developmentally regulated. The
identification when multiple antibodies are present, but they gene has three forms of exon 1 that are differentially spliced
are easy to inhibit with saliva (made isotonic) from secre- to give three transcripts—IGnTA, IGnTB, or IGnTC—each
tors or with commercially prepared Lewis substance. Lewis responsible for synthesis of I antigen in various tissues 6
antibodies occur in the sera of Le(a−b−) persons and may be (Fig. 6–5).72,74,75 For example, I antigen on RBCs is encoded
naturally occurring.15 by IGnTC, and expression of I antigen in lens epithelium is 75
Rare hemolytic transfusion reactions, due to transfu- encoded by IGnTB.
sion of Le(a+) RBCs to recipients with anti-Lea, have been The I-negative phenotype (adult i) in Taiwanese and in
reported.68 Because Le(a−) RBCs are found in almost 80% Japanese is associated with three different mutant alleles; all
of donors, it is easy to select Le(a−) RBCs for transfusion have a Gly348Glu change encoded in exon 3, with or without
to recipients with anti-Lea when the antibody is reactive at Arg383His (exon 3) or Gly334Arg (exon 2), or a large gene
37°C.69 Because Lewis antigens are not intrinsic to the RBC deletion encompassing exons 1B, 1C, and exons 2 and 3.72 In
membrane and are present in plasma, selection of Le(b−) whites, an Ala169Thr change, with or without Arg228Gln,
RBCs is unnecessary for recipients with anti-Leb, because encoded by IGnTC exon 1C causes adult i. In Japanese, the
Lewis antigens in donor plasma readily neutralize Lewis adult i phenotype has been associated with congenital cat-
antibodies in transfusion recipients.61 aracts,76,77 but this has not been observed in whites.78 The
The intimate relationship between Lewis and ABH anti- different genetic backgrounds explains these observations,
gens is revealed by antibodies such as anti-LebH, which reacts because the white mutation in exon 1C affects RBC expres-
best with O Le(b+) cells. Crossmatching ABO-identical sion of the I antigen, but normal I antigen is expressed in lens
RBCs (rather than O) usually provides compatible blood. epithelium. In contrast, mutations in exon 3 of IGnT found
Anti-LebA reacts best with A Le(b+) cells. These antibodies in Taiwanese and Japanese persons result in lack of transferase
are IgM and not clinically significant.
Anti-Lea and anti-Leb are not known to cause HDN,
because they are usually IgM, and because fetal RBCs type
as Le(a−b−).15
Pk Galα(1-4)Galβ(1-4)Glc-CER Galβ(1-4),GlcNAcβ(1-3)Galβ(1-4)Glu-CER
CD77/Gb3
P1 gene
P gene [α(1-4)galactosyltransferase]
[β(1-3)N-acetylgalacto-
sylaminyltransferase]
P P1
Gb4
GalNAcβ(1-3)Galα(1-4)Galβ(1-4)Glc-CER Galα(1-4)Galβ(1-4)GlcNAcβ(1-3)Galβ(1-4)Glu-CER
LKE
NeuAc α(2-3)Galβ(1-3) GalNAcβ(1-3)Galα(1-4)Galβ(1-4)Glc-CER
Incidence (%)
P1 79 94 P1, P, Pk*
P2 21 6 P, Pk*
P1k Very rare Very rare P1, Pk
P2k Very rare Very rare Pk 6
p Very rare Very rare None
* k
P is difficult to detect on these cells because it is converted to P. 77
This chapter summarizes four clinically significant blood from the maternal antibody reported by Levine and Stetson,
group systems that are defined by protein polymorphisms. the antigen responsible for HDN was named Rh. Later it was
The proteins that carry these blood group antigens were realized that the rabbit antiserum was not recognizing the
difficult to isolate because they are integral membrane pro- same antigen but was detecting an antigen found in greater
teins present as minor components of the total red blood amounts on Rh-positive than on Rh-negative RBCs.6 This
cell (RBC) protein. Biochemical techniques developed in antigen was named LW for Landsteiner and Wiener,1 and the
the 1970s and 1980s enabled their purification and partial original human specificity became known as anti-D.
amino acid sequencing. In the 1990s, the protein sequence As early as 1941, it was obvious that Rh was not a simple
data were used to construct nucleic acid probes for ampli- single antigen system. Fisher named the C and c antigens
fication and screening of bone marrow cDNA libraries to (A and B had been used for ABO) on the basis of the reactiv-
isolate the genes. In the past decade, there has been a con- ity of two antibodies that recognized antithetical antigens,
siderable increase in the amount of information about these and used the next letters of the alphabet, D and E, to define
blood group antigens; principally, the development and use antigens recognized by two additional antibodies.1 Anti-e,
of the polymerase chain reaction (PCR) has resulted in the which recognized the e antigen, was identified in 1945.7
rapid elucidation of the molecular basis for the antigens and
phenotypes. In addition, the structure and function of these
Nomenclature
membrane proteins is an active area of investigation.
II The Rh system has long been acknowledged as one of the
most complex blood group systems because of its large
80 Rh BLOOD GROUP SYSTEM number of antigens (45) and the heterogeneity of its anti-
bodies. The introduction of two different Rh nomenclatures
History reflected the differences in opinion concerning the num-
ber of genes that encoded these antigens. The Fisher-Race
The Rh system is second only to the ABO system in impor- nomenclature was based on the premise that three closely
tance in transfusion medicine. Rh antigens, especially D, are linked genes — C/c, E/e, and D — were responsible, whereas
highly immunogenic and can cause hemolytic disease of the the Wiener nomenclature (Rh-Hr) was based on the belief
newborn (HDN) and severe transfusion reactions. HDN was that a single gene encoded one agglutinogen that carried
first described by a French midwife in 1609 in a set of twins, several blood group factors.
of whom one was hydropic and stillborn, and the other was Even though neither theory was correct (there are two
jaundiced and died of kernicterus.1,2 In 1939, Levine and genes, RHD and RHCE, correctly proposed by Tippett8), the
Stetson3 described a woman who delivered a stillborn fetus Fisher-Race designation (CDE) for haplotypes is often pre-
and suffered a severe hemolytic reaction when transfused ferred for written communication, and a modified version of
with blood from her husband. Her serum agglutinated the Wiener’s nomenclature (the original form is nearly obsolete)
RBCs of her husband and 80 of 104 ABO-compatible donors. is preferred for spoken communication (Table 7–1). A capital
Subsequently, in 1941, Levine and colleagues4 correctly con- “R” indicates that D is present, and a lowercase “r” (or “little
cluded that the mother had been immunized by the fetus, r”) indicates that it is not. The C or c and E or e Rh anti-
which carried an antigen inherited from the father, and sug- gens carried with D are represented by subscripts: 1 for Ce
gested that the cause of the erythroblastosis fetalis was mater- (R1), 2 for cE (R2), 0 for ce (R0), and Z for CE (Rz). The CcEe
nal antibody in the fetal circulation. Attempts to immunize antigens present without D (r) are represented by superscript
rabbits against this new antigen were not successful.1 Levine symbols: “prime” for Ce (r’), “double-prime” for cE (r”) and
and colleagues did not name the antigen.5 “y” for CE (ry) (see Table 7–1). The “R” versus “r” terminol-
Meanwhile, Landsteiner and Wiener, in an effort to dis- ogy allows one to convey the common Rh antigens present
cover additional blood groups, injected rabbits and guinea on one chromosome in a single term (a phenotype). Dashes
pigs with rhesus monkey RBCs. The antiserum agglutinated are used to represent missing antigens of the rare deletion
not only rhesus cells but also the RBCs of 85% of a group of (or CE-depleted) phenotypes; for example, D– – (referred to
white subjects from New York, whom the researchers called as D dash, dash) lacks C/c and E/e antigens.
Rh positive; the remaining 15% were Rh negative.1 Because In 1962, Rosenfield and associates9 introduced numerical
the anti-Rhesus appeared to have reactivity indistinguishable designations for the Rh antigens to more accurately represent
RH, KELL, DUFFY, AND KIDD ANTIGENS AND ANTIBODIES
Table 7–1 Nomenclature and Prevalence for Rh Haplotypes
Prevalence (%)
Haplotype Based on Shorthand
Antigens Present for Haplotype White African American Asian
DCe R1 42 17 70
DcE R2 14 11 21
Dce R0 4 44 3
DCE RZ <0.01 <0.01 1
ce r 37 26 3
Ce r’ 2 2 2
cE r” 1 <0.01 <0.01
CE ry <0.01 <0.01 <0.01
the serologic data, to be free of genetic interpretation, and to or lack of expression of, RhAG results in a lack of Rh antigen
be more compatible for computer use (Table 7–2). However, expression (Rh-null) or a marked reduction of Rh antigen
this numerical nomenclature, with a few exceptions (Rh17, expression (Rh-mod).17 RhAG has one N-glycan chain that
Rh32, Rh33), is not widely used in the clinical laboratory. also carries ABO and Ii specificities (Fig. 7–1).
G 16W
antigen 16C
Figure 7–2 Top panel, Diagram of the RHD and RHCE genes, indicating the changes that resulted in the common RhCE polymorphisms. The
shared exon 2 of RHD and RHCE (shown as a black box) explains the expression of G antigen on RhCe and RhD proteins. Bottom panel, Examples
of some RHCE and RHD rearrangements.
II
Table 7–3 Molecular Basis of Some Rh Antigens, Partial D, and Unusual Phenotypes
84
Molecular Basis Gene Phenotype/Antigen/Genotype
Single point mutations RHD Partial D: DMH, DVII, D+G−, DFW, DHR, DVa, DHMi, DNU, DII, DNB, DHO
Weak D (previously called Du)
RHCE CX, CW, Rh−26, E type I, IV, V+,VS+
Multiple mutations RHD Partial D: DIIIa, DIVa, DVa, DFR type I
(gene conversions) RHCE E type III, IV, V+VS+
Rearranged gene(s) RHD RHD-CE-D Partial D: DIIIb, DIIIc, DIVb, DVa, DVI, DFR type II, DBT r”G, (Ce)Ce, (C)ces VS+V−
=N
RHCE RHCE-D-CE DHar, rG, R
RHD-CE E type II
RHD: RHCE RHD-CE: RHCE-D DCW−
RHD: RHCE-D-CE D− −, D••, Dc−
RHD: RHD-CE D••
RHCE-D: RHD-CE D••
Table 7–4 Composition (IgM and IgG Clones) and Reactivity of FDA-Licensed Anti-D Reagents with
Some Rh Variant RBCs That Can Result in D Typing Discrepancies
Reagent IgM Monoclonal IgG DVI DBT DHAR (Whites) Crawford (Blacks)
*
Result following slash denotes anti-D test result by the IAT, as permitted by the manufacturer.
RH, KELL, DUFFY, AND KIDD ANTIGENS AND ANTIBODIES
Lastly, an Rhce protein with an R154T mutation, designated Other modifications of RHCE, which are uncommon, are
ceRT, demonstrates weak reactivity with some anti-D monoclonal the hybrids rG, RN, and several E/e variants (see Fig. 7–2). RN
reagents, and the reactivity is enhanced at lower temperatures. RBCs are found in people of African origin and type as e-weak
Interestingly, this variant does not carry any D-specific amino (or negative) with polyclonal reagents, but are indistinguish-
acid but mimics a D-epitope (epD6) structure.47 able from “normal” e-positive RBCs with some monoclonal
anti-e. The E variants—EI, EII, and EIII—result either from
C/c and E/e Antigens
a point mutation (EI) or from gene conversion events that
There are four major allelic forms of RHCE: ce, Ce, cE, and lead to replacement of several extracellular RhcE amino acids
CE.14 C and c differ by four amino acids: Cys16Trp (cysteine with RhD residues (EII and EIII) and loss of some E epitope
at residue 16 replaced by tryptophan) encoded by exon 1, expression.52 Category EIV RBCs, which have an amino acid
and Ile60Leu, Ser68Asn, and Ser103Pro encoded by exon 2. substitution in an intracellular domain, do not lack E epit-
Only residue 103 is predicted to be extracellular; it is located opes but have reduced E expression.53 The very rare RH:-26
on the second loop of RhCE (see Fig. 7–1). The amino acids results from a Gly96Ser transmembrane amino acid change
encoded by exon 2 of RHC are identical to those encoded that abolishes Rh26 and weakens c expression.54
by exon 2 of RHD. At the genomic level, RHCe appears to
VARIATION IN e EXPRESSION AND SICKLE CELL PATIENTS
have arisen from transfer of exon 2 from RHD into RHce (see
Fig. 7–2). The shared exon 2 explains the expression of the G Variation in expression of the e antigen is not uncommon
antigen on both RhD and RhC proteins. and can result from several different mutations. Deletion of
E and e differ by one amino acid, Pro226Ala. This poly- the codon for Arg229, which is close to the Ala226 residue
morphism, predicted to reside on the fourth extracellular loop found in normal e expression,55 and substitution of a Cys
of the protein (see Fig. 7–1), is encoded by exon 5. A single residue for Trp at position 16 in the Rhce protein56 weaken
point mutation in RHce resulted in RHcE (see Fig. 7–2).14,48 expression of the antigen. Individuals of African ancestry
CW and CX antigens result from single amino acid changes, often have RHce genes that encode variant e antigens. The
encoded by exon 1, which are predicted to be located on the RBCs type as e positive, but they often make alloantibodies
first extracellular loop of the RhCE protein.49 with e-like specificities. The antibodies, designated anti-hrS,
The antigens V and VS are expressed on RBCs of more than -hrB, -RH18, and -RH34, are difficult to identify serologically,
30% of black persons. They are the result of a Leu245Val sub- are clinically significant, and have caused transfusion fatali-
stitution located in the predicted eighth transmembrane seg- ties.57 The prevalence of e variants in this population, together
ment of Rhce.50 The close location of the e antigen, Ala226 on with the incidence of sickle cell disease requiring transfusion
the fourth extracellular loop, suggests that Leu245Val causes support often provided by white donors with conventional
a local conformation change responsible for the weakened RHce, make the occurrence of alloanti-e in these patients not
expression of e antigen in many black persons who are V and uncommon. Some of the RH genetic backgrounds have now
VS positive. The V−VS+ phenotype results from a Gly336Cys been defined58 and include the RHCE haplotypes shown in
change on the 245Val background,51 and these alleles are Figure 7–3. All encode the Trp16Cys difference in exon 1, and 7
referred to as ceS (Fig. 7–3). Loss of VS expression (the V+VS− have additional changes, primarily localized to exon 5. The ceS
phenotype) is associated with additional amino acid changes allele is associated with RBCs that are hrB−, whereas individu- 85
and is characteristic of the ceAR haplotype (see Fig. 7–3). als homozygous for ceAR, ceMO, ceEK, and ceBI alleles lack
ceS
(hrB–)
D/Ce/D (D-negative, C-positive) W16C L245V G336C
ceMO
DIIIa
(hrS–)
N152T T201R F223V W16C V223F
DIII ceEK
Type 5 (hrS–)
L62F T201R F223V W16C M238V M267K
A137V R263G
N152T
ceBI
DIVa (hrS–)
W16C M238V A273V L378V
L62F N152T D350H
DAR I306V
ceAR
T201R F223V I342T (hrS–)
W16C M238V M267K
DOL
L245V R263G
M170T F223V
BLOOD BANKING the high-incidence hrS antigen (see Fig. 7–3). Importantly, Anti-c, clinically the most important Rh antibody after
because of the multiple molecular backgrounds responsible anti-D, may cause severe HDN. Anti-C, anti-E, and anti-e do
for the hrB− and hrS− phenotypes, some of which are not yet not often cause HDN, and when they do, it is usually mild.6,59
elucidated, the antibodies produced are not all serologically Autoantibodies to high-incidence Rh antigens often occur
compatible. This explains why it is difficult to find compat- in the sera of patients with warm autoimmune hemolytic
ible blood for patients with these antibodies, and often only anemia and in some cases of drug-induced autoimmune
rare deleted D– – RBCs appear compatible. As an additional hemolytic anemia. These autoantibodies are nonreactive
complication, these variant RHce can often be inherited with with Rh-null cells.60
an altered RHD (e.g., DAR, DAU, or DIIIA), so they can also
make anti-D (see Fig. 7–3).
Anti-D reagents
Rh Genotyping A large number of IgM, direct-agglutinating, anti-D mono-
clonals have been generated by immortalizing human B lym-
RH genotyping is a useful means to determine the Rh pheno- phocytes in vitro with Epstein-Barr virus. D-typing reagents
type of patients who have been recently transfused or whose licensed for use in the United States are a blend of monoclo-
RBCs are coated with IgG. RH genotyping in the prenatal nal IgM reactive at room temperature along with monoclonal
setting can be used to determine paternal RHD zygosity and or polyclonal IgG reactive by the IAT for the determination
to predict fetal D status to prevent invasive and expensive of weak D. Four different FDA-licensed reagents are avail-
monitoring for the possibility of HDN. The ethnic back- able for tube testing and one for gel (see Table 7–4). All but
ground of the parents is important to the design of the assay, two contain different IgM clones. The reactivity of each with
because the different molecular events responsible for D- variant D antigens may result in D typing discrepancies.
negative phenotypes must be considered. Testing of samples Importantly, the IgM anti-D component of these reagents
from the parents limits the possibility of misinterpretation. has been selected to not react with DVI RBCs (see Table 7–4).
RH genotyping can aid resolution of D typing discrep- DVI is the most common partial D found in white popula-
ancies. These often are the result of differences in manufac- tions, and these individuals often make anti-D when exposed
turers’ reagents, but in the donor setting they can be FDA to conventional D. Most agree that they should be classified
reportable. Genotyping can determine a specific weak D as Rh-negative for transfusion or Rh immune globulin. The
type, partial D category, or the presence of Del. IgG component in these reagents reacts with DVI RBCs in the
Genotyping to detect the inheritance of altered or variant IAT phase (see Table 7–4). DVI RBCs can stimulate produc-
e alleles, which are often linked to variant RHD, aids reso- tion of anti-D in an Rh-negative recipient and must be typed
lution of antibody specificities and is helpful to determine as Rh-positive as donors. The composition of current FDA-
alloantibody versus autoantibody specificity. This can be licensed reagents has prompted the movement away from
of significance, especially in sensitized sickle cell patients. weak D testing in the hospital and prenatal setting to classify
II Molecular genotyping can aid in the selection of compat- individuals with DVI RBCs, or some of the other partial D
ible blood for transfusion and ultimately improve long-term phenotypes (see Table 7–4), as Rh negative.
86 transfusion support. Currently, these investigations can Monoclonal anti-D has not yet been tested clinically for
sometimes be cumbersome, often requiring complete gene its ability to prevent immunization after pregnancy but has
sequencing. The development of automated, high-through- been shown to suppress D immunization in D-negative male
put platforms that sample many regions of both RHD and volunteers.61
RHCE, along with detailed algorithms for accurate interpre-
tation, are needed. Expression
Northern blot analysis indicated that Rh and RhAG mes-
Antibodies senger RNA (mRNA) are restricted to cells of erythroid and
myeloid lineage, but reverse transcriptase-PCR (RT-PCR)
Most Rh antibodies are IgG, subclasses IgG1 and IgG3 (IgG2 found Rh mRNA splicing isoforms in B and T lymphocytes
and IgG4 have also been detected), and some sera have an and monocytes.62 The significance of this observation is not
IgM component.59 Rh antibodies do not activate comple- yet known. During erythropoiesis, RhAG appears early (on
ment, although two rare exceptions have been reported. The CD34+ progenitors), but the Rh proteins appear later—RhCE
lack of complement activation by Rh antibodies is thought first, followed by RhD.63
to be due to the distance between antigens, but is probably
due to a lack of mobility.6 Reactivity of Rh antibodies is Evolution
enhanced by enzyme treatment of the test RBCs, and most The RHD and RHCE genes arose from an early duplication
react optimally at 37°C. of the erythrocyte RHAG gene. RH and RHAG have been
Anti-D can cause severe transfusion reactions and severe investigated in nonhuman primates64,65 and rodents,66–68 and
HDN. Approximately 80% to 85% of D-negative persons most, with the exception of gorillas and chimpanzees, have
make anti-D after exposure to D-positive RBCs. The lack of a RHAG and only one RH gene. The RH gene duplicated
response in 15% may be due to antigen dose, recipient HLA- in some common ancestor of gorillas, chimpanzees, and
DR alleles, and other as yet unknown genetic factors. Anti- humans, leading to RHCE and RHD. Chimpanzees and some
D was the most common Rh antibody, but its incidence has gorillas have three RH genes, indicating that a third duplica-
greatly diminished with the prophylactic use of Rh immune tion has taken place in these species.
globulin for prevention of HDN. ABO incompatibility
between the mother and the fetus has a partial protective Function
effect against immunization to D; this finding suggested the The predicted membrane structures of Rh and RhAG sug-
rationale for development of Rh immune globulin.59 gest that they are transport proteins, and the analysis of their
RH, KELL, DUFFY, AND KIDD ANTIGENS AND ANTIBODIES
amino acid sequence reveals distant similarity to ammo- cyte cytoskeleton and the lipid bilayer are understood to be
nium transporters in bacteria, fungi, and plants.69 Evidence through glycophorin C and band 3, an additional attach-
for ammonia transport by RhAG comes from yeast comple- ment site mediated by the Rh complex explains the Rh-null
mentation experiments,70,71 expression studies in Xenopus defect. Additional studies are needed to determine the pro-
oocytes,72 and direct evidence in RBCs.73 Rh and RhAG tein–protein associations and the dynamics of the assembly
homologues have been found in many organisms, includ- of the Rh-membrane complex.
ing the sponge (Geodia), the slime mold (Dictyostelium), the
fruit fly (Drosophila), and the frog (Xenopus),74 indicating
that they are conserved throughout evolution. THE KELL AND Kx SYSTEM
Nonerythroid homologues, designated RhCG and RhBG,
are found in the kidney,75 liver,76,77 testis, brain, gastrointes-
History
tinal tract,78 and skin79,80 and are localized to regions where
ammonium production and elimination are critical in mam- The Kell blood group system was discovered in 1946, just a
malian tissues, strongly suggesting a role for these proteins few weeks after the introduction of the antiglobulin test. The
in ammonia/ammonium homeostasis. When expressed in RBCs from a newborn baby who was thought to be suffering
Xenopus oocytes, RhBG and RhCG also mediate transport of from HDN gave a positive reaction in the direct antiglobulin
ammonia.81 In the kidney ammonium ions act as expendable test.84 The serum of the mother reacted with RBCs from her
cations that facilitate excretion of acids, and renal ammo- husband, her older child, and 9% of random donors. The
nium metabolism and transport are critical for acid-base system was named from Kelleher, the mother’s surname, and
balance. In the collecting segment and collecting duct, where the antigen is referred to as K (synonyms: Kell, K1). Three
large amounts of ammonia are excreted, RhBG and RhCG years later, the more common antigen, k (synonyms: Cellano,
are found on the basolateral and apical membranes, respec- K2), which has a high incidence in all populations, was iden-
tively, of the intercalated cells.76 These localization studies tified through the typing of large numbers of RBC samples
suggest that RhBG and RhCG are ideally situated to mediate with an antibody that had also caused a mild case of HDN.85
transepithelial movement of ammonium from the intersti- The Kell system remained a simple two-antigen system
tium to the lumen of the collecting duct. In support, mouse until 1957, when the antithetical Kpa and Kpb antigens were
collecting duct (mIMCD-3) cells, which show polarized reported, as was the K0 (Kell-null) phenotype.6 Subsequently,
expression of these proteins, demonstrate transporter- the number of Kell antigens has grown to 25, making Kell
mediated movement of ammonia.82 one of the most polymorphic blood group systems known.
The function of RhCE and RhD has not been determined.
Co-expression of RhCE/RhAG did not influence the rate or
Proteins
total substrate accumulated in oocytes compared to that seen
with expression of RhAG alone.83 Although further studies The Kell protein is a type II glycoprotein with an approxi-
are necessary to determine if RhCE/D are involved in mem- mate Mr of 93,000. It has a 665-amino acid carboxyl terminal 7
brane transport, RhCE/D may have lost transport function extracellular domain, a single 20-amino acid transmem-
and may have a structural role in the RBC membrane. brane domain, and a 47-amino acid N-terminal cytoplasmic 87
domain.86 The protein has five N-glycosylation sites and 15
Summary extracellular cysteine residues that cause folding through the
formation of multiple intrachain disulfide bonds (Fig. 7–4).
The molecular basis of many of the Rh antigens has now This explains why Kell blood group antigens are inactivated
been elucidated. The revelation that RhD and RhCE pro- when RBCs are treated with reducing agents, such as dithio-
teins differ by 35 amino acids explains why D antigen is so threitol and aminoethylisothiouronium bromide, which
immunogenic. In addition, exchanges between RHD and disrupt disulfide bonds.87 All Kell system antigens are car-
RHCE, mainly by gene conversion, have generated many ried on this glycoprotein, which is present at 3500 to 17,000
Rh polymorphisms. The proximity of the two genes on the copies per RBC.88,89 All but two (Jsa and Jsb) of the Kell
same chromosome probably affords greater opportunity for antigens are localized in the N-terminal half of the protein
exchange. This finding finally explains the myriad of antigens before residue 550, strongly suggesting that the C-terminal
observed in the Rh blood group system and gives interesting domain does not tolerate change and is functionally impor-
insight into the evolutionary history of duplicated genes and tant. Indeed, the Kell glycoprotein is a zinc endopeptidase,
the interactions that can take place between them. The com- and the C terminus contains a zinc-binding domain that is
plexity hampers molecular genotyping, and additional Rh the catalytic site.90,91
variants are still being discovered. The challenge is to develop Kx is a 444-amino acid, 37-kD protein that is linked by a
automated platforms that sample several regions of the genes disulfide bond to the Kell protein.17 Kx is predicted to span
for unequivocal interpretation. the membrane 10 times (see Fig. 7–4), is not glycosylated,
The discovery that members of the Rh family of proteins, and may be a membrane transport protein. RBCs lacking
RhAG, RhBG, and RhCG, are involved in ammonia/ammo- Kx have the McLeod phenotype, which is characterized by
nium transport and are ideally positioned in key tissues a marked reduction of Kell antigens, acanthocytosis, and
essential for ammonium elimination is a significant finding reduced in vivo RBC survival.
because it was long assumed that the high membrane per-
meability of ammonia would obviate the need for specific
Genes
transport pathways in mammalian cells.
RBC membrane protein–cytoskeleton and protein– The KEL gene has been localized on chromosome 7q33. It
protein interactions are an active area of investigation. consists of 19 exons spanning approximately 21.5 kb.92,93 The
Although the major attachment sites between the erythro- Kell antigens result from nucleotide mutations that cause
BLOOD BANKING Catalytic domain
very weak expression of Kell antigens. It is now evident that
he lacked Kx, which is important for expression of Kell, and
597
that lack of Kx is the basis for the McLeod syndrome. Males
Jsb/Jsa with the McLeod syndrome have muscular and neurologic
Leu597Pro 281 defects, including skeletal muscle wasting, elevated serum
Kpb/Kpa
Arg281Trp creatine phosphokinase, psychopathology, seizures with basal
k/K ganglia degeneration, and cardiomyopathy.90,100 Most symp-
Thr193Met 193 toms develop after the fourth decade of life. The syndrome is
very rare. Approximately 60 males have been identified, and
all but 2 have been white; however, because of the plethora of
symptoms, the syndrome is probably underdiagnosed.
Fifteen different mutations in the XK gene were found
COOH
Extracellular S in a study of 17 families; these mutations involve major and
minor deletions, point mutations, and splice site or frame-
Lipid shift mutations that result in the absence or truncation of
Bilayer Kx protein.17,100 At one time, chronic granulomatous disease
Intracellular
(CGD) was thought to be related to the McLeod syndrome,
but the gene controlling CGD is near the XK gene on the
NH2
X chromosome, and the small minority of patients with
Kx
NH2
Kell CGD who have the McLeod phenotype have X-chromosome
COOH deletions encompassing both genes.6,101
Figure 7–4 Kell and Kx proteins. Kell is a single-pass protein, but
Kx is predicted to span the red blood cell membrane 10 times. Kell
and Kx are linked by a disulfide bond, shown as -S—. The amino acids Antigens
that are responsible for the more common Kell antigens are shown.
The N-glycosylation sites are shown as Y. The hollow Y represents the The Kell system consists of five sets of high-incidence and
N-glycosylation site that is not present on the K (K1) protein. low-incidence antigens, as follows (the names of the high-
incidence antigens appear in boldface):
●
K and k
single amino acid substitutions in the protein (Table 7–5). ●
Kpa, Kpb, and Kpc
The lack of Kell antigens, K0, is caused by several different ●
Jsa and Jsb
molecular defects, including nucleotide deletion, defective ●
K11 and K17
splicing, premature stop codons, and amino acid substitu- ●
K14 and K24
II tions.91,94,95
The XK gene is on the short arm of the X chromosome at In addition, 14 independently expressed antigens, 3 low-
88 Xp21.96 XK has three exons, and mutations cause the McLeod incidence (Ula, K23, VLAN), and 11 high-incidence (Ku,
syndrome, which, because the gene is X linked, affects males. Km, K12, K13, K16, K18, K19, K22, Tou, KALT, KTIM),
Carrier females, because of X-chromosome inactivation, have been identified. The null phenotype, K0, lacks Kell
have two populations of RBCs (one of the McLeod pheno- antigens, and Kell-mod phenotypes have a weak expression
type and one normal), and the proportion varies from 5% of Kell antigens.
McLeod:95% normal to 85% McLeod:15% normal.97,98 Kell antigens show population variations (Table 7–6). K
has an incidence of 9% in white persons but is much less
McLeod Syndrome common in people of other ethnic backgrounds.1,58 Kpa and
K17 are also mainly found in white persons. Jsa is almost
When testing medical students in 1961, Allen and coworkers99 exclusively found in African Americans, with an incidence of
found that one of the students, Mr. McLeod, had RBCs with 20%. Ula is found in Finnish and Japanese persons.1,58
The molecular basis of most of the Kell antigens has been
determined (see Table 7–5).91,102 The K methionine substitu-
tion disrupts a glycosylation consensus sequence so that K has
Table 7–5 Molecular Basis of Antigens in the one less N-glycan than k.103 Jsa and Jsb are located within a
Kell Blood Group System cluster of cysteine residues,104 a finding that explains why they
are more susceptible than other Kell antigens to treatment
Antigen Amino Acid Position
with reducing agents.105 No Kell haplotype has been found to
k(K2) Threonine 193 express more than one low-incidence Kell antigen, not because
K(K1) Methionine of structural constraints, but because multiple expression
Kpa(K3) Tryptophan 281 would require more than one mutation encoding a recognized
Kpb(K4) Arginine
Kpc(K21) Glutamine
Kell system antigen to occur in the same gene.106
Jsa(K6) Proline 597 Weaker expression of Kell antigens is found when RBCs
Jsb(K7) Leucine carry Kpa in cis.107–109 Weak expression of Kell antigens can
K11 Valine 302 be inherited or can be acquired and transient. Inherited
K17 Alanine weak expression occurs when the Kell-associated Kx protein
K14 Arginine 180
K24 Proline is absent (McLeod phenotype), when glycophorins C and D
Ula Glutamic acid→Valine 494 are absent (Leach phenotype), or when a portion of the extra-
cellular domain of glycophorin C and D, specifically exon 3, is
RH, KELL, DUFFY, AND KIDD ANTIGENS AND ANTIBODIES
Table 7–6 Kell Phenotypes and Prevalence Evolution
Prevalence (%) Primate RBCs express the k antigen but not the K antigen,
indicating that the K mutation appeared in human lineage.
Phenotype White African American Chimpanzees are Js(a+b−),112 and Jsa is also present on
RBCs from Old World monkeys,113 suggesting that Jsb also
K−k+ 91 98
K+k+ 8.8 2
arose after human speciation. Kell protein is not found on
K+k− 0.2 Rare immunoblots of RBCs from sheep, goat, cattle, rabbit, horse,
Kp(a+b−) Rare 0 mouse, donkey, cat, dog, or rat.114
Kp(a−b+) 97.7 100
Kp(a+b+) 2.3 Rare
Kp(a−b−c+) 0.32 Japanese 0 Function
Js(a+b−) 0 1
Js(a−b+) 100 80 The Kell glycoprotein is a member of the M13 or neprilysin
Js(a+b+) Rare 19 family of zinc endopeptidases that cleave a variety of physi-
K11 High incidence ologically active peptides. One member of this family, endo-
K17 Low incidence
K14 High incidence
thelin converting enzyme 1 (ECE-1), is a membrane-bound
K24 Low incidence metalloprotease that catalyzes the proteolytic activation of
Ula Low incidence big endothelin-1 (big ET-1).115 Like ECE-1, Kell protein can
(2.6% in Finns, proteolytically cleave endothelins, specifically big ET-3 to gen-
0.46% in Japanese) erate ET-3, which is a potent vasoconstrictor.116 ET-3 is also
involved in the development of the enteric nervous system
and in migration of neural crest-derived cells. The biologic
role of the endothelins is not yet completely elucidated, but
deleted (some Gerbich-negative phenotypes).109,110 Transient they act on two G protein–coupled receptors, ETA and ETB,
depression of Kell system antigens has been associated with the which are found on many cells. Kell-null individuals, who lack
presence of autoantibodies mimicking alloantibodies in auto- Kell protein, do not have any obvious defect, so they do not
immune hemolytic anemia and with microbial infections. Kell immediately give insight into the biologic function of the Kell
expression was reduced in two cases of idiopathic thrombocy- protein. This lack of defect in Kell-null people may be because
topenic purpura but returned to normal after remission.6 other enzymes probably also cleave ET-3,116 and determining
whether Kell-null individuals have abnormal levels of plasma
endothelins is an active area of investigation.
Antibodies
Kell antigens are highly immunogenic, and anti K is common.
However, because more than 90% of donors are K negative, it DUFFY (Fy) BLOOD GROUP SYSTEM 7
is not difficult to find compatible blood for patients with anti-
K. The other Kell system antibodies are less common but are History 89
also usually IgG, and they have caused transfusion reactions
and HDN or neonatal anemia. In HDN due to Kell antibodies, The Duffy (Fya) blood group antigen was first reported in
neither maternal antibody titers nor amniotic bilirubin lev- 1950 by Cutbush and associates,117 who described the reac-
els are good predictors of the severity of the disease. Reports tivity of an antibody found in a hemophiliac male who
demonstrate that Kell antigens are expressed very early during had received multiple transfusions. This blood group sys-
erythropoiesis63 and that Kell antibodies can cause suppression tem bears the patient’s surname, Duffy, the last two letters
of erythropoiesis in vitro.111 This finding suggests that the low of which provide the abbreviated nomenclature (Fy). Fyb
level of bilirubin observed, in the presence of neonatal anemia, was found 1 year later.118 In 1975, Fy was identified as the
is due to Kell antibodies that bind to erythroid progenitors and receptor for the malarial parasite Plasmodium vivax.119 This
exert effects before hemoglobinization. discovery explained the predominance of the Fy(a−b−)
Anti-Ku is the antibody made by immunized K0 indi- (Fy-null) phenotype, which confers resistance to malarial
viduals, and Ku represents the high-incidence or “total-Kell” invasion, in persons originating from West Africa.
antigen. McLeod males with CGD make anti-Kx+Km; this
antibody reacts strongly with K0 cells, weaker with RBCs of Proteins
common Kell phenotype, and not at all with McLeod pheno-
type RBCs. Anti-Km, made by McLeod persons who do not The Fy protein is a transmembrane glycoprotein of 35 to
have CGD, reacts with RBCs of common Kell phenotypes 43 kD consisting of a glycosylated amino terminal region,
but not with K0 or McLeod RBCs, suggesting that it detects which protrudes from the membrane and has seven trans-
one or more epitopes requiring both the Kell and Kx pro- membrane-spanning domains (Fig. 7–5).120,121 In 1993, it was
teins. One case has been reported of a McLeod male without realized that Fy was the erythrocyte chemokine receptor that
CGD who made an apparent anti-Kx without the presence could bind interleukin-8 and monocyte chemotactic peptide-
of anti-Km.101 1 (MCP-1).122 The cloning of the FY gene123 confirmed that it
belongs to the conserved family of chemokine receptors.
Expression
Gene
Kell system antigens appear to be erythroid specific and are
expressed very early during erythropoiesis.63 Kx mRNA is The FY gene is located on the long arm of chromosome
found in muscle, heart, brain, and hematopoietic tissue.17 1q22–q23124 and spans 1.5 kb. The gene, which has only two
BLOOD BANKING NH2
Table 7–7 Duffy Phenotypes and Prevalence
29 18 Prevalence (%)
31 Fya/Fyb
Gly42Asp Phenotype Caucasians Blacks Chinese Japanese Thai
Fy6 40 42 Fy3
Fy(a+b−) 17 9 90.8 81.5 69
Extracellular Fy(a−b+) 34 22 0.3 0.9 3
Fy(a+b+) 49 1 8.9 17.6 28
Lipid Fy(a−b−)* Rare 68 0 0 0
Bilayer FyX 1.4 0 0 0 0
*
Intracellular Fy(a−b−), incidence in Israeli Arabs 25%; Israeli Jews, 4%
Fyx
Arg89Cys
The Fy(a−b−) phenotype in white persons is very rare
COOH (Table 7–7).1 One propositus, an Australian (AZ) woman,
Figure 7–5 The predicted seven-transmembrane domain structure of appears to be homozygous for a 14-bp deletion in FYA, which
the Duffy protein. The amino acid change responsible for the Fya/Fyb introduces a stop codon in the protein138; a Cree Indian
polymorphism, the mutation responsible for Fyx glycosylation sites, and female (Ye), a white female (NE), and a Lebanese Jewish
the regions where Fy3 and Fy6 map are indicated.
male (AA) carry different Trp to stop codon mutations.139
Because these mutations would result in a truncated protein,
these people would not be expected to express endothelial or
exons, contains two ATG codons. The upstream exon con- erythroid Fy protein. All four people made anti-Fy3.
tains the major start site.125,126 The same two-exon organiza- There are 13,000 to 14,000 Fy antigen sites per RBC,140
tion is also found in genes for other chemokine receptors.127 and Fya, Fyb, and Fy6 antigens are sensitive to proteolytic
enzyme treatment of antigen-positive RBCs, although Fy3 is
Antigens resistant.
The Fya and Fyb antigens are encoded by two allelic forms
Antibodies
of the gene, designated FYA and FYB, and are responsible
for the Fy(a+b−), Fy(a−b+), and Fy(a+b+) phenotypes.126 Fy antigens are estimated to be 40 times less immunogenic
They differ by a single amino acid located on the extracel- than K antigens, and most Fy antibodies arise from stimu-
lular domain (see Fig. 7–5). lation by blood transfusion. They are mostly IgG, subclass
II Fyx, which is a weak expression of Fyb, is found in white IgG1, and only rarely are IgM. Anti-Fyb is less common than
persons and is due to a single mutation in the FYB gene.128,129 anti-Fya, and Fy antibodies are often found in sera with
90 The amino acid change, which is located in the first intracel- other antibodies. Anti-Fy3 is made by rare white Fy(a−b−)
lular cytoplasmic loop, is associated with a decrease in the individuals.6 Anti-Fy6 is a mouse monoclonal antibody.127
amount of protein in the membrane and results in dimin-
ished expression of Fyb, Fy3, and Fy6 antigens as well as
Expression
reduced binding of chemokine.127,130,131
Fy3, as determined with one monoclonal anti-Fy3, is Duffy mRNA is present in kidney, spleen, heart, lung, muscle,
located on the third extracellular loop,132 whereas Fy6 maps duodenum, pancreas, placenta, and brain.123,127 Cells respon-
to the amino terminal loop of the protein. The aspartic acid sible for Fy expression in these tissues are the endothelial cells
at amino acid 25 and glutamic acid at 26 are critical for anti- lining postcapillary venules,141–143 except in the brain, where
Fy6 binding.133 expression is localized to the Purkinje cell neurons.144,145 The
The Fy(a−b−) phenotype found in African Americans is same polypeptide is expressed in endothelial cells and RBCs,
caused by a mutation in the promoter region of FYB (T>C but in brain, a larger, 8.5-kb mRNA is present. The function
at position −46), which disrupts a binding site for the ery- of Fy on neurons is an area of active investigation.
throid transcription factor GATA-1 and results in the loss of In the fetus, Fy antigens can be detected at 6 to 7 weeks’
Fy expression on RBCs.134,135 The Fy protein has also been gestation and are well developed at birth.1 The expression of
found on endothelial cells. Because the erythroid promoter these antigens was found to occur late during erythropoiesis
controls expression only in erythroid cells, expression of Fy and RBC maturation.63
proteins on endothelium is normal in these Fy(a−b−). All
African persons with a mutated GATA sequence to date have
Evolution
been shown to carry FYB; therefore, Fyb is expressed on their
nonerythroid tissues. This finding explains why Fy(a−b−) RBCs of monkeys and apes react with anti-Fyb, and the con-
individuals make anti-Fya but not anti-Fyb.136 It also is rel- served GT repeat sequences in the 3 flanking region of the
evant in the selection of antigen-matched units for Fy(a−b−) gene both suggest that FYB was the ancestral gene.146,147 The
patients with sickle cell disease, because they would not be first human divergence occurred when a mutation in the
expected to make anti-Fyb. Rare Fy(a−b−) who do make erythroid promoter region of FYB resulted in the loss of
anti-Fyb should be investigated at the genetic level, because Fyb expression on RBCs. This mutation conferred resistance
the molecular basis for this observation has not yet been to malaria infection in regions where P. vivax was endemic
explained. Lastly, a FYA allele with a mutated GATA sequence and selected for the high proportion of Fy(a−b−) in popula-
has been found in Papua New Guinea.137 tions of African ancestry. Later, a single nucleotide change
RH, KELL, DUFFY, AND KIDD ANTIGENS AND ANTIBODIES
in the FYB gene caused the FYA polymorphism in people of
European and Asian ancestry. Fy has been cloned from sev-
eral nonhuman primates, including chimpanzees, squirrel 211 Jka/Jkb
monkeys, and rhesus monkeys, and from cows, pigs, rabbits, Asp280Asn
and mice.127 Extracellular
Lipid
Function Bilayer
The importance of Fy as a receptor for the malarial parasite
Intracellular
P. vivax is well established, but its biologic role as a chemo-
kine receptor on RBCs, endothelial cells, and brain is not yet COOH
clear. The chemokine receptors are a family of proteins that
are receptors on target cells for the binding of chemokines. NH2
Chemokines are so named because they are cytokines that are
chemotactic (cause cell migration), and their receptors are Figure 7–6 Predicted 10-transmembrane domain structure of the
Kidd/urea transporter. The polymorphism responsible for the Kidd anti-
an active area of investigation.148 Chemokine receptors have gens and the site for the N-glycan are indicated.
been found principally on lymphocytes, where they are cou-
pled to G proteins and activate intracellular signaling path-
ways that regulate cell migration into tissues. Unlike other
chemokine receptors, Fy does not have a conserved amino third extracellular loop is large and carries an N-glycan at
acid DRY-motif in the second extracellular loop, and cells Asn211 (Fig. 7–6). The protein has 10 cysteine residues, but
transfected with Fy and stimulation with chemokines do only 1 is predicted to be extracellular, a fact that explains why
not mobilize free calcium.120 Also, Fy can bind chemokines the antigens are not sensitive to disulfide reagents. There is
from both the CXC (IL-8, MGSA) and the CC (RANTES, internal homology between the N-terminal and C-termi-
MCP-1, MIP-1) classes of chemokines.127 These features have nal halves of the protein, with each containing an LP box
led investigators to hypothesize that it may act as a scaven- (LPXXTXPF) characteristic of urea transporters.155 Isolation
ger or sink for excess chemokine release into the circula- of the Kidd protein from RBCs was elusive, and cloning of the
tion.149,150 If the function of Fy on RBCs is to scavenge excess Kidd gene was accomplished with primers complementary
chemokine, it may be that Fy(a−b−) individuals would be to the rabbit kidney urea transporter. A clone isolated from
more susceptible to septic shock143 or to cardiac damage a human bone marrow library with the PCR product was
after infarction. Renal allografts have been reported to have confirmed by in vitro transcription–translation to encode a
shorter survival in African American Fy(a−b−) recipients,151 protein that carried Kidd blood group antigens.156
and Fy has been shown to be upregulated in the kidney
during renal injury.152 7
Gene
Abundant expression of Fy on high endothelial venules
and sinusoids in the spleen,143 which is a site central to che- The Kidd blood group gene (HUT11 or SLC14A1) is located 91
mokine-induced leukocyte trafficking, suggests a role in at chromosome 18q11–q12.157,158 The HUT11 gene has 11
leukocyte migration into the tissues. Fy is also similar to exons and spans approximately 30 kb. Two alternative poly-
the receptor for endothelins (ETB), vasoactive proteins that adenylation sites generate transcripts of 4.4 and 2.0 kb, which
strongly influence vascular biology and that may also be appear to be used equally. Exons 4 to 11 encode the mature
mitogenic.127 protein.17,159
Antigens
THE KIDD BLOOD GROUP SYSTEM
Two antigens, Jka and Jkb, are responsible for the three com-
History mon phenotypes—Jk(a+b−), Jk(a−b+), and Jk(a+b+). The
two antigens are found with similar frequency in white per-
The Kidd blood group system was discovered in 1951, when sons but show large differences in other ethnic groups (Table
a “new” antibody in the serum of Mrs. Kidd (who also had 7–8).1 The Jk(a−b−) phenotype is rare but occurs with
anti-K) caused HDN in her sixth child. Jk came from the greater incidence in Asian and Polynesian people.
initials of the baby (John Kidd), because K had been used
previously for Kell.153 Anti-Jkb was found 2 years later in
the serum of a woman who had a transfusion reaction (she
also had anti-Fya).154 Kidd system antibodies are character-
istically found in sera with other blood group antibodies,
suggesting that Kidd antigens are not particularly immuno- Table 7–8 Kidd Phenotypes and Prevalence
genic. However, Kidd antibodies induce a rapid and robust
Prevalence(%)
anamnestic response that is responsible for their reputation
for causing severe delayed hemolytic transfusion reactions. Phenotype White African Asian
This chapter provides information for the 16 blood group systems N-glycan, and an approximate Mr of 43,000 on sodium
not covered in Chapters 6 and 7. The systems are discussed here dodecyl sulfate polyacrylamide gel electrophoresis (SDS-
in the order of their International Society of Blood Transfusion PAGE); there are approximately 1million copies of GPA per
(ISBT) numbers (see Table 5–3). Antibodies to antigens in these RBC. GPB has 11 potential sites for O-glycans, no N-glycan,
systems are less common than those described in the preceding an approximate Mr of 25,000 on SDS-PAGE, and there are
chapters, and information regarding their general clinical sig- approximately 200,000 copies per RBC.7,8
nificance, when known, is summarized in Tables 5–5 and 5–6. M and N antigens are carried on alternative forms of GPA
The red blood cell (RBC) membrane components carrying the and are the result of amino acid substitutions at residues 1
antigens of the systems described in this chapter are depicted and 5. The M antigen has Ser at position 1 and Gly at posi-
in Figure 8–1. The antigens of the Chido-Rodgers system are tion 5, whereas the N antigen has Leu at position 1 and Glu at
adsorbed onto the RBC and are not integral membrane compo- position 5. The first 26 amino acids of the N form of GPA are
nents; thus, they are not included in Figure 8–1. identical to those of GPB. Anti-N reagents prepared for use
in the clinical setting are formulated to detect the N antigen
on GPA but not the N on GPB (N).9–13 Using these reagents,
THE MNS SYSTEM (ISBT SYSTEM 002) human RBCs type as M+N−, M−N+, or M+N+.
S and s antigens are carried on alternative forms of GPB.
History At amino acid residue 29, Met is critical for the S antigen and
II Thr for the s antigen. Both antigens also involve the amino
The MNS system, discovered in 1927, was the second blood acids at residues 25, 28, 34, and 35.14 Persons who are S−s−,
96 group system to be recognized. The first two antigens, M and usually blacks, may be negative for the high-incidence anti-
N, were named from the second and fifth letters of the word gen U owing to a deletion of or an altered form of GYPB.15–17
immune, because the corresponding antibodies were pro- Because M/N and S/s are on homologous proteins that
duced through immunization of rabbits with human RBCs, are encoded by adjacent genes, they are inherited en bloc,
and it was thought that the first letter, I, might be confused accounting for linkage dysequilibrium between the antigens
with the number 1.1,2 The S antigen, the next to be identified, (Table 8–1).
was named from the city (Sydney) where the first anti-S was The MNS blood group system is highly polymorphic,
discovered.3 When the antithetical antigen was identified, the with 43 antigens.7 Many of the antigens are uncommon,
logical name s was used. The name for the high-incidence resulting from an amino acid substitution or multiple rear-
antigen U was derived from the “almost universal distribu- rangements between GYPA and GYPB (Table 8–2).7,15 Low-
tion of the new blood factor.”4 Many of the other antigens incidence antigens in the MNS blood group system are as
were named after the original family in which an antigen- follows, in alphabetical order: Cla, DANE, Dantu, ERIK, Far,
positive neonate suffered from hemolytic disease of the HAG, He, Hil, Hop, Hut, MARS, Me, Mg, Mia, MINY, Mit,
newborn (HDN). Mta, Mur, MUT, Mv, Nob, Nya, Or, Osa, Ria, sD, SAT, Sta, TSEN,
Vr, Vw.7 The rare null phenotypes of this system—En(a−),
Gene, Protein, and Antigens which lacks MN antigens; U−, which lacks Ss−antigens; and
Mk Mk, which lacks both MN and Ss antigens—most often
MNS antigens are carried on glycophorin A (GPA) and GPB, result from gene deletions.7,18 Some antigens that are asso-
which are encoded by homologous genes (GYPA and GYPB, ciated with the MNS system but are not numbered by the
respectively) on chromosome 4q28–q31.5 GYPA has seven ISBT Working Party on Terminology for Red Cell Surface
exons, and GYPB has five exons and one pseudo-exon. A Antigens are a consequence of altered glycosylation at resi-
third homologous gene, GYPE, completes this glycophorin dues 2, 3, and 4 of GPA. They include Tm, Sj, M1, Can, Sext,
gene family, but it is not clear whether the product of GYPE and Hu.19
is expressed on the RBC membrane. The glycophorin gene
cluster encompasses approximately 330 kb in the following
order: GYPA, GYPB, GYPE.6 Antibodies
Glycophorin A and GPB are single-pass membrane sialo- Anti-M and anti-N antibodies are usually cold-reactive, clini-
glycoproteins oriented with their N termini to the exterior cally insignificant antibodies that are naturally occurring; that
of the RBC. GPA has 15 potential sites for O-glycans, one is, present in persons who have not been previously transfused
OTHER BLOOD GROUP ANTIGENS AND ANTIBODIES
SINGLE-PASS PROTEINS MULTI-PASS PROTEINS GPI-LINKED
PROTEINS NH2
NH2
Extracellular
Cytoplasmic
COOH NH2 COOH
Figure 8–1 Diagram of the red blood cell membrane illustrates the type of membrane components that carry the blood group antigens described
herein. The figure does not show components carrying Chido/Rodgers antigens, because they are not integral membrane components, or most of
the components carrying the blood group collections, and high- and low-incidence antigens, because their structure is unknown.
*
Genes that result in more than 1 transcript (t).
tissues occurs early in erythroid differentiation, but the Gene, Protein, and Antigens
associated antigens are detectable only later in erythroid
development.8,24 Lutheran (LU), along with Secretor, provided the first example
of autosomal linkage in humans, the first example of autoso-
mal crossing over, and the first indication that crossing over
Evolution
in humans is more common in females than in males.35,36
Glycophorin A and the MN antigens are found on RBCs of all The LU gene, located on chromosome 19q13.2–q13.3,
anthropoid apes (chimpanzee, gorillas, orangutan, and gib- consists of 15 exons distributed over approximately 12 kb
bon) and Old World monkeys, but GPB has been found only of DNA. LU encodes the Lutheran (Lu) glycoprotein and a
in humans and the anthropoid apes. Only humans have S and s spliced form of the glycoprotein, basal cell adhesion molecule
II antigens. The glycophorin gene duplication that generated GPA (B-CAM). Lu glycoprotein passes through the RBC mem-
and GPB is not present in gibbons or orangutans and probably brane once with the N terminus oriented to the extracellular
98 occurred in a common ancestor of anthropoid apes.25 surface, and the cytoplasmic region of Lutheran glycoprotein
interacts with the membrane cytoskeleton.37 The protein is
predicted to have five disulfide-bonded, extracellular, Ig
Function
superfamily (IgSF) domains (two variable-region and three
Glycophorin A (also known as membrane inhibitor of reactive constant-region sets).38 The Lu glycoprotein has five sites for
lysis type II, MIRL II) may function as a complement regula- N-glycans and is O-glycosylated and has an approximate Mr
tor26 and is a receptor for bacteria, viruses, and Plasmodium of 85,000 on SDS-PAGE; there are 1500 to 4000 copies per
falciparum malaria parasites.9,27,28 It is a chaperone for band 3 RBC.39,40 The minor isoform (B-CAM) has an approximate
transport to the RBC membrane and is the major component Mr of 78,000, because it lacks the cytoplasmic tail and has an
contributing to the negatively charged RBC glycocalyx,8,29 increased expression on epithelial cancer cells.38,41
which may contribute antiadhesion properties.30–32 As single- The Lutheran system consists of four pairs of antigens
pass membrane proteins, the glycophorins have been con- (Lua/Lub, Lu6/Lu9, Lu8/Lu14, Aua/Aub) and 10 independent
sidered candidate mediators of transmembrane signaling in antigens. The Lua antigen is present in 5% of blacks and 8%
RBCs.33 Rare GPA and GPB null phenotypes—En(a−), U−, of whites, and the antithetical antigen, Lub, is present in more
and MkMk—have RBCs that survive normally; people with than 99% of random blood samples.7 Only Lu9 and Lu14 are
these rare phenotypes have no apparent health defects.8 low-incidence antigens; the others are either polymorphic
(Lua, Aua, and Aub) or of high incidence (Lub, Lu3, Lu4, Lu5,
Lu6, Lu7, Lu8, Lu11, Lu12, Lu13, Lu16, Lu17, and Lu20). The
THE LUTHERAN SYSTEM molecular basis of Lua/Lub is His77Arg, and that of Aua/Aub
(ISBT SYSTEM 005) is Thr539Ala.42,43
The Lua and Lub antigens, and most of the other Lutheran
antigens, are resistant to treatment by papain, ficin, sialidase,
History and low concentrations (50 mM) of dithiothreitol (DTT), and
Anti-Lua was first found in 1946 in the serum of a patient are sensitive to treatment by trypsin and α-chymotrypsin.7
who had received multiple transfusions, and it agglutinated Lutheran-null phenotypes have three genetic back-
8% of random samples.34 The system should have been grounds: homozygozity for a recessive gene mutation Lu,
named Lutteran, after the name of the donor of the Lu(a+) a dominant suppressor gene In(Lu) (inhibitor of Lutheran),
RBCs; however, the handwritten label on the blood sample and an X-linked recessive gene XS44 (Table 8–3). Only the
from the original donor was misread. autosomal recessive Lu(a−b−) is a true null phenotype;
OTHER BLOOD GROUP ANTIGENS AND ANTIBODIES
Table 8–3 Characteristics of Lu(a–b–) Phenotypes
Lu(a–b–) Phenotype Lutheran Antigens Make Anti-Lu3 CD44 CDw75 I/i Antigen
weak expression of Lutheran antigens are found on RBCs blast–extracellular matrix interactions in the marrow and
of the two other types. The original Lu(a−b−) phenotype reticulocyte release into the circulation.33
was discovered by the proposita herself when her own RBCs
were not agglutinated by anti-Lua or anti-Lub.44 Later, it was
Function
shown that these RBCs had weak expression of Lutheran
antigens and that the phenotype was inherited in a dominant Lutheran blood group protein and B-CAM bind laminin, a
manner and unlinked to LU.44 The presence of an inhibitor major component of basement membranes.49 The cytoplas-
gene, In(Lu),45 was proposed. In(Lu) is also associated with mic domain interacts with the erythroid skeleton through
reduced expression of CD44,46,47 and weak expression of P1, spectrin binding,37,50 and the nature of the extracellular and
AnWj, Inb, i, and MER2 blood group antigens.48 cytoplasmic domains suggests receptor and signal transduc-
tion functions. On RBCs, Lutheran glycoprotein may be
Antibodies involved in the adherence of RBCs to vascular endothelial
cells and in the vaso-occlusion that is characteristic of sickle
Antibodies in this system are rarely encountered because the cell disease.51 RBCs from patients with sickle cell disease
antigens are not highly immunogenic (see Table 5–5).21 Anti- express approximately 1½ times more Lutheran glycoprotein
Lua are usually IgG and reactive on the IAT, but they may than do normal RBCs, and have a proportional increase in
be IgM and may directly agglutinate Lu(a+) RBCs, giving binding to laminin.52 Lutheran is phosphorylated in sickle
characteristic stringy agglutinates surrounded by unaggluti- cells, and phosphorylation has been shown to regulate the
nated RBCs. Anti-Lua has not been implicated in transfusion adhesion to laminin.53 Although the function of the Lutheran
reactions and has rarely caused mild HDN. Anti-Aua and glycoprotein is not fully defined, its involvement in cell-to-
anti-Aub are rare, usually found in sera that contain other cell and cell-to-substrate adhesion is probable.
antibodies. They are IgG, react on the IAT, and can cause mild
transfusion reactions, but neither has caused HDN.
Several Lutheran antibodies are directed at antigens of high DIEGO SYSTEM (ISBT SYSTEM 010)
incidence; the one most frequently encountered is anti-Lub. 8
Anti-Lub are IgG and react by the IAT but can occasionally History
be IgM. Anti-Lub can cause mild transfusion reactions and 99
has rarely caused mild HDN. Lu(b−) blood should be used The Diego blood group system was named after the producer
for transfusion, but only about 1 in 500 donors are Lu(b−). of the first example of anti-Dia. The antibody, which had
Anti-Lu3 is found only in the serum of immunized people caused HDN in a Venezuelan baby, was reported in detail in
of the rare recessive Lu(a−b−) phenotype. The antibody is 1955. Anti-Dib was described in 1967, and for many years, the
usually IgG, is reactive on the IAT, and may cause a delayed system consisted of two antithetical antigens, Dia and Dib.20
transfusion reaction or HDN. Blood with the Lu(a−b−) phe- The finding that Dia and Dib are carried on band 3, the anion
notype should be used for transfusion of patients with these exchanger (AE1),54,55 was the beginning of the expansion of
antibodies. the Diego system. In 1995, Bruce and colleagues55,56 located
Antibodies to other high-incidence Lutheran antigens, the Wra and Wrb antigens on band 3. Since then, many low-
usually weak IgG antibodies, have not been reported to cause incidence antigens have been assigned to this system, which
HDN. With the exception of one example of anti-Lu6, which now consists of 18 antigens.57–59
was shown to destroy transfused Lu6 RBCs, these specifici-
ties have not caused transfusion reactions. If a patient has Gene, Protein, and Antigens
formed a Lutheran antibody directed at a high-incidence
antigen, it is important to test for compatible siblings and The gene encoding band 3 and the Diego antigens, SLC4A1
encourage the patient to donate blood for long-term storage (solute carrier family 4, anion exchanger member 1; DI; AE1;
when clinical status permits. Anti-Lu9 and anti-Lu14 define EPB-3), consists of 20 exons distributed over 18 kb of DNA
low-incidence antigens that have not been reported to cause and is located on chromosome 17q12–q21.60,61 There are
transfusion reactions or HDN.7 Most randomly collected more than 1 million copies of band 3 in the RBC membrane,
blood is antigen negative and compatible. making it the most abundant integral RBC protein. The gly-
coprotein passes through the RBC membrane multiple times
and has both the N terminus and the C terminus oriented to
Expression
the cytoplasm. Band 3 has one large N-glycan, on the fourth
Lutheran antigens are expressed weakly on cord RBCs. The extracellular loop, with repeating lactosaminyl groups that
antigens are present in various tissues, including brain, heart, accounts for the broad banding pattern of approximate Mr
kidney, liver, lung, pancreas, placenta, and skeletal muscle.38 95,000 to 105,000 on SDS-PAGE. The N-glycan of band 3
Lutheran glycoprotein is expressed late during erythroid carries over half of the red cell A, B, H, I, and i blood group
differentiation and may play a role in mediating erythro- activity.62
BLOOD BANKING The Dia/Dib polymorphism is located on the last extracel- the other letters of the name were already in use. The
lular loop and is defined by Pro854 (Dib) or Leu854 (Dia).55 logic was that if all the other letters had been used, “Why
The Dia antigen is rare in most populations but is polymor- not t?,” or “Why t?” (Yt) (M. Pickles, personal communication,
phic in people of Mongoloid ancestry. The incidence in 1999).
South American Indians may be as high as 54%, and 10% to
12% of Native Americans are Di(a+).63 The incidence of Dib Gene, Protein, and Antigens
is generally greater than 99.9%; however, the incidence of Dib
among Native Americans is reduced to 96% and is likely to ACHE encodes acetylcholinesterase (AChE), a glycosylphos-
be lower in those populations with a high incidence of Dia. phatidylinositol-linked (GPI-linked) glycoprotein that exists
The Wra/Wr b polymorphism is located on the fourth extra- as a dimer in the RBC membrane.73 The gene is located on
cellular loop, close to the insertion into the RBC membrane, chromosome 7q22 and consists of six exons.7,74,75 Alternative
and is defined by Lys658 (Wrb) or Glu658 (Wra); however, splicing results in different domains at the C terminus of
for expression, the Wrb antigen also requires the presence of AChE. AChE glycoprotein has N-glycans and O-glycans and
normal GPA.56,64,65 an approximate Mr of 160,000 (72,000 monomer) on SDS-
The other antigens in the Diego system—Wda, Rba, PAGE; there are 10,000 copies per RBC.7,73
WARR, ELO, Wu, Bpa, Moa, Hga, Vga, Swa, BOW, NFLD, Jna, Yta and Ytb antigens are antithetical and a consequence of
and KREP57—are of low incidence and are each associated an amino acid substitution on AChE, His353Asn (originally
with a single point mutation.58,59 The antigens are resistant numbered 322).76 Yta occurs with an incidence of more than
to treatment of RBCs by proteolytic enzymes, sialidase, DTT, 99% in random blood samples. Ytb has an incidence of 8% in
chloroquine, and acid.7 most populations, but of 20% or higher in Israelis.77
The antigens are sensitive to treatment of RBCs by papain,
ficin, α-chymotrypsin, and DTT but resistant to treatment
Antibodies by trypsin, sialidase, chloroquine, and acid.7
Diego antibodies are usually IgG that react on the IAT and do
not bind complement. These antibodies have caused transfu- Antibodies
sion reactions (usually delayed) and HDN.7,66
Yt antibodies usually are IgG, are reactive on IAT, and do
not bind complement. These antibodies have caused delayed
Expression transfusion reactions but not HDN.
Diego antigens are expressed on RBCs of newborns. Band 3,
in addition to its presence on RBCs, is present in the inter- Expression
calated cells of the distal and collecting tubules of the kidney
and on granulocytes.62 Acetylcholinesterase is expressed on hematopoietic and
II innervated tissue (including brain and muscle).78,79 The
antigens are expressed weakly on RBCs of newborns and
Evolution are absent from RBCs of people with paroxysmal nocturnal
100
A variant form of band 3, Memphis I (56Glu), has a faster hemoglobinuria (PNH) III.80
migration on SDS-PAGE than the more common form (56Lys).
Evidence suggests that the primordial gene encodes 56Glu and Evolution
that 56Lys is the result of a more recent mutation.67
A form of AChE has been found in the electric fish Torpedo
californica.81
Function
Band 3 is the major anion HCO3−Cl− exchanger in RBC Function
membrane. This function is critical for RBC CO2 uptake
from the tissues and release of CO2 in the lungs. In addition, Acetylcholinesterase is a well-characterized enzyme that
band 3 has an important structural role in the RBC mem- hydrolyzes acetylcholine and is an essential component of
brane through the interaction of its N-terminal domain cholinergic neurotransmission.78 The role of AChE in RBCs
with ankyrin, band 4.2, and band 4.1 in the membrane skel- is unknown,82 but the molecule is enzymatically active.73
eton.61,62 An altered form of band 3, with a deletion of amino
acid residues 400 through 408, is present in ovalocytes of
Southeast Asian people and causes the RBCs to be rigid.68–70 Xg BLOOD GROUP SYSTEM
Numerous mutations exist in the predicted cytoplasmic or (ISBT SYSTEM 012)
transmembrane domains of band 3 and give rise to heredi-
tary spherocytosis, congenital acanthocytosis, and distal History
renal tubular acidosis.59,71
Anti-Xga, discovered in 1962, detects an antigen encoded
by a locus on the X chromosome. The “X” was used because
Yt BLOOD GROUP SYSTEM of the association with the X chromosome, and the g stood
(ISBT SYSTEM 011) for Grand Rapids, Michigan, the home of the male patient
who had received multiple transfusions and who made
the first anti-Xga.83 Xga has been useful in linkage stud-
History
ies involving the X chromosome and in sex chromosome
The Yt system was named in 1956,72 the last letter of the aneuploidy, in which an abnormal number of X chromosomes
antibody producer’s name, Cartwright, being used because occurs.63
implicated in cell-to-cell adhesion events,85,88 activates a
Do(a+b−) + 0 + + + 18 11
Do(a.+b
. +) + + + + + 49 44
Do(a−b+) 0 + + + + 33 45
Gy(a−) 0 0 0 0 0 Rare 0
Hy− 0 Weak Weak 0 0 0 Rare
Jo(a−) Weak 0/Weak + Weak 0 0 Rare
Expression phenotype lack LW antigens and type LW(a−b−). The LWab
Expression Antibodies
Ge antigens are expressed on RBCs of newborns. Gerbich Antibodies in the Cromer system are usually IgG, are reactive
antigens are weak on protein 4.1-deficient RBCs because on the IAT, and do not bind complement. The antibodies have 8
the membranes of such cells have reduced levels of GPC and caused mild delayed transfusion reactions but not HDN.
GPD. The majority of RBC samples with Leach or Gerbich When a patient’s antibody is directed at a high-incidence 105
phenotypes have a weak expression of Kell blood group sys- antigen, it is important to test siblings in the quest for com-
tem antigens. GPC and GPD are expressed on erythroblasts patible blood and to urge the patient to donate blood for
and fetal liver and in several nonerythroid tissues, including long-term storage when clinical status permits.66
kidney, brain cerebellum, and ileum.7
Expression
Function
The Cromer antigens are expressed on RBCs of newborns.
Glycophorin C and GPD are possibly involved in RBC mem- DAF is preferentially expressed on the apical surface of
brane integrity via interaction with protein 4.1. Both glycopho- trophoblasts and may protect the conceptus from anti-
rins are markedly reduced in protein 4.1-deficient RBCs.141 body-mediated hemolysis.146 DAF is not expressed on RBCs
from patients with PNH III. Dr(a−) variant RBCs express
inherited Cromer antigens very weakly.
CROMER BLOOD GROUP SYSTEM Cromer antigens are present in the plasma and urine of
(ISBT SYSTEM 021) people with the corresponding antigen on their RBCs. This
soluble form of the antigens can be used for hemaggluti-
nation inhibition tests, although the urine requires prior
History
concentration.147
The Cromer system was named after the first antibody pro-
ducer, Mrs. Cromer. When the antibody was first identified Function
in 1965, it was believed to be anti-Gob; later, however, it was
recognized as a new specificity, and in 1975, it was renamed Decay accelerating factor prevents assembly and accelerates the
anti-Cra. decay of C3 and C5 convertases, decreasing the deposition of
C3 on the RBC surface and thereby reducing complement-
mediated hemolysis.143
Gene, Protein, and Antigens
Five of the six known people with the Inab phenotype,
Blood group antigens in the Cromer system are carried on the null of the system, do not have significant complement-
the complement regulatory protein, decay accelerating fac- induced lysis in vivo.148 Protein-losing intestinal disorders
tor (DAF, CD55). The DAF gene, located at chromosome have been reported to be associated with Inab, but there
BLOOD BANKING is not a clear disease association.33 Dra is the receptor for RBCs of patients with autoimmune diseases. CR1 is present
uropathogenic Escherichia coli.7 on B cells, a subset of T cells, monocytes, macrophages,
neutrophils, eosinophils, glomerular podocytes, and splenic
follicular dendritic cells.151
KNOPS BLOOD GROUP SYSTEM
(ISBT SYSTEM 022) Function
CR1 has an inhibitory effect on complement activation by
History
both the classical and alternative pathways. CR1 binds C3b
The antigens Kna, Knb, McCa, Sla, and Yka had long been and C4b, thereby mediating phagocytosis by neutrophils
grouped together for serologic reasons. In 1991, the Knops and monocytes. RBC CR1 is important in the processing of
blood group system was established when these antigens were immune complexes, binding them for transport to the liver
shown to be on complement receptor 1 (CR1). The system and spleen for removal from the circulation. The presence of
was named after Mrs. Knops, the first antibody producer. CR1 on other blood cells and tissues suggests that it has mul-
tiple functions.149 The CR1 copy number per RBC (and thus
antigen strength) is reduced in SLE, cold agglutinin disease,
Gene, Protein, and Antigens
PNH, hemolytic anemia, insulin-dependent diabetes melli-
Knops antigens are encoded by various forms of CR1.149 Like tus, acquired immunodeficiency syndrome, some malignant
DAF, the CR1 gene is located within the RCA cluster on chro- tumors, and any condition associated with increased clearance
mosome 1q32. CR1 has four allotypes: A, B, C, and D. The of immune complexes.
most common allotypes are A (82%) and B (18%); the other CR1, and the Sla antigen in particular, may act as a receptor
two are rare. for the malarial parasite Plasmodium falciparum, and thus,
CR1 (CD35) is an unusual protein with 30 short consen- the Sl(a−) phenotype may provide selective advantage.152
sus repeat domains. SDS-PAGE reveals the approximate Mr
of the CR1 allotypes as follows: 190,000 (C allotype), 220,000
(A allotype), 250,000 (B allotype), and 280,000 (D allotype). INDIAN BLOOD GROUP SYSTEM
Of 20 potential N-glycan sites, only 6 to 8 are usually occu- (ISBT SYSTEM 023)
pied; there are four cysteine residues. Each RBC contains 20
to 1500 copies of CR1. The CR1 glycoprotein passes through
History
the RBC membrane once with its N terminus toward the
extracellular surface. The molecular basis of the McCa/McCb The Ina antigen, reported in 1973, is on RBCs from 4% of
polymorphism is associated with a Lys1590Glu missense Indians from Bombay. This blood group system was named
mutation, and the Sla/Vil polymorphism with an Arg1601Gly because of its association with India.
II missense mutation.150 The antigens are weakened by treat-
ment of RBCs with ficin and papain, are sensitive to treatment Gene, Protein, and Antigens
106 by trypsin, α-chymotrypsin, and 200 mM of DTT, and are
resistant to sialidase, 50 mM of DTT, and acid.7 Ina and Inb, the antigens of the Indian system, are carried on
With the exception of the low-incidence antigen Knb, CD44 (synonyms: In(Lu)-related p80; HUTCH-1; H-CAM;
the antigens in this system are fairly common and have a GP90(HERMES), Pgp-1; ECRMIII; Ly-24; p85).153 The
similar prevalence (>90%) in different populations; how- CD44 gene is located at chromosome 11p13 and consists of
ever, Sla is present on RBCs of 98% of whites but on only at least 19 exons, 10 of which are variably spliced. The CD44
60% of blacks. glycoprotein passes through the RBC membrane once, and
Typing for Knops system antigens can be challenging the extracellular N terminus has six cysteine residues, six N-
because of the low level of expression on the RBCs of some glycan sites, four chondroitin sulfate sites, and potential sites
people as well as the lack of potent antisera. Disease processes for O-glycans. There are 2000 to 5000 copies of CD44 per
causing CR1 deficiency and, therefore, weak expression of RBC. SDS-PAGE shows that CD44 has an approximate Mr
Knops system antigens can lead to false-negative results. of 80,000 when reduced. The Ina/Inb polymorphism is due
Furthermore, the low level of expression can lead to variable to Pro46Arg on CD44.154 Inb is a common antigen, and Ina is
results in tests on different samples from the same patient. rare in white persons but has an incidence of 4% in Indians,
10% in Iranians, and nearly 12% in Arabs. The antigens are
sensitive to treatment of RBCs by proteases and DTT but
Antibodies
resistant to treatment with sialidase and acid.7
Antibodies in the Knops system are usually IgG and reactive The In(a−b−) phenotype was described in a patient with
on the IAT, and they do not bind complement. The antibod- a novel form of congenital dyserythropoietic anemia (CDA)
ies do not cause transfusion reactions or HDN; once identi- and CD44 deficiency,155 but it was not possible to ascertain
fied, they can be ignored for clinical purposes. Identification whether the phenotype was genetically determined or related
may be complicated by the fluctuation of antigen expression to the patient’s hematologic disorder. The RBCs of the patient
on RBCs. Anti-Kna is the most common antibody in white also typed AnWj− and Co(a−b−) and had a reduced level of
persons, and anti-Sla is the most common in black persons.20 LWab expression.
Expression Antibodies
The Knops antigens are weakly expressed on RBCs of new- Antibodies in the Indian system are usually IgG and reac-
borns, RBCs with the dominant Lu(a−b−) phenotype, and tive on the IAT, and they do not bind complement. Some
Expression
Expression Evolution
Indian antigens are weakly expressed on cord RBCs as well The Oka antigen is on RBCs from gorillas and chimpanzees
as on RBCs from people with the dominant Lu(a−b−) phe- but not on RBCs from rhesus monkeys, baboons, and mar-
notype and from pregnant women. CD44 is expressed on mosets.158 Homologues of the Oka glycoprotein have been
neutrophils, lymphocytes, monocytes, brain, breast, colon found in the rat (OX-47 or CE9), mouse (basigin), rabbit,
epithelium, gastric tissue, heart, kidney, liver, lung, placenta, and chicken (neurothelin or HT7).160
skin, spleen, thymus, and fibroblasts.
Joint fluid from patients with inflammatory synovitis has Function
higher than normal levels of soluble CD44.155 The serum
CD44 value is elevated in some patients with lymphoma. Human CD147 on tumor cells is thought to bind an unknown
ligand on fibroblasts, which stimulates their production of
collagenase and other extracellular matrix metalloprotein-
Function ases, thus enhancing tumor cell invasion and metastasis.162
In studies with CD147 knockout mice, RBCs were apparently
CD44 has a diverse range of biologic functions involving not compromised. CD147 may be involved in the function of
cell–cell and cell–matrix interactions in cells other than the blood-brain barrier163 and lymphocyte inactivation.164
RBCs.153 It is an adhesion molecule in lymphocytes, mono-
cytes, and some tumor cells. CD44 binds to hyaluronate
and other components of the extracellular matrix and is RAPH BLOOD GROUP SYSTEM
also involved in immune stimulation as well as signaling (ISBT SYSTEM 025)
between cells.157
History
A new polymorphism on RBCs was originally defined by
OK BLOOD GROUP SYSTEM
monoclonal antibodies (1D12, 2F7) and called MER2 (M for
(ISBT SYSTEM 024) monoclonal; ER for Eleanor Roosevelt, the name of the labo-
ratory producing the antibodies). Later, the polymorphism 8
History was also recognized by human polyclonal antibodies, and
Anti-Oka was first identified in 1979 in the serum of a when the MER2 antigen attained system status in 1999, the
system was named RAPH after the first patient to make the 107
Japanese woman (Mrs. Okbutso) who had received a trans-
fusion, and was therefore named after her. After the identifi- specificity.
cation of the gene encoding the Ok protein, the Oka antigen
attained system status in 1999.57 Gene, Protein, and Antigens
The MER2 antigen is encoded by a gene located on chro-
mosome 11p15, but it has not been cloned, and the molec-
Gene, Protein, and Antigens
ular basis of the antigen is not known. Ninety-two percent
The Oka blood group antigen is carried on CD147 and of English blood donors are MER2-positive, and 8% are
is encoded by the OK gene at 19pter-p13.2. CD147 (also MER2-negative. The antigen strength varies among different
known as extracellular matrix metalloproteinase inducer RBC samples. The antigen is resistant to treatment of RBCs
[EMMPRIN]), M6, OX-47, CE9, basigin, gp42, neurothelin, by papain, ficin, and sialidase but sensitive to treatment by
HT7, 5A11) is an N-glycosylated glycoprotein that passes trypsin, α-chymotrypsin, and DTT.7
through the RBC membrane once with its N terminus to
the extracellular surface. It is also a member of the IgSF. On Antibodies
SDS-PAGE, CD147 is shown to have an approximate Mr of
35,000 to 69,000. The Oka polymorphism is due to an amino The three examples of human anti-MER2 (anti-RAPH) are
acid substitution at residue 92 [Glu for Ok(a+) and Lys for IgG and reactive on the IAT, and two of the three antibodies
Ok(a−)].158,159 Oka is resistant to treatment of RBCs by pro- bind complement. These antibodies have not caused HDN
teases, sialidase, DTT, and acid.7 The eight known Ok(a−) nor transfusion reactions; indeed, two siblings with the anti-
probands are Japanese. body have received numerous crossmatch-incompatible RBC
transfusions without problems. The three antibody producers
were Indian Jews.165
Antibodies
Expression
The original example of anti-Oka is IgG and is reactive on
the IAT; it does not bind complement. This antibody caused The antigen is expressed on RBCs of newborns. MER2 is
reduced cell survival but not HDN. Only one other example expressed on fibroblasts, and its expression may be reduced
of human anti-Oka is known. on Lu(a−b−) RBCs of persons with the In(Lu) gene.
BLOOD BANKING
Function tion of the antigens by modification of methionine. Hoppe-Seylers Z
Physiol Chem 1980;361:145–152.
All three people (two probands) with anti-MER2 (anti- 15. Huang C-H, Blumenfeld OO. MNSs blood groups and major gly-
RAPH) had renal failure requiring dialysis. It is possible cophorins: Molecular basis for allelic variation. In Cartron J-P, Rouger
P (eds). Molecular Basis of Human Blood Group Antigens. New York,
that the protein carrying the MER2 antigen is required for Plenum Press, 1995, pp 153–188.
normal kidney function.165 16. Storry JR, Reid ME. Characterization of antibodies produced by S−s−
individuals. Transfusion 1996;36:512–516.
17. Greenwalt TJ, Sasaki T, Sanger R, et al. An allele of the S(s) blood group
OTHER ANTIGENS genes. Proc Natl Acad Sci USA 1954;40:1126–1129.
18. Daniels G. Human Blood Groups, 2nd ed. Oxford, Blackwell Science,
2002.
Table 5–4 lists antigens that are not included in blood group 19. Dahr W, Knuppertz G, Beyreuther K, et al. Studies on the structures
systems. Three of the blood group collections are carbohy- of the Tm, Sj, M1, Can, Sext and Hu blood group antigens. Biol Chem
drate antigens (Ii, GLOB, Unnamed). Antigens of the Ii and Hoppe-Seyler 1991;372:573–584.
20. Issitt PD, Anstee DJ. Applied Blood Group Serology, 4th ed. Durham,
GLOB collections are discussed in Chapter 6. The other two N.C., Montgomery Scientific Publications, 1998.
(Cost, Er) are presumed to be protein antigens, and the anti- 21. Mollison PL, Engelfriet CP, Contreras M. Blood Transfusion in Clini-
bodies to antigens in these systems are generally not clini- cal Medicine, 10th ed. Oxford, Blackwell Science, 1997.
cally significant. Antigens in the 700 series of low-incidence 22. Harvey J, Parsons SF, Anstee DJ, Bradley BA. Evidence for the occur-
rence of human erythrocyte membrane sialoglycoproteins in human
antigens occur in less than 1% of most populations and kidney endothelial cells. Vox Sang 1988;55:104–108.
have no known alleles. Antibodies to many of these low- 23. Anstee DJ, Holmes CH, Judson PA, Tanner MJA. The use of mono-
incidence antigens have caused HDN.7 Antigens in the 901 clonal antibodies to determine the distribution of red cell surface
series of high-incidence antigens occur in more than 90% proteins on cells and tissues. In Agre PC, Cartron J-P (eds). Protein
of people, have no known alleles, and cannot be placed in Blood Group Antigens of the Human Red Cell: Structure, Function,
and Clinical Significance. Baltimore, Johns Hopkins University Press,
a blood group system or collection. Antibodies to Vel, Lan, 1992, pp 170–181.
Ata, Jra, AnWj, PEL, ABTI, and MAM can cause HDN and 24. Southcott MJG, Tanner MJ, Anstee DJ. The expression of human
transfusion reactions.7,57 Finding blood for patients with any blood group antigens during erythropoiesis in a cell culture system.
of the antibodies to these antigens can be difficult, and the Blood 1999;93:4425–4435.
25. Blancher A, Reid ME, Socha WW. Cross-reactivity of antibodies to
patient should be encouraged to predeposit autologous units human and primate red cell antigens. Transf Med Rev 2000;14:161–179.
for long-term frozen storage. JMH and EMM are carried on 26. Tomita A, Radike EL, Parker CJ. Isolation of erythrocyte membrane
GPI-linked proteins, and antibodies to these antigens are of inhibitor of reactive lysis type II. Identification as glycophorin A.
little clinical concern. J Immunol 1993;151:3308–3323.
27. Hadley TJ, Miller LH, Haynes JD. Recognition of red cells by malaria
parasites: the role of erythrocyte-binding proteins. Transf Med Rev
1991;5:108–113.
REFERENCES 28. Miller LH. Impact of malaria on genetic polymorphism and genetic
diseases in Africans and African Americans. Proc Natl Acad Sci USA
II 1. Landsteiner K, Levine P. A new agglutinable factor differentiating 1994;91:2415–2419.
human blood. Proc Soc Exp Biol Med 1927;24:600–603. 29. Jentoft N. Why are proteins O-glycosylated? Trends Biochem Sci
108 2. Levine P. A review of Landsteiner’s contributions to human blood 1990;15:291–294.
groups. Transfusion 1961;1:45–52. 30. Groves JD, Tanner MJ. The effects of glycophorin A on the expression
3. Garratty G, Dzik WH, Issitt PD, et al. Terminology for blood group of the human red cell anion transporter (band 3) in Xenopus oocytes.
antigens and genes: historical origins and guidelines in the new mil- J Membr Biol 1994;140:81–88.
lennium. Transfusion 2000;40:477–489. 31. Groves JD, Tanner MJ. Role of N-glycosylation in the expression of human
4. Wiener AS, Unger LF, Gordon EG. Fatal hemolytic transfusion reac- band 3-mediated anion transport. Mol Membr Biol 1994;11:31–38.
tion caused by sensitization to a new blood factor U. JAMA 1953;153: 32. Groves JD, Tanner MJ. Glycophorin A facilitates the expression of
1444–1446. human band 3-mediated anion transport in Xenopus oocytes. J Biol
5. Rahuel C, London J, d’Auriol L, et al. Characterization of cDNA clones Chem 1992;267:22163–22170.
for human glycophorin A. Use for gene localization and for analysis of 33. Reid ME, Mohandas N. Red blood cell blood group antigens: Structure
normal or glycophorin-A-deficient (Finnish type) genomic DNA. Eur and function. Semin Hematol 2004;41:93–117.
J Biochem 1988;172:147–153. 34. Callender STE, Race RR. A serological and genetical study of multiple
6. Vignal A, London J, Rahuel C, Cartron J-P. Promoter sequence and antibodies formed in response to blood transfusion by a patient with
chromosomal organization of the genes encoding glycophorins A, B lupus erythematosus diffusus. Ann Eugen 1946;13:102.
and E. Gene 1990;95:289–293. 35. Cook PJL. The Lutheran-secretor recombination fraction in man:
7. Reid ME, Lomas-Francis C. Blood Group Antigen FactsBook, 2nd ed. a possible sex-difference. Ann Hum Genet 1965;28:393–401.
San Diego, Academic Press, 2003. 36. Mohr J. A search for linkage between the Lutheran blood group and
8. Chasis JA, Mohandas N. Red blood cell glycophorins. Blood 1992;80: other hereditary characters. Acta Path Microbiol Scand 1951;28:80–96.
1869–1879. 37. Parsons SF, Lee G, Spring FA, et al. Lutheran blood group glycopro-
9. Dahr W. Immunochemistry of sialoglycoproteins in human red blood tein and its newly characterized mouse homologue specifically bind
cell membranes. In Vengelen-Tyler V, Judd WJ (eds). Recent Advances α5 chain-containing human laminin with high affinity. Blood 2001;97:
in Blood Group Biochemistry. Arlington, Va., American Association of 312–320.
Blood Banks, 1986, pp 23–65. 38. Parsons SF, Mallinson G, Holmes CH, et al. The Lutheran blood group
10. Fukuda M. Molecular genetics of the glycophorin A gene cluster. glycoprotein, another member of the immunoglobulin superfamily, is
Semin Hematol 1993;30:138–151. widely expressed in human tissues and is developmentally regulated in
11. Anstee DJ, Spring FA, Parsons SF, et al. Molecular background of human liver. Proc Natl Acad Sci USA 1995;92:5496–5500.
human blood group antigens. In Hackel E, Tippett P (eds). Human 39. Parsons SF, Mallinson G, Judson PA, et al. Evidence that the Lub blood
Genetics 1994: A Revolution in Full Swing. Bethesda, Md., American group antigen is located on red cell membrane glycoproteins of 85 and
Association of Blood Banks, 1994, pp 1–52. 78 kd. Transfusion 1987;27:61–63.
12. Blanchard D. Human red cell glycophorins: biochemical and antigenic 40. Anstee DJ. Blood group-active surface molecules of the human red
properties. Transf Med Rev 1990;4:170–186. blood cell. Vox Sang 1990;58:1–20.
13. Reid ME. Some concepts relating to the molecular genetic basis of cer- 41. Rahuel C, Kim CL, Mattei MG, et al. A unique gene encodes spliceo-
tain MNS blood group antigens. Transf Med 1994;4:99–111. forms of the B-cell adhesion molecule cell surface glycoprotein of
14. Dahr W, Gielen W, Beyreuther K, Krüger J. Structure of the Ss blood epithelial cancer and of the Lutheran blood group glycoprotein. Blood
group antigens. I. Isolation of Ss-active glycopeptides and differentia- 1996;88:1865–1872.
OTHER BLOOD GROUP ANTIGENS AND ANTIBODIES
42. El Nemer W, Rahuel C, Colin Y, et al. Organization of the human LU 68. Jarolim P, Palek J, Amato D, et al. Deletion in erythrocyte band 3 gene
gene and molecular basis of the Lua/Lub blood group polymorphism. in malaria-resistant Southeast Asian ovalocytosis. Proc Natl Acad Sci
Blood 1997;89:4608–4616. USA 1991;88:11022–11026.
43. Parsons SF, Mallinson G, Daniels GL, et al. Use of domain-deletion 69. Tanner MJ, Bruce L, Martin PG, et al. Melanesian hereditary ovalocytes
mutants to locate Lutheran blood group antigens to each of the five have a deletion in red cell band 3. Blood 1991;78:2785–2786.
immunoglobulin superfamily domains of the Lutheran glycoprotein: 70. Schofield AE, Reardon DM, Tanner MJ. Defective anion transport
Elucidation of the molecular basis of the Lua/Lub and the Aua/Aub activity of the abnormal band 3 in hereditary ovalocytic red blood
polymorphisms. Blood 1997;89:4219–4225. cells. Nature 1992;355:836–838.
44. Crawford MN. The Lutheran blood group system: Serology and genetics. 71. Jarolim P, Shayakul C, Prabakaran D, et al. Autosomal dominant distal
In Pierce SR, Macpherson CR (eds). Blood Group Systems: Duffy, Kidd renal tubular acidosis is associated in three families with heterozygos-
and Lutheran. Arlington, Va., American Association of Blood Banks, ity for the R589H mutation in the AE1 (band 3) Cl−HCO3− exchanger.
1988, pp 93–117. J Biol Chem 1998;273:6380–6388.
45. Taliano V, Guevin RM, Tippett P. The genetics of a dominant inhibitor 72. Eaton BR, Morton JA, Pickles MM, White KE. A new antibody, anti-
of the Lutheran antigens. Vox Sang 1973;24:42–47. Yta, characterizing a blood group of high incidence. Br J Haematol
46. Telen MJ, Eisenbarth GS, Haynes BF. Human erythrocyte antigens. 1956;2:333–341.
Regulation of expression of a novel erythrocyte surface antigen by the 73. Spring FA, Gardner B, Anstee DJ. Evidence that the antigens of the Yt
inhibitor Lutheran In(Lu) gene. J Clin Invest 1983;71:1878–1886. blood group system are located on human erythrocyte acetylcholines-
47. Spring FA, Dalchau R, Daniels GL, et al. The Ina and Inb blood group terase. Blood 1992;80:2136–2141.
antigens are located on a glycoprotein of 80,000 MW (the CDw44 74. Coghlan G, Kaita H, Belcher E, et al. Evidence for genetic linkage
glycoprotein) whose expression is influenced by the In(Lu) gene. between the KEL and YT blood group loci. Vox Sang 1989;57:88–89.
Immunology 1988;64:37–43. 75. Getman DK, Eubanks JH, Camp S, et al. The human gene encoding
48. Poole J. Review: The Lutheran blood group system—1991. Immuno- acetylcholinesterase is located on the long arm of chromosome 7. Am
hematology 1992;8:1–8. J Hum Genet 1992;51:170–177.
49. El Nemer W, Gane P, Colin Y, et al. The Lutheran blood group glyco- 76. Bartels CF, Zelinski T, Lockridge O. Mutation at codon 322 in the
proteins, the erythroid receptors for laminin, are adhesion molecules. human acetylcholinesterase (ACHE) gene accounts for YT blood
J Biol Chem 1998;273:16686–16693. group polymorphism. Am J Hum Genet 1993;52:928–936.
50. Kroviarski Y, El NW, Gane P, et al. Direct interaction between the Lu/ 77. Levene C, Bar-Shany S, Manny N, et al. The Yt blood groups in Israeli
B-CAM adhesion glycoproteins and erythroid spectrin. Br J Haematol Jews, Arabs, and Druse. Transfusion 1987;27:471–474.
2004;126:255–264. 78. Taylor P. The cholinesterases. J Biol Chem 1991;266:4025–4028.
51. Udani M, Zen Q, Cottman M, et al. Basal cell adhesion molecule 79. Li Y, Camp S, Rachinsky TL, et al. Gene structure of mammalian ace-
Lutheran protein—The receptor critical for sickle cell adhesion to tylcholinesterase: Alternative exons dictate tissue-specific expression.
laminin. J Clin Invest 1998;101:2550–2558. J Biol Chem 1991;266:23083–23090.
52. Hines PC, Zen Q, Burney SN, et al. Novel epinephrine and cyclic AMP- 80. Telen MJ, Rosse WF, Parker CJ, et al. Evidence that several high-fre-
mediated activation of BCAM/Lu-dependent sickle (SS) RBC adhesion. quency human blood group antigens reside on phosphatidylinositol-
Blood 2003;101:3281–3287. linked erythrocyte membrane proteins. Blood 1990;75:1404–1407.
53. Gauthier E, Rahuel C, Wautier MP, et al. Protein kinase A-dependent 81. Sussman JL, Harel M, Frolow F, et al. Atomic structure of acetylcholin-
phosphorylation of Lutheran/basal cell adhesion molecule glycopro- esterase from Torpedo californica: a prototypic acetylcholine-binding
tein regulates cell adhesion to laminin α5. J Biol Chem 2005;280: protein. Science 1991;253:872–879.
30055–30062. 82. Lawson AA, Barr RD. Acetylcholinesterase in red blood cells. Am
54. Spring FA, Bruce LJ, Anstee DJ, Tanner MJ. A red cell band 3 variant J Hematol 1987;26:101–111.
with altered stilbene disulphonate binding is associated with the Diego 83. Mann JD, Cahan A, Gelb AG, et al. A sex-linked blood group. Lancet
(Dia) blood group antigen. Biochem J 1992;288:713–716. 1962;1:8–10.
8
55. Bruce LJ, Anstee DJ, Spring FA, Tanner MJ. Band 3 Memphis variant II. 84. Ellis NA, Tippett P, Petty A, et al. PBDX is the XG blood group gene.
Altered stilbene disulfonate binding and the Diego (Dia) blood group Nature Genet 1994;8:285–290. 109
antigen are associated with the human erythrocyte band 3 mutation 85. Tippett P, Ellis NA. The Xg blood group system: A review. Transf Med
Pro854→Leu. J Biol Chem 1994;269:16155–16158. Rev 1998;12:233–257.
56. Bruce LJ, Ring SM, Anstee DJ, et al. Changes in the blood group 86. Daniels GL, Anstee DJ, Cartron JP, et al. International Society of Blood
Wright antigens are associated with a mutation at amino acid 658 in Transfusion working party on terminology for red cell surface anti-
human erythrocyte band 3: a site of interaction between band 3 and gens: Vienna report. Vox Sang 2001;80:193–197.
glycophorin A under certain conditions. Blood 1995;85:541–547. 87. Schlossman SF, Bounsell L, Gilks W, et al. CD antigens 1993. Blood
57. Daniels GL, Anstee DJ, Cartron JP, et al. Terminology for red cell surface 1994;83:879–880.
antigens—Oslo report. Vox Sang 1999;77:52–57. 88. Gelin C, Aubrit F, Phalipon A, et al. The E2 antigen, a 32 kd glycopro-
58. Zelinski T. Erythrocyte band 3 antigens and the Diego blood group tein involved in T-cell adhesion processes, is the MIC2 gene product.
system. Transf Med Rev 1998;12:36–45. EMBO J 1989;8:3253–3259.
59. Jarolim P, Rubin HL, Zakova D, et al. Characterization of seven low 89. Bernard G, Breittmayer JP, de Matteis M, et al. Apoptosis of immature
incidence blood group antigens carried by erythrocyte band 3 protein. thymocytes mediated by E2/CD99. J Immunol 1997;158:2543–2550.
Blood 1998;92:4836–4843. 90. Lewis M, Kaita H, Chown B. Scianna blood group system. Vox Sang
60. Zelinski T, Coghlan G, White L, Philipps S. The Diego blood group 1974;27:261–264.
locus is located on chromosome 17q. Genomics 1993;17:665–666. 91. Xu H, Foltz L, Sha Y, et al. Cloning and characterization of human
61. Tanner MJ. Molecular and cellular biology of the erythrocyte anion erythroid membrane-associated protein, human ERMAP. Genomics
exchanger (AE1). Semin Hematol 1993;30:34–57. 2001;76:2–4.
62. Tanner MJA. The structure and function of band 3 (AE1): recent 92. Wagner FF, Poole J, Flegel WA. The Scianna antigens including Rd are
developments. Mol Membr Biol 1997;14:155–165. expressed by ERMAP. Blood 2003;101:752–757.
63. Race RR, Sanger R. Blood Groups in Man, 6th ed. Oxford, Blackwell 93. Spring FA, Herron R, Rowe G. An erythrocyte glycoprotein of appar-
Scientific, 1975. ent Mr 60,000 expresses the Sc1 and Sc2 antigens. Vox Sang 1990;
64. Blumenfeld OO, Huang C-H, Xie SS, Blancher A. The MNSs blood 58:122–125.
group system: molecular biology of glycophorins in humans and 94. DeMarco M, Uhl L, Fields L, et al. Hemolytic disease of the newborn
nonhuman primates. In Blancher A, Klein J, Socha WW (eds). Molecu- due to the Scianna antibody, anti-Sc2. Transfusion 1995;35:58–60.
lar Biology and Evolution of Blood Group and MHC Antigens in 95. Su YY, Gordon CT, Ye TZ, et al. Human ERMAP: An erythroid adhesion/
Primates. Berlin, Springer-Verlag, 1997, pp 113–146. receptor transmembrane protein. Blood Cells Mol Dis 2001;27:938–949.
65. Reid ME. Contribution of MNS to the study of glycophorin A and 96. Banks JA, Hemming N, Poole J. Evidence that the Gya, Hy and Joa
glycophorin B. Immunohematology 1999;15:5–9. antigens belong to the Dombrock blood group system. Vox Sang
66. Reid ME, Øyen R, Marsh WL. Summary of the clinical significance of 1995;68:177–182.
blood group alloantibodies. Semin Hematol 2000;37:197–216. 97. Gubin AN, Njoroge JM, Wojda U, et al. Identification of the Dom-
67. Palatnik M, Simoes ML, Alves ZM, Laranjeira NS. The 60 and 63 brock blood group glycoprotein as a polymorphic member of the
kda proteolytic peptides of the red cell membrane band-3 protein: ADP-ribosyltransferase gene family. Blood 2000;96:2621–2627.
their prevalence in human and nonhuman primates. Hum Genet 98. Eiberg H, Mohr J. Dombrock blood group (DO): assignment to
1990;86:126–130. chromosome 12p. Hum Genet 1996;98:518–521.
BLOOD BANKING 99. Koch-Nolte F, Haag F, Braren R, et al. Two novel human members of an 128. Lee G, Spring FA, Parsons SF, et al. Novel secreted isoform of adhe-
emerging mammalian gene family related to mono-ADP-ribosylating sion molecule ICAM-4: potential regulator of membrane-associated
bacterial toxins. Genomics 1997;39:370–376. ICAM-4 interactions. Blood 2003;101:1790–1797.
100. Koch-Nolte F, Haag F, Braren R, et al. Erratum (to Koch-Nolte et al. 129. O’Neill GJ, Yang SY, Tegoli J, et al. Chido and Rodgers blood groups
Genomics 1997;39:370–376). Genomics 1999;55:130. are distinct antigenic components of human complement C4. Nature
101. Rios M, Hue-Roye K, Øyen R, et al. Insights into the Holley-negative 1978;273:668–670.
and Joseph-negative phenotypes. Transfusion 2002;42:52–58. 130. Yu CY, Campbell RD, Porter RR. A structural model for the location
102. Rios M, Hue-Roye K, Lee AH, et al. DNA analysis for the Dombrock of the Rodgers and the Chido antigenic determinants and their cor-
polymorphism. Transfusion 2001;41:1143–1146. relation with the human complement component C4A/C4B isotypes.
103. Lucien N, Celton J-L, Le Pennec P-Y, et al. A short deletion within the Immunogenetics 1988;27:399–405.
blood group Dombrock locus causing a Do-null phenotype. Blood 131. Giles CM, Jones JW. A new antigenic determinant for C4 of relatively
2002;100:1063–1064. low frequency. Immunogenetics 1987;26:392–394.
104. Rios M, Storry JR, Hue-Roye K, et al. Two new molecular bases for the 132. Westhoff CM, Sipherd BD, Wylie DE, Toalson LD. Severe anaphylactic
Dombrock null phenotype. Br J Haematol 2002;117:765–767. reactions following transfusions of platelets to a patient with anti-Ch.
105. Preston GM, Agre P. Isolation of the cDNA for erythrocyte integral Transfusion 1992;32:576–579.
membrane protein of 28 kilodaltons: Member of an ancient channel 133. Tippett P, Storry JR, Walker PS, et al. Glycophorin A-deficient red cells
family. Proc Natl Acad Sci USA 1991;88:11110–11114. may have a weak expression of C4-bound Ch and Rg antigens. Immu-
106. Preston GM, Smith BL, Zeidel ML, et al. Mutations in aquaporin-1 in nohematology 1996;12:4–7.
phenotypically normal humans without functional CHIP water chan- 134. Moulds JM. Association of blood group antigens with immuno-
nels. Science 1994;265:1585–1587. logically important proteins. In Garratty G (ed). Immunobiology of
107. Chretien S, Catron JP. A single mutation inside the NPA motif of aqua- Transfusion Medicine. New York, Marcel Dekker, 1994, pp 273–297.
porin-1 found in a Colton-null phenotype [letter]. Blood 1999;93: 135. Reid ME, Spring FA. Molecular basis of glycophorin C variants and
4021–4023. their associated blood group antigens. Transf Med 1994;4:139–149.
108. Joshi SR, Wagner FF, Vasantha K, et al. An AQP1 null allele in an Indian 136. Colin Y, Rahuel C, London J, et al. Isolation of cDNA clones and com-
woman with Co (a−b−) phenotype and high-titer anti-Co3 associated plete amino acid sequence of human erythrocyte glycophorin C. J Biol
with mild HDN. Transfusion 2001;41:1273–1278. Chem 1986;261:229–233.
109. Smith BL, Preston GM, Spring FA, et al. Human red cell aquaporin 137. Chang S, Reid ME, Conboy J, et al. Molecular characterization of
CHIP. I. Molecular characterization of ABH and Colton blood group erythrocyte glycophorin C variants. Blood 1991;77:644–648.
antigens. J Clin Invest 1994;94:1043–1049. 138. Telen MJ, Le Van Kim C, Chung A, et al. Molecular basis for ellipto-
110. King LS, Agre P. Pathophysiology of the aquaporin water channels. cytosis associated with glycophorin C and D deficiency in the Leach
Ann Rev Physiol 1996;58:619–648. phenotype. Blood 1991;78:1603–1606.
111. King LS, Kozono D, Agre P. From structure to disease: the evolving tale 139. Winardi R, Reid M, Conboy J, Mohandas N. Molecular analysis of
of aquaporin biology. Nat Rev Mol Cell Biol 2004;5:687–698. glycophorin C deficiency in human erythrocytes. Blood 1993;81:
112. King LS, Choi M, Fernandez PC, et al. Defective urinary-concen- 2799–2803.
trating ability due to a complete deficiency of aquaporin-1. NEJM 140. Sistonen P. Some notions on clinical significance of anti-Lsa and
2001;345:175–179. independence of Ls from Colton, Kell, and Lewis blood group loci
113. King LS, Nielsen S, Agre P, Brown RH. Decreased pulmonary vascular (abstract). XIX Congress of the International Society of Blood Trans-
permeability in aquaporin-1-null humans. Proc Natl Acad Sci USA fusion 1986;652.
2002;99:1059–1063. 141. Alloisio N, Morle L, Bachir D, et al. Red cell membrane sialoglycopro-
114. Smith BL, Baumgarten R, Nielsen S, et al. Concurrent expression of tein b in homozygous and heterozygous 4.1(−) hereditary elliptocyto-
erythroid and renal aquaporin CHIP and appearance of water channel sis. Biochim Biophys Acta 1985;816:57–62.
II activity in perinatal rats. J Clin Invest 1993;92:2035–2041. 142. Post TW, Arce MA, Liszewski MK, et al. Structure of the gene for
115. Roudier N, Verbavatz JM, Maurel C, et al. Evidence for the presence human complement protein decay accelerating factor. J Immunol
110 of aquaporin-3 in human red blood cells. J Biol Chem 1998;273: 1990;144:740–744.
8407–8412. 143. Lublin DM, Atkinson JP. Decay-accelerating factor: Biochemistry,
116. Roudier N, Ripoche P, Gane P, et al. AQP3 deficiency in humans and molecular biology and function. Ann Rev Immunol 1989;7:35–58.
the molecular basis of a novel blood group system, GIL. J Biol Chem 144. Telen MJ, Rao N, Udani M, et al. Molecular mapping of the cromer
2002;277:45854–45859. blood group Cra and Tca epitopes of decay accelerating factor: toward
117. Landsteiner K, Wiener AS. An agglutinable factor in human blood the use of recombinant antigens in immunohematology. Blood
recognized by immune sera for rhesus blood. Proc Soc Exp Biol Med 1994;84:3205–3211.
1940;43:223. 145. Lublin DM, Kompelli S, Storry JR, Reid ME. Molecular basis of Cromer
118. Sistonen P, Tippett P. A “new” allele giving further insight into the LW blood group antigens. Transfusion 2000;40:208–213.
blood group system. Vox Sang 1982;42:252–255. 146. Holmes CH, Simpson KL, Wainwright SD, et al. Preferential expres-
119. Bailly P, Hermand P, Callebaut I, et al. The LW blood group glycopro- sion of the complement regulatory protein decay accelerating factor
tein is homologous to intercellular adhesion molecules. Proc Natl Acad at the fetomaternal interface during human pregnancy. J Immunol
Sci USA 1994;91:5306–5310. 1990;144:3099–3105.
120. Mallinson G, Martin PG, Anstee DJ, et al. Identification and partial 147. Daniels GL, Okubo Y, Yamaguchi H, et al. UMC, another Cromer-
characterization of the human erythrocyte membrane component(s) related blood group antigen. Transfusion 1989;29:794–797.
that express the antigens of the LW blood-group system. Biochem 148. Telen MJ, Green AM. The Inab phenotype: Characterization of
J 1986;234:649–652. the membrane protein and complement regulatory defect. Blood
121. Hermand P, Gane P, Mattei MG, et al. Molecular basis and expression of 1989;74:437–441.
the LWa/LW b blood group polymorphism. Blood 1995;86:1590–1594. 149. Ahearn JM, Fearon DT. Structure and function of the complement recep-
122. Bloy C, Hermand P, Blanchard D, et al. Surface orientation and anti- tors, CR1 (CD35) and CR2 (CD21). Adv Immunol 1989;46:183–219.
gen properties of Rh and LW polypeptides of the human erythrocyte 150. Moulds JM, Zimmerman PA, Doumbo OK, et al. Molecular identifica-
membrane. J Biol Chem 1990;265:21482–21487. tion of Knops blood group polymorphisms found in long homologous
123. Hermand P, Le Pennec PY, Rouger P, et al. Characterization of the gene region D of complement receptor 1. Blood 2001;97:2879–2885.
encoding the human LW blood group protein in LW+ and LW− pheno- 151. Rao N, Ferguson DJ, Lee SF, Telen MJ. Identification of human eryth-
types. Blood 1996;87:2962–2967. rocyte blood group antigens on the C3b/C4b receptor. J Immunol
124. Giles CM. The LW blood group: a review. Immunol Commun 1991;146:3502–3507.
1980;9:225–242. 152. Rowe JA, Moulds JM, Newbold CI, Miller LH. P. falciparum rosetting
125. Bailly P, Tontti E, Hermand P, et al. The red cell LW blood group pro- mediated by a parasite-variant erythrocyte membrane protein and
tein is an intercellular adhesion molecule which binds to CD11/CD18 complement-receptor 1. Nature 1997;388:292–295.
leukocyte integrins. Eur J Immunol 1995;25:3316–3320. 153. Telen MJ, Ferguson DJ. Relationship of Inb antigen to other antigens
126. Spring FA, Parsons SF, Ortlepp S, et al. Intercellular adhesion mol- on In(Lu)-related p80. Vox Sang 1990;58:118–121.
ecule-4 binds α4b1 and αV-family integrins through novel integrin- 154. Telen MJ, Udani M, Washington MK, et al. A blood group-related
binding mechanisms. Blood 2001;98:458–466. polymorphism of CD44 abolishes a hyaluronan-binding consen-
127. Hermand P, Gane P, Huet M, et al. Red cell ICAM-4 is a novel ligand for sus sequence without preventing hyaluronan binding. J Biol Chem
platelet-activated αIIbβ3 integrin. J Biol Chem 2003;278:4892–4898. 1996;271:7147–7153.
OTHER BLOOD GROUP ANTIGENS AND ANTIBODIES
155. Parsons SF, Jones J, Anstee DJ, et al. A novel form of congenital dys- 161. Anstee DJ, Spring FA. Red cell membrane glycoproteins with a broad
erythropoietic anemia associated with deficiency of erythroid CD44 tissue distribution. Transf Med Rev 1989;3:13–23.
and a unique blood group phenotype [In(a−b−), Co(a−b−)]. Blood 162. Biswas C, Zhang Y, DeCastro R, et al. The human tumor cell-derived
1994;83:860–868. collagenase stimulatory factor (renamed EMMPRIN) is a member of
156. Joshi SR. Immediate haemolytic transfusion reaction due to anti-Inb. the immunoglobulin superfamily. Cancer Res 1995;55:434–439.
Vox Sang 1992;63:232–233. 163. Seulberger H, Unger CM, Risau W. HT7, neurothelin, basigin,
157. Goldstein LA, Zhou DFH, Picker LJ, et al. A human lymphocyte hom- gp42 and OX-47—many names for one developmentally regulated
ing receptor, the Hermes antigen, is related to cartilage proteoglycan immunoglobulin-like surface glycoprotein on blood-brain barrier
core and link proteins. Cell 1989;56:1063–1072. endothelium, epithelial tissue barriers and neurons. Neurosci Lett
158. Williams BP, Daniels GL, et al. Biochemical and genetic analysis of the 1992;140:93–97.
OKa blood group antigen. Immunogenetics 1988;27:322–329. 164. Ghebrehiwet B, Lu PD, Zhang W, et al. Identification of functional
159. Spring FA, Holmes CH, Simpson KL, et al. The Oka blood group anti- domains on gC1Q-R, a cell surface protein that binds to the globular
gen is a marker for the M6 leukocyte activation antigen, the human “heads” of C1Q, using monoclonal antibodies and synthetic peptides.
homolog of OX-47 antigen, basigin and neurothelin, an immunoglob- Hybridoma 1996;15:333–342.
ulin superfamily molecule that is widely expressed in human cells and 165. Daniels GL, Levene C, Berrebi A, et al. Human alloantibodies detect-
tissues. Eur J Immunol 1997;27:891–897. ing a red cell antigen apparently identical to MER2. Vox Sang 1988;55:
160. Kasinrerk W, Fiebiger E, Stefanova I, et al. Human leukocyte activation 161–164.
antigen M6, a member of the Ig superfamily, is the species homologue
of rat OX-47, mouse basigin, and chicken HT7 molecule. J Immunol
1992;149:847–854.
111
Chapter 9
Human Platelet Antigens
Thomas J. Kunicki ● Diane J. Nugent
FcγR, Fc receptor γ chain; vWF, von Willebrand factor; VLA, very late activation (antigen); CD, cell differentiation (antigen).
NAIT
Incidence: 1 per 3000 in a retrospective study, 1 per 2200 births in one prospective study.
Maternal antibodies produced against paternal antigens on fetal platelets.
Similar to erythroblastosis fetalis, except that 50% of cases occur during first pregnancy.
Most frequently implicated antigens are HPA-1a and HPA-5b (United States/Europe).
In the case of responsiveness to HPA-1a, there is a high-risk association with HLA-DRB3*0101or DQB1*02.
In the case of responsiveness to HPA-6b, there is an increased association with HLA-DRB1*1501, DQA1*0102 or-DQB1*0602.
PTP
Nearly all of the reported patients have been females previously sensitized by pregnancy or transfusion (<5% were males).
Thrombocytopenia usually occurs 1 week after transfusion.
Homozygous HPA-1b individuals account for a majority (>60%) of cases.
High-risk association with HLA DRB3*0101or DQB1*02.
Enigmatically, the recipient’s antigen-negative platelets are destroyed by autologous antibody.
The laboratory diagnosis is normally made by a com- alloantigen haplotypes among different racial and ethnic
parison of maternal and paternal genotypes and a serologic populations (see Figs. 9–1 through 9–4). The data for four
search for antibodies in maternal plasma or serum that react of the most clinically prominent alloantigen haplotypes
with paternal platelets. are depicted, namely, HPA-1b, HPA-2b, HPA-3b, and
Only a fraction of those mothers negative for the platelet HPA-5b (ITGA2 haplotype 3). Each of these alloantigens
antigen in question deliver infants affected with NAIT. For represent one of a diallelic system. Consequently, the fre-
example, in the western world, responsiveness to HPA-1a is quencies of the immunogenic (less common alleles) are
most commonly the cause of NAIT, with severe thrombo- depicted, and those of the more common alleles can be
cytopenia due to maternal alloantibodies occurring in 1 per deduced.
1200 pregnancies, yet the frequency of homozygous HPA-1b HPA-1b has a relatively higher frequency among
mothers among white non-Hispanics is much higher (2%). A North Africans (0.26), whites of European background
key to understanding this discrepancy lies in the finding that (0.15), and black Americans (0.1), and is much rarer
responsiveness to HPA-1a shows a human leukocyte antigen among Asians (0.009), Aboriginals (0.008), and Native
(HLA) restriction.17 Individuals who are homozygous for Pro33 North Americans (<0.001) (Fig. 9–1). Not surprisingly,
(homozygous HPA-1b) and responsive to the predominant the most frequent antigen target in NAIT in white popu-
HPA-1a antigen are almost exclusively HLA DRB3*010117 or lations is HPA-1a (estimated at 78%).22 However, among
II DQB1*02.18 In the case of DRB3*0101, the calculated risk fac- Asians, anti-HPA-1a has never been shown to be involved
tor is 141, a risk level equivalent to that of the hallmark of HLA in NAIT.24
114 restriction in autoimmune disease, ankylosing spondylitis and HPA-2b has the highest frequencies among Central
HLA-B27.19 In contrast, responsiveness of homozygous HPA- Africans (0.3), North Africans (0.19), black Americans (0.17),
1a individuals to the HPA-1b allele is not linked to HLA.19,20 T and Native North Americans (0.12) (Fig. 9–2). A somewhat
cells are the likely candidates for providing HLA restriction in lower frequency is found among whites (0.09), native South
this case, and Maslanka and coworkers19 provided elegant evi- Americans (Amerindians) (0.07), Asians (0.05), and Southeast
dence that in one case of NAIT, T cells that share CDR3 motifs
are stimulated by peptides that contain the same Leu33 poly-
morphism that is recognized by anti-HPA-1a alloantibodies. HPA-1b
In five cases of alloimmunization against the less frequent
antigen, HPA-6b, there was a clear association between respon- White (16)
siveness and the major histocompatibility complex (MHC)
haplotype DRB1*1501, DQA1*0102, DQB1*0602.21 This allo-
immunization likely involves different HLA class II molecules Asian (14)
than those involved in immunization against HPA-1a.
Responsiveness to HPA-1a is not the sole cause of NAIT. In
a large study of 348 cases of clinically suspected NAIT,22 78% North African (2)
of serologically confirmed cases were due to anti-HPA-1a and Black American (4)
19% to anti-HPA-5b. All other specificities accounted for no
Amerindian (3)
more than 5% of cases. In reports from other laboratories, the
association of NAIT with other alloantigens, such as HPA-3a,
HPA-3b, HPA-1b, and HPA-2b, has been noted, but is rare.23 Aboriginal (2)
Obviously, differences in haplotype frequencies between racial
or ethnic populations will have an important impact on the 0 0.1 0.2 0.3
frequency of responsiveness to a particular alloantigen. Mean Allele Frequency
Figure 9–1 HPA-1b frequencies among selected racial/ethnic popula-
Haplotype Frequencies in Different Racial/ tions. The frequencies of HPA-1b (abscissa) among the populations listed
Ethnic Groups on the ordinate are depicted. Each data entry (bar) represents the mean
and standard deviation for the number of studies indicated in paren-
A substantial amount of information has been accumu- theses. References: White,188–202 Asian,190,203–215 North African,216,217
lated with respect to the relative frequency of the major black American,188–190,203 Amerindian,188,189,218 and Aboriginal.202,219
HUMAN PLATELET ANTIGENS
HPA-2b
Post-transfusion Purpura
White (18)
Post-transfusion purpura follows 7 to 10 days after an
South-East Asian (1) immunogenic blood (platelet) transfusion (see Table 9–3).
Asian (12) It most often affects previously nontransfused, multiparous
women. As with NAIT, there is an increased risk to develop
Central African (1) PTP among HLA-DR3–positive individuals, and HPA-1a is
North African (2) the antigen most often implicated (in European white popu-
lations).28,29
Black American (4)
Amerindian (3)
Immunochemistry of Platelet Alloantigens
Native American (1)
By convention, the designation human platelet antigen has
Aboriginal (1) been assigned to alloantigen systems in which the precise
0 0.1 0.2 0.3 0.4 polymorphism that accounts for the serologic difference
Mean Allele Frequency between alleles has been identified. Of 16 HPA systems
Figure 9–2 HPA-2b frequencies among selected racial/ethnic populations.
defined to date (see Table 9-2), 9 are expressed by the integ-
The frequencies of HPA-2b (abscissa) among the populations listed on the rin β3 subunit (HPA-1, HPA-4, HPA-6, HPA-7, HPA-8, HPA-
ordinate are depicted. Each data entry (bar) represents the mean and stan- 10W, HPA-11W, HPA-14W, and HPA-16W), 2 are localized
dard deviation for the number of studies indicated in parentheses. References: on the integrin αIIb subunit (HPA-3 and HPA-9W), 2 are
White,36,188–203 Southeast Asian,36 Asian,36,190,204,205,207–213,215 Central found on the integrin α2 subunit (HPA-5 and HPA-13W), 1
African,36 North African,36,216 black American,188–190,203 Amerindian,188,189,218
Native American,220 and Aboriginal.202 is expressed by the GPIbα (HPA-2), 1 by GPIbβ (HPA-12W),
and the last is located on the CD109 (HPA-15).
The vast majority of clinical episodes following alloim-
munization involve five of these systems: HPA-1, HPA-2,
Asians (0.03), and it is essentially nonexistent among HPA-3, HPA-5, and HPA-15.
Aboriginals (<0.001). The reports of HPA-2b frequency
among Asians have provided disparate values ranging from 0 HPA-1
to 0.1 (mean = 0.05). HPA-2a has been implicated in NAIT The HPA-1 alloantigen system is defined by the Leu33/Pro33
among both whites and Asians,23,25–27 but is much more com- polymorphism that is enclosed within a small 13–amino acid
mon among whites. loop formed by the pairing of Cys26 with Cys38. This region
HPA-3b appears at a very low frequency (0.05) uniquely of the molecule is held proximal to the distal cysteine-rich
among Aboriginals. Otherwise, HPA-3b seems to be promi- region in the middle of β3 by a long-range disulfide bond
nently represented among all other races tested: Asians linking Cys5 and Cys435.30 The complex structure of the β3 9
(0.47), whites (0.4), Amerindians (0.36), black Americans molecule and the sensitivity of the determinants to this
(0.34), and North Africans (0.32) (Fig. 9–3). HPA-3a is a fre- structure is the likely explanation for observed heterogene- 115
quent target for alloimmunization in NAIT and PTP among ity in binding properties of anti-HPA-1a alloantibodies.31
whites. Although all alloantibodies would bind to the denatured
The highest frequencies of HPA-5b are found among molecule or to recombinant amino-terminal segments of
Central Africans (0.25), Aboriginals (0.2), North Africans the molecule expressed in Escherichia coli,32 a subset of anti-
(0.18), and black Americans (0.17) (Fig. 9–4). Lower fre- bodies appear to require presentation of the antigenic loop
quencies are found among whites (0.09) and Southeast within a more native environment (e.g., the nondenatured
Asians (0.09), and the lowest frequencies occur among molecule).31Anti-HPA-1a antibodies inhibit clot retrac-
Asians (0.03), Amerindians (0.02), and Native North tion and platelet aggregation; in the latter case, presumably
Americans (0.02). because they block the binding of fibrinogen.33
Aboriginal (2)
0 0.1 0.2 0.3 0.4 0.5 0.6
Mean Allele Frequency
BLOOD BANKING HPA-5b to HPA-5(a/b) offspring, Panzer and colleagues40 determined
that there was no increased risk of maternal-fetal HPA-5
White (19) incompatibility among mothers whose genotype is 807T.
South-East Asian (1) There are at least 6 major haplotypes of the integrin α2
gene (ITGA2) that are defined by linkage disequilibrium
Asian (13)
between alleles at position –52 (C/T), 807 (C/T), and 1643
Central African (1) (A/G; HPA-5).36
North African (3) HPA-15
Black American (5) Kelton and coworkers41 initially described the Gov system, which
Amerindian (3) is now known to be carried by the 175-kDa glycosyl phospha-
tidylinositol-anchored glycoprotein CD109 and has been des-
Native American (1)
ignated HPA-15. Alloantibodies defining each of the two alleles
Aboriginal (2) have been detected in patients who had received multiple plate-
0 0.1 0.2 0.3 let transfusions41 as well as in patients who developed NAIT.42
Mean Allele Frequency
An A2108C SNP of the CD109 gene results in a Tyr703Ser
substitution that defines the HPA-15 alloantigen system.43 The
Figure 9–4 HPA-5b frequencies among selected racial/ethnic popu-
lations. The frequencies of HPA-5b (abscissa) among the populations
detection of antibodies against HPA-15 is complicated by varia-
listed on the ordinate are depicted. Each data entry (bar) represents tion in expression and instability of CD109 during preparation
the mean and standard deviation for the number of studies indicated and storage.44 Nonetheless, relative to all detected HPA-specific
in parentheses. References: White,36,188–203,220 Southeast Asian,36 antibodies, HPA-15 was found to account for roughly 6% of
Asian,36,189,190,204,205,207–213,215 Central African,36 North African,36,216,217 alloimmunization in a white population.44
black American,188–190,203,220 Amerindian,188,189,218 Native American,220
and Aboriginal.202,219
ISOANTIGENS
Other cells that express β3, as the β subunit of the vitro-
nectin receptor, including endothelial cells, fibroblasts, and Isoantibodies are produced against an epitope that is expressed
smooth muscle cells, also express HPA-1 epitopes.34 This by all normal individuals and is not polymorphic. In the area
could contribute to the complexity of the clinical symptoms of human platelet immunology, a classic example of isoim-
in alloimmune-mediated thrombocytopenia. At this time, munization occurs when a patient with an inherited defi-
little, if anything, is known about the involvement of tissues ciency of a membrane glycoprotein has been subjected to
other than platelets in these conditions. multiple platelet transfusions to correct a bleeding diathesis.
Glanzmann thrombasthenia and BSS are such inherited dis-
HPA-2 orders, wherein the individual either lacks or expresses an
II
Two previously described polymorphisms of the GP Ib-IX-V altered form of αIIbβ3 (Glanzmann thrombasthenia) or GP
complex, termed Ko and Sib, are now known to be reflec- Ib-IX-V (BSS), respectively. Isoantibodies developed by trans-
116
tions of two linked polymorphisms, one of which defines the fused patients do not distinguish any of the allelic forms of
diallelic system, HPA-2.35 In the GPIbα sequence, a Thr/Met the glycoproteins (such as HPA-3 or HPA-1 alloantigens on
polymorphism at residue 145 is associated with HPA-2a and αIIbβ3) but react with the platelets of all normal persons tested.
HPA-2b epitopes, respectively. Because the propositus does not express the platelet glycopro-
There are several major GPIbα gene (GP1BA) haplotypes tein that carries the epitope in question, these antibodies do
that can be defined by linkage disequilibrium between the alleles not bind to their own platelets.
at positions –5 (Kozak), 1018 (Thr145Met; HPA-2), and 1285
(variable number of tandem repeats [VNTR] A, B, C, and D).36 Isoantibodies in Glanzmann Thrombasthenia
HPA-3 Several cases have been documented in which patients with
Glanzmann thrombasthenia have produced antibodies spe-
The HPA-3 system is associated with an Ile843/Ser843 poly-
cific for αIIb, β3, or the αIIbβ3 complex. Recently, we defined
morphism of αIIb.37
an idiotype (OG) that is associated at high frequency with
HPA-5 isoantibodies specific for integrin subunit β3 generated by
Glanzmann thrombasthenia patients.45,46 Patient OG suf-
The HPA-5 system is located on the integrin subunit α2.38 The fered from persistent often-serious bleeding episodes as a
detection of this system was facilitated by the development result of both his Glanzmann thrombasthenia phenotype
of a highly sensitive murine monoclonal antibody-based and the fact that he had generated a very high-titered, IgG
monoclonal antibody immobilization of platelet antigen isoantibody inhibitor of platelet cohesion.46
assay.39 Like the preceding alloantigenic systems, the HPA-5
system is diallelic. Roughly 500 to 2000 copies of α2 are pres-
Isoantibodies in Bernard-Soulier Syndrome
ent on the surface of normal platelets, and each α2 molecule
expresses a single HPA-5 epitope.38 The integrin α2 is dis- Because BSS is less frequently encountered than Glanzmann
tributed on a wide variety of cells, but nothing is currently thrombasthenia, it follows that isoantibodies produced
known about the antigenicity of the HPA-5 determinants on in conjunction with this syndrome are also less frequently
receptors expressed by cell types other than platelets. encountered. In the only clearcut case of such an isoanti-
The density of the α2β1 integrin is known to correlate body,47 the isolated IgG impaired both normal platelet adhe-
with the α2 single-nucleotide polymorphism (SNP) C807T. sion to subendothelial elements and in vitro aggregation in
However, in a study of 79 HPA-5(a/a) mothers giving birth response to ristocetin and bovine factor VIII.
HUMAN PLATELET ANTIGENS
CD36 might directly interfere with platelet production. Later studies
examining the effect of ITP plasma on megakaryocytic growth
The frequency of platelet CD36 (GPIV) deficiency, which in culture confirmed that those plasmas containing antiplate-
occurs most often in Asians and blacks, is approximately 4%, let antibodies resulted in a direct suppression of replication.
and a subset of these individuals is at risk for immuniza- Of note, these studies once again suggested that there might
tion against CD36.48 In five cases of neonatal thrombocyto- be significant differences in the etiology of megakaryocytic
penia, isoantibodies reactive with CD36 were identified in suppression by plasma from children versus adults with ITP.
the mothers of the affected infants. Two black mothers were In addition, autoantigen specificity, especially those antibodies
homozygous for a T1264G substitution in the CD36 gene, directed against GPIb/IX complex in pediatric patients, may
and a European white mother was homozygous for a novel have a profound effect on thrombopoiesis.
deletion of exons 1 through 3. These findings and the results In childhood ITP, the mean yield of megakaryocytic cells
of prior reports indicate that isoimmunization against in culture was significantly reduced in the presence of patient
CD36 can cause neonatal isoimmune thrombocytopenia, plasmas containing either antiplatelet GPIb autoantibod-
refractoriness to platelet transfusions, and PTP.48 ies alone (p<0.001) or antibodies against both GPIb and
αIIbβ3(p<0.001), as compared to control plasmas.61 There was
AUTOANTIGENS AND IDIOPATHIC no significant difference between the mean megakaryocytic
THROMBOCYTOPENIC PURPURA yield in cultures containing control plasmas or patient plas-
mas with either no autoantibodies or autoantibodies specific
Autoimmune (or idiopathic) thrombocytopenic purpura solely for αIIbβ3. The ploidy distribution of megakaryocytic
(ITP) is the most frequently encountered form of immune precursors in cultures containing either control or ITP plasma
thrombocytopenia.49 This disorder can be classified as acute was not significantly different. The importance of autoanti-
or chronic on the basis of the duration of the thrombocyto- bodies in the suppression of megakaryocyte production was
penia, the chronic form persisting longer than 6 months. The demonstrated by platelet adsorption studies. Platelet absorp-
acute, self-limiting form occurs predominantly in children, tion of anti-GPIb autoantibodies in ITP plasmas resulted in a
often following a viral illness or immunization, and affects doubling of megakaryocyte production compared to that in
males and females with equal frequency. Although 10% to the presence of the same plasmas without absorption. On the
15% of childhood ITP lasts longer than 6 months, the chronic other hand, platelet absorption of normal control plasma had
form of immune thrombocytopenia is mainly an adult illness no effect on megakaryocyte yield.
and affects twice as many females as males. Life-threatening Studies with plasmas from adult ITP patients demonstrated
bleeding occurs in up to 1% of patients with ITP. The reason similar suppression of megakaryocytic production, with one
that some patients sustain severe hemorrhagic complications striking difference: The ploidy distribution of megakaryocytic
and others do not remains unexplained, but because of differ- precursors in cultures containing either control or ITP plasma
ences seen in the clinical expression of chronic and acute ITP, was also significantly different. ITP plasmas that suppressed in
it has been theorized that the mechanisms of disease for each vitro megakaryocyte production clearly reduced the percent- 9
form stem from different etiologies.50 Although most auto- ages of 4N, 8N, and 16N megakaryocytes as well.63 Houwerzijl
antibodies apparently induce thrombocytopenia, a minority and colleagues64 showed that megakaryocytes in bone marrow 117
can induce platelet dysfunction with or without an increase aspirates of adult ITP patients exhibit morphologic altera-
in platelet clearance.51,52 The proportion of the two types of tions characteristic of para-apoptosis: mitochondrial swelling
autoantibody, that which leads to platelet clearance and that with cytoplasmic vacuolization, distention of the demarcation
which blocks platelet function, may be a very important fac- membranes, and chromatin condensation within the nucleus.
tor controlling the unpredictable clinical severity of ITP. Importantly, they found that similar para-apoptotic changes
It had been hoped that the antigenic targets in acute and could be induced in normal CD34+ cells differentiated to
chronic forms of ITP might be distinct, so that antigen iden- megakaryocytes in the presence of ITP plasma. These authors
tity might one day be used as an early indicator of clinical suggested that the binding of antiplatelet autoantibodies to
outcome. Subsequent studies have shed more light on this megakaryocytes may play an important role in initiating the
issue and have demonstrated that autoantibody specificity in cascade of programmed cell death.64 These ploidy and para-
chronic versus acute ITP is quite similar, particularly in chil- apoptotic changes were not observed in cultures co-incubated
dren.53,54 Nonetheless, a distinction between antigen speci- with pediatric ITP plasmas despite their ability to suppress
ficity and an acute versus chronic course in ITP may yet be megakaryopoiesis.61 In addition, when megakaryocytes cul-
found in the early stages of the autoimmune response, par- tured with ITP plasma were also treated with FMK, a general
ticularly if additional risk factors such as cytokine response caspase inhibitor, the differential effect of adult and pediatric
and predisposing high-risk polymorphisms in immune ITP plasmas remained. This result would rule out a difference
response genes are included in the evaluation.55–57 in caspase-mediated apoptotic mechanisms between adults
and children as the cause of these findings.61–64
Suppression of Megakaryocyte Production It is more likely that a difference in autoantigen epitopes
and Maturation or the intensity of autoantibody binding to certain platelet
receptors accounts for the difference in apoptotic changes
One area where both autoantigen and age may play a role in seen in adult megakaryocytes that is absent in pediatric stud-
predicting chronicity in ITP is megakaryocyte expansion and ies. In addition, since the immune system is still maturing
thrombopoiesis.58–64 Initial studies by McMillan and cowork- in patients younger than age 5, one would expect to observe
ers58 showed that immunoglobulin G (IgG) produced in vitro increased variability in antibody specificity, affinity, and
by splenic cells from ITP patients would bind to megakary- isotypes among children, with an increased suppression of
ocytes but IgG produced by normal splenic cells or purified megakaryocytopoiesis compared to adults.61,62 Adolescent
normal sera IgG would not, thus suggesting that autoantibody ITP patients are much more likely to have autoantibodies
BLOOD BANKING specific for the GPIb/IX complex and a clinical course simi- normally internal but, at some point, are expressed on the
lar to that of adult ITP.57 Autoantibody binding to the GPIb/ surface of the platelet, perhaps as a result of activation or
IX complex at a certain stage of megakaryocyte development senescence, at which time the autoantibodies can bind and
may trigger signaling pathways that predispose to apoptotic trigger the removal of platelets from the circulation.
changes, resulting in a more profound and lasting suppression The report by Fujisawa and colleagues71 is one of many
of megakaryopoiesis and concomitant thrombocytopenia. suggesting that autoantibodies to αIIbβ3 may bind to cryptic
epitopes or neoantigens that are expressed on platelet activa-
tion or senescence.73 Platelets normally accumulate IgG on the
Glycoproteins as Platelet Autoantigens
membrane surface as they age over 7 to 10 days in the circu-
An extensive characterization of platelet target autoantigens lation and/or become activated, and then they are cleared by
has been accomplished within the past decade. It is now evi- reticuloendothelial cells in the spleen.74 Using a Fab′2 fraction
dent that both T and B cells mediate self-reactivity in immune of IgG, it has been shown that normal individuals have autoan-
thrombocytopenia. Under normal circumstances, immune tibodies in their sera that specifically bind to senescent antigens
recognition may serve to clear damaged or senescent cells, on red cells and platelets and can mediate clearance of spent
but in ITP patients, these processes result in the pathologic or dysfunctional cells.75,76 The role of these naturally occurring
destruction of platelets. In normal individuals, the network of autoantibodies in pathologic immune destruction of cells is
autoreactive T and B cells remains tightly regulated, allowing unclear. Animal studies suggest that the initial binding of the
the effective maintenance of platelet production and integrity. pathogenic antiplatelet autoantibody may trigger the expres-
In autoimmune states, following infection or environmental sion of senescent or cryptic antigens on the platelet membrane,
triggers, pathologic platelet autoreactivity results from an resulting in a second wave of immunoglobulin binding and
imbalance in T-cell signaling and regulation. This state allows clearance by naturally occurring autoantibodies.77
the clonal expansion and somatic mutation of B cells through More recently, additional autoantigenic epitopes have been
epitope spreading, resulting in pathologic autoantibodies. An identified on the β3 subunit. Nardi and coworkers78 identi-
understanding of the role of autoantigens in promoting and fied the peptide β349–66 as a site that is bound by a majority
perpetuating this response has been assisted by an extensive of affinity-purified antibodies isolated from serum immune
characterization of platelet glycoproteins and glycolipids.65,66 complexes of human immunodeficiency virus-1–infected
patients with immune thrombocytopenia. This is not a pep-
Integrin aIIbb3 tide region that is bound by serum IgG antibodies from con-
Integrin αIIbβ3 was the first platelet membrane component trol subjects or patients with the classic form of ITP.
to be identified as a dominant antigen in chronic ITP,67 and Identification of autoimmune epitopes has been aided
subsequent studies from several laboratories have confirmed by the development of the phage-display peptide library.
the important contribution of this receptor to the autoanti- Bowditch and coworkers79 used a filamentous phage library
genic makeup of the human platelet.49 An excellent review of that displayed random linear hexapeptides to identify pep-
II the specific assays used to identify serum antiplatelet anti- tide sequences recognized by platelet-specific autoantibod-
bodies and their autoantigen targets has recently been pub- ies. Plasma antibody eluates from ITP patients were used to
118 lished for readers interested in a detailed description of the select for phage-displaying autoantibody-reactive peptides.
technology and use of these assays in the diagnosis of ITP.68 They identified anti-αIIbβ3 antibody-specific phage encod-
ing the hexapeptide sequences Arg-Glu-Lys-Ala-Lys-Trp
Integrin Subunit b3 (REKAKW), Pro-Val-Val-Trp-Lys-Asn (PVVWKN), and Arg-
Initially, attempts to further localize autoepitopes on either Glu-Leu-Leu-Lys-Met (RELLKM). Each phage showed satu-
integrin subunit were more successful with regard to β3. rable dose-dependent binding to immobilized autoantibody,
Early on, Kekomaki and coworkers69 defined a prominent and binding could be blocked with purified αIIbβ3. The bind-
autoantigenic region as the 33-kDa chymotryptic fragment ing of plasma autoantibody to the phage encoding REKAKW
of β3 located within the cysteine-rich region of β3, which could be blocked with a synthetic peptide derived from the β3
was bound by both plasma autoantibodies and autoantibody cytoplasmic tail; however, binding to the PVVWKN-bearing
eluted from patients’ platelets. Using recombinant fusion phage was not inhibited. Using sequential overlapping pep-
proteins containing the extracellular disulfide-rich region of tides from the β3 cytoplasmic region, an epitope was localized
β3, Beardsley and coworkers70 more recently confirmed that to the sequence Arg-Ala-Arg-Ala-Lys-Trp (β3 734–739).
a disulfide-rich segment of β3 (residues 468–691) immedi- Using a similar phage peptide display approach, Gevorkian
ately external to the proposed membrane spanning domain and coworkers80 identified other mimotopes of platelet auto-
(residues 693–721) contains peptide target antigens for auto- antigens recognized by autoantibodies in sera from 20 ITP
antibodies from ITP patients. patients. These mimotopes exhibited potential homology
Fujisawa and colleagues71 also determined that plasma with sequences on both β3 and GPIb, although a specific
autoantibodies in 5 of 13 patients with chronic ITP bound sequence identity was not proposed.
to peptides representing β3 residues 721–744 or 742–762,
Integrin Subunit aIIb
namely, the carboxy-terminal region of β3, which is pre-
sumed to be located in the cytoplasm of the platelet. Their In a growing body of literature, it now appears that the αIIb mol-
role in the pathogenesis of immune thrombocytopenia is ecule may be the dominant target in the adult form of chronic
unclear because this portion of the β3 molecule is located ITP.81 Autoantibodies reactive with αIIb were identified in two
within the cytoplasm, rather than on the surface, of the rest- patients with chronic ITP,82,83 and in one of these patients,
ing platelet. However, internal autoantigens are the hallmark the antibody was subsequently shown to react with a chymo-
of many immune-mediated diseases, such as polymyosi- tryptic, 65-kDa, COOH-terminal fragment of the αIIb heavy
tis and systemic lupus erythematosus (SLE).72 As in those chain.83 Ethylenediaminetet-raacetic acid (EDTA)-dependent
diseases, one can postulate that certain autoantigens are autoantibodies represent a special category that are adsorbed
HUMAN PLATELET ANTIGENS
by autologous platelets when whole blood is drawn in EDTA. nogenic in humans. The first human monoclonal antibody
In one such case of EDTA-dependent “pseudothrombocyto- against a platelet glycoprotein was derived from an ITP
penia,” an IgM antibody was shown to bind to αIIb by immu- individual producing an antibody against β3.72 This human
noblot assay and crossed immunoelectrophoresis.84 monoclonal antibody detects a neoantigen associated with
In subsequent studies, Gevorkian and colleagues80,85 used β3 that is expressed only on stored or thrombin-activated
phage display to map an autoepitope on αIIb. A filamentous platelets. A number of human monoclonal autoantibodies
phage library was employed that displays random peptides, specific for the heavy chain of GPIb were generated from
11 amino acids in length, flanked on each side by a cysteine to the lymphocytes of an ITP patient with serum autoantibody
identify peptide sequences recognized by antiplatelet autoan- specific for GPIb.90 The heavy-chain variable region genes of
tibodies previously determined to block fibrinogen binding four of these antibodies have been sequenced and found to
to αIIbβ3. Phage expressing the sequence CTGRVPLGFEDLC be markedly homologous to human immunoglobulin, germ-
exhibited saturable dose-dependent binding to the immobi- line, heavy-chain variable region genes.90 Another human
lized autoantibody that could be blocked with purified αIIbβ3 monoclonal autoantibody was produced that is specific for
or αIIb, but not β3. This amino acid sequence exhibits partial the heavy chain of αIIb.73 The epitope recognized by this anti-
identity with amino acid residues 4–10 and 31–35 on αIIb. body (2E7) has been identified as a contiguous amino acid
These results suggest that the autoantibodies in question sequence with residues 231–238 with an immunodominant
bind to a combinatorial epitope within the amino-terminal tryptophan residue at position 235.91 This was the first time
35 amino acids of αIIb. that the precise epitope on αIIb or β3 recognized by a human
Other critical autoantigen epitopes have been identi- antibody had been identified.
fied on αIIb in recent years. Using a combination of lysates The αIIbβ3 molecule is a member of the greater family
from platelets with known protein mutations, Kosugi and of adhesion molecules or integrins, and αIIb has significant
colleagues86 demonstrated that one third of patients with homology with other integrin α chains. A comparison of the
chronic ITP produced anti-αIIbβ3 autoantibodies with a 220–238 sequences of αIIb and seven other integrin α sub-
unique specificity. The binding of these autoantibodies is units revealed significant homology between αIIb and other
inhibited by the single amino acid substitution (D163A) and α integrin chains, yet 2E7 did not bind to the other cell types
a 2-amino acid insertion in the W3 4–1 loop of αIIb. that express these related integrin family members, such as
Autoantigenic epitopes may be conformational, thereby endothelial cells, lymphocytes, monocytes, or neutrophils. In
dependent on the tertiary configuration of the molecule. Given the highly homologous region spanning residues 234–238,
the fluid nature of the platelet membrane and the remarkable αIIb differs from the other α chains only in position 235
morphologic changes that can occur on adhesion and activa- (W→L). Ironically, this tryptophan plays a critical role in
tion, it is not surprising that conformational epitopes might determining the immunogenic epitope recognized by 2E7.
be dynamically expressed. In one of the first descriptions of From the foregoing analysis, it is clear that further studies
this kind, Shadle and Barondes87 described a neoantigen on are required to determine the extent to which the production
αIIb, which is only revealed following the normal binding of of human autoantibodies to platelet glycoproteins is clonally 9
fibrinogen to its receptor, the αIIbβ3 complex. In addition, restricted. Given a selected number of idiotypes related to
serum autoantibodies have been described that demonstrate autoimmunity to αIIbβ3, one could potentially use the anti- 119
increased binding to platelets in the presence of calcium che- idiotype (anti-Id) to modulate immunization to αIIbβ3. Along
lators, such as EDTA, heparin, or other drugs.84 Some of this these lines, it has been reported that IVIG, which is routinely
autoantibody specificity may be directed to neoantigens and used to reverse acute thrombocytopenia in ITP, may contain
thus is enhanced following changes in membrane glycopro- anti-Id directed to idiotypes of autoantibody but not allo-
teins induced by agents such as EDTA or heparin. antibodies that recognize αIIbβ3.92 Other investigators have
defined a DM idiotype that is characteristic of human auto-
Human Monoclonal Autoantibodies antibodies that are specific for the GPIb heavy chain.93 The
One important question that needs to be resolved is the latter study clearly suggests that the repertoire of idiotypes
extent of the autoepitope repertoire on a given platelet glyco- expressed by human autoantibodies specific for membrane
protein antigen. The answer will have an impact on the fea- glycoproteins, such as those of the human platelet, will be
sibility of developing therapeutic and diagnostic measures narrowly defined and, thus, amenable to study.
based on epitope specificity. Because the autoantibodies that The production of human monoclonal autoantibodies
react with a given epitope are likely to share idiotypes, one or recombinant Fab fragments from patients with immune
can approach this question from two directions, analyzing the thrombocytopenia provides researchers with specific anti-
epitope repertoire or the idiotype repertoire. body probes to map target antigens. Human monoclonal
Two studies have addressed the extent of the autoantigen antibodies have an advantage over serum antibodies in that
repertoire on αIIbβ3 by analyzing the competitive binding they are monoclonal and circumvent problems associated
between human autoantibodies and murine monoclonal with using a polyclonal serum where many different, low
antibodies.88,89 However, the limited number of studies that titer autoantibodies may be found with reactivity against a
have used this approach have generated conflicting results, variety of epitopes.
and insufficient data are available to judge the size of the
GPIb/IX Complex
autoepitope repertoire on αIIbβ3. Additional analyses aimed
at epitope localization will be necessary, and perhaps novel Unlike the αIIbβ3 complex, autoantibodies directed against
approaches will expedite this task. One such novel approach GPIb are almost universally associated with pathologic
is the development of human monoclonal autoantibodies immune destruction of platelets. Thus far, no one has described
specific for αIIbβ3 and other platelet glycoproteins. naturally occurring antibodies reacting with GPIb in either
Human monoclonal antibodies are an alternative tool resting or activated platelets. Although only 15% to 30% of
in the search for glycoprotein epitopes that are autoimmu- ITP patients have autoantibodies that bind to GPIb, there is
BLOOD BANKING no background binding of antibodies in normal plasma so the a unique case of autoimmune platelet dysfunction following
characterization of this antigen is much more straightforward. myasthenia gravis.100 This autoantibody inhibited aggrega-
Autoantibodies to components of the GPIb complex are also tion of normal platelets induced by collagen or wheat germ
frequently encountered in adult chronic ITP. agglutinin. This was the first case wherein autoantibodies to
He and colleagues94 have made progress in the localiza- α2 were associated with a chronic hemorrhagic disorder.
tion of selected autoantigenic epitopes on the GPIb mole- Efforts to map the specific epitope on this receptor have
cule. Epitopes were most frequently found on a recombinant been unsuccessful even though a number of groups have sug-
fragment of Ibα corresponding to residues 240–485, and gested that the binding of these antibodies, since they impair
next most often on a fragment representing residues 1–247. function, might be near the binding site of collagen. Deckmyn
In the case of those antibodies reactive with the former and colleagues describe a plasma autoantibody directed
sequence, further epitope mapping identified the dominant against a protein comigrating with α2 and recognized by the
determinant as the 9-amino acid sequence TKEQTTFPP patient’s antibody when affinity-purified α2β1 was used as
(residues 333–341).94 In some cases in which autoantibody antigen.100 The α2β1 complex was immunoprecipitated from
to GPIb/IX was detected, the clinical presentation proved to a platelet lysate by the patient’s plasma, and purified platelet
be particularly severe and refractory to therapy.93 One case specific IgG from this patient inhibited aggregation of normal
of “pseudo-Bernard-Soulier syndrome” (dysfunction of the platelets induced by collagen or by wheat germ agglutinin.
Ib-IX-V receptor complex) was reported to be caused by an Dromigny and coworkers101 described a 48-year-old woman
autoantibody to GPIb.95 Finally, in a subset of childhood in whom they found increased bleeding times, impaired col-
ITP, that associated with Varicella zoster infection, the GPV lagen-induced platelet aggregation, and the presence of auto-
component of this receptor was found to be the dominant antibodies directed against α2β1 and the GPIb complex.
target of serum autoantibodies that do not crossreact with
GPVI
viral antigens.96 At the same time, in another group of chil-
dren with this disease, it was found that serum antibodies Sugiyama and coworkers102 and Moroi and coworkers103
specific for viral antigens can crossreact with normal platelet were the first to identify and characterize autoantibodies
antigens and may thus contribute to platelet clearance.97 from an ITP patient specific for GPVI, a 62-kDa membrane
The GPIbα heavy chain is a large molecule and contains glycoprotein present on both human and murine platelets
leucine-rich repeats and a region in the middle of the glyco- noncovalently associated with a coreceptor FcγR chain.
protein that is heavily glycosylated. The NH2-terminal region More recently, Boylan and colleagues104 described a patient
of GPIb is responsible for binding to von Willebrand antigen, with a mild bleeding disorder and a moderately reduced
mediating platelet adhesion. Theoretically, there could be a platelet count whose platelets were nonresponsive to colla-
variety of antigenic sites on this molecule. However, using gen, although no abnormalities were found in the genomic
anti-Id reagents, it appears that the repertoire of idiotypes DNA sequence of GP6 and the level of platelet GP6 mRNA
expressed by anti-GPIb antibodies from patients with ITP is was normal. The plasma of this patient was found to con-
II very narrowly defined.93 tain an autoantibody that binds specifically to GPVI. Because
shedding of GPVI on mouse platelets in vivo by infusion of
GPV
120 specific anti-GPVI antibody had already been described,105 it
In the mid-1980s, when screening for antigen-specific was proposed that in this patient the GPVI/FcγR complex is
autoantibodies became available, many researchers hoped that cleared from the platelet by this autoantibody, resulting in a
the patterns of target antigen reactivity might be predictive of chronic “acquired” deficiency of platelet GPVI.
the clinical course of ITP (i.e., acute versus chronic). In fact, This case represents a unique effect of platelet-specific
an initial report looking at eight children with acute ITP using autoantibodies and emphasizes that acquired defects in
immunoblot technique failed to demonstrate anti-β3 antibody platelet function can also be a pathologic outcome of auto-
even though this was the dominant antigenic target recognized immunity to platelets.
by autoantibodies from patients with chronic ITP.98 This was
Glycolipids as Autoantigens
followed by a report from the same laboratory99 demonstrat-
ing serum antibody reactivity to Glycoprotein V (GPV) in four A number of reports have implicated cardiolipin, lactosyl-
pediatric patients with thrombocytopenia associated with ceramide, and other glycosphingolipids (GSL) as autoan-
varicella (chickenpox) infection. GPV is a thrombin-sensitive, tigenic targets (Table 9–4).106,107 Van Vliet and colleagues107
85-kDa glycoprotein of unknown function on the platelet sur- analyzed the binding of serum IgG/IgM antibodies from 30
face. There was no evidence of immune complexes in these ITP patients to platelet glycosphingolipids separated by high-
patients or varicella antigens adsorbed to the surface of their performance thin-layer chromatography. Acidic GSL, namely
platelets. As all of the children with varicella-related ITP and sulfatides and gangliosides, were identified as the major
acute childhood form of ITP resolved their thrombocytope- targets of serum autoantibodies. Thirteen of the 30 sera, five
nia within 6 months, there was some hope that GPV reactivity with anticardiolipin antibodies, had antibodies that bound to
would be a marker for mild disease and these patients could
escape aggressive immunosuppressive therapy. Table 9–4 Glycolipid Antigens of Platelets
Integrin a2b1
Cardiolipin
Platelet integrin α2β1 is an important collagen receptor on Lactosylceramide
the platelet membrane. Although present in a much lower Glycosphingolipids (GSL)
concentration than αIIbβ3 or the GPIb complex, this recep- Acidic: sulfatides/gangliosides
Monogalactosyl sulfatide (16/6 idiotype)
tor has also been shown to be a target autoantigen in 10% to Neutral: globotriosyl ceramide
20% of chronic ITP patients.100 Moreover, serum IgG auto- Globotetraosyl ceramide
antibodies specific for integrin subunit α2 were identified in
HUMAN PLATELET ANTIGENS
sulfatides; four sera showed antibody binding to gangliosides.
inherent in the immune system that optimizes our response to
Koerner and colleagues106 employed a more efficient phase
external organisms and safeguards us against autoimmune dis-
partition separation of acidic GSL from neutral GSL and were
ease, one wonders why antibodies to platelet antigens appear
able to demonstrate that serum antibodies specific for neutral
to be preserved in our repertoire of immunoglobulins.
GSL were more characteristic of ITP. Two classes of GSL auto-
Kazatchkine and coworkers109 were among the first to
antigens were defined: those associated with general autoim-
demonstrate the presence of antiplatelet antibodies in normal
munity and detected in the sera of patients with either SLE
human serum and anti-Id antibodies in IVIG, suggesting that
or ITP, and those peculiar to platelet-specific autoimmunity
immunoglobulin reactive with common platelet antigens, such
and detected only in the sera of ITP patients. Two GSL forms
as the GPIb/IX complex and αIIbβ3, are critical components of
belong to the platelet-specific group, but they are present at
the naturally occurring autoantibody repertoire.109–115 These
minute levels, and further characterization awaits large-scale
observations, along with a growing body of evidence, suggest
purification. One half (6/12) of patients with ITP had serum
that antiplatelet antibodies are produced as a result of epit-
IgG or IgM antibodies that bound these platelet-specific GSL.
ope shift116,117 after exposure to certain infectious organisms
Sera from none of 10 patients with nonimmune thrombocy-
or external antigens. This response must be tightly controlled
topenia, none of 10 patients with SLE, and only 1 of 18 nor-
with a network of regulatory cells and plasma factors, such the
mal subjects gave positive reactions with the platelet-specific
anti-Id antibodies found in IVIG.115,118,119 In childhood ITP,
GSL group. The general GSL antigen group includes globotri-
which is seasonal and transient in nature, the regulatory net-
aosyl ceramide, globotetraosyl ceramide, and a third uniden-
work can usually restore itself to proper order within 6 months
tified neutral GSL. Antigens in the general group were bound
following the acute insult in a majority of patients.57
by IgG or IgM antibodies in the sera of 10 of 10 patients with
In animal models, loss of peripheral tolerance to self-anti-
SLE, 8 of 12 patients with ITP, and none of 10 patients with
gens is emerging as a possible mechanism for autoimmune
nonimmune thrombocytopenia or 18 control subjects. These
disease, with particular focus on the regulatory CD4+25+
findings provide compelling support for a role of neutral GSL
T-lymphocyte subset.120–122 These cells prevent autoimmune
as antigenic targets in selected cases of ITP.
disease by modulating self-reactive T cells and secretion of
suppressor cytokines, such as interleukin-10 (IL-10) and
Animal Models of ITP transforming growth factor beta (TGF-β).123,124 As summa-
rized in the following section, the finding of inherent T-cell
Many investigators have searched in animals for clinical cor-
abnormalities and alterations in the expression of accessory
relates of human ITP. These attempts have been largely unsuc-
molecules on autoreactive T cells provides a novel avenue for
cessful, but an exception may have recently evolved from
the development of immunotherapy in ITP.119,125
concerted studies in dogs. As pointed out in a recent series
of reports by Lewis and Meyers,108 there is now substantial
T Lymphocytes
evidence that ITP in dogs and ITP in humans are clinically
analogous syndromes. In 32 cases of canine ITP, increased As in other autoimmune diseases, elevated numbers of T-cell 9
platelet-bound IgG was detected in a majority of the animals receptor (TCR) γ/δ-positive T lymphocytes have been noted
(30 of 32), and immunoglobulin eluted from the platelets of in both acute and chronic ITP patients, as initially described 121
11 of 19 affected dogs bound to homologous normal canine by Ware and Howard.126 Analysis of the nucleotide sequences
platelets.108 Furthermore, immunoglobulin specific for integ- used by these TCR γ/δ T cells demonstrated a diverse set of
rin subunits αIIb and/or β3 was detected in sera of 4 of 17 dogs VDJC gene rearrangements, characteristic of a superantigen
tested. They have concluded that, in canine ITP, immuno- response. Further evidence for T-cell reactivity in ITP came
globulin bound to the surface of platelets is directed against from the observations that isolated T-cell clones showed in
host antigens and that the target antigen is frequently αIIbβ3. vitro proliferation against allogeneic platelets, the numbers of
Given the similarities between these findings and the cumula- Vβ8+ T cells were elevated, and platelet stimulation resulted
tive experience with human ITP, the canine model may prove in measurable IL-2 secretion.126 These results provided the
to be a valuable tool to further understand the pathogenesis first evidence that patients with ITP have platelet-reactive T
of the autoimmune response to platelet antigens. lymphocytes identifiable at the clonal level.
Recent attempts to create a murine model of ITP have pro- T-cell proliferative responses to platelet membrane gly-
duced initial results that are promising for researchers interested coproteins were examined in 21 patients with chronic ITP
in studying the pathophysiology of immune-mediated throm- or SLE, with or without thrombocytopenia, and 10 healthy
bocytopenias. Monoclonal antibodies of different IgG subclasses donors.127 T cells from all subjects failed to respond to deter-
directed against mouse αIIbβ3, β3, GPIba, GPIb/IX, GPV, and gent-soluble native glycoproteins, but tryptic peptides of
CD31 were generated. When injected into mice, monoclonal integrin αIIbβ3 stimulated T cells from nearly all subjects,
antibodies against GPIb/IX, GPV, CD31, and linear epitopes on including the normal individuals. Similar T-cell proliferation
β3 had mild and transient effects on platelet counts and induced in healthy individuals in response to αIIbβ3 had been described
no spontaneous bleeding. On the other hand, anti-GPIba mono- previously by Filion and colleagues.128 These findings implied
clonal antibodies induced profound irreversible thrombocyto- that some autoreactive T cells, directed against membrane
penia (to <3% of normal) by an Fc-independent mechanism.65 antigens present on platelets, were not necessarily eliminated
by intrathymic deletion but are present in a suppressed state
in normal individuals. In the ITP patients, the T-cell response
Immunologic Abnormalities in ITP
was restricted by HLA-DR, the responding T cells had a CD4+
Increasing insights into the etiology and pathophysiology of phenotype, and the proliferation was accelerated relative to
ITP have stimulated an examination of immune abnormalities normal subjects, suggesting in vivo T-cell activation.128 None
that distinguishes adult from childhood immune thrombocy- of the peripheral blood mononuclear cell culture supernatants
topenia. Given the elegant and extensive regulatory network from healthy donors contained a significant amount of IgG
BLOOD BANKING
anti-αIIbβ3 antibody, but all showed trypsin-digested αIIbβ3- following therapeutic intervention with the most common
induced T-cell proliferation. These findings led to the conclu- drugs, IVIG and anti-D immunoglobulin.57,132,141
sion that CD4+ HLA-DR-restricted T cells reactive with αIIbβ3 Bussel and coworkers141 compared the changes in IL-6, IL-
epitopes are involved in production of antiplatelet autoanti- 10, MCP-1, and TNF-α cytokine levels after treatment with
body and are implicated in the pathogenesis of chronic ITP. IVIG and anti-D in thrombocytopenic adults. There was an
T-cell activation, proliferation, and T-B cell cognate recog- increase in all but TNF-α within 2 hours after administration
nition in response to specific autoantigens require signaling of anti-D. After IVIG, there was a significant increase in IL-10
through a variety of membrane immune receptors. The MHC levels within 4 hours (but not 2 hours) post-therapy. It was
and TCR were recognized as major receptors for many years, concluded that the early increase in these macrophage-syn-
but it is now understood that there are accessory molecules that thesized cytokines following anti-D administration reflects the
facilitate or modulate the response and that these are potential substantial interaction of antibody-coated erythrocytes with
therapeutic targets in treating ITP.119,125,129 Antigen-present- macrophages.132 Mouzaki and coworkers measured a panel of
ing cells, which include B cells as well as macrophages, require Th0-, Th1-, and Th2-related cytokines over 0.5 to 5 years fol-
specific recognition and signaling through the receptors B7, lowing the onset of ITP and found that patients who did not
LFA-3, and ICAM-1, whereas the T cell employs the receptors convert from a Th1- to Th2-dominant cytokine profile were
CD28, LFA-1, and CD2.125,130 Increased expression of CD40L more likely to have a chronic or relapsing course of thrombo-
on platelets may be a normal response of platelets activated cytopenia.57 Interestingly, the same group of researchers found
by infection or other inflammatory mediators. However, this that adults with ITP were likely to have a striking polariza-
increase in CD40L has the potential to drive B-cell production tion toward Th1 phenotype, but this profile did not change in
of platelet autoantibody. Two research groups reported that response to disease status over time or with treatment.135 This is
administration of a humanized anti-CD40L antibody resulted in contrast to another group who found an association between
in increased T-cell tolerance in patients with ITP, suggesting improved platelet count and disease outcome in adults with a
that blockade of CD40L may prove to be a useful therapeu- lower Th1/Th2 ratio.134 Again, these differences may arise from
tic approach.125,131 Likewise, Zimmerman and colleagues132 disparities in population, differences in timing and technology
reported that IVIG and dexamethasone induced an alteration of assays, and differences in treatment regimens, which vary
of T-lymphocyte subsets and suppression of T-lymphocyte greatly between centers. A better understanding of the rel-
proliferation in vitro in ITP patients. Although equally effective evance of Th1/Th2 ratio and cytokine profiles will emerge with
in the treatment of ITP in children, they found that high-titer the establishment of coordinated clinical trials and the incor-
anti-D immunoglobulin caused significantly less inhibition poration into such trials of the newly recognized components
than IVIG or dexamethasone in children with chronic ITP, of the cellular immune response, such as CD4+25+ (Treg) cells
and anti-D did not affect T-lymphocyte subsets. and the antigen-specific idiotypic network.
Cytokines
II DRUG-INDUCED IMMUNE
Considerable work has been focused on the measurement
of cytokines in ITP in an effort to further characterize the THROMBOCYTOPENIA (QUININE/
122 influence of T-helper subsets (Th0, Th1, and Th2) in the QUINIDINE PURPURA)
autoimmune etiology of this disease.130,133 Unfortunately, the
majority of these studies draw conclusions based on a single Although drug-induced immune thrombocytopenia (DITP)
cytokine profile determination. Because cytokine signaling may be a complication of therapy employing a variety of
is dynamic and may fluctuate rapidly, it is difficult to draw drugs, it is most frequently seen in the United States with the
valid conclusions from a single assay. In addition, the pres- administration of quinine and quinidine.142 It has been pro-
ence of soluble cytokine receptors may confound the results posed that the following criteria be met before an individual
if these are not measured directly in parallel with cytokines. can be considered to have DITP: the patient is not throm-
To distinguish a normal versus pathologic immune response bocytopenic before administration of the drug; thrombocy-
in both acute and chronic ITP, multiple samples must be topenia follows drug ingestion and begins to reverse shortly
obtained over time, with attention paid to both the cytokines after cessation of drug; thrombocytopenia does not recur
and their soluble receptors. The referenced studies have after cessation of drug treatment; and all other causes of
incorporated multiple samples and controls in the overall thrombocytopenia are ruled out.143
analysis. Nonetheless, a degree of disparity remains between Cumulative evidence now favors a mechanism whereby
the results obtained in children57 versus those found in the drug induces the expression of a neoantigen on the
adults,134,135 which likely reflects a difference in the primary platelet surface144 that is recognized by circulating antibod-
pathophysiology of the thrombocytopenia. ies only in the presence of the drug. The observation that
In pediatric ITP patients, a Th1 type of cytokine response platelets from BSS patients (lacking Ib-IX-V) failed to lyse in
is the customary finding, characterized by very low IL-4 and the presence of drug-dependent antibody, specific drug, and
IL-6, but elevated levels of IL-2, interferon-γ, and TNF-β.136 complement, was the first indication that a specific platelet
In adults with chronic ITP, malignancy associated with ITP, antigen is recognized by such antibodies.145 This finding led
or the autoimmune lymphoproliferation syndrome associ- other laboratories to confirm that purified GPIb/IX would
ated with defects in the Fas apoptosis pathway, there are ele- compete for drug plus antibody and was therefore likely to
vated levels of IL-10, IL-11, IL-6, and IL-13.137–140 One would contain the antigenic epitope in question. Evidence of direct
expect that successful therapy might be associated with favor- binding of such antibodies to GPIb/IX was first provided by
able changes in cytokine profiles. Although many treatment Chong and coworkers,146 and Berndt and coworkers147 estab-
regimens have been efficacious in ITP, in only a few instances lished that the complex of both Ib and IX is likely required
has there been documentation of cytokine profile changes for maximum antigen expression.
Chong and coworkers148 showed that one quinine-depen-
David F. Stroncek
Centromere
Class II loci Class III loci Class I loci
DP DN DM DO DQ DR 21OH Bf Hsp70 B C E J A H G F
C4 C2 TNF
Expanded class II
region
DP DN DM DO DQ DR
B2 A2 B1 A1 A A B B B2 A2 B3 B1 A1 B1B2B3B4B5 ψ A
TAP1 TAP2
LMP2 LMP7
0 500 1,000
KBase pairs
Figure 10–1 Physical map of the HLA genetic complex, illustrating the clusters of genes according to the class of encoded gene products. The
symbol ψ represents four DRB pseudogenes, designated 7, 8, and 9. Other pseudogenes are shown in gray. (From Hoffman R, et al. Hematology:
Basic Principles and Practice, 4th ed. Philadelphia, Churchill Livingstone, 2005.)
II
130 are made up of a heavy and light chain. The genes encod- Class I
ing the HLA class I heavy chain and both the HLA class II 5⬘UT,L α1 α2 α3 TM
CY
3⬘UT
heavy and light chains are located in the MHC region. All
HLA class I molecules share the same light chain, β2-micro-
globulin, which is encoded outside the MHC on chromo-
some 15.7
Three separate genes encode the class I HLA-A, HLA-B,
and HLA-C heavy chains, HLA-A, HLA-B, and HLA-C, and
the three have a similar structure. Each contains eight exons
encoding for the α1, α2, and α3 extracytoplasmic domains, a
transmembrane domain, and a cytoplasmic tail (Fig. 10–2). Class II TM,CY,3⬘UT
The first exon of the class I α chain encodes a leader sequence. α 5⬘UT,L α1 α2 3⬘UT
Exons 2 through 4 are highly polymorphic and encode extra-
cellular domains, α1, α2, and α3, that are responsible for pep-
tide binding and T-cell receptor (TCR) engagement. Exon 5
encodes the transmembrane region, exons 6 and 7 the cyto-
plasmic tail, and exon 8 the 3′ untranslated region.
The HLA class II molecules are heterodimers composed
TM
of an α and β chain of approximately the same size. The
β
genes encoding the class II α and β chains have a similar 5⬘UT,L β1 β2 CY CY,3⬘UT
structure, but the α chain has 5 exons and the β chain has 6
exons. For both the α and β chain genes exon 1 encodes the
leader peptide, and exons 2 and 3 encode the two extracellu- Figure 10–2 The organization of class I and II MHC genes. 5′UT and
lar domains. For the β chain gene exon 4 encodes the trans- 3′UT are the untranslated regions in the 5′ and 3′ ends of the gene.
membrane domain, exon 5 the ctyoplasmic tail, and exon 6 L is the leader sequence, TM the exons encoding the transmembrane
domains, and CY the exons encoding the cytoplasmic tails. (From
the 3′ untranslated region. For the gene encoding the α chain Germain RN, Malissen B. Analysis of the expression and function of
exon 4 encodes the transmembrane domain and the cyto- class-II major histocompatibility complex-encoded molecules by DNA-
plasmic tail and exon 5 encodes the 3′ untranslated region. mediated gene transfer. Annu Rev Immunol 1986;4:281–315.)
HUMAN LEUKOCYTE AND GRANULOCYTE ANTIGENS AND ANTIBODIES
HLA-DRB1 HLA-DRB3,4,5
specificities specificities
DR1,
HLA-DRB1 HLA-DR6 HLA-DRB5 HLA-DRB9 HLA-DRA DR51
DR15,16
DR11,12
DR13,14 HLA-DRB1 HLA-DR2 HLA-DRB3 HLA-DRB9 HLA-DRA DR52
DR17,18
There are 5 isotypes of the class II molecules, which are according to mendelian principles. Because each individual
designated as HLA-DM, HLA-DO, HLA-DP, HLA-DQ, and inherits two of four possible haplotypes (two from each par-
HLA-DR. Within the class II region of the MHC, the α and β ent), there are four possible genotypes for each individual
genes encoding most of the isotypes are organized as pairs of and the probability of genotypic identity between two sib-
α and β chains that contribute to the same isotype. The genes lings is 25% (Fig. 10–4). Because HLA genes are inherited as
encoding the α chains are designated as A and those encod- haplotypes, most HLA phenotypically identical siblings are
ing the β chain B. The HLA-DM molecules are the product also HLA genotypically identical. In 2% of cases, recombi-
of HLA-DMA and HLA-DMB, HLA-DO of HLA-DOA and nant HLA haplotypes (a set of genes derived partially from
HLA-DOB, HLA-DP of HLA-DPA1 and HLA-DPB1, and two chromosomes through recombination) yield genotypes
HLA-DQ of DQA1 and DQB1 genes, respectively. that deviate from this rule.
There are several HLA class II pseudogenes that are not In large populations gene frequencies achieve equilibrium
10
expressed. There is one HLA-DPA pseudogene, HLA-DPA2, within a few generations unless selective pressure influences
and one HLA-DPB pseudogene, HLA-DPB2. There is also individuals’ survival and mating capacity (Hardy-Weinberg
one HLA-DQA pseudogene, HLA-DQA2, and two HLA- law). In equilibrium, gene prevalence is maintained based 131
DQB pseudogenes, HLA-DQB2 and HLA-DQB3. solely on its frequency. However, within specific populations
The expression of HLA-DR molecules is more compli- the presence of specific combinations of HLA antigens occurs
cated than the expression of the other HLA class II mole- with a higher frequency than expected from the prevalence of
cules. All HLA-DR α chains are encoded by the same DRA
gene, HLA-DRA1. However, there are several functional
HLA-DRB genes that encode the HLA-DR β chains and
several HLA-DRB pseudogenes. In addition, the number Father Mother
of HLA-DRB genes and pseudogenes present varies among
HLA Haplotype C
HLA Haplotype D
HLA Haplotype A
HLA Haplotype B
HLA Haplotype C
HLA Haplotype D
HLA Haplotype D
HLA Haplotype A
HLA Haplotype B
HLA Haplotype A
HLA Haplotype B
α2 α1 β1 α1
α3 β2m β2 α2
Figure 10–6 Schematic diagram of HLA class I and class II molecules, pointing to their structural similarity. In both HLA classes two immuno-
globulin-type domains reside close to the cell membrane (α3 and β2 for class I and α2 and β2 for class II). The other two domains project toward the
extracellular milieu with α-helices (α1 and α2 for class I and α1 and β1 for class II) and a platform of parallel β-sheets that form a peptide-binding
groove. (From Hoffman R, et al. Hematology: Basic Principles and Practice, 4th ed. Philadelphia, Churchill Livingstone, 2005.)
chain with β2-microglobulin.16,27 In addition to using algo-
first digit corresponds to the number of Ig-like domains in cells; therefore, they can preferentially expand in the pres-
the molecule and a “D” denotes domain. The D is followed by ence of allogeneic tissue.
either an “L” for long cytoplasmic tail, S for short cytoplasmic NK cells also express activating receptors that are respon-
tail, or “p” for a pseudogene. The last digit indicates the number sible for their lytic activity. These activating receptors include
of the KIR gene.92 KIRs with short cytoplasmic tails. Although the ligands for
Expression of KIR in individual NK cells is complex these activating receptors have not been identified, it is
because NK cells may express several members of the KIR possible that they are expressed primarily by activated or
family. The number of KIR genes in each haplotype varies proliferating cells. It is, therefore, possible that during the
among individuals. The most common haplotype, known as inflammatory process induced in allogeneic conditions, nor-
group A, is made up of 6 genes (2DL1, 2DL2 or 2DL3, 3DL1, mal cells may become activated by cytokines and express
3DL2, 2DS4, and 2DL4). Various KIR genes can recognize ligands responsible for NK activation in the absence of HLA
different HLA-A, HLA-B, and HLA-C molecules. HLA-C class I molecules reactive with the inhibitory receptors.89
antigens can be divided into two groups based on polymor- The relevance of KIR in haploidentical hematopoietic
phisms at amino acid positions 77 and 80 of their class I transplantation has been well studied. In haploidentical
II heavy chains. One group has asparagine (Asn) at position transplants several KIR–HLA class I allele combinations are
77 and lysine (Lys) at 80; the other has serine (Ser) at 77 and possible, including NK cells from the graft that express KIR
136 Asn at 80. Some KIRs recognize HLA-C antigens with Asn77 but do not interact with the donor’s HLA molecules (graft-
and Lys80; other KIRs recognize HLA-C antigens with Ser77 versus-host alloreactivity) and NK cells from the donor that
and Lys80. The polymorphism at position 80 is most impor- do not interact with the host’s HLA molecules (graft rejection
tant. Another group of KIRs reacts with HLA-B antigens that alloreactivity). The presence of graft-versus-host reactive NK
carry specific combinations of amino acids at positions 77 cells due to incompatibilities between donor and recipient
and 83 of the heavy chain that form HLA-Bw4. KIR recogni- (especially HLA-Cw families) has favorable effects in the
tion of HLA class I molecules is degenerate because a single outcome of acute myeloid leukemia.94 Alternatively, a good
KIR can interact with multiple HLA class I alleles. Another match may be present between graft NK cells and host’s HLA
NK inhibitory receptor (CD94-NKG2A) recognizes the non- and between host’s NK cells and graft HLA. In such a case no
classical HLA-E molecule. In addition, each of the KIR genes alloreactivity will occur. If the stem cell donor is incompat-
is extensively polymorphic. ible with the host’s NK cell repertoire, the host’s reactivity
Because the genes for KIR, HLA, and CD94-NKG2A may lead to graft rejection. Interestingly, alloreactive grafted
are located in separate chromosomes, they segregate NK cells seem to prevent GVHD while inducing GVN.89
independently; consequently, individuals may carry genes Like HLA, KIR genes are polymorphic, and their vari-
for KIRs for which there is no correspondent HLA ligand.93 ability is clustered in positions likely to impact the overall
Because HLA-E is expressed in all individuals, NK cells that structure of the molecule. The relevance of KIR gene poly-
bear the CD94-NKG2A receptor are not alloreactive. Because morphism in the outcome of bone marrow transplantation
the specificity of KIRs for their ligands is broad and each (BMT) is unclear. It appears that the risk of GVHD is high-
individual carries several KIRs, it is likely that in the majority est in the context of unrelated BMT when the recipient KIR
of cases all NK cells of a given person express at least one KIR genotype is “included” in the donor KIR genotype.95 These
that is specific for a self-HLA class I allele. Thus, in autolo- results show that compatibility between KIR genotypes
gous settings NK cells kill only aberrant cells that have lost themselves may influence the outcome of BMT.
HLA class I expression. In contrast, NK cells can kill alloge- Minor histocompatibility antigens (mHA) are polymor-
neic cells that do not express HLA class I alleles recognized phic molecules whose peptides contain variant sequences
by their KIR. Thus, by knowing the KIR repertoire of a given that are presented by HLA alleles. They have been shown
transplant recipient and the HLA type of the donor, it is the- to be targets of cytotoxic T lymphocytes that can lyse
oretically possible to predict the likelihood of an NK-medi- leukemia cells.96 In addition, some mHA are selectively
ated alloreaction. Importantly, it appears that “alloreactive” expressed by neoplastic cells.97 At present little is known
NK cells undergo proliferation on exposure to stimulatory about the identity of mHA epitopes in the context of
lar nomenclature for that allele.102 The 10th International
HLA phenotyping has also been applied to identify links required before enrollment into immunization protocols25
between a given disease and the genetic makeup of its car- because the response to the vaccine may be HLA-restricted.
riers.106 Strong associations are exemplified by birdshot Analysis of specific HLA–epitope combinations for immu-
uveitis, a disease occurring exclusively in HLA-A29 individu- nization protocols requires high-resolution typing because
als107; type I diabetes and other autoimmune diseases108–110; allelic variations may affect antigen presentation and the
or long-term survival of HIV-infected individuals.54 These immune response to vaccination.23,113
studies evaluated the role that genetic background may have Traditional serologic assessment of HLA antibodies and
contributed over environmental factors.106,111 HLA associa- antigens takes advantage of fetomaternal sensitization. In
tions are thought to be due to differential ability of distinct mammals, the progeny carries a full haplotype of paternal
alleles to present immunogenic epitopes107,108 or to the close origin, and pregnant women may develop antibodies against
linkage of HLA class I and II antigens to the HLA class III antigens expressed by the paternal haplotype. Maternal
region where potent immune modulators such as TNF-α are sera are collected at term and characterized by testing their
located.112 ability to kill HLA-bearing cell lines of known phenotype
HLA phenotype associations have also been suggested as in the presence of complement (complement-dependent
predictors of immune responsiveness of cancer to immune cytotoxicity, or CDC).114 CDC is then used for HLA typing
therapy.59 However, such associations have remained diffi- by exposing circulating cells expressing HLA class I (most
cult to reproduce. Most recently, patient HLA typing is being cells) and class II (predominantly B cells) antigens from the
HUMAN LEUKOCYTE AND GRANULOCYTE ANTIGENS AND ANTIBODIES
individual to be typed to previously characterized alloan- has become increasingly complex and burdensome. In addi-
tisera or monoclonal antibodies.115 tion, due to the important role that HLA molecules play in
Traditionally, when sera from sensitized patients are antigen presentation and the stringency of the relationship
screened for antibodies directed to HLA antigens, a HLA- between epitope and associated HLA allele, high-resolu-
phenotyped repository of cell lines are used in a CDC assay tion typing is increasingly requested for appropriate enroll-
to identify alloantibodies. The fraction of cell lines killed by ment of patients into immunization protocols aimed at the
the sera serves as a rough grade or indicator of the intensity of enhancement of T-cell responses. Therefore, high-resolution
allosensitization and is referred to as panel-reactive antibody HLA typing is increasingly in demand in clinical and experi-
(PRA) reactivity. mental settings.
Some antibodies activate complement and kill with poor Although oligonucleotide-based methods could theo-
efficiency: the cytotoxicity-negative-absorption-positive retically discriminate any known polymorphic site, they
(CYNAP) phenomenon.116 CYNAP may cause the underes- have two major limitations. First, they require a specific
timated allosensitization judged by cytotoxicity testing. By PCR for each allele investigated. Because each individual
modifying CDC with the addition of antihuman antibodies has only two alleles for each locus a disproportionately
capable of activating complement, CYNAP can be circum- large number of PCRs need to be performed to cover all
vented (augmented CDC). However, with this assay relatively possible polymorphisms in order to identify the two alleles
innocuous antibodies may cause overestimation of clinically in the individual tested. In addition, because both oligo-
relevant allosensitization.117 nucleotide-based methods are based on specific interac-
Other methods identify alloantibodies, including the tions with known oligonucleotide sequences unique to a
immobilization of HLA molecules on solid surface to cap- particular allele, they cannot identify unknown polymor-
ture soluble antibodies118–120 and a flow cytometry method phisms unless, by chance, the variation occurs within the
that uses a spectrum of microbeads coated with HLA mole- region spanned by one of the oligonucleotide primers or
cules of known type.121,122 An interlaboratory comparison of pairs used in the assay. Because of these limitations inter-
serum screening for HLA antibody determination suggested est is growing for definitive typing methods that yield
that enzyme-linked immunosorbent assay (ELISA) and flow conclusive information about the identity of the alleles
cytometry yield higher PRA activity values compared with typed. The most comprehensive method is sequence-based
CDC or augmented CDC.119 However, the study suggested a typing (SBT), however, its utilization has been limited by
lack of consistency among participant laboratories, leaving cost of equipment and reagents and by the high level of
unsolved which method most accurately defines clinically expertise and time required for the interpretation of each
relevant allosensitization; a panel of various methods may typing. In addition, some of the alleles and allele com-
be most informative. binations are difficult to resolve even with SBT.130 More
The CDC phenotyping assay is declining in interest in recently, high-throughput robotic SBT has been developed
the United States because most laboratories are switching to that allows sequencing of hundreds of genomic fragments
easier-to-handle and higher-resolution molecular methods. each day.131 Finally, new methods based on high-density 10
However, cellular immunologic methods remain valuable to array technology are being developed that may allow
characterize functional aspects of HLA molecules because extensive typing of known and unknown polymorphisms 139
molecular genotyping methods cannot define whether an on microchips.132
HLA allele is expressed.123 Thus, it is likely that immunologic High-resolution HLA genotyping methods yield high-
methods will continue to complement molecular methods resolution information of an individual’s HLA type. The
in the future.123 wealth of information is, however, counterbalanced by
The usefulness of conventional serologic typing has increased difficulty in identifying suitable HLA alleles dur-
been limited by the availability of allele-specific sera. Most ing donor–recipient pairing or accrual into immunization
importantly, because antibodies identify structural differ- protocols restricted to specific HLA–epitope combinations.
ences on the surface of HLA molecules, variations caused Thus, at present clinicians are faced with the daunting task of
by nucleotide polymorphism in areas that are not exposed, applying high-resolution typing results of unclear relevance
such as the peptide-binding groove of the HLA heavy chain, to clinical settings.133
are not detectable. However, the differences that are located
TESTING FOR ALLOSENSITIZATION AND DETERMINATION
in the peptide-binding groove are of functional significance
OF COMPATIBLE RECIPIENT–DONOR PAIRS
because they determine the specificity and affinity of peptide
binding124,125 and T-cell recognition of self and allogeneic Ideally, any cell containing a human product transfused or
target cells.24,126 transplanted between different individuals should be HLA
DNA-based typing methods directly determine the compatible. Yet, in most cases histocompatibility is not pro-
sequence of HLA molecules.127 Various polymerase chain spectively sought. Thus, patients who have received long-
reaction (PCR)-based methods have been described, among term therapies often become reactive to various antigens,
which sequence-specific primer (SSP)127,128 and sequence- including HLA. Transfusion of leukocyte-depleted products
specific oligonucleotide probe (SSOP)-based methods129 are has decreased the incidence of allosensitization, yet organ
the most universally used. The resolution of SSP and SSOP transplant candidates often develop alloreactivity before the
assays is limited only by the number of allele-specific primers transplant. To prevent rejection in organ transplant recipi-
or probes used to identify an ever growing number of alleles ents alloreactivity must be documented before transplanta-
(http://www.anthonynolan.com/HIG/index.htm). The real- tion. Alloreactive patients can still receive a transplant if the
ization of the richness of HLA polymorphism has led to a documentation provided states that the donor organ has no
proportional increase in the complexity of the assays utilized mismatched HLA antigens reactive with the patient’s anti-
to cover all possible alleles. As a consequence, accurate HLA bodies. Similarly, patients who have received repeated plate-
typing for donor and recipient matching in transplantation let transfusions may become allosensitized and consequently
BLOOD BANKING refractory to further transfusions unless platelets from HLA- the remaining 113 triplets have immunogenic potential.
compatible donors are used. With this algorithm it is possible to significantly broaden the
Obviously, the best compatibility among organ, hema- number of “molecularly matched” HLA alleles and therefore
topoietic progenitor cell, and platelet donors and recipients significantly increase the chances of identifying a compat-
is identical matching. However, particularly in the case of ible donor, particularly in those cases when the recipient has
rare HLA types, it is often impossible to identify a perfectly a rare HLA phenotype.
matched unrelated donor. Thus, other strategies have been In addition, if information on the specificity of HLA anti-
adopted to identify the best possible match or “compatible bodies in the recipient is known, HLAMatchmaker considers
mismatch.” Selection of compatible unrelated donor–recipi- triplets that are present in the panel of screening cells that
ent pairs is carried out through typing with serologic, cellular, give negative reactions with the recipient’s serum. These neg-
and molecular methods.134 With the increasing resolution of ative panel cells can be expected to share antigens with the
these typing methods, the chances of identifying compatible patients; however, other HLA antigens may be present and
donors based on full or partial HLA matching has become contain mismatched triplets that apparently are not immu-
increasingly low.133 To broaden compatibility-matching nogenic for that recipient. Such triplets are therefore accept-
criteria for donor–recipient pairs, matching systems have able and can be added to the algorithm for the identification
been developed based on shared public epitopes assigned of possible donors.
to crossreacting groups (CREGs)135 or shared amino acid Thus, HLAMatchmaker assesses HLA compatibility at a
polymorphisms defined through sequence information.136 molecular level by determining whether or not a triplet in a
Pre-existing alloantibodies further restrict the availability of given position of a mismatched HLA antigen is also found
compatible donors. Highly sensitized recipients with PRA in the same position in any of the recipient’s own HLA-
reactivity exceeding 85% of tested specificities represent a A, HLA-B, and HLA-C molecules. A shared triplet in the
particularly challenging group.137 An alternative approach to same position on a mismatched HLA antigen cannot elicit
the exclusion of alloreactive determinants is the inclusion of a specific antibody response in that patient. This hypoth-
“acceptable” antigen mismatches expressed in a panel of cells esis needs future testing as this strategy might represent a
that give negative reactions with the recipient sera,138 thereby revolutionary tool for the identification of potential donors.
extending the repertoire of possible donors. Unfortunately, Preliminary verification of the algorithm in a series of high
even these tools for the identification of unrelated, partially PRA renal patients suggested that this is a proper strategy, at
matched donor–recipient pairs often fail to identify a suit- least in highly sensitized renal transplant candidates waiting
able match. for kidneys from unrelated donors.139 HLAMatchmaker is
Recently Duquesnoy137 described a molecularly based also effective at selecting the best HLA-typed platelet com-
algorithm to identify histocompatible donor–recipient ponent to transfuse to alloimmunized thrombocytopenic
pairs called HLAMatchmaker. This method focuses on the patients.140
structural basis of HLA class I polymorphism so that com-
II patible HLA mismatches can be identified without extensive The HLA Molecules as Antigens
serum screening. This algorithm is based on the principle
that short amino acid sequences, triplets, characterizing and HLA Alloimmunization
140
polymorphic sites of the HLA molecules, are the critical Because of their high density on the surface of cells HLA
components of allosensitizing epitopes. Such amino acids molecules can become immunogenic when cells from one
reside in the α-helices and β-loops of the heavy chain. individual are exposed to another’s immune system. The
Because each HLA molecule expresses a characteristic string mechanism(s) leading to HLA allosensitization are believed
of these determinants, it is possible to characterize each mol- to follow two pathways. The first pathway is followed during
ecule according to the linear sequence of amino acid triplets most immune reactions and involves donor antigen uptake
present on its surface. Based on the reasonable assumption by donor antigen-presenting cells and presentation to recipi-
that none of the triplets present in the HLA repertoire of the ent lymphocytes (indirect pathway). In this case, the donor’s
recipient are self-immunogenic, it is possible through a pro- HLA molecules are processed into peptides through the
cess of “electronic recombination” to identify donors with exogenous pathway of antigen presentation and presented
HLA alleles different from the recipient’s but containing to recipient T cells as linear peptides.141,142 This pathway is
exclusively shared triplets. The selected HLA alleles will be believed to be responsible for the development of alloanti-
compatible because they do not contain any epitope absent bodies as well as helper T-cell responses, but its role in the
in the recipient. development of cytotoxic T-cell responses remains unclear.
In theory, a large number of triplets could occur if poly- Because this pathway depends on the presentation of donor
morphisms were randomly distributed. However, most HLA HLA molecules by the recipient’s HLA alleles, it may explain
molecules span conserved domains and only a total of 142 why the humoral response to HLA class I allodeterminants
different polymorphic triplets designate serologically defined correlates with the HLA phenotype of the recipient.143 The
HLA-A, HLA-B, and HLA-C antigens.137 Triplet polymor- indirect pathway of HLA allorecognition has been associated
phism can occur in 30 locations on HLA-A, 27 in HLA-B, with allograft rejection.144
and 19 in HLA-C chains. Because the HLAMatchmaker Because the function of HLA molecules is to present anti-
algorithm includes interlocus comparison, it is possible to genic determinants to T cells, it could be easily envisioned
accumulate the information into a single database. Among how minor changes in their structure could be misinter-
the 142 polymorphic triplets, 29 are polymorphic for one preted as antigenic epitopes. Thus, intact HLA molecules
class I locus but monomorphic for another. Such poly- residing on the surface of donor cells are a perfect target for T
morphic triplets cannot be immunogenic because they are cell–mediated allorecognition (direct pathway) by recipient
always present on the patient’s own HLA antigens, whereas cells either through direct cytotoxic effect of T cells against
HUMAN LEUKOCYTE AND GRANULOCYTE ANTIGENS AND ANTIBODIES
target cells or by the activation of helper T cells, through ents belong to the same CREG, alloreactivity is thought to
HLA class II engagement, which leads to stimulation of anti- be less likely to develop (Table 10–3). The predictive value
body-mediated immune responses.145 of CREG matching of donor–recipient pairs on transplant
Humoral alloresponses mediated most frequently by outcome or platelet transfusion results, however, remains
immunoglobulin M (IgM) are predominant when sensi- to be demonstrated.
tization involves infrequent allogeneic exposure because Not all subjects that have been exposed to HLA allo-
they require smaller amounts of antigenic material. T-cell antigens develop alloantibodies and, in fact, exposure to
responses become more predominant in the context of low doses of donor-specific HLA antigens through donor-
transplantation where the persistence of the continuous allo- specific blood transfusions may have a beneficial effect
geneic stimulation allows the expansion and sustenance of on the survival of organ allografts.146 Several hypotheses
alloreactive cytotoxic T cells. have been proposed for the capriciousness of allosensitiza-
Because antibodies and TCR have different requirements tion, including presence of regulatory immune responses
for engagement, epitopes recognized by T cells and antibod- or cytokine-mediated immune suppression. However, the
ies are different. Antibodies require interaction with a small mechanisms that modulate the quality and quantity of allo-
structure, including a limited number of amino acids; thus, immunity remain elusive; different aspects of this algorithm
any amino acid sequence combination on the surface of are discussed ad hoc in this chapter with particular attention
an HLA allele not present in the individual exposed to the to molecularly defined algorithms for the prediction of his-
alloreaction may represent an epitope. The TCR has a much tocompatibility.123,137,147,148
lower binding affinity for their ligand and binding requires HLA matching enhances the outcome of some types of
the complete interaction of the TCR with the antigenic pep- organ transplants. An analysis of more than 150,000 renal
tide as well as the α− and β−helices of the HLA class I heavy transplant recipients performed by centers participating
chain.12 Thus, although several B-cell epitopes recognized in the Collaborative Transplant Study showed that a com-
by antibodies can be identified in a given HLA molecule, plete mismatch (6 HLA-A+B+DR) had a 17% lower survival
generally the whole HLA molecule is necessary for T cell– expectation than no mismatch (p < 0.0001),149 and match-
dependent allorecognition. ing was particularly beneficial in patients with highly reac-
Two types of antibody-defined epitopes can be identified tive preformed alloantibodies. The same study suggested
according to their frequency among HLA alleles. Private that high-resolution matching based on molecular typing
epitopes are almost, but not totally, unique for each single improved graft survival. Similar results were observed in
serologically defined HLA antigen, and antibodies directed cardiac transplantation where HLA matching yielded signif-
to private epitopes are used for phenotyping. Private epit- icantly (p < 0.0001) better results. This is particularly impor-
opes are generally shared by all molecularly defined alleles tant because in most centers donor hearts are currently not
in a given family; for this reason fine differences among allocated according to HLA match. HLA matching was not
alleles within a general family cannot be distinguished beneficial in liver transplantation.150
serologically. Public epitopes are more widely distributed The induction of donor-specific hyporesponsiveness has 10
among serologically defined antigens and have been used been particularly well documented in the context of renal
to cluster distinct serologic families into groups. Public allotransplantation and may limit the need for immune 141
epitopes have an immune dominant character. Immune suppression. A recent randomized study suggested that
sera that identify public epitopes have been considered to pretransplant donor transfusions improved the survival of
be predictive of major CREGs. When donors and recipi- cadaver kidney grafts in patients receiving modern immune
*
North American white populations of European origin.
BLOOD BANKING suppressive regimens, although the mechanism remains occurrence of GVHD but is associated with an increased
unclear.151,152 Thus, although most centers presently do not risk of graft rejection and tumor/leukemia relapse.174 In
implement deliberate blood transfusions, the usefulness of fact, Weiden and colleagues175 observed that survivors of
this approach needs to be further investigated. severe acute GVHD had a significantly lower incidence of
Patients who have been previously exposed to HLA class tumor relapse compared with patients who did not expe-
I expressing heterologous cells and who have produced HLA rience GVHD. This association appeared mandatory and
antibodies often require platelet transfusions. Such patients it was felt that the beneficial GVN effect was inseparable
are refractory to random donor platelets and must be given from GVHD.
HLA-matched or “semi-matched” apheresis components.41–44 The risk of GVHD increases with genetic distancing
However, the provision of HLA-matched platelets does not between donor and recipient. Recipients of HPC trans-
always improve platelet recovery and survival. Possibly, the plantation from identical siblings have a lower chance of
ineffectiveness of some “HLA-matched” platelet transfusions developing GVHD compared with recipients of HPC trans-
is due, at least, in part to unrecognized HLA mismatches plants from HLA-matched unrelated donors and partial
between the donor and recipient resulting from low-reso- HLA-matched realated donors.176,177 However, although the
lution methods used for typing the donor and recipients. genetic closeness between donor and recipient appears to
Higher-resolution typing methods have been advocated, but decrease the risk of GVHD, it also decreases the therapeu-
it remains controversial whether molecularly based HLA typ- tic benefit of the GVN effect and increases the chances of
ing confers an advantage over serologic typing; this principle tumor relapse.
was recently questioned in the context of hematopoietic cell
Graft-versus-Neoplasia Effect
transplantation.153,154
It was first recognized that a beneficial collateral effect of
HLA as a Functional Mediator of Graft-versus- GVHD was the rejection of neoplastic cells by the donor
Host Disease and/or Graft-versus-Neoplasia immune system in the context of hematologic malignancies
Effect (graft-versus-leukemia effect).165 It was rapidly recognized that
Allogeneic or syngeneic hematopoietic progenitor cell (HPC) the graft-versus-leukemia effect could play a powerful thera-
transplantation is used to treat hematologic malignan- peutic role in the treatment of refractory malignant disorders,
cies155,157; hematologic disorders such as aplastic anemia,158 including some solid tumors (graft-versus-tumor effect).161
thalassemia,159 or myelodysplastic syndrome160; immune dif- Because the biology and clinical principles underlining the
ficiency states; and some types of cancers.161 The objective of two effects are likely similar, for simplicity, in this chapter we
HCP transplantation in malignancies is to cure the patient coin a unifying term: graft-versus-neoplasia effect.
by eradicating the neoplastic cells with high-dose myeloab- It is believed that the GVN reaction is the most potent
lative chemotherapy followed by restoration of hematopoi- form of tumor immunotherapy currently in clinical use.
esis through the transplantation of normal hematopoietic However, the mechanism(s) responsible for the GVN effect
II stem cells derived from HLA-compatible normal donors. are still poorly understood. T cells definitely play a fundamen-
This strategy, however, is characterized by the reaction by tal role in the initiation and maintenance of the alloreaction
142 the transplanted (donor’s) immune system toward the host’s toward neoplastic cells.178 In fact a sevenfold increase in the
normal cells, GVHD.162,164 The development of GVHD is chance of disease relapse was noted in patients with chronic
often associated with an immune reaction that preferentially myelogenous leukemia who received a T cell–depleted HPC
targets neoplastic cells (the GVN effect).161,165,171 GVHD and transplant compared with a subset of patients who had
the GVN effect can occur in the presence of a complete HLA received a T cell–replete HPC transplant but did not develop
match, suggesting that the HLA molecules themselves are not GVHD.172,173 This result suggests that GVHD is a biologic
the targets of the allosensitization reaction responsible for entity different from the GVN effect. In addition, on relapse
GVHD or the GVN effect. Instead cells involved with GVHD of chronic myelogenous leukemia, the administration of
and the GVN effect are induced by polymorphic molecules lymphocytes from the transplant donor can re-induce clini-
presented by the HLA molecules and expressed by the recipi- cal remission.179 Finally, leukemia-specific CD8+ T cells have
ent’s cells and recognized by the grafted immune cells. been identified among circulating lymphocytes at the time
of leukemia regression.180 NK cells also play a role in mediat-
Graft-versus-Host Disease
ing this phenomenon; clinical data suggests that mismatch of
Graft-versus-host disease represents the alloimmune reac- NK receptor and ligands during allogeneic BMT may be used
tion of donor lymphocytes against normal cells of the recipi- to enhance the GVN effect.89,181
ent. GVHD occurs predominantly in association with HPC Appreciation of the GVN effect has led to the develop-
transplantation because HPC transplantation also replaces ment of nonmyeloablative stem cell transplants designed to
the host humoral and cellular immune systems. GVHD is a immune suppress the host only to a level sufficient to permit
major complication of HPC transplantation but for a suc- engraftment of the donor immune cells to generate the GVN
cessful transplant a fine balance must be maintained between effect without inducing the serious complications associated
GVHD and graft rejection by modulating the level of post- with myeloablation.174 With nonmyeloablative HPC trans-
transplant immune suppression.169 In addition, other major plantation, GVHD and the GVN effect are being used as the
events associated with HPC transplantaion, such as post- primary antitumor therapy rather than high-dose chemo-
transplant infection, leukemic or tumor relapse, and other therapy. The use of nonmyeloablative HCP transplants and
regimen-related mortalities, are strongly influenced by the their associated GVN effect has gained popularity in the last
methods use to treat GVHD. decade to the point that this allogeneic-based immunothera-
T-cell depletion has been advocated as a means of pre- peutic approach has been advocated for several nonhema-
venting GVHD by decreasing the probability of cellular and tologic malignancies.174 The rationale is largely based on
humoral alloresponses.172,173 This strategy decreases the the assumption that the immune cell repertoire capable of
HUMAN LEUKOCYTE AND GRANULOCYTE ANTIGENS AND ANTIBODIES
recognizing cancer cells in the allogeneic context is broader are processed and presented in association with distinct
than that in the autologous system. Donor T cells may target HLA alleles. Even in clinical settings where this method of
not only tumor-specific antigens, but also allelic variants of vaccination or sensitization is used, high-resolution typ-
antigens expressed by the tumor such as mHA and, in the ing is desirable because it allows accurate interpretation of
case of HLA-mismatched transplants, HLA antigens dispa- immunization results by allowing a comparison between the
rate from the donor.182–184 Although there are several theories detailed genetic makeup of the individual receiving the vac-
about how GVN effect occurs, it remains unclear why allo-T cine and his or her antigen-specific immune response. Thus,
cells have a better chance of targeting tumor cells compared it is likely that in the future HLA laboratories will increas-
with the natural antitumor immunity described in patients ingly be required to provide high-resolution, definitive typ-
with several type of solid tumors. ing for enrollment of patients into vaccination protocols and
for subsequent interpretation of immune responses.
HLA and T Cell–Directed Immunization
Monitoring Immune Responses with
The past decade has witnessed remarkable progress in the
Tetrameric HLA–Peptide Complexes
identification and mapping of T-cell epitopes for various
infectious diseases and cancers. In particular, progress has The growing understanding of the molecular immunology
been made in mapping HLA-associated epitopes for HIV, of T-cell interactions with HLA–epitope complexes in the
CMV, and EBV.185–188 In addition, the molecular identifica- context of infectious disease, virally induced malignancies,
tion of tumor-associated antigens has yielded a large number and spontaneous tumors and interest in their treatment with
of epitopes that could be used to immunize against neopla- T cell–directed vaccines has sparked an interest in accurate
sia.189 A comprehensive discussion of these topics is beyond methods to quantify ex vivo the extent of antigen-specific
the scope of this chapter, but we address a few points fram- immune responses.197 The most widely used assays for the
ing the relevance of HLA in the context of T cell–directed enumeration of antigen-specific T cells include tetrameric
immunization. HLA–peptide complexes (tHLA), intracellular flow cytom-
The identification of T-cell epitopes led to two major etry staining for cytokines expressed on cognate peptide
areas of clinical investigation: active-specific immunization stimulation, detection of cytokine release by ELISPOT, and
to prevent or treat ongoing infections or cancer and the har- quantitative real-time PCR.198
vest and in vitro expansion of immunization-induced T cells Tetrameric HLA–peptide are complexes of four HLA
for adoptive transfer. In general, active immunization has molecules combined with a specific peptide and bound to
proven successful in inducing epitope-specific T cells easily a fluorochrome199 (Fig. 10–8). These complexes bind to the
detectable among circulating lymphocytes.113,190–192 However, complementary TCR and therefore identify antigen-specific
in most cases, the immunization-induced enhancement of T cells.200These tHLA can measure cellular responses against
T-cell function is not associated with clinical improvement. specific epitopes with sensitivity as low as 1/5000 CD8+ T
Although the reason for the clinical ineffectiveness of immu- cells. To synthesize tHLA molecules, soluble HLA heavy chain 10
nization-induced T cells is unclear, it has been postulated containing a biotinylation site and recombinant β2-micro-
that they may be quantitatively191 or qualitatively193 inad- globulin are synthesized and purified. They are then refolded 143
equate for eradicating disease. Therefore, a second strategy in the presence of the specific epitope and the monomer is
is being pursued, whereby the number of antigen-specific T isolated by gel filtration and biotinylated. Fluorescent strep-
cells is amplified in vitro for autologous or donor-derived tavidin is added to induce tetramerization. An aliquot of tet-
adoptive transfer. This second strategy has met some prom- ramer is added to the peripheral blood mononuclear cells
ising success in the context of ganciclovir-resistant CMV together with other antibodies for a more detailed character-
infection,194 EBV-induced post-transplant lymphoprolifera- ization of antigen-specific T cells.193 Analysis is performed
tive disorders,195 and metastatic melanoma.196 using a flow cytometer. Due to the specificity of vaccines,
Whether delivered as a primary form of therapy or to the HLA type of the patient and the specific peptide must be
prime in vivo T cells for further ex vivo expansion, epitope- identified and synthesized to provide the adequate tetramer.
specific vaccination is limited by the stringent requirement Analysis of tHLA offers many potential advantages over
for HLA allelic association or HLA restriction. Although other T-cell assays. This method is quantitative and enables
superfamilies of HLA alleles may share epitopes,18,19 in prac- an estimation of the avidity between TCR and peptide-
tical terms, clinically relevant HLA–epitope associations are loaded HLA class I molecules. In addition, tHLA staining
restricted to a few peptide–allele combinations for a given does not kill the labeled cells, allowing sorting of subpopula-
protein.23 Thus, patients considered for enrollment in immu- tions by flow cytometry for additional analysis or expansion
nization protocols are best served by high-resolution HLA for adoptive transfer. With tHLA, specific T cells can be ana-
typing to exclude subtypes with unproven immunogenic lyzed from blood samples without the prerequisite of in vitro
potential for a given epitope. culture, and all specific cytotoxic T lymphocytes are detected,
To overcome the stringent HLA requirements for success- regardless of their functional status.191,193
ful epitope-specific vaccination, some protocols have used
entire peptides, based on the assumption that the epitope Summary
repertoire of a protein can be adjusted to specific HLA phe-
notypes by a cellular process of self-selection that naturally The relevance of HLA in clinical pathology has broadened
couples peptides to HLA molecules according to their bind- from a predictor of allosensitization to a mediator of GVHD
ing affinity. The peptide(s) within the protein with the great- and the GVN effect. In addition, understanding of the mech-
est binding affinity are bound by the molecules. Although anism of action of various nonclassical HLA genes as well
theoretically sound, in practice the truth of this assumption as KIRs has opened a new field involving the study of the
depends on the efficiency with which individual molecules innate immune response in the context of transplantation.
BLOOD BANKING Figure 10–8 Schematic representation of the mecha-
nism of binding of tetrameric peptide–HLA complexes
(tHLA) to antigen-specific T cells. A, Binding of tHLA to
TCR. The tHLA consists of four HLA–peptide complexes
identical to the ones recognized by the T cell on the sur-
face of live cells. Each HLA molecule is modified to contain
one biotin molecule that serves as a bridge for binding to
tetravalent strepavidin molecules fluorescently labeled. B,
Actual FAC analysis result of CD8+, tHLA-positive T cells.
(Modified with permission from Monsurro V, Nagorsen D.
Immunotracking of specific cancer vaccine CD8+ lympho-
cytes. ASHI Quarterly 2003;26:100–102.)
Table 10–4 International Society of Blood Transfusion (ISBT) Human Neutrophil Antigen
Nomenclature
Antigen System Antigens Location Former Name Alleles
CR3, C3bi receptor; gp, Glycoprotein; ISBT, International Society of Blood Transfusion; LFA-1, leukocyte functional antigen-1.
II
Together with HLA, the study of immunogenetics is expe- 2a or NB1 in 1971.203 Reports of several other granulocyte
144
riencing rapid growth that is driven by the realization that antigens followed.
polymorphism in molecules such as mHA, KIR, cytokines, The current neutrophil antigen nomenclature was
and their receptors is an important hallmark of human established in 1998 by an International Society of Blood
immune pathology. Modern histocompatibility and immu- Transfusion (ISBT) Working Party204 (Table 10–4). In this
nogenetics laboratories must be able to adopt high-through- nomenclature antigen systems are referred to as human
put systems for the parallel assessment of all these variables neutrophil antigens, or HNA. The antigen systems are indi-
when addressing the genetic makeup of an individual in cor- cated by integers, and specific antigens within each system
relation with the natural or therapeutic history of his or her are designated alphabetically by date of publication. Alleles
disease. The comparison of donor and recipient protein pro- of the coding genes are named according to the Guidelines
filing and cytokine polymorphism may offer new insights on for Human Gene Nomenclature. Five neutrophil antigen sys-
the mechanism of rejection, GVHD, and GVN in transplant tems have been described: HNA-1, HNA-2, HNA-3, HNA-4,
recipients. Finally, the increased utilization of T cell–directed and HNA-5.
immunization protocols is driving an increase in demand for
high-resolution typing of HLA molecules to allow a more
accurate interpretation of clinical and immunologic results. Genetics, Structure, and Function
of Neutrophil Antigens
Genetics
HUMAN NEUTROPHIL ANTIGENS AND
THEIR CLINICAL SIGNIFICANCE HNA-1 antigens, which are located on the neutrophil low-
affinity Fc receptor, Fc-γ receptor IIIb (FcγRIIIb), are encoded
by the Fc-γ receptor gene, FCGR3B, which is located on chro-
Introduction
mosome 1q23.205–208 This area of chromosome 1 contains a
Lalezari and colleagues described the first neutrophil-specific cluster of two families of FcγR genes, FCGR2 and FCGR3. The
antigens. These antigens were designated “N” for neutrophil. FCGR3 family is made up of FCGR3A and FCGR3B, which are
Each antigen system was described alphabetically and each located adjacent to each other on chromosome 1q23. FCGR3B
allele was described numerically in order of discovery. They is highly homologous to FCGR3A, which encodes FcγRIIIa
identified human neutrophil antigen 1a (HNA-1a) or NA1 (Table 10–5). The most important difference between the two
in 1966; its allele, HNA-1b or NA2 in 1972201,202; and HNA- genes is a C to T change at position 733nt in FCGR3B that
HUMAN LEUKOCYTE AND GRANULOCYTE ANTIGENS AND ANTIBODIES
Table 10–5 Nucleotide Differences among the Genes Encoding the FcgRIIIa
and the HNA-1a, HNA-1b, and HNA-1c Antigens of FcgRIIIb
Molecule Gene 141 147 227 266 277 349 473 505 559 641 733
FcγRIIIb HNA-1a FCGR3B*1 AGG CTC AAC GCT GAC GTC GAC CAC GTT TCT TGA*
FcγRIIIb HNA-1b FCGR3B*2 AGC CTT AGC GCT AAC ATC GAC CAC GTT TCT TGA*
FcγRIIIb HNA-1c FCGR3B*3 AGC CTT AGC GAT AAC ATC GAC CAC GTT TCT TGA*
FcγRIIIa FCGR3A AGG CTC AGC GCT GAC ATC GGC TAC TTT TTT CGA
*
Stop codon.
Differences among genes are in bold.
creates a stop codon. As a result FCGR3A has 21 more amino is located on CD11b and is encoded by CD11B*1; HNA-5a
acids than FCGR3B. antigen is located on CD11a and is encoded by CD11A*1.213
The HNA-2a antigen is encoded by the gene CD177,
Structure and Function of Neutrophil Antigens
which is located on chromosome 19q13.2. CD177 belongs to
the Ly6 gene superfamily.209–211 The coding region of CD177 The HNA-1a antigens are located on FcγRIIIb or CD16. This
consists of 1311 bp that code for a protein of 416 and a sig- glycoprotein has 233 amino acids and is GPI-anchored.205,208
nal peptide of 21 amino acids.209 The predicted protein has The molecular weight of this heavily glycosylated protein
two cysteine-rich domains, three potential N-linked glyco- ranges from 50 to 80 kD; differences in N-glycosylation
sylation sites, and a potential ω-site for attachment of the account for the differences in molecular mass. FcγRIIIb has
glycosylphosphatidylinositol (GPI) anchor.209 N-linked carbohydrate side chains, and the number of side
The Ly6 gene superfamily is also known as the urokinase- chains differs between the HNA-1a and HNA-1b forms of
type plasminogen activator receptor (uPAR) or snake toxin the glycoprotein. The HNA-1a form of FcγRIIIb has 4 N-link
family. This superfamily is characterized by conserved cyste- side chains and it ranges in size from 50 to 65 kD; the HNA-
ine-rich domains. Typically these domains contain 70 to 100 1b form of FcγRIIIb has 6 N-linked side chains and is 65 to
amino acids, including 8 to 10 cysteine residues spaced at 80 kD.
conserved distances. The Ly6 superfamily includes two sub- FcγRIIIb is a low-affinity Fc receptor. It is the most abun-
families, one encodes GPI-anchored glycoproteins and the dant neutrophil low-affinity Fc-γ receptor. The low-affinity
other encodes secretory proteins without a GPI anchor. In Fc-γ receptors link humoral immunity to cellular immune
general, the GPI-anchored Ly6 proteins have domains with function; specifically, Fcγ receptors on effector cells recog- 10
10 cysteines and the secretory proteins have 8. The protein nize cytotoxic IgG molecules and immune complexes con-
encoded by CD177, NB1 gp, is an exception in that it is GPI taining IgG molecules. 145
anchored, but it has only six cysteine residues in its cysteine- HNA-2a antigen is on NB1 gp, which is located on neu-
rich domains.209 Most Ly6 proteins have one cysteine-rich trophil plasma membranes and secondary granules214,215 and
domain. Two exceptions are uPAR, which has three cysteine- is linked to the plasma membrane via a GPI anchor.215,216 The
rich domains and NB1 gp, which has two.212 Members of this molecular weight of NB1 gp is 58 to 64 kD and it contains N-
family tend to have little homology. At most, 20% to 30% of linked carbohydrate side chains.214,215 NB1 gp has been well
amino acids are conserved among members. The functions characterized, but its function is not known.
of these proteins are diverse, but not well understood. HNA-3 antigens are located on a 70 to 95 kD glycopro-
Genes in the Ly6 superfamily were first described in mice tein that has not yet been fully characterized.217 The function
and are widely used as markers of murine hematopoietic of this glycoprotein is not known. The HNA-4a and HNA-5
stem cells and T-cell differentiation. Ly6A/E (stem cell anti- antigens are located on the αmβ2 or CR3 and αlβ2 or LFA-1
gen 1, or Sca-1) is used as a marker of murine hematopoietic integrin molecules, which are important leukocyte adhesion
precursor cells. Among the Ly6 superfamily genes found in molecules. The integrin αmβ2 also serves as a receptor for
humans, CD59 or membrane inhibitor of reactive lysis and complement component C3bi.
CD87 or uPAR are best described. CD177 is most similar to
Expression of Neutrophil Antigens
uPAR; however, CD177 and uPAR are only approximately
30% homologous.209,210 FcγRIIIb and HNA-1 antigens are expressed only on neu-
The gene encoding HNA-3a has not yet been identified. trophils. FcγRIIIb, HNA-1a, and HNA-1b antigens are
HNA-4a and HNA-5a are encoded by genes belonging to expressed on all segmented neutrophils, about one half of
the integrin family of leukocyte cell adhesion molecules. neutrophilic metamyelocytes, and about 10% of neutro-
This family is made up of three members sharing a common philic myelocytes.218 Soluble FcγRIIIb is found in plasma
integrin β chain, β2, but a different α chain. The leukocyte and both the HNA-1a and HNA-1b forms of the glyco-
function-associated antigen (LFA-1) is a heterodimer made protein can be found in plasma.219,220 Neutrophils release
up of αl or CD11a and β2 or CD18 and is expressed on all FcγRIIIb when they are stimulated in vitro or in vivo.
leukocytes. The second member of this group, complement Soluble FcγRIIIb levels are increased when neutrophils are
receptor 3 (CR3) or Mac-1, is made up of αm or CD11b stimulated or granulopoiesis is increased.221 In healthy sub-
and β2. The other member, CD11c/CD18, is made up of αx jects given granulocyte colony-stimulating factor (G-CSF),
(CD11c) and β2. Mac-1 and CD11c/CD18 are expressed by FcγRIIIb expression by neutrophils is reduced and levels of
granulocytes, monocytes, and NK cells. The HNA-4a antigen soluble FcγRIIIb are increased.
BLOOD BANKING HNA-2a is a neutrophil-specific antigen. It is expressed The gene encoding the HNA-1c form FcγRIIIb, FCGR3B*3,
only on neutrophils, neutrophilic metamyelocytes, and is identical to FCGR3B*2 except for a C to A substitution at
myelocytes.218,222 Although some GPI-anchored proteins, nucleotide 266. The nucleotide substitution results in an
such as FcγRIIIb, are shed by stimulated neutrophils, NB1 gp Alanine to Aspartate change at amino acid 78 of FcγRIIIb230
is not, nor is soluble NB1 gp present in plasma.215 HNA-2a (see Table 10–5). In many cases FCGR3B*3 exists on the same
is unique in that it is expressed on subpopulations of neu- chromosome with a second or duplicate FCGR3B gene.238 One
trophils. The mean size of the HNA-2a-positive subpopula- group has found that in Danish people, FCGR3B*3 always
tion of neutrophils is 45% to 65%.215,223,224 The expression exists as a duplicate gene in association with FCGR3B*1.239
of HNA-2a is greater on neutrophils from women than However, in other populations duplicate FCGR3B*3 genes
men.224,225 The size of the HNA-2a-positive subpopulation of were associated with both FCGR3B*1 and FCGR3B*2.
neutrophils from women is approximately 60% to 70% com- Several other sequence variations in FCGR3B have been
pared to approximately 50% to 60% for men. The expression described.234,237 These chimeric alleles have single-base sub-
of HNA-2a falls with age in women, but remains constant in stitutions involving one of the five single nucleotide poly-
men.224 Neutrophil expression of HNA-2a is greater in preg- morphisms that distinguish FCGR3B*1 and FCGR3B*2.
nant women than in healthy female blood donors.226,227 FCGR3B alleles that more closely resembled FCGR3B*2 were
The surface expression of HNA-2a is slightly upregu- found more often in African Americans than in whites or
lated by treatment with the chemotactic peptide f-Met-Leu- Japanese.237
Phe.215,216 The administration of G-CSF to healthy subjects Neutrophils from some people do not express any HNA
for several days increases the proportion of neutrophils antigens.240–242 Blood cells from patients with paroxysmal
expressing HNA-2a to near 90%.228 Patients with polycythe- nocturnal hemoglobinuria (PNH) lack GPI-linked glyco-
mia vera have markedly increased levels of CD177 mRNA, proteins and their granulocytes express reduced amounts
but the levels of NB1 gp expressed by neutrophils from of FcγRIIIb and the HNA-1 antigens. In addition, genetic
patients with polycythemia vera have been reported to be deficiencies of granulocyte FcγRIIIb and HNA-1 anti-
normal.229 However, NB1 gp expression has not been studied gens have also been reported. With inherited deficiency of
thoroughly in patients with polycythemia vera. FcγRIIIb, the FCGR3B gene is deleted along with an adja-
Unlike the HNA-1 and HNA-2 antigens, the other neu- cent gene, FCGR2C.242 Among whites the incidence of indi-
trophil antigens HNA-3, HNA-4, and HNA-5 are expressed viduals homozygous for the FCGR3B gene deletion is about
by several other cell types. HNA-3a is expressed by neutro- 0.1%.243,244 However, among Africans and African Americans
phils, lymphocytes, platelets, endothelial cells, and kidney, the incidence is much higher; in one study 3 of 126 Africans
spleen, and placental cells.217 HNA-4a is expressed on gran- were found to have homozygous FCGR3B deletions231 and
ulocytes, monocytes, and NK cells; HNA-5a is expressed on in another 1 of 53 were found to have homozygous FCGR3B
all leukocytes. deletions.237
HNA-2a is expressed on neutrophils by approximately
II 97% of whites, 95% of African Americans, and 89% to 99%
Neutrophil Antigen Polymorphisms of Japanese.224,225,245 HNA-2a has been reported to have
146 and Clinical Significance an allele, NB2, but the product of this gene cannot be reli-
ably identified with alloantisera and no monoclonal anti-
Neutrophil Antigen Polymorphisms
body specific for NB2 has been identified.246 Several CD177
The neutrophil-specific HNA-1 antigen system is made up polymorphisms have been described. Two different groups
of the three alleles HNA-1a, HNA-1b, and HNA-1c230,231 (see sequenced CD177 independently. One group sequenced it as
Table 10–4). The gene frequencies of the three alleles vary the gene encoding HNA-2a and called the gene NB1209; the
widely among different racial groups. Among whites the fre- other group sequenced it as a gene overexpressed in granu-
quency of the gene encoding HNA-1a, FCGR3B*1, is between locytes from people with polycythemia rubra vera and called
0.30 and 0.37 and the frequency of the gene encoding HNA- the gene PRV-1.210 The NB1 and PRV-1 alleles differ at only
1b, FCGR3B*2, is from 0.63 to 0.70.231–236 In contrast, among four nucelotides.209 Bettinotti and colleagues used Human
Asian populations the FCGR3B*2 gene is more common. In Genomic Project databases to characterize the structure of
Japanese and Chinese populations the FCGR3B*1 gene fre- the PRV-1 and NB1 genes.211 They described the intron and
quency ranges from 0.60 to 0.66 and the FCGR3B*2 gene exon structure of NB1; however, they found only one gene,
frequency from 0.30 to 0.33.233,235–237 The gene frequency CD177, homologous to both PRV-1 and NB1, suggesting that
of the gene encoding HNA-1c, FCGR3B*3, also varies they are alleles of the same gene. In addition, they described a
among racial groups. FCGR3B*3 is expressed by neutrophils pseudogene homologous to exons 4 through 9 of CD177 and
from 4% to 5% of Caucasians and 25% to 38% of African adjacent to CD177 on 19q13.2.211
Americans.231 One of the four single nucleotide substitutions that dis-
The structures of FCGR3B*1, FCGR3B*2, and FCGR3B*3 tinguish the NB1 and PRV-1 alleles of CD177 affects the
alleles are very similar. The FCGR3B*1 gene differs from the size of the neutrophil population that expresses HNA-2a,
FCGR3B*2 gene by only five nucleotides in the coding region, but these single nucleotide polymorphisms are not respon-
at positions 141, 147, 227, 277, and 349205–208 (see Table 10-5). sible for the HNA-2a-negative phenotype.247,248 Instead, the
Four of the nucleotide changes result in changes in amino HNA-2a-negative neutrophil phenotype is due to a CD177
acid sequence between the HNA-1a and HNA-1b forms of transcription defect.249 HNA-2a genes from two women with
the glycoprotein. The fifth polymorphism at 147 is silent. HNA-2a-negative neutrophils who produced HNA-2a-spe-
The glycosylation pattern of the protein differs between the cific alloantibodies have been studied, and abnormal CD177
two antigens because of two nucleotide changes at bases 227 mDNA sequences of variable lengths were detected in both
and 277. The HNA-1b form has six N-linked glycosylation women.249 Their neutrophil mRNA had both exons and
sites and the HNA-1a form has four glycosylation sites. accessory sequences that were considered to be introns. Some
HUMAN LEUKOCYTE AND GRANULOCYTE ANTIGENS AND ANTIBODIES
cDNA containing the entire CD177 coding sequence was with PNH or chronic myelogenous leukemia has any clinical
identified, but all cDNA had some accessory sequences.249 significance.
HNA-3a has a gene frequency of 0.66, but the molecu- CD177 has become an important biomarker for polycy-
lar basis of this antigen is not known.217 HNA-4a has a phe- themia vera. The diagnosis of polycythemia vera can be dif-
notype frequency of 99.1% in whites and is due to a single ficult because several other clinical conditions are associated
nucleotide substitution of G to A at position 302 of CD11B.213 with increased hemoglobin levels. Hemoglobin levels can be
This substitution results in an Arg to His polymorphism at increased as a secondary response to reduced arterial oxy-
amino acid 61 of the α chain of CR3, αm. A second poly- gen levels, increased erythropoietin levels, and the presence
morphism of the β2 integrins, HNA-5a, was first described of high-oxygen affinity hemoglobin. It is important to dis-
as Onda. A chronically transfused male with aplastic anemia tinguish patients with secondary erythrocytosis from those
became alloimmunized to HNA-5a. HNA-5a is due to a G to with polycythemia vera, because if not treated, patients with
C single nucleotide substitution at position 2446 of CD11A. polycythemia vera can suffer serious thrombotic and hemor-
This change leads to an amino acid change of Arg to Thr at rhagic complications.
amino acid 766 of the α chain of LFA-1, αl.213 The measurement of neutrophil CD177 mRNA is being
used along with other assays by many centers to distinguish
Clinical Significance of Neutrophil Antigens patients with increased hemoglobin levels due to second-
Polymorphisms of HNA-1 have been found to have only a ary erythrocytosis from those with polycythemia vera.258,259
few clinical consequences. The deletion of the entire FCGR3B Quantitative real-time PCR is being used to measure neutro-
gene does not have major clinical consequences, and most phil CD177 mRNA levels, and several studies have found that
people with FcγRIIIb deficiency are healthy. However, too neutrophil levels of CD177 mRNA are increased in polycy-
few patients have been studied to identify a slight increase themia vera patients. Overall, 91% to 100% of patients with
in susceptibility to infection or autoimmune disease due to polycythemia vera have increased CD177 mRNA levels.258–260
FcγRIIIb deficiency. In a study of 21 people with FcγRIIIb In contrast, none of the patients with secondary erythrocy-
deficiency, 2 people were found to have autoimmune thy- tosis have increased CD177 mRNA levels.258,259
roiditis and 4 had multiple episodes of bacterial infec- Neutrophil CD177 mRNA levels are also increased in
tions.242 Other smaller studies and case reports have found some patients with myeloproliferative disorders related to
that despite their FcγRIIIb deficiency, all individuals, except polycythemia vera.259 Neutrophil CD177 mRNA is overex-
for the one person with systemic lupus erythematosus, were pressed in 30% to 60% of patients with essential thrombocy-
healthy, had no circulating immune complexes, and showed themia and in approximately 60% of patients with idiopathic
no increased susceptibility to infections. myelofibrosis. In patients with essential thrombocythemia
Despite the lack of serious illness in people with FcγRIIIb the overexpression of CD177 may affect the course of their
deficiencies, HNA-1 polymorphisms have some effect on disease. Essential thrombocythemia patients with increased
neutrophil function. Neutrophils that are homozygous for neutrophil CD177 mRNA levels have an increased incidence
HNA-1b have a lower affinity for IgG3 than granulocytes of thrombosis and bleeding.261 10
homozygous for HNA-1a.250 Neutrophils from people who CD177 mRNA overexpression in myeloproliferative dis-
are homozygous for HNA-1b phagocytize erythrocytes sen- orders is likely secondary to a point mutation in the Janus 147
sitized with IgG1 and IgG3 anti-Rh monoclonal antibod- kinase 2 (JAK2). Approximately 75% to 95% of patients with
ies251 as well as bacteria opsonized with IgG1 at a lower level polycythemia vera have a G to T substitution in JAK2, which
than granulocytes homozygous for HNA-1a.251,252 causes a Phenylalanine to be substituted for Valine at posi-
Several studies suggest that FCGR3B polymorphisms tion 617 of JAK2 (V617F).262–265 JAK2 is involved with signal-
affect the incidence and outcomes of some autoimmune ing of a number of hematopoietic growth factors, including
and inflammatory diseases, but the results of some of these G-CSF. The JAK2 V617F gain of function mutation and
studies seem contradictory. However, because FCGR3B is the stimulation of neutrophils by physiologic G-CSF likely
clustered with FCGR3A and FCGR2 on chromosome 1q22, leads to increased JAK2 activation, increased expression of
it is possible that some of these findings may be due in part secondary signaling molecules such as the transcription
to linkage disequilibrium among Fc receptors. Children factor STAT3, and increased CD177 mRNA transcription.
with chronic immune thrombocytopenia purpura are more Interestingly, CD177 mRNA is also overexpressed by neutro-
likely to be FCGR3B*1 homozygous than controls,253 but phils from healthy subjects given G-CSF.248
Spanish patients with systemic lupus erythematosus are
more likely to be FCGR3B*2 homozygous.254 Myasthenia Testing for Neutrophil Antibodies
gravis is more severe in FCGR3B*1 homozygous patients,255
but multiple sclerosis is more benign in FCGR3B*1 homo- and Neutrophil Antigen Typing
zygous patients.256 Patients with chronic granulomatous Many barriers surround neutrophil antibody testing and
disease who are FCGR3B*1 homozygous are less likely to antigen typing, including limited availability of reagents,
develop major gastrointestinal or genital urinary tract infec- lack of commercially available test kits, and the need to use
tious complications compared to those with chronic gran- fresh neutrophils. However, neutrophil antibodies remain
ulomatous disease who are heterozygous and FCGR3B*2 clinically important, and many laboratories test for these
homozygous.257 antibodies. As a result of the reduced risk of transmit-
The role of NB1gp in neutrophil function is unknown. ting infectious diseases via blood component transfusion,
The rare women who produce HNA-2a specific alloanti- attention has focused on TRALI. TRALI is now one of the
bodies and who lack NB1 gp are healthy. The expression of most frequent causes of transfusion-associated mortal-
HNA-2a is reduced on neutrophils from people with PNH ity. Neutrophil antibody testing, phenotyping, and cross-
and chronic myelogenous leukemia.218 It is not known if the matching are an important part of evaluation of patients
lack of expression of HNA-2a on neutrophils from patients and donors.
BLOOD BANKING for 3 hours at 22°C. Antibody binding is detected using
Neutrophil Antibody Testing
sheep erythrocytes coated with antihuman IgG. The anti-
Screening for neutrophil antibodies remains technically gen-coated plates can be prepared and stored for at least 1
challenging. Antibodies to some neutrophil antigens can year at −80°C. The assay has been shown to detect antibod-
be detected using solid phase assays; however, the clinically ies specific for HNA-1a, HNA-1b, HNA-2a, and HNA-3a.
important antibodies to HNA-1 antigens cannot. As a result, Although results of this assay look promising, it has yet to
intact neutrophils must be used for antibody screening assays. be compared extensively with other assays or tested in inter-
Unfortunately, neutrophils have a short life span so fresh national workshops.
neutrophils must be prepared daily for testing. Neutrophils
are prepared from fresh whole blood using density gradient Monoclonal Antibody Capture Assays
separation. Patient sera are tested in one of several assays The monoclonal antibody capture, or monoclonal anti-
against panels of neutrophils prepared from several donors body immobilization of neutrophil antigens (MAINA),
with known phenotypes to distinguish antibodies with differ- assay allows the detection of antibodies to specific neutro-
ent specificities. The presence of HLA antibodies with broad phil membrane glycoproteins. In this assay, neutrophils are
specificities can make the detection of neutrophil antibodies incubated with test sera, washed, incubated with a murine
difficult because neutrophils express class I antigens. HLA- monoclonal antibody to a specific neutrophil glycopro-
specific antibodies can be separated from neutrophil-specific tein, and washed again. The neutrophils are then dissolved
antibodies by absorbing serum with platelets. Alternatively, in a mild detergent. The soluble glycoprotein–monoclonal
monoclonal antibody capture assays can be used to test for antibody complex is “captured” in a well with an antibody
antibodies specific to neutrophil membrane glycoproteins. specific to mouse IgG fixed to the well bottom. An antibody
specific to human IgG conjugated to alkaline phosphatase is
Granulocyte Agglutination
added followed by a substrate, and the reaction is detected
In this assay antibodies cause neutrophils to actively aggluti- with a spectrophotometer.
nate. Cells and serum are incubated for 4 to 6 hours at 30°C. The MAINA assay can be used to detect antibodies to
When antibodies are present, neutrophils will clump. The FcγRIIIb (CD16), NB1 gp (CD177), LFA-1 (CD11a), and
granulocyte agglutination assay is very reliable but less sen- CR3 (CD11b). This assay will detect antibodies to HNA-1,
sitive than other assays. It can detect antibodies to HNA-1, HNA-2, HNA-4, and HNA-5. The use of neutrophils from
HNA-2, HNA-3, HNA-4, and HNA-5 antigens and it is the panels of donors with known HNA-1 phenotypes allows the
only assay that can identify antibodies specific for HNA-3a. identification of antibodies specific to HNA-1a and HNA-
1b. In addition, antibodies are sometimes detected that are
Granulocyte Immunofluorescence
directed to FcγRIIIb but are not specific to HNA-1a, HNA-
and Flow Cytometry
1b, or HNA-1c. The MAINA assay permits the recognition
In the granulocyte immunofluorescence assay, antigen–anti- of antibodies to specific neutrophil glycoproteins even when
II body reactions are detected using fluorescence-conjugated antibodies to HLA antigens are present.
secondary antibodies and a fluorescent microscope. Before
incubation with sera, neutrophils are treated with 1% para- Strategy for Antibody Detection
148
formaldehyde for 5 minutes at 20°C to 24°C to help prevent Most laboratories screen serum for neutrophil antibodies by
nonspecific binding of antibodies to neutrophil Fc recep- granulocyte agglutination and granulocyte immunofluores-
tors and to stabilize the cell membranes. Treated neutrophils cence or flow cytometry. The serum reactive with neutrophils
and patient sera are then incubated for 30 minutes at 37°C. must also be screened in an assay that can detect HLA anti-
Binding of antibody to the neutrophils is detected with a flu- bodies. If the serum is reactive with neutrophils and HLA
orochrome conjugated secondary antibody. To prevent non- antibodies are present, then the MAINA or a similar assay is
specific binding of the secondary antibody to neutrophil Fc used to determine if both HLA antibodies and antibodies to
receptors, F(ab′)2 secondary antibodies are used. The bind- neutrophil-specific antigens are present. Because the mono-
ing of antibodies in the test serum results in a uniform stain- clonal antibody capture assays sometimes identify antibod-
ing of the outside of the neutrophils. Strong reactions are ies that cannot be detected in other assays, some laboratories
readily distinguished, but considerable training is required to test all serum samples in MAINA assays.
distinguish weak reactions from background staining.
Testing for neutrophil antibodies with flow cytometry
is technically similar to the granulocyte immunofluores- Phenotyping and Genotyping
cence assay except that neutrophils are evaluated with a flow of Neutrophil Antigens
cytometer rather than a fluorescent microscope. The flow
cytometer can more readily compare the reactions of test Phenotyping
sera with positive and negative control sera than the fluores- Traditionally, neutrophil antigen phenotyping has been
cent microscope. performed using human alloantibodies in the granulocyte
agglutination or granulocyte immunofluorescence assays.
Mixed Passive Agglutination
However, alloantisera are difficult to obtain. Monoclonal anti-
The mixed passive agglutination assay uses a granulocyte bodies specific to HNA-1a, HNA-1b, and HNA-2a have been
antigen preparation for antibody screening. This assay described, are commercially available, and have been used to
allows granulocyte testing trays to be prepared in large phenotype neutrophils using flow cytometry. Phenotyping
batches and frozen until testing. Antigens are extracted with monoclonal antibodies and flow cytometry is faster and
from isolated neutrophils using 3% sucrose.266 The neutro- easier than phenotyping with alloantibodies because pheno-
phil extract is used to coat U-bottom Terasaki plates. Sera typing with monoclonal antibodies can be done with whole
to be tested is incubated with neutrophil extract in wells blood instead of isolated neutrophils.
Genotyping
II
156
B. Blood Donation
Chapter 11
Blood Donation and Collection
Gary Zeger ● Eileen Selogie ● Ira A. Shulman
Blood donation is critical to all of transfusion therapy, as it must be convincing and compelling to result in a scheduled
provides the starting product. In the United States and many appointment to donate blood.
other economically developed nations, all of the blood is given Once a donor has been recruited, the screening process is
by volunteer, nonremunerated donors. Donated whole blood carried out to make sure that the donation process will be safe
is then made into transfusable components, which include for the donor and that the collected blood will be safe for the
but are not limited to packed red blood cells (RBCs), plate- recipient. The prospective donor is initially given information
lets, and frozen plasma or cryoprecipitate. Other lesser utilized about criteria for eligibility for blood donation and about the
blood components, such as granulocytes and cryoprecipi- process itself. The screening process consists of a question-
tate-depleted plasma, also have important therapeutic value. naire that seeks to find medical conditions and behaviors
Individual plasma proteins, such as factor VIII, have been that might make donation unsafe for the donor or recipient.
manufactured using recombinant methods for years; however, Critical information is confirmed by direct verbal question-
there is no commercial product, single or combined, with the ing to ensure that the answers are accurate. If no disqualifying
clinical properties of frozen plasma. Each of these components information is uncovered during the screening process, a brief
make possible an extraordinary number of traditional and physical examination follows, which includes examination of
state-of-the-art medical therapies, including trauma surgery, antecubital veins, followed by measurement of body tempera-
organ transplantation, and cancer chemotherapy. At the time ture, donor hematocrit or hemoglobin, and heart rate.
of this writing, there are no clinically effective or available sub- After the venipuncture is performed, blood is collected,
stitutes for RBCs, platelets, or plasma in the United States. labeled, and temporarily stored until it can be transferred 11
In 2001, the National Blood Donor Resource Center esti- to a manufacturing center for further processing and distri-
mated that 8 million volunteer U.S. blood donors contributed bution. Specimen tubes are drawn at the time of collection 157
approximately 15 million whole blood donations per year, the for infectious disease testing; these tubes are sent for testing
majority of which were manufactured into separate compo- immediately after collection.
nents, such as RBCs, fresh frozen plasma, and platelets.1 These After the donation, donors receive oral fluids and remain
components allowed transfusion of 29 million blood compo- under observation for a period of time so that any post-dona-
nents in the United States. Thus, in the United States, the aver- tion reactions may be treated appropriately. Post-donation
age volunteer blood donor gives blood about 1.6 times a year. Of instructions are given to help the donor avoid untoward side
that total, almost 2% represent donations indicated for a specific effects. The donor is instructed to call the blood center with any
recipient other than the donor. These are generally referred to post-donation information, such as the development of wor-
as directed or designated donations. In addition, approximately risome physical symptoms or information remembered that
3% of blood donations are autologous: blood that is donated by would change the answers given during the screening process.
an individual for his or her own use, usually for a prescheduled
elective surgery.2–4 Although it is estimated that 60% of the adult
population in the United States is eligible to donate blood, at Donor Recruitment
present it is believed that less than 5% of the eligible population Maintaining an adequate blood supply is an ongoing challenge.
donates within any given year.5,6 The various reasons that some Attrition of blood donors due to older age and illness, imple-
people give blood readily and others do not have been studied mentation of new regulations resulting in deferrals, or other
for several decades, but the blood donation process and appli- reasons makes it difficult for blood collection centers to keep
cable statistics during this time have changed little, if at all. pace with the increasing demand for blood. Thus, the recruit-
ment of new blood donors must be ongoing and vigorous.
New exclusionary criteria and serologic testing make this task
THE PROCESS OF BLOOD DONATION increasingly difficult, as does the fact that newly recruited blood
donors are nearly twice as likely to have disqualifying medical
Blood donation can be divided into five processes that are conditions as are established blood donors.7
directly related to the donor: recruitment, screening, physical It is unacceptable to provide volunteer blood donors with
examination, collection, and post-donation care. monetary compensation (i.e., cash or cash equivalents), so
Recruitment of blood donors is a specialized task. It is the act of blood donation in the United States is voluntary.
often performed by telerecruiters, and the message delivered Thus, without paying donors for their time and blood, the
BLOOD BANKING formidable challenges of encouraging volunteer blood dona- of the general eligible blood donor population in certain
tion begin at the first step of the blood collection process: communities; the high prevalence of blood group B in
donor recruitment. the African American population; and the higher preva-
lence of African Americans with specific blood types (e.g.,
Sources of Donor Motivation antigen-negativity for a variety of RBC antigens to which
The most successful approach to recruitment of volunteer antibodies are frequently made in highly transfused pop-
blood donors has been an appeal to community responsibil- ulations, such as patients with sickle cell disease) that may
ity. Individuals often first learn about the need for donation be used for patients who have made antibodies to these
during blood shortages via public service announcements antigens.
and appeals for blood from newspapers, radio, and televi- Few publications adequately address the reasons why cer-
sion. Other donors become aware of the importance of tain minority populations do not donate blood at the same
blood donation when transfusions are needed for family and percentages as the white population. Much research is neces-
friends (or themselves). sary to understand the needs and wants of these important
Appeals after disasters tend to bring out community donors, so that these minority donors can be successfully
spirit in Americans. This was particularly evident after the recruited into the blood donor system.
September 11, 2001, terrorist attack on the New York World
Paid Donors
Trade Center and the Pentagon in Washington, D.C. In both
instances, blood donations vastly exceeded the local demand, Aside from paid plasma donors at centers that manufacture
due to the motivation of the entire community to contribute fractionated, licensed plasma products, it is not acceptable to
to their fellow Americans in need. provide monetary compensation (cash or cash equivalents)
Donating blood for a friend or relative (directed dona- to blood donors in the United States. In the early days of
tion) has proven to be an excellent motivator and has brought blood banking, paying for blood donors was a commonplace
many first-time blood donors into the system. Donating for and accepted practice. These donors were often motivated
one’s own use (autologous donation) has also been an effec- by a lack of funds to maintain drug or alcohol habits; subse-
tive motivator. quently, paid donors had a higher incidence of transfusion-
For whatever reason each donor is motivated to give transmissible diseases, particularly hepatitis, which infected
blood, he or she must be convinced that donation is truly many early blood recipients. In the 1970s, growing recogni-
necessary and will be appreciated. For this reason, appeals tion of this problem11 led the Food and Drug Administration
for blood should only be made when there is a significant (FDA), in its Code of Federal Regulations (CFR), to require
shortage. Once the donor has been motivated to donate, blood from paid donors to be labeled as such.12 As these
making his or her blood donation a convenient and pleasant “paid donor”-labeled units were considered to be undesir-
experience is critical to retaining that donor for subsequent able by clinicians and hospitals, the practice of paid whole
donations. Excellent customer service is the key to retaining blood donations effectively died out.
II blood donors. In some states, however, the shortage of single donor plate-
let concentrates collected by apheresis technology prompted
A Note about Minority Donors exceptions for these donations. Until January 1, 2003,13 a few
158
Latino Americans, particularly immigrants from Mexico, U.S. blood collection facilities continued to pay apheresis
are the largest growing demographic group in the United donors. Due to the previous stigma attached to paid blood
States.8 Adequate donor recruitment and collection among donation, these centers employed screening procedures that
this minority group is especially important because of the met or exceeded those of “all-volunteer” centers. Despite
high percentage of blood group O among Latinos. Because studies demonstrating that these donors had infectious dis-
group O individuals can only receive group O RBCs, a higher ease marker frequencies similar to, or better than, those of
percentage of group O blood is necessary in areas with large volunteer donors, these centers were eventually forced to
Latino populations. Specialized recruiting programs are cease paying plateletpheresis donors.14
important to attract and maintain these essential donors. Paid donation, however, is regularly utilized for the
Appeals in Spanish to Latino organizations and in the media recruitment of donors in the United States for commercial
are of key importance. In areas with large Latino populations, source plasma. This plasma is collected by apheresis and
an effort should be made to provide Spanish versions of all sent for further manufacture into various plasma-derived
donor materials. It is also advisable to have staff members products. Because the pooled plasma from these donations
who are conversant in Spanish or to have translators readily is effectively “sterilized” during the fractionation and man-
available. ufacturing process, there is less concern about the poten-
With the exception of Chagas disease, the incidence of tially increased risk for infectious disease transmission by
infectious disease markers among whole blood donors in using paid donors for “source plasma.” Most countries with
areas with large Latino populations is similar to that of other all-volunteer commercial plasma programs have struggled,
repeat whole blood donors.9 However, the seroprevalence of usually unsuccessfully, to meet their population’s plasma
Chagas disease among whole blood donors in Los Angeles is derivative needs.15
1 in 7200, versus 1 in 93,000 among plateletpheresis donors.10
Health Benefits of Whole Blood Donation
The significant difference in this seroprevalence is due to the
fact that very few Latino individuals donate apheresis plate- The proven health benefit to blood donors is the free
lets in Los Angeles. mini-physical examination and the infectious disease
African American donors are currently the second larg- screening testing performed at the time of donation.
est minority population in the United States.9 Recruitment Many donors might not have otherwise become aware of
and donation by African Americans is particularly impor- diseases such as hypertension, anemia, cardiac arrhyth-
tant, due to many factors: they make up a large percentage mia, hepatitis, or human immunodeficiency virus (HIV)
BLOOD DONATION AND COLLECTION
infection. This information alerts the donor to seek fur- Patients are sometimes encouraged to have friends and
ther appropriate medical diagnosis and treatment, and relatives donate blood to “replace” any that they might use.
may limit the transmission of infectious disease to others. This appears to be a reasonable recruitment strategy, as long
Aside from these benefits, a controversial hypothesis that as the patient is not made to feel stressed and anxious about
depletion of iron stores through whole blood donation finding replacement donors. It is most important for the
can improve cardiovascular status16,17 has been proposed; patient to understand that he or she will never be denied
more research needs to be performed prior to making blood because of inability to replace blood that has been, or
any claims regarding cardiac health benefits from blood might be, used. Poor communication, however, might cause
donations. the patient to put blood donation pressure on family, friends,
and acquaintances who may have valid reasons for not donating,
Donor Incentives thus potentially endangering the blood supply.
Improved screening and infectious disease testing meth-
ods used for donor blood have made widespread infec- Motivation by Free Testing
tious disease transmission by transfusion, as occurred in A serious concern throughout the blood industry following
the early days of paid blood donors, a thing of the past. the discovery of HIV in the early 1980s (and the subsequent
All-volunteer donor programs have become the base of knowledge that HIV was transmissible through transfu-
the blood collection establishment. Although blood donor sions), was that high-risk individuals would donate blood to
incentives such as t-shirts, gift certificates, and paid time obtain a free confidential HIV test. The concern that people
off are acceptable gifts, rewards that can easily be con- might now donate blood to receive free blood tests (a magnet
verted to cash (or cash itself) are not. The issue of donor effect) has been shown to be generally unfounded, at least for
motivation caused by incentives continues to be an area HIV p24 antigen testing.20 Nevertheless, donor centers gen-
of concern to the FDA and the American Association of erally make available a list of testing sites where confidential
Blood Banks (AABB). The CFR, in its definition of paid or anonymous HIV blood testing is available, to discourage a
and volunteer donors, states that “Benefits, such as time potential high-risk individual from donating.
off from work, membership in blood assurance programs,
and cancellation of nonreplacement fees that are not read- Factors for Success
ily convertible to cash, do not constitute payment within Any donor’s internal motivation will only provide a finite
the meaning of this paragraph.” The AABB18 and FDA19 amount of impetus for continued participation in the blood
have provided some guidance on donor incentives (Table donation process. It is the job of the entire blood collection
11–1). There is still a concern that a potential donor might team to make the donation process as pleasant as possible. If
be untruthful about high-risk behaviors for infectious they are successful, a hesitant first-time donor may be con-
disease to receive a gift being offered at a blood collection verted into a regular repeat blood donor. This is a worthwhile
site. For this reason, incentives should be provided for goal: regular, repeat blood donors are more reliable and have
simply attending a blood drive and attempting to donate, 11
less risk of infectious disease.
rather than the gift being given based on the condition of Making blood donation as convenient as possible is of
the actual donation. prime importance. After a national disaster, blood donors 159
Blood Credit Programs have stood in long lines for hours to donate blood for anony-
mous victims. In such times, the truly heroic nature of the
Blood credit programs, which in the past were more popular motivated blood donor is evident. Under more routine cir-
entities, are difficult to manage logistically and practically. The cumstances, inhospitable conditions and/or poor customer
implication that a blood donor will receive a credit that can service may almost certainly discourage a blood donor from
eventually be cashed in for “free” blood in the future is almost making a donation. A safe and convenient location is critical
always misleading. The credits are often symbolic “credits to to attract and retain repeat blood donors. Parking should be
the blood supply” and have no direct application to the donor, easily available and free. The waiting area should be clean
monetarily or otherwise. The logistics of a true crediting pro- and pleasant. Excessive waits are to be avoided, and donors
gram are generally prohibitive, because the time and place that should be given an accurate wait-time whenever possible.
the credits will be redeemed is unknown, and the involved Blood center staff should be professional, knowledgeable,
health care providers may not be party to the program. and courteous. Of particular importance is making sure
Cash payment or cash equivalent Tokens or prizes of nominal value (e.g., coffee cups, t-shirts, pins)
Tickets to concerts or sporting events where market Employee paid time off
for resale exists Raffle tickets, regardless of value of prize. Prize must not be
Music media not associated with product promotions transferable or readily convertible to cash
where market for resale exists Membership in blood assurance program
Transferable product discounts or coupons Medical tests performed at the time of donation
convertible to cash Scholarships transferred directly to academic institution
Vouchers for free medical tests Gift cards and gift certificates that are nontransferable, not
Scholarships paid directly to students redeemable for cash, and bear the donor’s name
From Compliance Policy Guide for FDA Staff and Industry, Chapter 2, Section 230.150. Issued May 7, 2002, revised November 22, 2005.
Available at http://www.fda.gov/ora/compliance_ref/cpg/cpgbio/cpg230=150final.htm. Last modified December 12, 2005. Accessed June 3, 2006.
BLOOD BANKING that a new donor understands the donation process and well-trained collections staff is important, because every-
fully knows what to expect. Donors appreciate honesty, and thing necessary for the blood drive must be properly set
unpleasant or painful surprises often provoke bad feelings. up and organized on site. Essential equipment and supplies
are brought by the mobile collection team, and any omis-
Blood Collection Sites: Fixed and Mobile sion may result in cancellation of the drive or unacceptable
Fixed site is a widely used term for a permanent or freestanding delays. Delays and cancellations of mobile blood drives can
blood collection center. The fixed site may be located in a hospi- lead to ill will between the sponsor and the blood collection
tal-based donor room or in a community blood center building. center, which may dampen the likelihood of another blood
The site should be clean and pleasant and must meet standards drive being sponsored by that group in the future.
of current Good Manufacturing Practices (GMP)21 for cleanli- An alternative to using space within a school or business for
ness, ventilation, space, and temperature. Donor confidentiality a mobile blood drive is a self-contained mobile unit, usually
must be maintained, and there must be compliance with the a specially adapted bus, typically of four- to six-bed capacity.
Health Insurance Portability and Accountability Act of 1996. These buses are most often used for small blood drives.
Compliance with these regulations requires a screening area that Mobile blood drives should be set up along the same basic
provides the donor with privacy to discuss the many personal principles as fixed sites, although a certain amount of flex-
questions on the donor-screening questionnaire. There must be ibility is often in order. Donor confidentiality concerns must
adequate room in the collection area for the phlebotomists to be adhered to as best as possible, often by use of portable
function freely, and there must be a “canteen,” or refreshment modular components to maintain privacy.
area, where the donor can be orally rehydrated and observed for Recruiting for mobile blood drives requires an entirely
post-donation reactions. Properly monitored storage areas must different approach than recruiting blood donors for a fixed
be available for storage of blood products and equipment. site. An individual from the sponsor group is often asked to
Most autologous and directed donations are performed organize the blood drive, by providing a personal message
at fixed sites. Donations that require apheresis technology, of support, hosting employee rallies, and designating orga-
such as plateletpheresis and granulocyte collections, are also nizers to work with a blood center representative to produce
typically performed at fixed sites, although new automated a plan for a productive and well-run blood drive. Sponsor
blood collection technology has allowed for the collection of organizers work on a personal level to recruit donors, who
multiple blood products by apheresis at mobile sites as well. sign up to donate on a particular day and time. A good
Fixed sites are generally less convenient for donors than are sponsor organizer will also do whatever is necessary to make
mobile sites, as they often require additional travel, parking, certain each donor arrives at the appointed time. A success-
and time. For this reason, a friendly, attractive, and professional ful drive is often followed by a recognition ceremony for all
staff is important. Most regular blood donors look forward to involved.
their visits and, in a sense, become part of the blood collec- After a first successful mobile blood drive with a spon-
tion “family.” Intensive telephone recruitment of repeat blood sor, future drives are generally easier to organize and run.
II donors is usually necessary for a fixed site to be successful. Setting up a first-time blood drive, however, requires a blood
Plateletpheresis donations are most often collected at collection center donor recruiter with excellent interpersonal
160 fixed sites. Regular plateletpheresis donors tend to differ from and organizational skills, because it is often not a simple pro-
whole blood donors in their levels of motivation and willing- cess to convince a sponsor to commit to a blood drive in the
ness to endure longer and more uncomfortable procedures to workplace, because of disruptions of work due to employees’
donate their blood platelets. These donors have their blood taking time away from their jobs to donate blood.
processed by a machine for as long as 2 hours, compared to the
7 or 8 minutes needed to complete a whole blood donation.
Plateletpheresis donors are able to donate more frequently Special Donations
(up to 24 times per year) than are whole blood donors (regu-
Autologous Donation
lar whole blood donors may only donate every 56 days). For
these reasons, positive relationships between platelet donors Autologous blood donation is blood donated for the
and blood center staff appear to play a more important role in donor’s own use, usually in preparation for an upcoming
plateletpheresis donor retention. These donors are generally elective surgery. The major impetus for autologous dona-
recruited from the ranks of repeat whole blood donors and tion is the donor’s perception of eliminating the risk of
tend to be quite steadfast and reliable. transfusion-transmitted viral disease, particularly HIV and
Mobile blood drives are the ultimate in convenience for hepatitis. Recognition of transfusion-transmitted HIV in
the blood donor. The donor room is essentially transported the mid-1980s greatly increased the utilization of autolo-
to the donor. The mobile blood collection team generally gous donation, which was used less frequently before that
arranges mobile blood drives with a sponsoring organiza- time. Another benefit of autologous donation is minimiza-
tion, often a business, school, hospital, public service orga- tion of exposure to allogeneic red cells and leukocyte anti-
nization, religious group, or military installation. Although gens that may stimulate alloantibody formation and create
it is generally easier and more cost effective to run a fixed future transfusion compatibility problems. Some literature
site, the convenience of a mobile drive brings many other- also suggests that allogeneic blood transfusion can lead to
wise “unavailable” blood donors into the system. Once these modulation of the recipient’s immune system.22–27
mobile site donors have had a positive and successful blood Because the autologous donor is also the patient who will
donation experience, it is often possible to bring them to a receive the donated product, deferral criteria are less strin-
fixed site for further donations, with effective and continuous gent than for allogeneic blood donation. For example, the
recruitment techniques. autologous donor can donate every 72 hours (and typically
An adequate area must be provided by the sponsoring no less than 72 hours before surgery), rather than at an interval
organization for the mobile team to set up. An experienced, of at least 56 days. Similarly, the minimum hemoglobin level
BLOOD DONATION AND COLLECTION
is lowered from 12.5 to 11 g/dL for autologous donors. When Autologous transfusion is not risk-free, so autologous
multiple autologous units are requested, it is best to begin units should never be transfused simply because they are
donation a few weeks in advance of the upcoming surgery. In available. However, individual clinicians’ thresholds for trans-
some cases, the donor is given supplemental iron or erythro- fusion of autologous blood may be somewhat lower than for
poietin injections to maintain hemoglobin levels during the allogeneic blood transfusions.31,32 Bacterial contamination
autologous donation process.28 remains a risk with autologous units, and clerical errors may
Collections staff who evaluate and draw autologous blood cause an autologous unit with positive infectious disease
donors must have more extensive training than those who han- markers to be transfused to an unintended recipient.33
dle only routine donations. This is in part due to the fact that Excessive wastage of unused autologous blood is often an
autologous donors tend to be older than allogeneic donors, issue, because unused units are very rarely, if ever, given to
resulting in more age-related health problems (which may thus other patients (i.e., “crossed over”). These units are allowed
increase the incidence of serious adverse reactions at the time to expire at the hospital and must be discarded. “Cross-over”
of donation). The frequency of severe donor reactions requir- has been discouraged, in part, because, as a group, autolo-
ing hospitalization, although quite low for all donors, is signifi- gous donors have a higher frequency of infectious disease
cantly higher among autologous donors than allogeneic donors markers than regular allogeneic donors.34 They may also
(1 in 17,000 versus 1 in 200,000).29 Blood center staff must also have underlying disease conditions that would make them
take into consideration the disease processes that made the unacceptable as donors for allogeneic blood transfusion.
elective surgeries necessary in the first place. Cardiac patients, Another factor making autologous units less desirable for
for example, may have arrhythmias or symptoms of vascular allogeneic transfusion is the lower hematocrit acceptable for
disease. Orthopedic patients often have mobility problems that autologous donation, which does not meet allogeneic cri-
would adversely affect their donation experience. Blood collec- teria and may provide a substandard (less potent) red cell
tion staff screeners should be especially mindful of identifying product.
those autologous donors who are at risk for ischemic heart dis- Because modern screening and testing methodologies
ease, cerebrovascular disease,30 and seizures. reduce the risk of transfusion-transmitted disease, the pri-
The donor history form is typically abbreviated for autol- mary medical indications for autologous donations have
ogous donation, insofar as risk factors for infectious disease been reduced; however, these donations are still often medi-
transmission are concerned. Acceptability criteria for autol- cally indicated, particularly for patients who have a rare
ogous donation often differ from routine allogeneic dona- blood type. Autologous donation is also beneficial as a
tion: there is a far broader list of health problems that make a means of supplementing the blood supply and does provide
donor acceptable for autologous donation that would neces- a degree of psychological benefit to patients who fear trans-
sitate deferral for allogeneic or routine directed donation. fusion-transmitted disease. Autologous donation may also
Bacterial contamination of the blood product remains introduce repeat donors into the system; however, the pro-
a risk, even for the autologous donor. Individuals with evi- cess tends not to be cost effective, as measured by traditional
dence of bacterial infection should be deferred from dona- cost-benefit estimations.35 Autologous donations will likely 11
tion until the condition is resolved. Blood collection staff continue to decrease in popularity, unless a frightening new
screeners should question the autologous donor regarding transfusion-transmitted pathogen, such as HIV, is discovered 161
signs or symptoms of infection (e.g., fever and antibiotic in the blood supply in the future.
use), indwelling catheters, and open wounds. Donors who Infectious disease testing of autologous blood and trans-
have had recent procedures that could lead to a transient fusion of units with positive infectious disease markers is
bacteremia, such as recent dental work or colonoscopy, are controversial. If the blood is collected in a hospital-based
typically deferred for at least 24 hours. donor room for use in that hospital only, infectious disease
Improved screening and infectious disease testing have testing is not mandated. Autologous blood drawn at a com-
significantly minimized the infectious disease risks of allo- munity blood center, however, must be fully tested (as for
geneic transfusion. However, many donors and physicians allogeneic units). Autologous units positive for infectious
continue to request autologous donation as a transfusion disease must be labeled with biohazard stickers.36,37 The
option. Autologous donations require more complicated AABB Standards require that if an autologus unit is to be
donor screening and collection procedures, associated logis- shipped to another facility and the unit tests positive for
tical problems, and associated higher costs. The autologous any marker of transfusion-transmitted disease, the shipping
unit must be specifically labeled for the designated patient, facility shall notify the receiving trasfusion service.38 It is the
and systems must be in place in both the blood center and prerogative of the hospital transfusion service whether to
the hospital that will guarantee that the blood arrives at the accept autologous blood components that are positive for
proper place, in the right condition, and in time for sur- infectious disease(s).
gery. Occasionally, autologous blood is not available for use, Some transfusion services agree to store and transfuse
due to surgery being delayed beyond the expiration of the autologous units that are confirmed positive for HIV, or
donated blood components or due to failure of proper com- hepatitis B or C (HBV, HCV). Evidence presented by the
munication between the collection center and the hospital College of American Pathologists (CAP) indicates that
staff. Positive infectious disease testing or clerical errors may many transfusion services either do not test autologous
also delay availability of autologous blood. At the hospital, blood for infectious disease markers or knowingly collect,
care must be taken to transfuse autologous blood before store, and transfuse infectious units.39 Although transfu-
allogeneic or directed donor blood. If an adverse effect is sion of these infected units may not present an obvious
attributed to an allogeneic or directed unit that, arguably, risk to the donor/patient, accidental needle-sticks and
would never have been transfused had the autologous unit splatters do put blood handlers at risk. Accidental trans-
been available for use and transfused first, medicolegal fusion of an infected autologous unit to the wrong recipi-
consequences may ensue. ent is possible. Storage of infectious blood components
BLOOD BANKING in hospital blood banks also presents some risk to other pressure to donate by friends and family members, might
patients, considering that at least 1 in every 25,000 blood not be truthful about risk factors for infectious disease.
products is transfused to the “wrong” individual. 40 In These donors would, therefore, present an increased risk of
1992, the CAP conducted a survey of 3852 hospital trans- infectious disease transmission to the recipient. The other
fusion services and found that 34 (0.9%) had issued one way of thinking suggested that individuals would be more
or more autologous blood products to the wrong patient careful about admitting potential risk factors when making
during the previous year, and that 20 of these units were such donations. Eventually it became evident that directed
actually transfused.40 An analysis of 256 licensed trans- donations are likely to be as safe as most first-time blood
fusion services by The New York State Department of donations, but not as safe as donations from repeat donors
Health, from 1990 through 1998, indicated that 1 in who have a history of safe donations.48 Although there is no
19,000 RBC units where transfused to the wrong patient evidence that directed donations are safer than routine vol-
or were of incorrect ABO group or Rh type. 41 In addition, unteer donations, the practice often does provide a psycho-
preliminary data indicate the frequency of infectious dis- logical sense of well-being for the patient and may alleviate
ease markers among autologous donors is significantly the feelings of helplessness that occur when a loved one is
higher than that of allogeneic donors (Table 11–2). This suffering from health care problems.
data, along with the decreasing benefits of autologous Blood from directed donors is collected and tested in
transfusion due to improved infectious disease testing of accordance with the same criteria that is in place for alloge-
allogeneic blood, make the practice of storing and trans- neic donations and hence can be “crossed-over” and used by
fusing infected units less attractive to hospital transfu- other patients when not required by the original intended
sion services. recipient. It is the choice of the hospital transfusion service
One possible reason why many transfusion services per- whether to utilize the practice of “crossing over.” This option,
mit storage of infectious autologous units is for fear of legal however, is important to recognize, because the blood types
action based on the Americans with Disabilities Act, which of the donor and/or intended recipients are often not known
affords to asymptomatic individuals infected with HIV a at the time of donation; thus, incompatible directed dona-
protected class status.42–46 There is a concern that not offer- tions are not uncommon. These units can be transfused to
ing autologous services to these donor/patients might be other patients, improving the overall blood supply. The prac-
interpreted as a violation of this act.47 tice of directed donation is also a valuable means of getting
donors into the system, because a sizable number of these
Directed Donations
donors go on to become repeat allogeneic donors. Rather
A directed donation is a blood donation made specifi- than creating a two-tiered system, as was initially feared,
cally for use by a designated patient. Directed donations directed donations tend to increase the amount of blood
are usually made by friends and family members of the available for all patients.
patient. These donations are typically manufactured into Directed donation presents a series of logistical problems
II RBCs; however, directed plateletpheresis donations are not not present in allogeneic donation. A physician’s order must
uncommon. Using new apheresis technology, a combina- be in place indicating the number and type (e.g., platelets,
162 tion of red cells and platelets (or plasma) can be donated RBCs) of directed units required. The blood types of the
in one sitting. directed donors and intended recipients are often incom-
Directed donation was initially discouraged by most patible. Additionally, directed units may not be available at
blood centers, for fear that the practice would institute the time of need, because the intended directed donor was
an inequitable two-tiered blood system in which well- unable to donate due to fear, time constraints, or exclusion-
connected patients would have access to a safe and ade- ary health conditions. Fully screened and motivated donors
quate blood supply while less fortunate patients might may be unable to donate due to inadequate venous access
have none. However, the discovery of HIV in the blood or technical errors. Directed donations testing positive for
supply in the early 1980s created so much demand that infectious disease are discarded. For these reasons and per-
today directed donations have become a routine part of haps others, directed donations may not be available for use
blood donation. as expected by the patient.
There were two schools of thought in the early days of Communication among donors, patients, clinicians, the
directed donations. One suggested that individuals, under blood center, and the hospital transfusion service is critical
HBV, hepatitis B virus; HCV, hepatitis C virus; HIV, human immunodeficiency virus.
*Systemwide collection data from American Red Cross for calendar year 2004. Personal
communication from Edward P. Notari IV, M.P.H, American Red Cross, Jerome H. Holland
Laboratory, ARCNET Data Center.
†
Wang B, Schreiber GB, Glynn SA, et al. Retrovirus Epidemiology Donor Study: Does
prevalence of transfusion-transmissible viral infection reflect corresponding incidence in United
States blood donors? Transfusion 2005;45:1089–1096.
BLOOD DONATION AND COLLECTION
for a successful directed donation program. A system must viated physical evaluation of donor blood pressure, pulse,
be in place to allow the patient and attending physician to temperature, and venous access. The donor’s hematocrit or
know how many directed donor units are available for trans- hemoglobin levels are also evaluated at this time.
fusion, so that more donors can be recruited if necessary. Donor screening and blood collection must be conducted
Good communication and successful procedures for directed under specific rules found in the CFR, as well as in applicable
donation programs avoid last-minute misunderstandings, in FDA guidelines and memoranda. In addition, the AABB, the
circumstances in which the anticipated number of directed preeminent nongovernmental organization involved with
units is not available when needed. transfusion medicine in the United States, issues a publica-
tion, the Standards for Blood Banks and Transfusion Services
Medically Indicated Directed Donations (Standards),52 which is adhered to by the majority of American
Directed donations are not safer than allogeneic donations, but blood centers and has been adopted into law, in varying degrees,
they do increase the blood supply and provide a sense of secu- by many states. The AABB Standards are upgraded regularly, to
rity to the recipients. Most directed donations are not clinically keep pace with current trends in transfusion medicine and the
necessary. However, circumstances do exist that require directed most recent federal regulations. Websites for the AABB (www.
donations or in which a directed donation offers medical ben- aabb.org) and the FDA (www.fda.gov/cber) are good sources
efit. Using the same directed donor to provide small volumes for the most up-to-date transfusion-related regulations and
of blood at regular intervals to neonates should reduce the risk information. Additionally, state and local regulations regard-
of transfusion-transmitted diseases that would presumably be ing blood collection practices often apply. Qualifying donor
present with the use of multiple donors.49 Similarly, one can requirements, as stipulated in the most recent edition of the
use a small group of donors for chronically transfused patients AABB Standards, are listed in Table 11–3.
(e.g., patients with sickle cell anemia or thalassemia).50 Patients
requiring rare blood types also benefit from specific directed Donor Identification and the Deferred
donations, often from a blood relative. With proper authoriza- Donor Registry (DDR)
tion, the frequency of these medically indicated donations can Proper identification, often photographic, is required to
be increased beyond that which would be acceptable for rou- confirm the donor’s identity before donation.53 This infor-
tine donation. In these instances, the slight potential donor risk mation is important if it becomes necessary to track down
is offset by the benefit to the recipient. and notify the donor of any positive infectious disease test
HLA-Matched Platelet Donors results. Proper identification is also required to perform a
“lookback” study, to investigate whether a donor may have
Some blood collection centers test their plateletpheresis transmitted an infectious disease, unknown at the time of
donors for human leukocyte antigen (HLA) type and store donation, to a blood recipient. Donors are asked if they have
the data in a computerized database to have a readily avail- ever donated under any other name, possibly a maiden name
able pool of donors to treat patients who have developed or nickname, which would make it difficult to confirm previ-
anti-HLA alloantibodies and require HLA-compatible plate- ous donations. Correct personal identifiers are also necessary 11
lets. The advent of platelet crossmatching techniques and to calculate if adequate time has passed between donations.
the reduced frequency of alloimmunization, possibly due to Computers are becoming a mainstay of donor screen- 163
leukoreduced blood products, have made the availability of ing and tracking, but they can only work properly if sup-
a large HLA-typed donor pool less necessary than in previ- plied with accurate information. As an added precaution, the
ous years. However, orders for HLA-matched plateletphere- donor’s name is compared against a database of individuals,
sis products are still made, and most blood centers still offer the deferred donor registry (DDR), who have been disquali-
this option to their hospital customers. fied from donating in the past, usually due to a positive infec-
tious disease marker.54 This database can be maintained with
Donors with Hemochromatosis
computers or by using manual methods such as microfiche.
Therapeutic phlebotomy is an accepted modality for pre- During the early days of the HIV epidemic in the early and
venting iron overload and subsequent organ damage for mid-1980s, the DDR was instituted as a precaution against
patients with hereditary hemochromatosis. Some blood col- individuals falsifying information to donate blood to obtain
lection centers, in the United States and elsewhere, have used a free HIV test. The use of the DDR is still in effect today.
units collected from individuals with hemochromatosis for
allogeneic transfusion. These donors must meet all other
allogeneic criteria. Recently the FDA has sanctioned this pro-
The Donor History Questionnaire (DHQ)
cess by allowing variances for blood centers to collect blood The donor history questionnaire is an extensive series of ques-
from these individuals, provided certain donor follow-up tions, often quite personal, designed to minimize the chance of
and other stringent criteria are met.51 adverse consequences to the blood donor and ensure a safe and
potent blood product for the recipient. Questions are typically
phrased in a “yes-no” format, other than the few open-ended
Donor Screening
questions regarding health care problems. The questionnaire
Blood donors are carefully screened to minimize the risk of must comply with requirements of the CFR and Standards.
adverse consequences to the donor and to the recipient The AABB has developed a questionnaire that fulfills these
of the transfused blood. The screening process is made up of requirements, which has been adopted, to some extent, by
two distinct steps. The first is the donor history question- most blood centers in the United States (Table 11–4).55
naire (DHQ), a series of questions designed to expose A blood collection staff screener is required to answer any of
potential health problems that might lead to adverse effects the donor’s questions and makes sure the forms are accurate and
to the donor or blood recipient. The second step is an abbre- complete. It is crucial that the DHQ be completed properly: a false
BLOOD BANKING
Table 11–3 AABB Standards Requirements for Donor Qualification*
*
Reference Standard 5.4.1A- Requirements for Allogeneic Donor Qualification.
†
FDA Guidance for Industry, January 9, 2002. Revised Preventative to Reduce the Possible Risk of Transmission of Creutzfeldt-Jakob disease
(CJD) and variant Creutzfeldt-Jakob disease (vCJD) by Blood and Blood Products.
‡
AABB Association Bulletin 05–11. Interim Standard for Standards for Blood Banks and Transfusion Services (23rd edition). Sept. 30, 2005.
BLOOD DONATION AND COLLECTION
Table 11–3 AABB Standards Requirements for Donor Qualification—Continued
§
FDA Memorandum, April 23, 1992, Revised Recommendation for the Prevention of Human Immunodeficiency Virus (HIV) Transmission by
Blood and Blood Products.
||
FDA Memorandum, December 11, 1996. Interim Recommendations for Deferral of Donors at Increased Risk for HIV-1 Group O Infection.
#
FDA Guidance for Industry, June 2005, Assessing Donor Suitability and Blood and Blood Product Safety in Cases of Known of Suspected West
Nile Virus.
**
The Department of Defense has recommended a 24-month deferral. Department of Defense Memorandum, October 14, 1999, “Deferral of
Service Members Stationed in Possible Malaria Areas in the Republic of Korea,” and February 28, 2001 update.
††
http://www.cdc.gov/travel
or missing entry must be corrected before the blood is released since the last donation. Data presented to the FDA Blood 11
for transfusion. For additional clarity, screeners may be required Products Advisory Committee on March 18, 2005, showed that
to confirm certain critical questions verbally.56 For autologous a significant number of donors desire faster processing with a 165
donors, who may have special health problems, it is often wise to less complicated interview. This data demonstrated no indica-
have a well-trained registered nurse participate in the screening. tion that the abbreviated questionnaire increases blood safety
Other than the “yes” or “no” questions, the DHQ uses “cap- risk.58 At this time, the FDA has not accepted the aDHQ and has
ture questions” that cover a variety of broad topics. When an requested that the AABB Donor History Task Force develop a
affirmative answer is given to a particular question, additional pre-implementation study of the aDHQ (currently in progress).
follow-up questions are asked by the screener to obtain addi- The minimum age for blood donation is typically age 17, but
tional information. For example, the question “Have you ever laws vary from state to state. Collections teams must follow local
had any type of cancer, including leukemia?” often serves as a regulations and make certain proper consent is obtained. In some
capture question that would elicit further information. states, parental notification and/or consent may also be necessary.
Additionally, to ensure that donors who self-administer a
Donor History Questionnaire and Donor Safety
paper DHQ maintain focus, several “attention” questions are
included. They serve to indicate if a donor is actually paying For the majority of blood donors, the blood volume lost
attention to the DHQ questions. The following is an example at donation is restored within 48 to 72 hours. With nor-
of one of the attention questions: mal vascular elasticity, blood pressure is maintained and
In the past 6 weeks, have you been pregnant or are you adverse reactions are kept at a minimum. Experienced
pregnant now? (Males check “I am male.”) screeners are more conservative with individuals who have
An inappropriate answer to the question would be a male a history of hypertension, diabetes, atherosclerosis, or other
answering “yes” or “no.” Each blood center must define the vascular diseases that can interfere with the normal physi-
action of the screener when a donor inappropriately answers ologic response to acute blood loss, which might precipi-
the attention questions. Attention questions may not be tate a hypotensive vasovagal reaction. Although rare, an
necessary when using other techniques to assure donor focus, acute drop in blood pressure could precipitate symptoms
such as an audiovisual computer-assisted self-interviewing of otherwise occult coronary artery or cerebrovascular dis-
system. In recent years, methods of computerized data entry ease. The incidence of these disorders increases with age,
have become more common.57 so it is wise to be cautious with elderly donors or smaller
Several blood centers have investigated the use of an abbre- donors with lesser blood volumes, for whom the acute loss
viated donor history questionnaire (aDHQ) for repeat donors. of a pint of blood provides relatively more severe strain to
The aDHQ eliminates nonrepeatable event questions; identifies the circulatory system.
recent changes in health, travel, or behavior; and retains ques- Open-ended questions about a donor’s general state of
tions about risk-associated activities that might have changed health, medications, previous surgeries, or current health care
BLOOD BANKING
Table 11–4 AABB Full-Length Donor History Questionnaire, Version 1.1, June 2005*
Yes No
Are you
1. Feeling healthy and well today? ❒ ❒
2. Currently taking an antibiotic? ❒ ❒
3. Currently taking any other medication for an infection? ❒ ❒
Please read the Medication Deferral List.
4. Are you now taking or have you ever taken any medications on the ❒ ❒
Medication Deferral List?
5. Have you read the educational materials? ❒ ❒
In the past 48 hours
6. Have you taken aspirin or anything that has aspirin in it? ❒ ❒
In the past 6 weeks
7. Female donors: Have you been pregnant or are you pregnant now? ❒ ❒ ❒ I am male
(Males: check “I am male.”)
In the past 8 weeks have you
8. Donated blood, platelets, or plasma? ❒ ❒
9. Had any vaccinations or other shots? ❒ ❒
10. Had contact with someone who had a smallpox vaccination? ❒ ❒
In the past 16 weeks
11. Have you donated a double unit of red cells using an apheresis machine? ❒ ❒
In the past 12 months have you
12. Had a blood transfusion? ❒ ❒
13. Had a transplant such as organ, tissue, or bone marrow? ❒ ❒
14. Had a graft such as bone or skin? ❒ ❒
15. Come into contact with someone else’s blood? ❒ ❒
16. Had an accidental needle-stick? ❒ ❒
17. Had sexual contact with anyone who has HIV/AIDS or has had a positive ❒ ❒
test for the HIV/AIDS virus?
18. Had sexual contact with a prostitute or anyone else who takes money or ❒ ❒
drugs or other payment for sex?
19. Had sexual contact with anyone who has ever used needles to take drugs ❒ ❒
or steroids, or anything not prescribed by their doctor?
20. Had sexual contact with anyone who has hemophilia or has used clotting ❒ ❒
II factor concentrates?
21. Female donors: Had sexual contact with a male who has ever had sexual ❒ ❒ ❒ I am male
166 contact with another male? (Males: check “I am male.”)
22. Had sexual contact with a person who has hepatitis? ❒ ❒
23. Lived with a person who has hepatitis? ❒ ❒
24. Had a tattoo? ❒ ❒
25. Had ear or body piercing? ❒ ❒
26. Had or been treated for syphilis or gonorrhea? ❒ ❒
27. Been in juvenile detention, lockup, jail, or prison for more than 72 hours? ❒ ❒
In the past three years have you
28. Been outside the United States or Canada? ❒ ❒
From 1980 through 1996,
29. Did you spend time that adds up to three (3) months or more in the ❒ ❒
United Kingdom? (Review list of countries in the UK)
30. Were you a member of the U.S. military, a civilian military employee, ❒ ❒
or a dependent of a member of the U.S. military?
From 1980 to the present, did you
31. Spend time that adds up to five (5) years or more in Europe? (Review list ❒ ❒
of countries in Europe.)
32. Receive a blood transfusion in the United Kingdom? (Review list of ❒ ❒
countries in the UK.)
From 1977 to the present, have you
33. Received money, drugs, or other payment for sex? ❒ ❒
34. Male donors: had sexual contact with another male, even once? ❒ ❒ ❒ I am female
(Females: check “I am female.”)
Have you EVER
35. Had a positive test for the HIV/AIDS virus? ❒ ❒
36. Used needles to take drugs, steroids, or anything not prescribed by ❒ ❒
your doctor?
37. Used clotting factor concentrates? ❒ ❒
38. Had hepatitis? ❒ ❒
39. Had malaria? ❒ ❒
40. Had Chagas disease? ❒ ❒
*
Final Guidance from FDA not yet released on this version. Current version is 1.0 from April 23, 2004.
BLOOD DONATION AND COLLECTION
Table 11–4 AABB Full-Length Donor History Questionnaire, Version 1.1, June 2005—Continued
Yes No
problems serve to elicit potential risk factors that require careful transmitted viral disease, the window period (the time during
evaluation prior to donation. Additional questions target specific which a recently infected donor is infectious but tests negative
medical conditions such as cardiac, lung, liver, and blood dis- for infectious disease markers) contributes to a low-level risk.
eases; pregnancy; and cancer. A detailed discussion of the many The introduction of sensitive nucleic acid amplification tech-
disease states that would affect donation is beyond the scope of nology (NAT) has shortened the window period for HCV and
this chapter; however, cardiovascular disease, cerebrovascular HIV, when compared to previous HCV and HIV antibody or
disease, and seizure disorders are some of the primary reasons HIV p24 antigen testing. However, no matter how small the
for which deferral may be indicated. Many properly controlled risk, it is unlikely that testing will ever detect every infected
medical problems, such as thyroid disease, hypertension, mild blood donor, so reliance on screening cannot diminish.
seizure disorders, diabetes, and certain heart conditions, such as Some behaviors, which have been associated with
mitral valve prolapse, may not interfere with blood donation. It increased risk for HIV or hepatitis infection, result in indefi-
is good policy to refer difficult cases to the blood center’s medi- nite deferral. These include intravenous drug use and pros-
cal director for a final decision as to whether it is safe for the titution. Male homosexual or bisexual activity, often defined
donor to donate blood. If it is not clear whether an individual as “men having sex with other men, even once, since 1977,”
meets criteria for donation, it is generally wise to err on the side is also considered high-risk behavior and results in indefinite
of donor safety and defer the donor. deferral from allogeneic donation. Other behaviors, such as
Medications are rarely of significance from the aspect having sexual contact with a prostitute or an intravenous
of donor safety. Although angiotensin-converting enzyme drug user, require a 12-month deferral from the last contact.
inhibitors have fostered concern about potential hypoten- The DHQ is the first line of defense against such patho-
sive episodes, medications often serve to alert the screener to gens as malaria and the agents responsible for Chagas disease 11
health problems that otherwise may have been inadvertently and Creutzfeldt-Jakob disease (CJD), for which no practical
left out of the donor history. Use of a coronary artery dila- screening tests are now available in the United States.59,60 167
tor, for example, would indicate a history of ischemic heart Donor centers should have information to indicate areas
disease, which might have gone unmentioned. Experienced where malaria is endemic.61 Potential donors who have trav-
screeners are often impressed by the lack of information eled to an endemic area are deferred for 12 months from the
that many individuals have about their own health his- date of return. Donors with a history of malaria are deferred
tory. Some pre-existing disease states are not mentioned for 3 years if symptoms do not recur.
by donors when completing the DHQ, as are medications Transmission of variant CJD (vCJD) through blood trans-
and the reasons for which the medications are being taken. fusion has been a controversial topic in recent years.62, 63 There
This may be a matter of denial or may be symptomatic of a has been documented disease transmission of classic CJD
language problem. The latter is of increasing significance as through dura mater grafts, pituitary-derived human growth
our donor population becomes more diverse. hormone, and ineffectively sterilized electroencephalogram
electrodes. Animal studies and a few reported cases in humans
Donor History Questionnaire suggest strongly that vCJD can be transmitted by blood trans-
and Recipient Safety fusion. Although the risk seems to be very low, the magnitude
Most significantly, the DHQ exists to protect the transfu- is unknown, due to the long (e.g., 30-year) incubation period.
sion recipient. The driving force for much of the DHQ Those stricken with vCJD by eating contaminated meat prod-
involves screening for transfusion-transmissible infectious ucts during the 1990s became symptomatic and died in just a
diseases. The greatest danger exists for diseases in which an few years.
undetected, asymptomatic carrier state exists at the time of To deal with this difficult problem, geographical screen-
donation. Often there are geographic or behavioral indica- ing is now in place to exclude donors who have spent time
tors that place a donor at increased risk for transmitting in countries where cases of vCJD have been known to occur,
these diseases. The DHQ seeks to identify these risk factors including the United Kingdom (UK) and much of Europe.
to reduce the chance of disease transmission. Potential donors are indefinitely deferred who have spent a
Viral diseases are the most tested for transfusion-transmit- cumulative 3 months in the United Kingdom between 1980
ted pathogens. Most notorious among these is HIV, which and 1996, the period when unsafe cattle feeding practices led
devastated the blood supply in the early to mid 1980s. HBV to an outbreak of bovine spongiform encephalopathy, or “mad
and HCV (once called non-A, non-B hepatitis) have also caused cow” disease. Ingestion of infected beef at that time is believed
significant morbidity and mortality in transfusion recipients. to responsible for a number of cases of human vCJD.64 Also
Although sophisticated infectious disease testing method- indefinitely deferred are those who spent a cumulative 6
ologies have significantly reduced the rate of transfusion- months on European military bases or a cumulative 5 years
BLOOD BANKING in certain European countries other than the UK, as are those Vaccination with the vaccinia virus requires a minimum
who received a blood transfusion in the UK from 1980 to the of a 21-day deferral; complicated additional criteria have
present.65 Recipients of pituitary-derived growth hormone are been established related to vaccine-related complications
deferred as are individuals with known exposure to vCJD or and whether the scab separates spontaneously. There are
CJD. Until practical mass serological screening becomes avail- also deferrals for individuals who may have come in con-
able, geographical exclusions may stay in place and thousands tact with a vaccine recipient. For more information, see
of otherwise eligible donors will be excluded. FDA Guidance (http://www.fda.gov/cber/gdlns/smpoxde-
Exclusionary criteria are changed and updated regularly fquar.htm#iv).
based on new threats to the blood supply. Deferral criteria Medications taken by potential donors present a sig-
may change based on improved serologic testing, and better nificant concern for the donor screeners. Medication defer-
understanding of the disease processes involved. Criteria to rals fall into three major categories: those that might have
avoid severe acute respiratory syndrome66 were in effect at adverse effects on the blood recipient, those that are taken
one time, and Gulf War veterans were deferred from blood for a medical condition that might make donation unaccept-
donation from 1991 to 1993 to avoid the transmission of able, and those that would reduce the effectiveness of a blood
leishmaniasis.67–70 Donors with a history of blood-borne product.
parasites such as babesiosis and Chagas disease are permanently Some drugs are teratogenic and could cause birth defects
deferred. if transfused to pregnant women. These medications include
Diseases that present with severe clinical symptoms, such finasteride (Proscar, Propecia) and isotretinoin (Accutane),
as hepatitis A, tend not to require special screening, because each of which requires a 1-month deferral after the last dose.
the victims are generally too sick to donate. Standards allow Leflunomide (Arava) and dutasteride (Avodart) require
prospective donors who have a history of hepatitis before age 3-month and 6-month deferrals, respectively. Etretinate
11 to be eligible for donation, provided no other cause for (Tegison) requires an indefinite deferral. Drugs that might
deferral exists. transmit infections, such as human pituitary-derived growth
A number of DHQ questions deal with hematologic dis- hormone, which is associated with CJD, are cause for indefi-
ease, leukemia, and previous use of clotting factors. These nite deferral. In the 1980s, prior to improved screening and
conditions may cause abnormalities in RBCs, platelets, or purification processes, hemophiliac recipients of pooled
plasma proteins that may lead to substandard blood products clotting factor concentrates were at very high risk for trans-
being produced. mitting HIV and hepatitis.
Exposure, or even potential exposure, to another indi- Antibiotic use may indicate an active bacterial infection.
vidual’s blood requires a 12-month deferral. The rationale Associated subclinical bacteremia may result in a contami-
for the 12-month deferral period is that the vast major- nated blood product, which could cause serious, even fatal,
ity of transfusion-transmitted infectious diseases would consequences to the recipient.
manifest positive serologic markers within 1 year’s time. Medications may interfere with the quality of certain
II Besides blood transfusion and accidental needle-sticks, blood products. This is especially true of medications that
other sources of blood exposure include human bites and inhibit platelet function, such as aspirin. Platelet donors
168 acupuncture, tattoos, and piercings performed with non- must have not taken aspirin within 36 hours of donation.
sterile instruments. Nonsterile body piercing (including Other medications interfering with platelet function include
ears) has become increasingly problematic in recent years clopidogrel (Plavix) and ticlopidine (Ticlid).
due to the increase in popularity of piercings and tattoos. The CFR and AABB Standards provide some specific
The screener should ask if the procedure was done using guidelines with which blood collections facilities must adhere,
sterile techniques. but it is impossible to attempt to provide specific guidance
Vaccinations provide another area of concern for donor for all situations. This most often becomes an issue when
screening. This is particularly true for live-attenuated viral evaluating donors with underlying health care problems. Not
vaccines, which could theoretically infect immunocompro- only might a particular condition require medical director
mised individuals. Recipients of rubella and varicella zos- consideration, but often the degree of clinical severity of that
ter vaccines are deferred for 4 weeks. A 2-week deferral is condition must also be considered.72 For example, an active
required of recipients of rubeola (measles), polio (Sabin/oral), case of rheumatoid arthritis may require deferral, whereas a
mumps, typhoid (oral), and yellow fever vaccines. Individuals history of rheumatoid arthritis may not. In these instances,
vaccinated for exposure to HBV or rabies are deferred for 12 a collection center must develop its own procedures and
months to avoid the remote possibility of disease transmission. criteria. Although not every circumstance can be dealt with
The American Red Cross requires a 7-day deferral period for in a comprehensive manner, some centers have developed
routine (not exposure-related) HBV vaccination; the deferral comprehensive procedures to standardize this as best as pos-
time varies among different organizations. Vaccination with sible.73 It is occasionally necessary, however, particularly for
nonviable agents such as toxoids and nonviable antigenic autologous donations, to have a properly credentialed medi-
material requires no deferral. Donors who are vaccinated cal director review ambiguous situations and make informed
with experimental vaccines should be carefully evaluated and decisions on a case-by-case basis. The medical director may
deferred for at least 12 months if there is any doubt as to the also be needed to resolve issues about medications, poten-
safety of the agent.71 tial risk factors, and any other circumstance where a medical
The September 11, 2001, terrorist attacks on the doctor’s decision is needed.
United States and subsequent attacks of anthrax, real or
Donor Consent and Additional Information
hoax, through the postal system created concern about
bioterrorism. The United States’ vulnerability to a small- It is necessary to obtain informed consent prior to dona-
pox attack resulted in plans for mass vaccinations that tion. To help ensure that the prospective donor is properly
have caused concern in the blood banking community. informed, educational material about the donation process is
BLOOD DONATION AND COLLECTION
distributed, including information about screening, phlebot- medications are acceptable if the blood pressure is controlled
omy, and potential donation-related complications. The goal and the donor otherwise meets donation criteria.
of this material is for the donor to understand the reasons The prospective donor’s temperature should be less
for self-deferral and the importance of self-deferral when than 99.6°F. An elevated temperature may indicate a disease
appropriate. The notification process for positive serologic process that might affect the blood recipient (i.e., bacterial
tests may be explained as well as donor confidentiality issues. contamination of the product) and may require medical
AABB Standards requires review of information about the attention for the donor.
symptoms of AIDS, and the possibility of infectious disease The potential donor’s antecubital fossae are evaluated
transmission through blood transfusion. This reading mate- for acceptable venous access. The skin is checked for rashes,
rial should be in language simple enough that every donor scars, or other lesions that would make phlebotomy unac-
can comprehend it.74 When local demographics demand, it ceptable. The presence of “tracks” indicative of intravenous
may be wise to accommodate donors with materials written drug abuse also leads to deferral.
in a language that they can understand; otherwise, translators Determination of hemoglobin/hematocrit level is an
may be necessary. The donor must acknowledge that these essential part of the donation process. Whole blood collec-
materials were read and that all questions were answered. tion from an anemic donor jeopardizes the donor and pro-
vides a substandard product for the transfusion recipient.
Donor Physical Examination and Hematocrit
Standards and the CFR require that donors have a minimum
A brief physical examination is performed to help ensure the hemoglobin of 12.5 g/dL or hematocrit of 38% or greater.
donor’s suitability for blood donation. The physical exami- Venous blood or finger pricks are common methods for
nation consists of evaluation of the donor’s pulse, blood obtaining blood for hemoglobin/hematocrit determination.
pressure, temperature, and weight. There is also an inspection Earlobe sampling, once a commonly used method, has been
of antecubital fossae as sites of venous access. proven inaccurate and is not longer acceptable according to
Generally, a donor must weigh at least 110 pounds to Standards.77–79 Hematocrit determinations are often done
undergo routine donation. Standards allows donation using the manual microcapillary tube method or a portable
of 10.5 mL of whole blood for every kilogram of donor point of care technology. Some facilities use the copper sul-
weight, but most centers purchase blood bags with a pre- fate method, which relies on the specific gravity of blood rel-
measured amount of anticoagulant/preservative. These ative to copper sulfate, to determine whether a blood sample
have a specified minimum and maximum amount of blood has an adequate hemoglobin level.
that can be collected for the anticoagulant and preserva-
tives to function according to manufacturer’s specifica- Confidential Unit Exclusion
tions. For this reason, the volume of whole blood donated Blood donors with unacceptable risk factors for transfusion-
is fairly constant: approximately 1 pint. Assuring that the transmitted disease might be coerced into blood donation.
donor is of the minimum weight is important because a This situation may arise at an institutional blood drive or
smaller donor, with a smaller blood volume, will suffer with directed donations for a friend or family member. The 11
greater relative stress to the circulatory system. This will donor may deny risk factors, such as homosexuality or drug
increase the likelihood of an adverse reaction, possibly loss abuse, for fear that a breach of confidentiality might allow 169
of consciousness or seizures. Underweight donors are par- this behavior to become widely known. The confidential unit
ticularly likely among younger individuals, who also have a exclusion (CUE) provides the opportunity for a donor to
higher incidence of adverse donation reactions than older request, with confidentiality guaranteed, that unacceptable
individuals.75,76 Although the weight is generally noted donated blood not be used for transfusion. This procedure
from the history, it is wise to actually weigh those donors is accomplished by affording the donor a CUE card contain-
who appear to be underweight. There is no maximum ing the unit number and a means of indicating, if necessary,
weight for blood donation, but the donor’s weight should that the unit not be used for transfusion. The CUE is then
not exceed the maximum capacity of the center’s collection placed in a locked box to be reviewed at a later time, at which
equipment. time excluded units are earmarked for destruction. In addi-
The donor’s pulse must be regular and between 50 and tion, centers can provide a designated telephone number
100 beats per minute. A lower pulse is sometimes acceptable so that a donor can call back with additional information,
in athletic donors, although a medical director’s approval when necessary, that would exclude a unit for transfusion.
may be necessary. A physician should also evaluate first-time When first used, the infectious disease marker frequency
donors with an irregular pulse. Allogeneic donors should be of units designated for nonuse by CUE was significantly
deferred if there is any question about the potential donor’s higher than other units: as many as 20% of anti-HIV posi-
cardiovascular fitness. For autologous donors with irregu- tive units would have been diverted from transfusion due
lar heartbeats, consultation or written permission from the to the CUE.80 Improved donor education, infectious disease
patient’s physician may be necessary. testing, and screening techniques have made the CUE a less
The donor’s blood pressure must be less than 180/100 valuable tool.81–83 Many CUEs are the result of donors mis-
mmHg on the day of donation. First-time donors may understanding the CUE directions and inadvertently checking
present with elevated blood pressure due to donation-related the wrong box.84,85 For this reason, after investigation, some
anxiety. Allowing them a few moments’ rest may lower the centers will destroy an excluded unit but will continue to
blood pressure to acceptable levels. Screeners should also allow the donor to donate blood.86
be wary of low blood pressure in donors of small stature,
advanced age, or with vascular disease, such as diabetes or Infectious Disease Testing
atherosclerosis. These individuals may not tolerate acute Although donor screening plays an undeniable role in main-
blood loss as well as a normotensive donor. Blood pressure taining a safe blood supply, infectious disease testing remains
BLOOD BANKING the gold standard. As of 2005, blood is routinely screened for Collection
the following disease markers:
Prevenipuncture Procedures
Hepatitis B surface antigen (HBsAg)
Hepatitis B core antibody (anti-HBc) The collection process begins with accurate identification of
Hepatitis C virus antibody (anti-HCV) the blood donor. This is especially important in large centers
HIV-1 and HIV-2 antibody (anti-HIV-1 and anti-HIV-2) where screening and phlebotomy are done in separate areas
HTLV-I and HTLV-II antibody (anti-HTLV-I and anti- and by different staff. A unique identification number is
HTLV-II) placed on the collection bags, paperwork, and the pilot tubes
Nucleic acid amplification testing (NAT) for HIV-1 and HCV collected for serologic testing. The phlebotomist applies mild
NAT for West Nile virus (WNV) pressure over the upper arm, usually with a blood pressure
Serologic test for syphilis cuff or tourniquet. The increased venous pressure engorges
the veins in the anticubital fossa, making them easier to
Although hepatitis plagued the blood supply for many detect for phlebotomy. Once a vein is selected, the skin is
years, it was not until the mid-1980s and the discovery that thoroughly disinfected, often using a two-step procedure
HIV was transmitted through blood transfusion that blood utilizing soap and iodine solutions.95 After the skin has been
safety became a national obsession. Currently, the risk of disinfected, the phlebotomist performs the venipuncture
transfusion-transmitted HIV may be as low as 1 in 2 mil- and the collection begins.
lion units.87 Despite this, HIV is the most feared transfusion-
transmitted disease. Whole Blood Collection
Risk of HBV contamination may be as low as 1 in 100,000. Whole blood is collected by means of venous pressure and
NAT testing for HBV is still not in widespread use, although gravity. Usually, the phlebotomy needle comes attached to a
it is under consideration.88,89 The debate, in part, consists of preconfigured bag system containing a premeasured amount
whether it is more cost effective to spend public funds on of anticoagulant and preservative. A number of different
HBV vaccination programs or for NAT testing of the blood anticoagulant/preservative preparations are available. The
supply, and whether NAT testing would be more sensitive for maximum liquid storage time for any RBC unit is currently
“window period” detection of donors who have lost HBV 42 days. The number of bags in the collection set depends on
antigenemia.90,91 intentions for further manufacture: whole blood, packed cells
The anti-HBc test is somewhat controversial.92 It was and plasma, or packed cells, plasma, and platelets. Multiple
once considered a surrogate marker for HIV, but improved small bags can be attached if the blood is designated for
serologic testing for HIV would seem to make anti-HBc pediatric transfusion. It is possible, through a sterile docking
unnecessary. The high false-positive rate for this marker fur- device, to add additional bags to a set. It is also possible to
ther limits its usefulness. However, its proponents suggest manually adjust the amount of anticoagulant in a collection
that anti-HBc detects those donors who remain infectious of bag for an underweight donor, but this requires significant
II HBV who have lost HBsAg positivity. time and expertise and is not done in most centers.
Hepatitis C virus was once considered responsible for a The collection bag is often placed on a trip-scale, which
170 transfusion-transmitted hepatitis rate of up to 10%. Current impedes further blood flow once the desired amount (usu-
estimates of HCV transmission may be as low as 1 in 2 mil- ally 450 or 500 mL) has been drawn. The blood is agitated
lion units. during collection, either manually or with an automated
The test for antibodies to HTLV is also controversial.93 device, to ensure adequate mixing with the anticoagulant-
Early on, HTLV-II was considered a possible etiologic agent preservative mixture in the bag. Many facilities choose to uti-
of hairy-cell leukemia. This has been proven not to be the lize a blood collection system that diverts the initial aliquot
case, and HTLV-II is currently not associated with any dis- of donor blood into an integrally connected pouch. This
ease process. HTLV-I is associated with adult T-cell leuke- diversion reduces the possibility that a skin plug or core cut
mia/lymphoma in Japan and HTLV-associated myelopathy with the needle during phlebotomy, possibly harboring bac-
and tropical spastic paraparesis, chronic demyelinating teria, will contaminate the collection bag. Blood in the diver-
diseases found in the Caribbean. The confirmatory test for sion pouch can be used for blood typing and viral marker
HTLV is often time consuming and expensive. Many screen- testing without increasing blood loss associated with dona-
ing test results are not confirmed, however; the donated tion. This technique (diversion pouch) may reduce bacterial
blood is destroyed and the donor is alarmed for no appar- contamination rates in blood components overall by about
ent reason.94 Because HTLV-associated diseases are highly 40%, with the highest reduction observed for common skin
unusual in the United States, the necessity of HTLV as a contaminants.96,97
screening test is unclear.
The most recent threat to the blood supply is West Nile Apheresis Platelet Collections
virus. WNV is a mosquito-borne pathogen known to cause Plateletpheresis is a sophisticated technology by which blood is
meningoencephalitis. Once it was recognized as a threat to processed by an apheresis machine that uses centrifugation to
the blood supply in 2003, NAT became available in fairly remove a selected component of the blood and returns the rest
short order. In 2005, WNV seems to be an increasing threat, to the donor. The most common use of this technology is for col-
but the highly successful screening program has reduced its lection of apheresis platelets. Platelet donors are usually recruited
transmission by transfusion significantly. from the ranks of whole blood donors. The minimum plate-
Serologic testing for syphilis is another somewhat contro- let count required to donate apheresis platelets is 150,000/μL.
versial test because transfusion-related transmission of this Apheresis platelet donors can donate more frequently than
disease has been documented only recently. Some consider it whole blood donors: AABB Standards limits apheresis platelet
to have value as an indicator of lifestyle problems that might donations to no more than twice in a 7-day period and no more
make blood donation undesirable. than 24 times per year. The apheresis procedure is more rigor-
BLOOD DONATION AND COLLECTION
ous than whole blood collection because the donor must remain for frequent donors to help avoid these effects. These adverse
connected to the apheresis machine for an extended period, effects tend to be less of a problem for infrequent donors and
often 1 to 2 hours. Another difficulty is the high incidence of with collections using pentastarch. Although traces of heta-
hypocalcemic reactions, due to the calcium-binding anticoagu- starch may be detected in the donor for years, there has been
lant used to keep blood from clotting in the machine. no demonstrated clinical significance. Combined premedi-
Platelets collected through apheresis technology have cation with G-CSF, dexamethasone, and collection with the
some advantages over random donor platelets (RDPs, col- use of hetastarch have been reported to provide a product
lected by centrifugation from individual whole blood units) with a granulocyte yield of 4.1 to 10.8 × 1010 compared to 2.1
because 1 apheresis platelet unit is the equivalent of 6 to 10 to 2.6 × 1010 using dexamethasone alone.104
RDPs. This decreases the risk of transfusion-transmitted Granulocytapheresis products should be administered at
disease and allergic transfusion reactions. If an apheresis least daily to an adult patient to achieve a physiologic dose,
donor’s platelet count (and patience) is sufficient, a double and this must be repeated for a number days. This require-
or even triple product can be collected at one sitting. Many ment creates serious logistical problems and is one reason
apheresis platelet technologies provide a leukocyte-reduced why (along with the cost) granulocytapheresis is not more
product. frequently utilized. Improved antibiotic therapy and the
high incidence of adverse effects in the recipients of gran-
Granulocytapheresis ulocytapheresis products, including pulmonary reactions
Granulocytapheresis produces a product of concentrated and leukocyte alloimmunization, have further decreased the
neutrophils using apheresis technology. Granulocytes are functionality of this therapy.105,106
used to treat neutropenic patients with infections that
Hematopoietic Progenitor Cells, Apheresis
are not responding to antibiotics. Donors are placed on
an apheresis machine with granulocytapheresis capabili- Collection of hematopoietic progenitor cells by apheresis
ties. Not all machines can perform granulocytapheresis: (HPC-A), often referred to as peripheral blood stem cells or
many of the newer machines are specialized for collection stem cells, has become increasingly prevalent over the past
of platelets or plasma. The machine must be properly pro- decade. The main advantage of HPC-A over hematopoietic
grammed and the operator specially trained for granulo- progenitor cells, marrow (HPC-M) is that adequate HPC-A
cyte collections. can be collected to support several courses of high-dose
A significant challenge to granulocytapheresis therapy chemotherapy. Furthermore, transplantation of autologous
is collecting enough granulocytes to produce a therapeu- HPC-A results in a more rapid hematopoietic recovery com-
tic response. Granulocytapheresis donors are premedicated pared to autologous HPC-M. HPCs are mobilized into the
with corticosteroids and/or granulocyte-colony stimulating donor’s peripheral blood from the bone marrow with
factor (G-CSF) before collection, to maximize granulocyte the use of recombinant colony-stimulating factors, either
yield. These medications cause release of marginated granu- G-CSF, granulocyte-macrophage colony-stimulating factor,
locytes from the spleen and major blood vessels, markedly or a combination of the two. 11
increasing the peripheral blood granulocyte count before Autologous HPC-A collections may be performed fol-
donation.98 Higher peripheral granulocyte blood counts lowing, without, or in conjunction with chemotherapy. 171
result in larger granulocyte collections. Apheresis is performed for several days as necessary, until an
Regimens for premedication vary. An example dosage is adequate stem cell dose is achieved. The adequacy is deter-
5 to 10 μg/kg administered subcutaneously, 12 hours before mined by the CD34+ dose (the number of CD34+ cells per
the collection procedure. It is recommended that one consult kilogram of recipient body weight). It is generally agreed that
the manufacturer’s package insert for specific dosage guide- a minimum dose of 2.5 × 106/kg CD34+ cells is necessary for
lines.99–101 Corticosteroids are usually given as prednisone successful engraftment.107 In most autologous collections,
or dexamethasone. The latter is given as a dose of 8 to 12 venous access is obtained through a dual- or triple-lumen
mg, depending on the donor’s weight, at intervals of 4 and catheter.
12 hours before the collection procedure. A uniform dose of Collection of allogeneic HPC-A from HLA-matched
450 μg of G-CSF coupled with 8 mg of dexamethasone, both relatives is primarily performed using G-CSF mobilization.
given 12 hours before collection, has been shown to be as Clinical trials have suggested that a dose of 2.0 × 106/kg
effective as larger combined doses.102 CD34+ cells is a minimum threshold for transplantation.108
Producing a granulocytapheresis product that is not Collections of HPC-A can be stored unmodified or can
heavily contaminated with RBCs has also been a challenge. be processed further with the intent of improving outcomes.
The density of granulocytes is only slightly lower than that of These include purging of cancer cells utilizing monoclonal
RBCs, which makes it difficult to produce a clean separation. antibodies,109 CD34+ selection techniques,110 and ex vivo
RBCs in the granulocyte product must be compatible with the expansion111 (i.e., culture techniques).
recipient to avoid the possibility of an acute hemolytic trans-
Plasmapheresis
fusion reaction. To help remedy the problem of red cell con-
tamination, differential sedimentation is enhanced by use of Most plasma for transfusion is produced by centrifugation
rouleaux-inducing agents. Both hetastarch and pentastarch of a unit of whole blood, which produces a unit of platelet-
are employed for this purpose, although hetastarch is more rich plasma and a unit of RBCs. Plasma can also be obtained
widely used for its higher granulocyte yields.103 Hetastarch, using apheresis technology. Apheresis plasma is usually col-
however, is less rapidly cleared from the body and accumu- lected from group AB donors, which is of particular value
lates more readily in the extravascular space, which can lead because it can be transfused to patients with any blood type,
to localized edema, headache, and fluid retention in repeat although several investigators have raised concerns regarding
donors, often family members, who donate regularly over a the impact of ABO nonidentical blood product transfusions
period of several days. It may be wise to use reduced dosages on patient outcomes.112 The advantage of apheresis plasma
BLOOD BANKING technology is the ability to collect larger units, often called 12% occur after the donor has left the collection site. This
jumbo plasma. Larger units are desirable because fewer units underscores the importance of closely observing donors
are required per dose, which lessens the chance of infectious even after the donation has been completed without inci-
disease transmission and allergic reactions per recipient. For dent. From 30% to 45% of the syncopal reactions include
blood centers collecting blood for transfusion, plasma col- involuntary tetany or tonic-clonic convulsive movements.125
lections usually occur at an interval of 4 weeks or greater and These usually last less than 30 seconds; however, 20% may
must be compliant with FDA guidelines for “infrequent plas- last longer, up to a minute or two. These can also progress to
mapheresis.”113 These donors must meet whole blood crite- full-blown tonic-clonic seizures with associated urine incon-
ria and are limited to an annual maximum of 12 L of plasma tinence. Prolonged hyperventilation can rarely lead to tetany
(14.4 L if at least 80 kg). without syncope as a result of hypocapnea leading to hypo-
“Source” plasma is collected in large quantities by com- calcemia. Severe vasovagal reactions may resemble shock
mercial firms for fractionation into plasma derivatives used to clinically, except that the pulse is slow rather than fast. When
produce reagents and other plasma products. Source plasma the blood pressure becomes extremely low the donor often
is not for direct human transfusion. Because source plasma becomes pale and even cyanotic.
donors are often paid and tend to be aggressively recruited, Mild vasovagal reactions are treated by elevating the
specific regulatory requirements have been designed to pro- donor’s legs above his or her heart, helping to improve
tect the frequent plasmapheresis donor. These donors can blood flow to the brain. Ammonium salts, cold neck com-
donate a maximum of twice in a 7-day period, with at least presses, and reassurance are often all that is necessary, and
2 days between donations. Frequent donors require periodic the donation can proceed. Treatment with intravenous flu-
physical examination by a physician and periodic determina- ids or medications is usually unnecessary if the donor does
tion of serum protein levels. not have an underlying medical condition such as coronary
artery or cerebrovascular disease. Experienced staff can help
Multiple Apheresis Products avoid a severe vasovagal reaction by recognizing and treat-
New apheresis technologies now make possible the collec- ing a reaction in the early stages.126,127 Treatment of more
tion of multiple products in a single donation when appro- severe vasovagal reactions, which may proceed to loss of
priate donor criteria are met. A double red cell product can consciousness and seizure activity, require that the dona-
be collected every 16 weeks if the donor meets the specified tion be stopped and the needle be withdrawn to prevent
weight and hematocrit criteria. Some donors appreciate the local tissue injury from convulsive movements. The main
convenience of donating less frequently. The collection of risk associated with syncopal vasovagal reaction is trauma,
multiple products from a single donation helps to maintain particularly head trauma. Fractures and other significant
an adequate blood supply and is often more cost efficient for injuries have been reported.114 Some centers keep tongue
the collection facility. protectors available, but damage to the tongue from convul-
sive movements is rare. The typical time for recovery from a
II vasovagal reaction is 5 to 30 minutes.
Adverse Donor Reactions and Injuries
On occasion, a particularly severe reaction may require
172 The vasovagal reaction is the most common systemic donor additional resources. Hospital-based donor rooms may have
reaction, occurring in 2% to 3% of donors.114–117 These reac- access to an emergency response team with a crash cart to
tions often present as loss of consciousness due to a drop in deal with these situations. For remote centers without an on-
blood pressure without a normal compensatory increase in site emergency response team, it may be prudent to call para-
heart rate. Vasovagal reactions are 5 to 10 times more fre- medics and have the donor transferred to a local emergency
quent in younger donors (8% to 11%),118 making careful room if necessary. Maintaining a crash cart in the donor
observation especially important at high school and college center with emergency life support equipment and medica-
blood drives. Other predisposing factors include first-time tions is controversial. Centers that rarely see a severe reaction
donor status,119 low weight,120 and a history of a previous may have little or no actual experience with advanced life
donation reaction.121,122 An anxiety-related psychosomatic support procedures despite formal accreditation and may be
component appears to be present because vasovagal reac- unfamiliar with the available emergency equipment when
tions have occurred before donation and epidemic fainting needed. It may be better to have an emergency backup sys-
is known to occur. tem, often calling 911, than to have an crash cart and not
Vasovagal reactions often occur with short warning, know how to use it.
during or immediately or after phlebotomy. Experienced Vasovagal reactions may recur within the next several
phlebotomists know to look for lightheadedness, weakness, hours and so donors who have had one should be advised
pallor, nausea, and diaphoresis. Excessive anxiety, often to use caution when driving or operating heavy machinery.
manifested by nervous talkativeness and hyperventilation, Donors with severe or multiple reactions should be discour-
can precipitate a vasovagal reaction. Hyperventilation can aged from attempting to give blood again in the near future.
result in respiratory alkalosis and hypocalcemia, which can Even donors who have had severe vasovagal reactions tend to
help precipitate a vasovagal reaction.123 In these instances, recover spontaneously. No reports of deaths caused by blood
having the donor breathe into a paper bag may increase car- donation-related vasovagal reaction appear in the medical lit-
bon dioxide levels, reversing the alkalosis and hypocalcemia. erature, although there are reports of cardiac arrest in patients
A calm, assuring demeanor by the phlebotomist will also do after venipuncture for blood sample collection.128,129
much to alleviate anxiety. The sudden drop in blood pressure caused by a vasovagal
Approximately 5% of vasovagal reactions are syncopal reaction may evoke an ischemic event in donors with occlu-
and progress to loss of consciousness in about 0.08% to sive atherosclerotic vascular disease. These reactions tend to
0.34% of donors.124 Syncopal reactions tend to occur after be rare, likely due to successful donor screening techniques.114
phlebotomy—about 60% occur at the refreshment table and The risk, of course, is higher for individuals with pre-existing
BLOOD DONATION AND COLLECTION
disease, so these donors should be accepted only if their dis- adverse effects precipitated by acute volume loss are less pro-
ease is stable and even then with caution. Some degree of risk nounced. The two-needle continuous flow technology draws
may be acceptable for autologous donation, which is com- blood from one vein at the same rate as processed blood is
monplace for elective cardiac bypass surgery (if the patient returned in the other. Intermittent single-needle technol-
does not have unstable angina).30,130 Taking chances with a ogy removes and replaces blood in very small increments.
routine allogeneic donor is unacceptable, and if there is any The decreased rate of vasovagal reactions among apheresis
real uncertainty, the donor should be deferred. donors may also be due to the fact that these donors tend
Cerebrovascular disease presents a special problem to be older and more experienced with the donation pro-
because loss of consciousness and mild seizure activity may cess. As with whole blood donation, the frequency of adverse
be manifestations of both a vasovagal reaction and cerebro- reactions is higher in first-time apheresis donors. Vasovagal
vascular ischemia. For a donor with a known history of cere- reactions, for instance, range from 2.0% in first-time to 0.5%
brovascular disease, distinguishing a donation reaction from in repeat donations.
a stroke or transient ischemic attack may be difficult. For the Other than mild citrate reactions, adverse events are seen
same reasons, it is generally a good idea to be cautious about with an overall frequency of 2.18% (428 of 19,611 dona-
accepting a donor with a poorly controlled seizure disorder. tions from 17 centers).139 The most common adverse events
Minor local tissue injury at the venipuncture site is a well- are related to the venipuncture, with a frequency of 1.30%.
known complication of any venipuncuture. Postphlebotomy Palpable hematoma accounted for 88% of these events. The
bruising is the most common adverse donor event. Two risk of a hematoma is higher for a plateletpheresis procedure
studies of outpatient phlebotomy suggest that the incidence than for a whole blood donation (0.3% for the latter; Table
of bruising may range from 9% to 16%.131,132 Hematomas 11-5). This may be due to the extended time frame that the
occur less commonly: approximately 0.3% at the time of needle is in place compared with the 10 to 15 minutes for a
donation and an additional 0.05% reported by the donor whole blood collection.
later. Hematomas, bruising, and soreness can usually be Other adverse effects of apheresis include RBC hemoly-
treated with compresses and acetaminophen. These hematomas sis, air emboli, clots, and leaks. These have been reported but
generally resolve within 2 weeks. are extremely rare with improved apheresis technology and
Accidental arterial punctures are rare.125 They may pre- better-trained operators. The incidence of adverse apheresis
sent as an unusually rapid phlebotomy with bright red reactions varies among institutions, which may be due to
blood and a pulsating needle. On recognition, the phlebot- donor selection, operator training, or even record keeping.
omy should be stopped and pressure should be applied for One program has reported an overall 0.81% frequency of
at least 10 minutes. The donor should be closely observed adverse events; of 19,736 procedures, 47 (0.24%) were rated
for an extended period. If there is any question about effec- as serious.140 Seven of these 47 donors required transfer to an
tive hemostasis, competent medical consultation should be emergency department.
obtained. Brachial artery pseudoaneurysm,133,134 arteriove- A single unit plateletpheresis procedure will cause a drop
nous fistula,135 and compartment syndrome136 are possible in the donor’s platelet count of approximately 30,000 to 11
sequelae of arterial puncture. All three complications are 50,000 platelets/μL, although a return to prepheresis levels
rare but do require surgical repair. will usually occur in a few days.141,142 In some donors, fre- 173
Nerve damage due to a hematoma or direct trauma is an quent plateletpheresis may cause a gradual drop in plate-
unusual event. Reports suggestive of nerve damage occur in lets, such that collections must be discontinued, or possibly
approximately 1 of every 6000 blood donors.137,138 Symptoms donations can be scheduled with longer intervals between
may include pain and paresthesias at the venipuncture site donations.143 The platelet count usually recovers in these
extending into the donor’s hand, fingers, or shoulder area. donors over several months without treatment. Donors
A hematoma is present in approximately 25% of cases. with persistent thrombocytopenia are deferred from platelet
Approximately 40% recover in a few days and 70% recover donation.
completely within 30 days. The remaining 30% take as long Lymphocytes, which have a density similar to platelets,
as 9 months to recover. A few donors have a small area of are often lost during a plateletpheresis procedure. In theory,
persistent numbness even after 9 months. Rarely, significant frequent plateletpheresis could remove enough lymphocytes,
permanent neurologic damage occurs. especially long-lived T lymphocytes, to cause immune dys-
Bactericidal skin cleansing solutions and adhesive tape function. At this time, however, no adverse clinical effects
can cause local allergic reactions and irritation despite exten- have been observed in healthy donors.144
sive prerelease testing and FDA approval. For iodine-sensi- The same adverse effects of plateletpheresis are also be
tive donors, alternative solutions can be used. Post-donation seen with granulocytapheresis. In addition, the sedimenting
hemostasis can be attained with pressure dressings rather agents used to effectively remove granulocytes and the medi-
than adhesive tape. cations used to stimulate the donors’ granulocyte counts have
A common adverse effect of apheresis donation is hypo- additional side effects. Small doses of corticocosteroids have
calcemia, due to the calcium-binding citrate anticoagulant been reported to cause insomnia in up to 25% of donors.145
used to keep blood from clotting in the apheresis machine Frequent granulocyte donors should be free of peptic ulcers,
tubing. Return of citrated blood to the donor may cause diabetes mellitus, hypertension, glaucoma, and other diseases
lightheadedness and perioral and peripheral paresthesias. exacerbated by prolonged use of corticosteroids. A combined
These are readily treated by ingestion of calcium-contain- dose of G-CSF and dexamethasone may cause side effects in
ing antacid tablets. On machines that allow more extensive as many as 72% of donors; these are commonly insomnia
operator control, the rate of citrate infusion can be lowered (30%), mild bone pain (41%), and headaches (30%). The
to help diminish these reactions. latter two are readily relieved by analgesics.146 Sedimentation
Apheresis removes a limited amount of blood from the agents, usually hetastarch, may cause fluid retention and
donor at any given time, so vasovagal reactions and other allergic reactions.
BLOOD BANKING
Table 11–5 The Incidence of Whole-Blood Donor Complications and of outside Medical Care
Arm injuries
Bruise NA 22.7 See hematoma
Sore arm NA 10.0 Unknown
Hematoma 0.35 1.7 0.57 (1/17,500)
Nerve irritation 0.02 0.9 0.46 (1/21,700)
Local allergy 0.5 (estimate) NA Unknown
Arterial puncture 0.01 NA 0.07 (1/142,900)
Thrombophlebitis 0.002 (estimate) NA 0.01 (1/75,000)
Thrombosis Very rare NA <0.001
Local infection 0.002 (estimate) NA <0.01 (1/225,000)
Systemic
Fatigue NA 7.8 Unknown
Vasovagal reaction 2.5 7.0 1.08 (1/9300)
Syncope 0.08–0.34 NA See above
Syncope with injury 0.01–0.05 NA See above
Nausea, vomiting NA 0.4 Unknown
MI, stroke, etc. Very rare Very rare Unknown
Total (donors) 3.5 37 2.94 (1/3400)||
||
MI, myocardial infarction; NA, not applicable.
From Newman BH. Blood donor complications after whole-blood donation. Curr Opin Hematol 2004;Sep;11(5):339–345.
†
Newman BH. Donor reactions and injuries from whole-blood donation. Transfusion Med Rev 1997;11:64–75.
‡
Newman BH, Pichette S, Pichette D, et al. Adverse effect in blood donors after whole-blood donation: a study of 1,000 blood donors
interviewed 3 weeks after whole-blood donation. Transfusion 2003;43:598–603.
§
Newman B, Crooks N, Zhou L, et al. Whole-blood donor complications leading to outside medical care: National overview and a detailed
review at one blood center [abstract]. Transfusion 2004;44(Suppl):77A.
||
Includes an unlisted category of “other.”
HBsAg Repeatedly reactive HBsAg confirmed positive by neutralization (or neutralization not done) is found
in a subsequent donation whose prior test results were nonreactive OR
Repeatedly reactive HBsAg, negative neutralization AND repeatedly reactive HBcore is found in
a subsequent donation whose prior test results were nonreactive
HBcore Repeatedly reactive anti-HBcore, and the test result of the second test method is reactive in
a subsequent donation whose prior test results were nonreactive OR
Repeatedly reactive anti-HBcore and repeatedly reactive HBsAg in a subsequent donation whose prior
test results were nonreactive
HTLV-I/II Repeatedly reactive anti-HTLV-I/II and the second test result is repeatedly reactive or not performed
is found in a subsequent donation, whose prior test results were nonreactive or not previously tested
Anti-HIV-1,2 Repeatedly reactive anti-HIV-1,2 confirmed positive HIV-1 Western Blot OR
HIV-2 EIA reactive found in a subsequent donation not previously tested or whose prior test results
were nonreactive
HIV-1 NAT NAT reactive AND HIV-1 Western Blot (or IFA) indeterminate positive or HIV-2 EIA reactive is found
in a subsequent donation whose prior test results were nonreactive or prior donation was not tested
Anti-HCV Repeatedly reactive anti-HCV with a supplemental test result of positive, indeterminate OR no
supplemental test performed found in a subsequent donation whose prior donation was not
previously tested with the currently licensed test, or whose prior test results were nonreactive
HCV NAT NAT reactive with a supplemental test result of positive, indeterminate OR no supplemental test
performed found in a subsequent donation whose prior donation was not previously tested,
or whose prior test results were nonreactive
WNV Current donation sample has a reactive WNV NAT. Relevant collections include those occurring
between 120 days prior to the date of the reactive test and 120 days after the date of the
reactive test.
Donor reports a diagnosis of West Nile virus occurring between 14 days prior to the onset of illness
and up to and including 120 days subsequent to the onset of illness or diagnosis, whichever is the
later date.
Donor reports unexplained febrile illness with headache or symptoms suggestive of WNV infection
between June 1 and November 30, and Medical Director has determined this represents likely
infection by WNV.
A report is received regarding possible transmission of WNV by a blood component received within
the 120 days prior to the onset of symptoms, or a WNV- fatality in a transfused recipient. Prompt
quarantine and retrieval for in-date components collected from the donor of suspect donation in
the period between 120 days before the suspect donation and up to and including 120 days after 11
the suspect donation must be performed.
CJD or vCJD Subsequent to donation, the donor:
Risk factors is diagnosed with Creutzfeldt-Jakob disease 175
indicates a family history of Creutzfeldt-Jakob disease
acknowledges receipt of human pituitary-derived growth hormone (HGH)
acknowledges receipt of a dura mater transplant
indicates having spent a total time of 3 months or more in the United Kingdom from 1980
through 1996
indicates receipt of injectable products from cattle in BSE-endemic countries
acknowledges spending a total time of 5 years or more in Europe, including time spent in the UK
If a member of the U.S. military, a civilian military employee or a dependent of a member of the
US military and spent a total of 6 months or more associated with a military base in any of
the following countries:
From 1980 through 1990 in Belgium, the Netherlands, or Germany
From 1980 through 1996 in Spain, Portugal, Turkey, Italy, or Greece
AIDS-related If a donor implicated in the investigation of transfusion-associated AIDS has a reactive test result for
anti-HIV-1, 2 and/or HIV-1 antigen.
If information is received that a patient with AIDS has previously donated blood.
High-risk behavior Post-donation information becomes available regarding a donor who would have been deferred
(e.g., travel, had the information been known at the time of donation.
vaccination malaria,
tattoo, blood
exposure)
BLOOD BANKING
Compliance Issues On identification of a unit meeting lookback criteria, a facil-
ity must search records for prior donations by the same indi-
The management of unexpected events may often be referred vidual. A sample worksheet may be seen in Figure 11–1. The
to as error or deviation management. Because humans are falli- highest priority should be placed on the most recently donated
ble, limiting the incidence of errors to absolute zero can never units. This should be done within 72 hours, so that any unit(s)
be achieved.154, 155 Well-prepared organizations approach error remaining in inventory may be immediately quarantined.
management from a systems level, anticipating that errors and Policies must include consignee notification, so that any com-
deviations will occur, and prepare system defenses for dealing ponents shipped to other facilities may be immediately quar-
with their inevitable occurrence rather than focusing on blam- antined and returned. A sample notification letter and product
ing individuals for forgetfulness or inattention.154 disposition record may be seen in Figures 11–2 and 11–3.
If the implicated donor has donated on many occasions,
Lookbacks lookback notification should be started with the most recent
Being human, blood donors themselves represent a source of recipients. Reasonable and timely attempts must be made to
deviation. With the incubation time between exposure and notify transfusion recipients, particularly if a lookback is due
onset of disease, some individuals are unaware that they may to HIV156 or HCV,157 so that recipients may obtain testing,
be infectious to others. To identify these individuals, a donor counseling, and medical referral as needed.
center must develop procedures to notify the recipients of
Recalls and Market Withdrawals
blood or components of previous donations when a donor
becomes confirmed positive for an infectious disease marker, Errors can also result from improper testing, incorrect label-
or when a statement of high-risk information dis closed at ing of components, improperly interpreting test results,
a subsequent time determines that the donor was actually improperly using equipment, or failure to follow the manu-
ineligible at the time of their donation. Identification of facturers’ directions or facility procedures. These kinds of
persons who may have received blood or components from errors may result in recalls or market withdrawals.
such donors is referred to as lookback. Examples of causes for Recalls are defined as actions taken by a facility to remove
lookback may be seen in Table 11–6. a product from the market.158,159 Recalls may be conducted
Whole Blood
Red Cells
Fresh Frozen Plasma
Platelets
Recovered Plasma
Cryoprecipitate
Other information (Patient primary physician notifications, final outcomes, response from consignees). List by component. Use other side as
necessary and attach all related documents.
Unit Status Legend: INV = Inventory; QU = Quarantined; TF = Transfused; DS = Destroyed; SHIP = Shipped
SOP # attachment #
Effective date
John Doe, MD
Director, Blood Bank
General Hospital
Anywhere, CA 90000
This confirms our telephone notification that your hospital received a blood component
from a donor that was subsequently found to be confirmed positive for HIV-1. All test
results from prior donations were nonreactive, including those for the blood component
shipped to your facility.
On November 30, 2005, at 11:25am, we telephoned your facility and spoke to Jane Doe
and conveyed the following information:
Please complete and return the enclosed Lookback Product Disposition Record.
Maintain a copy of the record for your files.
If you have any questions concerning this matter, please contact the Somewhere Donor
Center at (555) 123-4567 x891.
Sincerely,
177
on a facility’s own initiative, by FDA request, or by FDA A Class II recall is a situation in which use of or exposure
order under statutory authority. FDA guidelines categorize to a violative product may cause temporary or medi-
all recalls into one of three classes according to the level of cally reversible adverse health consequences or where
hazard involved. the probability of serious adverse health consequences is
remote. An example of products in this category would
A Class I recall is a situation in which there is a rea- be a unit later found to be collected from a donor whose
sonable probability that the use of or exposure to a hemoglobin did not meet the minimum criteria.
violative product will cause serious adverse health A Class III recall is a situation in which use of or exposure to a
consequences or death. An example of products in violative product is not likely to cause adverse health con-
this category would be a unit issued that was found to sequences. Examples of products in this category might be
be HIV positive. those that do not meet FDA labeling regulations.
BLOOD BANKING
o Transfused
o Returned
o Discarded (Reason)
o Transferred to another facility (Please complete the information on the receiving facility below)
Name
II
Street Address
178 City/State/Zip
Phone ( )
A market withdrawal occurs when a product has a minor program to improve the effectiveness of the FDA’s regulatory
violation that would not be subject to FDA legal action. The program. FDA replaced the term error and accident with the
firm removes the product from the market or corrects the term biological product deviation (BPD).
violation. For example, a product removed from the mar- Licensed blood establishments, unlicensed blood estab-
ket due to tampering, without evidence of manufacturing or lishments, registered blood establishments, and transfusion
distribution problems, would be a market withdrawal. services are required to report to the FDA all BPDs. BPDs are
defined as any event associated with manufacturing of blood
Biological Product Deviation Reporting or blood components that EITHER:
On November 7, 2000, the FDA published a final rule160 to
amend the requirements of reporting errors and accidents in 1. Represent a deviation from current GMPs, applicable
manufacturing of products. This rule was issued as part of a regulations, or established specifications that may
BLOOD DONATION AND COLLECTION
affect the safety, purity, or potency of that product http://www.bloodservices.ca/Centreps/Internet/UW_V502_MainEn-
OR gine.nsf/page/E_NR2005-09-07_IpsosReid_Touched+by+system?
OpenDocument. Accessed September 8, 2005.
2. Represent an unexpected or unforeseeable event that 6. Newman BH. Whole-blood donation: blood donor suitability and
may affect the safety, purity, or potency of that product adverse events. Curr Hematol Rep 2004;3:437–443.
AND 7. U.S. General Accounting Office. blood Supply: Availability of Blood
• Occurs in your facility or a facility under contract to (GAO/HEHS-99-187R). Washington, D.C., U.S. General Accounting
Office, September 20, 1999.
you AND 8. U.S. Department of State. Hispanics Replace African Americans as
• Involves a distributed blood or blood component. Largest U.S. Minority Group. January 23, 2003. Available at http://
usinfo.state.gov/usa/diversity/a012303.htm.
Post-donation information is considered reportable to 9. Glynn SA, Schreiber GB, Busch MP, et al. Demographic characteris-
the FDA as a BPD if the donor should have been deferred tics, unreported risk behaviors, and the prevalence and incidence of
had the information been known at the time of donation viral infections: a comparison of apheresis and whole blood donation.
and the safety, purity, or potency of the product could be Transfusion 1998;38:350–358.
10. Leiby DA, Herron RM Jr, Read EJ, et al. Trypanosoma cruzi in Los
affected. Post-donation information also includes informa- Angeles and Miami blood donors: impact of evolving donor demo-
tion that a blood center obtains when it adds new donor graphics on seroprevalence and implications for transfusion transmis-
history questions. sion. Transfusion 2002;42:549–555.
In many cases blood establishments cannot control post- 11. Eastlund T. Monetary blood donation incentives and the risk of trans-
donation information. For example, a donor may call after fusion-transmitted infection. Transfusion 1998;38:874–882.
12. Code of Federal Regulations. 21 CFR 606.121(c)(5). Washington, D.C.,
donating to report a post-donation illness, or information U.S. Government Printing Office, April 1, 2005.
obtained post-donation about exposure to a disease or a sex 13. California Health and Safety Code. 1626(d). Available at http://www.
partner at high risk. Reports of post-donation information leginfo.ca.gov.
continue to represent the largest percentage of BPDs submit- 14. Strauss RG. Blood donations, safety, and incentives. Transfusion
2001;41:165–167.
ted by blood and plasma establishments (71%). In 88% of the 15. Barker LF, Westphal RG. Voluntary, nonremunerated blood donation:
reports the donor was aware of the information at the time still a world health goal? Transfusion 1998;38:803–806.
they were interviewed, but failed to provide the information 16. Sullivan JL. Iron and the sex differences in heart disease risk. Lancet
during the interview. Most often (91%), the donor center staff 1981;1:1293–1294.
is made aware during a subsequent donation interview.161 17. Meyers DG. The iron hypothesis: does iron play a role in atherosclerosis?
Transfusion 2000;40:1023–1029.
18. Wallas CH, Lipton KS. Donor Incentives––A Report of the AABB
Board of Directors (Association Bulletin 94-6). Bethesda, Md., Ameri-
CONCLUSION can Association of Blood Banks, October14, 1994.
19. U.S. Food and Drug Administration. Compliance Policy Guidance
for FDA Staff and Industry, Chapter 2, Section 230.150. Issued May
The products of blood donation—RBCs, platelets, and 7, 2002, revised November 22, 2005. Available at http://www.fda.gov/
plasma—are a vital resource making possible modern medi- ora/compliance—ref/cpg/cpgbio/cpg230=150final.htm. Last modified
cine. At this time, there are no comparable substitutes. A debt December 12, 2005.
20. Busch M, Stramer S. The efficiency of HIV p24 antigen screening of
11
of gratitude is owed to those blood donors who give their
blood, for no monetary compensation, to anonymously help US blood donors: projections versus reality. Infus Ther Transfus Med
1998;25:194–197. 179
save the lives of others. 21. Code of Federal Regulations. 21 CFR 606.40. Washington, D.C., U.S.
At one time, all that was needed to collect blood was a col- Government Printing Office, April 1, 2005.
lection set and a donor with suitable veins. Since then, thou- 22. Dellinger EP, Anaya DA. Infectious and immunologic consequences of
sands of transfusion medicine professionals have transformed blood transfusion. Crit Care (London) 2004;8(Suppl 2):S18–S23.
23. Blumberg N. Deleterious clinical effects of transfusion immuno-
the early donor rooms to a multibillion dollar industry collect- modulation: proven beyond a reasonable doubt. Transfusion 2005;45
ing tens of millions of blood products per year. Their cease- (2 Suppl):33S–39S.
less efforts to improve transfusion safety have had remarkable 24. Raghavan M, Marik PE. Anemia, allogenic blood transfusion, and
success on the well-being of the blood recipient, donor, and immunomodulation in the critically ill. Chest 2005;127:295–307.
even the collections staff. Unremitting efforts to improve 25. Blajchman MA. Immunomodulation and blood transfusion. Am J
Therap 2002;9:389–395.
donor screening and serologic testing for infectious disease are 26. Vamvakas EC. Blajchman MA. Deleterious clinical effects of transfu-
needed to keep ahead of potential threats to the blood supply. sion-associated immunomodulation: fact or fiction? Blood 2001;97:
For these reasons, change occurs very rapidly in the world 1180–1195.
of blood banking, and all decision-making processes require 27. Rouger P. Transfusion induced immunomodulation: myth or reality?
Transfus Clin Biol 2004;11:115–116.
checking for the latest information available, which can be 28. Goodnough LT, Monk TG, Andriole GL. Erythropoietin therapy.
obtained through the AABB or FDA websites. NEJM 1997;336:933–938.
29. Popovsky MA, Whitaker B, Arnold N. Severe outcomes to allogeneic
REFERENCES and autologous blood donation: frequency and characterization.
Transfusion 1995;35:734–737.
1. American Association of Blood Banks. Available at http://www.aabb. 30. Kiyama H, Ohshima N, Imazeki T. Safety and efficacy of blood dona-
org/content/All_Blood/Facts_About_Blood_and_Blood_Banking/ tion prior to elective cardiac surgery in anemic patients. Japan J Tho-
aabb_ faqs.htm. Last modified April12, 2006. rac Cardiovasc Surg 2000;48:101–105.
2. Sullivan MT, Wallace EL, Umana WO, Schreiber GB. Trends in the 31. From California Blood Bank Society. Available at http://www.cbbsweb.
collection and transfusion of blood in the United States, 1987–1997 org/enf/2001/adtxtrigger.html. Last revised March 1, 2001.
[abstract]. Transfusion 1999;39(Suppl):1S. 32. Dupuis JY, Bart B, Bryson G, Robblee J. Transfusion practices among
3. Brecher ME, Goodnough LT. The rise and fall of preoperative autolo- patients who did and did not predonate autologous blood before elec-
gous blood donation (editorial). Transfusion 2001;41:1459–1462. tive cardiac surgery. Can Med Assoc J 1999;160:97–1002.
4. Goodnough LT, Brecher ME, Kanter MH, AuBuchon JP. Transfusion med- 33. From California Blood Bank Society. Available at http://www.cbbsweb.
icine. First of two parts—blood transfusion. NEJM1999;340:438–447. org/enf/2004/ad_positivetests.html. Last modified August 5, 2005.
5. Canadian Blood Services News Release. Ipsos-Reid poll: over half of 34. Grossman BJ, Stewart NC, Grindon AJ. Increased risk of a positive
Canadians say they or their family have needed blood. Less than four test for antibody to hepatitis B core antigen (anti-HBc) in autologous
percent of eligible population donated blood last year. Available at blood donors. Transfusion 1988;28:283–285.
BLOOD BANKING 35. Etchason J, Petz L, Keeler E, et al. The cost-effectiveness of preoperative 65. U.S. Department of HHS, Food and Drug Administration, Center for Bio-
autologous blood donations. NEJM 1995;332:719–724. logics Evaluation and Research. Guidance for Industry: Revised Preventive
36. Code of Federal Regulations. 21 CFR 610.40(d). Washington, D.C., Measures to Reduce the Possible Risk of Transmission of Creutzfeldt-
U.S. Government Printing Office, April 1, 2005. Jakob Disease (CJD) and Variant Creutzfeldt-Jakob Disease (vCJD) by
37. Code of Federal Regulations. 21 CFR 606.121. Washington, D.C., U.S. Blood and Blood Products. Federal Register Docket No. 97D-0318. Wash-
Government Printing Office, April 1, 2005. ington, D. C., Government Printing Office, January 16, 2002.
38. Silva MA (ed). Standards for Blood Banks and Transfusion Services, 66. Schmidt M, Brixner V, Ruster B, et al. NAT screening of blood donors for
23rd ed. Bethesda, Md., American Association of Blood Banks, 2004. severe acute respiratory syndrome coronavirus can potentially prevent
Standard 5.8.5.1.2. transfusion associated transmissions. Transfusion 2004;44:470–475.
39. Shulman IA, Osby M. Storage and transfusion of infected autologous 67. le Fichoux Y, Quaranta JF, Aufeuvre JP, et al. Occurrence of Leishma-
blood or components. Arch Pathol Lab Med 2005;129:981–983. nia infantum parasitemia in asymptomatic blood donors living in an
40. Shulman IA. 1992 CAP Surveys. Comprehensive Transfusion Medicine area of endemicity in southern France. J Clin Microbiol 1999;37:1953–
Survey, 1992 Set J-C. Interlaboratory Comparison Program. North- 1957.
field, Ill., College of American Pathologists, 1992. 68. Eastman RT, Barrett LK, Dupuis K, et al. Leishmania inactivation in
41. Linden JV, Wagner K, Voytovich AE, Sheehan J. Transfusion errors human pheresis platelets by a psoralen (amotosalen HCl) and long-
in New York State: an analysis of 10 years’ experience. Transfusion wavelength ultraviolet irradiation. Transfusion 2005;45:1459–1463.
2000;40:1207–1213. 69. Pomper GJ, Wu Y, Snyder EL. Risks of transfusion-transmitted infec-
42. Pub. Law No. 101–336, 104 Stat. 327 (1990). Codified at 42 U.S.C. tions: 2003. Curr Opin Hematol 2003;10:412–418.
§12101–12213. 70. Busch MP, Kleinman SH, Nemo GJ. Current and emerging infectious
43. Rothstein RF. Bragdon v. Abbott—Supreme Court Decision Addresses risks of blood transfusions. JAMA 2003;289:959–962.
Application of Americans with Disabilities Act to Individuals with HIV. 71. Mintz PD, Liption KS. Interim Standard for Standards for Blood Banks and
Houston, Tex., University of Houston Health Law & Policy Institute, June Transfusion Services, 23rd ed. AABB Association Bulletin #05–07. Bethesda,
26, 1998. Md., American Association of Blood Banks, September 30, 2005.
44. Bragdon v. Abbott et al. (97-156). Cornell Law School Legal Informa- 72. Newman B. Blood donor suitability and allogeneic whole blood dona-
tion Institute Supreme Court Selection. Available at http://supct.law. tion. Transfus Med Rev 2001;15:234–244.
cornell.edu/supct/html/97-156.ZS.html. Accessed June 3, 2006. 73. Department of Defense. Available at http://www.militaryblood.dod. mil/
45. Mintz PD. Participation of HIV-infected patients in autologous blood library/policies/downloads/conditions_list.doc. Revised July 5, 2005.
programs. JAMA 1993;269:2892–2894. 74. Mayo DJ, Rose AM, Matchett SE, et al. Screening potential blood donors at
46. Yomtovian R, Kelly C, Bracey AW, et al. Procurement and transfusion risk for human immunodeficiency virus. Transfusion 1991; 31:466–474.
of human immunodeficiency virus-positive or untested autologous 75. Newmand BH. Vasovagal reactions in high school students: findings
blood units: issues and concerns: A report prepared by the Autologous relative to race, risk factor synergism, female sex, and non-high school
Transfusion Committee of the American Association of Blood Banks. participants. Transfusion 2002;42:1557–1560.
Transfusion 1995;35:353–361. 76. Trouern-Trend JJ, Cable RG, Badon SJ, et al. A case-controlled multicenter
47. The ADA, HIV, and autologous blood donation. Association Bulletin study of vasovagal reaction in blood donors: influence of sex, age, donation
98–5. Bethesda, Md., American Association of Blood Banks, 1998. status, weight, blood pressure, and pulse. Transfusion 1999;39:316–320.
48. Williams AE, Wu Y, Kleinman SH. The declining use and comparative 77. Coburn TJ, Miller WV, Parrill WD. Unacceptable variability of hemo-
seroprevalence of directed whole blood donations [abstract]. Transfu- globin estimation on samples obtained from ear punctures. Transfu-
sion 2000;40(Suppl):5S. sion 1977;17:265–268.
49. Strauss RG, Barnes A, Blanchette VS, et al. Directed and limited-exposure 78. Avoy DR, Canuel ML, Otton BM, et al. Hemoglobin screening in
blood donations for infants and children. Transfusion 1990; 30:68–72. prospective blood donors: a comparison of methods. Transfusion
50. Hare VW, Liles BA, Crandall LW, Nufer CN. “Partners for Life”—a 1977;17:261–264.
II safer therapy for chronically transfused children [abstract]. Transfu- 79. Newman B. Very anemic donors pass copper sulfate screening test [let-
sion 1994;34(Suppl):92S. ter]. Transfusion 1997;37:670–671.
180 51. U.S. Food and Drug Administration. Guidance for industry: variances for 80. Nusbacher J, Chiavetta J, Naiman R, et al. Evaluation of a confidential
blood collection from individuals with hereditary hemochromatosis. FDA method of excluding blood donors exposed to human immunodefi-
Docket No. 00D-1618. Federal Register August 23, 2001;66(164). ciency virus. Transfusion 1986;26:539–541.
52. Silva MA (ed). Standards for Blood Banks and Transfusion Services, 81. Peterson LR, Lackritz E, Lewis WF, et al. The effectiveness of the confi-
23rd ed. Bethesda, Md, American Association of Blood Banks, 2004. dential unit exclusion option. Transfusion 1994;34:865–869.
53. From California Blood Bank Society. Available at http://www.cbbsweb. 82. Korelitz JJ, Williams AE, Busch MP, et al. Demographic characteristics
org/enf/2003/donor_id.html. Last modified July 5, 2003. and prevalence of serologic markers among donors who use the con-
54. Grindon AJ, Norrell S, Robertson WR, et al. Predonation determina- fidential unit exclusion process: The Retrovirus Epidemiology Donor
tion of donor eligibility [abstract]. Transfusion 1995;35(Suppl):72S. Study. Transfusion 1994;34:870–876.
55. From U.S. Food and Drug Administration. Available at http://www. 83. Zou S, Notari EP, Musavi F, Dodd RY. Current impact of the confiden-
fda.gov/cber/gdlns/donorhistques.htm. Accessed July 5, 2006. tial unit exclusion option. Transfusion 2004;44651–44657.
56. Silvergleid AJ, Leparc GF, Schmidt PJ. Impact of explicit questions about 84. Kean CA, Hsueh Y, Querrin JJ, et al. A study of confidential unit exclu-
high-risk activities on donor attitudes and donor deferral patterns: results sion. Transfusion 1990;30:707–709.
in two community blood centers. Transfusion 1989;29:362–364. 85. Kessler D, Valinsky JE, Bianco C. Sensitivity and specificity of confi-
57. Zuck TF, Cumming PD, Wallace EL. Computer-assisted audiovisual dential unit exclusion (CUE)-does it work? [abstract] Transfusion
health history self-interviewing. Results of the pilot study of the Hox- 1993;33(Suppl):35S.
worth Quality Donor System. Transfusion 2001;41:1469–1474. 86. From California Blood Bank Society. Available at http://www.cbbsweb.
58. From U.S. Food and Drug Administration. Available at http://www.fda.gov/ org/enf//2003/donor_cue_deferral.html. Last modified April 14, 2003.
ohrms/dockets/ac/05/slides/2005-4096S2_03.ppt. Accessed July 5, 2006. 87. Goodnough LT, Shander A, Brecher ME. Transfusion medicine: look-
59. Brown P, Cervenakova L. The modern landscape of transfusion-related ing to the future. Lancet 2003;361:161–169.
iatrogenic Creutzfeldt-Jakob disease and blood screening tests. Curr 88. Kleinman SH, Strong DM, Tegtmeier GG, et al. Hepatitis B virus
Opin Hematol 2004;11:351–356. (HBV) DNA screening of blood donations in minipools with the
60. McCullough J, Anderson D, Brookie D, et al. Consensus conference COBAS AmpliScreen HBV test. Transfusion 2005;45:1247–1257.
on vCJD screening of blood donors: report of the panel. Transfusion 89. Busch MP. Should HBV DNA NAT replace HBsAg and/or anti-HBc
2004;44:675–683. screening of blood donors? Transfus Clin Biol 2004;11:26–32.
61. National Center for Infectious Diseases. Available at http://www2. 90. Ringwald J, Mertz I, Zimmermann R, et al. Hepatitis B virus vaccination of
ncid.cdc.gov/travel/yb/utils/ybGet.asp?section=dis&obj=index.htm blood donors—what costs may be expected? Transfus Med 2005;15:83–92.
&cssNav=browseoyb. Accessed July 5, 2006. 91. Marshall DA, Kleinman SH, Wong JB, et al. Cost-effectiveness of
62. Dodd RY. Bovine spongiform encephalopathy, variant CJD, and blood nucleic acid test screening of volunteer blood donations for hepati-
transfusion: beefer madness? Transfusion 2004;44:628–630. tis B, hepatitis C and human immunodeficiency virus in the United
63. Ironside JW, Head MW. Variant Creutzfeldt-Jakob disease: risk of transmis- States. Vox Sang 2004;86:28–40.
sion by blood and blood products. Haemophilia 2004;10(Suppl 4): 64–69. 92. Muhlbacher A, Zdunek D, Melchior W, Michl U. Is infective blood
64. Boulton F. The impact of variant CJD on transfusion practices in the donation missed without screening for antibody to hepatitis B core
UK. Transf Apheresis Sci 2003;28:107–116. antigen and/or hepatitis B virus DNA? Vox Sang 2001;81:139.
BLOOD DONATION AND COLLECTION
93. Witt D, Kuramoto K, Kemper M, Holland P. Utility of prospective 122. Brown H, McCormack P. An analysis of vasomotor phenomena
study of donors deferred as HTLV indeterminate. [Letter] Vox Sang (faints) occurring in blood donors. BMJ 1942;1:1–5.
2000;78:130–131. 123. McHenry LC, Fazekas JF, Sullivan JF. Cerebral hemodynamics of syn-
94. False-positive serologic tests for human T-cell lymphotropic virus type cope. Am J Med Sci 1961;241:173.
I among blood donors following influenza vaccination, 1992. MMWR 124. Williams GE. Syncopal reactions in blood donors. BMJ 1942;1:783–
1993;42:173–175. 786.
95. Brecher M (ed). Technical Manual, 15th ed. Bethesda, Md., American 125. Newman B. Donor reactions and injuries from whole blood dona-
Association of Blood Banks, 2005, pp 800–801. tions. Transfus Med Rev 1997;11:64–75.
96. de Korte D, Marcelis JH, Verhoeven AJ, Soeterboek AM. Diversion of 126. Williams GE. Syncopal reactions in blood donors. BMJ 1942;1:783–786.
first blood volume results in a reduction of bacterial contamination 127. Ogata H, Iinuma N, Nagashima K, et al. Vasovagal reactions in blood
for whole-blood collections. Vox Sang 2002;83:13–16. donors. Transfusion 1980;20:679–683.
97. U.S. Department of Health and Human Services, Advisory Committee on 128. Engel GL. Psychologic stress, vasodepressor (vasovagal) syncope, and
Blood Safety and Availability. April 2004 Presentations. Available at http:// sudden death. Ann Intern Med 1978;89:403–412.
www.hhs.gov/bloodsafety/presentations/Jaroslav.pdf. April 7–8, 2004. 129. Tizes R. Cardiac arrest following routine venipuncture. JAMA
98. Lord LI, Bronchud MH, Owens S, et al. The kinetics of human granu- 1976;236:1846–1847.
lopoiesis following treatment with granulocyte colony-stimulating 130. Yoda M, Nonoyama M, Shimakura T. Autologous blood donation
factor in vivo. Proc Natl Acad Sci USA 1989;86:9499–9503. before elective off-pump coronary artery bypass grafting. Surgery
99. Neupogen (filgrastim) prescribing information. Amgen, Inc. Thou- Today 2004;34:21–23.
sand Oaks, Calif., Issued December 20, 2004. 131. Galena HJ. Complications occurring from diagnostic venipuncture.
100. Neulasta (pegfilgrastim) prescribing information Amgen, Inc. Thou- J Fam Pract 1992;34:582–584.
sand Oaks, Calif., Issued December 20, 2004. 132. Howanitz PJ, Cembrowski GS, Bachner P. Laboratory phlebotomy. College
101. Granocyte (lenograstim) patient information leaflet. Chugai Pharma of American Pathology Q-probe study of patient satisfaction and compli-
UK Limited, Tokyo, Japan. Revision November 2001. cation in 23,783 patients. Arch Pathol Lab Med 1991;115: 867–872.
102. Liles WC, Rodger E, Dale DC. Combined administration of G-CSF 133. Newman B. Arterial punctures in whole blood donors. Transfusion
and dexamethasone for the mobilization of granulocytes in normal 2001;41:1390–1392.
donors: optimization of dosing. Transfusion 2000;40:643–644. 134. Kumar S, Agnihotri SK, Khanna SK. Brachial artery pseudoaneurysm
103. Lee J-H, Leitman SF, Klein HG. A controlled comparison of the effi- following blood donation. Transfusion 1995;35:791.
cacy of hetastarch and pentastarch in granulocyte collections by cen- 135. Lung J, Wilson S. Development of arteriovenous fistula following
trifugal leukapheresis. Blood 1995;86:4662–4666. blood donation. Transfusion 1971;11:145–146.
104. Burgstaler EA. Blood component collection by apheresis. D01-10.1002/ 136. Gibble J, Ness P, Anderson G, Conry-Cantilena C. Compartment syn-
jca.20043. drome and hand amputation after whole blood phlebotomy: report of
105. Robinson SP, Marks DI. Granulocyte transfusions in the G-CSF era. a case [abstract]. Transfusion 1999;39(Suppl):30S.
Where do we stand? Bone Marrow Transplant 2004;34:839–846. 137. Newman BH, Waxman DA. Blood donation-related neurologic needle
106. Stanworth S, Massey E, Hyde C, et al. Granulocyte transfusions for injury: evaluation of 2 years’ worth of data from a large blood center.
treating infections in patients with neutropenia or neutrophil dysfunc- Transfusion 1996;36:213–215.
tion. Cochrane Database Syst Rev 2005;D005339. 138. Berry PR, Wallis WE. Venipuncture nerve injuries. Lancet 1977;1:
107. Jillella AP, Ustun C. What is the optimum number of C34+ periph- 1236–1237.
eral blood stem cells for an autologous transplant? Stem Cells Dev 139. McLeod BC, Price TH, Owen H, et al. Frequency of immediate
2004;13:597–606. adverse effects associated with apheresis donation. Transfusion 1998;
108. Singhal S, Powles R, Treleaven J, et al. A low CD34+ cell dose results 38:938–943.
in higher mortality and poorer survival after blood or marrow stem 140. Despotis GJ, Goodnough LT, Dynis M, et al. Adverse events in platelet
cell transplantation from HLA-identical siblings: should 2 × 106 apheresis donors: a multivariate analysis in a hospital-based program.
11
CD34+ cells/kg be considered the minimum threshold? Bone Mar- Vox Sang 1999;77:24–32.
row Transplant 2000;26:489–496. 141. Katz AJ, Genco PV, Blumberg N, et al. Platelet collection and transfusion 181
109. Feller N, van der Pol MA, Waaijman T, et al. Immunologic purging of using the Fenwal CS-3000 cell separator. Transfusion 1981;21:560–563.
autologous peripheral blood stem cell products based on CD34 and 142. Simon TL, Sierra ER, Ferdinando B, et al. Collection of platelets with
CD133 expression can be effectively and safely applied in half of the a new cell separator and their storage in a citrate-plasticized container.
acute myeloid leukemia patients. Clin Cancer Res 2005;11:4793–4801. Transfusion 1991;31:335–339.
110. Kawabata Y, Hirokawa M, Komatsuda A, Sawada K. Clinical applica- 143. Lazarus EF, Browning J, Norman J, et al. Sustained decreases in plate-
tions of CD34+ cell-selected peripheral blood stem cells. Ther Apher let count associated with multiple, regular plateletpheresis donations.
Dial 2003;7:298–304. Transfusion 2001;41:756–761.
111. Ziegler BL, Kanz L. Expansion of stem and progenitor cells. Curr Opin 144. McCullough J. Introduction to apheresis donations including his-
Hematol 1998;5:434–440. tory and general principles. In McLeod BC, Price TH, Drew MJ (eds).
112. Heal JM, Liesveld JL, Phillips GL, Blumberg N. What would Karl Apheresis: Principles and Practice. Bethesda, Md., American Associa-
Landsteiner do? The ABO blood group and stem cell transplantation. tion of Blood Banks, 1997, p 40.
Bone Marrow Transplant 2005;36:747–755. 145. Leitman SF, Oblitas JM. Optimization of granulocytapheresis mobili-
113. FDA Memorandum: Revision of FDA Memorandum of August 27, zation regimens using granulocyte colony stimulating factor (G-CSF)
1982: Requirements for Infrequent Plasmapheresis Donors. Bethesda, and dexamethasone [abstract]. Transfusion 197;37(Suppl):67S.
Md., Food and Drug Administration, 1995. 146. Price TH, Bowden RA, Boeckh M, et al. Phase I/II trial of neutrophil
114. Boynton MH, Taylor ES. Complications arising in donors in a mass transfusions from donors stimulated with G-CSF and dexamethasone
blood procurement project. Am J Med Sci 1945;209:421–436. for treatment of patients with infections in hematopoietic stem cell
115. Fainting in blood donors. A report to the Medical Research Council transplantation. Blood 2000;95:3302–3309.
prepared by a subcommittee of the Blood Transfusion Research Com- 147. Occupational Safety and Health Administration. Occupational expo-
mittee. BMJ 1944;1:279–283. sure to bloodborne pathogens: needlesticks and other sharps injuries:
116. Tomasulo PA, Anderson AJ, Paluso MB, et al. A study of criteria for Final rule. Fed Register 2001;66:5317–5325.
blood donor deferral. Transfusion 1980;20:511–518. 148. Page PL. Risk of hepatitis B exposure in regional blood services. Trans-
117. Kasprisin DO, Glynn SH, Taylor F, et al. Moderate and severe reactions fusion 1987;27:242–244.
in blood donors. Transfusion 1992;32:23–26. 149. From California Blood Bank Society. Available at http://www.cbbsweb.
118. Khan W, Newman B. Comparison of donor reaction rates in high- org/enf/2001/gloves.html, last modified November 27, 2001 and http://
school, college, and general blood drives [abstract]. Transfusion www.cbbsweb.org/enf/2005/gloves2.html, last modified August1, 2005.
1999;39(Suppl):31S. 150. Code of Federal Regulations. 29 CFR 1910.1030. Washington, D.C.,
119. Trouern-Trend J, Cable R, Badon S, et al. Vasovagal reaction in blood U.S. Government Printing Office, July 1, 2003.
donors: Influence of gender, age, donation status, weight, blood pres- 151. Grindon AJ, Keelan LT, Lenes BA. HIV post-exposure prophy-
sure, and pulse. A case-controlled multicenter study. Transfusion laxis for blood center healthcare workers [abstract]. Transfusion
1999;39:316–320. 1998;38(Suppl):109S.
120. Poles FC, Boycott M. Syncope in blood donors. Lancet 1942;2:531–535. 152. Updated U.S. Public Health Service guidelines for the management of
121. Maloney WC, Lonnergan LR, McClintock JK, et al. Syncope in blood occupational exposure to HIV and recommendations for postexpo-
donors. NEJM 1946;234:114–118. sure prophylaxis. MMWR 2005;54(RR09):1–17.
BLOOD BANKING 153. Updated U.S. Public Health Service guidelines for the management of and the notification of consignees and blood recipients of donor test
occupational exposures to HBV, HCV, and HIV and recommendations results for anti-HCV. Federal Register Docket No. 98D-0143. Washing-
for postexposure prophylaxis. MMWR 2001;50(RR11):1–52. ton, D.C., Government Printing Office, September 23, 1998.
154. Reason J. Human error: Models and management. BMJ 2000;320: 158. Nordenberg T. Recalls: FDA, industry cooperate to protect consumers.
768–770. FDA Consumer Magazine 1995;29:24–27.
155. Reason J. Beyond the organisational accident: The need for “error wis- 159. U.S. Food and Drug Administration, Center for Food Safety and
dom” on the frontline. Qual Saf Health Care 2004;13:28–33. Applied Nutrition. FDA recall policies. Industry Affairs Staff Brochure,
156. U.S. Food and Drung Administation, FDA 21 CFR 610, 46 and 47 June 2002. Avaiable at http://www.cfsan.fda.gov/~Ird/recall2.html.
Medicare and Medicaid programs: hospital standard for potentially Accessed June 4, 2006.
HIV infectious blood and blood products. Washington, D.C., Govern- 160. U.S. Department of Health and Human Services. U.S. Food and Drug
ment Printing Office, September 9, 1996, revised April 1, 2005. Administration. Final Rule: Reporting of Biological Product Deviations
157. U.S. Department of Health and Human Services, U.S. Food and Drug in Manufacturing. Federal Register Docket No. 97N-0242. Washington
Administration, Center for Biologics Evaluation and Research Guid- D.C. Government Printing Office, November 7, 2000.
ance for Industry. Current good manufacturing practice for blood 161. U.S. Food and Drug Administation, Center for Biologics and Research.
and blood components: (1) Quarantine and disposition of units from Biological Product Devitation Reports—Annual Summary for Fiscal
prior collections from donors with repeatedly reactive screening tests Year 2004. Avilable at http://www.fda.gov/cber/biodev/bpdrfy04.htm.
for antibody to Hepatitis C Virus (anti-HCV);(2) Supplemental testing, Last modified April 28, 2005.
II
182
Chapter 12
Blood Manufacturing: Component
Preparation, Storage, and Transportation
Shealynn B. Harris ● Christopher D. Hillyer
Skin Antisepsis
Modern transfusion medicine in highly developed nations
is based on the use of components, including cellular and Bacterial contamination of blood products is a leading cause
plasma components, prepared or manufactured from whole of transfusion-related morbidity and mortality.1 Because the
blood collected from a volunteer donor. This is often called donor is most often the source of bacterial contamination, either
component therapy and offers both therapeutic and economic through bacteremia or skin flora, methods to reduce the risk of
advantages, including the provision of concentrated products contamination begin with proper donor screening, arm inspec-
for specific and targeted transfusion management and the tion, and skin antisepsis.2 For all types of collection procedures,
cost-effective and efficient use of this valuable and often donor preparation starts with the inspection of both arms for
limited resource. Along with the benefits of component therapy, signs of intravenous drug use or skin lesions that pose an infec-
however, there are also challenges posed by the complex man- tious disease risk.3 This is followed by the appropriate application
ufacturing process required for component production. of one of the FDA approved skin disinfection methods for blood
In this chapter, the essentials of the blood manufacturing collection.4 Most methods specify use of an iodine disinfectant,
process will be discussed, with an emphasis placed on blood either povidone iodine or tincture of iodine. Because iodine can
collection, component preparation, labeling, and storage. result in skin reactions in sensitive individuals, the FDA-approved
Although they are mentioned in this chapter, the areas of regu- 2% chlorhexidine gluconate in 70% isopropyl alcohol method
lation, donor eligibility, quality control, infectious disease test- is recommended by the AABB for iodine-allergic donors.5 The
ing, and clinical use of components are addressed in greater use of green soap and 70% isopropyl alcohol for iodine-allergic
detail in dedicated chapters elsewhere in this textbook. donors is no longer recommended by the AABB, based on data 12
that show significantly more residual skin flora after cleaning
as compared to the other currently approved skin antisepsis 183
BLOOD COLLECTION methods listed previously.6
Donor Care and Complications
Donor Preparation and Care
Care of the donor is essential to a safe and effective dona-
Essentially, the manufacturing of components can be tion process. Because the collection staff serve as the primary
thought of as beginning with the collection of blood from interface between donors and the blood community, it is
an eligible volunteer donor by whole blood phlebotomy imperative that staff act in a courteous, professional manner
or automated apheresis. Indeed, the process of blood col- to help promote donor satisfaction and repeat donations. In
lection is regulated under the CFR and is subject to AABB addition, to ensure donor safety, collection staff must also be
Standards. Accordingly, establishments that collect blood aware of potential donor reactions and must be trained in
for transfusion or “for further manufacture” need to have collection facility procedures for reaction management.
adequate and appropriate standard operating procedures Venipuncture, intravascular volume shifts, and apher-
(SOPs), training, and equipment validation for donor esis-specific citrate use present risks for donation compli-
collection. An effective donor selection, preparation, and cations.7,8 For apheresis donation, the overall frequency
collection process serves several purposes: it ensures the of complications is approximately 2%, with the majority
quality of the final products and components; it minimizes related to mild hypocalcemia from the citrate anticoagulant.9
risks to both the donor and transfusion recipient; and it In contrast, for whole blood donation, the overall frequency
enhances donor satisfaction, which increases the likelihood of complications is at least 10 times greater (20% to 40%),
of repeat donations. with bruising, arm discomfort, fatigue, vasovagal events, and
Prior to collection, all allogeneic blood donors must be nerve irritation, listed in order of decreasing frequency, as
accurately identified, provide informed consent, and be the complications most often reported.7
evaluated for general suitability requirements, which include In general, for the common complications just listed,
criteria for age, weight, donation interval, medical and preg- donors typically do not require significant medical interven-
nancy history, infectious diseases and related risk behavior, tion and readily recover without lasting sequelae. However,
hemoglobin level, temperature, pulse, and blood pressure. If for donors with phlebotomy-related cutaneous nerve irrita-
a donor is accepted for collection, the donor’s identity must tion or injury (incidence, 0.02%–0.9%), a study by Newman
be effectively linked to all resulting components and test and Waxman10 showed that approximately 30% will experi-
samples by a unique alphanumeric identifier. ence symptoms, typically mild numbness, lasting 1 month or
BLOOD BANKING longer, and the majority of these donors will request a physi- primary collection bag with an anticoagulant-preservative
cian consultation. In the same study, approximately 25% of solution. For routine donations, a collection kit that contains
nerve injury cases were associated with a hematoma, indicat- a 450-mL or 500-mL primary bag is used to accommodate
ing improper phlebotomy technique as a contributing fac- the standard whole blood collection. In addition, 250-mL
tor. However, in another report by Newman,11 approximately and 450-mL bags can be used for special collections, such as
40% of nerve injuries occurred after an uncomplicated autologous donations and low-volume units. In the United
phlebotomy, which suggests that phlebotomy-related nerve States, 250-mL collection kits are not licensed for volunteer
injury may not be completely preventable. Furthermore, or directed donation and must be reserved for autologous
an anatomical study by Horowitz12 of the cutaneous nerve- donors. Depending on the kit manufacturer and process-
superficial vein relationships at common venipuncture sites ing requirements, the integrated system may include one or
suggests that cutaneous nerves often overlie superficial veins, more of the following: a diversion pouch, satellite bags for
which increases the likelihood for mild nerve injury, even component preparation (i.e., double, triple, or quad packs),
with proper phlebotomy technique. a bag with additive solution for red blood cell (RBC) storage,
Rarely, phlebotomy-related nerve injury is more severe, and an in-line leukoreduction filter.
resulting in a chronic, disabling neuropathic pain syndrome, Some collection kit accessories are provided separately
Complex Regional Pain Syndrome Type 2 (CRPS-II), which and can be attached to the primary collection kit through
is characterized by shooting, burning, or electrical pain often a sterile connection device (SCD), which applies a sterile
associated with motor dysfunction.13 In contrast to donors with weld between two pieces of compatible tubing. To effectively
mild nerve injury, 80% of donors with CRPS-II report traumatic maintain the integrity of the closed system, SCDs must be
phlebotomy with multiple attempts or hematoma formation.13 used as recommended, and the sterile weld must be checked
Although the vast majority of donor complications do not for defects. If there is a defective connection, the system is
require emergency intervention or hospitalization, very seri- considered open, and the product expiration is affected.20
ous events, defined by the need for hospitalization, do occur Guidelines for the proper use of SCDs are provided by the
and are more likely with apheresis donation (1 in 20,000)14 FDA in a guidance document.21
than with whole blood donation (1 in 200,000).15 Because Manufacturers of apheresis equipment supply custom-
serious adverse events can take place during or immedi- ized apheresis collection and storage kits that meet the
ately after donation, a plan for managing these events must unique specifications of each apheresis instrument and type
be in place at the collection facility. Collection staff should of component collection. In general, apheresis disposables
be familiar with signs and symptoms of donation compli- incorporate a specialized, integrated system of tubing and
cations and should closely monitor donors throughout the bags that allows for closed, sterile collection. Similar to whole
donation process.16 blood collection kits, the type of blood bags in an apheresis
As complications can occur or be noted after the donor kit depends on the type of component collected and the
leaves the collection facility, donors should be provided with component’s storage requirements.
II postphlebotomy information,17 which includes care of the
Diversion Pouches
venipuncture site, limits on strenuous physical activity, dis-
184 cussion of the need for increased fluid consumption, and Diversion pouches have been recently introduced as a means
contact instructions in the event of an adverse reaction. to help further reduce bacterial contamination from donor
skin flora. Even with proper skin antisepsis during donor
Donor Test Samples preparation, residual deep-seated bacteria can be introduced
Current standards require that donors be tested for multiple into the collected blood through coring of the skin during
infectious disease markers, ABO and Rh type, and if previ- venipuncture.22 Several studies have shown that the collection
ously transfused or pregnant, red cell alloantibodies.18 These of the initial volume of blood into an in-line diversion pouch
testing requirements necessitate the collection of several vac- can provide an additional safeguard against skin flora con-
uum tubes of blood from the donor. To eliminate the need tamination. By using an in vitro model with the common skin
for a second donor venipuncture, most collection and stor- contaminant Staphylococcus aureus, Wagner and colleagues23
age kits include a convenient method for donor sampling, demonstrated that the diversion of the first 21 mL to 42 mL of
either a diversion pouch or an in-line sampling system. Once blood flow results in a 95% to 98% reduction of bacteria in the
test samples are obtained, the tubes must be properly labeled final collection volume. Another study by de Korte and associ-
with the donor identifier to ensure effective traceability and ated24 showed a significant overall decrease in Staphylococcus
proper result reporting. species in whole blood collections (0.14% to 0.03%, p = 0.015)
with the use of a 10-mL diversion system. Several manufactur-
Collection and Storage Kits
ers offer an FDA-approved, in-line diversion pouch as part of
In addition to donor-derived bacteria, environmental patho- a blood collection kit. In addition to reducing bacterial con-
gens, such as Serratia marcecsens, encountered during blood tamination, these pouches have been designed to accommo-
manufacturing and transport serve as a potential contami- date sufficient volume for donor sampling for required and
nation source.2,19 A closed manufacturing system helps standard infectious disease and serological testing.
to ensure that all surfaces in direct contact with the blood
Blood Bag Plastics: Polymers and Plasticizers
remain sterile and pyrogen free throughout blood manu-
facturing and storage. Closed collection, processing, and All blood collection and storage systems are made of dis-
storage kits (also known as disposables) have been devel- posable plastic. There are many types of plastics available;
oped as an effective means of preventing the introduction of however, only a few plastic materials meet the specifications
environmental pathogens. for blood processing and storage. The fundamental struc-
Whole blood collection and storage kits incorporate, at ture of a plastic is the polymer, which is a repeatedly linked
a minimum, a 16-gauge phlebotomy needle, tubing, and a chain of a simple base chemical called a monomer.25 To be
BLOOD MANUFACTURING
functional, polymers require additives, such as plasticizers, to products. Further studies of DEHP toxicity in rodents have
increase flexibility and stability. In general, the unique prop- suggested the potential for developmental and reproductive
erties of each type of plastic are determined by the monomer, effects in humans, although human toxicity has not been
chain length, and additives. conclusively proven.41 The National Toxicology Program’s
Plastic materials for blood bags must have particular Center for Risks to Human Reproduction provides a recent
qualities. These qualities include, depending on the bag’s extensive review of DEHP toxicity studies and suggests the
intended use, adequate flexibility and strength to withstand need for additional evidence to conclusively prove DEHP’s
centrifugation and handling; temperature resistance for toxic effects.41 As stated in the report, the primary concern
both steam sterilization and freezing; limited toxicity to the is the effect of DEHP in neonates and children who are
transfusion recipient; compatibility with cells and plasma to chronically transfused or undergo blood volume exchange
reduce component adulteration; selective permeability for and are exposed to a high cumulative dose of DEHP. For
cellular gas exchange; water and pathogen impermeability; example, with a single-procedure double-volume exchange
and transparency for effective product visualization.25 Of all in a neonate, the dose of DEHP is approximately 1800 μg/kg
available polymers, polyvinylchloride (PVC) has been shown bw/day; and for a neonate treated with replacement transfu-
to be the most compatible with component manufacturing. sions, the average dose is 300 μg/kg bw/day. For comparison,
During the past 50 years, most innovations in blood bag the current maximum allowable occupational inhalational
materials have focused on improvements in PVC with the exposure to DEHP in an adult is 700 μg/kg bw/day. At this
incorporation of different plasticizers. The two most com- time, the potential toxicity of DEHP is still unclear, and cur-
mon plasticizers in current use for PVC blood bags are di-2- rently, DEHP remains the most common plasticizer used for
ethylhexylphthalate (DEHP) and tri-2-ethylhexyl trimellitate blood bags.
(TEHTM).25 In addition to increasing PVC flexibility, plasti-
Anticoagulant and Preservative Solutions
cizers act to improve gas exchange across the bag barrier and
help to stabilize the RBC membrane during storage. Under storage conditions, RBCs show predictable, time-
Effective gas exchange is especially important for maintain- dependent adverse changes, which together are known as the
ing platelet viability during storage. Platelets are particularly red cell storage lesion (Table 12–1). Key consequences of the
sensitive to an acidic environment, with major shape changes storage lesion are a reduction in post-transfusion RBC sur-
and cell death occurring at a pH less than 6.0.26 In a gas-imper- vival and impairment of oxygen transport, which are effects
meable container, the high metabolic rate of room-temperature that lead to an overall decrease in transfusion efficacy.42
platelets causes the pH to drop rapidly as lactic acid accumu- Although the cellular and molecular events that lead to the
lates with oxygen consumption and compensatory anaerobic storage lesion have not been completely elucidated, certain
metabolism.27 To meet the need for enhanced platelet viability principles are known. RBCs are anucleate, and thus, they
and longer storage, platelet storage bags in current use (“sec- lack the synthetic machinery to renew structural proteins
ond-generation bags”) are made with selectively permeable and enzymes. Without replacement of structural proteins
plastics, thinner walls, and increased surface area for better and enzymes to maintain the RBC membrane and meta- 12
gas exchange.28–31 In addition to PVC-based bags, a polyole- bolic machinery, RBCs undergo senescence during storage.
fin-based bag has been shown to effectively maintain platelet RBCs are also devoid of mitochondria, the organelle required 185
viability and is approved by the FDA for platelet storage.28 for oxidative phosphorylation. To produce ATP for cellular
For RBCs, in addition to the benefit of improved gas energy, RBCs rely on anaerobic metabolism (the Embden-
exchange, plastic containers also appear to improve RBC sta- Meyerhof pathway), which requires a continuous supply of
bility during storage. This property was initially suggested by glucose and adenine and produces large quantities of lactic
early work that showed significantly reduced hemolysis and acid and hydrogen ions.43 With acid accumulation and
osmotic fragility of blood stored in plastic bags compared decreasing pH, enzymes and structural proteins undergo
to blood stored in glass containers.32 It was later shown that changes that compromise cell metabolism and stability.
RBCs stored in the presence of the plasticizer DEHP have In addition to these structural and metabolic changes,
improved in vitro survival,33 decreased microvesicle forma- RBC function is affected with storage. To effectively deliver
tion,34 and better post-transfusion survival35 compared to oxygen, RBCs require sufficient levels of 2,3-diphosphoglyc-
RBCs stored in the absence of DEHP. Additional studies have erate (2,3-DPG). By competing with oxygen for hemoglobin
shown that DEHP, a highly lipophilic compound, leaches binding sites, 2,3-DPG acts to displace oxygen from hemo-
out of PVC-DEHP containers and into the RBC membrane globin, thus making the oxygen more readily available for
during storage,36,37 thus stabilizing the RBC membrane and tissue uptake. During storage, the concentration of 2,3-DPG
prolonging RBC survival. Given the superior quality of RBCs in RBCs declines. With low levels of 2,3-DPG, hemoglobin
stored in DEHP and the lack of a commercially available binds oxygen more avidly, and the oxygen delivery function
alternative, PVC-DEHP is currently the only plastic used for of the stored RBCs is reduced. Within approximately 3 weeks
RBC storage.25 of storage, RBC 2,3-DPG concentration falls to below 10%.44
The use of DEHP is not without concern, and the rela- This effect is reversed after transfusion, however, because the
tive safety of PVC-DEHP has been questioned for the past transfused RBCs replenish 2,3-DPG within 24 to 48 hours in
20 years.38 Early work by Jacobson and colleagues,39 using a circulation.45,46
rhesus monkey chronic platelet transfusion model, suggested Two important methods for RBC preservation are refrig-
that the chronic transfusion of DEHP-stored platelets is asso- eration and the use of supplemental anticoagulant-preserva-
ciated with an increased rate of hepatic dysfunction and liver tive solutions. With refrigeration (1°C–6°C), the metabolic
histopathology compared to chronic transfusion with non– rate of RBCs is reduced significantly, and senescence is
DEHP-stored platelets. A postmortem study by Hillman and delayed.47 However, to effectively extend red cell storage
coworkers40 also showed accumulation of DEHP in the tis- beyond 5 days, a preservative solution is necessary.48 The
sues of critically ill human neonates who had received blood efficacy of a preservative in maintaining stored red cells is
BLOOD BANKING
Table 12–1 Biochemical Changes of Stored Red Blood Cells (RBC Storage Lesion)
Whole Red Blood Whole Red Blood Red Blood Red Blood Red Blood
Variable Whole Blood Blood Cells Blood Cells Cells Cells Cells
Days of storage 0 21 0 0 35 35 42 42 42
No. of viable cells 100 80 100 100 79 71 76(64–85) 84 80
(24 hours post-
transfusion)
pH (measured 7.20 6.84 7.60 7.55 6.98 6.71 6.6 6.5 6.5
at 37°C)
ATP (% of initial 100 86 100 100 56(±16) 45(±12) 60 59 68.5
value)
2,3-DPG (% of 100 44 100 100 <10 <10 <5 <10 <5
initial value)
Plasma K+ (mmol/L) 3.9 21 4.20 5.10 27.30 78.50‡ 50 46 45.6
Plasma hemoglobin 17 191 82 78 461 658.0‡ N/A 386 N/A
(mg/L)
% Hemolysis N/A N/A N/A N/A N/A N/A 0.5 0.9 0.6
*
Based on information supplied by the manufacturer.
†
Simon TL, Hunt WC, Garry PJ. Iron supplementation for menstruating female blood donors. Transfusion 1984;24:469–472.
‡
Values for plasma hemoglobin and potassium concentrations may appear somewhat high in 35-day stored RBC units; the total plasma in
these units is only about 70 mL. From American Association of Blood Banks. Technical Manual, 14th ed. Bethesda, MD, AABB, 2002.
determined by assessing the preservative-stored RBC viabil- blood collection is less than 300 mL, the volume of antico-
ity post-transfusion. Specifically, in order to meet established agulant-preservative solution must be adjusted to maintain
regulations and standards, 75% of the stored RBCs must an approximate anticoagulant-to-blood ratio of 1.4:10.
remain in circulation at 24 hours after transfusion,42 as mea- To extend the expiration time to 42 days, RBCs are sus-
sured by an in vivo 51Cr red cell radiolabeling and transfu- pended in an additive solution (100 mL) that is transferred
sion method.49 to the primary bag after the plasma is expressed from the
The common anticoagulant-preservative solutions used whole blood collection. Additive solutions decrease in vitro
in transfusion practice are acid citrate dextrose (ACD-A), hemolysis by stabilizing the RBC membrane, either through
II citrate phosphate dextrose (CPD and CP2D), and citrate the action of mannitol or citrate.42 With an approximate
phosphate dextrose adenine CPDA-1. RBCs suspended in hematocrit of 60% for additive-stored RBCs, these solutions
186 these solutions are stored at a hematocrit (Hct) between 70% also act to enhance transfusion flow and reduce product
and 80%. As seen in Table 12–2, anticoagulant-preservative administration time by decreasing the viscosity of packed
solutions vary in content and the approved length of RBC RBCs. Listed in Table 12–3, the currently approved addi-
storage, with the RBC storage limit for CPD and CP2D at 21 tive solutions include the saline-adenine-glucose-mannitol
days and CPDA-1 at 35 days. Additive solutions (AS-1, AS-3, (SAGM) formulations, AS-1 (Adsol) and AS-5 (Optisol),
AS-5), discussed in the following section, are approved for and a non–mannitol-based solution, AS-3 (Nutricel). AS-1 and
RBC storage up to 42 days. AS-5 can be coupled with any of the anticoagulant-preservative
Each component of the preservative solution has a spe- solutions in a whole blood collection system; however,
cific metabolic support function. Dextrose and adenine AS-3 requires supplemental glucose and must be paired with
serve as substrates for ATP production, and phosphate acts CP2D for a whole blood collection kit.43
as a pH buffer and substrate for 2,3-DPG formation. For a Current efforts to develop media that prolong RBC storage
whole blood collection kit, the volume of the anticoagulant- beyond 6 weeks have led to a new generation of experimen-
preservative solution is specific for the collection volume. tal additive solutions (EASs) that are under investigation.50-52
For a 450-mL primary collection bag, the anticoagulant- Although most EASs increase storage to between 7 and 10
preservative solution is 63 mL, and for a 500-mL primary weeks, reports by Meryman and associates53 and Hess and
collection bag, the solution volume is 70 mL. If the whole associates54 provide evidence for effective liquid RBC storage
ACD, acid citrate dextrose; CPD, citrate phosphate dextrose; CPDA-1, citrate phosphate dextrose adenine.
BLOOD MANUFACTURING
Table 12–3 Content of Additive Solutions (mM)
for up to 12 weeks. The potential benefits of prolonged stor- donation is 50 kg or 110 lb (10.5 mL/kg × 50 kg = 525 mL).
age with EASs include reduction of allogeneic RBC outdate To prevent overcollection, a practical method for monitoring
rates; decreased donor exposure for premature infants who volume during donation is to weigh the whole blood unit.
require repeat, small-volume maintenance transfusions; and For a normal donor hematocrit, the specific gravity of whole
an increased collection window for autologous donors. The blood is 1.053,66 so a commonly used conversion factor to
primary limitation of these new solutions is the inability calculate the collected volume (mL) from the unit weight (g)
to adequately preserve 2,3-DPG levels. Because most RBC is 1.06 g/mL.
transfusions are given between 12 and 21 days of storage, the Another physiologic consequence of blood donation
value of extended RBC storage for routine transfusion prac- is the temporary reduction of donor RBC mass. Because
tice is debatable; however, extended storage may be of benefit most of the body’s iron is carried in RBCs, loss of RBC mass
in rural outposts and military settings.54 after repeated donations can lead to anemia, especially in
As previously mentioned, although 2,3-DPG levels replen- premenopausal females.67 To reduce the risk of anemia, an
ish within 48 hours of transfusion, stored RBCs older than 2 allogeneic donor is required to have a hemoglobin level of
weeks have reduced oxygen transport function immediately greater than or equal to 12.5 g/dL (Hct 38%) and is limited
after transfusion. This property of “older” RBCs may have to donating once every 8 weeks, unless examined and cleared
implications for effective resuscitation of certain patients by a physician to donate more frequently, which in practice
with more urgent or greater oxygen delivery needs, such as occurs only in unusual situations with directed donors.
those in cardiac surgery, trauma, massive transfusion, and
Donors with Hereditary Hemochromatosis
critical care settings.55-59 Thus, research efforts for improving
additive solutions continue to focus on methods for improv- Whole blood collected for the purpose of therapeutic phle-
ing 2,3-DPG preservation. A recent report by Högman and botomy must be labeled with the donor’s disease if the blood
colleagues60 provides promising data for a modified EAS, is used for transfusion.68 Although therapeutic collections 12
Erythro-Sol 2, which maintains RBC 2,3-DPG levels at a from otherwise healthy individuals with hereditary hemo-
normal level over a 2-week storage period. Another study by chromatosis are considered safe for allogeneic transfusion, 187
Kurup and coworkers61 compared the RBC storage proper- these units are often rejected by hospital transfusion ser-
ties of SAGM and a modified SAGM solution. Although the vices due to the required disease label.69 Recently, however,
SAGM showed a 99% decrease in 2,3-DPG levels by day 28, the the FDA introduced the option for a variance that allows
modified SAGM maintained day 28 RBC 2,3-DPG levels near collection facilities to exclude the disease designation from
baseline, with an overall decrease of less than 1%. Although whole blood units obtained by therapeutic phlebotomy from
EASs have received approval in Europe,62 an extended storage individuals with hereditary hemochromatosis.70 Under the
solution is not currently approved in the United States. variance, the donor must meet the standard allogeneic donor
suitability requirements, and the collection facility must not
charge a fee for the collection. In addition, to collect from a
Whole Blood Collection
donor more frequently than every 8 weeks, a physician pre-
scription for therapeutic phlebotomy or a physical exam and
Donor Requirements physician’s certification of good health is required.
In addition to enhancing transfusion recipient safety, some
Collection Procedure
of the donor eligibility requirements for allogeneic and
directed whole blood donation are established for the pur- Phlebotomy is performed using the integral donor nee-
pose of enhancing donor safety and include limits for donor dle supplied with the sterile collection and storage kit.
volume deficit, weight, hemoglobin, and donation frequency. Immediately prior to phlebotomy, a tourniquet or blood
Studies of controlled blood loss in normal, healthy adult vol- pressure cuff is applied and is kept in position throughout
unteers indicate that up to 15% of a donor’s blood volume collection. To limit trauma to subcutaneous tissue, a single
(BV) can be safely removed without significant physiologic venipuncture is performed, and the needle is secured to pre-
signs or symptoms.63,64 Thus, the maximum allowable intra- vent movement or displacement. If signs of an arterial punc-
vascular volume deficit for a single donation is 10.5 mL/kg ture are present, such as rapid filling of the blood bag (<4
(15% BV), which includes the volume for test samples and minutes) and bright red, pulsating blood flow, the procedure
in residual tubing.65 In the United States, the standard vol- should be immediately discontinued, the needle removed,
ume for a whole blood donation is 450 mL to 500 mL ± 10%. and extended pressure applied.7
With test samples, the entire collected blood volume for a Because the internal surface of the collection tubing is
typical donation is approximately 525 mL. Therefore, the not filled with an anticoagulant, adequate blood flow must
standard minimum donor weight for routine whole blood be maintained to prevent coagulation factor activation and
BLOOD BANKING clotting prior to the blood entering the primary collection composition is abnormal, the donor is ineligible until the
bag. As the blood enters the primary container, it should be values normalize. Frequent donors can donate a maximum
mixed frequently with the anticoagulant to avoid clot forma- of 2 times in a week, as long as the collections occur at least
tion. With sufficient flow, the duration of a whole blood col- 2 days apart. The FDA-approved volume limits for a single,
lection is approximately 10 minutes. The effect of donation automated plasma collection are specific for each apheresis
time on whole blood–derived plasma and platelets does not instrument and provided by the instrument manufacturer.80
appear to be significant until the donation exceeds 15 min- The annual limits for maximum donated plasma volume are
utes.71,72 For whole blood collections that exceed 15 minutes, 12 liters (110 to 175 lb) and 14.4 liters (>175 lb).75
data suggest increased thrombin generation in the plasma For plateletpheresis, a platelet count is not required for
and decreased platelet counts in the platelet concentrates. the first collection and is only required for subsequent col-
Although there is no FDA regulation for maximum collec- lections if the donation interval is more frequent than every
tion time, based on these data, it is recommended that whole 4 weeks.81 If the platelet count drops below 150,000/μL, the
blood collections that exceed 15 minutes be restricted to red donor usually is temporarily deferred until the count rises
cell production and not be used for platelet concentrates or above 150,000/μL. Regardless of platelet count, a donor
plasma components for transfusion.73 who has ingested aspirin or aspirin-containing medications
within 36 hours prior to collection is ineligible for platelet-
Automated Apheresis Collection pheresis donation, because aspirin irreversibly inhibits plate-
Apheresis is the process by which a desired component of let function. The frequency limits for plateletpheresis are two
whole blood is separated and collected, and the unselected collections in a 7-day period with at least a 48-hour interval
constituents of whole blood are returned to the donor. between collections, and a maximum of 24 collections per-
The types of blood components collected by apheresis formed in a year.74 The total volume limits (excluding anti-
of volunteer donors include plasma (plasmapheresis), coagulant) for a single, automated plateletpheresis collection
platelets (plateletpheresis), RBCs (erythrocytapheresis), are 500 mL (110 to 175 lb) and 600 mL (>175 lb). As with
and leukocytes (leukapheresis). Although manual apheresis automated plasma collection, the annual volume limits for
methods are available, automated apheresis is more conve- automated plateletpheresis are 12 liters (110 to175 lb) and
nient and provides better product safety and quality. Hence, 14.4 liters (>175 lb).
manual apheresis is not routinely employed for component At the time of this writing, the FDA has posted a draft
collection. guidance document for automated platelet collection, and if
accepted for implementation, it will replace previous guide-
Donation Requirements lines.82 Proposed changes to the previous guidelines include
In addition to meeting the standard eligibility requirements additional eligibility criteria for platelet-inhibiting medica-
for allogeneic blood donation, routine apheresis donors must tions, such as Clopidogrel, Ticlopidine, and NSAIDs, and an
also meet criteria specific for the type of apheresis procedure extension of the aspirin deferral from 3 days to 5 days post-
II and frequency of donation. For all apheresis collections, a ingestion. In addition, because modern apheresis instru-
donor is not eligible if a whole blood donation or equivalent ments are capable of more than one platelet collection, new
188 apheresis RBC loss (>200 mL) occurred within the previous recommendations for double and triple platelet donors have
8 weeks, unless the extracorporeal RBC volume during the been proposed.
current apheresis procedure is less than 100 mL.74–76 If the RBC Automated RBC apheresis donor requirements, includ-
loss for a single procedure is greater than 300 mL, the donor ing height, weight, and hemoglobin/hematocrit, are defined
is deferred for 16 weeks.76 The collection facility must monitor for each FDA-approved instrument and provided by the
total annual donor RBC loss from all apheresis and whole manufacturer. In addition to instrument-specific criteria,
blood donations to ensure that the RBC volume does not there are general donor suitability requirements for auto-
exceed that allowed for cumulative yearly whole blood dona- mated RBC collections.76 For single apheresis collections
tion.77 Regardless of standard requirements, when a specific (RBC volume of 200 mL to 300 mL), the requirements are
component is deemed to be of particular value to a specific similar to whole blood donation (e.g., limited to 8-week
recipient (e.g., human leukocyte antigen [HLA]-matched intervals). However, the criteria for double RBC collections
platelets), an otherwise ineligible donor may donate if (RBC volume > 300 mL) are somewhat different. A donor
approved by the medical director.78 hemoglobin or hematocrit level is required prior to each
For automated plasmapheresis, in addition to the routine donation, as measured by a quantitative method and not
donor test samples, the FDA requires the collection of a blood by the copper sulfate (CuSO4) method. This preprocedure
sample on the day of the first physical exam for donation or value, along with the donor’s height, weight, and gender, is
on the day of the first plasmapheresis procedure.79 Tests on used to determine a predicted postprocedure hemoglobin
the sample include a total protein quantification and plasma or hematocrit for donor eligibility.76 If a drop in hemoglo-
or serum immunoglobulin composition (e.g., serum protein bin of <10g/dL or in hematocrit of <30% is predicted, the
electrophoresis). Additional donor qualifications are defined donor should not be considered for a double RBC collection.
by donation interval. “Infrequent” plasmapheresis is per- Instrument manufacturers provide device-specific nomo-
formed every 4 weeks or less frequently, and “frequent” plas- grams and formulae to assist collection facilities with this
mapheresis is performed at an interval of 4 weeks, or more determination. If a double RBC collection is performed, the
frequently.75 For frequent donors, a serum or plasma protein donor is deferred from whole blood or apheresis collection
study is required prior to each donation, and the cumula- for 16 weeks.
tive results must be evaluated by a qualified, licensed physi- Several apheresis devices are capable of collecting mul-
cian within 21 days of the sample draw to determine donor tiple concurrent components (see Table 12–4). For multi-
acceptability for subsequent plasmapheresis collections. If the component collection, the donor must meet eligibility and
total protein level is less than 6 g/dL or the immunoglobulin donation frequency criteria for each collected component. In
BLOOD MANUFACTURING
Table 12–4 Components That Can Be Collected from Various Instruments
Fenwal CS3000 × × ×
Fenwal CS3000 + × × ×
Baxter Amicus × × ×
Fenwal Autopheresis C ×
COBE Spectra × × ×
Gambro Trima V4 × × × ×
Gambro Trima Accel × × × ×
Haemonetics LN9000 × × ×
Haemonetics MCS + LN8150 × × ×
Haemonetics PCS-2 ×
Baxter Alyx ×
Fresenius AS104 ×
cPlasma, concurrent plasma; concurrent, more than one component can be collected; cRBC, concurrent 1 unit red blood cells; Gran,
granulocytes; PH, plateletapheresis (single, double, triple); 2-RBC, double unit RBC; V4, software version 4.
From Burgstaler EA. Blood component collection by apheresis. J Clin Apher, May 6, 2005 [Epub ahead of print].
addition, the combined donor plasma and RBC losses must Collection Procedure
be determined and must not exceed the device-specific limits For automated apheresis, the donor is connected by intra-
approved by the FDA.83 venous access to a programmable device that draws, pro-
Granulocyte collection is a more involved process com- cesses, collects, and returns blood to the donor all in a
pared to other automated apheresis donations and neces- single donation session. To separate a component, most
sitates special donor preparation. To obtain sufficient automated devices employ differential centrifugation,
granulocyte dose (>1.0 × 1010), donors are treated with which separates by specific gravity. Instruments that col-
leukocyte mobilizing agents, such as corticosteroids and lect by centrifugation are designed to process blood by
granulocyte colony-stimulating factor (G-CSF), prior to continuous flow centrifugation (CFC), with an uninter-
granulocyte collection (see Chapter 24).84 Although there rupted separation and return circuit, or by intermittent
are limited data on the long-term effects of mobilization in flow centrifugation (IFC), with alternating separation
healthy donors, several studies indicate that in the short- and return cycles. The process of cell separation during
term there is minimal risk, even after repeat donations.85–87 centrifugation isolates leukocytes for “process leukore-
However, donors often experience side effects, such as duction” during the apheresis collection. In addition, 12
headache, bone pain, myalgias, and arthralgias after mobi- depending on the manufacturer, the apheresis dispos-
lization, particularly with G-CSF administration. These able kit may include an in-line filter for leukoreduction. 189
symptoms are generally transient (<24 hours) and well Leukoreduction can also be performed off-line with a
tolerated by donors.88 In addition to precollection leuko- sterile-docked leukoreduction filter. Automated apheresis
cyte mobilization, an erythrocyte sedimenting agent, such instruments also differ in terms of required intravenous
as hydroxyethyl starch (HES), is added during granulocyte access (single- or double-arm) and collection capability
apheresis for more effective granulocyte–red cell separa- (single- or multiple-component collection). See Table
tion. The addition of HES improves granulocyte yield and 12–4 for a summary of current available instruments and
reduces donor RBC loss during the procedure.89 For donors, collection capabilities. Because each apheresis instrument
the disadvantages of HES include the immediate side effects has a unique design and function, a disposable collection
of intravascular volume expansion, such as headache and kit that meets the specifications of an individual instru-
pulmonary edema, and the potential cumulative toxic side ment and type of collection is required.
effects, such as severe pruritus and coagulopathy.90 Because As with whole blood donation, the maximum extracor-
HES persists for weeks to years in circulation,91,92 the addi- poreal blood volume allowed at any time during apheresis is
tive volume of HES for a repeat donor must be routinely 10.5 mL/kg. Most instruments are equipped with a control
determined and monitored for cumulative toxic dose.93 panel for the apheresis operator to input donor and collec-
As part of the eligibility assessment for granulocyte dona- tion data, calculate donor blood volume, and monitor fluids
tion, donors should be evaluated for medical conditions and flow rate during the procedure. To maintain relatively
that may be exacerbated by HES and leukocyte-mobiliz- constant volume during apheresis, normal saline is often
ing drugs. In addition, donors should receive information used both to prime the circuit and to balance fluid shifts. A
about the potential side effects of these drugs as part of the citrate-based anticoagulant, such as ACD-A, is also added to
informed consent process. As with other apheresis collec- prevent clotting and to maintain extracorporeal flow during
tions, the standard allogeneic donor eligibility criteria and apheresis. Citrate, which chelates calcium, can cause symp-
infectious disease testing requirements apply to granulo- tomatic hypocalcemia, which can result in seizures, hypoten-
cyte collection, though the release of collected granulocytes sion, and fatal arrhythmias if left untreated.8 In practice, these
to the transfusing facility often occurs before all test results severe reactions are unusual, because early signs of hypocal-
have been obtained, as these cells lose viability quickly, cemia, such as perioral tingling, are monitored in donors. If
within 8 to 24 hours of collection. In addition, the auto- signs and symptoms of hypocalcemia are noted, temporary
mated platelet collection frequency limits and cumulative cessation of the procedure and reduction of apheresis flow
RBC loss limits apply to granulocyte collection. rate are usually effective.8 If symptoms persist, oral calcium
BLOOD BANKING supplementation may be given, or with more severe symp- For plasma and RBC production, the unit is immediately
toms, an intravenous calcium solution is administered. stored at 1°C to 6°C or transported in a refrigerated system
Quality control criteria for apheresis component collec- that cools the unit toward the range of 1°C to 10°C to main-
tions are established to ensure consistent and adequate prod- tain the optimal storage temperature for RBCs. If fresh frozen
uct dose. For apheresis RBCs, the method employed should plasma (FFP) is to be produced, the plasma is separated from
result in a mean hemoglobin of 60 g in the final product, and the whole blood unit within 8 hours of collection and stored
at least 95% of the units tested should have a hemoglobin of in a freezer at −18°C or colder. A whole blood unit intended
50 g.94 For leukoreduced, apheresis red cells, the procedure for plasma frozen within 24 hours of phlebotomy (FP24) is
should result in a mean hemoglobin of 51 g, with fewer than processed and frozen between 8 and 24 hours after collection.
5 × 106 residual white cells in the final component, and 95% In the case of platelet production, refrigeration leads to
of the units tested should have a hemoglobin of 42.5 g.95 For changes in the platelet membrane that result in poor in vivo
plateletpheresis quality control, the method should result in platelet recovery after transfusion.97–99 Thus, if the whole
a final platelet count of 3 × 1011 in at least 90% of the units blood unit is intended for production of a platelet concen-
tested, and a residual white cell count of fewer than 5 × 106 trate, then the unit is transported in a container capable of
in 95% of the units tested.95 maintaining a temperature as close as possible to the optimal
temperature range for platelet viability, 22°C to 24°C.98 At
the component preparation laboratory, the whole blood unit
COMPONENT PREPARATION is processed to form a platelet concentrate within 8 hours
AND STORAGE of collection, and the remaining RBCs are then immediately
refrigerated.
Components are prepared from the processing of a whole In addition to whole blood transport and storage, collec-
blood donation or via automated apheresis collection meth- tion facilities must take into account the whole blood separa-
ods, as just described. Component preparation from whole tion method to be employed when planning for component
blood entails distinct manufacturing steps for collection, preparation. Whole blood is separated by differential cen-
transportation, and separation, each of which involves unique trifugation based on the specific gravity (relative density)
logistical issues. With automated apheresis, the whole blood of the blood constituents. The relative density of blood con-
collection and component separation steps are integrated stituents ranges from most dense to least dense, as follows:
into a single process, as described in the previous section. RBCs, white blood cells (WBCs), platelets, and plasma. The
After collection and separation, individual components degree of separation and component yield depend primarily
can be further modified by leukoreduction, irradiation, or on the centrifuge rotor size, speed (g- force or rpm), and spin
washing to reduce the likelihood of certain transfusion- time.100 A temperature-controlled centrifuge is required to
related complications. Each component and modified com- maintain the proper temperature range for the specific type
ponent has specific processing, transportation, and storage of component being prepared.
II requirements, established to optimize product quality, safety, The following general methods are used in the United
and therapeutic efficacy. Table 12–5 provides a summary States for whole blood separation. If a whole blood unit is
190 of components, modified components, and manufacturing intended for platelet production, the centrifuge temperature
requirements. is set at 20°C, and a soft spin (low g- force) is used to separate
In the following sections, component preparation, modi- the whole blood into platelet-rich plasma (PRP) and RBCs.
fication, and storage are discussed. The clinical and thera- The PRP is then manually expressed with a spring-loaded
peutic aspects of blood transfusion are presented in greater plasma expresser, through a top port of the primary bag into
detail in the dedicated chapters of this textbook. a satellite bag. The remaining RBCs in the primary bag are
refrigerated at 1°C to 6°C. The PRP is further processed at
20°C by a hard spin (high g- force) to separate the plasma
Whole Blood Processing from the platelets. This method of platelet concentrate pro-
With the advent of component therapy, the use of whole blood duction is often referred to as the PRP method, as opposed to
for transfusion in the United States has become uncommon. the buffy coat method used in European countries.
Whole blood is still reported to be a replacement for mas- If the whole blood is not intended for platelet production,
sive blood loss, such as in trauma and transplant cases.96 In the unit is centrifuged at 4°C with a hard spin to separate the
practice, however, intravenous fluids and blood components whole blood into a platelet-poor plasma (PPP) layer, a buffy
serve as the standard therapy for massive resuscitation, and coat layer containing platelets and WBCs, and a packed RBC
whole blood is rarely used for this purpose. The primary layer. The PPP is then expressed into a satellite bag, leaving
exception is autologous blood, which, for simplicity, is often the majority of the platelets and WBCs with the RBCs in the
transfused as whole blood. Currently, the primary purpose primary bag. The RBCs are then immediately refrigerated at
of whole blood is as source material for blood component 1°C to 6°C.
preparation. Pooled Buffy Coat Method
For collection facilities, the logistics of whole blood pro-
cessing depend on the components that are planned to be In most European countries and Canada, the buffy coat
produced from the whole blood unit. If the whole blood method is applied to the processing of whole blood for plate-
unit is to be processed into plasma and RBCs only, the unit let production. With this method, a specialized primary col-
is transported on ice; however, if the whole blood unit is lection bag with ports on the bottom and top, referred to
intended for platelet production, then the unit is maintained as the bottom-and-top system (BAT), is used. A whole blood
at room temperature until after separation of the platelet-rich unit is first centrifuged to produce a buffy coat (platelets
plasma from the RBCs. and WBCs). With a semiautomated extractor, plasma is then
Table 12–5 Requirements for Storage, Transportation, and Expiration
12
BLOOD MANUFACTURING
192
II
BLOOD BANKING
12
BLOOD MANUFACTURING
BLOOD BANKING expressed out of the top port, and the RBCs are expressed rejuvenate RBCs that have been in refrigerated storage less
out of the bottom port.101 The buffy coat platelet concentrate than 6 days.106
(BC-PC) remaining in the primary bag contains the majority For rejuvenation of a single unit of stored RBCs, a 50-mL
of platelets and WBCs, along with approximately 30 mL of vial of rejuvenating solution is added through a sterile, dis-
plasma and 30 mL of the RBCs. A pool of BC-PCs is prepared posable, Y-type connection and transfer set supplied by the
by suspending 4 to 6 units in a platelet additive solution. The rejuvenating solution manufacturer. Once suspended in the
BC-PC pool is centrifuged, and the platelet-rich supernatant rejuvenating solution, the unit is sealed in a waterproof, dou-
is passed through a leukoreduction filter and stored.102 With ble plastic bag overwrap and incubated in a 37 °C water bath
a pool of 6 BC-PC units, this method will result in more than for 60 minutes. After rejuvenation, the RBCs are prepared for
3 × 1011 platelets in the final product greater than 75% of same-day transfusion (CPD and CPDA-1 units) or processed
the time.103 Compared to the PRP method, advantages of the for frozen storage (CPD, CPDA-1, and CPD/AS-1 units).
BC-PC method include greater plasma volume for plasma For same-day transfusion, the rejuvenated unit is imme-
components, the ability to perform bacterial testing on stored diately washed with an unbuffered saline solution in an
pools (likely increasing the effective and allowed utilization approved cell washer to remove the residual metabolites
of these components to 6 or 7 days), and improved metabolic hypoxanthine, uric acid, inosine, and inorganic phosphates
stability of the platelets.102 The disadvantage of this method prior to transfusion.107 Outdated, rejuvenated, washed RBCs
is the increased RBC loss (10% to 15% loss) that occurs dur- that are stored at 4 °C for up to 3 days have been shown to
ing processing.102 An automated method for BC-PC process- have an approximate in vitro RBC recovery of 95% and a 24-
ing is now available that results in increased efficiency, more hour, post-transfusion survival rate of greater than 75%.107
consistent product volume, and better platelet yield than the Despite these favorable data, the open washing procedure,
manual BC-PC method.104 which has the potential for pathogen contamination of the
product, limits the approved storage of refrigerated, washed,
Red Blood Cells rejuvenated RBCs to 24 hours.
For cryopreservation by freezing, the rejuvenated RBCs
The primary indication for RBC transfusion is the restora- in the primary collection bag are not immediately washed,
tion of oxygen-carrying capacity in such conditions as blood but they are instead further processed for frozen storage. The
loss, anemia, or hemoglobinopathy. RBCs are prepared either RBCs are centrifuged to separate the rejuvenating solution,
by whole blood processing (red blood cells) or by automated which is then expressed into a transfer bag and discarded.
erythrocytapheresis (apheresis red blood cells). Often, RBCs Following separation, the RBCs are immediately prepared
are further modified by leukoreduction to reduce the risk of with a 40% glycerol solution and stored at −80 °C. Prior to
cytomegalovirus (CMV) transmission, alloimmunization, transfusion, these rejuvenated, frozen RBCs are deglycero-
and febrile nonhemolytic transfusion reactions. For special lized and washed. The freeze–thaw–wash cycle results in an
circumstances, RBCs can be irradiated to prevent graft- approximate in vitro RBC recovery of 90% and a 24-hour
II versus-host disease (GVHD) or washed to prevent severe post-transfusion survival rate of greater than 75%. As for
allergic reactions. For routine storage, RBCs are refrigerated RBC function, the rejuvenated, frozen, washed units have
194 between 1°C and 6°C and stored for 21 days (CPD, CP2D), normal or above-normal 2,3-DPG levels and adequate or
35 days (CPDA-1), or 42 days (AS-1, AS-3, AS-5). RBCs that improved oxygen delivery.108,109
have passed the expiration by up to 3 days can be rejuvenated
for immediate transfusion or frozen storage. Frozen and Deglycerolized Red Blood Cells
Cryopreservation by freezing slows or suspends most meta-
Rejuvenated Red Blood Cells
bolic functions and thus limits cellular deterioration dur-
Within the first 2 weeks of refrigerated storage, RBC ATP ing prolonged storage. When processed correctly, studies of
and 2,3-DPG are depleted.43 To increase intracellular stored, frozen RBCs have shown acceptable post-thaw in vitro
ATP and 2,3-DPG to normal or above-normal levels, refrig- viability and function after several decades of storage.110–113
erated RBCs can be rejuvenated through a treatment process The major limitation to cryopreservation for RBC storage is
with the FDA-approved solution Rejuvesol (enCyte Systems, the cryoinjury effect that can occur during the freeze–thaw
Braintree, Mass.). This rejuvenating solution contains the process.114 To minimize the effects of cryoinjury, a perme-
substrates inosine, adenine, phosphate, and pyruvate for able or nonpermeable cryoprotective agent is added to the
RBC ATP and 2,3-DPG biosynthesis. CPD and CPDA-1 cell solution prior to cryopreservation. Glycerol, a permeable
RBCs are approved for rejuvenation up to 3 days after expi- cryoprotectant, is the recommended cryopreservative for
ration. After rejuvenation, CPD and CPDA-1 RBCs can be transfusable RBCs, because it is considered to be nontoxic to
stored in the frozen state for up to 10 years, or they can humans and has proven efficacy.
be transfused within 24 hours of rejuvenation. CPD/AS-1 There are two general methods employed for the prepara-
(Adsol) RBCs are approved for rejuvenation and cryopreser- tion of frozen RBCs, the low glycerol/rapid cooling technique
vation within the 42-day storage period and may be stored and the high glycerol/slow cooling technique. Because glyc-
in the frozen state for up to 3 years. As opposed to CPD and erolized RBCs are hypertonic and hemolyze on contact with
CPDA-1 units, CPD/AS-1 units are not approved for reju- plasma, both freezing methods require a specialized post-
venation after expiration or for same-day transfusion after thaw–wash procedure to remove excess glycerol (deglycero-
rejuvenation. In addition, the rejuvenation of RBCs stored lization) prior to transfusion.
in an additive system other than CPD/AS-1 (i.e., AS-3 or With the low glyercol/rapid cooling method, which is
AS-5) is not currently approved by the FDA. Because reju- primarily used by European blood manufacturers, RBCs are
venation early in storage results in excessive 2,3-DPG lev- suspended in a 15% to 20% weight/volume (wt/vol) solu-
els that can potentially impair the oxygen delivery function tion of glycerol, cooled at a rate of greater than −100 °C/min
of the transfused cells in vivo,105 it is not recommended to by immersion in −197 °C liquid nitrogen, and stored at
BLOOD MANUFACTURING
temperatures below −150 °C in liquid nitrogen or nitrogen
Platelets
vapor.115,116 To prepare for transfusion, the frozen RBCs are
rapidly thawed in a 42 °C to 45 °C water bath and deglycero- Platelet transfusion is indicated for prophylactic or hemostatic
lized. Due to the required extreme storage conditions, RBCs therapy in individuals with acquired or congenital throm-
frozen with low glycerol are not as amenable to transporta- bocytopenia or thrombocytopathy. Platelet components are
tion, and thus, they are typically thawed at the storage facility prepared from either whole-blood donation or apheresis
and shipped to the transfusing facility in the liquid state.117 collection. For whole blood–derived platelets (Platelets), the
In the United States, RBCs that have been in refriger- preparation method must yield at least 5.5 × 1010 platelets
ated storage for 6 days or less or have been rejuvenated per unit in at least 75% of the units tested by the manufac-
are approved for glycerolization and frozen storage. For turing facility. With whole blood–derived platelets, typically
the high glycerol/slow cooling process, RBCs are prepared 4 to 6 units are pooled immediately prior to transfusion,
with a 40% wt/vol glycerol solution, slowly cooled at a rate for a dose comparable to a single apheresis-derived platelet
of approximately −1°C/min, and stored at −80 °C in metal unit. For apheresis-derived platelets (Platelets, Pheresis), the
or cardboard containers in a mechanical freezer. There are collection method must yield at least 3 × 1011 platelets per
two process variations for the 40% glycerol method: the unit in at least 90% of the units tested by the manufactur-
Meryman-Hornblower technique and the Valeri or Naval ing facility. Large-yield plateletpheresis collections can be
Blood Research Laboratory (NBRL) technique.112,118 With divided into two separate apheresis-derived platelets units,
the traditional Meryman-Hornblower method, the excess also known as splits, as long as the platelet count in each unit
glycerol solution is not removed from the RBCs prior to is greater than or equal to 3 × 1011. The platelet storage con-
freezing, so the post-thaw deglycerolization step requires a ditions, such as plasma volume, agitation, and temperature,
large-volume wash to remove the excess glycerol. With the must be optimized to maintain an adequate pH of greater
Valeri or NBRL variation for glycerolization, the glycerolized than or equal to 6.2 in 90% of the units tested at the end of
unit is centrifuged, and the supernatant glycerol is discarded the approved storage period. As with RBCs, platelet compo-
prior to freezing. Thus, after thawing, there is a smaller volume nents are often modified by leukoreduction, and for special
of glycerol to remove and a less-extensive deglycerolization conditions, platelets can be further modified by irradiation
process compared to the Meryman-Hornblower method. or washing. In addition, apheresis-derived platelets can be
For transfusion, the frozen RBC unit is initially thawed cross-matched or HLA-matched for alloimmunized indi-
in a 37°C water bath or dry warmer and then deglycerolized. viduals who exhibit poor platelet increments with random
With the traditional, open method of deglycerolization, the platelet transfusions.
unit is diluted with hypertonic saline, and then washed with Compared to other components, platelets are more sensi-
approximately 2 liters of 1.6% saline solution. After adequate tive to variable storage conditions. Although refrigerated or
washing, the RBCs are suspended in isotonic saline for trans- frozen storage would seem ideal for slowing cellular deterio-
fusion. Frozen RBCs that have been deglycerolized in an open ration and retarding bacterial growth, cold storage of plate-
system are approved for transfusion within 24 hours. The lets results in significantly worse post-transfusion in vivo 12
24-hour expiration of deglycerolized units and the tedious, survival relative to room temperature–stored platelets.97–99
semiautomated deglycerolization process have presented Furthermore, platelets that remain undisturbed during stor- 195
both logistic and work-flow challenges for the routine use age are less efficacious than platelets that are agitated on a
of frozen units.117 Recently, however, a more efficient, auto- horizontal rocker, a process that helps ensure optimal gas
mated, closed glycerolization-deglycerolization system (ACP exchange and maintenance of pH.125 Thus, for optimum
215, Haemonetics, Braintree, Mass.) has been approved by storage, platelets must be gently agitated and maintained at
the FDA.119,120 This system utilizes a disposable processing a temperature range from 20 °C to 24 °C. During transport,
set, sterile connector device, and an in-line 0.22-μ filter to the period of time that platelets remain without gentle agi-
prevent bacterial contamination during processing. After tation should not exceed 24 hours, because interruption of
deglycerolization, the processed RBCs are resuspended in agitation for greater than 1 day results in significant damage
AS-3, instead of isotonic saline, and are approved for refrig- to the component.126
erated storage up to 14 days.121 Frozen RBCs glycerolized by A major challenge with platelet storage is that the room
the Meryman-Hornblower method cannot be deglycerolized temperature environment and nutrient-rich plasma favor
on the ACP 215, because the system in unable to process the bacterial proliferation.127 To reduce the risk of significant
larger volume of supernatant glycerol in these units. bacterial burden, the approved storage period for platelets
Unlike low-glycerol frozen RBCs, RBCs frozen in 40% is limited to 5 days after collection,128 even though plate-
glycerol are better suited for transport, because these units are lets stored for longer periods show acceptable in vivo sur-
more tolerant of temperature fluctuations and can be effec- vival.31,129 Even with 5-day storage, bacterial contamination
tively transported on dry ice.122 One factor that complicates of platelets with subsequent transfusion-related bacteremia
the transportation of frozen units is the increased fragility or sepsis had been a leading cause of transfusion morbidity
of blood bags that occurs at approximately −78 °C, which and mortality.1 To address this issue, in the 23rd edition of
makes the units prone to breakage.123 Traditionally, RBCs the Standard for Blood Banks and Transfusion Services, AABB
have been stored in polyolefin bags inside metal containers introduced a standard for blood establishments to have
(Meryman-Hornblower method). With transport, frozen methods in place that limit and detect bacterial contamina-
units stored in this manner have an incidence of breakage tion of platelets.
of greater than 30%.124 An alternative, improved method for Given the nonspecific nature of the standard, many meth-
storage and transport, developed at the NBRL, uses special- ods are employed for bacterial testing, including Gram stain
ized PVC-based freezer bags protected in corrugated card- or bacterial culture, point-of-care testing by pH/glucose
board holders, which reduces the rate of transportation determination, and visualization of a lack of swirling, which
breakage to less than 3%.124 is a sign of nonviable platelets.1 A major limitation of the
BLOOD BANKING nonculture techniques is the lack of sensitivity (106 to 107 testing of whole blood–derived versus apheresis–derived
CFU/mL) compared to bacterial culture systems (102 to 103 platelets. Prestorage pooling of whole blood–derived plate-
CFU/mL or less).130–133 Consequently, more false negatives lets provides sufficient sample volume for more sensitive
should be expected with nonculture systems. Although more automated bacterial culture methods, thus increasing the
sensitive, the bacterial culture method has the limitations of likelihood of bacterial detection and improving the safety of
a larger volume of sample required for testing and a longer whole blood–derived platelets. Consequently, in the United
wait for test results. In the case of the latter limitation, the States there is renewed interest in prestorage pooling of whole
platelet unit may be distributed and transfused prior to the blood–derived platelets.141–143 Recently, the FDA approved a
reported culture result, so the bacterial status at the time or system for prestorage leukoreduction and pooling of whole
point of transfusion may be unknown. Typically, point-of- blood–derived platelets that contains an integrated bacterial
care methodologies are more commonly employed for whole culture system for automated bacterial detection (Acrodose
blood–derived platelets, because the volume necessary for PL, Pall Corporation, East Hills, NY). With improved tech-
bacterial culture constitutes a more significant loss to this nology, the availability and use of prestorage pooled platelets
smaller-volume platelet product. Apheresis-derived platelets, in the United States is likely to increase.
as a larger-volume product, are more amenable to sampling
for bacterial culture, and thus are more often tested by the Plasma
bacterial culture method, using a sample obtained within
24 to 48 hours of storage. With all methodologies, if a true Plasma components are obtained by separation from a cen-
positive result is obtained, the facility is obligated to quaran- trifuged whole blood unit, from a plasmapheresis dona-
tine the product, identify the contaminating organism, and tion, or as a by-product of plateletpheresis. After collection,
if already distributed, notify the customer and retrieve the plasma can be processed as components for transfusion (FFP
product if not transfused. If the product has been transfused, and PF24) or as source material for further manufacture
the clinical service must be notified for proper assessment (recovered plasma and source plasma) into injectable prod-
and treatment of the transfusion recipient. ucts (i.e., plasma derivatives) or noninjectable products (i.e.,
reagents). In general, plasma from a healthy blood donor
Extended (7-Day) Platelet Storage contains normal levels of proteins, immunoglobulins, and
Room temperature platelets have been stored for up to 2 coagulation factors. However, plasma units can vary some-
weeks with acceptable therapeutic efficacy.102 Advantages of what in content and appearance based on the donor’s diet,
extended storage of platelets are improved logistics, increased medication intake, and physiology. For example, plasma from
inventory flexibility, and decreased cost associated with female donors on estrogen-based contraceptives or hormone
reduced outdate rates.134 Although 7-day storage was previ- replacement therapy can appear green due to an estrogen-
ously approved by the FDA in the 1980s,31 concern about the related elevation in ceruloplasmin levels.144,145
increased risk of transfusion-related sepsis with prolonged As a transfused product, plasma components primar-
II storage at room temperature led the FDA to limit platelet ily serve as a source of coagulation factors and are indi-
storage to 5 days. However, with the increased sensitivity cated in the management of coagulopathies associated
196 of bacterial detection systems for platelet testing, extended with liver disease, warfarin therapy, disseminated intra-
storage of platelets has again become a possibility.133,135 The vascular coagulation, massive transfusion, and congenital
FDA’s position on 7-day platelet storage is currently in evo- factor deficiencies.146,147 In addition, plasma components
lution. In 2003, the FDA approved, with restrictions, the use are used as replacement fluids during therapeutic plasma
of specific plateletpheresis collection and storage systems for exchange.
7-day storage of platelets (COBE Spectra Apheresis System
Plasma Components for Transfusion:
and Trima Automated Blood Collection System, Gambro
FFP and PF24
BCT, Inc., Lakewood, Colo.). At this time, facilities cleared
by the FDA for 7-day storage must comply with a specific FFP and FP24 are indicated for the treatment of disorders
protocol for bacterial testing, and approved facilities must of secondary hemostasis associated with multiple coagula-
participate in a postmarket surveillance study of the applied tion factor deficiencies. For most indications, FFP and FP24
protocol.136 are essentially comparable in therapeutic dose of coagula-
tion factors, except that FFP has significantly higher factor
Prestorage Pooling of Platelets VIII activity.148 Because the activities of coagulation factors
Studies of prestorage pooled, whole blood–derived plate- in stored plasma are dependent on temperature and length
lets show satisfactory in vitro and in vivo properties.137–139 of storage, the processing and storage requirements for FFP
Although a common practice in Europe with buffy coat and FP24 are established to optimize coagulation factor
platelets, prestorage pooling of whole blood–derived platelet levels. In general, during refrigerated whole blood storage,
units has not been approved in the United States due to con- coagulation factor levels remain relatively stable for up to 24
cerns about increased risk of bacterial growth in prestorage hours.148 However, although remaining within the hemo-
platelet pools.140 With the introduction of required bacterial static range, factor VIII activity significantly decreases by
testing of platelets in the United States, it has become appar- 20% to 50% within 24 hours of whole blood storage.149–151
ent that the most sensitive bacterial detection methods are Factor V exhibits variable stability during 24-hour whole
based on automated bacterial culture. Given the relatively blood storage, with some studies showing an insignificant
large sample volume required for bacterial culture and the decrease in activity152 and others reporting a decrease in
increased number of units to test for a single transfusion, activity of greater than 10% from baseline.148,153
such methods are not routinely applied for bacterial test- Due to the significant decrease in factor VIII activity
ing of whole blood–derived platelets. Typically, due to these between 8 and 24 hours of refrigerated storage, FFP collected
constraints, less sensitive methods are employed for bacterial in CPD, CP2D, and CPDA-1 is required to be frozen within
BLOOD MANUFACTURING
8 hours of collection. FFP collected in ACD is required to be to be tested for human T-lymphotropic virus (HTLV)
frozen within 6 hours of phlebotomy. If maintained at −18 °C 1 and 2.160 In addition, in contrast to allogeneic donors,
or colder, FFP is approved for storage for up to 1 year. If donors of source plasma are not required to have a negative
cleared by the FDA, facilities may store FFP for up to 7 years result for antibody to hepatitis B core antigen (HBcAg).160
at −65 °C or colder. For plasma processed and frozen within Furthermore, donors that have a reactive serological test for
8 to 24 hours of collection, the component is designated as syphilis can undergo plasmapheresis for source plasma only
PF24. PF24 is approved for storage at −18 °C for up to 1 year if the donor is being treated for syphilis and the plasma is
after collection. During transport, FFP and PF24 are main- specifically designated for further manufacture into control
tained in the frozen state by packaging with dry ice in an reagents for syphilis testing.161
insulated container. Source plasma that is prepared for injectable products
For transfusion, FFP and PF24 are thawed between 30 °C to must be labeled “Caution: For Manufacturing Use Only.” For
37 °C in a water bath for approximately 30 minutes or rapidly noninjectable products, the source plasma product must be
thawed in an FDA-approved microwave device for approxi- labeled “Caution: For Use in Manufacturing Noninjectable
mately 6 minutes.154,155 Although less expensive than microwave Products Only.”160 If intended for injectable products, source
devices, water baths contribute to a longer thaw time compared plasma is approved for storage at −20 °C or colder for up to
to microwave devices, which can impact the immediate provi- 10 years.
sion of thawed plasma in situations that require urgent need,
particularly in the trauma setting. In addition, although the fro- Cryoprecipitate and Cryoprecipitate-
zen unit is placed in a waterproof plastic overwrap bag prior
to immersion in the water bath, the baths contain nonsterile Reduced Plasma
water, which may contaminate the entry ports of the unit if the Cryoprecipitated antihemophilic factor (cryoprecipitate AHF,
unit is not sufficiently protected. Although microwave devices cryoprecipitate, cryo) is indicated for the treatment of coagu-
are more sterile and result in a more rapid turnaround time, lopathy associated with hypofibrinogenemia, dysfibrinogen-
these devices are relatively expensive. Furthermore, if not prop- emia, factor XIII deficiency, uremic thrombocytopathy, and
erly maintained, microwave devices have the potential to pro- tissue plasminogen activator (tPA) therapy.162 Cryoprecipitate
duce temperature “hot spots” in the unit during thawing, which is also used in surgical settings as a source of fibrinogen,
can damage plasma proteins. Once thawed, FFP and PF24 are admixed with thrombin, for topically applied fibrin glue.
approved for storage at 1 °C to 6 °C for up to 24 hours. If pre- In the past, prior to the development of pasteurized, puri-
pared in a closed system and not transfused within 24 hours, fied plasma derivatives and recombinant factor concentrates,
thawed FFP and PF24 can be relabeled as “Thawed Plasma” and cryoprecipitate was a treatment for hemophilia A and von
stored at 1 °C to 6 °C for up to 5 days. Although prolonged stor- Willebrand disease (vWD). However, given the improved
age of refrigerated plasma is associated with cold activation of safety, purity, and efficacy of factor concentrates, cryopre-
the coagulation system and decreased levels of some coagula- cipitate is no longer a first-line therapy for hemophilia A
tion factors,156,157 this has a clinically insignificant effect, as coag- or vWD when appropriate factor concentrates are available. 12
ulation factor levels remain well within the acceptable range for Cryoprecipitate-reduced plasma (also called cryo-poor superna-
adequate hemostasis.150,158 tant) is employed as a plasma replacement fluid in therapeutic 197
plasma exchange for the treatment of TTP. Cryoprecipitate-
Plasma for Further Manufacture: Recovered
reduced plasma is not an FDA-licensed product.
Plasma and Source Plasma
A unit of cryoprecipitate contains the following concen-
Plasma that is separated from a unit of whole blood up to trated, cold, insoluble plasma proteins: fibrinogen (150 mg
5 days after the whole blood unit expiration is labeled as to 250 mg), factor VIII (80 IU to 120 IU), fibronectin (30 mg
“Liquid Plasma” and stored between 1 °C and 6 °C. Plasma to 60 mg), factor XIII (40 IU to 60 IU), and von Willebrand
that is separated from a whole blood unit greater than 24 factor (80 IU).163 At a minimum, a cryoprecipitate unit is
hours after collection and up to 5 days after the whole blood required to have a fibrinogen level of 150 mg and a factor
unit expiration is labeled “Plasma” and stored at −18 °C VIII activity of 80 IU. To ensure an adequate concentration
or colder. FFP that has exceeded the 12-month expiration of plasma constituents, a unit of cryoprecipitate is prepared
can also be relabeled as “Plasma” and stored at −18 °C for from an FFP unit that contains a volume of at least 200 mL.
an additional 4 years. Liquid plasma is approved for use up The unit of FFP is slowly thawed, in a refrigerator or water
to 5 days after the expiration of the whole blood unit from bath, to a range of 1 °C to 6 °C. At this temperature, a precipi-
which it was separated, and plasma is approved for storage tate of cold, insoluble plasma proteins forms. After centrifu-
up to 5 years from the whole blood collection date. Although gation at 4 °C to concentrate and pellet the insoluble proteins,
liquid plasma and plasma are licensed for transfusion, these the supernatant plasma, known as cryoprecipitate-reduced
products are often relabeled as “Recovered Plasma” and des- plasma, is expressed into a satellite bag, and approximately
ignated for further manufacture. As an unlicensed product, 15 mL is reserved for the cryoprecipitate unit. To maintain
recovered plasma can only be shipped to another manufac- adequate factor VIII levels, cryoprecipitate must be refrozen
turer if the shipping facility has a “short supply agreement” within 1 hour of thawing of the FFP unit. For cryoprecipi-
with the receiving manufacturer.159 tate-reduced plasma, the unit is required to be refrozen with
Source plasma is a licensed product that is collected by plas- 24 hours of FFP thawing. Both cryoprecipitate and cryopre-
mapheresis for the purpose of fractionation into injectable cipitate-reduced plasma are approved for storage at −18 °C
or noninjectable plasma products. Donors of source plasma or colder for up to 1 year from the original collection date.
must meet the requirements for allogeneic plasmapheresis For cryoprecipitate transfusion, approximately 8 units are
donors, including infectious disease testing, plasma pro- thawed in a 30 °C to 37 °C water bath and then pooled into
tein evaluation, and donation interval limitations. The only a single transfer bag. The pooled cryoprecipitate product is
exception is that donors of source plasma are not required required to be stored at room temperature (20 °C to 24 °C)
BLOOD BANKING and is approved for administration within 4 hours of pool- apheresis-derived and whole blood–derived RBCs, the RBC
ing. A single unit of thawed cryoprecipitate, stored at room loss from filtration must not exceed 15% of the original
temperature, is approved for use within 6 hours of thawing. unit. The most common quality control method employed
Thawed cryoprecipitate-reduced plasma is approved for for white blood cell enumeration is light microscopy using
storage at 1 °C to 6 °C for up to 5 days. a Nageotte chamber.169 Although this manual method is
relatively inexpensive, it is labor-intensive and results in a
high degree of interindividual variation. Automated flow
Granulocytes
cytometry and microfluorometry hemocytometers provide
The general indication for granulocyte transfusion is treat- a more accurate and precise method for WBC enumera-
ment of a life-threatening fungal and bacterial infection in tion.170 Although more expensive, automated devices offer
an individual with severe but reversible neutropenia, typi- greater convenience and are being increasingly employed
cally resulting from myeloablative chemotherapy that has by blood centers for leukofiltration quality control.
not adequately responded to antibiotic therapy after 48 to
Washing
72 hours (see Chapter 24).84 Granulocytes are collected by
leukapheresis from an eligible donor who has received a leu- Washed cellular products are indicated for individuals
kocyte-mobilizing agent, such as corticosteroids and G-CSF, who have a history of recurrent, severe allergic transfusion
prior to the procedure. Although granulocytes for transfu- reactions.171 In addition, for neonatal transfusion, washed
sion are not an FDA-licensed product, AABB requires that maternal platelets are indicated for neonatal alloimmune
a minimum of 1.0 × 1010 granulocytes per unit are present thrombocytopenia (NAIT),172 and washed RBCs, previously
in at least 75% of the products tested. In addition, because stored in additive solution, have been suggested as an option
a granulocyte component contains greater than 2 mL of red for neonatal massive transfusion.173 Cell washing can be per-
blood cells, the donor and recipient ABO type and cross- formed by manual or automated methods. Automated cell
match must be compatible. washing devices contain a specialized centrifugation device
Starting immediately after collection, granulocyte func- with fluid input and output channels that remove super-
tion deteriorates rapidly; thus, a granulocyte component natant waste from the cells and add normal saline (0.9%)
should be transfused as soon as possible after collection, solution for cell resuspension. Manufacturers of cell-wash-
usually within 8 to 24 hours.164 Because granulocyte com- ing systems provide a customized, disposable processing kit
ponents contain viable T lymphocytes and are transfused to that meets the specifications of the cell-washing device. With
immunocompromised individuals, all granulocyte products a 1-L to 2-L saline wash procedure, greater than 99% of the
should be irradiated to prevent GVHD. When transported supernatant plasma is removed from the component, with
or temporarily stored, the component should be kept at approximately 90% recovery for washed platelets174 and 80%
room temperature (20 °C to 24 °C) without agitation.165 In to 85% recovery for washed RBCs.173,175 This modification
addition, for obvious reasons, a granulocyte product should process is discussed in additional detail in Chapter 29.
II not be leukoreduced or transfused through a bedside leu- Current cell-washing devices operate as an open system,
koreduction filter. Although not an FDA-licensed product, so the expiration of the washed component is affected. For
198 granulocytes are considered a “product under development” refrigerated, washed red blood cells, the expiration is 24
under 21 CFR 601.21,166 and thus are exempt from require- hours, and for room temperature, washed platelets, the expi-
ments for licensing for interstate transport and may be ration is to 4 hours. The shortened expiration time presents
shipped from one state to another as long as the product is a logistic challenge for washed platelet transfusion, because
not “introduced into interstate commerce.” an additional platelet “rest” step of approximately 1 hour is
typically performed prior to release172; therefore, the actual
allowable time to transfusion is between 2 and 3 hours.
Modification Processes Finally, washed components are not FDA licensed, and thus
they are not approved for interstate transport.
Leukoreduction
Irradiation
Leukoreduction involves the removal of WBCs from cellular
components to reduce the risk of HLA alloimmunization, Irradiation of blood components is performed to pre-
CMV transmission, and febrile nonhemolytic transfusion vent transfusion-associated graft-versus-host disease (TA-
reactions.167 Leukoreduction can be performed by filtration GVHD), a fatal transfusion complication. TA-GVHD occurs
prior to component storage (prestorage leukoreduction) or when donor T lymphocytes, present in cellular blood prod-
during the transfusion (bedside filtration). For apheresis- ucts, engraft, proliferate, and destroy the target organs in
derived platelets, leukoreduction is often performed by cell a susceptible transfusion recipient. Irradiation of cellular
separation during the apheresis collection. For whole blood, products, with either gamma rays or x-rays, prevents TA-
whole blood–derived platelets, and RBCs, leukoreduction GVHD by damaging donor T-lymphocyte DNA, with resul-
is performed using third-generation leukoreduction filters, tant inhibition of donor T-lymphocyte proliferation within
which are commercially available as an integral part of a the recipient.176
whole blood collection kit or as a separate, sterilely docked Blood component irradiation is considered a manufac-
device. This modification process is discussed in detail in turing process, so a facility that performs irradiation must
Chapter 26. be registered with the FDA. Free-standing irradiators used
Quality control standards for leukocyte-reduced com- in blood manufacturing contain either a radioactive cesium
ponents require that 95% of the units sampled contain (137Cs) or cobalt (60Co) source. Because the intensity of the
fewer than 5 × 106 WBCs per unit for apheresis-derived radioactive source decays with time, the dose of radiation to
platelets and RBCs and fewer than 8.3 × 105 WBCs per unit which the blood product is exposed, measured in centiGray
for whole blood–derived platelets.168 In addition, for both (1 cGy = 1 rad), is dependent on the source half-life. Thus,
BLOOD MANUFACTURING
to obtain a predicted, fixed radiation dose, the radioisotope and future introduction of radio frequency identification
decay is calculated to determine radiation exposure time. (RFID) technology.184 Both the FDA and AABB endorse
In addition, the irradiator must be recalibrated periodically, the ICCBBA’s “United States Industry Consensus Standard
semiannually for 60Co irradiators and annually for 137Cs irra- for the Uniform Labeling of Blood and Blood Components
diators, to ensure adequate delivered dose.177 Although a dose Using ISBT 128,” and they have committed to a time line for
of 1500 cGy significantly reduces T-lymphocyte proliferation, a ISBT 128 implementation by the U.S. blood industry.185,186
dose of approximately 2500 cGy has been shown to completely Effective November 1, 2006, the 24th edition of the AABB
inhibit T-lymphocyte proliferation, based on studies using Standards for Blood Banks and Transfusion Services will
the sensitive limiting dilution assay (LDA).178,179 Accordingly, require accredited facilities to have a written plan for the
quality control standards require that the minimum radiation transition to and implementation of ISBT 128; and effec-
dose delivered to the component be 2500 cGy, with a mini- tive May 1, 2008, the 25th edition will require accredited
mum dose of 1500 cGy at any point in the blood product. This facilities to have fully implemented ISBT 128.186 In addi-
modification process is discussed in detail in Chapter 28. tion to blood and blood products, ISBT 128 will be applied
Validation of the irradiation process is performed by to labeling for cellular therapy products and tissues, with
a dose-mapping procedure, which measures the dose of an effective implementation date of September 1, 2008.186
radiation delivered to specific points over the entire area of In light of these developments, the following section will
a simulated blood product. For dose mapping, a phantom, provide a general overview of the ISBT 128 labeling system,
composed of water or plastic and equipped with strategically with reference to ABC Codabar for comparison purposes
placed dosimeters, is positioned inside the irradiation cham- only. The reader is encouraged to visit the ICCBBA website
ber. On irradiation, each dosimeter absorbs a specific dose (http://www.isbt128.org) and the AABB website (http://
of radiation, which is then measured and mapped. Available www.aabb.org) for additional resources on ISBT 128 imple-
dose-mapping systems for blood irradiators employ one of mentation.
the following: radiochromic film dosimeters, thermolumi- In general, the attributes of ISBT 128 include an enhanced
nescent dosimeters (TLD chips), or metal oxide semicon- unique donor identification system; an improved data struc-
ductor field effect transistors (MOSFET). Along with dose ture for critical information (e.g., blood group, expiration
validation, a process for irradiation confirmation is necessary. date); an internationally recognized product description
This process is performed by affixing a label that contains database; an eye-readable, machine-readable uniform label-
radiosensitive film to the blood product prior to irradiation. ing format; and a more accurate, flexible bar coding system.187
On irradiation, the film changes from transparent to opaque, In addition, unlike ABC Codabar, ISBT 128 supports con-
providing visual evidence that the product was irradiated. catenation, which is the ability to integrate data from paired
With irradiation, the expiration date and time of granu- bar codes into a single data message.184 Furthermore, ISBT
locytes and platelets are unaffected. However, the shelf life of 128 contains check characters to further reduce scanning
RBCs is shortened due to irradiation-induced acceleration of errors.184 A sample ISBT 128 label is shown in Figure 12–1,
adverse storage effects, such as decreased in vivo survival180,181 which illustrates the standardized four-quadrant layout and 12
and increased potassium leakage.182 For irradiated RBCs, the the key label features of the ISBT 128 system.
expiration date is 28 days from the day of irradiation or the Located in the upper left-hand corner is the unique donor 199
original expiration date, whichever comes first. identification code, which is a 13-character data set composed
LABELING
SHIPPING
REFERENCES
Shipping and receiving diagnostic specimens and blood prod- 1. Hillyer CD, Josephson CD, Blajchman MA, et al. Bacterial contamina-
ucts is an essential part of the blood manufacturing process. tion of blood components: risks, strategies, and regulation: Joint ASH
Regulations for the interstate transport of biological mate- and AABB educational session in transfusion medicine. Hematology
rials are established by the Department of Transportation (Am Soc Hematol Educ Program) 2003:575–589.
2. Wagner SJ. Transfusion-transmitted bacterial infection: risks, sources
(DOT) for ground transport, the International Air Transport and interventions. Vox Sang 2004;86:157–163.
Association (IATA) for air transport, and the United States 3. Code of Federal Regulations, 21 CFR 604.3. U.S. Government Printing
Postal Service (USPS) for mail transport. Diagnostic speci- Office, Washington, D.C., 2005 (revised annually).
mens and components that have a high likelihood of contain- 4. U.S. Food and Drug Administration (CBER). Options for Arm Prepa-
ing an infectious agent must be packaged in compliance with ration. Available at http://www.fda.gov/cber/blood/armpreprev.htm.
Accessed October 30, 2005.
42 CFR 72 and shipped in accordance to the requirements 5. AABB Association Bulletin #03–07 and #03–10, August 2003.
set forth by the IATA, USPS, and the Centers for Disease 6. Goldman M, Roy G, Frechette N, et al. Evaluation of donor skin disin-
Control and Prevention (CDC).189 In general, the shipping fection methods. Transfusion 1997;37:309–312.
BLOOD MANUFACTURING
7. Newman BH. Blood donor complications after whole-blood donation. 38. Rubin RJ, Ness PM. What price progress? An update on vinyl plastic
Curr Opin Hematol 2004;11:339–345. bags. Transfusion 1989;29:358–361.
8. Winters JL. Complications of donor apheresis. J Clin Apheresis 2005 39. Jacobson MS, Kevy SV, Grand RJ. Effects of a plasticizer leached
(Epub ahead of print). from polyvinyl chloride on the subhuman primate: a consequence of
9. McLeod BC, Price TH, Owen H, et al. Frequency of immediate chronic transfusion therapy. J Lab Clin Med 1977;89:1066–1079.
adverse effects associated with apheresis donation. Transfusion 1998; 40. Hillman LS, Goodwin SL, Sherman WR. Identification and measure-
38:938–943. ment of plasticizer in neonatal tissues after umbilical catheters and
10. Newman BH, Waxman DA. Blood donation-related neurologic needle blood products. N Engl J Med 1975;292:381–386.
injury: evaluation of 2 years’ worth of data from a large blood center. 41. Kavlock R, Boekelheide K, Chapin R, et al. NTP Center for the Evalua-
Transfusion 1996;36:213–215. tion of Risks to Human Reproduction: phthalates expert panel report
11. Newman B. Venipuncture nerve injuries after whole-blood donation. on the reproductive and developmental toxicity of di(2-ethylhexyl)
Transfusion 2001;41:571–572. phthalate. Reprod Toxicol 2002;16:529–653.
12. Horowitz SH. Venipuncture-induced causalgia: anatomic relations of 42. Hess JR, Greenwalt TG. Storage of red blood cells: new approaches.
upper extremity superficial veins and nerves, and clinical consider- Transfus Med Rev 2002;16:283–295.
ations. Transfusion 2000;40:1036–1040. 43. Scott KL, Lecak J, Acker JP. Biopreservation of red blood cells: past,
13. Horowitz SH. Venipuncture-induced neuropathic pain: the clinical present, and future. Transfus Med Rev 2005;19:127–142.
syndrome, with comparisons to experimental nerve injury models. 44. Valeri CR, Collins FB. Physiologic effects of 2,3-DPG-depleted red
Pain 2001;94:225–229. cells with high affinity for oxygen. J Appl Physiol 1971;31:823–827.
14. Despotis GJ, Goodnough LT, Dynis M, et al. Adverse events in platelet 45. Beutler E, Wood L. The in vivo regeneration of red cell 2,3 diphospho-
apheresis donors: a multivariate analysis in a hospital-based program. glyceric acid (DPG) after transfusion of stored blood. J Lab Clin Med
Vox Sang 1999;77:24–32. 1969;74:300–304.
15. Popovsky MA, Whitaker B, Arnold NL. Severe outcomes of alloge- 46. Valeri CR, Hirsch NM. Restoration in vivo of erythrocyte adenosine
neic and autologous blood donation: frequency and characterization. triphosphate, 2,3-diphosphoglycerate, potassium ion, and sodium
Transfusion 1995;35:734–737. ion concentrations following the transfusion of acid-citrate-dextrose-
16. Standard 5.3.2: Silva MA (ed). Standards for Blood Bank and Transfu- stored human red blood cells. J Lab Clin Med 1969;73:722–733.
sion Services, 23rd Edition. Bethesda, MD: AABB, 2004. 47. Högman CF. Preparation and preservation of red cells. Vox Sang
17. Standard 5.3.3: Silva MA (ed). Standards for Blood Bank and Transfu- 1998;74(Suppl 2):177–187.
sion Services, 23rd Edition. Bethesda, MD: AABB, 2004. 48. Kendrick D. Chapter IX. Preservative Solutions Blood. Program in
18. Standard 5.8: Silva MA (ed). Standards for Blood Bank and Transfu- World War II. Washington, D.C. Office of the Surgeon General, Depart-
sion Services, 23rd Edition. Bethesda, MD: AABB, 2004. ment of the Army, U.S. Government Printing Office, 1964.
19. Szewzyk U, Szewzyk R, Stenstrom TA. Growth and survival of Serratia 49. Moroff G, Sohmer PR, Button LN. Proposed standardization of meth-
marcescens under aerobic and anaerobic conditions in the presence of ods for determining the 24-hour survival of stored red cells. Transfu-
materials from blood bags. J Clin Microbiol 1993;31:1826–1830. sion 1984;24:109–114.
20. Standard 5.7.2: Silva MA (ed). Standards for Blood Bank and Transfu- 50. Högman CF, Eriksson L, Gong J, et al. Half-strength citrate CPD com-
sion Services, 23rd Edition. Bethesda, MD: AABB, 2004. bined with a new additive solution for improved storage of red blood
21. FDA Guidance for Industry. Use of sterile connecting devices in blood cells suitable for clinical use. Vox Sang 1993;65:271–278.
bank practices, November 2000. Available at http://www.fda.gov/cber/ 51. Hess JR, Rugg N, Knapp AD, et al. Successful storage of RBCs for 10
gdlns/bbconn.htm. Accessed October 30, 2005. weeks in a new additive solution. Transfusion 2000;40:1012–1016.
22. Blajchman MA. Bacterial contamination of cellular blood compo- 52. Hess JR, Rugg N, Knapp AD, et al. Successful storage of RBCs for 9
nents: risks, sources and control. Vox Sang 2004;87(Suppl 1):98–103. weeks in a new additive solution. Transfusion 2000;40:1007–1011.
23. Wagner SJ, Robinette D, Friedman LI, et al. Diversion of initial blood 53. Meryman HT, Hornblower ML, Syring RL. Prolonged storage of red
flow to prevent whole-blood contamination by skin surface bacteria: cells at 4 degrees C. Transfusion 1986;26:500–505.
12
an in vitro model. Transfusion 2000;40:335–338. 54. Hess JR, Hill HR, Oliver CK, et al. Twelve-week RBC storage. Transfu-
24. de Korte D, Marcelis JH, Verhoeven AJ, et al. Diversion of first blood sion 2003;43:867–872. 201
volume results in a reduction of bacterial contamination for whole- 55. Ho J, Sibbald WJ, Chin-Yee IH. Effects of storage on efficacy of red
blood collections. Vox Sang 2002;83:13–16. cell transfusion: when is it not safe? Crit Care Med 2003;31(12 Suppl):
25. Carmen R. The selection of plastic materials for blood bags. Transfus S687–S697.
Med Rev 1993;7:1–10. 56. Leal-Noval SR, Jara-Lopez I, Garcia-Garmendia JL, et al. Influence of
26. Murphy S, Sayar SN, Gardner FH. Storage of platelet concentrates at erythrocyte concentrate storage time on postsurgical morbidity in car-
22°C. Blood 1970;35:549–557. diac surgery patients. Anesthesiology 2003;98:815–822.
27. Murphy S, Gardner FH. Platelet storage at 22°C: role of gas trans- 57. Zallen G, Offner PJ, Moore EE, et al. Age of transfused blood is an
port across plastic containers in maintenance of viability. Blood independent risk factor for postinjury multiple organ failure. Am J
1975;46:209–218. Surg 1999;178:570–572.
28. Murphy S, Kahn RA, Holme S, et al. Improved storage of platelets for 58. Marik PE, Sibbald WJ. Effect of stored-blood transfusion on oxygen
transfusion in a new container. Blood 1982;60:194–200. delivery in patients with sepsis. JAMA 1993;269:3024–3029.
29. Holme S, Heaton A, Momoda G. Evaluation of a new, more oxygen- 59. Purdy FR, Tweeddale MG, Merrick PM. Association of mortality
permeable, polyvinylchloride container. Transfusion 1989;29:159–164. with age of blood transfused in septic ICU patients. Can J Anaesth
30. Snyder EL, Pope C, Ferri PM, et al. The effect of mode of agitation and 1997;44:1256–1261.
type of plastic bag on storage characteristics and in vivo kinetics of 60. Hogman CF, Knutson F, Loof H, et al. Improved maintenance of 2,3
platelet concentrates. Transfusion 1986;26:125–130. DPG and ATP in RBCs stored in a modified additive solution. Transfu-
31. Hogge DE, Thompson BW, Schiffer CA. Platelet storage for 7 days in sion 2002;42:824–829.
second-generation blood bags. Transfusion 1986;26:131–135. 61. Kurup PA, Arun P, Gayathri NS, et al. Modified formulation of CPDA
32. Sasakawa S, Tokunaga E. Physical and chemical changes of ACD-pre- for storage of whole blood, and of SAGM for storage of red blood cells,
served blood: a comparison of blood in glass bottles and plastic bags. to maintain the concentration of 2,3-diphosphoglycerate. Vox Sang
Vox Sang 1976;31:199–210. 2003;85:253–261.
33. Horowitz B, Stryker MH, Waldman AA, et al. Stabilization of red 62. Hogman CF, Eriksson L, Wallvik J, et al. Clinical and laboratory expe-
blood cells by the plasticizer, diethylhexylphthalate. Vox Sang 1985; rience with erythrocyte and platelet preparations from a 0.5CPD
48:150–155. Erythro-Sol opti system. Vox Sang 1997;73:212–219.
34. Estep TN, Pedersen RA, Miller TJ, et al. Characterization of erythro- 63. Ebert R, Stead EA, Gibson JG. Response of normal subjects to acute
cyte quality during the refrigerated storage of whole blood containing blood loss. Arch Intern Med. 1941;68:578.
di-(2-ethylhexyl) phthalate. Blood 1984;64:1270–1276. 64. Howarth S, Sharpey-Schafer EP. Low blood-pressure phases following
35. AuBuchon JP, Estep TN, Davey RJ. The effect of the plasticizer di-2- haemorrhage. Lancet 1947;1:18.
ethylhexyl phthalate on the survival of stored RBCs. Blood 1988;71: 65. Standard 5.6.6.1.3: Silva MA (ed): Standards for Blood Bank and
448–452. Transfusion Services, 23rd Edition. Bethesda, MD: AABB, 2004.
36. Jaeger RJ, Rubin RJ. Contamination of blood stored in plastic packs. 66. Trudnowski RJ, Rico RC. Specific gravity of blood and plasma at 4 and
Lancet 1970;2:151. 37°C. Clin Chem 1974;20:615–616.
37. Rock G, Tocchi M, Ganz PR, et al. Incorporation of plasticizer into red 67. Simon TL, Hunt WC, Garry PJ. Iron supplementation for menstruat-
cells during storage. Transfusion 1984;24:493–498. ing female blood donors. Transfusion 1984;24:469–472.
BLOOD BANKING 68. Code of Federal Regulations, 21 CFR 640.3(d): U.S. Government 99. Slichter SJ, Harker LA. Preparation and storage of platelet concen-
Printing Office, Washington, D.C. (revised annually). trates. I. Factors influencing the harvest of viable platelets from whole
69. Jeffrey G, Adams PC: Blood from patients with hereditary hemochro- blood. Br J Haematol 1976;34:395–402.
matosis—a wasted resource. Transfusion 1999;39:549–550. 100. Kahn RA, Cossette I, Friedman LI. Optimum centrifugation condi-
70. FDA Guidance for Industry: Variances for collections from individu- tions for the preparation of platelet and plasma products. Transfusion
als with hereditary hemochromatosis, August 2001. Available at http:// 1976;16:162–165.
www.fda.gov/cber/gdlns/hemchrom.pdf. Accessed October 30, 2005. 101. Hogman CF, Eriksson L, Hedlund K, et al. The bottom and top system:
71. Huh YO, Lichtiger B, Giacco GG, et al. Effect of donation time on a new technique for blood component preparation and storage. Vox
platelet concentrates and fresh-frozen plasma. An in vitro study. Vox Sang 1988;55:211–217.
Sang 1989;56:21–24. 102. Murphy S. Platelets from pooled buffy coats: an update. Transfusion
72. Reiss RF, Katz AJ. Platelets and factor VIII as functions of blood collec- 2005;45:634–639.
tion time. Transfusion 1976;16:229–231. 103. Murphy S, Heaton WA, Rebulla P. Platelet production in the Old
73. Brecher M, ed. Technical Manual, 14th ed. Bethesda, MD: AABB, 2002. World—and the New. Transfusion 1996;36:751–754.
74. U.S. Food and Drug Administration (CBER). Revised Guidelines for 104. Janetzko K, Kluter H, van Waeg G, et al. Fully automated process-
the Collection of Platelets, Pheresis. Memorandum, October 1988. ing of buffy-coat-derived pooled platelet concentrates. Transfusion
75. U.S. Food and Drug Administration (CBER). Revisions of FDA Mem- 2004;44:1052–1058.
orandum of August 27, 1982: Requirements for Infrequent Plasma- 105. Valeri CR. Use of rejuvenation solutions in blood preservation. Crit
pheresis Donors. Memorandum, March 1995. Rev Clin Lab Sci 1982;17:299–374.
76. FDA Guidance for Industry. Recommendations for collecting red 106. Package Insert. Rejuvesol. Braintree, Mass. enCyte Systems, 1997.
blood cells by automated apheresis methods, January 2001. Available 107. Valeri CR, Gray AD, Cassidy GP, et al. The 24-hour posttransfusion
at http://www.fda.gov/cber/gdlns/rbcautoph2.pdf. Accessed October survival, oxygen transport function, and residual hemolysis of human
30, 2005. outdated-rejuvenated red cell concentrates after washing and storage
77. Standard 5.5.3.4: Silva MA (ed). Standards for Blood Bank and Trans- at 4°C for 24 to 72 hours. Transfusion 1984;24:323–326.
fusion Services, 23rd Edition. Bethesda, MD: AABB, 2004. 108. Szymanski IO, Teno RA, Lockwood WB, et al. Effect of rejuvenation and
78. Standard 5.5.1: Silva MA (ed). Standards for Blood Bank and Transfu- frozen storage on 42-day-old AS-1 RBCs. Transfusion 2001;41:550–555.
sion Services, 23rd Edition. Bethesda, MD: AABB, 2004. 109. Valeri CR, Zaroulis CG, Vecchione JJ, et al. Therapeutic effectiveness
79. Code of Federal Regulations, 21 CFR 640.65, U.S. Government Print- and safety of outdated human red blood cells rejuvenated to restore
ing Office, Washington, D.C., 2005 (revised annually). oxygen transport function to normal, frozen for 3 to 4 years at –80°C,
80. U.S. Food and Drug Administration (CBER). Volume Limits for Auto- washed, and stored at 4°C for 24 hours prior to rapid infusion. Trans-
mated Collection of Source Plasma. Memorandum, November 1992. fusion 1980;20:159–170.
81. Standard 5.5.3.5: Silva MA (ed). Standards for Blood Bank and Trans- 110. Valeri CR, Ragno G, Pivacek LE, et al. An experiment with glycerol-
fusion Services, 23rd Edition. Bethesda, MD: AABB, 2004. frozen red blood cells stored at –80°C for up to 37 years. Vox Sang
82. FDA Draft Guidance for Industry. Collection of platelets by automated 2000;79:168–174.
methods. September 2005. (http://www.fda.gov/cber/gdlns/platelet- 111. Umlas J, Jacobson M, Kevy SV: Suitable survival and half-life of red cells
auto.pdf). Accessed November 18, 2005. after frozen storage in excess of 10 years. Transfusion 1991;31:648–649.
83. Standard 5.5.4: Silva MA (ed). Standards for Blood Bank and Transfu- 112. Valeri CR, Pivacek LE, Gray AD, et al. The safety and therapeutic effec-
sion Services, 23rd Edition. Bethesda, MD: AABB, 2004. tiveness of human red cells stored at -80°C for as long as 21 years.
84. Briones MA, Josephson CD, Hillyer CD. Granulocyte transfusion: Transfusion 1989;29:429–437.
Revisited. Curr Hematol Rep 2003;2:522–527. 113. Lecak J, Scott K, Young C, et al. Evaluation of red blood cells stored at
85. Stroncek DF, Clay ME, Herr G, et al. Blood counts in healthy donors -80°C in excess of 10 years. Transfusion 2004;44:1306–1313.
1 year after the collection of granulocyte-colony-stimulating factor- 114. Mazur P. Freezing of living cells: mechanisms and implications. Am J
II mobilized progenitor cells and the results of a second mobilization and Physiol 1984;247(3 Pt 1):C125–C142.
collection. Transfusion 1997;3:304–308. 115. Rowe AW, Eyster E, Kellner A. Liquid nitrogen preservation of red
202 86. MacHida U, Tojo A, Takahashi S, et al. The effect of granulocyte col- blood cells for transfusion: a low glycerol-rapid freeze procedure.
ony-stimulating factor administration in healthy donors before bone Cryobiology 1968;5:119–128.
marrow harvesting. Br J Haematol 2000;108:747–753. 116. Valeri CR. Simplification of the methods for adding and removing
87. Liles WC, Huang JE, Llewellyn C, et al. A comparative trial of granu- glycerol during freeze-preservation of human red blood cells with the
locyte-colony-stimulating factor and dexamethasone, separately and high or low glycerol methods: biochemical modification prior to freez-
in combination, for the mobilization of neutrophils in the peripheral ing. Transfusion 1975;15:195–218.
blood of normal volunteers. Transfusion 1997;37:182–187. 117. Hess JR. Red cell freezing and its impact on the supply chain. Transfus
88. Einsele H, Northoff H, Neumeister B. Granulocyte transfusion. Vox Med 2004;14:1–8.
Sang 2004;87(Suppl 2):205–208. 118. Meryman HT, Hornblower M. A method for freezing and wash-
89. Lee JH, Leitman SF, Klein HG. A controlled comparison of the efficacy ing red blood cells using a high glycerol concentration. Transfusion
of hetastarch and pentastarch in granulocyte collections by centrifugal 1972;12:145–156.
leukapheresis. Blood 1995;86:4662–4666. 119. Valeri CR, Ragno G, Pivacek L, et al. In vivo survival of apheresis
90. Strauss RG. Review of the effects of hydroxyethyl starch on the blood RBCs, frozen with 40-percent (wt/vol) glycerol, deglycerolized in the
coagulation system. Transfusion 1981;21:299–302. ACP 215, and stored at 4°C in AS-3 for up to 21 days. Transfusion
91. Trivedi SM, Humphrey RL, Braine HG, et al. Hydroxyethyl starch 2001;41:928–932.
serum levels in leukapheresis donors measured by modified periodic 120. Valeri CR, Ragno G, Pivacek LE, et al. A multicenter study of in vitro
acid-Schiff staining technique. Transfusion 1984;24:260–263. and in vivo values in human RBCs frozen with 40-percent (wt/vol)
92. Maguire LC, Strauss RG, Koepke JA, et al. The elimination of hydroxy- glycerol and stored after deglycerolization for 15 days at 4°C in
ethyl starch from the blood of donors experiencing single or multiple AS-3: assessment of RBC processing in the ACP 215. Transfusion
intermittent-flow centrifugation leukapheresis. Transfusion 1981;21: 2001;41:933–939.
347–353. 121. Valeri CR, Srey R, Tilahun D, et al. The in vitro quality of red blood
93. Standard 5.6.6.1.2.1: Silva MA (ed). Standards for Blood Bank and cells frozen with 40 percent (wt/vol) glycerol at -80°C for 14 years,
Transfusion Services, 23rd Edition. Bethesda, MD: AABB, 2004. deglycerolized with the Haemonetics ACP 215, and stored at 4°C in
94. Standard 5.7.5.8: Silva MA (ed). Standards for Blood Bank and Trans- additive solution-1 or additive solution-3 for up to 3 weeks. Transfu-
fusion Services, 23rd Edition. Bethesda, MD: AABB, 2004. sion 2004;44:990–995.
95. Standard 5.7.5.8.1: Silva MA (ed). Standards for Blood Bank and 122. Valeri CR, Pivacek LE, Cassidy GP, et al. In vitro and in vivo measure-
Transfusion Services, 23rd Edition. Bethesda, MD: AABB, 2004. ments of human RBCs frozen with glycerol and subjected to various
96. Laine E, Steadman R, Calhoun L, et al. Comparison of RBCs and storage temperatures before deglycerolization and storage at 4°C for 3
FFP with whole blood during liver transplant surgery. Transfusion days. Transfusion 2001;41:401–405.
2003;43:322–327. 123. Hmel PJ, Kennedy A, Quiles JG, et al. Physical and thermal properties
97. Murphy S, Gardner FH. Effect of storage temperature on maintenance of blood storage bags: implications for shipping frozen components
of platelet viability—deleterious effect of refrigerated storage. N Engl J on dry ice. Transfusion 2002;42:836–846.
Med 1969;280:1094–1098. 124. Valeri CR, Lane JP, Srey R, et al. Incidence of breakage of human RBCs
98. Murphy S, Gardner FH. Platelet storage at 22°C; metabolic, morpho- frozen with 40-percent wt/vol glycerol using two different methods for
logic, and functional studies. J Clin Invest 1971;50:370–377. storage at -80°C. Transfusion 2003;43:411–414.
BLOOD MANUFACTURING
125. Bannai M, Mazda T, Sasakawa S. The effects of pH and agitation on 152. O’Neill EM, Rowley J, Hansson-Wicher M, et al. Effect of 24-
platelet preservation. Transfusion 1985;25:57–59. hour whole-blood storage on plasma clotting factors. Transfusion
126. Hunter S, Nixon J, Murphy S. The effect of the interruption of agi- 1999;39:488–491.
tation on platelet quality during storage for transfusion. Transfusion 153. Sohmer PR, Bolin RB, Scott RL, et al. Effect of delayed refrigeration
2001;41:809–814. on plasma factors in whole blood collected in CPDA-2. Transfusion
127. Brecher ME, Holland PV, Pineda AA, et al. Growth of bacteria in inoc- 1982;22:488–490.
ulated platelets: implications for bacteria detection and the extension 154. Churchill WH, Schmidt B, Lindsey J, et al. Thawing fresh frozen plasma
of platelet storage. Transfusion 2000;40:1308–1312. in a microwave oven. A comparison with thawing in a 37°C waterbath.
128. Goldman M, Blajchman MA. Blood product-associated bacterial sep- Am J Clin Pathol 1992;97:227–232.
sis. Transfus Med Rev 1991;5:73–83. 155. Rock G, Tackaberry ES, Dunn JG, et al. Rapid controlled thawing
129. Dumont LJ, AuBuchon JP, Whitley P, et al. Seven-day storage of single- of fresh-frozen plasma in a modified microwave oven. Transfusion
donor platelets: recovery and survival in an autologous transfusion 1984;24:60–65.
study. Transfusion 2002;42:847–854. 156. Suontaka AM, Silveira A, Soderstrom T, et al. Occurrence of cold
130. Wagner SJ, Robinette D. Evaluation of swirling, pH, and glucose tests activation of transfusion plasma during storage at +4°C. Vox Sang
for the detection of bacterial contamination in platelet concentrates. 2005;88:172–180.
Transfusion 1996;36:989–993. 157. Blomback M, Chmielewska J, Netre C, et al. Activation of blood coag-
131. Chambers LA, Long KF, Wissel ME, et al. Multisite trial of a pH paper ulation, fibrinolytic and kallikrein systems during storage of plasma.
and a pH value of less than 7.0 to screen whole-blood-derived platelets: Vox Sang 1984;47:355–342.
implications for bacterial detection. Transfusion 2004;44:1261–1263. 158. Buchta C, Felfernig M, Hocker P, et al. Stability of coagulation factors
132. Hay SN, Brecher ME. Validation of pH and glucose determination in thawed, solvent/detergent-treated plasma during storage at 4°C for
for bacteria detection screening in platelet concentrates stored in the 6 days. Vox Sang 2004;87:182–186.
Terumo Teruflex XT612 platelet container. Transfusion 2004;44:1395. 159. Code of Federal Regulations, 21 CFR 601.22: U.S. Government Pub-
133. Brecher ME, Means N, Jere CS, et al. Evaluation of an automated cul- lishing Office, Washington, D.C. (revised annually).
ture system for detecting bacterial contamination of platelets: an anal- 160. Code of Federal Regulations 21 CFR 610.40: U.S. Government Print-
ysis with 15 contaminating organisms. Transfusion 2001;41:477–482. ing Office, Washington, D.C. (revised annually).
134. Larsen CP, Ezligini F, Hermansen NO, et al. Six years’ experience of 161. Code of Federal Regulations, 21 CFR 640.65 2005: U.S. Government
using the BacT/ALERT system to screen all platelet concentrates, and Printing Office, Washington, D.C. (revised annually).
additional testing of outdated platelet concentrates to estimate the fre- 162. Pantanowitz L, Kruskall MS, Uhl L. Cryoprecipitate. Patterns of use.
quency of false-negative results. Vox Sang 2005;88:93–97. Am J Clin Pathol 2003;119:874–881.
135. AuBuchon JP, Cooper LK, Leach MF, et al. Experience with universal 163. Hughes C, Thomas KB, Schiff P, et al. Effect of delayed blood process-
bacterial culturing to detect contamination of apheresis platelet units ing on the yield of factor VIII in cryoprecipitate and factor VIII con-
in a hospital transfusion service. Transfusion 2002;42:855–861. centrate. Transfusion 1988;28:566–570.
136. Vostal JG. FDA’s current considerations for 7 day platelets and bacte- 164. McCullough J, Weiblen BJ, Fine D. Effects of storage of granulocytes
rial detection in single and pooled platelet products, January 25, 2005. on their fate in vivo. Transfusion 1983;23:20–24.
Available at http://www.hhs.gov/bloodsafety/presentations/Vostal- FDA 165. Glasser L, Lane TA, McCullough J, et al. Neutrophil concentrates:
Positionfor7dayplatelets.pdf. Accessed October 30, 2005. functional considerations, storage, and quality control. J Clin Apher
137. Moroff G, Holme S, Dabay MH, et al. Storage of pools of six and eight 1983;1:179–184.
platelet concentrates. Transfusion 1993;33:374–378. 166. FDA Compliance Program Manual. Chapter 42: Blood and Blood
138. Snyder EL, Stack G, Napychank P, et al. Storage of pooled platelet con- Products: Inspection of Licensed and Unlicensed Blood Banks, Brokers,
centrates. In vitro and in vivo analysis. Transfusion 1989;29:390–395. Reference Laboratories, and Contractors - 7342.001. FDA Office of
139. Ferrer F, Rivera J, Corral J, et al. Evaluation of pooled platelet concen- Regulatory Affairs. Available at http://www.fda.gov/cber/7342001bld.
trates using prestorage versus poststorage WBC reduction: impact of htm. Accessed October 12, 2005.
12
filtration timing. Transfusion 2000;40:781–788. 167. Ratko TA, Cummings JP, Oberman HA, et al. Evidence-based recom-
140. Wagner SJ, Robinette D, Nazario M, et al. Bacteria levels in compo- mendations for the use of WBC-reduced cellular blood components. 203
nents prepared from deliberately inoculated whole blood held for 8 or Transfusion 2001;41:1310–1319.
24 hours at 20 to 24 degrees C. Transfusion 1995;35:911–916. 168. Standard 5.7.4.1 Silva MA (ed). Standards for Blood Bank and Trans-
141. Sweeney JD, Kouttab NM, Holme S, et al. Prestorage pooled whole- fusion Services, 23rd Edition. Bethesda, Md, AABB, 2004.
blood-derived leukoreduced platelets stored for seven days, preserve 169. Lutz P, Dzik WH. Large-volume hemocytometer chamber for accurate
acceptable quality and do not show evidence of a mixed lymphocyte counting of white cells (WBCs) in WBC-reduced platelets: Validation
reaction. Transfusion 2004;44:1212–1219. and application for quality control of WBC-reduced platelets prepared
142. Heddle NM, Barty RL, Sigouin CS, et al. In vitro evaluation of prestor- by apheresis and filtration. Transfusion 1993;33:409–412.
age pooled leukoreduced whole blood-derived platelets stored for up 170. Dzik S, Moroff G, Dumont L. A multicenter study evaluating three
to 7 days. Transfusion 2005;45:904–910. methods for counting residual WBCs in WBC-reduced blood compo-
143. Heddle NM, Cook RJ, Blajchman MA, et al. Assessing the effectiveness nents: nageotte hemocytometry, flow cytometry, and microfluorom-
of whole blood-derived platelets stored as a pool: a randomized block etry. Transfusion 2000;40:513–520.
noninferiority trial. Transfusion 2005;45:896–903. 171. Buck SA, Kickler TS, McGuire M, et al. The utility of platelet wash-
144. Tovey LA, Lathe GH. Caeruloplasmin and green plasma in women ing using an automated procedure for severe platelet allergic reactions.
taking oral contraceptives, in pregnant women, and in patients with Transfusion 1987;27:391–393.
rheumatoid arthritis. Lancet 1968;2:596–600. 172. Vesilind GW, Simpson MB, Shifman MA, et al. Evaluation of a cen-
145. Wolf P, Enlander D, Dalziel J, et al. Green plasma in blood donors. N trifugal blood cell processor for washing platelet concentrates. Trans-
Engl J Med 1969;281:205. fusion 1988;28:46–51.
146. American Society of Anesthesiologists Task Force on Blood Compo- 173. Weisbach V, Riego W, Strasser E, et al. The in vitro quality of washed,
nent Therapy. Practice guidelines for blood component therapy. Anes- prestorage leucocyte-depleted red blood cell concentrates. Vox Sang
thesiology 1996;84:732–747. 2004;87:19–26.
147. O’Shaughnessy DF, Atterbury C, Bolton Maggs P, et al. Guidelines for 174. Kalmin ND, Brown DJ. Platelet washing with a blood cell processor.
the use of fresh-frozen plasma, cryoprecipitate and cryosupernatant. Transfusion 1982;22:125–127.
Br J Haematol 2004;126:11–28. 175. Toth CB, Kramer J, Pinter J, et al. IgA content of washed red blood cell
148. Cardigan R, Lawrie AS, Mackie IJ, et al. The quality of fresh-frozen concentrates. Vox Sang 1998;74:13–14.
plasma produced from whole blood stored at 4°C overnight. Transfu- 176. Orlin JB, Ellis MH. Transfusion-associated graft-versus-host disease.
sion 2005;45:1342–1348. Curr Opin Hematol 1997;4:442–448.
149. Weisert O, Jeremic M. Preservation of coagulation factors V and VIII 177. FDA Guidance for Industry. Gamma irradiation of blood and blood
during collection and subsequent storage of bank blood in ACD-A and components: a pilot program for licensing, February 2000. Available
CPD solutions. Vox Sang 1973;24:126–133. at http://www.fda.gov/cber/gdlns/gamma.pdf. Accessed November 18,
150. Nilsson L, Hedner U, Nilsson IM, et al. Shelf-life of bank blood and 2005.
stored plasma with special reference to coagulation factors. Transfu- 178. Pelszynski MM, Moroff G, Luban NL, et al. Effect of gamma irradia-
sion 1983;23:377–381. tion of red blood cell units on T-cell inactivation as assessed by limiting
151. Kakaiya RM, Morse EE, Panek S. Labile coagulation factors in thawed dilution analysis: Implications for preventing transfusion-associated
fresh frozen plasma prepared by two methods. Vox Sang 1984;46:44–46. graft-versus-host disease. Blood 1994;83:1683–1689.
BLOOD BANKING 179. Luban NL, Drothler D, Moroff G, et al. Irradiation of platelet com- 185. FDA Guidance for Industry. Recognition and use of a standard for the
ponents: Inhibition of lymphocyte proliferation assessed by limiting- uniform labeling of blood and blood components, June 2000. Available
dilution analysis. Transfusion 2000;40:348–352. at http://www.fda.gov/cber/gdlns/unilabbld.pdf. Accessed November
180. Davey RJ, McCoy NC, Yu M, et al. The effect of prestorage irradiation 18, 2005.
on posttransfusion red cell survival. Transfusion 1992;32:525–528. 186. AABB Association Bulletin #05–12, October 12, 2005.
181. Mintz PD, Anderson G. Effect of gamma irradiation on the in vivo 187. Ashford P (ed). An Introduction to ISBT 128, 2nd Edition: ICCBBA,
recovery of stored red blood cells. Ann Clin Lab Sci 1993;23:216–220. Inc, 2002.
182. Hillyer CD, Tiegerman KO, Berkman EM. Evaluation of the red cell 188. Federal Register. 68 FR 146 July 30, 2003.
storage lesion after irradiation in filtered packed red cell units. Trans- 189. Federal Register. 67 FR 53118, August 14, 2002. Available at http://www.
fusion 1991;31:497–499. cdc.gov/od/ohs/pdffiles/DOTHazMat8-14-02.pdf. Accessed November
183. U.S. Food and Drug Administration (CBER). Guideline for the 18, 2005.
Uniform Labeling of Blood and Blood Components, August 1985. 190. George VM, Pringle TC, Kline L, et al. Development and evaluation of a
Available at http://www.fda.gov/cber/gdlns/unilabel.pdf. Accessed shipping system for platelet components. Transfusion 1996;36:335–338.
November 18, 2005.
184. Wallas CH. Closing the technology gap with supermarkets: implemen-
tation of ISBT 128. Transfusion 2005;45:1054–1055.
II
204
Chapter 13
Red Blood Cell Metabolism
during Storage: Basic Principles
and Practical Aspects
John R. Hess ● Ginine M. Beyer
BRIEF HISTORICAL REVIEW period of hope that the transfusion hepatitis problem was
solved rapidly terminated in disillusion when it was found
The risk of disease transmission by blood transfusion is that most of the additional reactive results were false positive
extremely low, particularly in developed nations. This situ- and that post-transfusion hepatitis continued to occur. As a
ation is a consequence of a number of interlocking safety result of the observations on specificity of the radioimmuno-
measures, including the selection of safe populations from assay, confirmatory testing was implemented. Subsequently,
which donors are drawn, careful donor questioning, labora- testing for antibodies to the hepatitis B core antigen (anti-
tory testing, record keeping, and the use of quality systems HBc) was licensed as a means to further reduce HBV infec-
and good manufacturing practices. These approaches have tivity, although it was initially introduced as a surrogate test
evolved over the years, and perhaps the most change has for other forms of hepatitis. In 2004 and 2005, clinical trials
been seen in the area of blood testing. Indeed, blood collec- of nucleic acid amplification tests (NAT) for HBV DNA were
tors have now added nucleic acid amplification techniques completed, and one test had been licensed as of July 2005.
to the battery of tests directed toward the safety of the blood Recognition of the hepatitis A virus and development of
supply. However, although transmission of known agents diagnostic tests for infection with this virus led to the defini-
has almost been eliminated, residual fear of the unknown, tion of most cases of residual post-transfusion hepatitis as
fueled by a continuing stream of newly emerging or newly non-A, non-B (NANB).4 Extensive studies were performed to
recognized microbes, continues. characterize the agent(s) of NANB hepatitis, but such stud-
II Although some form of blood transfusion has been used ies were without any real success until 1989.5 Many attempts
for well over a century, organized testing approaches did not were also made to identify donor characteristics that might
212 really commence until the 1940s. At that time, the transmis- be used for screening. As a result of two of these studies,
sibility of syphilis by this route was recognized, and test- recommendations were made to screen donors for elevated
ing (albeit using nonspecific methods) was implemented. levels of alanine aminotransferase (ALT) in the serum, and a
However, viral hepatitis was the biggest unresolved concern small number of institutions adopted this measure around
from the 1940s to about 1970. Clinical hepatitis was recog- 1982.6,7 Continuing studies also implicated antibodies to the
nized as an almost inevitable consequence of transfusion, hepatitis B core antigen (anti-HBc) as another surrogate for
but little could be done to prevent it. The greatest benefits NANB infectivity.8,9
during that time were derived from epidemiologic studies Further measures to reduce the incidence of post-transfu-
that established the increased infection risk attributable to sion NANB hepatitis were largely set aside as a result of the
commercial donors and prisoners. There were also a number emergence of acquired immunodeficiency syndrome (AIDS)
of attempts to use liver function tests to screen blood donors and recognition of its transmissibility by blood components
for risk of hepatitis transmission, but none of the tests was and plasma fractions between 1981 and 1984. Although the
broadly adopted. potential value of the anti-HBc test as a surrogate for AIDS
In 1965, Blumberg and colleagues1 described the Australia infectivity was discussed, this approach was adopted in only
antigen, mistakenly identifying it as an allotypic protein.2 a limited number of blood establishments. Recognition of
Subsequently, the antigen was recognized as a component of the human immunodeficiency virus (HIV) as the infec-
the hepatitis B virus (HBV), fortuitously produced in consid- tious agent of AIDS by Gallo and Montagnier and their col-
erable excess during both acute and chronic infection. This leagues10,11 led to the rapid development and introduction of
discovery underlies the entire history of specific serologic screening tests for antibodies to the virus. Testing started in
testing for transfusion-transmissible infections. Between March 1985. Because of the persistent nature of HIV infec-
1969 and 1972, all blood-collecting establishments in the tion, the presence of antibodies to HIV was closely linked to
United States adopted some form of testing for hepatitis B infectivity, and it is now clear that the introduction of this
surface antigen (HBsAg), using agar gel diffusion, counter- test played a major role in the almost total elimination of
electrophoresis, or rheophoresis. Subsequently, Ling and transfusion-transmitted HIV and AIDS. A problem with the
Overby3 reported on a radioimmunoassay for HBsAg, and exclusive use of anti-HIV as a viral marker is the fact that,
this assay was rapidly adopted. Interestingly, the insensitive early in infection, infectious virus can circulate before the
gel precipitation methods seemed to affect only about 20% of appearance of detectable levels of antibody. Thus, there has
post-transfusion hepatitis, and the frequency of radioimmu- been a continuing process of improvement of antibody tests
noassay–positive samples was about fivefold greater. A brief and implementation of additional tests in order to reduce the
INFECTIOUS DISEASE TESTING
length of this infectious window period and reduce risk. Not Unexpectedly, in 2002 it became apparent that West Nile
only have antibody tests become much more sensitive, but virus (WNV), which had entered the United States for the
also tests for HIV p24 antigen and now for HIV RNA have first time in 1999, was readily transmissible by blood trans-
been introduced. On licensure of NAT, the p24 antigen test fusion. Although infection with WNV is acute, the frequency
was shown to be superfluous and has been discontinued. of infection was so great that a significant number of blood
The first human retrovirus to be recognized was the donors did give blood in the brief viremic, asymptomatic
human T-lymphotropic virus 1 (HTLV-1).12 This virus is an phase. As a result of these observations, tests for WNV RNA
oncovirus, and its epidemiology is characterized by extreme were developed and implemented within a relatively short
geographic clustering in Japan, the Caribbean, and parts of time (about 9 months). Well over 1000 viremic and poten-
Africa. Clinical outcomes of infection with this virus are rela- tially infectious donations were identified in the first 2 years
tively uncommon but include the serious T-cell lymphoma/ of testing.14
leukemia and a neurologic disease called HTLV-associated
myelopathy, otherwise termed tropical spastic paraparesis.
Subsequently, a closely related virus, HTLV-2, was charac- APPROACH TO TESTING
terized. Its epidemiology is less well defined, but it appears
to be naturally endemic in certain indigenous populations In the United States, blood, blood components, and plasma
in the Americas and also circulates among users of injected products are classified as biologics. By extension, tests used
drugs. The first serologic tests to be developed were designed in the preparation of blood are also classified as biolog-
to detect antibodies to HTLV-1, but as a result of cross-reac- ics and, as such, are regulated by the U.S. Food and Drug
tivity, they also identified the majority of HTLV-2 infections. Administration (FDA). This licensure involves extensive
Epidemiologic studies in the U.S. donor population revealed clinical trials and stringent regulatory oversight. However,
a 0.025% prevalence of anti-HTLV,13 and this figure, along these procedures are somewhat cumbersome and involve
with the potentially serious outcomes of infection (which considerable resource use by manufacturers and regulators.
had been shown to occur through transfusion), led to the As a consequence, the selection of available tests in the United
implementation of donor testing in 1988. States may differ from that in other parts of the world. A list-
During the same period, there was a reevaluation of the ing of FDA-licensed tests may be found on the FDA website
severity of long-term outcomes of NANB hepatitis. As a (http://www.fda.gov/cber/products/testkits.htm). It should
result of this reevaluation and of a heightened awareness of be noted that some of these tests may no longer be available.
blood safety, blood establishments adopted testing for both Typical results from the use of these tests in a large voluntary
elevated ALT levels and anti-HBc, in the expectation that donor population are presented in Table 14–1.
these tests together might reduce the incidence of transfu- There is a broad commonality in the implementation
sion-transmitted NANB hepatitis by about 60%.8,9 This of all tests used to ensure the safety of blood and blood
projection turned out to be quite accurate. components. Every donation is tested, and current require-
However, the most important advance in management ments in the United States are that the test be performed on 14
of NANB was the cloning of a portion of the genome of a a sample drawn at the time of donation. With the exception
virus termed hepatitis C virus. This advance permitted the of testing for viral nucleic acids, which is described later, 213
development of a test for antibodies to the virus based on there are three phases of the testing. Each sample is tested
peptides expressed from the viral genome. Use of this test singly; nonreactive results are considered to be negative
revealed that almost all NANB hepatitis was due to hepati- for the marker, and the corresponding blood unit (and its
tis C. The first-generation test was implemented for donor components) may be issued for transfusion. If the result
screening in 1989 and 1990. Subsequently, with the expres- is reactive, the sample is retested in duplicate. If both of
sion of additional peptides, versions 2.0 and 3.0 of the test the repeated test results are nonreactive, the sample is classi-
were successively implemented, with consequent gains in fied as negative, and the unit, or its components, is released.
sensitivity. Nevertheless, because of a significant infectious However, if one or both of the repeated results is reactive, the
window period, NAT for HCV has also been implemented. sample is classified as repeatedly reactive, and the blood unit
As a result, and because none of the supposed non-A, non-B, cannot be issued for transfusion. This practice was initiated
non-C hepatitis viruses were shown to be pathogenic, testing when it was recognized that radioimmunoassays (and sub-
for ALT has also been discontinued. sequently enzyme immunoassays [EIAs]) were subject to
Table 14–1 Prevalence and Incidence of Transfusion-Transmissible Infections among American Red
Cross Blood Donors over the Two-year Period 2000–2001
Number Confirmed Rate per Hundred Number Confirmed Rate per Hundred
Positive Thousand Positive Thousand Person-years
Adapted from Dodd RY, Notari EP, Stramer SL. Current prevalence and incidence of infectious disease markers and estimated window-period
risk in the American Red Cross blood donor population. Transfusion 2002;42:975–979.
BLOOD BANKING nonrepeatable reactive results related to minor contamina- automated instrument systems are also subject to change
tion with the labeled conjugate. and improvement. In some cases, these changes improve the
The final phase of testing is the application of confirma- performance characteristics of the tests, but they also reduce
tory or supplementary tests, largely intended to ensure that the need for human intervention and improve adherence to
the donor is properly advised about the significance of the the requirements of good manufacturing practice. Third, a
screening test results. Although imperfect, this phase of test- better understanding of the early (window) phase of infec-
ing markedly improves the accuracy of donor notification tion has led to the addition of new test methods for some
and counseling. In certain circumstances, these additional agents. For example, the measures for HIV now include a test
tests may also be used to support reentry of donors whose for antibodies to the virus, and a test for viral nucleic acids.
screening test results are definitively false positive. In some In addition, a number of tests are supplemented to detect
countries, a second screening test is used in place of different additional strains or subtypes of a given agent. Fourth, con-
technology. In the United States, it has proved effective to use tinuing concern about blood safety is likely to lead to the
this second EIA strategy to reduce the number of confirma- implementation of tests for other agents including imported
tory tests that have to be performed.15 The strategy requires diseases such as Chagas disease17 and emerging infections
that the two EIA tests have essentially the same sensitivity. such as babesiosis.
Despite the use of supplementary or confirmatory testing,
the general rule in the United States is that a donor is indefi-
nitely deferred on the basis of a repeatedly reactive screening WINDOW PERIOD
test result.
There are some exceptions to this generalized algorithm. At least until the late 1980s, the serologic tests in use were
For example, there is no supplementary or confirmatory considered to have adequate sensitivity for detecting well-
test for antibodies to the HBc antigen. In this case, a donor established, chronic infection. Comparison of detection rates
is not permanently deferred until he or she is found to be among first-time and repeat donors generally revealed that
repeatedly reactive on a second occasion. Recent data, how- about 50% of all confirmed positive test results were found
ever, suggest that this policy is of little value and will become among first-time donors, although they contributed only
even less so as the specificity of tests for anti-HBc improves. 20% of all donations. Concern about the continued occur-
Somewhat similarly, a donor does not need to be deferred rence of transfusion-transmitted HIV infection led to much
the first time a test for anti–HTLV-1 or -2 is repeatedly reac- closer evaluation of the risk of infectivity during the early
tive, provided the supplementary test result is nonreactive or phases of infection, particularly the viremic but antibody-
indeterminate. negative window period. Indeed, in a study by Petersen and
Starting in 1999, blood collection establishments in the colleagues,18 HIV infectivity was demonstrated in one in five
United States implemented NAT for HCV and HIV RNA in seronegative donations collected from individuals who were
donor samples. This testing was performed on small (i.e., HIV antibody positive at their next donation. Mathematic
II 16 or 24) pools of samples. In some cases, primary testing modeling showed that the period of such infectivity averaged
was performed using a multiplex system.16 Consequently, 45 days as of 1990.18
214 the testing algorithms differ significantly from those out- Knowledge of the length of the window period and of
lined previously. A second round of testing is used to resolve the incidence of new infection in the donor population also
pools, followed, in some cases, by another test to differentiate permitted estimation of the residual risk of infection from
HIV from HCV. In addition, a second, different NAT proce- transfusion. The availability of closely spaced, sequential
dure is used to confirm or support the results of the primary samples from commercial plasma donors permitted careful
test. Initially, all NAT for donor blood in the United States characterization of the window period and of the impact of
was performed under Investigational New Drug (IND) pro- additional tests. These studies showed that as the sensitivity
tocols, but tests were licensed in 2002. Table 14–2 outlines of antibody tests was increased, the HIV window period was
the results of NAT for the U.S. blood donor population from reduced to about 22 days.19 The impact of the addition of
March 2000 through April 2002.16 tests for the HIV-1 p24 antigen and HIV RNA was also esti-
Infectious disease testing is not static; there is continu- mated.19 Similarly, the dynamics of early infection with HCV
ing improvement in four broad areas. First, individual tests and HBV are now well defined.20 Although data published in
are being modified to ensure improved sensitivity and speci- 2002 suggested that the benefits of NAT were consistent with
ficity. This process is perhaps best exemplified by the rapid projections,21 this was not the case for HIV-1 p24 antigen,
progression of anti-HCV tests from version 1.0 to version which was detected much less frequently than anticipated.
3.0. Second, the formats of the tests and the nature of the It is possible that because the HIV antigen peak corresponds
Table 14–2 Results of NAT for HIV and HCV, USA March 1999–April 2002, and for WNV, 2002–2003
*
Data represents collections only during the period that WNV was detected among donors.
Adapted from Stramer SL, Fang CT, Foster GA, et al. West Nile virus among blood donors in the United States, 2003 and 2004. N Engl J Med
2005;253:451–459; Stramer SL, Glynn SA, Kleinman SH, et al. Detection of HIV-1 and HCV infections among antibody-negative blood donors by
nucleic acid-amplification testing. N Eng J Med 2004;351:760–768.
INFECTIOUS DISEASE TESTING
with the symptomatic phase of early acute HIV infection, the characterization of the agent in question. The classic exam-
affected donors feel unwell and are less likely to present for ples of this approach are the use of nontreponemal tests for
donation. syphilis and the implementation of testing for HBsAg. In
fact, both of these approaches turned out to be appropriate
and successful, but we are now in a much better position to
SCIENTIFIC BASIS FOR TEST SELECTION select methods on the basis of the properties of the infec-
tious agent itself, and of its pathology and natural history.
Determining a Need for Testing Broadly speaking, the approaches are to identify the infec-
tious agent itself, or to focus on the body’s reaction to infec-
There are a number of levels of decision making relative to tion with the agent.
testing for transfusion-transmissible infections. At the high-
est level is the determination to test for a given agent. Many
Direct Identification of Infectious Agents
factors are considered in this process, including the potential
frequency of transmission of the disease, its severity, and the Direct detection of the agent itself may involve techniques as
efficacy of treatment. This is perhaps the most difficult deci- simple as visualization, or as complex as nucleic acid ampli-
sion to make, and there is no clear guidance, methodology, or fication. Obviously, direct visualization is unlikely to be
prioritization process in place. In fact, in the United States, sensitive enough to support current needs for blood safety,
there are no clear decision-making hierarchies for implement- although it is used in some circumstances—for example,
ing a new test, although final authority for requiring a test evaluation of donor malaria infection in parts of the devel-
rests with the FDA. During the history of blood testing, there oping world. Most current technologies in widespread use
have been examples of voluntary initiation of tests, either involve the immunologic detection of specific antigens and
by a local decision in one or more blood establishments, or amplification of nucleic acids. Antigen testing has undoubt-
through professional bodies, such as the AABB. In the first edly been of greatest use up to the present time and has dem-
of these categories, a number of individual blood establish- onstrated its greatest efficacy in the detection of infection
ments independently initiated early testing for ALT, anti- with HBV. Its efficacy is directly attributable to an essentially
HBc, or HIV-1 p24 antigen. In the past, the AABB, through unique property of HBV infection, which is the synthesis of
its standard-setting procedures, has called for testing for ALT a great excess of the viral surface antigen. This, in effect, is an
and anti-HBc, and more recently, for methods of detecting amplification mechanism, because immunologically based
bacteria in platelet concentrates. In addition, even if it is clear tests can usually only detect analytes at levels well above the
that a preventative measure is required for a given agent, there picogram-per-mL range. It is not to be expected that many
is no process to make sure that a test will be developed and other infectious agents will be as readily detected through
made available. Conversely, it appears that the development their antigenic components. Nevertheless, HBV has been
of a test will not necessarily lead to its adoption, at least in the considered to be a model, and a great deal of effort has been
short term. However, the development and implementation expended on a search for detectable soluble antigens from 14
of a test for WNV RNA appears to be a good example of a other transfusion-transmissible viruses. In fact, methods to
case in which the need was clear, and all stakeholders worked identify antigens of both HIV and HCV have been success- 215
together to achieve an effective outcome.14 fully developed and deployed. However, the antigenic levels
Although it would seem appropriate to base a decision have been low, and their sensitivity in detecting infectivity
about test implementation on science and cost-benefit anal- is greatly surpassed by nucleic acid testing. For example,
yses, there is now a much greater tendency to invoke a vari- the maximum sensitivity of a test for HCV core antigen
ant of the precautionary principle. In effect, this tendency is equivalent to about 30,000 to 50,000 genome copies of
translates to implementing some measure to reduce transfu- RNA, although this sensitivity can be increased by the use
sion risk before data are available to define the efficacy of of immune complex dissociation techniques.24 Another issue
such a measure. The major constraint is that the measure with antigen detection is that, in general, antigens cease to
should not have any deleterious effect. It has been argued be readily detectable once a significant level of antibody has
that the use of this approach has been successful in the case developed. Thus, in the absence of methods using immune
of donor deferral policies for risk of vCJD.22 However, in complex dissociation, antigens are detectable only dur-
this case, the argument was that measures had been taken ing the early acute phases of infection. (This is not the case
before the disease had clearly been shown to be transmissible for HBV, because the levels of antigen synthesized during
by transfusion. The selected measures had not been shown chronic infection are enough to overwhelm the antibody
to be efficacious, so the argument for success was based on response and lead to a situation of antigen excess.) Thus,
the finding of transmission, rather than a demonstration with the exception of HBsAg, antigen detection methods are
that the measures actually prevented it. If the precautionary used to supplement other detection procedures by provid-
principle is used to promote the use of unproven methods, ing for earlier detection of infection and thus reducing the
it is important to recognize that the principle is not absolute length of the infectious window period. A test for HIV p24
and that certain constraints should be considered before it is antigen was in use for this purpose for a number of years in
invoked.23 the United States and some other countries, but at least in
the United States it has been supplanted by the much more
Selecting a Test Methodology sensitive NAT. Some countries have adopted a test for HCV
core antigen instead of implementing the more costly and
Of more scientific interest is the selection of an appropriate complex NAT. An evolving trend is the development of so-
methodology to be used to test for any given agent. In the called combo assays, which permit the detection of both the
past, this selection was often based on the historic availabil- viral antigen and the antibody. Such tests offer operational
ity of a given technology and its linkage to the discovery or advantages, but they are generally somewhat less sensitive
BLOOD BANKING than individual tests alone. In general, it should not be antic- fined to persistent infections, such as HCV, HIV, or Chagas
ipated that blood product contamination by any emerging disease. In contrast, as pointed out above, antibody testing
infection will be effectively prevented solely by testing for its does not offer any benefit for the prevention of acute infec-
antigens. tions, as exemplified by WNV, where it appears that infectiv-
Nucleic acid amplification tests are extremely sensitive, ity diminishes or disappears as antibody becomes detectable.
with the theoretical possibility of detecting a single molecule Clearly, in such circumstances, direct detection of the agent
of nucleic acid, although this level is rarely achieved in prac- is the preferred approach. More broadly, the use of any
tice. Conceptually, NAT was introduced into donor testing given test approach for donor screening should be based on
as a way to increase safety by identifying individuals in the knowledge of the natural history of each particular infection,
viremic window period. In large part, it appears to have ful- rather than on the availability of a test method per se.
filled this expectation, reducing the estimated risk of trans- At the time of writing, there is interest in an entirely dif-
fusion-transmitted HCV from 1 per 200,000 units to 1 per ferent approach to the detection of infection, based on pro-
1,390,000 units, and that of HIV from 1 per 1,048,000 units teomics. Modern technology permits the characterization of
to 1 per 1,525,000 units.21,25 However, it is clear that NAT actual or relative concentrations of a very large number of
is insufficiently sensitive to identify every infectious blood proteins (for example, in the plasma). Some of these protein
unit, and it is currently seen as a supplement to conventional levels may change as a result of infection with a particular
immunologic tests for antibodies to HIV and HCV. NAT for organism, and conceptually, at least, it may be possible to
HBV DNA is now available and, at the time of writing, one identify an infection through a particular pattern of change
test has been licensed for routine use, but such use is has not in the levels or distribution of these proteins. Rapid, simul-
been mandated by the FDA. This is the case largely because taneous screening of large numbers of proteins, along with
there appears to be relatively little benefit to implementing intensive computational capabilities, makes this identifica-
this test on pooled samples, relative to the use of highly sen- tion possible. However, this approach has yet to be shown
sitive tests for HBsAg. Nevertheless, a continuing question to be appropriately sensitive and specific, or even simple
is whether NAT could displace currently used serologic tests enough, for routine use.
for any of these viruses. Certainly, it is clear that a proportion
of donor samples may have confirmed positive test results
Selecting a Specific Test
for antibodies, but they may be negative for RNA or DNA.
This result occurs among about 20% of anti-HCV–positive, The final level of test selection is the choice of which partic-
and about 0.4% of anti-HIV–positive donors (SL Stramer, ular manufacturer’s test to use. Although it may be argued
unpublished observations). In some, but not all, cases, these that in many parts of the world tests must be approved by
findings do represent resolution of infection and viral clear- a regulatory authority, this requirement does not ensure
ance. Overall, at the time of writing, it seems unlikely that that all approved tests have equivalent performance char-
conventional test methods will be eliminated. acteristics. This fact is clearly demonstrated by the differ-
II The value of NAT for managing emerging transfusion- ences in analytic sensitivity for licensed (and unlicensed)
transmissible infections has been amply illustrated by the HBsAg tests reported by the FDA.26 Consequently, a process
216 U.S. experience with WNV. NAT for this virus was developed of evaluation of test characteristics is desirable in order to
by two manufacturers and implemented within 9 months inform purchase decisions, in addition to issues of cost and
of the recognition of need. It is certainly easier to develop convenience.
a nucleic acid-based assay than an immunoassay, because it For blood screening tests, the key performance parameters
does not require the expression of antigens or the production are sensitivity and specificity. In the case of most markers, ana-
of antibodies. However, NAT may not necessarily be the most lytic sensitivity directly impacts epidemiologic sensitivity. In
effective solution for all emerging infections. It has been very other words, a more sensitive test will identify more infected
successful in the case of WNV because the infection itself is donors. Perhaps the most important aspect of sensitivity is
acute, and it appears that the period of blood-borne infectiv- the ability to detect infection at the earliest time, because
ity is restricted to the earliest phases of infection, prior to this detection reduces the length of the infectious window
the development of significant levels of antibodies. In fact, period and hence the risk of transmitting infection to recipi-
there do not appear to have been any well-documented cases ents. Sensitivity for early infection is best evaluated against
of transfusion transmission of WNV from donors who had panels of samples taken during seroconversion. Such panels
developed antibodies to WNV.14 are commercially available and have usually been collected
from donors of plasma for further manufacture. It should,
Detection of Antibody and Other however, be noted that in an environment where multiple
tests are used for any given infectious agent, the sensitivity of
Responses to Infection the overall test system is most important. For example, use of
The other broad approach to testing for infectivity is by NAT may compensate for an antibody test that may not be
assessing the body’s response to the agent. Most frequently, the most sensitive one available. Another aspect of sensitiv-
of course, this assessment is done by detecting antibodies, ity is the ability of a given test to detect different serotypes,
although other methods have been used or considered. The genotypes, or mutants of the agents of concern. Most avail-
major benefit of testing for antibodies is that the immune able tests have been designed to have optimal sensitivity for
response is a form of amplification: infection with, or expo- the strains of agent that predominate in the United States.
sure to, a modest titer of a microorganism generates levels of Even though a particular strain or genotype is very rare,
circulating antibodies that are readily detectable by relatively there may be benefits to using tests with broader sensitivity.
simple procedures. Antibody tests have been highly effective For example, in the United States, potential donors must be
as a measure to control transfusion-transmitted infection deferred if they have a history of travel, residence, or sexual
with a number of agents, but this effectiveness is really con- contact in certain African countries where HIV-1 type O is
INFECTIOUS DISEASE TESTING
present, unless a test with FDA-validated claims for group O predictive value of the test is also low. Thus, as with any other
detection is used. donor screening test, any repeatedly reactive finding should
Test specificity is another critical parameter. Tests with be confirmed before the donor is notified. Manufacturers of
poor specificity result in unnecessary loss of products and HBsAg test kits also provide confirmatory reagents consist-
their donors, and they require performance of additional ing of specific antibodies to HBsAg. The test is repeated in the
confirmatory tests. Furthermore, blot-based confirma- presence of the antibody, and a control test is also run using
tory tests frequently generate indeterminate results that are normal serum. If the added antibody inhibits the test signal
hard to interpret and which cause unnecessary alarm when by 50% or more relative to the control, the reactive result
donors are notified. The very low prevalence of some infec- is considered to be confirmed as positive. There are some
tions among blood donors magnifies the impact of poor test additional steps that are required in specified circumstances
specificity. For example, the positive predictive value of a (such as a strong primary test result). It should be noted that
repeat reactive test result for HIV is currently around 5% in very weak reactive signals may also give false-positive results
the U.S. donor population, even though current tests have in this confirmatory procedure.
good to excellent specificity. A test’s specificity is best evalu- As indicated earlier, testing for anti-HBc was uniformly
ated by testing many nonreactive samples, preferably from introduced in the United States in 1986 and 1987. It was
donor populations. originally used to screen for NANB hepatitis infectivity as
Ideally, assessments of sensitivity and specificity should a result of some epidemiologic correlation between HBV
be made in field conditions. However, doing so may be and HCV infection, but this use is no longer considered
unrealistic, and decisions may have to be based on available valid or appropriate. In fact, anti-HBc testing was licensed
information. Although comparative evaluations of test kits by the FDA as an additional measure to improve safety with
are occasionally published in the open literature, every kit respect to HBV. The core antigen is actually the viral capsid,
approved in the United States includes a product insert with and antibodies appear early in infection, but after HBsAg.
summary information of the results of the clinical trials that Originally, it was hypothesized that anti-HBc was the only
define the claimed performance characteristics for the test. marker detectable during a window period after the decline
Tests should be expected to meet these performance claims in of HBsAg and before the appearance of the corresponding
routine use. Clinical trials for licensure of tests for donation protective antibody (anti-HBs). Current test methods are so
screening require evaluations of test sensitivity and specific- sensitive that this is probably no longer the case. Nonetheless,
ity in field conditions, along with evaluations of the impact there are both old and contemporary data suggesting that
of samples with interfering substances. Sensitivity evalua- a minority of donations with anti-HBc as the sole marker
tions include the use of a wide variety of positive samples of HBV infection may be infectious.27–30 However, the pres-
representing different genotypes. In addition, a population ence of significant levels of anti-HBs seems to negate the
of high-risk samples is usually tested. Specificity is assessed risk of any infectivity through the blood (although not from
by testing 10,000 or more routine donor samples. Samples a transplanted liver).31 Consequently, in Japan, a country
giving positive results must be further investigated and the with a high prevalence of HBV infection, donors are tested 14
donors evaluated by follow-up. for HBsAg, anti-HBc, and anti-HBs. All units with detect-
able HBsAg are discarded, as are those with anti-HBc in the 217
absence of anti-HBs. However, units with both antibodies
TESTS BY AGENT are used for transfusion.32
There are two approaches to anti-HBc testing. The ini-
Hepatitis B Virus tial test that was developed was an inhibition immunoas-
say, which is still available. The solid phase has recombinant
Hepatitis B Immunoassays HBc antigen as a capture reagent, and the probe is a labeled,
The primary means for the detection of hepatitis B infectiv- partially purified anti-HBc antibody. The presence of anti-
ity is the test for HBsAg. This antigen is the first serologic HBc in the test sample inhibits the signal. Perhaps surpris-
marker to appear during acute HBV infection, and it per- ingly, this test has rather poor specificity, in part because of
sists during active, chronic infection. It is produced in the reduction-sensitive interfering molecules in some samples.33
cytoplasm of HBV-infected cells, and it is often present in Newer versions of the test incorporate reductants, which
the serum at high levels, even up to micrograms per mil- clearly increase the specificity of the test. However, the inhi-
liliter. The HBsAg represents the viral coat and is found in bition procedure is also quite dependent on good technique,
the serum or plasma in the form of self-assembled spheres and reproducibility is not optimal. A second test, a direct
and tubules 22 nm in diameter. It is readily detected by sim- antiglobulin assay for anti-HBc, is also available; it appears to
ple, sandwich-type immunoassays using animal antibodies be more specific than the inhibition procedure. There is no
to HBsAg (anti-HBs) as a solid-phase capture reagent and formal confirmatory procedure for anti-HBc, but the use of
a conjugated anti-HBs as a probe. Conventional assays use two different tests may help to increase the predictive value
either a bead or a microplate well as the solid phase, but of a reactive result.
tests (particularly automated ones) are increasingly based
Nucleic Acid Amplification Testing
on microparticle substrates. Similarly, enzyme conjugates
for Hepatitis B Virus
and chromogenic detection methods are being replaced by
chemiluminescent labels. A number of NAT procedures for HBV DNA have been
Currently available tests have an analytic sensitivity on the developed and, in some cases, put into routine use, particu-
order of 0.5 ng/mL, with a range of 0.08 ng/mL to 0.7 ng/mL. larly in Japan and parts of Germany.34,35 In the United States,
The epidemiologic sensitivity and specificity of the test are two commercial procedures have been developed and evalu-
high. Nevertheless, because the prevalence of positive HBsAg ated in clinical trials.36 At the time of writing (2005), one of
findings in the donor population is quite low, the positive these procedures has been licensed by the FDA. However,
BLOOD BANKING the FDA did not specifically require its use. Clinical trials a test sample, and any adherent antibodies are detected and
of the licensed test using minipool procedures did result visualized by use of an antiglobulin conjugate and a chro-
in the detection of HBV DNA–positive, serology-negative mogenic substrate. The test is designed to complement the
donations at a frequency of about 1 per 340,000 donations. version 3.0 EIA and consequently uses essentially the same
In general, minipool testing for HBV DNA is considered to peptides. Clearly, the test was developed to dissect out and
have a sensitivity similar to that of the most sensitive tests for identify the presence of antibodies to different viral epit-
HBsAg, and there is considerable debate about the real safety opes. In common with the Western blot, this approach has
benefits of minipool NAT for HBV in the United States. It two deficiencies: (1) it is a subjective test, and (2) it does not
must be remembered that anti-HBc testing of donors is in generate outcomes with unequivocal interpretations. That is,
routine use in the United States, and studies have shown that results may be defined as nonreactive, positive, or indeter-
fewer than 1% of anti-HBc-reactive donations have detect- minate. A positive test is defined on the basis of at least two
able HBV DNA. Consequently, minipool NAT for HBV may bands with a density greater than or equal to that of the low-
have more value in an environment where anti-HBc testing positive control, but in the absence of an SOD band. A nega-
is not in use. It is considered likely that the use of individual tive result is defined as one or more bands with a density less
donation testing for HBV DNA would offer some increased than that of the low-positive control. A result is defined as
safety, because the increased sensitivity may decrease the indeterminate if there is only one band with a density greater
length of the window period. Finally, HBV NAT of plasma for than or equal to that of the weak-positive control or if an
further manufacture has been implemented using relatively SOD band is present, irrespective of other band patterns.
large pools of samples (i.e., 512 to 1200). An indeterminate result does not clearly establish the pres-
ence or absence of current or prior infection with HCV. In
Other Hepatitis B Virus Tests testing donors, about 20% of EIA repeat-reactive donations
There are other tests for markers of HBV infection, but, in are found to be indeterminate, but only about 1% of these
general, they have little current relevance to transfusion med- donations are positive for HCV RNA by polymerase chain
icine, at least in the United States. As pointed out earlier, tests reaction (PCR).37,38
for anti-HBs are available, and a positive result usually (but
Hepatitis C Virus Nucleic Acid
not always) signifies a resolved infection and the absence of
Amplification Testing
circulating virus.32 Indeed, the presence of anti-HBs may not
signify complete elimination of HBV, because the virus may At the time of writing, two commercially available NAT pro-
still be present in the liver; transplanted organs from donors cedures are in use for testing donor blood for evidence of
with anti-HBs can result in HBV infection in a susceptible HCV infectivity.16 One is a PCR procedure that is conducted
recipient.31 Nonetheless, anti-HBs testing may be used to on RNA extracted by standard chemical methods from pools
identify anti-HBc–reactive donations that are safe for trans- of 24 plasma samples. The other approach uses a transcrip-
fusion, as in Japan. There are commercially available tests tion-mediated amplification procedure on nucleic acid prep-
II for immunoglobulin M anti-HBc, and this test is useful for arations made by a solid-phase, probe-capture method. Both
diagnosis of acute HBV infection. Tests for the hepatitis B e approaches are successful in detecting HCV antibody–nonre-
218 antigen are of some value in defining the severity of an infec- active, RNA-positive window–phase donations. Furthermore,
tion, but the value of a test for the corresponding antibody both methods seem to have the same performance character-
(anti-HBe) is less clear. istics when the different sensitivities of the version 2.0 and 3.0
EIAs are accounted for. Following infection, HCV RNA levels
increase rapidly and reach values of 105 to 107 copies/mL dur-
Hepatitis C Virus ing the period about 40 to 50 days before the appearance of
HCV Immunoassays detectable HCV antibody. In a review of national experience
with NAT, Stramer and colleagues16 showed that the detec-
The primary screening and diagnostic test for HCV infec- tion rate for HCV RNA was about 1 in every 230,000 sero-
tion is an antiglobulin EIA for antibodies to the virus. As of negative donations.16 Accepted testing algorithms for blood
the end of 2005, only two tests were available in the United donor screening differ somewhat from those for diagnosis
States. Both use immobilized, recombinant viral antigens as or screening of the general population, inasmuch as the NAT
the capture reagent and an anti-immunoglobulin conjugate procedures have not been formally accepted as part of the
as the probe. One test is defined as a version 2.0 test, and the confirmatory process for donors. This subject is discussed in
other as a version 3.0. Both tests use antigens representing more detail in Chapter 44.
the core and NS3/NS4 regions of the viral genome, and the Although not in widespread use, immunologic tests for
3.0 version includes an NS5 peptide. In comparison, the first the HCV core antigen have been developed, with a sensitivity
test to be licensed (termed 1.0) used only a single peptide that is approximately equivalent to the detection of 30,000 to
in the capture reagent. This HCV 1.0 test, although a con- 50,000 copies of HCV RNA. Although less sensitive than NAT
siderable advance on surrogate testing, lacked both sensitiv- for HCV RNA, core antigen tests nevertheless are capable of
ity and specificity. Subsequent versions are much improved substantially reducing the infectious window period.
with respect to both of these characteristics.
One licensed supplementary test for anti-HCV is available
in the United States. This test is an immunoassay consisting Human Immunodeficiency Virus
of a nitrocellulose paper strip bearing HCV peptides at spec-
ified locations. There are also controls to ensure that the test Immunoassays for Human Immunodeficiency
Virus Antibodies
is performed properly and to identify reactions due to anti-
bodies to superoxide dismutase (SOD), a carrier protein for As a direct result of the discovery of HIV, the causative
expression of some of the peptides. The strip is exposed to agent of AIDS, the first tests for antibodies to this virus were
Human Immunodeficiency Virus
II
222
C. Regulatory, Quality, and Legal
Principles
Chapter 15
Regulatory Principles and Issues Central
to Blood Banking and Transfusion
Medicine
Shealynn B. Harris ● Christopher D. Hillyer
BLOOD AND BLOOD MANUFACTURING: regulatory oversight of all biological products, including blood,
THE FDA vaccines, cellular therapies, gene therapies, tissues, and related
devices.4,7 CBER’s research operations are further divided into
Background offices and divisions. The Office of Blood Research and Review
II (OBRR) conducts research activities related to blood and blood
The federal government has regulatory authority over safety through two divisions, the Division of Hematology
224 drugs, biologics, and devices, as mandated by the Federal and the Division of Emerging and Transfusion Transmitted
Food, Drug, and Cosmetic (FDC) Act of 1938 and the Public Diseases.8
Health Services (PHS) Act of 1944.2 The FDA, an official
agency of the Department of Health and Human Services
Regulatory Code
(HHS), is the responsible body for the regulatory oversight
and compliance enforcement of its regulations for drugs, Although considered a biological product from an FDA
biologics, and devices. Along with its regulatory duties, the organizational standpoint, in terms of legislation, blood is
FDA also engages in scientific and research endeavors, and viewed as both a drug and a biological product. Accordingly,
it serves as an information resource for industry, health care, the legal requirements for blood products and blood man-
and the public. ufacturing are contained in the FDC Act, which applies to
drugs and devices, and the PHS Act, which applies to bio-
Regulatory Organization logics. The specific federal statutes and amendments that
pertain to blood and blood manufacturing are codified in
The FDA is organized into offices, which manage and Title 21 of the CFR in Parts 210 and 211 (drugs) and 600 to
direct FDA operations, and specialized program centers, 680 (biologics).9
which oversee the specific products regulated by the FDA.3 The entities and processes subject to this regulatory code
For all FDA-regulated products and industries, the Office are broadly defined by the FDA within the CFR and more
for Regulatory Affairs (ORA) serves as the lead office for clearly described in related compliance documents. According
compliance assurance.4,5 The ORA’s principle duty is the to the FDA, a blood establishment is “a place of business under
management and coordination of FDA regulatory field one management at one general physical location.” The term
operations. These operations include routine inspections includes among others, human blood and plasma “donor
of registered and licensed facilities and compliance enforce- centers, blood banks, transfusion services, other blood
ment activities, such as assistance with legal and criminal manufacturers and independent laboratories that engage
investigations.6 in quality control and testing for registered blood product
Each regulated product and industry is also assigned to a establishments.”10 For the purposes of FDA registration and
specialized program center that is dedicated to non–field-based compliance oversight, the types of blood establishments, as
compliance management, regulatory guidance, and research listed in Table 15–2, are more specifically defined by the FDA
efforts. The Center for Biologics Evaluation and Research in terms of the extent of an establishment’s manufacturing
(CBER) is the specialized program center committed to processes and level of manufacturing complexity.
REGULATORY PRINCIPLES AND ISSUES CENTRAL TO BLOOD BANKING AND TRANSFUSION MEDICINE
Table 15–2 FDA-Defined Blood Establishments and Registration Requirements
Blood Bank, Blood Center (foreign A facility, sometimes located within Required to register
and domestic) a hospital, that engages in the
manufacture of blood and blood
components including:
collection
processing of blood components product
testing compatibility testing
storing
distributing of blood products to consignees
Blood and Plasma Broker Takes physical possession of blood products Required to register
or engages in any manufacturing step. Exempt if activities restricted to the
Arranges for the sale or shipment of product. arrangement of sale or shipment
of products
Component Preparation Facility Prepares components from blood collected Required to register through parent
at a mobile or fixed collection site and establishment
operates under the control of a parent blood
bank or blood center
Contractor Person or entity that performs part or all of Required to register
the steps in the manufacture of a licensed Manufacturer and contractor share
product or that performs a service for responsibility for product quality,
a blood or blood component manufacturer however the manufacturer is
ultimately responsible
Distribution Center or Depot Stores blood and blood components under Required to register through parent
specific, controlled conditions for establishment
redistribution (intrastate or interstate) to
final users and operates under the control
of a parent blood bank or blood center
Donor (Collection) Center Fixed location that collects blood from donors Required to register through parent
by manual or automated methods and establishment
operates under the control of a parent blood
bank or blood center. May operate blood
mobiles or mobile blood drives
Hospital Transfusion Service Performs compatibility testing of (cross Not required to register if the parent
matching) for blood and blood hospital participates in the Medicare
components, but does not: reimbursement program
routinely collect allogeneic or autologous Inspection responsibility granted to
blood process whole blood into the CMS through a 1980 MOU
15
componentswash components, prepare
plasma cryoprecipitate reduced or 225
leukocyte reduced products, or irradiate
blood products
May perform:
product pooling of platelets or
cryoprecipitate
compatibility testing
transfusion product thawing
division of products
preparation of recovered plasma or red
blood cells from whole blood collected
by a blood bank
Indian Health Service (IHS) Hospital Blood establishments maintained within the IHS
See Blood Bank and Hospital Transfusion Service
Military Blood Bank and Blood establishments, foreign and domestic, All foreign and domestic
Transfusion Service maintained by the Department of Defense establishments must register
Inspections are NOT unannounced;
inspectors must notify respective
military contacts 30 days before
initiating an inspection
Testing Laboratory Perform testing for a blood bank which may Required to register
include testing for: No registration is required if only
infectious diseases testing patient samples or
donor suitability and donor reentry performing syphilis confirmatory tests
product quality
Veterans Health Administration Blood banks and hospital transfusion services All VHAMC blood banks and hospital
Medical Center (VHAMC) maintained within the VHAMC transfusion services must register
Other Blood Establishment Nonhospital-affiliated establishments that Required to register
collect blood or prepare blood cells, serum,
or plasma for “further manufacture” into
drugs or devices
Adapted from FDA Compliance Program Manual. Chapter 42: Blood and Blood Products: Inspection of Licensed and Unlicensed Blood Banks,
Brokers, Reference Laboratories, and Contractors - 7342.001. FDA Office of Regulatory Affairs. Available at http://www.fda.gov/cber/7342001bld.
htm. Accessed Oct. 12, 2005.
BLOOD BANKING In order to operate legally in the United States, a blood manufacturing operations (see Table 15–2).13 Manufacturers
establishment must comply with the specific blood manu- can register by completing either a paper Form FD-2830
facturing regulations outlined in the CFR, which include (Blood Establishment Registration and Product Listing) or
general criteria for licensing, registration, product quality, the electronic blood establishment registration (eBER) form
and component production. In addition to the CFR, which is on the FDA website. Certain blood establishments and blood
updated annually, the FDA announces changes to regulations distributors are exempt from CBER registration; these are
in the Federal Register, an official, daily publication of the facilities that do not engage in FDA-defined manufacturing
federal government in which finalized rules and regulations, activities that require a higher level of technical complexity
proposed regulations, and notices of regulatory agencies are or proficiency.14 Manufacturing steps that are considered
posted. The final rules and regulations that are published in exempt are routine compatibility testing, product pooling,
the Federal Register are incorporated into the CFR with the product thawing, and transfusion. A blood establishment
annual update. Electronic copies of both the CFR and recent that does not collect blood and restricts its operations to the
editions of the Federal Register are available on the FDA above activities (e.g., a hospital transfusion service) is deemed
website (http://www.fda.gov). exempt if its parent hospital participates in the Medicare
The regulatory requirements in the CFR are stated in broad, reimbursement program. However, if the blood establish-
general terms. To assist blood establishments with interpre- ment performs more complex steps, as above, including but
tation, standardization, and implementation of regulatory not limited to washing, irradiation, or preparation of plasma
requirements, CBER publishes guidance documents that pro- cryoprecipitate–reduced or leukocyte-reduced products and
vide background, explanation, interpretation and recommen- is not affiliated with a hospital in the Medicare reimburse-
dations for more specific industry procedures and processes. ment program, it is required to register with CBER. Exempt
Although not technically legally binding, guidance documents establishments are surveyed by the Center for Medicare and
are considered by many to be standards for the blood industry Medicaid Services (CMS) for regulatory compliance and
and current good manufacturing practice (cGMP). Guidance are not routinely inspected by the FDA15 unless just cause
documents for the blood industry are available on the CBER requires FDA inspection (e.g., a transfusion-related fatality).
website (http://www.fda.gov/cber/guidelines.htm).
Licensure
In addition to being registered, if a blood establishment
Current Good Manufacturing
engages in interstate commerce of a blood product, under
Practice Section 351 of the PHS Act, it must obtain a biologics
With regard to blood and blood establishments, relevant sec- license.16 In December 1999, in an effort to improve regu-
tions of the CFR focus primarily on cGMP. In general, cGMP latory efficiency as mandated by the FDA Modernization
requirements are set forth to ensure the safety, quality, iden- Act of 1997 (FDAMA), the previously separate establish-
tity, potency, and purity of drugs and biological products, ment license application (ELA) and product license applica-
II also referred to as SQuIPP, in the blood industry.11,12 More tion (PLA) were consolidated into a single biologics license
specifically, cGMP regulations are established for key areas of application (BLA).17 CBER oversees the BLA process, which
226 the blood manufacturing process that affect SQuIPP (Table requires submission of specified documents and a prelicense
15–3). If a blood establishment fails to comply with cGMP, the inspection of the blood establishment. Once an establishment
manufactured product may be considered adulterated, and is licensed, any intentional changes to the manufacturing
the establishment is subject to regulatory action. To effectively process that may impact product safety, purity, and potency
oversee blood establishment compliance with cGMP, the FDA must be reported to the FDA through CBER. Manufacturing
employs several controls, which include registration, licensing, changes are categorized into major, moderate, and minor,
adverse event investigations, and facility inspections. according to the effect that the change will have on the prod-
uct. The reporting action that the manufacturer takes is
based on the category of change. CBER provides a guidance
Blood Establishment Registration document to help blood establishments with the process for
and Licensure reporting a change to an approved BLA.18
Registration
Facility Inspections and Compliance
Under the FDC Act, all blood establishments are required
Enforcement
to register with the FDA and submit a list of all prod-
ucts commercially distributed within 5 days of initiating All registered, unlicensed and registered, licensed blood
establishments are required to submit to routine, unan-
nounced FDA inspections,19 which are performed at a mini-
Table 15–3 A Summary of the Key cGMP mum of every 2 years. The only exception to unannounced
Areas Referred to in 21 CFR 210, 211, and 600 inspections is for military blood establishments, for which
the FDA must give at least 30 days notice prior to the inspec-
Organization and Personnel tion.15 The primary objective of the inspection is to ensure
Buildings and Facilities
Equipment
SQuIPP by evaluating compliance with cGMP and thus the
Components, Containers, and Closures appropriate sections of the CFR. To focus on critical areas of
Production and Processing blood safety, the FDA has identified five “blood systems” and
Packaging and Labeling five “layers of safety” (listed below) on which their inspec-
Quarantine, Storage, and Distribution tion strategy is based.15 The five blood systems are Quality
Testing
Records and Reports Assurance, Donor (Suitability) Eligibility, Product Testing,
Returned and Salvaged Products Quarantine/Inventory Management, and Production and
Processing. The five layers of safety are Donor Screening,
REGULATORY PRINCIPLES AND ISSUES CENTRAL TO BLOOD BANKING AND TRANSFUSION MEDICINE
Donor Deferral, Product Testing, Quarantining, and authority over clinical laboratory facilities and, more spe-
Monitoring and Investigating Problems. The extent of the cifically, set forth the requirements for certification of such
inspection is determined by the blood establishment’s spe- facilities to operate legally in the United States. Under Section
cific manufacturing activities and is categorized by levels. A 353 of the PHS Act, the federal government introduced the
Level I inspection is a comprehensive assessment and requires following key certification requirements: (1) submission of
an evaluation of compliance with all five systems. A Level records for types of tests performed, methodologies, and
II inspection is a more streamlined process that involves a personnel training and experience; (2) accreditation by
review of three of the five blood systems. a government-approved, nonprofit, private accreditation
On arrival for inspection of a blood establishment, the ORA body; and (3) participation in a proficiency testing program
investigator must provide a Form 482 (Notice of Inspection) through a government-approved, nonprofit, private orga-
that identifies the facility, the inspector, and the legal statutes nization.24,25 Furthermore, the PHS Act granted authority
by which the inspection is authorized.20 This form must be to the Department of Health and Human Services (HHS)
signed by a facility representative or employee, often an indi- for the oversight of clinical laboratory certification, devel-
vidual of fairly high rank, indicating his or her recognition opment and promulgation of quality standards for clinical
of the official nature of the inspection. On completion of the laboratory test performance, and inspection of clinical labo-
inspection, the inspector submits Form 483 (Inspectional ratory facilities and records.
Observations), which is a list of any observations of possible In 1988, Congress passed the Clinical Laboratory
compliance violations.20 This form is also signed by a facility Improvement Amendments of 1988 (CLIA or CLIA 88) in an
official and acknowledges receipt of Form 483. Form 483 is effort to ensure the continuous improvement of the quality
not a final document of noncompliance, and the manufacturer of clinical laboratory test performance, with a primary focus
may submit any objection to the observed violations or a plan on the enhancement of test accuracy, reliability, and result
for corrective action. Then, in consultation with ORA regional turnaround time.26 Furthermore, CLIA established param-
(or national, if needed) leadership and after final review, a for- eters for the applicability and extent of clinical laboratory
mal inspection report is issued. After this compliance evalua- regulations by categorizing tests into levels of complexity.
tion is completed, and if the facility is found to be in violation,
the ORA may then begin compliance enforcement. The ORA’s Regulatory Organization
enforcement activities include advisory, administrative, and/or
judicial actions.21–23 The primary advisory action is a warning The HHS is charged with overseeing and providing services
letter, which specifies significant regulatory violations and the that relate to public health and welfare. The responsibilities
necessity for prompt and adequate corrective action. If appro- for specific regulatory oversight functions and public services
priate action is not taken by the manufacturer, or the action are divided among the twelve agencies within the HHS, each
does not satisfactorily correct the violation, the ORA can then of which has a designated focus area. With regard to regula-
proceed with administrative and/or judicial actions, which tion of clinical laboratory testing, the HHS has charged its
include citations, license revocation or suspension, product Centers for Medicare and Medicaid Services (CMS), formerly 15
recall or destruction, civil penalties, seizures, court action, and the Health Care Financing Administration (HCFA), with the
criminal prosecution. authority for general CLIA implementation and oversight.26 227
In addition to the above enforcement activities, if a blood In addition, the HHS has given specific CLIA responsibilities
establishment fails to comply with regulations after multiple to its Food and Drug Administration (FDA) for test complex-
enforcement actions, the FDA may file a “Consent Decree of ity categorization, and to its Centers for Disease Control and
Permanent Injunction” (also known as a “Consent Decree”), Prevention (CDC) for convening the Clinical Laboratory
which is a court order approved by a judge. A Consent Decree Improvement Advisory Committee (CLIAC), which provides
is a legally binding agreement, overseen by the court, between technical and scientific advice that relates to CLIA.26
the FDA and a blood establishment that contains specific The responsibilities and functions of CMS are divided
requirements and timelines for corrective action of regula- into four “programs,” which, in addition to the CLIA pro-
tory violations. The Consent Decree defines the way in which gram, include Medicare, Medicaid, and the State Children’s
the establishment must conduct its business while under the Health Insurance Programs (SCHIP). CMS performs its field
Consent Decree and allows for more frequent and in-depth operations through regional offices, which also serve as the
FDA inspections. In addition, the Consent Decree defines liaisons with state and local agencies.27 Through its Center
the penalties, such as fines, placed on the establishment for Medicaid and State Operations (CMSO), CMS also part-
while it is under the Consent Decree. If the establishment ners with state and local governments to manage CMS pro-
satisfactorily meets the requirements outlined in the Consent grams. In terms of CLIA, this partnership is most evident in
Decree, the establishment will be released from the Con- functions that relate to compliance oversight, in particular,
sent Decree and considered “in good standing” by the FDA. the performance of clinical laboratory inspections or surveys
Because the Consent Decree is a court order, if the establish- by state survey agencies.28 Individual states may apply to be
ment fails to comply with the Consent Decree, it is considered deemed as “exempt” under CLIA. When exempted, the state
“in contempt of court,” which is a criminal offense. takes on the burden of clinical laboratory regulation. At the
time of this writing, there are only two CLIA exempt states:
Washington and New York.
LABORATORY TESTING: CLIA AND CMS
The Regulatory Code and Regulated Entities
Background
The regulations that pertain to clinical laboratory testing
With the passage of the Public Health Service (PHS) Act are codified in Title 42 of the CFR. CLIA defines a labo-
of 1944, the federal government established regulatory ratory as “any facility that performs laboratory testing on
BLOOD BANKING specimens for the diagnosis, prevention, treatment of dis- are obligated to submit to routine performance surveys. CMS
ease, or impairment of, or assessment of health.”25 All labo- follows an “outcome-oriented survey process,” which focuses
ratories that perform even a single test on human specimens on assessing the overall quality of laboratory performance
for these purposes, even those entities that do not bill to and not on an in-depth evaluation of each specific regula-
Medicare, are subject to federal regulation under CLIA tory requirement.33 The survey is typically an on-site inspec-
and must register with the CLIA program.25 The primary tion of the clinical laboratory; however, if the laboratory has
exception is a laboratory that performs testing for research shown consistently good performance, the survey can be in
purposes only. Thus, all laboratories within blood banking the form of a self-assessment, known as an Alternate Quality
and transfusion services, except research laboratories, are Assessment Survey (AQAS),34 which does not require an on
regulated under CLIA. site inspection. Surveys are performed by the survey agency
in the state in which the clinical laboratory resides or, in
some cases, by the CMS regional office.
Test Complexity and CLIA Certification
If a clinical laboratory fails to comply with regulations,
The applicability of specific CLIA regulations and the extent maintains operations that pose an immediate threat to pub-
of regulatory oversight are based on the complexity of the lic welfare, and/or fails to show improvement in test per-
test(s) performed by the clinical laboratory. Based on the formance deficiencies, then CMS is obligated under CLIA
technical demands of a test and the expertise required for the to proceed with enforcement actions. These enforcement
performance of a test, a test method is placed into one of the actions include, but are not limited to, on-site monitoring of
following categories: waived complexity, moderate complex- the laboratory by the state, mandatory retraining and tech-
ity, or high complexity.26 For each moderate- and high-com- nical assistance, civil penalties (fines), suspension or revoca-
plexity test, CLIA outlines requirements for quality control, tion of CLIA certification, suspension of Medicare/Medicaid
quality assurance, personnel qualifications, proficiency test- payments, cancellation of Medicare/Medicaid approval, and
ing, and patient sample and result management. CLIA does court order of injunction to refrain from violative operations
not specify such requirements for waived tests, for example, or practices.35
a point-of-care rapid diagnostic test for influenza performed
in a physician’s office. Although required to register with
the CLIA program, clinical laboratories that perform only SUMMARY
waived tests, also known as “waived laboratories,” are only
obligated to follow the test manufacturer’s instructions.30 In summary, regulation is a government’s authority over a
Furthermore, these waived laboratories are not subject to nongovernment body’s operations and processes for the
routine inspections under CLIA.31 general intention of protecting public health and welfare.
All clinical laboratories, even those that perform only Compliance with regulations, which are the legal require-
waived tests, are required to apply to CMS for CLIA certifi- ments established through the enactment of legislation,
II cation by submitting the “CLIA Application of Certification” is nonvoluntary, and failure to comply with regulations is
(Form CMS-116) to the CMS-affiliated state agency in the justification for civil and/or criminal action. Blood banks
228 state in which the clinical laboratory resides.32 The applica- and transfusion services in the United States are subject to
tion review focuses on the assessment of the complexity of regulatory oversight by a variety of government agencies,
test(s) performed by the clinical laboratory, the determina- with specific regulatory influence from the FDA for blood
tion of the type of certificate to be issued, and the establish- and blood manufacturing and the CMS for clinical labora-
ment of fees to be collected. The types of CLIA certificates tory performance. Regulations are codified in the CFR and
are listed in Table 15–4. As a “user fee–funded” government organized into Titles according to the specific regulatory
program, all costs that relate to CLIA compliance, including agency. Given that regulatory code is continuously updated,
registration, certification, and inspection fees, are required to to ensure compliance, regulated entities must remain up-to-
be covered by the regulated laboratory. date on regulations and changes to regulatory code. Lastly,
because the area of regulation is closely linked to the areas of
accreditation and quality standards (Chapter 16) and legal
Compliance Oversight
practices (Chapter 17), a review of these subjects provides
To ensure compliance with CLIA regulations, clinical labora- a more comprehensive view of regulation in the context of
tories that perform moderate- and/or high-complexity tests blood banking and transfusion practice.
Certificate of Waiver. This certificate is issued to a laboratory to perform only waived tests.
Certificate for Provider-Performed Microscopy Procedures (PPMP). This certificate is issued to a laboratory in which a physician,
midlevel practitioner, or dentist performs no tests other than the microscopy procedures. This certificate permits the laboratory
to also perform waived tests.
Certificate of Registration. This is a certificate issued to a laboratory that enables it to conduct moderate- or high-complexity
laboratory testing, or both, until it is determined by survey to be in compliance with the CLIA regulations.
Certificate of Compliance. This certificate is issued to a laboratory after an inspection that finds the laboratory is in compliance
with all applicable CLIA requirements.
Certificate of Accreditation. This is a certificate that is issued to a laboratory on the basis of the laboratory’s accreditation by
an accreditation organization approved by HCFA.
Process
Corrective action Validation ABO typing
Quality control Antibody screen that promotes an environment with minimal varia-
Audit Component preparation tion. Ensuring that SOPs exist and that they accurately
Labeling
Root cause Errors
Release into inventory
describe each procedure is a basic function of the QA
Accidents
Complaints Transport unit. Each SOP must be reviewed, signed, and indexed
Deviation Transfusion reactions Storage in a master copy of the SOPs. SOPs must be available
to employees who perform the tasks. SOPs must be
Product
updated promptly to reflect changes in processes, and
Compatibility testing
Crossmatch
these modifications must be appropriately documented.
Issue SOPs must exist for QA unit activities that define its role
Transport in reviewing, approving, and authorizing all SOPs.
Transfusion
2. Training and education. The QA unit should assist in
Patient developing, reviewing, and approving training and edu-
Figure 16–1 Continuous quality improvement.
cational programs for all personnel. This responsibility
includes new employee orientation; cGMP, SOP, and QA
training; and technical, supervisory, managerial, and
cannot be designed to be perfect. It is therefore necessary to computer system training. The QA unit is ultimately
implement an evolutionary refining process to detect devia- responsible for ensuring that personnel are appropri-
tions and to put in place appropriate corrective actions (Fig. ately qualified and trained to perform their tasks. Written
16–1). The FDA requires that blood establishments have in documentation of training must be on file.
place procedures for thoroughly searching for signs of devia- 3. Competence evaluation. The QA unit should imple-
tions in processes. These deviations may become apparent ment a formal, regular competence evaluation program
during validation of a new process, during quality control to ensure that staff maintain the skills to perform their
(QC) and quality assurance (QA) audits, or through occur- tasks. This program should include direct observation
rences such as errors, accidents, complaints, and transfusion of performance of routine and QC procedures, review
reactions. Establishments must have systems in place to inves- of worksheets, preventive maintenance records, and
tigate deviations, to determine their root cause, and to put written tests to assess theory and knowledge of SOPs.
in place corrective action. Central to this concept is the idea Remedial action and retraining should be documented
that staff and customers must be encouraged to report devia- in the personnel records.
tions; indeed, they should be encouraged to search actively 4. Proficiency testing. The QA unit should review and
for mistakes without fear of retribution. This view recognizes 16
monitor proficiency testing procedures and results to
deviations as opportunities to improve and to provide better ensure adequate evaluation of test methods, equipment,
services. Well-designed processes performed by well-trained and personnel competence. Proficiency testing should 233
staff should ideally not allow deviations. When they occur, it be performed by the same personnel who perform the
is critical to thoroughly evaluate the problem and the involved tasks routinely and should include backup or alternative
processes to ensure systemic quality improvement. testing (e.g., manual procedures performed during com-
puter downtime). There should also be a written plan for
The Role of the Quality Assurance Unit remedial action in the case of unsuccessful proficiency
testing performance.
Ensuring the safety of blood products requires the implemen- 5. Validation. The QA unit is responsible for ensuring
tation of effective control over manufacturing processes and that validation protocols are designed prospectively,
systems. In 1995 the FDA published guidelines that require performed, and evaluated, and that written validation
blood establishments to develop written QA programs with reports are prepared. Although validation of test meth-
an emphasis on error prevention rather than on retrospec- ods is a standard practice in the clinical laboratory, the
tive detection.18 QA is the sum of activities planned and extension of these principles to manufacturing pro-
performed to provide confidence that all systems and their cesses, computer systems, and facilities poses unique
elements that influence the quality of a product are function- challenges. Validation requires documented evidence to
ing as expected and can be relied on. QA functions include demonstrate that the system is performing as designed
QC procedures and audits, as well as ensuring that standards and with the expected degree of accuracy and repro-
are in place for facilities, personnel, procedures, equipment, ducibility. Complaints, errors, accidents, and problems
testing, and record keeping. at critical control points should be reviewed to deter-
All blood establishments are required to have a QA func- mine the need for revalidation, according to the FDA’s
tion, either as a QA unit in large establishments or a single Guideline on General Principles of Process Validation.19
individual in small transfusion services. Whatever the size, 6. Equipment and computers. Equipment, in particular,
the QA unit is required to be separate from the management should undergo installation qualification by the establish-
of production and to report directly to the responsible head ment (not the manufacturer). This is a form of validation
or the designated qualified person in charge of the operation. that establishes “confidence that process equipment and
The QA unit has final responsibility for the quality of prod- ancillary systems are capable of consistently operating
ucts released and must have the power to stop production within established limits and tolerances.” There should
and release of product if it deems necessary. be written procedures for equipment qualification,
BLOOD BANKING validation, maintenance, calibration, and monitoring. A investigation of complaints, errors, accidents, or adverse
major piece of equipment that is easily overlooked in the reactions. This is seen as a vital source of information
blood bank is the computer-based information system. on which continuous quality improvement efforts can
The FDA has recognized that computers play a central be focused through a system of corrective actions and
role in practically all blood bank processes and that the fine-tuning of existing processes.
risk of deviation caused by computer errors is high. For 8. Records management. The QA unit is responsible for
this reason, blood bank information systems are consid- ensuring that all records, including computer records,
ered medical devices in their own right and are directly are adequately stored and held for the appropriate peri-
regulated by the Center for Biologics Evaluation and ods as defined by the FDA. Systems of storage, espe-
Research (CBER) at the FDA, requiring 510(k) approval cially computerized systems, must be fully validated and
before being marketed. This requirement places a unique reviewed as necessary to ensure completeness.
burden on the manufacturers of blood bank software to 9. Lot release procedures. Each component released by a blood
demonstrate good software engineering practices and establishment represents one lot of product and bears a
to document and validate all aspects of their systems. lot number, usually the unit number assigned at the time
Furthermore, blood establishments are required to of collection. QA procedures must ensure that all records
validate all aspects of software on-site, an exhaustive pertaining to manufacture are reviewed for accuracy, com-
process that was initially described in an FDA draft guid- pleteness, and compliance with existing standards before
ance, although a finalized guidance document had not release. A second person should review the significant
yet been issued at the time of this writing.20 As with other steps. Labeling procedures, especially, are considered criti-
processes in the blood bank, there must be a system in cal control points and must be tightly controlled.
place to document complaints and problems with com- 10. QA audits. A QA audit is one mechanism for evaluating the
puter software and hardware, with documented investi- effectiveness of the total quality system. Comprehensive
gation, root cause analysis, and corrective actions. New audits should be conducted periodically in accordance
versions of software require revalidation. Many institu- with written procedures and should consist of a review
tions find that they require a dedicated computer person of a statistically significant number of records. On occa-
to perform these tasks despite the purchase of com- sion, focused audits may be conducted when quality
mercial software systems. The situation for homegrown problems have been identified or to monitor a particu-
computer programs has been less clear historically, as lar critical control point. Individuals conducting audits
these programs are not for commercial sale. Homegrown must have sufficient knowledge of and expertise in the
systems may be used without preapproval, but the FDA process under review but must not be responsible for
has made it clear that they expect the same standards to the processes being audited. There should be a written
apply to homegrown computer system design as to com- report documenting audit procedures and results; this
mercial vendors. Although FDA authorities do not rou- written report should include a review by the responsible
II tinely inspect the manufacturers of homegrown systems, head or other designated qualified person to evaluate the
they enforce the full weight of the law should major defi- results of the audit so that suitable corrective action can
234 ciencies in the computer system become apparent dur- be implemented. QA audits should be constructed using
ing routine blood bank inspection. a systems approach and may include review of donor
7. Error/accident reports, complaints, and adverse reactions. suitability, blood collection, manufacturing, product
Licensed establishments are required to report to CBER testing, storage and distribution, lot release, and com-
all errors and accidents that affect the safety, quality, puters. QA audits should evaluate critical control points
identity, potency, or purity (”SQuIPP”) of a blood prod- and key elements in each system, and each establishment
uct. The FDA has announced that nonlicensed estab- should customize its audits for its own systems.
lishments, including hospital transfusion services, must
report similar occurrences, now termed biological prod-
uct deviations (BPDs), if the product has left the con- RECALLS AND WITHDRAWALS:
trol of the establishment by the time the occurrence is PRACTICAL APPLICATION OF
detected.21 It is the role of the QA unit to investigate all THE CONCEPT OF SQuIPP
complaints, errors, and accidents to determine whether
there is a need to report each occurrence. Similarly, all Continuous quality improvement exercises aim to use all
adverse reactions must be fully investigated and docu- sources of information to identify BPDs, with a view to
mented, and recalls and withdrawals must be handled improving processes and systems. In the endeavor to iden-
according to established procedures. In particular, tify deviations, problematic products may be identified
transfusion reactions must be fully investigated, and in that have been released from inventory or even transfused,
cases of suspected bacterial contamination, a full review requiring a consideration of the safety of the recipient. If
of manufacturing procedures must be undertaken. the deviation from the standard was the result of an error
Fatalities that are related to either the donation process in the collection or processing of the unit by the collect-
or the receipt of blood products are reportable within ing facility, in violation of the laws overseen and enforced
24 hours to a special hotline at the FDA. This reporting by the FDA, especially violations of cGMPs, “and against
must be followed within a week by a full written report which the agency would initiate legal action,”22 the process is
describing the occurrence and its subsequent investi- called a biologic recall. Recalls may be conducted voluntarily
gation. This written report, in turn, usually triggers a on a firm’s own initiative, by FDA request, or by FDA order
focused FDA inspection. under statutory authority.23 All recalls must also be reported
As outlined earlier, an essential element of CQI is feed- to the FDA, regardless of who initiated the action. The FDA
back into the QA system of knowledge acquired through subsequently classifies recalls as follows: A Class I recall is a
QUALITY ASSURANCE, CONTROL AND IMPROVEMENT, AND ACCREDITATION
situation in which there is a reasonable probability that the carrying transfusion-transmissible disease, they are deferred.
use of or exposure to a violative product will cause serious Had the center known of the blood exposure at the time of
adverse health consequences or death. A Class II recall is a the earlier donation, the donor would have been deferred;
situation in which use of or exposure to a violative product therefore, components distributed from the donation 3
may cause temporary or medically reversible adverse health months ago must be withdrawn.
consequences, or where the probability of serious adverse Quality represents conformance of a product or process
health consequences is remote. A Class III recall is a situa- with preestablished specifications or standards. If it is dis-
tion in which use of or exposure to a violative product is not covered after the release of a leukoreduced component that
likely to cause adverse health consequences.22 These recalls the leukoreduction quality control for that component failed,
are then reported publicly by the FDA several months later.24 then the component must be recalled.
If large, multiunit recalls are excluded, the more common Identity indicates the need to ensure that the identifica-
blood center recalls occur typically at an approximate rate of tion of the donor, the unit, and its components is certain
1 in 5000 components distributed.24 throughout the collection, processing, and labeling steps. For
A market withdrawal occurs when a product has a minor instance, if a unit of red blood cells, acceptable in every other
violation that would not be subject to FDA legal action (i.e., way, is stated on the label to be negative for Kell (K1) when it
not an “actionable violation”). The firm removes the prod- is not, the unit is “misbranded,” and it could be harmful to a
uct from the market or corrects the violation. For example, a recipient whose serum contains anti-K1.
product removed from the market due to tampering, without Purity deals with freedom from extraneous contaminating
evidence of manufacturing or distribution problems, would matter in a product. If, after a recipient febrile transfusion
be a market withdrawal. For the blood center, withdrawals reaction, the residue from a unit of platelets is cultured and
are most often triggered by information provided by a blood is found to contain microorganisms, the other components
donor or a third party after donation that, if known at the from that collection should be withdrawn, because they may
time of donation, would have led to deferral of the donor. also have been contaminated.
An example would be the withdrawal, in the late 1990s, of Potency represents the ability of a product to produce the
blood products that were obtained from donors at risk of desired effect. Here, loss of potency typically indicates a fail-
Creutzfeldt-Jakob Disease (CJD), or variant CJD. The FDA ure of quality control. The findings that fewer than 75% of
became aware of a theoretical risk of transfusion transmis- platelets units have 5.5 × 1010 platelets and that red blood
sion of these ailments and advised that previously collected, cell recovery after leukoreduction of red blood cells is less
“in-date” products from donors at risk should be with- than the required amount represent a loss of potency. If these
drawn. Likewise, in the fall of 2002, the discovery that West products had been distributed, discovery of their deficiencies
Nile virus could be transmitted by blood products led to the would have led to a recall.
later, voluntary withdrawal of most frozen blood products
from certain localities collected during the period of highest Actions Taken by the Blood Center
risk of transmission, on the grounds that they might be con- 16
taminated. These products met all the required standards of When the collecting facility discovers that a product that does
blood safety at the time of manufacture, but they were later not meet SQuIPP criteria has been released, staff members 235
considered unsafe and required withdrawal. Blood compo- must work rapidly to notify the receiving transfusion service.
nent–related market withdrawals occur with approximately For in-date products that may still be in the hospital inven-
the same frequency as recalls.25 tory, this notification is done typically by telephone, to allow
In deciding whether postdistribution information war- immediate quarantine of the suspect product, followed by
rants action on the part of a blood establishment, it is help- written notification. The center must also quarantine any
ful to consider SQuIPP (Table 16–2). Any deviation from in-date products in its inventory. Because the information
established procedures or deferral criteria that would lead is typically obtained at the time of a subsequent attempted
to reduced safety in the product (i.e., greater potential for donation, platelet and red blood cell components are usually
harmful effects) generally leads to withdrawal or recall. outdated, but frozen plasma, frozen red blood cells, and cryo-
For instance, a presenting donor may state that he received precipitate may still be in the center or hospital inventory. The
a needlestick exposure to blood 4 months earlier, before a center in-date inventory must be identified and quarantined
donation 3 months ago. Because donors with exposures quickly. Hospitals must be notified of in-date components
to blood represent a group that may have a greater risk of as soon as possible, with written notification following soon
afterward. When the hospital returns suspect components,
the center usually destroys them or corrects the underlying
problem with appropriate documentation. If the postdona-
Table 16–2 SQuIPP tion information would lead to donor deferral in the future,
the center must place the donor in the deferral registry and
Use SQuIPP to help determine whether a product needs to must notify the donor of this action, to prevent collection and
be recalled or withdrawn distribution of any future donations. The center must docu-
Safety: does the product have a greater potential than
usual for harmful effects?
ment its actions and must seek from the consignee information
Quality: does the product conform to preestablished regarding the final disposition of components.
specifications or standards?
Identity: is the identity of the final product and its
labeling certain? Actions Taken by the Hospital
Purity: is the product free of extraneous contaminating Transfusion Service
matter?
Potency: can the product produce the desired effect? When notified that products are being recalled or with-
drawn, the hospital must act immediately to quarantine any
BLOOD BANKING such products in the inventory. The hospital must ensure the blood. The goal of test improvement has been to increase
that proper steps are taken and documented. When products sensitivity to the extent that the infectious window period
have been transfused (75% of recalls in one hospital were is reduced as much as possible. Because an infectious win-
found to be for components already transfused),25 the hospi- dow period still exists, and may have been longer in the past,
tal transfusion service must consider carefully the appropri- when a donor is found to have a newly acquired marker of
ate steps to take, in concert with hospital administration and an infectious disease, the more recent, previously seronega-
risk management staff. For some recalls or withdrawals, no tive donations of that donor must be regarded as suspect.31
further action may be necessary. For instance, when a unit The risk of a previous donation’s ability to transmit disease
that was incorrectly labeled as K1 negative was subsequently depends on the incidence of the disease and on the length
transfused to a recipient who had no anti-K1, no clinician of the infectious window period, the latter a function of the
or patient notification would be needed, merely a note for sensitivity of the test used. For HIV or HCV, using nucleic
the record of the facts. For other situations, the transfusion acid amplification testing, the infectious window period
service may wish to notify the clinician but advise him or may be a few days, at most.30 Notification of recipients of
her that patient notification is probably unnecessary. For previously donated seronegative units is important, so that
instance, red blood cells from a donor who had traveled to the recipients may be tested and, if they are positive for the
a malarial area and that were transfused 6 months earlier agent, counseled and treated appropriately.
would have caused transfusion-induced malaria within 50
days of transfusion, and if no febrile disease developed, these Actions by the Center and the Hospital
red blood cells would represent no additional risk.26 Finally,
in some cases, the patient should be notified, particularly The blood center must identify previously donated units at
when there is some risk of infectious disease transmission, risk from the implicated donor and must determine the com-
so that investigation, counseling, and treatment can be con- ponents made from these units and the consignees to whom
sidered. Patient notification may be required, for instance, these components were shipped. The center must then notify
for possible transmission of human immunodeficiency virus the hospital of the situation and must obtain information
(HIV) or hepatitis C virus (HCV).27 In each case in which a about the final disposition of the components. If in-date
component has been transfused, the transfusion service must components from such units may be available, the consignee
balance the evidence that an infectious agent was transmit- should be notified as soon as possible after the detection of a
ted through transfusion, the potential benefit to the recipi- repeat reactive screening test.
ent of receiving this information, and the potential harm to The specifics of the actions required vary by disease. For
the recipient when no diagnostic test exists that can resolve HIV and human T-cell leukemia/lymphoma virus (HTLV),
whether such transmission has occurred.28 If the transfusion the center searches for the first seronegative (or untested)
service physician finds the information sent to be insuffi- unit previously donated and for all units donated for at least 1
cient, sometimes more information can be obtained from the year before that donation.27 The center must identify all com-
II blood center. Fortunately, most recall and market withdrawal ponents made from the suspect units and must determine
notifications have little clinical significance.25,29 where they were sent. If any in-date components exist, these
236 must be identified and quarantined if they are in the center.
If these components have been distributed, the consignees of
LOOKBACK such components should be notified within a few days of the
positive test result. The hospital, in turn, must ensure that
When a donor unit is found to be positive for an infectious the patient is notified (typically through the clinician, but if
disease marker, several actions must be taken by the collect- necessary, by the hospital directly) within a few weeks. For
ing blood center. Obviously, the implicated unit must be HIV, even if the patient is deceased, the patient’s next of kin
destroyed, and this disposition must be recorded. In addition, must be notified.27 For HTLV, in contrast, the clinician can
donor-related actions are required (notification, counseling, make the decision not to notify the patient or the recipient,
and placement in a donor deferral registry), and action may if the decision is warranted clinically. Both the center and the
need to be taken to ensure that previous units donated by hospital must document notification attempts.
that donor have not transmitted disease. The latter process is For HCV lookback on a donor found to be seropositive,
termed lookback. the same process is used. The center must identify units 12
When determining the risk of transfusion transmission months before the last seronegative unit, locate the consign-
of disease, prevalence, measured for transfusion-trans- ees of components made from those units and shipped to
mitted diseases by an infectious disease marker, is usually the consignee, and notify the consignee within 45 days of the
of less concern than incidence, or the new acquisition and repeat reactive donation. The consignee must ensure that the
development of disease, because blood with a marker indi- patient is notified of the need for HCV testing and of the
cating prevalence is usually readily identified by testing and availability of testing and counseling. There must be docu-
is discarded. For a newly acquired, incident viral agent, the mentation of three attempts to notify the recipient within 12
concern is about transmission of disease during the window weeks of notification by the center.32
period, defined as the time between donor infection and A massive retrospective HCV lookback of units donated
the appearance of detectable markers of disease, a time in between 1990 and 1999, on which lookback had not already
which current tests may not identify an infectious unit. The been performed, was begun in 1999 and was completed by
window period may be conveniently divided into two parts: 2001. This targeted (identification of specific blood recipi-
an eclipse period, when the agent is undergoing prolifera- ents) lookback process may not be very cost effective. One
tion in host tissue but is not present in sufficient numbers estimate is that of approximately 300,000 recipients to be
in the blood to be infectious,30 and the subsequent infec- notified, perhaps only 5000 to 10,000 will be living and
tious period, when circulating virus can be transmitted by will have newly recognized infections, and 1500 will ben-
efit from therapy.33 Interim data from the U.S. Centers for
Today’s transfusion medicine and cellular therapy profes- These causes of action are generally based on at least three
sionals are challenged with striking an appropriate balance different theories of liability: (1) negligence, (2) strict liabil-
between optimizing and maintaining the safety of blood ity, and (3) breach of implied warranty. Appreciating the
donors and recipients and effectively managing the legal distinctions among the applicable theories is important to
risks inherent in the operation of blood-collection and minimize the risk of liability.
blood-transfusion facilities. This chapter provides health
care professionals with an overview of the growing number
of legal issues that affect the transfusion medicine and cellu- NEGLIGENCE
lar therapy communities and offers suggestions as to how to
mitigate the associated risks in order to better serve donors Although the law of negligence varies significantly from
and patients. state to state, the basic elements of a negligence claim are
Because of the evolution of the law in these areas, more as follows: the defendant owed a duty of care to the plain-
cases relate to the collection and transfusion of blood and tiff; the defendant breached the duty; the plaintiff ’s injury
blood components than to the collection, processing, stor- was directly or proximately caused by the breach; and the
age, and administration of cellular therapy products. For this plaintiff suffered damages as a result.1 In short, the theory
reason, this chapter focuses on cases relating to the transfu- of negligence holds that the provider of medical services is
sion of blood and blood components. As a practical matter, responsible for providing safe products and services that will
II however, many of the legal issues that have affected transfu- not harm the patient and that are consistent with the prevail-
sion medicine will similarly affect the cellular therapies field. ing standards of care.
238 Although the cases discussed in this chapter are intended Historically, negligence was grounded in fault-based liabil-
to be illustrative of legal concepts, they are not intended to ity: the negligent actions had to be proven to cause the injury.
replace the advice of competent legal counsel. Beginning with the Restatement (Second) of Torts §402A
(1964), a new basis for liability was widely adopted largely to
hold manufacturers accountable for poorly designed products.
LEGAL RISKS IN TRANSFUSION This was termed strict liability or liability without fault.2 To find
MEDICINE a manufacturer strictly liable, one need prove only that the
injury was due to the product’s design, regardless of whether
the manufacturer was at fault. The impetus for the adoption
Theories of Liability
of the theory was to encourage manufacturers to make safer
A plaintiff complaining that he or she was wrongly injured products by requiring them to absorb the real costs of the
by a transfusion typically looks to include the greatest num- consequences of the unsafe product. However, recognizing
ber of defendants, based on as many causes of action as pos- that blood is a living tissue, inherently variable and incapable
sible. For example, defendants may include clinician(s) who of being rendered uniform or completely safe, virtually every
transfused blood and treated the patient, the hospital where state has adopted special “Blood Shield Statutes” exempting
the transfusion and treatment were provided, the blood cen- blood from strict liability standards. Even in jurisdictions that
ter that collected the blood, and possibly the organization have not yet adopted blood shield statutes, courts can decide
that set the applicable standards. that strict liability should not apply to blood collection and
The claims may include failure properly to warn of the storage. A notable case is described in Box 17–1.
risks of treatment or surgery, failure adequately to discuss
alternatives to transfusion such as autologous or directed
donation, failure adequately to warn of the potential risks Elements of Negligence
inherent in transfusion, negligent hiring and supervision
Duty of Care
of staff, failure adequately to test the donor for transfu-
sion-transmitted disease, failure to develop or implement an The first element in a successful negligence claim is establish-
appropriate “lookback” program for notifying recipients of ing that the defendant owed a duty of care to the individual
infected blood, negligent standard setting, failure to trans- who is claiming harm. Traditionally, in cases involving the
fuse at the proper rate, and negligently administering trans- provision of health care services, a doctor-patient relation-
fusions that were not medically necessary. ship between the defendant and plaintiff must be established
LEGAL PRINCIPLES AND ISSUES CENTRAL TO TRANSFUSION MEDICINE
BOX 17–1 Illustrative Case BOX 17–3 Illustrative Case
Nestor v. Hospital Pavia3 involved a patient who was transfused In Smith v. American Red Cross,8 a blood-transfusion recipient
during August 2001 and was later diagnosed with hepatitis C. who was infected by AIDS-contaminated blood in April 1984
The patient alleged that both the local blood bank and the hos- and subsequently died, brought an action claiming, among
pital were negligent as they provided contaminated blood and other things, that the American Red Cross (Red Cross) did not
therefore were strictly liable for marketing a defective product, promptly notify the recipient that the donor of the blood with
despite the performance of the most sensitive, specific, licensed, which the recipient was transfused tested positive for human
and unlicensed tests available at that time. The court decided to immunodeficiency virus (HIV). Rejecting a charge that Red
follow the vast majority of states with blood-shield statutes and Cross should have notified her immediately after the donor
determined that, as a matter of law, strict liability should not tested HIV positive in August 1985 instead of waiting until
apply to blood collection and storage. it developed its “lookback” program in July 1986, the court
pointed out that the plaintiff had suffered no damages resulting
from the delay because no allegation was made that the recipi-
ent spread the disease to anyone else, and, in 1985, no early-
before a duty of care may be attached, although courts are treatment protocols existed for those infected with HIV.
not unanimous in defining the parameters of the duty of
care. Courts in a few states, including Arizona, Maryland,
Mississippi, Ohio, and Washington, have broadened the
duty of care between physicians and patients to situations in standard setting organization (American Association of
which no traditional doctor-patient relationship exists. Blood Banks; [AABB]) and a plaintiff who had contracted
One of the most striking cases illustrating this trend is human immunodeficiency virus (HIV) from a transfusion,
Stanley v. McCarver4 detailed in Box 17–2. Citing cases in the even though no direct relationship was found between the
District of Columbia,5 Michigan,6 and New Jersey,7 the Arizona parties. The court concluded that AABB owed a duty of care
Supreme Court noted that “[t]he requirement of a formal- to an individual treated by one of its voluntary members
ized relationship between the parties has been quietly erod- because the association had voluntarily assumed responsi-
ing in several jurisdictions.”4 The court held that, in deciding bility for the nation’s blood supply by inviting blood banks,
whether a duty exists with no formal doctor-patient relation- hospitals, physicians, and the public to rely on its Standards.
ship, it is appropriate to consider “whether the doctor was in At the time of this writing, this logic has been adopted in
a unique position to prevent harm, the burden of preventing only one other jurisdiction.10
harm, whether the plaintiff relied upon the doctor’s diagnosis In general, the majority of courts require some greater
or interpretation, the closeness of the connection between the connection or relationship between the health care provider
defendant’s conduct and the injury suffered, the degree of cer- and the recipient of health care services. One court, refus-
tainty the plaintiff has suffered or will suffer harm, the skill or ing to impose liability on AABB as a standard setting orga-
reputation of the actors, and public policy.”4 nization, specifically recognized that “adverse consequences 17
In the blood-transfusion setting, some courts have rec- to the public by chilling scientific and medical debate on
ognized a duty in even more attenuated relationships, such important issues and leaving these matters to the often slow 239
as between the blood provider and the recipient of the and cumbersome processes of governmental agencies” sig-
blood. Examples of this duty arise in “lookback” cases, which nificantly weighed against finding AABB liable under the
involve the notification of recipients of components from circumstances presented.11
donors who have previously tested positive for an infectious
agent. Once a blood collector becomes aware that one of its
Breach
donors has a disease transmissible by blood transfusion, the
provider has a duty to develop and implement an appropri- Once a plaintiff establishes the existence of a legal duty, the
ate program to notify the recipient of any component that plaintiff must prove that the defendant breached the stan-
might transmit the disease. An important example of this dard of care. The breach may be a failure to diagnose, a delay
attenuated level of relationship is described in Box 17–3. in diagnosis, improper treatment, a negligent rate of infu-
In Snyder v. AABB,9 a New Jersey court went to great sion, failure to obtain informed consent, and/or inferior care,
lengths in finding that a duty of care existed between a including substandard surgery or any number of other pur-
ported acts of malfeasance. A review of transfusion-related
case law reveals that courts apply three distinct standards
of care: medical negligence, ordinary negligence, and pro-
BOX 17–2 Illustrative Case fessional negligence. Subtle but important differences exist
among these standards, and the outcome of an individual
In Stanley v. McCarver,4 a nursing home contracted with a radi- lawsuit is often based on which standard is applied.
ologist to read chest radiographs of prospective employees of Medical Negligence: The standard of care applicable in a
the nursing home to screen for tuberculosis. The radiologist
medical negligence action requires that a physician exercise
never met with the plaintiff but noted a spot on her chest radio-
graph and, in his report to the nursing home, recommended
the degree of knowledge and care ordinarily possessed and
further evaluation. Although the nursing home’s procedures exercised by other members of the profession acting under
called for notifying a prospective employee of radiographic similar conditions and circumstances. Expert testimony is
results within 72 hours of receiving the report, the plaintiff was generally required to establish medical negligence.
never notified of the defendant’s report. Relatively soon there- Ordinary Negligence: Ordinary negligence is a standard of
after, she was diagnosed with lung cancer and ultimately died care that can be assessed by a “reasonable man of ordinary
a few years later. prudence.”12 Thus, ordinary negligence cases do not require
expert testimony. In general, plaintiffs prefer to frame their
BLOOD BANKING cases as ordinary negligence actions rather than as medi-
cal negligence cases specifically to avoid the added burden BOX 17–4 Illustrative Case
of introducing expert testimony. This approach allows the In 2003, a New Jersey court awarded $300,000 in damages to
defendant’s actions to be judged by the finder of fact, often a patient who, after a blood test for human immunodeficiency
a jury, who measures the actions against what a reasonable virus (HIV), was incorrectly informed by a doctor that he was
person would do in similar circumstances. Very few cases HIV positive.17 The patient brought a medical malpractice
involving donor screening or donor testing have been suc- action against the doctor and hospital where the patient was
cessfully asserted as a matter of simple negligence, although treated, claiming that, as a result of the misdiagnosis, he became
other blood center and transfusion service activities have depressed and suffered from physical and psychological injuries.
been judged against the ordinary negligence standard. The court found that, although the hospital was not negligent,
Professional Negligence: In the blood banking and trans- the doctor had deviated from the standard of care by failing to
fusion setting, most jurisdictions view the production and give the plaintiff pretest and posttest counseling, by misinter-
preting the test results, by incorrectly advising the plaintiff that
safeguarding of blood as a professional activity to be judged he was HIV positive, and by giving the results over the telephone
against the actions of a “reasonable professional” (the pro- rather than informing the plaintiff in person.
fessional standard) as opposed to a reasonable person, as
in ordinary negligence cases mentioned earlier. In its most
traditional form, the professional standard requires that an and humiliation. In determining the amount of damages, it
expert offer testimony on the prevailing standard of care in is appropriate to consider past, present, and future economic
the field. This evidence typically includes government regula- and noneconomic damages.
tions and applicable private standards and guidelines. Under Plaintiffs sometimes claim that their fear of contract-
this same standard, a single expert may not second-guess an ing a disease is a compensable injury.15 These types of cases
entire profession.13 Some courts, however, have broadened arose originally in the asbestos context, and plaintiffs have
the evidence that can be introduced to establish the prevail- had mixed success expanding this cause of action into the
ing standard of care, including evidence about practices of transfusion arena. Pennsylvania, for example, does not allow
other hospitals and transfusion services. This is most likely monetary damages for an asymptomatic plaintiff ’s fear of
to occur in situations in which the standard of practice in the contracting acquired immunodeficiency syndrome (AIDS)
field is not settled. For most professional negligence cases, in the absence of actual exposure to the disease.16 The oppo-
the less consensus that exists within the medical community site occurred in New Jersey, as described in Box 17–4.
concerning a specific course of action, the greater the oppor-
tunity for divergent outcomes in litigation.
Causation
CONTRACTUAL CAUSES OF ACTION:
IMPLIED WARRANTY
The third element of negligence is establishing causation, in
II other words, determining that the defendant’s breach of the Some plaintiffs apply a broad approach to litigation and sue
duty of care was the actual or proximate cause of the plain- not only on negligence theories, but also on theories founded
tiff ’s injury.14 In the transfusion setting, this means that the in contract. Typically, such claims allege that the physician
240
plaintiff must prove that it is more likely than not that the promised certain results from procedures or treatments, the
blood transfusion caused some form of injury, for example, patient is unsatisfied with the results, and the physician’s fail-
HIV infection. Where the hospital, physician, or transfusion ure to produce the promised results gives rise to an action for
service can show that the plaintiff engaged in high-risk activ- breach of contract or breach of warranty.
ities before the transfusion or that the particular morbidity In one such case, a blood bank that supplied the wrong
or mortality could have been caused by a different source, it type of blood, resulting in a patient’s death, was held not
is more difficult to satisfy the element of causation. liable under the doctrine of implied warranty, because the
When considering causation, the concept of foreseeability transfused blood was not in any way unwholesome or defec-
is relevant to determining whether certain actions or inac- tive.18 The court reasoned that the patient’s death resulted
tions constitute negligence. When the manner in which an from the negligence of an employee who recorded the wrong
injury occurs is so improbable or unpredictable that the blood-type number on the decedent’s card. In the court’s
defendant could not have “foreseen” it, the injury is not opinion, supplying blood for a fee constitutes the provision
actionable. As cases discussed later in this chapter illustrate, of a service, rather than the sale of a product and, accord-
the more foreseeable an untoward outcome is or was, the ingly, breach of warranty did not apply.
greater the potential exposure to liability. The best way to prevent breach of contract or breach of
Damages warranty claims is to avoid promising specific results; man-
aging patient expectations is not only sound practice, but
The final element of negligence is proving that the plain- also can help to protect against certain causes of action.
tiff has suffered actual damages; a patient must have been
injured as a result of the breach of the standard of care. The
proper measure of damages is the amount that will compen- DONOR SCREENING, DONOR TESTING,
sate the plaintiff for the injury proximately caused by the AND COMPONENT PROCESSING
defendant. These are called compensatory damages and are
further subdivided into economic and noneconomic dam- In the field of transfusion medicine, most legal cases brought
ages. Possible economic damages can include lost wages and against blood centers, hospitals, and physicians involve chal-
medical expenses. Noneconomic damages are subjective and lenges to donor-screening and infectious disease–testing
include pain and suffering, physical impairment, emotional practices. Although blood center and transfusion medicine
harm, inconvenience, loss of society and companionship, professionals regard donor screening and, in particular,
LEGAL PRINCIPLES AND ISSUES CENTRAL TO TRANSFUSION MEDICINE
donor questioning as relatively noninclusive and nonspe- BOX 17–5 Illustrative Case
cific for mitigating the risk of transfusion-transmitted dis-
ease, plaintiffs have often made this issue the focus of their In Boland v. Montefiore,14 a patient diagnosed with myelogenous
complaints. In general, donor-screening complaints have leukemia underwent a bone marrow transplant and was told by
centered around alleged “failures” in the donor-screening his treating physician that he required outpatient transfusions
process, notably either the failure to ask a donor a specific to increase his platelet count. The hospital blood bank direc-
question or a failure to rely on direct questioning to screen tor responsible for the management of outpatient transfusions
performed in the blood center was present in the blood bank
the donor for a particular risk.
during the transfusions. The first two units were transfused
Infectious disease–screening testing cases most often without incident; however, during transfusion of the third
revolve around the failure to implement an available test, unit, the patient experienced shortness of breath. Although all
typically a test that is a “surrogate marker” for populations appropriate actions were taken, including stopping the transfu-
at risk for transmitting a specific agent but that does not sion and providing emergency care, the patient suffered cardiac
directly identify the infectious agent. The period between the arrest and died. The patient’s estate filed suit against the blood
emergence of infectious disease and the development of a center for negligent screening and testing of the units and fail-
test for the agent responsible for transmission of the disease ure to prepare properly the blood administered to the plaintiff.
often presents the greatest risk of liability for a defendant. The treating and transfusing physicians, as well as the hospital,
Additionally, in these same cases, the closer the injury is tem- were sued for conscious pain and suffering, wrongful death, and
failure to obtain informed consent.
porally to the implementation of testing for the agent, the
greater the risk of liability. This concept is best illustrated by
the early history of HIV and blood transfusion. The emer-
gence of the first incidences of HIV related to blood transfu-
sion occurred in 1981,19 and the licensure of the first HIV test decedent’s “treatment was consistent with good and accept-
occurred on March 2, 1985.20 This several-year period gave able medical practice.”
rise to a significant number of transfusion-transmitted (TT)- In other jurisdictions, however, the question of a defen-
HIV cases aimed at standard-setting organizations, blood dant’s negligence concerning donor screening or testing
centers and hospitals, transfusion medicine professionals in has been judged under a different application of the pro-
both blood centers and hospitals, and clinicians involved in fessional negligence standard. Specifically, the highest state
the patients’ care. Courts have come to different conclusions court in Colorado overturned a verdict rendered in favor of
about the liability of these parties for alleged deficiencies in the blood center and held that a blood collector’s compli-
donor screening and blood-donation testing to prevent the ance with governmental regulations and the standards and
transmission of disease through blood transfusion. guidelines of a professional association was not conclusive
In most jurisdictions, courts have applied the professional proof that additional precautions were not required, because
negligence standard and required only that the actions of some other blood centers were using “additional precaution-
blood-collection facilities be measured against the conduct ary measures.”25 When evidence concerning the appropriate 17
of other members of the defendant’s profession, through the standard of care is broadened to include not only the tes-
testimony of expert witnesses. In those jurisdictions, blood timony of expert witnesses but also the practices of other 241
collectors whose donor-screening and donation-testing individual blood centers, courts have determined that suffi-
procedures conformed to the standard of care in the profes- cient conflicting evidence exists about the standard to create
sion and, in particular, the recommendations of standard- a question of fact to be determined by a jury.
setting organizations such as AABB, the Food and Drug Even if a traditional professional negligence standard is
Administration (FDA), and the Centers for Disease Control applied, once the testimony of experts creates a question of
and Prevention (CDC) have not been held responsible for material fact for the jury, the specific instruction to the jury
transmission of a disease. In these cases, courts reject alle- on the standard becomes exquisitely important in determin-
gations of alleged failure to screen donors13,21 and failure to ing the outcome of the case.
implement tests not yet adopted by standard-setting organi- In Ray v. American National Red Cross (ANRC),26 the
zations or mandated by government agencies.22,23 plaintiff argued that the ANRC was obliged to act as a reason-
This articulation of the professional negligence standard able blood bank with the knowledge and skill level similar to
was most recently applied in a case involving transfusion- that of the ANRC, whereas the ANRC contended that it had
related acute lung injury (TRALI). Although not an infec- only to act according to industry standards. In applying the
tious disease risk, TRALI is, as of today, the most commonly ordinary negligence standard, the appellate court reasoned
reported cause of transfusion-related death in the United that, “particularly where the Red Cross is in a position to
States, surpassing deaths caused by ABO incompatibility substantially establish the industry standard, it may not use
and bacterial contamination.24 Unfortunately, because this that standard as a safe harbor to insulate its activities from
clinical syndrome remains poorly defined and its pathophys- scrutiny if competent evidence demonstrates that it would
iology is not understood, the medical community lacks con- have been reasonable for the Red Cross to adhere to a higher
sensus as to the appropriate way to prevent the syndrome. An standard or a different practice.”26
interesting and illustrative case is presented in Box 17–5. As a practical matter, the outcome of any case that chal-
The court specifically noted that the plaintiff ’s malprac- lenges donor-screening or -testing practices will largely
tice contentions with respect to the blood center were that the depend on the specific facts presented, including the specific
blood products administered to the patient should have been time and circumstances of the harm or injury. The Boland
screened for antibodies and that the entire blood-banking case (see Box 17–5), which involved a patient who died of
community was negligent for failing to follow this practice. TRALI and allegations of a failure to screen donors in 1995,
The court dismissed the case, finding that the plaintiff had was successfully defended.14 Whether that same case could be
failed adequately to dispute the defendants’ proof that the successfully defended today is uncertain, as several options
BLOOD BANKING for mitigating the risk of TRALI have recently been presented What is not as clear is the effect of divergent practices on
and discussed at a consensus conference.27 Although the a defendant’s ability to avoid a risk of litigation. The increas-
panel stopped short of recommending that any specific inter- ing trend toward the use of single-donor platelets, the debate
vention practice be adopted, some major blood-transfusion over the necessity of testing for aerobic as well as anaero-
systems, including the National Blood Service in the United bic bacteria, and the discrepancies in the relative predictive
Kingdom, have implemented practices to mitigate the risk of value of available test methods33 may constitute grounds
TRALI.28 As practices in the profession change, so too does for negligence claims. The true measure of potential legal
the standard of care against which facilities and individuals exposure, however, will be the incidence and severity of the
are judged. adverse reactions relating to the transfusion of bacterially
A few of the changing standards of practice are discussed contaminated units. Current anecdotal survey data and data
later. from some larger facilities suggest that morbidity and mor-
tality directly relating to failures to implement two-bottle
culture systems, eliminate or reduce the use of pooled plate-
CHANGING STANDARDS OF CARE IN lets, or substitute more sensitive and specific testing will be
TRANSFUSION MEDICINE–RELATED limited.
CASES
West Nile Virus and Variant
Bacterial Contamination Creutzfeldt-Jakob Disease
Prior to 1995, virtually no filed cases involved negligence in The transmission of West Nile virus (WNV) or variant
testing or screening to prevent transfusion-associated mor- Creutzfeldt-Jakob Disease (vCJD) through blood transfu-
bidity and mortality from bacterial contamination of platelet sion is often proffered as a source of potential liability for
or red blood cell units. One of the few reported cases of bac- blood collectors, hospitals, and transfusing physicians. These
terial contamination calls into question not the failure to test, two threats and the blood community response to each are
but the failure to provide directed donations as an alternative illustrative of two separate approaches to mitigating trans-
to conventional transfusions. In 1995, a woman who received fusion risk. As a practical matter, however, the blood com-
a transfusion of two units of packed red blood cells in the munity’s rapid and system-wide response to both of these
course of coronary bypass surgery developed sepsis and died threats mitigates the risk of liability for blood-transfusion
the day after her surgery.29 It was later determined that her professionals.
death was caused by an extremely rare blood-borne bacterial The management of the WNV threat to blood safety
infection. The woman’s family claimed that the medical facil- serves as a model for effective threat containment.
ity was negligent in refusing their request for directed dona- Beginning with the first diagnosis of potential TT-WNV
tion of blood and that this refusal was the proximate cause in 2002,34 through the implementation of donor screen-
II of death. The judge awarded the woman’s family $400,000; ing35 and inventory management to reduce the risk of dis-
however, the medical facility appealed, and the finding was tribution of infected units36,37 to the implementation of
242 reversed. The appellate court specifically determined that pooled38 and then individual nucleic acid testing (NAT)
the applicable standard of care should reflect the community in high-incidence locations,39 the response of the govern-
standard of care, including FDA and/or AABB recommenda- ment and the private sector was swift, coordinated, and
tions and standards, not just the hospital’s internal policies. evidence based.40 Although a number of blood recipi-
Then, in 1997, the family of a man who died of trans- ents were infected with WNV through transfusion, the
fusion-associated sepsis brought suit against a hospital in broad-based development of national recommendations
Providence, Rhode Island, and was awarded $5.6 million.30 for donor screening and inventory management, as well
Although the plaintiff alleged that the hospital was negli- as the rapid development and implementation of testing
gent in failing to prevent transfusion of the contaminated for the virus, established a clear national standard of prac-
unit, the jury appeared to base its award on the alleged fail- tice against which blood-center and transfusion-service
ure of the hospital to recognize and treat the infection. With response could be measured.
the implementation of AABB Standard 5.1.5.1,31 requiring The management of vCJD as a risk of blood transfusion
that blood collectors and transfusion services have in place has proven to be more difficult. The primary method for
a method both to limit and to detect bacterial contamina- controlling the risk of vCJD has been the implementation
tion in platelet components, undoubtedly increased atten- of geographic and time-defined donor exclusions.41 As a
tion will be focused on adverse reactions in patients related method for mitigating risk, these deferrals, which are based
to undetected bacterial contamination. in part on scientific data and in part on the need to maintain
Given evolving technologies, ongoing determination of public confidence in the safety of the blood supply, are nei-
residual risk, and the lack of uniform methods for meet- ther sensitive nor specific. Yet, with the publication of FDA
ing AABB Standard 5.1.5.1, the possibility of exposure to a guidance on the issue42 and the adoption of donor-exclu-
lawsuit for injuries relating to death or injury resulting from sion criteria in other countries, it is unlikely that a plaintiff
bacterial contamination is real. Bacterial contamination of who contracted vCJD through blood collected and issued
platelets has been one of the greatest transfusion-transmit- by a facility that followed these recommendations could
ted infectious risks in the United States, and is now signifi- successfully argue negligence.
cantly higher than the risk of transfusion-transmitted viral The primary reason that a case alleging the negligent
infection.32 Although lack of consensus appears to be present transmission of vCJD through blood transfusion is likely to
as to the best way to eradicate the risk, transfusion-medicine fail is the difficulty of proving causation, and, specifically,
professionals clearly now have a duty to try to minimize the associating transmission of the disease with a particular unit,
risk through available means. as no specific test exists for the pathologic prion at the time
LEGAL PRINCIPLES AND ISSUES CENTRAL TO TRANSFUSION MEDICINE
of this writing. In addition, given the extremely low preva- BOX 17–6 Illustrative Cases
lence of vCJD in the human population, it is evident that
only a small number of cases, if any, will occur. McDonnell v. American National Red Cross51 revolved around
a donor who fainted immediately after blood donation while
walking unescorted from the donation table across a stone floor
Component Processing to the snack table and suffered a concussion and a postconcus-
In the blood-transfusion field, significant discussion has sion syndrome. The court, in an unpublished opinion, adopted
ordinary negligence as the applicable standard of care rather
occurred over the potential for increased liability relating
than medical malpractice, because the complaint took issue
to the failure to adopt certain component-processing steps. with the general safety and operation procedures surrounding
As an example, increasing consensus is found that certain the blood donation—facts within the common knowledge and
patients will benefit from receiving leukocyte-reduced red experience of the jury—and did not raise questions requiring
blood cells. The failure to prescribe leukocyte-reduced red medical judgment.
cells for patients who have previously experienced febrile non- On the contrary, Baumann v. American National Red Cross52
hemolytic transfusion reactions, those who need cytomega- involved a donor whose right median nerve was injured dur-
lovirus (CMV)-safe units, or those who need to have their ing phlebotomy. The plaintiff sued the American National Red
risk of human leukocyte antigen (HLA) alloimmunization Cross under an ordinary negligence theory, alleging failure to
limited can create a potential risk of liability for a transfus- warn of the potential for injury to donors during phlebotomy
and negligence in the phlebotomy procedure. The court held
ing physician. Conversely, the obligation to provide leuko-
that the claim should have been brought under the state’s medi-
reduced units to every patient (universal leukoreduction) cal malpractice statute, reasoning that the standard of care gov-
has not been clearly established. Conflicting, underpowered, erning these types of claims clearly involves medical learning or
and inconclusive studies,43–46 as well as the very public and principles and is not within the knowledge of most laypeople.
documented discussion over the cost/benefit ratio of leuko-
reduction,47 suggest that similar debates over leukoreduction
would be repeated in any litigation asserting the obligation
to provide leukoreduced units to all patients. In addition,
because many of the asserted benefits of leukoreduction are reactions at the donation site, and to provide donors with
difficult to capture in a single episode (e.g., the prevention postdonation care and contact information. Most facilities
of immunomodulation or the transmission of viruses not also have adopted a policy of paying for immediate emer-
invariably associated with symptomatic disease) and because gency care in the event an injury occurs.
many of the risks of not using leukocyte-reduced units can Such preventive measures are critically important because,
be masked or corrected through the administration of other as previously noted, areas of activity in blood collection and
drugs (e.g., acetaminophen for fever reduction), both causa- transfusion services may be analyzed under principles of ordi-
tion and the specific injury to the plaintiff would likely to be nary negligence, as opposed to the professional negligence
difficult to establish. standard. Instructive examples are described in Box 17–6. 17
The utility of irradiation of blood components is another Although it is difficult to predict which standard will be
area in which consensus has not yet developed. Most physi- applied in all situations, a blood-collection facility will be 243
cians recognize the utility of blood-component irradiation to best protected by an annual review of its policies and pro-
prevent graft-versus-host disease (GVHD) in certain at-risk cedures against what a reasonable donor might understand
patients.48 The literature strongly supports the conclusion and expect.
that patients with hematologic malignancies, patients with
congenital immunodeficiencies, patients who are receiv-
Informed Consent
ing allogeneic bone marrow transplants, and patients who
receive intrauterine or exchange transfusion should receive The doctrine of informed consent has its origins in the legal
irradiated blood components (see Chapter 28). However, theory of battery or nonconsensual touching.53 Although this
some documented medical cases exist in which the imple- legal theory is sometimes referenced in case law involving the
mentation of universal irradiation would have prevented the failure to obtain informed consent, more recent cases base
death of patients not known to be at risk.49,50 Although some the legal theory on negligence.54 Whichever theory serves
in the transfusion-medicine community have advocated the as the basis for the analysis, it is clear that the ethical prin-
adoption of universal irradiation of blood components, oth- ciple of autonomy, or the individual’s right to make choices
ers debate whether the high severity and low frequency of about personal matters, is well ingrained in the legal system.
these adverse events warrant the implementation of univer- Multiple cases in virtually every jurisdiction reinforce the
sal irradiation. Nonetheless, facilities must keep abreast of legal obligation to obtain informed consent for most medical
the growing body of literature on patients who may benefit treatments or interventions, including blood transfusion.
from receiving irradiated units. With the exception of federal regulations applicable
to patients participating in research protocols, the law of
Donor Injury informed consent is governed by state law, which varies from
jurisdiction to jurisdiction. Some states, notably California,
With an overall frequency of donor complications reported New Jersey, and Pennsylvania, have enacted statutes that
as approximating 20% on average and including severities of delineate specific steps that must be taken in the informed-
reactions from bruising to death, blood-collection facilities consent process. In other states, informed-consent require-
have policies and procedures in place to minimize the risk of ments have been developed through case law.
both donor injury and litigation resulting from injury. Best Failure to obtain informed consent also has become an
practices require a facility to continually update the informa- issue in clinical trials. In a relatively recent case, Wright v.
tion in the consent for donation, to have a plan for adverse Fred Hutchinson Cancer Research Center,55 patients and their
BLOOD BANKING families brought a class action against the hospital and four Courts in several jurisdictions have relied on additional
physicians after 80 of 82 patients receiving care under a cer- theories of liability, including the “corporate negligence the-
tain protocol, designed to reduce the risk of GVHD, died. ory,”64 under which a hospital may be liable for failing prop-
The plaintiffs claimed that the center failed to inform them erly to oversee the treatment of patients or require physicians
adequately about the risks inherent in the T-cell–depletion with hospital privileges to obtain informed consent. Another
protocol. Although the case never went to trial, apparently theory is “respondeat superior,” under which an employer is
because of the inability to certify the class, one of the evi- responsible for its employees and, particularly, has a duty to
dentiary rulings precluded summary judgment in favor of exercise reasonable care in selecting, retaining, and supervis-
the cancer center, creating the inference that the center could ing medical staff.65 In addition, an institution that voluntarily
have been found liable for failure to inform. assumes this duty, but fails effectively to obtain the consent,
As informed consent for blood transfusion has become the can be held liable for injury relating to those failures, even in
standard of practice, the elements of effective informed con- the absence of an original duty to obtain consent.
sent have become more standardized. The process includes Separate from any specific duty to obtain informed con-
(1) disclosure of the risks and benefits of transfusion or sent, hospitals do have a duty to educate their physicians
related therapies, (2) presentation of the potential alterna- about risks, including the risks of transfusion. Hospitals are
tives, (3) an opportunity to ask questions of learned profes- independently responsible for providing information suffi-
sionals,56 and (4) documentation of the consent. Although cient to allow a treating physician to carry out his or her duty
these elements are straightforward, identifying which risks to obtain informed consent. In the transfusion setting, this
must be disclosed has been the subject of significant case law means that hospitals must have information available to phy-
and has resulted in the development of two different stan- sicians on transfusion risks, as well as available alternatives to
dards. The first and older standard requires the disclosure allogeneic transfusions.66
of risks that are considered to be material to the reasonable
physician.57 This standard, however, has been replaced in a Alternatives
majority of jurisdictions by a newer standard, which requires
the disclosure of risks that would be considered to be mate- One of the elements of informed consent is the obligation
rial to a reasonable patient.58 to advise the patient of alternatives to the recommended
The adoption of the “reasonable patient standard” has therapy. This particular element raises some interesting
introduced additional issues for transfusion-medicine pro- questions about the obligation to provide information about
fessionals. Specifically, (1) the relevant risks must be dis- alternatives to conventional transfusion therapy. Courts
closed, (2) the disclosure must be in language understandable have had numerous opportunities to consider the liability of
to the patient, and (3) a new informed-consent discussion is physicians for failing to recommend alternative procedures,
required if a substantial change in circumstances, whether such as autologous and directed donations, as well as other
medical or legal occurs.56 alternatives to allogeneic transfusion. This complex issue is
II based primarily on two questions: the standard of practice
Who Has the Duty to Obtain in the community (including the availability of the alternate
244 therapy), and the hospital or blood bank’s adherence to its
Informed Consent?
own policies and procedures.
Courts have not been consistent in deciding who specifi- In one case, a California court held that a blood center
cally has the obligation to provide the information and to could not be held liable for failing to disclose the option
obtain informed consent. Generally, the obligation to obtain of a directed-donation program because the plaintiff had
informed consent has been considered to be that of the treat- failed to show that these programs were standard practice
ing physician, who is the professional most likely to have in the community.66 Interestingly, the blood center in this
developed a relationship with the patient. In the transfusion case did have a directed-donation program, but not for
setting, however, this question is more complex, and some patients with the plaintiff ’s specific illness, which required
courts have reasoned that a physician who refers a patient multiple transfusions.
for surgery is not responsible for obtaining informed con- The outcome of these cases is often directly affected by
sent for transfusion; rather, it is the surgeon who performs the expert testimony presented. In Spann v. Irwin Memorial
the procedure who has the duty to obtain informed con- Blood Centers,66 the plaintiffs failed to offer testimony regard-
sent.59 This line of cases does not automatically implicate all ing the standard of practice in the community. If expert
surgeons assisting in the surgery. Some cases have held the testimony supporting the general availability of directed
primary surgeon responsible, while refusing to hold assist- donations or other alternative therapies had been presented,
ing physicians responsible for the failure to obtain informed it is possible that the testimony could have been sufficient
consent.60 In some circumstances, however, physicians may to defeat a motion for summary judgment. This would have
delegate the duty to obtain informed consent to other health created a question of fact for the jury, provided that the fail-
care providers.61 ure to follow the standard practice was causally connected
For the most part, hospitals have not been held to owe to the injury. It is, however, noteworthy that hospitals and
a duty to patients to obtain informed consent,62 because blood-collection facilities have successfully defended against
the obligation to do so is generally that of the physician allegations that the failure to provide directed donations
who ordered the transfusion. However, as the requirement constitutes negligence by introducing evidence that directed
to obtain informed consent becomes incorporated into donations are not necessarily safer, and possibly even less
the requirements of standard-setting organizations like safe, than regular donations.67
the Joint Commission on the Accreditation of Healthcare In other cases, the interplay between the testimony of
Organizations (JCAHO)63 hospitals increasingly will be held experts and the availability of the optional therapy deter-
accountable for ensuring that informed consent is obtained. mines the outcome of the case. In Doe v. Johnston,68 the
LEGAL PRINCIPLES AND ISSUES CENTRAL TO TRANSFUSION MEDICINE
plaintiff offered testimony on the superiority of autologous BOX 17–7 Illustrative Case
transfusion, but conflicting evidence over the “reasonable
availability” of the procedure in 1985 created a question of Oiler v. Willke70 demonstrates what can happen when expert
fact sufficient to uphold the trial court’s denial of a request testimony and a broad interpretation of a reasonably foresee-
for a directed verdict in favor of the plaintiff. able injury coincide in the same case. Oiler involved a plate-
As the type and availability of alternative therapies to let transfusion that allegedly resulted in the transmission of
blood transfusion continue to grow, the obligation to inform human immunodeficiency virus (HIV). The plaintiff presented
expert testimony that the platelet transfusion was unnecessary.
potential recipients about the availability of these therapies
Although the transfusion was administered in 1980, at a time
will increase. In addition to preoperative autologous dona- when the transmissibility of HIV through blood transfusion
tion, many hospitals now offer intraoperative blood sal- was not yet established, the court ruled that the plaintiff had
vage, postoperative blood salvage, and acute normovolemic established a genuine issue of fact as to whether the transmis-
hemodilution. Not every alternative is either appropriate for, sion of HIV was foreseeable because the plaintiff had presented
or available to, every patient. To the extent that an option is evidence that many viruses, including hepatitis, were known to
available and appropriate for a particular procedure, and the be transmissible through blood transfusion.
patient is not informed of the option, any complication from
an allogeneic unit can create potential liability for a physician
or hospital.
Less clear is whether a hospital would be held liable
for failing to make these alternative therapies available for negligence required the transfusion that transmitted HIV.72
patients. Case law from the HIV litigation, discussed ear- In other cases, applying a narrower interpretation of foresee-
lier, suggests that a hospital would not be required to offer ability, courts have granted summary-judgment motions in
alternative therapies that are so new that they have not yet favor of defendant physicians who provided negligent treat-
become accepted as the norm. To the extent that a hospital ment that required transfusion because the outcome of the
offering bloodless surgery is located in or near other hospi- transfusion was not foreseeable.
tals, its physicians could rely on the fact that the patient has
the option to choose to have a procedure performed at a dif- Blood Administration
ferent location. It is possible, however, that an expert witness
testifying that the availability of these alternatives is now the Now that the risk of transfusion-transmitted diseases has
standard of care could create a question of fact for the jury. dramatically decreased because of donor screening and
In still other cases involving an alleged failure to inform increasingly sensitive testing, more attention is being placed
a plaintiff of alternative procedures, the determining issue on injuries resulting from improper administration of blood
is not the availability of the alternative, but a hospital or and blood components. Currently, one of the leading causes
blood center’s adherence to its existing internal policies. In of death relating to transfusion is mistransfusion, or admin-
Doe v. American National Red Cross,69 a hospital that failed istering the wrong unit to the wrong patient. Although not 17
to inform a plaintiff about the option of directed donations many cases have been reported, plaintiffs are most likely to
was denied summary judgment, at least in part, because a be successful if they allege medical negligence in the transfu- 245
policy allowing such donations existed within the hospital. sion of the unit. In Walker v. Humana Medical Corporation,73
Again, the outcome of many of these cases is fact-specific a phlebotomist failed to draw blood from the proper patient.
and depended not only on the availability of an alterna- The resulting transfusion caused a reaction due to the
tive therapy, but also on the question whether the available incompatible blood, and the court found sufficient evidence
therapy would be appropriate for the specific patient. of wantonness, or knowledgeable malfeasance, to present the
facts to a jury.
Theories of breach of contract or of implied warranty
Medically Necessary Transfusions
generally will not apply in these cases because usually it can-
Beginning in the 1980s, many plaintiffs brought cases against not be demonstrated that the unit of blood itself is defective.
physicians and hospitals based on the theory that the trans- Although logic suggests that every case of mistransfusion
fusion that caused injury to the patient was not medically results in a legal action, few legal cases or rulings involve
necessary. The lack of reference standards or universally these types of injuries, presumably because most of these
accepted comprehensive transfusion guidelines, as well as cases are settled by hospitals out of court. Given the existence
regional variations in transfusion practice, contributed to of devices that can prevent mistransfusion, constructing a
the unpredictable outcome of these cases. In addition, expert successful legal defense to these cases is difficult.
testimony is critical to defining the appropriate standard of
care in these cases. One of the most interesting and troubling
cases for transfusing physicians is presented in Box 17–7. PRIVACY AND SECURITY CONCERNS
Once a case challenging the medical necessity of a partic- UNDER HIPAA
ular transfusion becomes a battle of the experts, the outcome
is difficult to predict, but juries have found hospitals and
Invasion of Privacy
physicians liable for injuries on the basis of expert testimony
about the necessity of the transfusion.71 Case dicta, as well as Invasion of privacy is the unwarranted appropriation or
experience, underscore the importance of hospital policies exploitation of one’s personality, publicizing one’s private
and hospital transfusion guidelines, as well as the review of affairs with which the public has no legitimate concern, or
transfusion practice in the hospital transfusion committee. wrongful intrusion into one’s private activities, in such a
Some courts, applying a broad definition of foreseeability, manner as to cause mental suffering, shame, or humiliation
have allowed cases to proceed against physicians whose prior to a person of ordinary sensibilities.
BLOOD BANKING
Right to Privacy trative, physical, and technical safeguards under the HIPAA
Security Rule,80 which are designed to safeguard electronically
Although not explicitly stated in the Constitution, “a right to protected health information. Generally, covered entities may
be left alone” began to emerge in the late 1800s. This right has use protected health information only as it relates to treat-
evolved into a liberty of personal autonomy protected by the ment, payment, or health care operations.81 Covered entities
14th amendment of the United States Constitution. Along must make reasonable efforts to limit the use or disclosure
with the constitutional right of privacy, states have adopted of, and requests for, patient health information (PHI) to the
statutory rights of privacy that limit access to personal infor- minimal amount necessary to accomplish the intended pur-
mation, and the Federal Trade Commission (FTC) protects pose. Covered entities are required to develop written policies
the public’s financial privacy through legislation such as the and procedures that are summarized in a “Notice of Privacy
Right to Financial Privacy Act of 197874 and the more recent Practices” that is publicly available.82 Noncompliance with the
Financial Services Modernization Act of 1999.75 Privacy Rules may result in civil and criminal penalties.83
Most recently, federal and state privacy laws protect
health-related information.
Business Associates
Even facilities that are not covered entities under HIPAA may
HIPAA have to comply with certain requirements if they are “busi-
ness associates.” Covered Entities are required to enter into
With the passage of the Health Insurance Portability and business associate agreements with persons and organiza-
Accountability Act of 1996 (HIPAA),76 as amended, which tions that perform functions or activities on their behalf and
became effective for most “covered entities” on April 14, receive PHI in connection with these services for other than
2003, practices relating to the use and disclosure of medi- treatment purposes. Common examples of business associ-
cal information have been subject to increasing attention. ates are accountants, consultants, administrators, and finan-
The federal privacy regulations that implement portions of cial services. HIPAA specifically includes accreditation as an
HIPAA (”Privacy Rules”) were designed to combat fraud and activity, giving rise to a business associate relationship.
abuse in health care, standardize the electronic exchange of Business associate agreements principally are designed
administrative and financial data, and protect the privacy to ensure that business associates and those acting on their
and security of individual health information. behalf use the PHI only for the purposes that they are engaged
to perform or as otherwise required by law; safeguard the
Covered Entities information from disclosure and misuse; promptly report
any disclosure to the Covered Entity; and assist the Covered
“Covered Entities” generally include health plans, health care Entity in complying with HIPAA requirements.
clearinghouses, and health care providers.77 A health care
II provider is a covered entity under HIPAA if it (1) meets the
Criminal Penalties
definition of “health care provider;” and (2) transmits health
246 information in electronic form in connection with covered In June 2005, the U.S. Department of Justice (DOJ) clarified
transactions. Independent blood suppliers who do not per- who can be held criminally liable under HIPAA. Covered
form patient-related health care activities are specifically entities and those individuals that “knowingly” obtain or
exempt from the provision of the act, as they engage in “activ- disclose individually identifiable health information in viola-
ities related to the procurement or banking of blood, sperm, tion of the applicable regulations may face criminal liability.
organs, or any other tissue for administration to patients.”78 The DOJ interpreted the “knowingly” element of the HIPAA
However, an entity’s status as a “covered entity” requires statute for criminal liability as requiring only knowledge of
a careful, fact-specific analysis of all of the activities pro- the actions that constitute an offense. Specific knowledge
vided by the facility in question. The Department of Health of an action being in violation of the HIPAA statute is not
and Human Services (HHS) has stated, “the procurement required.84
or banking of organs, blood (including autologous blood) In 2004, the U.S. Attorney in Seattle announced that
… or any other tissue or human product is not considered a hospital phlebotomist was being indicted for violating
to be health care under this rule and the organizations that the HIPAA privacy law.85 The phlebotomist had allegedly
perform such activities would not be considered health care accessed the medical records of a patient with a terminal
providers when conducting these functions.”79 cancer condition, obtained credit cards in the patient’s name,
Under this definition, if a facility is involved in activities and run up more than $9,000 in charges. In a statement to
other than the procurement and banking of blood, such as lab- the court, the patient said he “lost a year of life both mentally
oratory testing used to diagnose or otherwise treat a patient, as and physically dealing with the stress” that resulted from the
opposed to simply performing blood screening, it may be pro- phlebotomist’s actions. The guilty individual signed a plea
viding patient-related services that bring it under the definition agreement and was sentenced to 16 months in jail. At the
of “health care provider” under HIPAA. It is not clear whether time, the DOJ trumpeted the first HIPAA criminal prosecu-
a facility that performs cross-matching and other compatibil- tion. The DOJ site announced: “This case should serve as a
ity testing would be considered a health care provider. To be reminder that misuse of patient information may result in
considered a covered entity, the health care provider must also criminal prosecution.”85 Under its new legal opinion, how-
meet the second criterion: it must be transmitting protected ever, the phlebotomist could not be prosecuted further
health information in electronic form in connection with one under HIPAA because, arguably, he did not knowingly vio-
of the “standard transactions” listed in the rules. late HIPAA. Whereas HIPAA protects the health information
Covered entities must comply with numerous and extensive of individuals, it does not create a private cause of action
requirements under the Privacy Rules, including adminis- for those aggrieved. State law, however, may provide other
LEGAL PRINCIPLES AND ISSUES CENTRAL TO TRANSFUSION MEDICINE
avenues of liability. Generally, HIPAA does not preempt state Preventive Measures
privacy laws or professional licensure requirements, so that
determining compliance requires examining the relevant One theory to be considered is that patients who have suf-
state’s confidentiality laws.86 fered from medical errors are motivated not so much by
the prospect of economic compensation, as by the desire to
ensure that the error is not repeated. A pilot mediation con-
EXTRAPOLATION OF THE TRANSFUSION ducted under the sponsorship of the Massachusetts Board of
LEGAL EXPERIENCE Registration in Medicine confirms this theory and strongly
suggests that private dispute resolution through mediation
The legal precedents established in transfusion cases have improves the outcome for all concerned. As Ed Dauer, one
application to other areas. Notably, the collection, process- of mediation’s key proponents suggests, “For its part, media-
ing, manipulation, and administration of cellular therapy tion, when properly employed, can be private, integrative,
products raise legal issues similar to those presented in trans- safe, nonjudgmental, and flexible in scope, process, and
fusion cases. Although the actual requirements for cellular outcome. It can be a safe harbor with therapeutic potential,
therapy donor screening and unit testing may ultimately dif- and can offer its participants the opportunity to address the
fer from those applicable to blood donors and blood com- source as well as the consequence of the immediate problem.
ponents, the case law concerning the appropriate negligence Mediation may, in short, offer a process whose traditional
standard and the legal requirements for defining the stan- attributes are consistent with, rather than antithetical to, the
dard of care against which actions are to be measured will requisites of quality improvement.”87
influence the outcome of cellular therapy litigation. Some lawyers have embraced variations of these con-
An additional legal risk in the collection, processing and cepts in the form of “good-faith conferences” and “apol-
administration of cellular therapy products, not generally ogy meetings.” The good-faith conference is geared toward
present in transfusion medicine, is the relative efficacy of furnishing the plaintiff with the opportunity to vent his or
processing protocols for cellular therapy products. Unlike her anger at the defendant. The conference may be a condi-
blood components, which are basically licensed generic bio- tion to settlement, or it may follow settlement. No apology
logics, most cellular therapy products are developed under is promised or expected, but apologies often develop dur-
highly individualized collection, processing, manipulation, ing well-managed meetings. The apology meeting is simi-
freezing, and/or thawing protocols. Although some of these lar, except that an apology from the defendant is expected,
protocols and methods are protected intellectual property, along with a therapeutic discussion.
others are in the public domain and thus are available for Both types of meetings must be carefully planned and skill-
use. One of the areas of legal concern will be the extent to fully executed. It is imperative that all parties have a certain
which these different processes are associated with varying level of trust in one another and that all are committed to con-
engraftment survival rates. As in the transfusion field, the ducting the meeting in a civil and nonthreatening manner.
testimony of experts will be key to the outcome of litigation 17
in this area. Litigation Strategies
In the cellular therapies arena, the proliferation of clini- 247
cal trials will also draw increased attention to the informed- It is not uncommon for plaintiffs in transfusion-related
consent process. The failure to have in place or to follow a litigation to file suit against as many defendants as ethically
well-thought-out process will most certainly lead to legal possible to maximize the chances of finding a culpable “deep
problems. In addition, given the vulnerability of expectant pocket.” Typically, the treating physician and other medical
parents, issues related to informed consent for cord blood professionals involved in the administration of transfusions,
donation, particularly in cord blood banks that collect and along with the hospital, the blood bank, and even AABB, will
store these units for use by the family, will pose legal con- be named as defendants. Although the first reaction among
cerns for institutions that provide this service. co-defendants is to point fingers and assign blame, a more
Finally, facilities that store cord units for future use by the productive approach often is to work together on a common
family may find that issues surrounding the obligation to store defense. Not only can this cooperative approach result in
and, in some cases, dispose of these collected units may raise stronger defenses, but it can significantly reduce legal bills,
contract, as well as negligence, issues. Although the intricacies as well.
of contract law are beyond the scope of this chapter, facilities
that store cord blood units should have all storage contracts Affirmative Defenses
carefully reviewed by legal counsel. Even the best contracts
cannot completely protect a facility against legal liability in Even if a plaintiff is able to prove all four elements of negli-
the event that a facility negligently loses or renders unusable gence, a defendant can raise defenses that minimize or defeat
a cord blood unit. In this event, as in the case of blood donor the plaintiff ’s claims. One such defense is that the plaintiff
injuries, the actions or inactions of the facility are likely to be failed to bring the lawsuit within the legal time limits set by
judged against an ordinary negligence standard. the jurisdiction where the case is being brought. This defense
is referred to as the statute-of-limitations defense.
Another affirmative defense available in many jurisdic-
DEFENDING AGAINST LEGAL RISKS tions is that the actions of the plaintiff contributed to the
cause of the injury. This theory is called contributory negli-
In addition to preventive measures for avoiding litigation, liti- gence or comparative negligence on the part of the plaintiff.
gation strategies can help expedite resolution of issues and, Contributory or comparative negligence is a legally contrib-
if all else fails, affirmative defenses can be used effectively to uting cause, in addition to the negligence of the defendant, in
defend against litigation. bringing about the plaintiff ’s harm.
BLOOD BANKING A related affirmative defense is that an unforeseeable 17. Doe v. Arts, 823 A.2d 855 (N.J. Super. 2003).
intervening event caused the injury, rather than the defen- 18. Goelz v. J.K. & Susie L. Wadley Research Institute & Blood Bank, 350
S.W.2d 573 (Tex. Civ. App. 1961).
dant’s negligence. 19. Pneumocystis carinii pneumonia among persons with hemophilia
A. MMWR 1982;31:365.
20. Health and Human Services: New Release, March 4, 1985.
LESSONS LEARNED 21. Kirkendall v. Harbor Ins. Co., 698 F.Supp. 768 (W.D. Ark. 1988), aff ’d,
887 F.2d 857 (8th Cir. 1989).
22. Doe v. American Nat. Red Cross, 866 F.Supp. 242 (D. Md. 1994).
By applying the difficult lessons learned from the HIV- 23. Zaccone v. American Red Cross, 872 F.Supp. 457 (N.D. Ohio 1994).
related transfusion litigation to cases appearing on the hori- 24. Silliman C, Ambruso R, Boshkov L: Transfusion-related acute lung
zon, it is possible to avoid litigation, or at least minimize its injury. Blood 2005;105:2266–2273.
impact. One key lesson is the importance to medical facilities 25. United Blood Services v. Quintana, 827 P.2d 509, 520–21 (Colo. 1992).
26. Ray v. American Nat. Red Cross, 696 A.2d 399 (D.C. App. 1997).
of promptly implementing government recommendations 27. Kleinman S, Caulfield T, Chan P, et al: Toward an understanding of
and standards adopted by private organizations. Another les- transfusion-related acute lung injury: Statement of a consensus panel.
son is to consider all available scientific evidence continually Transfusion 2004;44:1774–1789.
and to keep abreast of best practices, especially those relating 28. National Blood Service, Hospitals & Science Website: Update on TRALI:
reducing the risk. Blood Matters Winter 2003/4:14.
to new techniques, technologies, and medical challenges, to 29. Quijano v. United States, 325 F.3d 564 (5th Cir. 2003).
anticipate the adoption of a practice as the standard of care. 30. McClear JA: Family wins tainted-blood suit. The Detroit News, October
Equally important, particularly in situations in which 16, 1997.
lack of consensus exists regarding how to manage a particu- 31. AABB. Standards for blood banks and transfusion services 22d ed., 2003.
lar risk, is the demonstration that a facility is knowledgeable 32. AABB Association Bulletin 05–02: Bacterial contamination of plate-
lets: summary for clinicians on potential management issues related to
about the risks and has considered the available options. transfusion recipients and blood donors. AABB Bacterial Contamina-
Finally, efforts such as consensus conferences and work- tion Task Force, February 23, 2005.
shops that attempt to establish a consensus or practice within 33. Silva M: Summary of results of AABB bacterial contamination task force
the professional community are extremely important. Even survey. Advisory Committee on Blood Safety and Availability Committee,
Department of Health and Human Services, http://www.hhs.gov/blood-
such simple efforts as the development and adoption of a safety/transcripts/ACBSA_Transcript_Jan_25_2005.pdf, January 25, 2005.
uniform donor-history questionnaire can ensure the recog- 34. West Nile virus activity in the United States, September 26–October
nition of clear standards of care. 2, 2002 and investigations of West Nile virus infections in recipients
Minimizing the risk of legal liability in the transfusion of blood transfusion and organ transplantation. MMWR 2002;51:
setting is not difficult, but in today’s litigious society, elimi- 39–884.
35. AABB: Information regarding West Nile virus, www.aabb.org/press-
nating that risk entirely is not possible. The most important room/in_the_news/wnwnv100302.htm, October 3, 2002.
preventive measures have been and remain using best prac- 36. AABB Association Bulletin 02–09: Further information relating to
tices, engaging in effective risk communication with patients, voluntary withdrawal of frozen product to mitigate the risk of trans-
and taking swift corrective action when problems inevitably fusion of West Nile virus through blood transfusion: Statement of the
American Association of Blood Banks, America’s Blood Centers, and the
II occur. Keeping current with standards of practice and effec- American Red Cross, December 18, 2002.
tively educating staff about those standards is essential to 37. AABB Association Bulletin 02–10: Update on testing of frozen products
248 fending off legal claims. By taking these proactive steps, it for West Nile virus: Statement of the American Association of Blood
will be possible to demonstrate, to the extent possible, that Banks, America’s Blood Centers, and the American Red Cross, Decem-
the physician or facility conformed to the currently accepted ber 24, 2000.
38. AABB Association Bulletin 03–06: Update on FDA West Nile virus rec-
standard of care in the transfusion and cellular therapy com- ommendations, May 13, 2003.
munities. This will allow the professional to spend much less 39. AABB Association Bulletin 04–04: Joint statement of the American
time defending past activities and devote considerably more Association of Blood Banks, America’s Blood Centers, and American
valuable time to serving donors and patients. Red Cross on implementation of individual donation nucleic acid
amplification testing for West Nile virus, June 4, 2004.
40. Nakhasi H: Development of West Nile virus testing and donor screening
as a model for screening bioterrorist agents. www.fda.gov/oc/initiatives/
REFERENCES criticalpath/wnv.html, July 15, 2004.
41. FDA Talk Paper: New precautionary measures to reduce the theoretical
1. Speiser SM, Krause CF, Gans AW: The American Law of Torts. Roches- risk of new variant CJD from blood products, T99-38, http://www.fda.
ter, N.Y.: Lawyers Cooperative Publishing Co, 1.4, 1983. gov/bbs/topics/ANSWERS/ANS00970.html, August 17, 1999.
2. Restatement (Second) of Torts § 402A American Law Institute, 1964. 42. FDA Guidance for industry: revised preventive measures to reduce the
3. Nestor v. Hospital Pavia, et al: 2005 WL 348313 (D.P.R. 2005). possible risk of Creutzfeldt-Jakob Disease (CJD) or variant Creutzfeldt-
4. Stanley v. McCarver, 92 P.3d 849 (Ariz. 2004). Jakob Disease (vCJD) by blood and blood products. January 9, 2002.
5. Betesh v. United States, 400 F.Supp. 238 (D.D.C. 1974). 43. Blajchman MA: Allogeneic blood transfusions, immunomodulation,
6. Dyer v. Trachtman, 679 N.W.2d 311 (Mich. 2004). and postoperative bacterial infection: do we have the answers yet?
7. Reed v. Bojarski, 764 A.2d 433 (N.J. 2001). Transfusion 1997;37:121–125.
8. Smith v. American Red Cross, 886 F.Supp. 1494 (E.D. Mo. 1995). 44. Vamvakas EC: Transfusion-association cancer recurrence and postop-
9. Snyder v. American Ass’n of Blood Banks, 676 A.2d 1036 (N.J. 1996). erative infection: meta-analysis of randomized, controlled clinical trials.
10. Weigand v. University Hosp. of New York University Med. Ctr., 659 Transfusion 1996;36:175–186.
N.Y.S.2d 395 (N.Y. 1997). 45. McAlister FA, Clark HD, Wells PS, Laupacis A: Perioperative allogeneic
11. N.N.V. v. American Ass’n of Blood Banks, 75 Cal.App.4th 1358, 1373 blood transfusion does not cause adverse sequelae in patients with cancer:
(Cal.App.4th Dist. 1999). a meta-analysis of unconfounded studies. Br J Surg 1998;85:171–178.
12. Keeton WP, Dobbs DB: Prosser and Keeton on Torts, 5th ed. West Grove, 46. van de Watering LMG, Herrnans J, Houbiers JGA, et al: Beneficial effects
St. Paul, Minn.:1984. of leukocyte depletion of transfused blood on postoperative complica-
13. Osborn v. Irwin Memorial Blood Bank, 5 Cal.App.4th 234 (Cal. App. 1st tions in patients undergoing cardiac surgery: a randomized clinical trial.
Dist. 1992), review denied (July 9, 1992). Circulation 1998;97:562–568.
14. Boland v. Montefiore, 800 N.Y.S.2d 343 (N.Y. Super. 2005). 47. AABB Association Bulletin 99–7: Leukocyte reduction. August 12, 1999.
15. Simmons KC: Recovery for emotional distress based on fear of contract- 48. Hillyer C, Silberstein L, Ness P, Anderson K (eds.): Blood Banking and
ing HIV or AIDS. ALR 5th 59:535, 1998. Transfusion Medicine: Basic Principles and Transfusion Practice. Phila-
16. Lobowitz v. Albert Einstein Medical Center, 623 A.2d 3 (Pa. Super. 1993). delphia, Churchill Livingstone/Elsevier Science, 2003:255.
LEGAL PRINCIPLES AND ISSUES CENTRAL TO TRANSFUSION MEDICINE
49. King KE, Ness PM. Prevention of transfusion-associated graft-versus- 66. Spann v. Irwin Memorial Blood Centers, 40 Cal.Rptr.2d 360 (1st Dist.
host disease in undiagnosed severe combined immunodeficiency syn- 1995).
drome [Abstract]. Blood 1996;88:528a. 67. Hoemke v. New York Blood Center, 912 F.2d 550 (2d Cir. 1990).
50. Maung ZT, Wood AC, Jackson GH, et al. Transfusion-associated graft- 68. Doe v. Johnston, 476 N.W.2d 28 (Iowa 1991), reh’g denied (October 23,
versus-host disease in fludarabine-treated B-chronic lymphocytic leu- 1991).
kaemia. Br J Haematol 1994;88:649–652. 69. Doe v. American Nat. Red Cross, 848 F.Supp.1228 (S.D.W.Va. 1994).
51. McDonnell v. American Nat. Red Cross, 2005 WL 1335128 (Mich. App. 70. Oiler v. Willke, 642 N.E.2d 667 (Ohio App. 4th Dist. 1994).
June 7, 2005) (unpublished opinion). 71. Jeanne v. Hawkes Hosp. of Mt. Carmel, 598 N.E.2d 1174 (Ohio App.10th
52. Baumann v. American Nat. Red Cross, 262 F.Supp.2d 965, 968 (C.D. Ill. Dist. 1991), case dismissed, 579 N.E.2d 210 (1991).
2003). 72. Doe v. United States, 737 F.Supp. 155 (D.R.I. 1990).
53. Paulsen v. Gundersen, 260 N.W. 448 (Wis. 1935). 73. Walker v. Humana Medical Corp., 415 So.2d 1107 (Ala. Civ. App.), on
54. Scaria v. St. Paul Fire & Marine Ins. Co., 227 N.W.2d 647 (Wis. 1975). remand, 423 So.2d 891 (June 9, 1982).
55. Wright v. Fred Hutchinson Cancer Research Center, 206 FRD 679 (W.D. 74. 12 U.S.C.A. §§ 3401, et seq. (November 10, 1978).
Wash. 2002). 75. 15 U.S.C.A. §§ 6701, et seq. (November 12, 1999).
56. Stowell C: Informed Consent for Blood Transfusion. Bethesda, Md.: 76. Public Law 104–191(HR 3103) (August 21, 1996).
American Association of Blood Banks, 1997:4–5. 77. 45 C.F.R. § 160.103 (2003).
57. Sawyer v. Methodist Hospital, 522 F.2d 1102 (6th Cir. 1975). 78. 65 Fed. Reg. 82462, 82571–82572 (December 28, 2000).
58. Canterbury v. Spence, 464 F.2d 772 (D.C. Cir.), cert. denied, 409 U.S. 79. 65 Fed. Reg. 82477 (December 28, 2000).
1064 (1972). 80. 45 C.F.R. §§ 164.308 et seq. (current through November 17, 2005).
59. Douglass v. Alton Ochsner Medical Foundation, 696 So.2d 136, (La. Ct. 81. 45 C.F.R. § 164.506 (August 14, 2002).
App. 5th Cir. 1997). 82. 45 C.F.R. § 164.530 (August 14, 2002).
60. Jones v. Philadelphia College of Osteopathic Medicine, 813 F.Supp. 1125 83. 42 U.S.C. § 1320d–5 (December 27, 2001).
(E.D. Pa. 1993). 84. Memorandum Opinion for the General Counsel Department of Health
61. Smogor v. Enke, 874 F.2d 295 (5th Cir. 1989). and Human Services and the Senior Counsel to the Deputy Attorney
62. Goss v. Oklahoma Blood Institute, 856 P.2d 998 (Okla. App. 1990). General (June 1, 2005).
63. JCAHO: Comprehensive Accreditation Manual for Hospitals. Oak- 85. Scott M: HIPAA gavel drops: A message to healthcare. Radiology Today
brook Terrace, Ill.: Joint Commission on Accreditation of Healthcare Nov. 22, 2004;24:38.
Organizations, 2005. 86. 65 Fed. Reg. 82566 (December 28, 2000).
64. Whittington v. Episcopal Hosp., 768 A.2d 1144 (Pa. Super. 2001). 87. Dauer EA, Marcus LJ: Adapting mediation to link resolution of medi-
65. Howell v. Spokane & Inland Empire Blood Bank, 785 P.2d 815, 823 (Wash. cal malpractice disputes with health care quality improvement. Law &
1990). Contemp Probs 1997;60:185–218.
17
249
D. Specific Blood Components
Chapter 18
Packed Red Blood Cells and
Related Products
Sally A. Campbell-Lee ● Paul M. Ness
Landsteiner1 discovered the ABO system in 1901, more than donor tubing, and several attached polyvinyl resin bags.5
100 years ago. Since that time, the development of a safe and Within the polyvinyl resin bags is 63 mL of, most commonly,
effective anticoagulant-preservative solution2 and the efforts citrate, phosphate, and dextrose (CPD) or citrate, phosphate,
in World War II that led to current methods of organized dextrose, and adenine (CPDA-1) anticoagulant-preservative,
blood collection3 have promoted transfusion medicine to an which provides a shelf-life of 35 days when the contents are
essential aspect of clinical medicine. The appropriate use of stored at 4°C. Including the anticoagulant-preservative, the
red blood cell concentrates (commonly referred to as packed volume of a unit of whole blood is approximately 510 mL
red blood cells, pRBCs) depends on knowledge of the physi- (450 mL of blood plus 63 mL of anticoagulant),6 although
ology of RBC transfusion and the therapeutic options of var- many blood centers now collect 500 mL of donor blood.
ious component manipulations (Table 18–1). This chapter Within 24 hours of collection, the platelets (if refrigerated)
covers the manufacturing, storage, and indications for whole and granulocytes are dysfunctional, and several coagulation
blood, pRBCs, and related products. factors are at suboptimal levels.7 To optimize factor activ-
II ity and platelet recovery, preparation of fresh frozen plasma
(FFP) and platelet concentrates must occur within 8 hours.8
250 COLLECTION One of the problems with the standard method of collecting
whole blood is that the amount of anticoagulant-preservative is
In 2001, nearly 15 million units of whole blood were collected predetermined, but the exact amount of whole blood collected
in the United States.4 Whole blood, composed of RBCs, leu- is not. Because of variability in donor hematocrits, different
kocytes, platelets, and plasma, is collected into a closed, sterile RBC masses are collected per unit, resulting in variable RBC
system. The system consists of a phlebotomy needle, integral mass provided per unit transfused.
CMV, cytomegalovirus; IgA, immunoglobulin A; RBC, red blood cell; TAGVHD, transfusion-associated graft-versus-host disease.
to that of 8 to 10 platelet units13; this effect appears to be
Substance Content (mmol/unit) ACD-A* CPDA-1* CPD* CP2D* Adsol (AS-1)† Nutri-cell (AS-3)†
*
pRBC storage solutions.
†
pRBC additive solutions.
From Hogman CF. Preparation and preservation of red cells. Vox Sang 1998;74(suppl 2):177.
PACKED RED BLOOD CELLS AND RELATED PRODUCTS
The function of erythrocyte 2,3-DPG is to bind to deoxy- hematocrit is about 62%, the Adsol-preserved pRBCs are
hemoglobin and facilitate oxygen transport. When 2,3-DPG less viscous. The lower viscosity results in potentially faster
binds to deoxyhemoglobin, the deoxyhemoglobin molecule flow rates, which are beneficial in emergency situations.32
is stabilized, and the equilibrium between deoxyhemoglobin A second commonly used additive solution, AS-3, contains
and oxyhemoglobin shifts toward deoxyhemoglobin. This sodium chloride, phosphate, adenine, and glucose and is
interaction shifts the oxygen-dissociation curve to the right, used similarly to AS-1.
decreasing the oxygen affinity of hemoglobin and enhancing Other additive solutions have been investigated. A solu-
oxygen delivery to tissues.28 Therefore, with decreased 2,3- tion containing adenine, dextrose, mannitol, sodium citrate,
DPG levels, the oxygen dissociation curve is shifted to the ammonium chloride, and inorganic phosphate was studied.33
left, decreasing oxygen delivery to tissues. Levels of ATP were higher in this test solution, compared
Although it may be feared that transfused RBCs beyond a with those in Adsol, over 84 days of storage. 2,3-DPG lev-
certain date of storage are of limited benefit to the patient, it els were higher in the test solution, but not significantly, and
has been shown that 2,3-DPG is rapidly regenerated in trans- hemolysis was higher in the Adsol units. It was thought that
fused RBCs, with nearly complete restoration after 1 day. the higher ATP concentration in the test solution was due to
In addition, in hypoxia, lactic acid is produced, decreasing the the ammonium or phosphate ions or both.30 In vivo survival
pH and thus shifting the oxygen-dissociation curve back to the was demonstrated in a second study by using a modifica-
right.29 An increase in cardiac output also occurs with hypoxia, tion of the previous test solution (less ammonium chloride).
increasing oxygen delivery. Therefor, for most patients requir- The 24-hour chromium-51 viability was superior to that of
ing transfusion, decreased 2,3-DPG is of little consequence. Adsol at 8 or 9 weeks of storage. Preparation for transfusion
For the patient who is in shock and cannot increase cardiac required removal of supernatant with one washing step.34
output to compensate, the current preservative solutions may This additive solution appears to be suitable for extending
not be optimal because no added component is available to pRBC shelf life, but confirmatory studies are necessary.
slow the decrease in 2,3-DPG levels. Rejuvenation solutions can restore some intracellular ATP
and 2,3-DPG lost during storage. The current FDA-licensed
solution contains pyruvate, inosine, phosphate, and adenine.
Additive and Rejuvenation Solutions This solution may be added only to pRBCs prepared from
In 1983 an RBC additive solution, Adsol (Baxter, Round Lake, whole blood collected in CPD or CPDA-1. It may be added
Ill.), also referred to as AS-1, was approved for use. Adsol con- at any point between 3 days after collection and 3 days after
sists of adenine (to help maintain ATP during storage), dex- expiration. This solution is not intended for intravenous use,
trose, saline, and mannitol. It contains 60% more adenine and and pRBCs must be washed before administration or used
approximately 2.5 times as much glucose as CPDA-1.30 The when thawing frozen units.8 Rejuvenation of units stored
addition of mannitol prevents excessive hemolysis over the in AS-1 or AS-3 has also been studied.35 In one experiment,
storage period.31 The increased glucose allows an adequate sup- rejuvenation of the AS-1 and AS-3 pRBCs resulted in above-
ply of energy for the RBCs beyond 35 days, and blood stored normal levels of ATP after one treatment but suboptimal lev- 18
with Adsol is outdated in 42 days (Tables 18–2 and 18–3).8 This els of 2,3-DPG. A second treatment raised the 2,3-DPG levels
decreases the number of pRBCs lost because of outdating and to normal. The authors suggested that this may be due to the 253
is helpful in shipping and storing autologous blood. increased adenine in the storage solutions; the conversion of
One of the added benefits of this preservative solution is 1,3-DPG to 3-phosphoglycerate is favored, decreasing 1,3-
that the amount of plasma recovered from a unit of whole DPG availability to make 2,3-DPG.35 Currently, only units
blood can be maximized. Whole blood is collected into a sys- stored in CPD or CPDA-1 may be used for rejuvenation.
tem of multiple closed bags containing CPD. The whole blood
is then centrifuged for separation, and sufficient plasma is Temperature
removed to raise the hematocrit to approximately 85%. Then
100 mL of Adsol preservative solution is added to the pRBCs, RBC concentrates must be stored between 1° C and 6° C.8
with a resultant hematocrit of 60% to 70%. Because the final Storage at this temperature slows RBC metabolism and facil-
itates extended storage in blood banks. Liquid blood storage
in CPDA-1 allows a shelf-life of up to 35 days. Even at 4–8° C,
significant chemical changes take place during the storage
period that may have clinical consequences for some patients.
Table 18–3 42-Day Poststorage pRBC These changes are collectively known as the storage lesion.
Characteristics after Resuspension in AS-3 No clinically significant change occurs in the plasma lev-
Characteristic Prestorage Poststorage
els of sodium and chloride. However, the plasma potassium
increases nearly eightfold over 28 days.36 At a temperature of
pH 6.8 6.4 4–8° C, the sodium-potassium pump is essentially nonfunc-
ATP (μmol/g Hb) 4.1 2.9 tional, and intracellular and extracellular levels gradually
DPG (μmol/g Hb) 9.0 0.3 equilibrate. In addition, the hemolysis that occurs during the
Potassium (mEq/L) 2.4 63
Glucose (mg/dL) 608 402 storage period results in increased potassium in the superna-
Plasma Hb (mg/dL) 39 372 tant. However, because the total volume of plasma in pRBCs
Hemolysis (%) — 0.61 is low (approximately 70 mL), the total potassium burden
is only about 5.5 mEq at product expiration. Therefore, the
ATP, adenosine triphosphate; DPG, 2,3-diphosphoglycerate; Hb, potassium load is rarely a clinical problem except in the set-
hemoglobin; pRBC, packed red blood cell.
Modified from Holme S, Elfath MD, Whitley P: Evaluation of in ting of preexisting hyperkalemia and renal failure or for very
vivo and in vitro quality of apheresis-collected RBC stored for 42 days. sick neonates. In these situations, fresher units of RBCs or
Vox Sang 1998;75:212–217. washed RBCs may be used.
BLOOD BANKING
Storage Containers as a cryoprotective agent. Because glycerol binds water, the
formation of ice spicules from the solvent water within the
The use of vinyl plastic blood bags and tubing in transfu- unit, which would damage the RBCs, is prevented.46 Another
sion medicine is advantageous in the collection, processing, theory concerning the efficacy of glycerol is that it prevents
storage, and dispensing of blood components. In the early cellular hypotonicity or hypertonicity, which may enhance
1970s, however, reports surfaced about the potential toxic- cell lysis. The blood of donors who have sickle cell trait is
ity of blood stored in bags with di(2-ethylhexyl)phthalate unsuitable for freezing because hemolysis occurs during
(DEHP). DEHP, the chemical that allows the vinyl plastic routine deglycerolization.47
to be pliable, is referred to as a plasticizer. DEHP is added Of three methods of freezing RBCs,48 the most common
in large quantities to the plastic, approximately 40% by one in the United States is the high-glycerol (40% to 50%)
weight. It is not bound to the plastic but is dissolved in it. As method. A low-glycerol method exists, but it has several dis-
a result, DEHP can leak into blood stored in the container advantages: liquid nitrogen must be used for storage, and the
and be transfused along with the blood. In 1970, Jaeger and metal containers in which the plastic blood bags are placed
Rubin37 reported that 5 to 7 mg of DEHP could be isolated before freezing can cause explosions if the liquid nitrogen
per 100 mL of blood. In addition, two patients were found leaks.49 With a third method, agglomeration, the cells are
to have DEHP at levels ranging from 0.069 to 0.270 mg per deglycerolized with a low-ionic-strength saline solution.
gram dry weight of tissue. A more recent evaluation of the The cells clump and sediment in the bag, after which the
levels of DEHP in stored blood components found that supernatant is removed, and the cells are washed.50
as storage time increased, the amount of DEHP detected When the glycerol solution has been added, the cells are
ranged from 6.8 to 36.5 μg/mL in RBC concentrates. Whole frozen and stored at −65 ° C or below in a suitable freezer
blood products had the highest DEHP levels, compared with or in liquid nitrogen with a gas-phase temperature below
RBC concentrates, irradiated RBC concentrates, FFP, and −120 ° C.8 To be transfused, the cells must be thawed and
platelet products.38 Concern over the potential toxic effects deglycerolized. One of the initial drawbacks of the high-
of DEHP in humans has fueled a great deal of research and glycerol method was that the procedure of thawing and
much debate. DEHP, identified as a carcinogen in rats and removing the glycerol had many practical limitations. A
mice,39 is ubiquitous in the environment. It caused a form simpler method of processing these cells for transfusion was
of shock lung leading to death when administered to rats in developed by Meryman and Hornblower.50 Their method
intravenous form,40 caused testicular atrophy in rats given requires only two cycles of centrifugation, washing with saline,
dietary doses,41 and led to lung injury in dogs and baboons and resuspension with isotonic saline containing glucose.
transfused with stored blood.42 The remaining product contains few white cells or plate-
Because of an association with hepatomegaly, DEHP lets, and 99.9% of the plasma is removed by the extensive
has also been linked to potential hepatocarcinogenicity. washing during processing.51 More than 90% of the donor’s
Hepatomegaly has been shown to be caused by proliferation RBCs are recovered.52 The 24-hour posttransfusion survival
II of cellular organelles called peroxisomes. Peroxisomes are has been shown to be 85% to 90%.53
involved in the β-oxidation of fatty acids, producing hydro- Posttransfusion survival and oxygen-carrying capacity
gen peroxide, which has been suggested to be the causative of RBCs are affected by the amount of time spent between
254
agent in the carcinogenicity of DEHP.43 With all of this infor- donation, refrigeration, and freezing.46 Frozen cells have
mation, however, no direct causal link has been established been shown to maintain prefreezing ATP and 2,3-DPG
between DEHP and cancer in humans. levels. The standard is to freeze within 6 days of collection,
Because of the concerns about toxicity, alternative mate- before these factors become significantly depleted. When it
rials for blood-storage containers have been under investi- is necessary to freeze older units, rejuvenation with a solu-
gation for some time. A study comparing another plastic, tion containing pyruvate, glucose, phosphate, and adenine
poly(ethylene-co-ethyl acrylate) (EEA), with polyvinyl chlo- can be used.54
ride (PVC) containing DEHP44 found that blood stored in The major advantage of frozen RBCs is that rare blood
EEA containers had higher plasma hemoglobin and greater types, such as (Oh) Bombay, can be stored. Patients with rare
susceptibility to osmotic lysis than did blood stored in PVC phenotypes may make autologous donations that can be fro-
containers. When DEHP was added to EEA containers, blood zen for later use. Cells from autologous donors can also be
stored in EEA containers without DEHP had greater RBC frozen if more units are required than can be collected in
osmotic fragility than did blood in EEA with DEHP or PVC the maximum 42-day liquid storage period or if surgery is
containers. postponed.
PVC plasticized with butyryl-n-trihexyl-citrate For patients who become alloimmunized to multiple
(BTHC) has been introduced in place of DEHP. Use of clinically significant RBC antigens, frozen RBCs from donors
BTHC has not become widespread. Less BTHC than with rare phenotypes are useful. Among these patients are
DEHP leaches into the bag contents, and excellent 24- multiply transfused patients with sickle cell anemia who have
hour posttransfusion RBC recovery occurs with minimal multiple alloantibodies. African-Americans frequently lack
hemolysis.45 antigens found on most donor RBCs from whites. During
blood drives targeting the African-American community for
the benefit of such patients, it is helpful to phenotype these
FROZEN RED BLOOD CELL RBCs and freeze the more uncommon types.
CONCENTRATES Because 99.9% of plasma is removed in processing frozen
RBCs, patients who may have adverse events related to plasma
RBCs can be frozen for long-term storage, at least 10 years, components may also benefit from the use of frozen RBCs.
and probably longer for certain indications.8 After pRBCs are For example, immunoglobulin A (IgA)-deficient patients
prepared from whole blood, they are treated with glycerol with anti-IgA antibodies may have anaphylactoid reactions
PACKED RED BLOOD CELLS AND RELATED PRODUCTS
when exposed to donor plasma. In the past, patients who LEUKOCYTE-REDUCED RED BLOOD
had multiple febrile nonhemolytic transfusion reactions that
CELL CONCENTRATES
persisted despite removal of the buffy coat and treatment
with medications also benefited from the more complete
removal of cytokine-laden plasma and the leukoreduction Leukocyte-reduced RBCs are defined by the AABB as having
that thawed frozen RBCs offered. Newer leukocyte reduction less than 5 × 106 leukocytes in the final component. Early
techniques have largely replaced this indication. Because of techniques of leukocyte reduction involved centrifugation,
the high cost and cumbersome nature of freeze-thaw proce- washing with saline, and removal of the buffy coat. A second-
dures, other more routine uses of frozen RBCs are difficult generation technique known as the spin-cool-filter method
to justify. If a simpler means of preparation of frozen blood was introduced in the 1980s.59 This method requires use of
were available that avoided the limitations of the open sys- 1-week-old RBCs, which are centrifuged and then cooled for
tems now used, more widespread use could be envisioned. 4 hours to enhance microaggregate formation before passage
through a microaggregate filter. Currently, filtration can be
performed at the bedside or in the laboratory with attachable
CYTOMEGALOVIRUS-SERONEGATIVE RED filters that reduce leukocytes more than 99.9% with less than
BLOOD CELL CONCENTRATES 10% depletion of RBCs.60
Leukocyte-reduced RBCs are indicated primarily in the
Cytomegalovirus (CMV) is a double-stranded DNA her- setting of repeated febrile nonhemolytic transfusion reac-
pesvirus (human herpesvirus 5) that can be transmitted by tions. It was previously thought that these reactions were
transfusion. Forty percent to 100% of adults are seroposi- mediated only by antibodies to foreign leukocyte antigens.61
tive for CMV, depending on socioeconomic status and geo- Now increasing evidence suggests that cytokines produced by
graphic region.55 Persistent and latent infection can result, the leukocytes during storage also cause these reactions.62,63
as well as reactivation and reinfection. The first report of Prestorage leukocyte reduction has also been shown to
transfusion-transmitted CMV described a syndrome seen in reduce substantially the incidence of febrile nonhemolytic
patients 3 to 8 weeks after cardiopulmonary bypass. The syn- transfusion reactions.64
drome consisted of fever, lymphocytosis, and splenomegaly.56 A second indication for leukocyte-reduced RBCs is the
A congenital syndrome including petechiae, hepatospleno- prevention of sensitization to human leukocyte antigens
megaly, jaundice, and microcephaly has also been identi- (HLAs) in bone-marrow transplant recipients and other
fied. Postnatal infection in children can cause hepatitis. In patients who require frequent platelet transfusion. The Trial
immunocompromised adults, infection with CMV can result to Reduce Alloimmunization to Platelets (TRAP) study65
in interstitial pneumonitis, hepatitis, encephalitis, gastroen- demonstrated a reduction in alloimmunization in acute
teritis, thrombocytopenia, or leukopenia; in certain patients myelogenous leukemia patients who received leukocyte-
receiving transplants, these conditions are associated with reduced blood components. Platelet refractoriness, although
a high fatality rate. In immunocompetent adults, fever and low in the control group (16%), was reduced among patients 18
hepatitis may result, but most patients have an asymptomatic receiving leukocyte-reduced products (7%).
mononucleosis.55 As a third indication, CMV transmission is mitigated
255
Transfusion-transmitted CMV is of concern in immu- as compared with the transfusion of unscreened blood prod-
nocompromised patients. CMV-seronegative blood is often ucts66–68 and may be comparable (as mentioned earlier), or
requested for CMV-negative bone-marrow transplant can- better than the use of seronegative products. At the time of
didates or recipients, in utero transfusions, low-birth-weight this writing, it is unclear whether using both leukoreduced,
premature infants of CMV-negative mothers, CMV-negative seronegative units is superior to using either alone. Last, the
pregnant women, and rare cases of human immunodefi- use of nucleic acid testing, though only in use on a research
ciency virus–positive, CMV-negative patients. CMV-nega- basis, may provide additional safety (see Chapter 46).
tive recipients of solid organ transplants from CMV-negative Finally, leukocyte-reduced RBCs may have another indi-
donors are among the patients who may benefit but in whom cation. The transfusion of allogeneic blood is thought by
the risk of using blood products not tested for CMV is not many scientists to be immunosuppressive, an effect termed
well established.52 transfusion-related immune modulation (TRIM). Several
The sites of CMV latency are thought to include CD34- studies appear to document this effect, but some controversy
positive progenitor cells and CD13- and CD14-positive remains. A correlation between pretransplantation alloge-
monocytes.55 Thus transfusion-transmitted disease can be neic RBC transfusions and improved renal allograft survival
mitigated by removal of leukocytes in the pRBCs. A study has been known for many years.69,70 Initial reports appeared
published in 1995 compared bedside leukoreduction and before the availability of cyclosporine and other improve-
CMV-seronegative blood products in bone-marrow trans- ments in immunosuppression that made this phenomenon
plant recipients and suggested that filtration is an effective less clinically relevant. However, a subsequent study includ-
alternative to CMV-seronegative blood for the prevention ing patients receiving modern immunosuppressive regimens
of transfusion-transmitted CMV. A follow-up of 142 bone- demonstrated that recipients of three unmodified allogeneic
marrow transplant recipients found that in 62 CMV-serone- RBC transfusions had 90% 1-year and 79% 5-year graft sur-
gative recipients of bone marrow from CMV-seronegative vivals, compared with 82% and 70% 1- and 5-year survivals,
donors, supported with the use of leukocyte-reduced blood respectively, for patients who received no transfusion, which
products, no documented CMV infection occurred.57 The suggests continued importance.71
American Association of Blood Banks (AABB) has sug- Perioperative allogeneic RBC transfusion may also have
gested that both approaches, leukocyte reduction and the an adverse effect on tumor recurrence.72–74 The data on
use of CMV-seronegative blood, are essentially equivalent this effect are more controversial; Blajchman75 observed
in the prevention of CMV transmission.58 that approximately 50% of the nonrandomized studies
BLOOD BANKING indicate an adverse effect of transfusion on tumor recur- WASHED PRBCs
rence. Patients with colorectal cancer who received peri-
operative transfusions were shown to have longer hospital Washed RBCs are prepared with isotonic saline by either
stays than those who did not receive transfusion,76 but the manual or automated methods. Automation is more effi-
effect in this study was attributed to a higher incidence of cient, resulting in loss of fewer RBCs with each wash cycle.
postoperative infection. In surgical patients, perioperative Because washing takes place in an open system, the product
transfusion may also predispose to bacterial infection. A must be used within 24 hours.
dose-response relation appears to exist between transfu- Washing RBCs removes plasma proteins, some leuko-
sion and the probability of infection; transfusion is the cytes, and remaining platelets. This product is indicated
best predictor of infection, over such factors as extent of for patients who have had recurrent severe allergic trans-
trauma, degree of blood loss, and presence of wound con- fusion reactions that are not prevented by antihistamines.
tamination,77 although confounding factors have not been Recipient IgE antibodies to donor plasma proteins mediate
eliminated in many studies. these reactions. Washed RBCs are also indicated for IgA-defi-
The mechanism for these effects may be related to the cient patients who have formed anti-IgA antibodies. In these
transfusion of contaminating white blood cells. The donor patients, transfusion of blood products containing plasma
white blood cells could cause a downregulation of cellular with IgA can result in anaphylaxis.83
immunity, mediated by secretion of T-helper 2 cytokines In patients with paroxysmal nocturnal hemoglobinuria
and inhibitors (interleukin-4, interleukin-10, and trans- (PNH), a rare disorder in which RBCs are unusually sensi-
forming growth factor-β), with resultant inhibition of the tive to complement lysis, transfusion of washed RBCs has
T-helper 1 response.78 The potential benefit of leukocyte been advocated to prevent hemolysis. However, a report by
reduction appears to be supported by reports that patients Brecher and Taswell84 on 23 patients with PNH seen over a
having colorectal surgery who received leukocyte-reduced 38-year period appears to show that this is a needless prac-
RBC transfusions had fewer infections and shorter hospi- tice.84 A total of 431 RBC products (94 whole blood, 208
tal stays than did those who received unmodified RBCs.78 pRBCs, 80 leukocyte-reduced RBCs, 38 washed RBCs, 5 fro-
Animal studies also appear to support this theory. In rab- zen RBCs, and 6 intraoperatively salvaged units) were trans-
bits inoculated with VX-2 tumor cells, those that received fused with only one episode of hemolysis after transfusion.
unmodified allogeneic RBC transfusions had significantly This single event was associated with transfusion of group O
more pulmonary metastases than did those that received whole blood to a group AB individual. Although the need for
99.8% leukocyte-reduced blood.79 washed RBCs in PNH is questionable, this disorder is rare,
An editorial concerning TRIM stated, “prestorage WBC and changing established transfusion protocols may not be
reduction is an intervention that is virtually risk-free and justified.
that, except for its cost, has no down side … the decision
to implement universal prestorage WBC reduction need
II not be delayed until further evidence of efficacy becomes IRRADIATED PRBCs
available.”80 The suggestion that leukocyte-reduced RBCs
256 decrease the unwanted immunosuppressive effects of trans- RBCs are commonly irradiated by using a cesium 137 source.
fusion may be further proof that conversion to a leukocyte- A dose of at least 2500 cGy must be delivered to each unit,
reduced blood supply enhances patients’ safety. and quality-control standards have been published to ensure
In light of these indications for leukoreduced RBCs, that this dose is achieved with blood bank irradiators.85
many centers have moved to universal prestorage leukore- After irradiation, storage time is decreased to a maximum
duction. Some hospitals, however, use a selective leukode- of 28 days8 because of shortened RBC survival and increased
pletion policy based on lack of direct convincing evidence potassium leakage. Please see Chapter 28 for additional and
that leukoreduction offers benefits to patients not in cer- detailed information.
tain categories. Data in support of this practice is available The purpose of irradiation of cellular blood products in
from a prospective randomized trial, which showed no transfusion medicine is to inactivate immunocompetent lym-
benefit to conversion to a universally leukodepleted inven- phocytes. Irradiated RBCs are indicated for the prevention
tory, except for specific indications, such as a reduction in of transfusion-associated graft-versus-host disease in immu-
febrile transfusion reactions.81 The cited study did, how- nocompromised patients, a frequently fatal complication.
ever, show a trend toward increased safety with leukore- Neonates, patients with hematologic malignancies, patients
duced units and may have been underpowered. One of the with aplastic anemia, bone-marrow transplant recipients,
pitfalls of selective leukodepletion is the difficulty of rec- and patients with congenital immune deficiency are suscep-
ognizing which patients require leukodepletion in complex tible to transfusion-associated graft-versus-host disease.86
clinical settings; in addition, this study did not address the Graft-versus-host disease is also a potential hazard of directed
potential adverse long-term consequences of CMV infec- donation from first-degree relatives who share HLA haplo-
tion or alloimmunization in the recipients. Without a com- types.87 See Chapter 53 for additional information regarding
munity standard for universal leukodepletion, patients in this phenomenon.
groups that do qualify for certain protocols may be over-
looked and will not receive these components.82 Based on REFERENCES
these concerns, the authors feel that universal leukodeple-
tion offers safety advantages for all patients and guarantees 1. Landsteiner K. Ueber Agglutinationserscheinungen normalen men-
schlichen Blutes. Wein Klin Wochenschr 1901;14:1132–1134.
the receipt of leukoreduced blood in such cases in which 2. Loutit JF, Mollison PL. Advantages of a disodium-citrate-glucose mix-
the specific indication is not recognized or may not have ture as a blood preservative. Br Med J 1943;2:744–745.
yet occurred. 3. Schmidt PJ. Charles Drew, a legend in our time. Transfusion 1997;37:234–236.
PACKED RED BLOOD CELLS AND RELATED PRODUCTS
4. National Blood Data Resource Center. Comprehensive Report on Blood 36. Latham J, Bove J, Weirich F. Chemical and hematologic changes in
Collection and Transfusion in the United States in 2001. Bethesda, Md., stored CPDA-1 blood. Transfusion 1982;22:158–159.
National Blood Data Resource Center, 2002. 37. Jaeger R, Rubin R. Contamination of blood stored in plastic packs. Lan-
5. Walter C, Murphy W. A closed gravity technique for the preservation of cet 1970;29:151.
whole blood in ACD solution utilizing plastic equipment. Surg Gynecol 38. Inoue K, Kawaguchi M, Yamanaka R, et al. Evaluation and analysis of
Obstet 1952;94:687–692. exposure levels of di(2-ethylhexyl)phthalate from blood bags. Clin
6. Brecher M (ed). Technical Manual, 15th ed. Bethesda, Md., American Chim Acta 2005;358:159–166.
Association of Blood Banks, 1999. 39. Kluwe W, Haseman J, Douglas J, Huff J. The carcinogenicity of dietary
7. Baldini M, Costea N, Dameschek W. The viability of stored human di-(2-ethylhexyl)phthalate (DEHP) in Fischer 344 rats and B6C3F1
platelets. Blood 1960;16:1669–1692. mice. J Toxicol Environ Health 1982;10:797–815.
8. Standards for Blood Banks and Transfusion Services, 23rd ed. Bethesda, 40. Schulz CO, Rubin RJ, Hutchins GM. Acute lung toxicity and sudden
Md., American Association of Blood Banks, 2005. death in rats following the intravenous administration of the plasticizer,
9. Knutson F, Rider J, Franck V, et al. A new apheresis procedure for di-(2-ethylhexyl)phthalate, solubilised with tween surfactants. Toxicol
the preparation of high-quality red cells and plasma. Transfusion Appl Pharmacol 1975;33:514–525.
1999;39:565–571. 41. Gray T, Butterworth K. Testicular atrophy produced by phthalate esters.
10. Holme S, Elfath M, Whitley P. Evaluation of in vivo and in vitro quality of Arch Toxicol Suppl 1980;4:452–455.
apheresis-collected RBC stored for 42 days. Vox Sang 1998;75:212–217. 42. Bennet SH, Creelhoed GW, Aaron RK, et al. Pulmonary injury resulting
11. Shi P, Ness P. Two-unit red cell apheresis and its potential advantages from perfusion with stored bank blood in the baboon and dog. J Surg
over traditional whole-blood donation. Transfusion 1999;39:218–225. Res 1972;13:295–306.
12. Bandarenko N, Rose M, Kowalsky R, et al. In vivo and in vitro character- 43. Rubin RJ, Ness PM. What price progress? An update on vinyl plastic
istics of double units of RBCs collected by apheresis with a single in-line blood bags. Transfusion 1989;29:358–361.
WBC-reduction filter. Transfusion 2001;41:1373–1377. 44. Horowitz B, Stryker M, Waldman A, et al. Stabilization of red blood cells
13. Lavee J, Martinowitz U, Mohr R, et al. The effect of transfusion of fresh by the plasticizer, diethylhexylphthalate. Vox Sang 1985;48:150–155.
whole blood versus platelet concentrates after cardiac operations. J Tho- 45. Hogman CF, Eriksson L, Ericson A, Reppucci AJ. Storage of saline-
rac Cardiovasc Surg 1989;97:204–212. adenine-glucose- mannitol-suspe red cells in a new plastic container:
14. Golan M, Modan M, Lavee J, et al. Transfusion of fresh whole blood polyvinylchloride plasticized with butyryl-n-trihexyl-citrate. Transfu-
stored (4°C) for short period fails to improve platelet aggregation of sion 1991;31:26–29.
extracellular matrix and clinical hemostasis after cardiopulmonary 46. Huggins C. Preparation and usefulness of frozen blood. Annu Rev Med
bypass. J Thorac Cardiovasc Surg 1990;99:354–360. 1985;36:499–503.
15. Schmidt P. Whole blood transfusion. Transfusion 1984;24:368–369. 47. Meryman HT, Hornblower M. Freezing and deglycerolizing sickle-trait
16. Leslie SD, Toy P. Laboratory hemostatic abnormalities in massively red blood cells. Transfusion 1976;16:627–632.
transfused patients given red blood cells and crystalloid. Am J Clin 48. Boral L, Henry JB. Clinical Diagnosis and Management by Laboratory
Pathol 1991;96:770–773. Methods, 19th ed. Philadelphia, WB Saunders, 1996, p 803.
17. National Institute of Health Consensus Conference. Fresh frozen 49. Akerblom O, Hogman CF. Frozen blood: A method for low-glycerol,
plasma: indications and rules. Transfus Med Rev 1987;1:201–204. liquid nitrogen freezing allowing different postthaw deglycerolization
18. Boral L, Henry JB. Clinical Diagnosis and Management by Laboratory procedures. Transfusion 1974;14:16–26.
Methods, 19th ed. Philadelphia, WB Saunders, 1996, p 802. 50. Meryman HT, Hornblower M. A simplified procedure for deglyceroliz-
19. Loutit JF, Mollison PL. Advantages of a disodium-citrate-glucose mix- ing red blood cells frozen in a high glycerol concentration. Transfusion
ture as a blood preservative. Br Med J 1943;2:744–745. 1977;17:438–442.
20. Atchison SR, Rettke SR, Fromme GA, et al. Plasma glucose concentra- 51. Contreras TJ, Valeri CR. A comparison of methods to wash liquid-stored
tions during liver transplantation. Mayo Clin Proc 1989;64:241–245. red blood cells and red blood cells frozen with high or low concentra-
21. Cheng KW, Chen CL, Cheng YF, et al. Dextrose in the banked blood tion of glycerol. Transfusion 1976;16:539–565.
18
products does not seem to affect the blood glucose levels in patients 52. Sayers M. Transfusion-transmitted viral infections other than hepati-
undergoing liver transplantation. World J Gastroenterol 2005;11:87–89. tis and human immunodeficiency virus infection: cytomegalovirus, 257
22. Rous P, Turner JR. The preservation of living red blood cells in vitro. Epstein-Barr virus, human herpes virus 6 and human parvovirus B19.
J Exp Med 1996;23:219–248. Arch Pathol Lab Med 1994;118:346–349.
23. Howland W, Bellville J, Zucker M, et al. Massive blood replacement: Failure 53. Valeri CR. Factors influencing the 24 hour post-transfusion survival
to observe citrate intoxication. Surg Gynecol Obstet 1957;105:529–540. and oxygen transport function of previously frozen red cells preserved
24. Moore GL, Peck CC, Sohmer PR, Zuck TF. Some properties of blood with 40% glycerol and frozen at −80°C. Transfusion 1974;14:1–15.
stored in anticoagulant CPDA-1 solution. Transfusion 1981;21:135–137. 54. Valeri CR, Zaroulis CG. Rejuvenation and freezing of outdated stored
25. Rapoport S. Dimensional, osmotic and clinical changes of erythrocytes human red cells. N Engl J Med 1972;287:1307–1313.
in stored blood I: Blood preserved in sodium citrate, neutral and acid 55. Pamphilon DH, Rider JR, Barbara JAJ, Williamson LM. Prevention of transfu-
citrate-glucose (ACD) mixtures. J Clin Invest 1947;26:591–615. sion-transmitted cytomegalovirus infection. Transfus Med 1999;9:115–123.
26. Nakao K, Wada T, Kamiyama T, et al. A direct relationship between ade- 56. Kreel I, Zarroff LI, Canter JW. A syndrome following total body perfu-
nosine triphosphate level and in vivo viability of erythrocytes. Nature sion. Surg Gynecol Obstet 1960;111:317–321.
1962;194:877–878. 57. Pamphilon DH, Foot ABM, Adeodu A, et al. Prophylaxis and prevention
27. Hogman C. Preparation and preservation of red cells. Vox Sang 1998; of CMV infection in bone marrow allograft recipients: leucodepleted
74(suppl 2):177–187. platelets are equivalent to those from CMV seronegative donors. Bone
28. Harken A. The surgical significance of the oxyhemoglobin dissociation Marrow Transplant 1999;23(suppl 1):S66.
curve. Surg Gynecol Obstet 1997;144:935–955. 58. American Association of Blood Banks. Leukocyte Reduction for the
29. Beutler E. What is the clinical importance of alterations of the hemoglo- Prevention of Transfusion-transmitted Cytomegalovirus (AABB Asso-
bin oxygen affinity in preserved blood, especially as produced by varia- ciation Bulletin 97–2, 10–12). Bethesda, Md., American Association of
tions of red cell 2,3-DPG content? Vox Sang 1078;34:1130. Blood Banks Press, 1997.
30. Mollison PL, Engelfriet CP, Contreras M. Blood Transfusion in Clinical 59. Meryman HT, Hornblower M. The preparation of red cells depleted of
Medicine, 10th ed. Boston, Blackwell Scientific, 1997, p 249. leukocytes: Review and evaluation. Transfusion 1986;26:101–106.
31. Hogman CF. Additive system approach in blood transfusion: birth of 60. Dzik WH. Leukoreduced blood components: Laboratory and
the SAG and Sagman systems. Vox Sang 1986;51:339–340. clinical aspects. In Rossi EC, Simon TL, Moss GS, et al. (eds): Principles
32. Heaton A, Miripol J, Aster R, et al. Use of Adsol preservation solution for of Transfusion Medicine. Baltimore, Williams & Wilkins, 1995, p 353.
prolonged storage of low viscosity AS-1 red blood cells. Br J Haematol 61. Payne R. Leukocyte agglutinins in human sera. Arch Intern Med 1957;
1984;57:467–478. 99:587–606.
33. Greenwalt TJ, McGuinness CG, Dumaswala UJ, Carter HW. Studies in 62. Davenport RD, Kunkel SL. Cytokine roles in hemolytic and non-hemo-
red blood cell preservation, 3: A phosphate-ammonium-adenine addi- lytic transfusion reactions. Transfus Med Rev 1994;7:157–168.
tive solution. Vox Sang 1990;58:94–99. 63. Heddle NM, Kelton JG. Febrile non-hemolytic transfusion reactions.
34. Greenwalt TJ, Dumaswala UJ, Dhingra N, et al. Studies in red blood cell In Popovsky MA (ed): Transfusion Reactions. Bethesda, Md., AABB
preservation, 7: In vivo and in vitro studies with a modified phosphate- Press, 1996, p 45.
ammonium additive solution. Vox Sang 1993;65:87–94. 64. King KE, Shirey RS, Thoman SK, et al. Universal leukoreduction
35. Brecher ME, Zylstra-Halling VW, Pineda AA. Rejuvenation of erythro- decreases the incidence of febrile nonhemolytic transfusion reactions to
cytes preserved with AS-1 and AS-3. Am J Clin Pathol 1991;96:767–769. RBCs. Transfusion 2004;44:25–29.
BLOOD BANKING 65. The Trial to Reduce Alloimmunization to Platelets Study Group. 76. Vamvakas EC, Carven JH. Allogeneic blood transfusion, hospital charges,
Leukocyte reduction and ultraviolet B irradiation of platelets to prevent and length of hospitalization: a study of 487 consecutive patients undergo-
alloimmunization and refractoriness to platelet transfusions. N Engl ing colorectal cancer resection. Arch Pathol Lab Med 1998;122:145–151.
J Med 1997;337:1861–1869. 77. Blumberg N, Heal J. Transfusion immunomodulation. In Anderson KC,
66. Bowden RA, Slichter SJ, Sayers M, et al. A comparison of filtered leuko- Ness PM (eds): Scientific Basis of Transfusion Medicine, 2nd ed. Phila-
cyte-reduced and cytomegalovirus (CMV) seronegative blood products delphia, WB Saunders, 2000, p 430.
for the prevention of transfusion-associated CMV infection after mar- 78. Blumberg N, Heal JM. Blood transfusion immunomodulation: the
row transplant. Blood 1995;86:3598–3603. silent epidemic. Arch Pathol Lab Med 1998;122:117–119.
67. Narvios AB, de Lima M, Shah H, Lichtiger B. Transfusion of leukore- 79. Blajchman MA, Bardossy L, Carmen R, et al. Allogeneic blood trans-
duced cellular blood components from cytomegalovirus-unscreened fusion-induced enhancement of tumor growth: two animal models
donors in allogeneic hematopoietic transplant recipients: analysis of 72 showing amelioration by leukocyte reduction and passive transfer using
recipients. Bone Marrow Transplant 2005;36:499–501. spleen cells. Blood 1993;81:1880–1882.
68. Laupacis A, Brown J, Costello B, et al. Prevention of prosttransfusion 80. Blajchman MA. Transfusion-associated immunomodulation and uni-
CMV in the era of universal WBC reduction: a consensus statement. versal white cell reduction: are we putting the cart before the horse?
Transfusion 2001;41:560–569. Transfusion 1999;39:665–670.
69. Opelz G, Sengar DP, Mickey MR, et al. Effect of blood transfusions on 81. Dzik WH, Anderson JK, O’Neill EM, et al. A prospective, randomized
subsequent kidney transplants. Transplant Proc 1973;5:253–259. clinical trial of universal WBC reduction. Transfusion 2002;42:1114–
70. Blajchman MA, Singal DP. Renal transplantation: the role of red blood 1122.
cell antigens, histocompatibility antigens and blood transfusions on 82. Ness PM, Lipton KS. Selective transfusion protocols: errors and acci-
renal allograft survival. Transfus Med Rev 1983;3:171–179. dents waiting to happen. Transfusion 2001;41:713–715.
71. Opelz G, Vanrentergehm Y, Kirste G, et al. Prospective evaluation of pre- 83. Vyas GN, Holmdahl L, Perkins HA, Fudenberg HH. Serologic speci-
transplant blood transfusion in cadaver kidney recipients. Transplanta- ficity of human anti-IgA and its significance in transfusion. Blood
tion 1997;63:964–697. 1969;34:573–581.
72. Schriemer PA, Longnecker DE, Mintz PD. The possible immunosup- 84. Brecher ME, Taswell HF. Paroxysmal nocturnal hemoglobinuria and
pressive effects of perioperative blood transfusion in cancer patients. the transfusion of washed red cells: a myth revisited. Transfusion
Anesthesiology 1988;68:422–428. 1989;29:681–685.
73. Van Aken WG. Does perioperative blood transfusion promote tumor 85. Moroff G, Leitman SF, Luban NLC. Principles of blood irradiation, dose
growth? Transfus Med Rev 1989;3:243–252. validation, and quality control. Transfusion 1997;37:1084–1092.
74. Heiss MM, Jauch KW, Delanoff C, et al. Blood transfusion modulated 86. Leitman SF, Holland PV. Irradiation of blood products: indications and
tumor recurrence: A randomized study of autologous versus homologous guidelines. Transfusion 1985;25:293–303.
blood transfusion in colorectal cancer. J Clin Oncol 1994;12:1859–1867. 87. Thaler M, Shamiss A, Orgad S, et al. The role of blood from HLA-homo-
75. Blajchman MA. Immunomodulatory effects of allogeneic blood trans- zygous donors in fatal transfusion associated graft versus host disease
fusions: Clinical manifestations and mechanisms. Vox Sang 1988;74 after open heart surgery. N Engl J Med 1989;321:25–28.
(suppl 2):315.
II
258
Chapter 19
Fresh Frozen Plasma and Related Products
Robert L. Crookes ● Christopher D. Hillyer
Component therapy has had a profound impact on the components, of which C3 (1.2 mg/dL) predominates quanti-
practice of transfusion medicine. The extraction of various tatively. In addition, plasma proteins include those involved
constituents, including plasma, from whole blood has led in maintaining normal rheologic properties of blood (e.g.,
to increased efficacy and economic utilization of the blood coagulation and fibrinolytic proteins).4–6
supply. When only the component that is needed is trans-
fused, the patient is spared untoward effects of other blood
components. Plasma, both in the circulation and for transfu- PLASMA COLLECTION, QUALITY,
sion, contains a variety of organic and inorganic elements AND PROCESSING
with therapeutic value as described subsequently. However,
the isolation, purification, and preparation for injection or Plasma can be obtained through centrifugation of whole
transfusion of some specific plasma constituents [e.g., factor blood, single-donor plasmapheresis, or as a by-product of
VIII (FVIII), albumin, immunoglobulins] have limited the cytapheresis [e.g., platelet or red blood cell (RBC); concur-
use of fresh frozen plasma (FFP) in clinical practice. The use rent plasma]. One unit of plasma is defined as the amount
of FFP (and related plasma components) is now reserved for of plasma obtained from centrifugation of 1 unit of whole
conditions requiring therapy in which replacement of mul- blood, and it usually contains 180 to 300 mL. When plasma is
tiple plasma constituents is needed or for which the specific obtained through single-donor plasmapheresis, the amount
constituent is not commercially available in a purified inject- may be 2 to 3 times greater than that obtained from whole
able or transfusable form. blood processing (500 to 800 mL). Most plasma that will be
used as FFP or FP24 (see later) is collected from whole blood
(or equivalent) donors, and thus donor safety, screening and 19
PHYSIOLOGIC ROLE OF PLASMA testing measures apply.
The rapidity with which plasma is collected and stored 259
Plasma is the aqueous component of blood in which cellu- determines its quality and subsequent use.7,8 Thus several
lar elements and macromolecules are transported through- plasma products are available at various centers. Single-donor
out the body and other constituents are maintained in a plasmapheresis can produce source plasma (or single-donor
dynamic equilibrium with the extravascular compartment. plasma) when it is stored at or below −18°C at variable times
The composition of plasma is influenced by gender, age, diet, from its collection. Plasma prepared from whole-blood or
and other individual and environmental characteristics.1,2 apheresis collections and stored frozen within 8 hours of its
The major component of plasma is water, which constitutes collection (at −18°C or colder) is called fresh frozen plasma
approximately 85% to 90% of the plasma volume. The sol- (FFP). Plasma separated from whole-blood donations and
ute component constitutes 0.3 mol/L, of which about 30% is frozen below –18°C within 24 hours of collection (FP24)
made up of proteins, with colloids, crystalloids, clotting fac- shows good retention of relevant coagulation factor activity.
tors, hormones, vitamins, and trace elements making up the However, compared with historic data on FFP frozen within 8
rest. Normal human plasma has a density of 1.055 to 1.063 g/ hours, fibrinogen, FV, FVIII, and FXI were shown to be reduced
mL and a pH that varies between 7.33 and 7.43 with respec- in FP24 by 12%, 15%, 23%, and 7%, respectively.9 This may
tive temperature changes between 37°C and 4°C.3 be significant in certain clinical circumstances, such as in the
Although human plasma contains a multitude of sub- treatment of neonates with coagulopathies. In general, how-
stances including ionic and nonionic solutes, the practice ever, FFP and FP24 are used interchangeably in many facilities
of transfusion medicine has exploited plasma mainly for throughout the United States, especially at large tertiary care
its protein content. It is estimated that human plasma con- hospitals with busy level 1 trauma centers.10 Thawed plasma is
tains more than 700 different proteins with various physi- prepared in a closed system from FFP or FP24. Centrifugation
ologic characteristics and functions. Although some 120 or sedimentation of whole blood can also produce recovered
proteins have been isolated, only a few are available for clini- plasma, which is obtained from a whole-blood donation: liq-
cal use. The most abundant plasma protein is albumin, with uid plasma, when collected and stored refrigerated within 5
a concentration between 3500 and 5000 mg/dL. Albumin days of the expiration date of whole blood: Cryo-poor plasma
is responsible for maintaining colloid oncotic pressure and (cryosupernatant plasma; CSP), which is the plasma product
serves as a major transport protein for endogenous and exog- remaining after the cryoprecipitate fraction is extracted from
enous substances. Plasma proteins with immunologic func- FFP through cold precipitation.
tions include the immunoglobulin (Ig) family, of which IgG Both source plasma and recovered plasma may be used
(5 to 14 mg/dL) is the most abundant, and the complement in the manufacture of various plasma derivatives. Source
BLOOD BANKING plasma can also be administered as component therapy, tion is present.17 Leukocyte depletion of FFP may also result
whereas recovered plasma, because of the less-stringent con- in losses of coagulation factor activity and in increases in
ditions of collection, processing, and storage, does not meet markers of coagulation activation, depending on the type of
the standards for coagulation factor concentrations and filter used. This is unlikely to be clinically significant unless
therefore is not used as component therapy. Liquid plasma subsequent processing of plasma (such as pathogen inactiva-
is no longer used in the United States. Donor retested plasma tion) results in further losses of coagulation factors.21
(FFP-DR) is FFP, or FP24, which has been placed in “quar-
antine” and released for transfusion only after the donor has
donated a subsequent blood donation, which has been tested CLINICAL CONSIDERATIONS
(“retested”) and found to be negative for markers of transfu-
sion-transmissible infection. The quarantine period is at least In clinical practice, FFP has been used as an exogenous
56 days. FFP-DR is considered a safer product, as subsequent source of proteins, specifically albumin, immunoglobulins,
testing of the blood donor is expected to detect and eliminate coagulation factors, and certain protease inhibitors. With the
plasma donations that are in an infectious window period.11 development of the fractionation method of Cohn and col-
Virally inactivated plasma is pooled plasma subjected to leagues22 and the consequent capability to administer indi-
a solvent-detergent process and is termed solvent-detergent vidual plasma components, the use of FFP has been reserved
plasma (SD plasma). This method is highly efficient at inac- for situations in which it is necessary to replace either mul-
tivating viruses by using a combination of organic solvent, tiple plasma constituents (e.g., multiple factor deficiency)
tri(n-butyl)phosphate (TNBP), and a nonionic detergent, or a plasma constituent not yet isolated. The randomized
Triton X-100. The SD method is virucidal against lipid- controlled trial evidence base for the clinical use of FFP is,
enveloped viruses, including human immunodeficiency virus however, limited.23 Indications and contraindications are
(HIV) types 1 and 2, hepatitis viruses (HBV, HCV, HGV), listed in Table 19–1.
human T-cell leukemia/lymphoma virus (HTLV) types I and
II, vesicular stomatitis virus, Sindbis virus, and Sendai virus,
but does not inactivate parvovirus or hepatitis A virus. THE USE OF FFP IN THE MANAGEMENT
SD plasma is pooled from approximately 2500 donors, OF COAGULOPATHIES
which results in a standard unit of 200 mL of SD plasma with
a coagulation factor profile similar to that of FFP.12–16 In part Patients with a known underlying coagulopathy, as ascer-
due to its cost, and in part due to the lack of availability of tained by an increase in the prothrombin time (PT >16 sec-
virally inactivated RBC and platelet units, SD plasma is not onds) or partial thromboplastin time (PTT > 55 seconds)
used in the United States. exceeding 1.5 to 1.8 times the control value, have an increased
SD plasma is, however, not simply a virally inactivated risk for clinically significant bleeding. In these patients,
equivalent of FFP.17 Levels of factor V, factor VIII, protein administration of FFP has decreased the risk of bleeding and
II S, antiplasmin, and antitrypsin are lower in SD plasma, and reduced or stopped active bleeding. Although FFP contains
this is of potential clinical significance.17–20 The serpin-type all coagulation factors at normal plasma concentrations, it
260 serine proteinase inhibitors such as antiplasmin, antitrypsin, must be recognized that its administration in physiologically
and antithrombin, have a flexible reactive-site loop that can tolerable quantities results in only a 20% to 30% increase in
convert from the active conformation to the inactive latent or the levels of coagulation factors.
polymerized conformations when exposed to heat or deter-
gents or both. Comparisons of conformational stability and Coagulopathy of Liver Dysfunction
inhibitory activity have shown that, in SD plasma, virtually
100% of antiplasmin and approximately 50% of antitrypsin The liver is essential in maintaining normal hemostasis. First,
are in either the latent or polymerized conformation and lack as the principal site of protein synthesis, it supplies the major-
inhibitory activity, whereas in FFP, only the active conforma- ity of proteins involved in the coagulation and fibrinolytic
Indications Contraindications
*
When factor concentrates are not available.
HELLP, hemolysis, elevated liver enzymes, and low platelets; HUS, hemolytic-uremic syndrome; pRBCs, packed red blood cells; TPE, therapeutic
plasma exchange; TTP, thrombotic thrombocytopenic purpura.
FRESH FROZEN PLASMA AND RELATED PRODUCTS
pathways and their regulators (except FVIII, von Willebrand It is well recognized that patients manifesting coagulop-
factor [vWF], tissue plasminogen activator, and plasmino- athy associated with liver disease or hepatic surgery are at
gen activator inhibitor). Second, a process important in the increased risk for bleeding, especially during invasive proce-
normal function of certain coagulation proteins (e.g., pro- dures (e.g., liver biopsy, paracentesis). Therefore, FFP, either
thrombin; FVII, FIX, and FX; proteins C and S), vitamin alone or in conjunction with other products (e.g., platelets,
K–dependent γ-carboxylation of glutamic acid residues, takes prothrombin complex, or antithrombin III concentrate), has
place in the liver. Third, the hepatic reticulum endothelial been used to control active bleeding and to decrease the risk
system is also involved in the clearance of activated coagula- of bleeding complications during invasive procedures.39–43
tion factors, activation complexes, and fibrin and fibrinogen The efficacy of FFP in these situations is assessed clinically
degradation by-products. Therefore the coagulopathy asso- because laboratory evidence such as normalization of PT
ciated with hepatocellular injury has a complex pathogen- or PTT or improvement of the thromboelastogram may
esis including ineffective protein synthesis, consumption of be lacking with the administration of usual quantities of
coagulation factors and inhibitors, and impaired clearance of FFP.39,43,44 Recombinant activated factor VII (rFVIIa) is an
activated coagulation complexes.24–26 In addition, end-stage antihemophilic factor that has shown promise in treating
liver disease is associated with a variable degree of thrombo- coagulopathy in liver disease and, in conjunction with FFP,
cytopenia secondary to hypersplenism27 and a multifactorial is effective in transiently correcting laboratory parameters of
thrombocytopathy.28 coagulopathy in patients with fulminant hepatic failure.45 In
Impairment of hemostasis may also be encountered in these patients, it facilitates the performance of invasive pro-
hepatic surgery, including partial hepatic resection,29 ortho- cedures and is associated with less frequent anasarca com-
topic liver transplantation (OLT), and peritoneovenous or pared with conventional therapy.
LeVeen shunt placement. The coagulation disorder most
frequently observed in these situations is acute and chronic Congenital Coagulation Factor Deficiency
disseminated intravascular coagulopathy (DIC) with a pre-
dominant fibrinolytic component.30,31 The coagulopathy The introduction of specific coagulation factor concentrates
associated with OLT is usually more severe because of the has limited the use of FFP. Currently, FFP is indicated for
underlying coagulopathy predating the transplantation, the patients with rare familial isolated factor deficiencies (e.g.,
profound hyperfibrinolysis characterizing the posttransplan- FV and FXI deficiency) for which no factor concentrate is
tation anhepatic phase, and, in some instances, the coagu- commercially available.46–49 In combined FV-FVIII defi-
lopathy induced by massive transfusion during surgery.32–35 ciency, FFP is the only source of FV, and either FFP infu-
The number of units of blood products transfused during sions or therapeutic plasma exchange (TPE) with FFP has
OLT operations has decreased significantly for reasons that been used to treat bleeding episodes. Congenital combined
include the use of aprotinin to block fibrinolysis and the vitamin K–dependent factor deficiency can be treated with
selective use of the thromboelastogram to monitor coagu- prothrombin complex concentrates, FFP, or occasionally
lation intraoperatively.36 In liver transplantation patients, high doses of vitamin K.50 Dosages and schedules of FFP 19
whole blood, when compared with component therapy administration for congenital factor deficiencies and pos-
(RBCs and FFP), has been reported to provide equally effec- sible alternatives are listed in Table 19–2. 261
tive replacement therapy for blood loss and is associated with
fewer donor exposures.37 Plasmapheresis using FFP, or FFP Warfarin-induced Coagulopathy
and albumin (50%), as replacement fluid may provide an
effective treatment option for primary hepatic allograft non- Vitamin K is a cofactor in the γ-carboxylation of the termi-
function immediately after liver transplantation and may nal glutamic acid residues of certain coagulation (e.g., fac-
obviate the need for retransplantation.38 tors II, VII, IX, and X) and regulatory (e.g., protein C and S)
19
269
Chapter 20
Cryoprecipitate and Related Products
Leon L. Su ● Lennart E. Lögdberg
Cryoprecipitate (cryoprecipitated antihemophilic factor, Cryo) to be obtained from whole-blood collection or plasmapher-
is the common term for the cold-insoluble plasma proteins esis, although the vast majority of cryoprecipitate used for
that precipitate in fresh frozen plasma (FFP) when it is thawed transfusion in the United States comes from whole-blood
to between 1°C and 6°C and then centrifuged, collected, and donors. Plasma donors are required to meet similar and
refrozen. It is one of the four main blood components avail- additional eligibility requirements as whole-blood allogeneic
able in virtually every blood bank in the developed world, donors. A sample of blood collected at the time of donation
but paradoxically, it is likely the component most underuti- is tested for ABO phenotype and communicable infectious
lized even in well-recognized clinical circumstances. diseases. Pathogen-reduction techniques such as nanofil-
The main constituents of cryoprecipitate are the high- tration, solvent-detergent treatment, and methylene blue-
molecular-weight proteins fibrinogen, factor VIII, von photoactivation also may be used to treat source plasma to
Willebrand factor (vWF), factor XIII, and fibronectin. reduce further the risk of infectious disease transmission.
Initially developed in the mid-1960s by Judith Pool and col- Preparation of plasma begins with separation of blood
leagues,1,2 cryoprecipitate revolutionized the treatment of cells by centrifugation to create a cell-free product. If the col-
hemophilia and von Willebrand disease (vWD), previously lection is in CPD, CP2D, or CPDA-1, separated plasma must
dependent on whole blood or fresh plasma. Cryoprecipitate, be frozen at a temperature less than or equal to −18°C within
because of its concentrated factor VIII and factor VIII-vWF 8 hours of collection. For blood collected in ACD, sepa-
complex, could be pooled, and thus its transfusion could rated plasma should be frozen within 6 hours of collection.
provide sufficient factor to prevent or treat serious bleeding Once frozen, cryoprecipitate can be made from FFP within
with minimized risk of volume overload. In the decades after 12 months of collection. The process resumes when FFP is
II the introduction of cryoprecipitate, purified and recom- thawed at 1°C to 6°C to allow formation of the cryoprecipi-
binant factor concentrates were developed and eventually tate. A final separation step using centrifugation or a plasma
270
supplanted cryoprecipitate as the primary treatment for expressor removes the residual precipitate from the thawed
hemophilia and vWD in most cases. plasma, where it is then placed in individual, sterile bags and
Today, cryoprecipitate use has been redefined. It is used resuspended in 10 to 15 mL (up to 20 mL in “wet cryo”) of
primarily to treat acquired fibrinogen deficiencies associ- plasma.4–8 After the extraction process, cryoprecipitate must
ated with active bleeding and either surgery or trauma with be refrozen within 1 hour.
dilution from massive transfusion and fluid resuscitation. It In addition to the traditional method of cryoprecipita-
also is used in congenital deficiency syndromes associated tion production outlined earlier, efforts to improve effi-
with bleeding or as prophylaxis against anticipated bleed- ciency and quality control in component processing led to
ing. Finally, cryoprecipitate is used as a source of fibrinogen, the development of a new automated device capable of gen-
factor XIII, and fibronectin in the commercial production erating cryoprecipitate from plasma. The CryoSeal FS system
of tissue sealants as well as the institutional production of (Thermogenesis Corp. Rancho Cordova, Calif.), currently
allogeneic and autologous fibrin glues. Whereas it is no lon- not for sale in the United States and pending FDA approval,
ger routinely used as the primary treatment for hemophilia, is a compact, computerized device designed to produce fibrin
vWD, and uremic coagulopathy, cryoprecipitate continues to sealant in the autologous preoperative setting.9 Alternatively,
have a role as second-line therapy in instances in which pri- the CryoSeal system is an automated device with tightly
mary treatment is ineffective or not available. controlled temperature cycling capable of producing cryo-
precipitate within 1 hour.10,11 The fibrinogen and factor VIII
content in cryoprecipitate prepared by the CryoSeal system is
PRODUCTION, PROCESSING, STORAGE, comparable to that with traditional production methods.10,11
AND QUALITY CONTROL Advantages of the CryoSeal system over manual preparation
include standardization and faster production of cryopre-
The source and quality of plasma used to make cryoprecip- cipitate.
itate is controlled by guidelines set forth by the Food and Current American Association of Blood Banks (AABB)
Drug Administration (FDA) as published in the Code of Standards require cryoprecipitate to be stored at −18°C,
Federal Regulations (CFR).3 Governing the quality of source transported in a frozen state, and used within 12 months
plasma ensures that production of cryoprecipitate contains of the original date of collection.12 As a final measure to
adequate amounts of biologically active product while mini- assure quality of cryoprecipitate, the CFR dictates that qual-
mizing contamination from infectious and cellular compo- ity of antihemophilic factor and fibrinogen should be tested
nents. Current guidelines from the CFR allow source plasma monthly on at least four representative containers of cryo-
precipitate.3 With every quality-assurance testing, the AABB
Primary
Acquired/congenital hypofibrinogenemia16,38–39,43–44,47
Massive transfusion with bleeding48–51
Component for tissue sealants or fibrin glues23–24,72–92
Factor XIII deficiency60–65
Reversal of thrombolytic therapy93,97
Secondary
Hemophilia A4,94
von Willebrand disease34,52–59
Uremic coagulopathy66–71
Common Misuses97,98
Fibrinogen replacement with normal fibrinogen levels and no evidence of increased dysfunction or consumption
Reversal of warfarin therapy
Treatment of impaired surgical hemostasis in the absence of hypofibrinogenemia
Treatment of hepatic coagulopathy with multiple factor deficiencies
Common Underutilization48–51
Massive transfusion with bleeding
In a review, Humphries41 reported that the most common acetate [1-deamino (8-d-arginine) vasopressin] (DDAVP)
use of cryoprecipitate as fibrinogen replacement is in acquired enhances the release of vWF from the endothelial cells and
fibrinogen deficiency. Dose and frequency of administration augments platelet activity. DDAVP is considered effective
depend on the rate of consumption or destruction of fibrin- therapy in most types of vWD. In approximately 10% to 20%
ogen, and this can be assessed by monitoring the fibrinogen of patients with vWD, however, DDAVP is either ineffective
level. In certain coagulopathies, repletion of other coagula- (types 1 and 3) or contraindicated (types 2B and 2N), or the
tion proteins in addition to fibrinogen may necessitate the development of tachyphylaxis prevents its prolonged use. In
use of FFP.16,42–47 these circumstances, cryoprecipitate can provide adequate
replacement of both vWF and factor VIII:C to shorten the
Massive Transfusion bleeding time and to control clinical bleeding.54–57 In the past
20 years, lyophilized and pasteurized plasma concentrates
The coagulopathy associated with massive transfusion in (e.g., intermediate-purity and high-purity plasma-derived
bleeding is often multifactorial. However, the primary cause factor VIII concentrates) rich in high-molecular-weight 20
is usually coagulation factor deficiencies from dilution and vWF and factor VIII, have been developed and have largely
increased factor consumption, as in disseminated intravas- replaced cryoprecipitate in the treatment of vWD.58,59 273
cular coagulation or normal coagulation.48 Situations of
major blood loss with massive transfusion often lead to an
Factor XIII Deficiency
initial decrease in fibrinogen levels below 100 mg/dL after
loss of 1.5 blood volumes followed by a decrease to less than Through the dual action of thrombin, fibrinogen is con-
25% activity of labile coagulation factors after loss of two verted into fibrin, and factor XIII is activated; this process
blood volumes.49,50 In some cases, rapid consumption and facilitates the cross-linking of polymerized fibrin and α2-
dilution of fibrinogen is associated with massive transfusion antiplasmin to fibrin clots. In the neonatal period, defi-
and bleeding, FFP dosing without additional cryoprecipi- ciency of this hemostatic function (factor XIII deficiency)
tate may not be sufficient to maintain hemostatic fibrinogen manifests as protracted umbilical stump bleeding, whereas
levels. Although FFP does contain fibrinogen, the amount later in life, postsurgical bleeding and delayed or abnormal
provided this way may be insufficient to maintain adequate (keloid) wound healing may be observed. Factor XIII also
levels during massive transfusion and bleeding, in turn lead- has been linked to the development of implantation tissue
ing to delayed correction and excessive volumes.51 As a result, during pregnancy, and its deficiency is believed to be the
use of cryoprecipitate should always be considered in mas- cause of repeated miscarriages.60 In these cases of factor XIII
sive transfusion with bleeding and should be given early in deficiency, prophylactic replacement therapy with plasma-
the course along with FFP. derived factor XIII concentrates such as Fibrogammin P61,62
(available in certain countries but currently not licensed in
von Willebrand Disease the United States) or cryoprecipitate is feasible.63–65
Tisseel Crosseal
Approved Indications
As adjunct to hemostasis in cardiopulmonary bypass surgery As adjunct to hemostasis in liver surgery
As a hemostatic agent in fully heparinized patients undergoing
cardiopulmonary bypass surgery
As adjunct in the closure of colostomies
In the control of bleeding associated with splenic injury
Antifibrinolytic Agent
Bovine aprotinin Tranexamic acid
Contraindications
Patients with a known hypersensitivity to bovine proteins or Patients with known anaphylactic or severe systemic
reactions to human blood products reactions to human blood products
Injection into circulation or tissues Surgical procedures in which contact with cerebrospinal
Massive and brisk arterial bleed fluid or dura mater could occur
Injection into circulation or tissues
Massive and brisk arterial bleeding
Potential Adverse Reactions
Severe allergic/anaphylactoid reactions Severe allergic/anaphylactoid reactions
Thromboembolic events (if injected into circulation or tissue) Transmission of infectious agents (e.g., viruses)
Transmission of infectious agents (e.g., viruses)
CRYOPRECIPITATE AND RELATED PRODUCTS
studies have yet to prove or disprove the use of fibronectin SUMMARY
on wound healing, and its use in this area remains unclear.
Cryoprecipitate is given primarily to treat acquired and con-
genital hypofibrinogenemia and is used in the production of
COMMON MISUSES AND fibrin glues and commercial sealants. Less frequently, it may
UNDERUTILIZATION OF CRYOPRECIPITATE be used to treat factor XIII deficiency and to reverse thrombo-
lytic therapy. Cryoprecipitate may also be used as secondary
Audits of the use of cryoprecipitate have been reported in treatment for hemophilia, vWD, and uremic coagulopathy
the literature and demonstrate that still considerable misuse or as initial treatment in situations in which specific factor
of cryoprecipitate occurs in the clinical setting. Recent stud- concentrates or DDAVP are unavailable. Underutilization of
ies show patterns of inappropriate use involving up to 24% cryoprecipitate in clinically appropriate circumstances and
to 62% of all cryoprecipitate orders.97,98 Commonly encoun- inappropriate use continue to occur at a high rate, according
tered misuses include fibrinogen replacement in patients to institutional audits. This pattern of use underscores the
with normal fibrinogen levels or who do not have current need for continued education of ordering physicians on the
laboratory results and no evidence of increased dysfunction appropriate use of cryoprecipitate.
or consumption, reversal of warfarin therapy, treatment of
impaired surgical hemostasis in the absence of hypofibrino-
gemia, and the treatment of hepatic coagulopathy with mul- REFERENCES
tiple factor deficiencies.
In addition, underutilization of cryoprecipitate in the 1. Pool JG, Shannon AE. Production of high potency concentrates of anti-
hemophilic globulin in a closed bag system assay in vitro and in vivo.
setting of massive transfusion with bleeding commonly N Engl J Med 1965;273:1443–1447.
occurs in the intraoperative setting. The inappropriately low 2. Hershgold EJ, Pool JG, Pappenhagen AR. The potent antihemophilic
use of cryoprecipitate, and in some cases FFP, in the mas- globulin concentrate derived from a cold insoluble fraction of human
sive transfusion setting may stem from a once-held notion plasma: characterization and further data on preparation and clinical
trial. J Lab Clin Med 1966;67:13–32.
that coagulation deficiency occurred late in massive transfu- 3. Code of Federal Regulations, 21 CFR 640 Subpart F. Washington, D.C.,
sion and after the development of thrombocytopenia. This Government Printing Office, 2004.
belief, although probably valid when whole blood was used 4. American Association of Blood Banks, America’s Blood Centers, American
predominantly in massive transfusion, no longer reflects Red Cross. Circular of information for the use of human blood and blood
current practice, because today’s blood products are mostly components, 2002, 26–28. http://www.fda.gov/cber/gdlns/circbld.pdf.
5. Brecher ME (ed). Technical Manual, 15th ed. Bethesda, Md., American
packed red blood cells (pRBCs) with significantly less plasma Association of Blood Banks, 2005.
volumes.48,50 As a result, coagulation deficiencies appear very 6. Myllyla G. Factors determining quality of plasma. Vox Sang 1998;74:507.
early on in massive transfusion, with fibrinogen being one 7. Allersma DP, Imambaks RMR, Meerhof LJ. Effect of whole blood stor-
of the first coagulation proteins to fall below hemostatic lev- age on factor VIII recovery in fresh frozen plasma and cryoprecipitate.
Vox Sang 1996;71:150.
els.49,50 Recent observations in the clinical setting suggest that 8. Hughes C, Thomas KB, Schiff P, et al. Effect of delayed blood processing
20
during massive transfusion with bleeding, fibrinogen levels on the yield of factor VIII in cryoprecipitate and factor VIII concentrate.
below 100 mg/dL may still persist despite repletion with Transfusion 1988;28:566. 275
pRBCs, FFP, and platelets, resulting in a persistent coagu- 9. Buchta C, Dettke M, Funovics PT, et al. Fibrin sealant produced by
lopathy. Recognition of the rapid consumption and dilution the CryoSeal® FS System: Product chemistry, material properties and
possible preparation in the autologous preoperative setting. Vox Sang
of fibrinogen in the setting of massive transfusion and bleed- 2004;86:257–262.
ing underscores the importance of frequent monitoring of 10. Gosselin RC, Larkin E, Owings JT, et al. CryoSealTM System, a new device
fibrinogen levels and vigilant use of cryoprecipitate with for generating cryoprecipitate from plasma. Clin Chem 1997;43:1782–
other blood products to maintain hemostasis. 1783.
11. Rock G, Berger R, Lange J, et al. A novel, automated method of tempera-
The incidence and pattern of inappropriate and insuffi- ture cycling to produce cryoprecipitate. Transfusion 2001;41:232–235.
cient use of cryoprecipitate signify an alarming misunder- 12. AABB. 5.1.8A Standards for Blood Banks and Transfusion Services,
standing of the use of cryoprecipitate among some clinicians 23rd ed. Bethesda, Md., American Association of Blood Banks, 2005,
and suggest the need for further education with respect to pp 58–59.
acceptable applications of this blood component. 13. AABB. 5.7.5.14 Standards for Blood Banks and Transfusion Services,
23rd ed. Bethesda, Md., American Association of Blood Banks, 2005,
p 30.
14. Saxena S, Odono V, Francis RB Jr, et al. Storage of thawed cryoprecipi-
POTENTIAL ADVERSE REACTIONS tated AHF is better at room temperature than at 1˚C to 6˚C for factor
VIII content. Arch Pathol Lab Med 1991;115:343–345.
Cryoprecipitate administration may be associated with 15. Sherman LA. In vivo stability of cryoprecipitate fibrinogen. Transfusion
1977;17:19–61.
nonspecific side effects related to its protein, cytokine, and 16. Ness PM, Perkins HA. Cryoprecipitate as a reliable source of fibrinogen
isohemagglutinin content. Adverse reactions include fever, replacement. JAMA 1979;241:1690.
chills, and allergic reactions of variable severity. In situations 17. Thompson HW, Touris S, Giambartolomei S, et al. Treatment of con-
in which large volumes of ABO-incompatible cryoprecipitate genital afibrinogenemia with cryoprecipitate collected through a plas-
mapheresis program using dedicated donors. J Clin Apheresis 1998;13:
are given, adult recipients may develop a positive direct anti- 143–145.
globulin test (DAT), and in rare circumstances, hemolytic 18. Basu S, Marini CP, Bauman FG, et al. Comparative study of biological
anemia.4,99 Untoward effects of tissue fibrin adhesives such glues: cryoprecipitate glue, two-component fibrin sealant, and “French”
as home-made fibrin glues have been related to the source glue. Ann Thorac Surg 1995;60:1255–1262.
of thrombin, especially of bovine origin. These include 19. Cohn SM, Feinstein AJ, Nicholas JM, et al. Recipe for poor man’s fibrin
glue. J Trauma 1998;44:507.
the development of allergic-type reactions, some as severe 20. Berguer R, Staerkel RL, Moore EE, et al. Warning: Fatal reaction to
as anaphylaxis, and factor V antibodies, which may lead to the use of fibrin glue in deep hepatic wounds: case reports. J Trauma
abnormal prothrombin times and postsurgical bleeding.20–22 1991;126:432.
BLOOD BANKING 21. Streiff MB, Ness PM. Acquired FV inhibitors: a needless iatrogenic com- 51. Stainsby D, MacLennan S, Hamilton PJ. Management of massive blood
plication of bovine thrombin exposure. Transfusion 2002;42:18–25. loss: a template guide. Br J Anaesth 2000;85:387–491.
22. Ortel TL, Mercer MC, Thames EH, et al. Immunologic impact and 52. Fischer BE, Kramer G, Mitterer A, et al. Effect of multimerization of
clinical outcomes after surgical exposure to bovine thrombin. Ann Surg human and recombinant von Willebrand factor on platelet aggregation,
2001;233:88–96. binding to collagen and binding of coagulation factor VIII. Thromb Res
23. Tisseel VH, Kit Insert. Two-component fibrin sealant, vapor-heated, 1996;84:55.
kit, manufactured by Osterreichisches Institut für Haemoderivate GES, 53. Sadler JE, Matsushita T, Dong Z, et al. Molecular mechanism and clas-
M.B.H., Subsidiary of Immuno AG. Distributed by Baxter Healthcare sification of von Willebrand disease. Thromb Haemost 1995;74:161.
Corporation, Deerfield, Ill., 1998. 54. Rodeghiero F, Castaman G, diBona E, et al. Consistency of responses
24. American Red Cross. Crosseal fibrin sealant (human) package insert. to repeated DDAVP infusions in patients with von Willebrand’s disease
Washington, DC, American Red Cross, 2003. and hemophilia A. Blood 1989;74:1997.
25. Dodd RA, Cornwell R, Holm NE, et al. The Vivostat® application system: 55. Lusher JM. Response to 1-deamino-8-arginine vasopressin in von Wil-
a comparison with conventional fibrin sealant application systems. Tech lebrand disease. Haemostasis 1994;24:276.
Health Care 2002;10:401–411. 56. Castaman G, Rodeghiero F. Current management of von Willebrand’s
26. Allain JP. Non-factor VIII related constituents in concentrates. Scand disease: practical therapeutics. Drugs 1995;602:14.
J Haematol 1984;33:173–180. 57. Mannucci PM. Biochemical characteristics of therapeutic plasma con-
27. Hornsey VS, Young DA, Docherty A, et al. Cryoprecipitate prepared centrates used in the treatment of von Willebrand disease. Haemostasis
from plasma treated with methylene blue plus light: increasing the 1994;24:285.
fibrinogen concentration. Transfus Med 2004;14:569. 58. Mannucci PM, Tenconi PM, Castaman G, et al. Comparison of four
28. Farrugia A, Grasso S, Douglas S, et al. Modulation of fibrinogen content virus-inactivated plasma concentrates for treatment of severe von Wil-
in cryoprecipitate by temperature manipulation during plasma process- lebrand disease: a cross-over randomized trial. Blood 1992;79:3130.
ing. Transfusion 1992;32:755–759. 59. Rodeghiero F, Castaman G, Meyer D, et al. Replacement therapy with
29. Farrugia A, Prowse C. Studies on the procurement of blood coagulation virus-inactivated plasma concentrates in von Willebrand disease. Vox
factor VIII: effects of plasma freezing rate and storage conditions on Sang 1992;62:193.
cryoprecipitate quality. J Clin Pathol 1985;38:433–437. 60. Asahina T, Kobayashi T, Okada Y, et al. Maternal blood coagulation fac-
30. DePalma L, Criss VR, Luban NL. The preparation of fibrinogen con- tor XIII is associated with the development of cytotrophoblastic shell.
centrate for use as fibrin glue by four different methods. Transfusion Placenta 2000;21:388–393.
1993;33:717–720. 61. Gootenberg JE. Factor concentrates for the treatment of factor XIII defi-
31. Keeling DM, Luddington R, Allain JP, et al. Cryoprecipitate prepared ciency. Curr Opin Hematol 1998;5:372–375.
from plasma virally inactivated by the solvent detergent method. Br 62. Dreyfus M, Arnuti BB, Borg P, et al. Safety and efficacy of Fibrogammin
J Haematol 1997;96:194–197. P®)for the treatment of patients with severe FXIII deficiency [Abstract].
32. Piedras J, Sanchez-Montero PE, Herrera FM, et al. Effect of plasma J Thromb Haemost 2003;1(suppl 1):abstract P0299.
freezing temperature, anticoagulant and time of storage on factor VIII: 63. Stirling D, Ludlam CA. Therapeutic concentrates for the treatment of
C activity in cryoprecipitate. Arch Med Res 1993;24:13–26. congenital deficiencies of factors VII, XI, XIII. Semin Thromb Hemost
33. Seghatchian J, Krailadsiri P. What’s happening? The quality of methylene 1993;19:48.
blue treated FFP and cryo. Transfus Apheresis Sci 2001;25:327–331. 64. Fear JD, Miloszewski KJA, Losowsky MS. The half life of factor XIII in the
34. Pomper GJ, Rick ME, Epstein JS, et al. Management of severe vWD with management of inherited deficiency. Thromb Haemost 1983;49:102.
cryoprecipitate collected by repeated apheresis of a single dedicated 65. Rodeghiero F, Castaman GC, Di Bona E, et al. Successful pregnancy in a
donor. Transfusion 2003;43:1515–1521. woman with congenital factor XIII deficiency treated with substitutive
35. Saxena S, Odono V, Francis RB Jr, et al. Can storage of thawed cryo- therapy. Blut 1987;55:45.
precipitate be extended to more than six hours? Am J Clin Pathol 66. Remuzzi G. Bleeding disorders in uremia: pathophysiology and treat-
II 1990;94:203–206. ment. Adv Nephrol 1989;18:171.
36. Spivey MA, Jeter EK, Lazarchick J, et al. Postfiltration factor VIII and 67. Andrassy K, Ritz E. Uremia as a cause of bleeding. Am J Nephrol
276 fibrinogen levels in cryoprecipitate stored at room temperature and at 1985;5:313.
1 to 6˚C. Transfusion 1992;32:340–343. 68. Janson PA, Jubelirer SJ, Weinstein MJ, et al. Treatment of the bleeding
37. Fresh-frozen Plasma, Cryoprecipitate, and Platelets Administration tendency in uremia with cryoprecipitate. N Engl J Med 1980;303:1318.
Practice Guidelines, Development Task Force of the College of Ameri- 69. Maierhoter W, Adams MB, Kleinman JG, et al. Treatment of the bleed-
can Pathologists. Practice parameter for the use of fresh-frozen plasma, ing tendency in uremia with cryoprecipitate [Letter]. N Engl J Med
cryoprecipitate, and platelets. JAMA 1994;271:777–781. 1981;305:645.
38. Galanakis DK. Fibrinogen anomalies and disease: a clinical update. 70. Juhl A. DDAVP, cryoprecipitate, and highly “purified” factor VIII con-
Hematol Oncol Clin North Am 1992;6:1171. centrate in uremia. Nephron 1986;43:305.
39. Mammen EF. Fibrinogen abnormalities. Semin Thromb Hemost 71. Triulzi DJ, Blumberg N. Variability in response to cryoprecipitate treat-
1983;9:1. ment for hemostatic defects in uremia. Yale J Biol Med 1990;63:1.
40. Forbes CD, Madhok R. Genetic disorders of blood coagulation: clini- 72. Noun R, Elias D, Balladur P, et al. Fibrin glue effectiveness and tolerance
cal manifestation and management. In Ratnoff OD, Forbes CD (eds): after elective liver resection: a randomized trial. Hepatogastroenterol-
Disorders of Hemostasis. Philadelphia, WB Saunders, 1991, p 141. ogy 1996;43:221–224.
41. Humphries JE. Transfusion therapy in acquired coagulopathies. Trans- 73. Kohno H, Nagasue N, Chang YC, et al. Comparison of topical hemo-
fus Med 1994;8:1181. static agents in elective hepatic resection: a clinical prospective random-
42. Nusbacher J. Blood transfusion support in liver transplantation. Trans- ized trial. World J Surg 1992;16:966–969.
fusion 1991;5:207. 74. Shirai T, Shimota H, Chida K, et al. Anaphylaxis to aprotinin in fibrin
43. Colman RW, Robboy SJ, Minna JD. Disseminated intravascular coagu- sealant. Intern Med 2005;44:1088–1089.
lation: a reappraisal. Annu Rev Med 1979;30:359. 75. Sierra DH. Fibrin sealant adhesive systems: a review of their chemis-
44. Francis JL, Armstrong DJ. Acquired dysfibrinogenemia in liver disease. try, material properties and clinical applications. J Biomater Appl 1993;
J Clin Pathol 1982;35:667. 7:309.
45. Gilabert J, Estelles A, Asnar J, et al. Abruptio placentae and disseminated 76. Spotnitz WD. Fibrin sealant in the United States: clinical use at the Uni-
intravascular coagulation. Acta Obstet Gynecol Scand 1985;64:35. versity of Virginia. Thromb Haemost 1995;74:482.
46. Sutton DM, Hauser R, Kulapongs P, et al. Intravascular coagulation in 77. Martinowtz U, Spotnitz WD. Fibrin tissue adhesives. Thromb Haemost
abruptio placentae. Am J Obstet Gynecol 1971;109:604. 1997;78:661.
47. Al-Mondhiry H, Ehmann WC. Congenital afibrinogenemia. Am 78. Radosevich M, Goubran HA, Burnouf T. Fibrin sealant: Scientific ratio-
J Hematol 1994;46:343–347. nale, production methods, properties, and current clinical use. Vox Sang
48. Ho AM, Karmakar MK, Dion PW. Are we giving enough coagulation fac- 1997;72:133.
tors during major trauma resuscitation? Am J Surg 2005;190:479–484. 79. Martinowitz U, Schulman S, Horoszowski H, et al. Role of fibrin seal-
49. Ciavarella D, Reed RL, Counts RB, et al. Clotting factor levels and ants in surgical procedures on patients with hemostatic disorders. Clin
the risk of diffuse microvascular bleeding in the massively transfused Orthop 1996;328:65.
patient. Br J Haematol 1987;67:365–368. 80. Saltz R, Dimick A, Harris C, et al. Application of autologous fibrin glue
50. Hiippala ST, Myllyla GJ, Vahtera EM. Hemostatic factors and replace- in burn wounds. J Burn Care Rehabil 1989;10:504.
ment of major blood loss with plasma-poor red cell concentrates. 81. Saltz R, Guzman G. Aesthetic reconstruction of burned hands. Plast
Anesth Analg 1995;81:360–365. Surg 1992;11:23.
CRYOPRECIPITATE AND RELATED PRODUCTS
82. Marchac D, Pugash E, Gault D. The use of sprayed fibrin glue for face 91. Reddy UM, Shah SS, Nemiroff RL, et al. In vitro sealing of punctured
lifts. Eur J Plast Surg 1987;10:139. fetal membranes: potential treatment for midtrimester premature rup-
83. Vogel A, O’Grady K, Toriumi DM. Surgical tissue adhesives in facial ture of membranes. Am J Obstet Gynecol 2001;185:1090–1093.
plastic and reconstructive surgery. Facial Plast Surg 1993;9:76. 92. Quintero RA, Morales WJ, Allen M, et al. Treatment of iatrogenic previ-
84. Shirai T, Amano J, Takabe K. Thoracoscopic diagnosis and treatment of able premature rupture of membranes with intra-amniotic injection of
chylothorax after pneumonectomy. Ann Thorac Surg 1991;52:306. platelets and cryoprecipitate (amniopatch): preliminary experience. Am
85. Shaffrey CI, Spotnitz WD, Shaffrey ME, et al. Neurosurgical applica- J Obstet Gynecol 1999;181:744–749.
tions of fibrin glue: augmentation of dural closure in 134 patients. Neu- 93. Sane DC, Califf RM, Topol EJ. Bleeding during thrombolytic therapy for
rosurgery 1990;26:207. acute myocardial infarction: mechanisms and management. Ann Intern
86. Martinowitz U, Ozer Y, Varon D, et al. Fibrin sealing in nerve repair. Med 1989;111:1010–1022.
Thromb Haemost 1993;69:1287. 94. Srivastava A. Factor replacement therapy in haemophilia: are there
87. Egkher E, Spangler H, Spangler HP. Indications and limits of fibrin models for developing countries? Haemophilia 2003;9:391.
adhesive applied to traumatological patients. In Schlag G, Redl H (eds): 95. Powell FS, Doran JE. Current status of fibronectin in transfusion medi-
Fibrin Sealant in Operative Medicine: Traumatology and Orthopaedics. cine: focus on clinical studies. Vox Sang 1991;60:193–202.
Berlin, Springer, 1986, pp 7, 144. 96. Steinbaum SS, Cucinell S. Effects of cryoprecipitate on the healing of
88. Lagoutte FM, Gauthier L, Comte PRM. A fibrin sealant for perforated chronic wounds. Mil Med 1994;159:105.
and pre-perforated corneal ulcers. Br J Ophthalmol 1989;73:757. 97. Pantanowitz L, Kruskall M, Uhl L. Cryoprecipitate patterns of use. Am
89. Kang DR. Fibrin tissue adhesive use in costal cartilage laryngotracheal J Clin Pathol 2003;119:874–881.
reconstruction. Paper presented at the Cambridge Symposium on 98. Schofield WN, Rubin GL, Dean MG. Appropriateness of platelet, fresh
Tissue Sealants: Current Practice, Future Uses, La Jolla, Calif., 1996. frozen plasma and cryoprecipitate transfusion in New South Wales pub-
90. Fischer CP, Sonda LP, Dionko AC. Further experience with cryopre- lic hospitals. Med J Aust 2003;178:117–121.
cipitate coagulum in renal calculus surgery: a review of 60 cases. J Urol 99. Goodnight SH Jr. Cryoprecipitate and fibrinogen [Editorial]. JAMA
1981;126:432. 1979;241:1716–1717.
20
277
Chapter 21
Albumin
Elizabeth E. Culler ● Lennart E. Lögdberg
From Code of Federal Regulations. Title 21 CFR 640. Washington, D.C., U.S. Government Printing Office, 2005 (revised annually).
ALBUMIN
hemodynamic responses in volunteers receiving either recom- situations. When evaluating the appropriate dose of albumin
binant human albumin or human serum albumin.24 Thirty to administer, study investigators have relied on parameters
participants in the double-blind, randomized trial received such as serum albumin level, urine output, pulse, blood
intravenous doses of 10 g on day 1, 20 g on day 22, and 50 g pressure, hematocrit, and degree of venous and pulmonary
on day 43 of either recombinant human albumin or human congestion.26
serum albumin. There were no significant differences in safety
or tolerability between the two products. The serum albumin,
Rate
colloid osmotic pressure, and hematocrit pre- and postinfusion
were measured and were not significantly different between There are no guidelines addressing the optimal infusion rate
the groups receiving either recombinant human albumin or for albumin solutions. The infusion rate should be based on
human serum albumin. Since recombinant human albumin is the patient’s condition and is limited only by the capacity of
virus- and prion-free, there is incentive for manufacturers to the administration set when an albumin infusion is needed
continue the clinical development of this product. emergently. Because high rates of albumin infusion can cause
circulatory overload and pulmonary edema, 5% albumin
solutions are commonly started at a rate of 1 to 2 mL/min
DOSING AND ADMINISTRATION and are not usually infused at a faster rate than 4 mL/min,
and 25% albumin solutions are not infused at rates faster
than 1 mL/min.26
Choosing a Product
Albumin solutions with protein concentrations of 5%, 20%, Administration
and 25% are currently available on the U.S. market. Because
PPF solutions contain a greater proportion of proteins other Albumin should be inspected for turbidity prior to admin-
than albumin, they are not often used. Thus, the focus of this istration. Although albumin does not have to be infused
section is on albumin solutions. through a filter, some manufacturers either recommend or
The infusion of 5% albumin solutions, which are iso- include a filter in administration sets to be used during albu-
oncotic with human plasma, increases the plasma volume by min administration. Hospital policy also may require the use
the volume of albumin solution infused,8 while the infusion of a filter.15 Administration must begin within 4 hours of
of the hyperoncotic 25% albumin solutions causes the plasma entry into the container.1
volume to expand by 3.5 times the volume of albumin solu- Because blood group isohemagglutinins are removed
tion infused.25 For treating pediatric patients or volume- or from albumin products during preparation, albumin is given
sodium-sensitive patients, the more concentrated 20% or without regard to ABO type. The Code of Federal Regulations
25% albumin solutions are more commonly used, whereas (CFR) does not address the measurement of isohemag-
the 5% albumin solutions are especially useful for hypovole- glutinin titers in albumin products; however, the European
mic patients.26 Dehydrated patients usually require additional Pharmacopoeia states that plasma products intended for 21
fluids along with the 20% or 25% albumin solutions.1 intravenous use should have an isohemagglutinin titer of
If necessary, a 5% albumin solution can be prepared by less than 1:64.12 281
diluting a 20% or 25% albumin solution with either normal
saline or 5% dextrose. Sterile water should not be used to
dilute albumin because the resulting hypotonic solution can CLINICAL CONSIDERATIONS
cause hemolysis when infused. This has occurred in at least
10 patients with one reported death.27 When large volumes Hypoalbuminemia may result from the decreased produc-
of diluted albumin are required, normal saline is the dilu- tion, altered distribution, increased metabolism, or exces-
ent of choice, since the infusion of large volumes of albumin sive loss of albumin. Some causes of decreased albumin
diluted with dextrose 5% can cause hyponatremia leading to production are liver dysfunction, malnutrition, and mal-
cerebral edema.28 absorption. Approximately 40 cases of congenital analbu-
When choosing a product, clinicians must also con- minemia (defined as HSA < 1 g/L) have been reported in
sider the aluminum concentration. As mentioned in the the literature; however, this disorder is usually associated
section on Potential Adverse Effects, albumin products with only mild signs and symptoms.32 Hypoalbuminemia
contain small amounts of aluminum, which can accu- also may result from a redistribution of albumin to the
mulate in premature infants or in patients with chronic extravascular space as a result of increased vascular per-
renal failure.10 Talecris Biotherapeutics produces low-alu- meability, as seen in inflammatory states. Thyrotoxicosis
minum formulations of 5%, 20%, and 25% human albu- and pancreatitis are two conditions associated with
min, each of which has an aluminum content of less than increased albumin catabolism. Increased albumin loss
200 μg/L.29–31 occurs in patients with protein-losing gastroenteropathy
and nephrotic syndrome.
Dose
Indication Guidelines for Albumin Usage
According to the American Hospital Formulary Service
(AHFS) Drug Information guide, a typical initial adult dose Although decreased albumin levels are present in many
of albumin is 25 g, which can be repeated in 15 to 30 minutes conditions, albumin infusion is not usually required to
depending on the patient’s response. Up to 250 g of albu- treat hypoalbuminemia. Rather, albumin infusions are used
min may be infused in a 48-hour period.26 However, because therapeutically for plasma expansion. Historically, this led
studies have used different end points to assess clinical to broad indications and widespread use of the product.
improvement after the administration of albumin, no stan- Later studies showed albumin to be ineffective in many of
dard dose of albumin can be recommended to fit all clinical these uses, leading to a still-ongoing evolution toward more
BLOOD BANKING
Table 21–2 UHC Guidelines on Albumin Usage
II
From Technology assessment: albumin, nonprotein colloid, and crystalloid solutions. Oak Brook, JH, University HealthSystem Consortium,
2000. Aboulghar M, Evers JH, Al-Inany H. Intravenous albumin for preventing severe ovarian hyperstimulation syndrome. Cochrane Database
282 Syst Rev 2002:CD001302.
conservative indication guidelines such as the most recent Alternatives to Albumin for Plasma
UHC guidelines for the use of albumin (summarized and
Expansion
updated in Table 21–2), nonprotein colloid, and crystal-
loid solutions.5 Reflecting the liberal guidelines of the past, The two major categories of products that may be used for
a 2003 report that examined albumin prescribing patterns plasma expansion are crystalloids (e.g., 0.9% sodium chlo-
in 53 member institutions of the University HealthSystem ride, Ringer’s lactate) and colloids, including protein (e.g.,
Consortium found that albumin was inappropriately used in albumin) and nonprotein substances (e.g., dextrans, gela-
57.8% of adult patients and 52.2% of pediatric patients. In tins, and starches). Crystalloids have not demonstrated a
the report, which collected data on 1649 adult and 23 pedi- definite clinical advantage over albumin, but are consider-
atric patients receiving albumin, two of the most common ably cheaper and are therefore widely used as first-line treat-
inappropriate uses of albumin were for intradialytic blood ment.33 Although nonprotein colloids have been associated
pressure support (159 patients) and for serum albumin val- with side effects such as alterations in hemostatic laboratory
ues of <2 g/dL (142 patients).6 The consensus of the panelists results, pruritus, and, rarely, with severe head and back pain,
involved in creating the 2000 UHC guidelines for albumin they also are less expensive than albumin and are preferred
use was that the available evidence did not support the use by some.34–39 In those clinical conditions in which plasma
of albumin in these situations. Table 21–3 provides a list of expansion through albumin or nonprotein colloids has
common misuses of therapeutic albumin infusions. demonstrated equivalent patient outcomes, the 2000 UHC
In the following sections, we discuss alternatives to the use of guidelines recommend the use of the latter due to their lower
albumin for plasma expansion and then summarize some of the cost. Neither crystalloids nor colloids can be substituted for
well-recognized clinical indications for albumin infusions, not- red blood cells when oxygen-carrying capacity is needed or
ing that such infusions are often a secondary treatment option. for platelets or plasma when coagulopathy exists.5
ALBUMIN
refractory to first-line treatment, which consists of a sodium-
Table 21–3 Common Misuses of Albumin
restricted diet and high-dose diuretics.51 These patients’
Infusions
treatment options include serial therapeutic paracenteses,
transjugular intrahepatic portosystemic stent shunt (TIPS),
peritoneovenous shunts, and liver transplantation.51
For cirrhotic patients with refractory ascites who require
serial therapeutic paracenteses, controversy exists concern-
ing whether volume expansion is useful and, if so, which
expander is most effective. A study by Gines and colleagues
found that paracentesis in cirrhotic patients with tense
ascites without albumin infusion resulted in significant
From Technology assessment: albumin, nonprotein colloid, increases in blood urea nitrogen, plasma renin activity,
and crystalloid solutions. Oak Brook, Ill: University HealthSystem
Consortium, 2000. Tanzi M, Gardner M, Megellas M, et al.
and plasma aldosterone concentration, whereas patients
Evaluation of the appropriate use of albumin in adult and pediatric who received postparacentesis albumin did not experience
patients. Am J Health Syst Pharm 2003;60:1330–1335. Tarin those changes.52 There were no significant differences in
Remohi MJ, Sanchez Arcos A, Santos Ramos B, et al. Costs related mortality between the groups. On the other hand, it has
to inappropriate use of albumin in Spain. Ann Pharmacother been shown that a single paracentesis of 4 to 6 L may be
2000;34:1198–1205.
performed as a short-term option without albumin infu-
sion in patients with tense, diuretic-resistant ascites.51,53
Thus, given the safety of paracentesis of smaller volumes,
Clinical Usage of Albumin Infusions the guidelines released by the American Association for the
Study of Liver Diseases (AASLD) in 1998 suggest that post-
Therapeutic Plasma Exchange paracentesis albumin infusion is unnecessary for removed
Albumin is commonly used as the replacement fluid in volumes of less than 4 to 5 L but that an albumin infusion
therapeutic plasma exchange (TPE) unless a condition can be considered for larger volume paracenteses.51
exists that specifically requires factors present in fresh Alternatives to serial paracentesis with albumin infusion
frozen plasma (FFP).40 Albumin solutions have a lower have been studied, and none has clearly demonstrated a
probability of viral transmission and a decreased risk of better patient outcome. Albumin infusion more effectively
citrate-induced hypocalcemia than FFP.41 According to prevents hemodynamic deterioration than the infusion of
the Circular of Information, FFP should not be used when other plasma expanders such as dextran 70 and polygel-
other volume expanders can safely and adequately replace ine after large-volume paracentesis.54 Results are conflict-
blood volume.42 ing when comparing treatment of cirrhotic patients with
Although albumin is a frequently used replacement ascites refractory to diuretic treatment, with either TIPS
fluid in TPE, cryopoor plasma or FFP are the preferred or large-volume paracenteses followed by albumin, using 21
replacement fluids in thrombotic thrombocytopenic pur- survival without transplantation as outcome.55–57 Because
pura (TTP) and related disorders.43 In TTP, a metallopro- no other treatment has demonstrated a superior patient 283
tease (ADAMTS-13) that usually cleaves von Willebrand outcome, the 1998 AASLD guideline advising that albumin
factor (vWF) into multimers is rendered ineffective by infusion be considered for large-volume paracentesis still
antibody inhibitors or by mutations in the ADAMTS-13 seems applicable.51
gene.44–48 This results in the accumulation of ultralarge von
Cirrhosis and Spontaneous
Willebrand factor multimers, which interact with platelets,
Bacterial Peritonitis
causing aggregation.45,47 Plasma exchange treats the disease
by removing the metalloprotease inhibitor and the ultra- Albumin infusion may be beneficial for patients with cirrho-
large von Willebrand factor multimers and by replacing sis and spontaneous bacterial peritonitis, as demonstrated
functional ADAMTS-13 metalloprotease through FFP or in a study in which such patients received either antibiotics
cryopoor plasma. Cryopoor plasma contains fewer ultra- or antibiotics plus albumin infusion.58 The latter group had
large von Willebrand factor multimers than FFP does and less renal impairment and a lower mortality rate. This clini-
is, therefore, preferred by some.43 cal benefit may due to the thiol-related antioxidant effect of
The use of FFP as the TPE replacement fluid may also be albumin.59
considered in patients undergoing treatment with angioten-
Nephrotic Syndrome
sin-converting enzyme (ACE) inhibitors. In such patients,
plasma exchange with albumin is associated with atypi- The nephrotic syndrome is caused by increased permeability
cal reactions such as flushing, hypotension, dyspnea, and of the glomerular capillary basement membranes, resulting
bradycardia.49 in a urine protein excretion rate of greater than 3.5 g/24 hr.60
Nephrotic syndrome is associated with hypoalbuminemia,
Paracentesis
edema, renal dysfunction, and hyperlipidemia. The standard
Ascites associated with cirrhosis follows elevated portal treatment consists of corticosteroid and cytotoxic therapies
pressure that results from increased intrahepatic vascular to treat the underlying disease and diuretic therapy with
resistance, leading to increased nitric oxide levels and a sodium-restricted diet to reduce peripheral edema and
systemic arterial vasodilatation.50 In patients with ascites, improve quality of life. A few patients may become refrac-
the vasoconstrictor systems become activated in response tory to these treatments. In such patients, investigators have
and the kidney retains sodium, causing ascites and attempted to increase diuresis by administering albumin in
edema.50 Approximately 10% of patients with ascites are combination with furosemide with some success.61 In one
BLOOD BANKING
study in which patients with hypoalbuminemia received and colleagues demonstrated that the rate of albumin loss
either furosemide alone, albumin alone, or a mixture of the from the vasculature to the tissue spaces was markedly
two, the latter regimen led to modest increases in sodium increased in critically ill patients, such as those with cachec-
and volume excretion.62 In contrast, more recent studies in tic cancer or septic shock and those who had undergone car-
nephrotic patients with hypoalbuminemia found no benefit diac surgery,75 suggesting that albumin administration may
in combining albumin with furosemide.63,64 In fact, several not be beneficial in these patient populations. Albumin infu-
detrimental effects have been linked to combined furose- sion may also have detrimental effects and can cause renal
mide-albumin treatment, including response delays, frequent dysfunction, decreased sodium clearance, and increased free
relapse to primary immunosuppressive therapy,65 hyperten- water clearance in patients with hypovolemic shock.76 It has
sion, respiratory distress, and electrolyte abnormalities.66 been suggested that albumin inhibits platelet aggregation
Given the above limitations, combining 25% albumin with and enhances antithrombin III activity.77,78
diuretic drugs is primarily indicated for patients with nephrotic In the late 1990s, a meta-analysis of 30 randomized con-
syndrome refractory to standard diuretic therapy with a trolled trials compared the use of albumin versus crystal-
sodium-restricted diet. Accordingly, the University Hospital loids or no albumin and found that when albumin was used
Consortium Guidelines for the Use of Albumin, No-Protein to treat patients with hypovolemia, burns, or hypoalbumin-
Colloid, and Crystalloid Solutions recommend the short-term emia the risk of death was 6% higher than in patients not
use of albumin with diuretics only for such refractory patients treated with albumin.33 The results prompted the FDA to
with acute, severe peripheral, or pulmonary edema.67 issue a letter to health care providers on August 19, 1998,
urging discretion in the use of albumin in the critically ill
Ovarian Hyperstimulation Syndrome population.79 Subsequently, a review of 17 studies found no
Ovarian hyperstimulation syndrome (OHSS) is a compli- difference in the incidence of pulmonary edema, length of
cation of ovulation induction that occurs after the addi- hospital stay, and mortality in adult patients receiving crys-
tional administration of human menopausal gonadotrophin talloids compared to those receiving albumin.80 A meta-
(hMG) but rarely after the use of clomiphene citrate alone.68 analysis of 55 trials conducted by Wilkes and Navickis81
The incidence of severe OHSS is estimated to occur in 0.5% also did not show a significant difference in the mortality
to 5% of in vitro fertilization cycles.69 OHSS is graded as rate when albumin was administered versus crystalloids, no
mild, moderate, or severe; in the last case, it can be fatal.70,71 albumin, or lower doses of albumin. The patient popula-
The pathophysiology of the syndrome is not yet clearly tions studied included high-risk neonates, burn patients,
defined, but it is thought that the ovaries secrete vasoactive patients with ascites, patients with hypoalbuminemia, and
substances when final follicular maturation occurs, causing trauma patients, among others. The more recent Saline ver-
increased capillary permeability.72 This results in the move- sus Albumin Fluid Evaluation (SAFE) trial82 supported the
ment of protein-rich fluid out of the intravascular space, and conclusion of the Choi review and the Wilkes and Navickis
patients can have vomiting, diarrhea, large ovarian cysts, meta-analysis.80,81 The SAFE-trial randomly assigned 6997
II thromboembolism, ascites, hydrothorax, hemoconcentra- ICU patients to receive either 4% albumin or saline for
tion, oliguria, and anasarca. fluid resuscitation and found no significant differences in
284 Although the panelists involved in making the 2000 UHC outcome in the number of days of mechanical ventilation,
guidelines concluded that there was not enough information number of days in the ICU, length of the hospital stay, or in
on the pathophysiology of OHSS to recommend the use of mortality rate at 28 days. The study included subgroups of
albumin to prevent it, a recent meta-analysis demonstrated patients with trauma, severe sepsis, and acute respiratory
that albumin administration was effective in the prevention distress syndrome (ARDS); however, the study had insuf-
of severe OHSS.73 The meta-analysis of five randomized ficient power to draw conclusions regarding albumin use
controlled trials compared the use of human albumin with in these populations.82 These studies prompted the FDA’s
placebo or no treatment on patient outcome. The albumin Blood Products Advisory Committee (BPAC) to release
dose ranged from 10 to 50 g and was given at 2 hours before, an information sheet on May 16, 2005, indicating that the
1 hour before, or just after oocyte retrieval. Albumin infu- SAFE study resolved previous safety concerns and urging
sion was estimated to prevent one case of severe OHSS for further studies on the use of albumin in burn patients and
every 18 women at risk. in patients with traumatic brain injury and septic shock.83
As an alternative treatment to albumin infusion, some An updated meta-analysis from the Cochrane Injuries
clinicians have attempted to prevent OHSS by withholding Group concludes that that there is no evidence that albu-
gonadotropins (so-called coasting). A retrospective study min reduces mortality to a greater extent than much less
comparing intravenous albumin and coasting found that expensive and equally safe options such as crystalloids in
the latter was as effective as albumin in preventing OHSS the overall critically ill patient population.84 Table 21–2 lists
in high-risk patients but that pregnancy rates were lower.74 subsets of critically ill patients (e.g., burn patients, patients
Although prophylactic albumin infusion has been shown in hemorrhagic or nonhemorrhagic shock, patients in par-
to prevent severe OHSS, it is unknown whether therapeutic ticular postoperative situations, etc.) who may benefit from
albumin infusion for women with an established diagnosis albumin administration according to UHC guidelines.
of severe OHSS is effective. In these situations, albumin is typically used when other
treatments are ineffective.
Resuscitation and Volume Expansion
in Critically Ill Patients
In the past, albumin was given for volume expansion in criti- POTENTIAL ADVERSE REACTIONS
cally ill patients because it was assumed to be more effec-
tive than crystalloids at increasing plasma volume while The incidence of adverse reactions to albumin infusions is
minimizing interstitial volume expansion. However, Fleck approximately 1 in 6600 infusions, with only 1 in 30,000
infusions being life threatening.39 Most adverse reactions
ALBUMIN
to P. cepacia, Escherichia coli, Bacillus subtilis, Candida albi-
to albumin are mild and are either allergic in nature or are cans, and Staphylococcus epidermidis also are able to grow in
related to albumin’s function as a volume expander. Some of 25% albumin. In the experiment, P. cepacia remained viable
the mild reactions that can occur include nausea, vomiting, in sealed vials of albumin kept at room temperature for
increased salivation, chills, and febrile reactions.85 Owing to 17 months after inoculation.93 Although albumin solutions
albumin’s role in increasing colloidal osmotic pressure and can support the growth of many types of bacteria, bacterial
intravascular volume, rapid infusion can result in circulatory contamination of albumin products is rarely reported in the
overload, pulmonary edema, and decreases in hematocrit literature.
and hemoglobin. Albumin, with its high negative charge,
binds calcium and can also cause complications related to
hypocalcemia.86 In addition, because albumin contains alu- CONCLUSION
minum in trace amounts, large doses can cause aluminum
to accumulate in patients with chronic renal failure and lead The development of therapeutic albumin formulations
to hypercalcemia, vitamin D–refractory osteodystrophy, was critical for soldiers in need of volume support during
anemia, and severe progressive encephalopathy.8 World War II. Over the decades following that war, albumin
PPF preparations can cause allergic reactions and reac- was used in a wide variety of settings despite the relative
tions related to intravascular volume expansion as well. PPF lack of published evidence supporting its use in those situ-
differs from other albumin-containing solutions because it ations. This led to the establishment of guidelines in 1975
includes a larger percentage of proteins other than albumin. that recommended more conservative use of albumin.4 Since
PPF has been associated with hypocoagulability, which could that time, studies have compared albumin with alternative
be due to the platelet factor-4 and β-thromboglobulin pres- fluids for volume expansion. When crystalloid administra-
ent in the preparations. Owing to a higher concentration of tion was compared with albumin use in several clinical situ-
contaminating proteins such as PKA, PPF causes more hypo- ations, no significant difference was demonstrated in terms
tensive episodes than albumin does and has been associated of patient outcome.80–82 Thus, usage guidelines for albumin
with metabolic acidosis in patients with renal dysfunction.21 have become even more conservative. Since albumin is more
expensive than crystalloids and is a plasma derivative with
related risks, crystalloids currently serve as first-line therapy
INFECTIOUS POTENTIAL for plasma expansion in most cases. The exceptions are those
clinical scenarios in which albumin has demonstrated a sig-
Because plasma derivatives are made from pooled plasma nificant clinical benefit over crystalloids, including large-scale
from thousands of donors, reduction of infectious disease therapeutic plasma exchange. At present, albumin is a valu-
transmission is an important issue in plasma processing. The able second-line treatment in many patients with conditions
process of cold ethanol fractionation significantly reduces refractory to other plasma expanders.
the concentration of viruses in plasma fractions. The pas- 21
teurization process also limits the transmission of infectious
agents by denaturing viral proteins and nucleic acids, inac- REFERENCES 285
tivating the viruses.87–90 When plasma is deliberately spiked
with viruses, the manufacturing process for the production 1. Code of Federal Regulations. Title 21 CFR 640. Washington, D.C., U.S.
Government Printing Office, 2005 (revised annually).
of 25% albumin results in a global viral reduction (log10) of 2. Kendrick DB. The Bovine and Human Albumin Programs. In Coates JB
≥17.8 for HIV; ≥16.3 for bovine viral diarrhea virus (BVDV), (ed). Blood Program in World War II. Washington, D.C., U.S. Govern-
which is a model virus for HCV; and ≥16.4 for pseudorabies ment Printing Office, 1964.
virus (PRV), which is a model virus for HBV.91 The process 3. Kendrick DB. Byproducts of Plasma Fractionation. In Coates JB (ed).
results in a global viral reduction (log10) of 14.9 for reovi- Blood Program in World War II. Washington, D.C., U.S. Government
Printing Office, 1964.
rus, which is a small, nonenveloped virus; 7.8 for hepatitis A 4. Sgouris JT, Rene A (eds). Proceedings of the Workshop on Albumin.
virus (HAV); and 6.8 for porcine parvovirus (PPV), which is Washington, D.C., U.S. Government Printing Office, 1975.
a model virus for parvovirus B19.91 There has never been a 5. Technology assessment: albumin, nonprotein colloid, and crystalloid
report of HIV or hepatitis C transmission through albumin solutions. Oak Brook, Ill., University HealthSystem Consortium, 2000.
6. Tanzi M, Gardner M, Megellas M, et al. Evaluation of the appropriate
infusion. One hepatitis B outbreak was reported in relation use of albumin in adult and pediatric patients. Am J Health Syst Pharm
to PPF infusion in 1976,92 but it probably resulted from a 2003;60:1330–1335.
lack of uniform heating of the bulk product during the pas- 7. Tarin Remohi MJ, Sanchez Arcos A, Santos Ramos B, et al. Costs
teurization process. Since 1977, the FDA has required that related to inappropriate use of albumin in Spain. Ann Pharmacother
the heating step take place after the product is placed in 2000;34:1198–1205.
8. Immuno, U.S., Inc. Package insert for Albumin (Human) 5%, 1998.
individual containers.1 Since that time, there have been no 9. Rothschild MA, Oratz M, Schreiber SS. Serum albumin. Hepatology
other reports of hepatitis B transmission through albumin 1988;8:385–401.
products. Since albumin is acellular, cell-associated viruses 10. Matejtschuk P, Dash CH, Gascoigne EW. Production of human albu-
such as CMV and Epstein-Barr virus are eliminated from min solution: a continually developing colloid. Br J Anaesth 2000;85:
887–895.
the final product. 11. Tullis JL. Albumin. 1. Background and use. JAMA 1977;237:355–360.
Cold ethanol fractionation and pasteurization reduces 12. WHO Recommendations for the Production, Control, and Regulation
but does not eliminate the risk of bacterial contamination of Human Plasma for Fractionation. Geneva, Switzerland, World Health
of albumin products. An outbreak of Pseudomonas bacte- Organization, 2005.
remia occurred in seven patients who had received albumin 13. Plasma Protein Therapeutics Association, Available at http://www.plas-
mainfo.org. Accessed Jan. 3, 2006.
from the same lot in 1973. When 190 albumin vials from 14. Testimony on Plasma Fractionator Industry FDA Regulation of Blood
the suspected lot were cultured, one vial grew Pseudomonas Safety by Thomas D. Roslewicz. Available at www.hhs.gov/asl/testify/
cepacia. A subsequent experiment showed that, in addition t970605b.html. Accessed Jan. 3, 2006.
BLOOD BANKING 15. Baxter Albumin Therapy. Available at www.albumintherapy.com. 46. Zheng X, Chung D, Takayama TK, et al. Structure of von Willebrand fac-
Accessed Jan. 3, 2006. tor-cleaving protease (ADAMTS13), a metalloprotease involved in throm-
16. Janeway CA. Human serum albumin: Historical review. In Sgouris JT, botic thrombocytopenic purpura. J Biol Chem 2001;276:41059–41063.
Rene A (eds). Proceedings of the Workshop on Albumin. Washington, 47. Tsai HM, Lian EC. Antibodies to von Willebrand factor-cleaving pro-
D.C., U.S. Government Printing Office, 1975. tease in acute thrombotic thrombocytopenic purpura. N Engl J Med
17. Cohn EJ, Strong LE, Hughes WL, et al. Preparation and properties of 1998;339:1585–1594.
serum and plasma proteins. IV. A system for the separation into frac- 48. Furlan M, Robles R, Galbusera M, et al. von Willebrand factor-cleaving
tions of the protein and lipoprotein components of biological tissues protease in thrombotic thrombocytopenic purpura and the hemolytic-
and fluids. J Am Chem Soc 1946;68:459–475. uremic syndrome. N Engl J Med 339:1578–84, 1998.
18. Finlayson JS. Physical and biochemical properties of human albumin. 49. Owen HG, Brecher ME. Atypical reactions associated with use of
In Sgouris JT, Rene A (eds). Proceedings of the Workshop on Albumin. angiotensin-converting enzyme inhibitors and apheresis. Transfusion
Washington, D.C., U.S. Government Printing Office, 1975. 1994;34:891–894.
19. Kendrick DB. The plasma program. In Coates JB (ed). Blood Program in 50. Martin PY, Gines P, Schrier RW. Nitric oxide as a mediator of hemody-
World War II. Washington, D.C., U.S. Government Printing Office, 1964. namic abnormalities and sodium and water retention in cirrhosis. N
20. Curling JM, Berglof J, Lindquist LO, Eriksson S. A chromatographic Engl J Med 1998;339:533–541.
procedure for the purification of human plasma albumin. Vox Sang 51. Runyon BA. Management of adult patients with ascites due to cirrhosis.
1977;33:97–107. Hepatology 2004;39:841–856.
21. Finlayson J. Albumin products. Semin Thromb Hemost 1980;6:85–120. 52. Gines P, Tito L, Arroyo V, et al. Randomized comparative study of thera-
22. Quirk AV, Geisow MJ, Woodrow JR, et al. Production of recombinant peutic paracentesis with and without intravenous albumin in cirrhosis.
human serum albumin from Saccharomyces cerevisiae. Biotechnol Appl Gastroenterology 1988;94:1493–1502.
Biochem 1989;11:273–287. 53. Peltekian KM, Wong F, Liu PP, et al. Cardiovascular, renal, and neu-
23. Kobayashi K, Nakamura N, Sumi A, et al. The development of recombi- rohumoral responses to single large-volume paracentesis in patients
nant human serum albumin. Ther Apher 1998;2:257–262. with cirrhosis and diuretic-resistant ascites. Am J Gastroenterol 1997;
24. Bosse D, Praus M, Kiessling P, et al. Phase I comparability of recom- 92:394–399.
binant human albumin and human serum albumin. J Clin Pharmacol 54. Gines A, Fernandez-Esparrach G, Monescillo A, et al. Randomized trial
2005;45:57–67. comparing albumin, dextran 70, and polygeline in cirrhotic patients with
25. Immuno, U.S., Inc. Package insert for Albumin (Human) 25%, 1998. ascites treated by paracentesis. Gastroenterology 1996;111:1002–1010.
26. McEvoy GK (ed). AHFS Drug Information. Bethesda, Md., American 55. Salerno F, Merli M, Riggio O, et al. Randomized controlled study of
Society of Health-System Pharmacists, Inc., 2005. TIPS versus paracentesis plus albumin in cirrhosis with severe ascites.
27. Hemolysis Associated with 25% Human Albumin Diluted with Sterile Hepatology 2004;40:629–635.
Water—United States, 1994–1998. MMWR 1999;48:157–159. 56. Gines P, Uriz J, Calahorra B, et al. Transjugular intrahepatic portosys-
28. USP DI Drug Information for the Healthcare Professional, 25th ed. temic shunting versus paracentesis plus albumin for refractory ascites
Available at http://online.statref.com/document.aspx?fxid=6&docid=2. in cirrhosis. Gastroenterology 2002;123:1839–1847.
Accessed Jan. 3, 2006. 57. Rossle M, Ochs A, Gulberg V, et al. A comparison of paracentesis
29. Talecris Biotherapeutics, Inc. Package insert for Plasbumin-5 (Low Alu- and transjugular intrahepatic portosystemic shunting in patients with
minum), Albumin (Human) 5%, USP, 2005. ascites. N Engl J Med 2000;342:1701–1707.
30. Talecris Biotherapeutics, Inc. Plasbumin-20 (Low Aluminum), Albumin 58. Sort P, Navasa M, Arroyo V, et al. Effect of intravenous albumin on renal
(Human) 20%, USP, 2005. impairment and mortality in patients with cirrhosis and spontaneous
31. Talecris Biotherapeutics, Inc. Plasbumin-25 (Low Aluminum), Albumin bacterial peritonitis. N Engl J Med 1999;341:403–409.
(Human) 25%, USP, 2005. 59. Quinlan GJ, Margarson MP, Mumby S, et al. Administration of albumin
32. Koot BG, Houwen R, Pot DJ, Nauta J. Congenital analbuminaemia: bio- to patients with sepsis syndrome: a possible beneficial role in plasma
II chemical and clinical implications—a case report and literature review. thiol repletion. Clin Sci (Lond) 1998;95:459–465.
Eur J Pediatr 2004;163:664–670. 60. Orth SR, Ritz E. The nephrotic syndrome. N Engl J Med 1998;338:
286 33. Human albumin administration in critically ill patients: systematic 1202–1211.
review of randomised controlled trials. Cochrane Injuries Group Albu- 61. Inoue M, Okajima K, Itoh K, et al. Mechanism of furosemide resistance
min Reviewers. BMJ 1998;317:235–240. in analbuminemic rats and hypoalbuminemic patients. Kidney Int
34. Strauss RG, Pennell BJ, Stump DC. A randomized, blinded trial com- 1987;32:198–203.
paring the hemostatic effects of pentastarch versus hetastarch. Transfu- 62. Fliser D, Zurbruggen I, Mutschler E, et al. Coadministration of albumin
sion 2002;42:27–36. and furosemide in patients with the nephrotic syndrome. Kidney Int
35. Owen HG, Brecher ME. Partial colloid starch replacement for therapeu- 1999;55:629–634.
tic plasma exchange. J Clin Apher 1997;12:87–92. 63. Chalasani N, Gorski JC, Horlander JC Sr, et al. Effects of albumin/furo-
36. Kimme P, Jannsen B, Ledin T, et al. High incidence of pruritus after large semide mixtures on responses to furosemide in hypoalbuminemic
doses of hydroxyethyl starch (HES) infusions. Acta Anaesthesiol Scand patients. J Am Soc Nephrol 2001;12:1010–1016.
2001;45:686–689. 64. Akcicek F, Yalniz T, Basci A, et al. Diuretic effect of frusemide in patients
37. Roberts JS, Bratton SL. Colloid volume expanders: problems, pitfalls with nephrotic syndrome: is it potentiated by intravenous albumin?
and possibilities. Drugs 1998;55:621–630. BMJ 1995;310:162–163.
38. Vercueil A, Grocott MP, Mythen MG. Physiology, pharmacology, and ratio- 65. Yoshimura A, Ideura T, Iwasaki S, et al. Aggravation of minimal change
nale for colloid administration for the maintenance of effective hemody- nephrotic syndrome by administration of human albumin. Clin Nephrol
namic stability in critically ill patients. Transfus Med Rev 2005;19:93–109. 1992;37:109–114.
39. Nearman HS, Herman ML. Toxic effects of colloids in the intensive care 66. Haws RM, Baum M. Efficacy of albumin and diuretic therapy in chil-
unit. Crit Care Clin 1991;7:713–723. dren with nephrotic syndrome. Pediatrics 1993;91:1142–1146.
40. Leblond PF, Rock G, Herbert CA. The use of plasma as a replacement 67. Vermeulen LC Jr, Ratko TA, Erstad BL, et al. A paradigm for consensus:
fluid in plasma exchange. Canadian Apheresis Group. Transfusion The University Hospital Consortium guidelines for the use of albu-
1998;38:834–838. min, nonprotein colloid, and crystalloid solutions. Arch Intern Med
41. Watson DK, Penny AF, Marshall RW, Robinson EA. Citrate induced 1995;155:373–379.
hypocalcaemia during cell separation. Br J Haematol 1980;44:503–507. 68. Ovarian hyperstimulation syndrome. Fertil Steril 2004;82:S81–S86.
42. American Association of Blood Banks, America’s Blood Centers, and 69. Klemetti R, Sevon T, Gissler M, Hemminki E. Complications of IVF and
American Red Cross Circular of Information for the Use of Human ovulation induction. Hum Reprod 2005;20:3293–3300.
Blood and Blood Components, July 2002. 70. Rabau E, David A, Serr DM, et al. Human menopausal gonadotropins
43. Obrador GT, Zeigler ZR, Shadduck RK, et al. Effectiveness of cryosuper- for anovulation and sterility: results of 7 years of treatment. Am J Obstet
natant therapy in refractory and chronic relapsing thrombotic throm- Gynecol 1967;98:92–98.
bocytopenic purpura. Am J Hematol 1993;42:217–220. 71. Schenker JG, Weinstein D. Ovarian hyperstimulation syndrome: a
44. Levy GG, Nichols WC, Lian EC, et al. Mutations in a member of the current survey. Fertil Steril 1978;30:255–268.
ADAMTS gene family cause thrombotic thrombocytopenic purpura. 72. Budev MM, Arroliga AC, Falcone T. Ovarian hyperstimulation
Nature 2001;413:488–494. syndrome. Crit Care Med 2005;33:S301–S306.
45. Tao Z, Peng Y, Nolasco L, et al. Role of the CUB-1 domain in dock- 73. Aboulghar M, Evers JH, Al-Inany H. Intravenous albumin for prevent-
ing ADAMTS-13 to unusually large Von Willebrand factor in flowing ing severe ovarian hyperstimulation syndrome. Cochrane Database Syst
blood. Blood 2005;106:4139–4145. Rev 2002:CD001302.
ALBUMIN
74. Chen CD, Chao KH, Yang JH, et al. Comparison of coasting and intrave- 84. Alderson P, Bunn F, Lefebvre C, et al. Human albumin solution for
nous albumin in the prevention of ovarian hyperstimulation syndrome. resuscitation and volume expansion in critically ill patients. Cochrane
Fertil Steril 2003;80:86–90. Database Syst Rev 2004;CD001208.
75. Fleck A, Raines G, Hawker F, et al. Increased vascular permeability: 85. Aventis Behring L.L.C. Package insert for Albuminar-25 Albumin
a major cause of hypoalbuminaemia in disease and injury. Lancet (Human) U.S.P., 25%, 2001.
1985;1:781–784. 86. Gales BJ, Erstad BL. Adverse reactions to human serum albumin. Ann
76. Moon MR, Lucas CE, Ledgerwood AM, et al. Free water clearance after Pharmacother 1993;27:87–94.
supplemental albumin resuscitation for shock. Circ Shock 1989;28:1–8. 87. Erstad BL. Viral infectivity of albumin and plasma protein fraction.
77. Joorgensen KA, Stoffersen E. Heparin like activity of albumin. Thromb Pharmacotherapy 1996;16:996–1001.
Res 1979;16:569–574. 88. McClelland DB. Safety of human albumin as a constituent of biologic
78. Jorgensen KA, Stoffersen E. On the inhibitory effect of albumin on therapeutic products. Transfusion 1998;38:690–699.
platelet aggregation. Thromb Res 1980;17:13–18. 89. Yei S, Yu MW, Tankersley DL. Partitioning of hepatitis C virus during
79. Letter to Healthcare Providers. Available at http://www.fda.gov/cber/ltr/ Cohn-Oncley fractionation of plasma. Transfusion 1992;32:824–828.
albumin.htm. Accessed Jan. 3, 2006. 90. Scheiblauer H, Nubling M, Willkommen H, Lower J. Prevalence of
80. Choi PT, Yip G, Quinonez LG, Cook DJ. Crystalloids vs. colloids in fluid hepatitis C virus in plasma pools and the effectiveness of cold ethanol
resuscitation: a systematic review. Crit Care Med 1999;27:200–210. fractionation. Clin Ther 1996;18:59–70.
81. Wilkes MM, Navickis RJ. Patient survival after human albumin admin- 91. Cai K, Gierman TM, Hotta J, et al. Ensuring the biologic safety of
istration: a meta-analysis of randomized, controlled trials. Ann Intern plasma-derived therapeutic proteins: detection, inactivation, and
Med 2001;135:149–164. removal of pathogens. BioDrugs 2005;19:79–96.
82. Finfer S, Bellomo R, Boyce N, et al. A comparison of albumin and 92. Pattison CP, Klein CA, Leger RT, et al. An outbreak of type B hepatitis
saline for fluid resuscitation in the intensive care unit. N Engl J Med associated with transfusion of plasma protein fraction. Am J Epidemiol
2004;350:2247–2256. 1976;103:399–407.
83. Safety of Albumin Administration in Critically Ill Patients. Available at 93. Steere AC. Adverse reactions to albumin caused by bacterial contami-
http://www.fda.gov/cber/infosheets/albsaf051605.htm. Accessed Jan. 3, nation. In Sgouris JT, Rene A (eds). Proceedings of the Workshop on
2006. Albumin. Washington, D.C., U.S. Government Printing Office, 1975.
21
287
Chapter 22
IVIG and Derivatives
Elizabeth E. Culler ● Lennart E. Lögdberg
INTRODUCTION At present, the FDA has approved the use of IVIG for six
indications with nine products on the market.5–13 However,
The immunoglobulins (Igs)1 constitute a major class of because of its wide-ranging immunomodulatory effects,
structurally similar globular plasma proteins. Owing to their IVIG increasingly is being used off-label to treat a diverse
antibody activity they are the principal effector molecules of assortment of diseases.14,15 At present, the off-label use of
specific humoral immunity, but they also function as broad IVIG exceeds the product’s use for FDA-approved indica-
regulators of immune system activity. The elucidation of the tions.15 Similarly, a large number of hyper-Ig products are
biochemical structure of Ig proteins and their corresponding approved for their respective niche markets.
genes unraveled the molecular basis of antibody activity and The development of therapeutic MABs represents a revo-
diversity in the 1960s and 1970s.2 lution in immunotherapy with all but 2 of the 17 marketed
The diversity of Ig genes is known to be inherited in the MAB products achieving FDA-approval in the last decade.
germline, then magnified by recombination during embryo- Indications for these products include the treatment of
nal development and expressed in each individual as a poly- organ allograft rejection, autoimmune or allergic diseases,
clonal B-lymphocyte repertoire.2 Each B-lymphocyte clone and some types of malignant tumors. With over 400 addi-
carries its own unique cell surface Ig molecule and therefore a tional MABs in clinical trials, Paul Ehrlich’s now century-old
single antibody specificity. During an immune response, the dream of creating “magic bullets” for selective treatment of
provoking agent (antigen) interacts with and activates the B- diseases is being realized.16
cell clones that have a cell surface Ig containing a matching This chapter reviews the characteristics and clinical uses
antibody combining site. Following activation, these B-cell of the various Ig preparations that are FDA-approved for
II clones differentiate into plasma cells, each of which mass- therapeutic administration in the United States.
produces and releases soluble Ig molecules with the same or
288 improved (by somatic mutation and further antigen-driven
selection) antibody specificities as the parental B-cell clone. DEVELOPMENT OF IMMUNOGLOBULINS
The progress with respect to our molecular understanding FOR THERAPEUTIC USE: A SHORT
of Ig has been paralleled, in the last few decades, by improved HISTORY
methods to select, produce, purify, and manufacture thera-
peutic Ig products. These methods include development of Igs were first detected as antibody activity more than a cen-
the hybridoma technology, allowing production of homoge- tury ago when, in 1890, Emil von Behring and Shibasabo
nous Ig molecules with predefined, single-antigen specificities Kitasato discovered that exposure of animals to diphtheria
(monoclonal antibodies [MABs]).3 At present, there is a wide toxin induced them to produce soluble factors in plasma/
array of such products on the market, including both poly- serum that could transfer protection against diphtheria to
clonal and monoclonal preparations. The polyclonal products “nonimmune” humans.17 The factor responsible for this
represent either attempts to capture representative sets of poly- passive immunity was called “antitoxin.” A decade later, Paul
clonal Ig (mainly, intravenous immunoglobulin[IVIG]), or Ehrlich recognized that humans can also actively mount
polyclonal Ig selected for particularly high titers of antibodies their own similar “antibody” responses specific to toxins
against a given antigen (hyperimmune globulin or hyper-Ig). or bacteria, providing the body with active immunity to the
Therapeutic MABs have mostly reached the market in the last relevant pathogen.16 He further suggested that such “magic
decade.4 Each MAB product has been selected to react with a bullets” might be involved in warding off tumors. Before the
specific pharmacologic target, usually a cell surface receptor or development of antibiotics and other antimicrobial drugs,
a soluble bioactive mediator. passive serotherapy for infectious diseases became the first
Today, polyclonal Ig products have a wide range of therapeutic clinical use of Ig.
uses most falling within three broad clinical categories. IVIG is the During World War II, cold ethanol fractionation was
main product for two of these: replacement therapy for patients used to mass produce the protein fraction responsible for
with congenital or acquired deficiencies of humoral immunity antibody activity (later shown to be polyclonal Ig).18 As IV
(e.g., hypo- or agammaglobulinemia) and immunomodulation. agents, however, these preparations caused chills, fever, and
Hyper-Ig preparations, by comparison, are therapies of choice shock, later shown to be due to the presence of high-molecu-
for transferring specific passive immunity, particularly used in four lar-weight aggregates in the product.19 Intramuscular (IM)
situations: (1) prophylaxis or treatment of infectious diseases; (2) administration of Ig using lower doses became the predomi-
neutralization of toxins, venoms, or drug overdoses; (3) preven- nant mode of delivery and resulted in fewer side effects. In
tion of alloimmunization; and (4) selective immunosuppression. fact, this was the route used by Bruton when he successfully
treated a patient with congenital agammaglobulinemia,20 of B lymphocytes from their progenitor cells,3 combinato-
Product Manufacturer How Supplied Method of Production Additive Osmolarity or Osmolality pH IgA Content
Carimune ZLB Behring Lyophilized, 1, 3, 6, Cold ethanol fractionation 1.67 g sucrose/g of protein, 192–1074 mOsm/kg 6.4–6.8 Trace
NF IV or 12 g and depth filtration, <20 mg sodium/g of (depending on
pH4/pepsin, nanofiltration protein concentration and diluent
used)
Flebogamma Grifols Liquid, 10 mL (0.5 g), Cold ethanol fractionation, 50 mg/mL of D-sorbitol, 240–350 mOsm/L 5.0–6.0 <0.05 mg/mL
5% 50 mL (2.5 g), 100 mL polyethylene glycol <3.2 mEq sodium/L,
(5.0 g), 200 mL precipitation, ion ≤6 mg PEG/mL
(10.0 g) exchange chromato-
graphy, pasteurization
at 60°C for 10 hours
Gammagard Baxter Lyophilized, 2.5 g, 5 g, Cold ethanol fractionation, 20 mg/mL of glucose, 5% 636 mOsm/L, 10% 6.4–7.2 ≤2.2 μg/mL
S/D Healthcare 10 g ultrafiltration, ion- 22.5 mg/mL glycine, 1250 mOsm/L
exchange chromatography, 3 mg/mL albumin,
solvent/detergent 8.5 mg sodium/mL,
treatment 2 mg PEG/mL
Gammagard Baxter Liquid, 10 mL (1.0 g), Cold ethanol fractionation, 0.25 mol/L of glycine 240–300 mOsmol/kg 4.6–5.1 37 μg/mL
liquid 10% Healthcare 25 mL (2.5 g), 50 mL chromatography, solvent/
(5.0 g), 100 mL detergent treatment,
(10.0 g), 200 mL nanofiltration, low pH
(20.0 g) incubation
Gammar-P IV ZLB Behring Lyophilized, 1 g, 2.5 g, Cold ethanol fractionation, 50 mg/mL of sucrose, 5% 309 mOsm/L, 10% 6.4–7.2 <25 μg/mL
5 g, 10 g pasteurization at 60°C for 3% albumin, 5 mg 600 mOsm/L
10 hours sodium/mL
Gamunex Talecris Liquid, 10 mL (1.0 g), Cold ethanol fractionation, 0.16–0.24 mol/L of glycine, 258 mOsm/kg 4.0–4.5 0.046 mg/mL
10% Biothera- 25 mL (2.5 g), 50 mL caprylate precipitation/ trace sodium
peutics (5.0 g), 100 mL cloth filtration, caprylate
(10.0 g), 200 mL incubation, depth
(20.0 g) filtration, column
chromatography, low pH
incubation
Iveegam EN Baxter Lyophilized, 0.5 g, 1 g, Cold ethanol fractionation, 50 mg/mL of glucose, ≥240 mOsm/L 6.4–7.2 Trace
Healthcare 2.5 g, 5 g 12% alcohol precipitation, 3 mg sodium/mL,
ion exchange chromato- <0.5 g PEG/dL
graphy, incubation with
immobilized hydrolases,
and PEG precipitation
Octagam Octapharma Liquid, 20 mL (1 g), Cold ethanol fractionation, 100 mg/mL of maltose, 310–380 mOsmol/kg 5.1–6.0 ≤1 mg/mL
USA 50 mL (2.5 g), 100 mL ultrafiltration, ≤30 mmol sodium/L
(5 g), 200 mL (10 g) chromatography, S/D
treatment, pH 4 treatment
Polygam S/D Baxter Lyophilized, 1 g, 3 g, Cold ethanol fractionation, 1.67 g sucrose/g of protein, 192–1074 mOsm/kg 6.4–6.8 Trace
Healthcare 6 g, 12 g ultrafiltration, ion- <20 mg sodium/g of (depending on
exchange chromatography, protein concentration and diluent
solvent/detergent used)
treatment
*
IVIG should never be frozen.
Data from refs. 5 through 13.
Sucrose The FDA issued a warning letter stating that the administration of IVIGs containing sucrose may increase the
risk of development of acute renal failure.89
Sorbitol Patients with hereditary fructose intolerance who receive sorbitol- or fructose-containing solutions may
develop irreversible multiple organ failure.
Glucose Use caution when using glucose-containing preparations in patients with diabetes or renal dysfunction
and in the elderly.
Glycine Glycine-containing IVIG products are associated with increased frequency of vasomotor events.
Maltose Some blood glucose monitoring systems may interpret maltose as glucose and give falsely elevated results,
which can result in an iatrogenic insulin overdose.
Sodium The concentration of sodium in the product is dependent on the concentration to which the product is
reconstituted and the diluent used for reconstitution. Caution should be used when products with high
sodium concentrations are given to patients with heart failure or renal dysfunction, neonates, young
children, the elderly, and those at risk for thromboembolism.
pH Use caution when administering low pH products to patients with compromised acid-base compensatory
methods such as neonates or patients with renal dysfunction. There are reports that low-pH preparations
may be associated with thrombophlebitis.
Osmolality and The osmolality and osmolarity of the product should be taken into account when treating patients with heart
osmolarity disease or renal dysfunction and for young children, the elderly, and those at risk for thromboembolism.
Volume Use caution when infusing IVIG to volume-sensitive patients. Some formulations can be diluted to various
concentrations, but when a smaller volume is used, the trade-off is that the osmolarity of the solution
increases. Some volume-sensitive populations include patients with renal dysfunction and heart disease,
the elderly, neonates, and small children.
advantages in dosing.34 SC administration is an off-label specific humoral immunity (e.g., common variable immu-
use but may be helpful when venous access is difficult and nodeficiency, X-linked agammaglobulinemia, and severe
offers the convenience of home self-administration by the combined immunodeficiency) can result in decreased lev-
patient. When IVIG and SC Ig therapy were compared in els of IgG through disorders in B-cell differentiation or in
a randomized trial, there were no significant differences in antibody production and can predispose patients to recur-
clinical efficacy or in the incidence of adverse reactions.35 rent sinus infections, otitis media, and bronchiectasis.
Oral Ig preparations have also been developed and have Encapsulated bacteria such as Streptococcus pneumoniae, 22
been administered experimentally with the intent of neu- Neisseria meningitidis, and Haemophilus influenzae are
tralizing pathogenic microorganisms in the gastrointes- usually the causative agents.39 Prophylactic administration 293
tinal tract; however, current evidence does not support of immunoglobulins has been the mainstay of therapy to
their routine use.36,37 reduce the number and duration of infections.39 The typi-
cal maintenance dose for adults is 400 to 600 mg/kg IVIG
THERAPEUTIC USES at monthly intervals.5–13
Because of the differences in preparation procedures, the A study comparing high-dose (600 mg/kg every 4 weeks
FDA requires that each company demonstrate the effi- for adults or 800 mg/kg every 4 weeks for children) and low-
cacy of its product in clinical trials for each indication. 15 dose (300 mg/kg every 4 weeks for adults or 400 mg/kg every
IVIG is generally used to either replace antibodies (in pri- 4 weeks for children) IVIG therapy found that high-dose
mary or secondary immunodeficiency syndromes) or as therapy reduced the number and duration of infections.40 No
an immunomodulatory agent in autoimmune disorders. target serum IgG trough level has been established; however,
Table 22–3 lists the FDA-approved indications for each a trough level of greater than 500 mg/dL was more protective
product along with the manufacturer’s recommended against infection than a trough level of less than 500 mg/dL in
dose. The high-dose regimens are not recommended for one study.41 Patients with active infections may require more
volume-sensitive patients. Although each IVIG prepara- frequent IVIG administration until the infection is cleared,
tion has not been approved by the FDA for all of the six followed by maintenance replacement therapy.39,40,42 The use
uses, many clinicians use the preparations interchange- of IVIG and antibiotics in patients with primary immunodefi-
ably. Besides the FDA-approved uses, IVIG has a wide ciency syndromes has significantly reduced the incidence and
variety of off-label uses. It is estimated that more than duration of infections in these patients, markedly improving
half of the IVIG prescribed annually is for a non-FDA- their quality of life.
approved indication.38 Secondary Immunodeficiencies. Secondary immunode-
ficiency syndromes are acquired disorders of the immune
FDA-Approved Indications system, which may be caused by a hematologic malignancy,
Primary Immunodeficiency Syndromes. Primary by an infection, or by immunosuppressive therapy. IVIG is
immunodeficiency syndromes are genetically deter- approved by the FDA in the treatment of secondary immuno-
mined disorders that may affect humoral immunity, cel- deficiencies caused by chronic lymphocytic leukemia (CLL),
lular immunity, phagocytosis, complement function, or a HIV infection, and immunosuppressive therapy related to
combination of immune activities.39 Disorders affecting bone marrow transplantation.
II
BLOOD BANKING
294
Table 22–3 Recommended Doses for FDA-approved Indications for IVIG5–13
Secondary Prevention of
Immunodeficiency Pediatric HIV GVHD and Infection Kawasaki
Product Primary Immunodeficiency ITP Due to CLL Infection in Adult BMT Syndrome
Carimune 200 mg/kg once a month; 400 mg/kg for 2–5 consecutive days. × × × ×
dose may be increased If platelet count does not respond,
to 300 mg/kg, or a single infusion of 400 mg/kg may
frequency of dosing may be given. If response is still
be increased inadequate, a single infusion of
800–1000 mg/kg may be given.
Flebogamma 300–600 mg/kg body × × × × ×
weight every 3–4 weeks
Gammagard 300–600 mg/kg body A single dose of 1000 mg/kg. The need 400 mg/kg every × × Either a single
S/D weight every 3-4 for additional doses can be 3–4 weeks 1000 mg/kg dose or
weeks determined by clinical response a dose of 400 mg/kg
and platelet count. Up to three for 4 consecutive
separate doses may be given on days beginning
alternate days if required. within 7 days of the
onset of fever. Also
administer aspirin
therapy (80–100 mg/
kg/day in four
divided doses)
Gammagard 300–600 mg/kg every × × × × ×
liquid 10% 3–4 weeks
Gammar-P IV Children and adolescents: × × × × ×
starting doses of
200 mg/kg every 3 to
4 weeks Adults:
200 mg/kg to 400 mg/kg
given every 3–4 weeks
Gamunex 300 and 600 mg/kg every Either two doses of 1000 mg/kg given × × × ×
3–4 weeks on consecutive days, or five doses of
400 mg/kg given daily for 5 days
Iveegam EN 200 mg/kg per month; the × × × × Either 400 mg/kg
dose may be increased daily for
up to fourfold, or intervals 4 consecutive days
between infusions can or a single dose of
be shortened 2000 mg/kg given
over a 10-hour
period within
10 days of disease
onset. Also, a dose
of 100 mg/kg of
aspirin should be
administered daily
through the 14th
day of illness, then
3–5 mg/kg each day
for 5 weeks
Octagam 300–600 mg/kg every × × × × ×
3–4 weeks
302
Table 22–4 Monoclonal Antibodies Approved by FDA for Therapeutic Uses in the United States
Category Subcategory Composition Target-Specificity Generic Name/Trade Name Main Clinical Application Area FDA Approval
Figure 23–1 Methods for preparation of platelet-rich plasma (PRP)– and buffy coat (BC)–derived platelet concentrates.
PLATELETS AND RELATED PRODUCTS
Similar studies have compared AD-Plts with BC WB-Plts like glucose and fatty acids, acts as a substrate for platelet
in terms of their in vitro hemostatic characteristics. Data metabolism and provides the highly desirable side effect of
from these studies support the contention that AD-Plts bicarbonate production during its consumption, which helps
possess better in vitro hemostatic parameters including pres- buffer pH. Glucose-free crystalloid solutions do not appear
ervation of ADP and collagen-induced platelet aggregation to maintain concentrate pH and are associated with reduced
and shorter closure times of the PFA 100 test system.41,42 in vivo recovery after 5 days of storage.50 Other investigators
In vivo studies to ascertain therapeutic differences in these have suggested placing various additives in the platelet stor-
preparations remain to be performed. age bag that inhibit platelet and coagulation factor activation
such as prostaglandin E1, theophylline, aprotinin, and hiru-
din.51–53 Although platelet shelf life can possibly be extended
Anticoagulant/Preservative Solutions with these additives, the potential for harmful side effects,
especially in infants and pregnant women, has slowed their
Citrate-based Storage Solutions commercial development. Moreover, these solutions could
Platelets are prepared from whole blood drawn into one not be deployed to extend the shelf life of stored platelets
of several anticoagulant-preservative solutions, the most without the availability of suitably robust methods to detect
common being citrate-phosphate-dextrose solutions (CPD, and limit bacterial contamination.
CP2D) and CPD with adenine (CPDA-1). Citrate anticoagu-
lates blood according to its ability to chelate calcium and thus
inhibit the coagulation cascade. Phosphate serves as a buffer Storage Temperature
and dextrose as a source of energy. Adenine, which improves
Current Storage Standards
red cell survival by increasing cellular ATP levels, serves no
purpose in platelet storage. EDTA cannot be used because Liquid platelet concentrates were originally stored at 4° C until
it causes platelet structural changes associated with reduced the late 1960s when it was discovered that products stored at
in vivo viability and is toxic to humans. Heparin is not an room temperature had longer in vivo survival and greater
acceptable alternative because it activates platelets, causing hemostatic efficacy than those stored at the colder temper-
platelet clumping. It also causes systemic anticoagulation ature.54–56 In fact, the shelf life of liquid platelets stored at
through its effect on antithrombin. The amount of citrate in 22° C was extended from 5 to 7 days in 1984 by the U.S. Food
the CPD or CPDA-1 preservative does not generally produce and Drug Administration (FDA) owing to a combination of
systemic anticoagulation or hypocalcemia in blood recipi- improved plastic storage containers and increased storage
ents because it is quickly metabolized to bicarbonate in the temperature. Within 2 years, however, a marked increased in
liver. AD-Plts are prepared in ACD-A (citric acid, trisodium the incidence of platelet transfusion–related bacterial sepsis
citrate, dextrose) preservative using automated cell collectors. prompted the FDA to reduce the maximum storage period
Citric acid is used to provide a lower pH. In some apheresis back to its current 5-day limit.
systems, platelets are collected as platelet-rich plasma and do The optimal liquid platelet storage temperature appears 23
not require resuspension, whereas other systems yield a con- to be 20° C to 24° C with continuous gentle agitation. While
centrated platelet pellet that must be resuspended. AD-Plts colder storage temperatures demonstrably slow bacterial 311
are stored similarly to WB-Plts. growth, platelets stored in the cold become activated and
lose their normal discoid shape.57 Development of spheri-
Alternative Storage Solutions cal morphology on exposure to cold has long been recog-
Although platelets stored in citrated plasma are satisfactory nized as evidence of irreversible physical damage.58 Changes
for clinical use, several investigators have evaluated other in platelet morphology have been observed even following
solutions that could extend the shelf life of platelet concen- short-term (24 hours) exposure to temperatures below 20° C,
trates or improve their capacity to circulate in vivo after infu- as may occur during platelet transport. The mechanism for
sion.43 The majority of these studies have evaluated storage these changes may be related to the release of calcium ions
media alternatives in PRP WB-Plts although synthetic storage from platelet dense bodies or to influx of calcium across the
medium has also been examined for stored BC WB-Plts.38,44 platelet membrane, resulting in actin filament assembly and,
The scientific basis for this came out of work performed in subsequently, platelet activation.59,60 Chilled platelets have
the 1990s demonstrating that the rate of in vitro platelet recently been shown to undergo rapid clearing from the cir-
aging at 22° C was slowed to less than half of that associated culation by hepatic macrophages. This clearance is mediated
with in vivo aging at 37° C, as determined by isotope label- by the αMβ2 integrin (a complement receptor type 3/Mac-
ing.45 The reduced rate of aging closely paralleled the con- 1) that binds to clustered GP1b receptors on the surface of
comitant reduction in platelet metabolic rate. Thus, in theory, chilled platelets, resulting in rapid platelet phagocytosis.61
the normal 9- to 10-day platelet life span could be extended Binding occurs through a lectin-mediated interaction with
to 18 to 20 days or more under optimal storage conditions. exposed β-N-acetylglucosamine (β-GlcNAc) residues on
Proposed interventional targets to achieve “optimal storage the GP1b complex. Enzymatic galactosylation of exposed
conditions” have included methods to (1) reduce platelet β-GlcNAc residues, however, has been shown to inhibit
activation during collection, component preparation, and αMβ2 binding, resulting in normalization of chilled platelet
storage; (2) reduce platelet metabolism,46 thereby decreasing intravascular circulation time.62
the rates of in-storage glucose consumption and lactate pro-
Investigative Strategies for Extended
duction; and (3) ensure that glucose is not exhausted in the
Cold Storage
storage medium prior to component outdate.47
Most alternatives studied or currently under develop- If changes in the platelet cytoskeleton can be prevented,
ment are composed of buffered salt solutions containing long-term storage of liquid platelets could become fea-
various additives such as acetate and gluconate.47–49 Acetate, sible. Platelets treated with cytochalasin B, an inhibitor of
BLOOD BANKING
new actin filament assembly, do not develop pseudopods adequately maintained for 5 days by storing 50 to 65 mL of
or undergo spreading when cooled to 4° C.63 Human plate- PCS in more gas-permeable blood bags made of either PVC
lets stored for 21 days at 4° C with EGTA-AM, a cytoplasmic with a trimellitate, non-DEHP plasticizer such as TOTM
Ca2+ chelator, and cytochalasin B remain responsive to ADP (Fenwal PL1240 or Cutter CLX),69 or blow-molded polyole-
and thrombin in the presence of exogenous calcium. Other fin (Fenwal PL732).70 Other second-generation platelet stor-
strategies have been developed to allow for preservation of age bags were composed of thin-film PVC with a 2-DEHP
platelet morphology and function during cold storage. These plasticizer (XT-612; Terumo), and PVC with a citrate-based
include trehalose, a disaccharide used during freeze-drying, non-DEHP plasticizer (butyryl-tri-hexyl citrate [BTHC];
and the antifreeze glycoproteins (AFGPs). The latter proteins Fenwal PL2209). The latter compound, PL-2209, allows stor-
are isolated from fish that have adapted to survive in the age of 5–7 × 1011 platelets/bag in 400 mL of plasma for 5 days
cold temperatures found in the polar regions. AFGPs reduce at 20° C to 24° C with acceptable in vitro and in vivo storage
platelet activation that occurs when human platelets are characteristics.71
stored in the cold.57 Despite these observations, the future
Plasticizers and Potential Toxicity
role of cold-stored platelets in clinical transfusion practice
remains unclear. Despite their use since the 1950s, there remain concerns
regarding the potential toxicity of storage bags composed of
Cryopreserved Platelets PVC plastic with a DEHP plasticizer, which readily leaches
Similarly, the use of cryopreserved platelets remains limited into the blood product.72–74 These concerns, which include
in clinical practice. Platelets frozen in a cryoprotectant solu- the possible carcinogenicity of DEHP and its monoethylhexyl
tion of 5% to 6% dimethyl sulfoxide (DMSO) can be stored phthalate (MEHP) metabolite, have led to the development
for a period of years at −80° C. Post-thaw ex vivo recovery of various alternative plasticizers used in blood storage con-
has been reported to be 75%, with an in vivo 1-hour recov- tainers.75,76 Studies have suggested that transfusing plate-
ery in normal volunteers of 33% and 8-day survival.64 The lets stored in BTHC containing bags, for example, is safe.77
clinical utility of frozen platelets, however, is diminished by BTHC leaches into plasma at a level 60% to 70% lower than
the lack of trained personnel and facilities at most hospitals does DEHP. Moreover, BTHC differs from the phthalate
for their preparation, storage, and reconstitution.65 Thus, plasticizer in that BTHC is metabolized to physiologic com-
widespread adoption of the use of cryopreserved platelets pounds like citric acid, butyric acid, and hexanol. Extensive
currently appears unlikely. Some investigators, however, toxicology testing has shown that BTHC has a very low level
support their use in specialized circumstances, such as for of toxicity and that the plasticizer and its degradation prod-
autologous or allogeneic donation for subsequent transfu- ucts are rapidly eliminated from the body by pulmonary,
sions to manage treatment-related thrombocytopenia in fecal, and urinary routes. Although gas permeability of the
patients with malignancies.66 material is slightly less than that of PL-732 plastic, there is
sufficient transmission of oxygen into, and carbon dioxide
II out of, the container to ensure aerobic oxidative metabolism
Platelet Storage Containers and maintenance of an acceptable pH.
312 Development of Gas-Permeable Plastic
Storage Bags Platelet Storage Agitation
The critical importance of oxygen-permeable containers was In the late 1970s, it was found that platelets stored with gentle
recognized shortly after the development of plastic bags for the agitation maintained better morphology and in vitro func-
storage of platelet concentrates. It was appreciated that if oxygen tionality than platelets stored undisturbed.78 Thus, platelets
supply to the platelet was inadequate, then metabolism would are now stored with continuous gentle agitation. Rotators are
shift from primarily aerobic to the anaerobic glycolytic path- available in a face-over-face (circular) angle of rotation or in a
way. This pathway produces high concentrations of lactic acid, flatbed configuration.37 However, not all agitators are appro-
which reacts with the bicarbonate buffer in the plasma. When priate to store platelets collected in certain plastic containers.
bicarbonate is exhausted, at levels of 20 to 25 mmol/L of lactic For example, platelets stored in PL732 blow-molded, polyolefin
acid, there is a rapid reduction in pH with a resultant loss of bags using 6-rpm elliptical rotators showed decreased post-
platelet viability. Adequate entrance of oxygen into the storage transfusion recovery and survival.70 These findings were
bag, conversely, allows platelets to maintain energy metabolism thought to be related to platelet–plastic storage bag interac-
through mitochondrial oxidative phosphorylation. tions at various shear stresses created with agitation. Platelet
Plastic bags originally developed to store platelet concen- storage using any available storage bag–rotator combination
trates were composed of polyvinyl chloride (PVC) contain- is associated with increasing levels of CD62P over time and
ing a di-2-ethylhexyl phthalate (DEHP) plasticizer and did progressive release of β-thromboglobulin from α-granules.
not permit storage of platelets beyond 3 days. The walls of Agitation is also associated with discharge of cytosolic lac-
such plastic bags failed to allow sufficient CO2 and oxygen tate dehydrogenase (LDH), suggesting that some degree of
exchange for platelets to sustain aerobic metabolism. After platelet lysis occurs during agitation.79
3 days of storage, anaerobic metabolism would produce
enough lactic acid that the pH of concentrates routinely fell
below 6.0. These changes markedly reduced in vivo platelet Apheresis-Derived Platelets
recovery and survival.33,67,68
Development of Apheresis Collection
Storing platelets in gas-permeable storage bags that
Technologies
permitted the influx of oxygen and the efflux of carbon
dioxide was found to ameliorate these untoward effects. The development of automated instruments for platelet col-
It was later discovered that platelet concentrate pH can be lection was, in part, motivated by the need to collect large
numbers of platelets from a single donor.80 It was clear during
*Data from American Association of Blood Banks. Technical manual, 14th ed. Bethesda, Md.: American Association of Blood Banks, 2002.
†
Data from Guide to the Preparation, Use and Quality Assurance of Blood Components, 10th ed. Strasbourg, Council of Europe Publishing,
2004.
‡
In 95% of units tested.
§
Standard met if 90% of units tested fall within indicated values.
ıı
Volume must maintain product within specified pH for duration of storage.
#
Platelet number within limits that comply with validated preparation and preservation conditions.
gram to document adequate LR.116 Only a small number of
GRANULOCYTES
traditionally been used for this purpose, and in modern cell
If the patient is CMV seronegative, the donor should also
separating machines this allows for 30% to 50% efficiency
be CMV seronegative, since most granulocyte recipients
in granulocyte collection. Two preparations of hydroxy-
are in a patient population that requires CMV-safe compo-
ethyl starch are available for this purpose. Hetastarch, a
nents. The incidence of CMV transmission appears to be
high-molecular-weight compound, was the substance used
higher with granulocyte concentrates than with other cel-
originally, but it was shown to persist in the circulation
lular blood products because the latent virus resides in the
for months. This persistence led to a concern for its safety,
leukocytes,15,16 and the risk of transmission can be nearly
although long-term effects were not actually observed. The
eliminated by selecting seronegative donors. Needless to say,
lower molecular weight version, pentastarch, was shown to
leukocyte reduction (leukoreduction), a technique available
have a much more rapid elimination time and appeared to
to render other blood components relatively CMV-safe, is
function equivalently for the purposes of granulocyte col-
not appropriate for granulocyte concentrates.
lection.32 Many collection centers switched to pentastarch
LEUKOCYTE COMPATIBILITY as a result of these findings. More recent controlled studies
showed, however, that collection efficiency is much greater
If the patient is not alloimmunized, it is probably not nec- when the high-molecular-weight preparation of hydroxy-
essary to select granulocyte donors on the basis of HLA or ethyl starch is used.33 Whichever agent is used, it must be
granulocyte typing or to perform leukocyte compatibility infused continuously for as long as the collection proceeds in
testing.17 However, convincing evidence indicates that allo- order to maintain the effect.
immunized recipients who are transfused with incompatible
leukocytes are more likely to experience adverse pulmonary DONOR STIMULATION
reactions or febrile transfusion reactions.18–22 In addition, In an effort to increase the number of granulocytes collected,
the infused granulocytes will be rapidly cleared from the cir- donors are routinely stimulated with corticosteroids to mobi-
culation and will be ineffective.17,18,21,22–26 The difficulty is in lize cells from the marrow storage pool and to increase the
knowing which patients are alloimmunized. Leukocyte anti- circulating granulocyte count, thereby increasing the number
bodies detected in the laboratory do not necessarily correlate of granulocytes that can be collected. Numerous stimulation
well with clinical evidence of alloimmunization, such as the regimens have been proposed, the most successful using up
ability of the transfused cells to circulate or to accumulate at to 60 mg prednisone or 8 mg dexamethasone, which raises
sites of inflammation. In addition, reliable detection of clini- the donor’s neutrophil count two- to threefold from baseline
cally significant antibodies requires that a panel of sophis- values.34,35 Short-term administration of such doses of cor-
ticated tests be performed,19,24 tests that are not available in ticosteroids are well tolerated by most donors; persons with
most institutions. In the absence of such results, a common medical contraindications to such medications, such as active
approach is to attempt to gauge the likelihood of alloim- peptic ulcer disease or diabetes, should not be donors. Even
munization on the basis of information such as the patient’s higher donor granulocyte counts can be obtained by stimula- 24
history of febrile transfusion reactions, response to random tion with G-CSF, a strategy discussed later in this chapter.
donor platelet transfusions, and results of a lymphocytotoxic
Cell Concentrate 343
antibody screen. If these results are normal, it is not likely
that the patient is alloimmunized. Significant abnormalities Granulocyte concentrates are not licensed by the FDA and
in these parameters do not necessarily mean that the patient therefore have no defined regulatory specifications. AABB
will have difficulty with granulocyte transfusions, but they do standards require that at least 75% of concentrates contain
suggest that the physician proceed gingerly and also consider at least 10 × 109 granulocytes,36 a value intended to serve as
further clinical or laboratory evaluation. a benchmark for adequate collection technique but not to
Collection Procedure imply that 10 × 109 granulocytes is an adequate clinical dose.
With the techniques discussed earlier, including adequate
Granulocytes are collected from donors by a leukapheresis corticosteroid stimulation of the donor, mean yields of 20–
procedure. Historically, two methods have been used: filtra- 30 × 109 granulocytes are typically achieved. Depending on
tion and centrifugation. In the former, the donor’s blood the particular cell separator used, the granulocytes are sus-
was passed over nylon wool filters to which the neutrophils pended in 200 to 400 mL plasma and contain 10 to 30 mL red
adhered; the cells were subsequently eluted from the col- blood cells and 1–6 × 1011 platelets.
umns. Although large numbers of granulocytes could be The functional capabilities of granulocytes obtained by
collected by this technique, subsequent studies showed that these techniques have been the subject of many reports. In
the cells were functionally impaired.27–29 In addition, the vitro and in vivo studies have shown repeatedly that cells col-
process itself activated complement and was associated with lected by centrifugal apheresis are normal or nearly normal
transfusion reactions in recipients and with occasional and functionally, and the characteristics are not compromised by
sometimes serious adverse effects in donors.30,31 This tech- the use of hydroxyethyl starch or by stimulation of normal
nique is no longer in use. For centrifugal leukapheresis, the donors with corticosteroids.27,37,38
cells are separated from other components by centrifuga-
tion. Numerous acceptable cell separators are on the market
for this purpose. Usually, 7 to 10 L of the donor’s blood are Storage of Neutrophils
processed in a procedure that takes approximately 3 hours. After collection, granulocytes rapidly undergo apoptosis,3,39,40
and this process greatly limits the ability to store the cells
RED CELL SEDIMENTATION
before transfusion. Although the hardier cell functions per-
It is necessary to add a red blood cell (RBC) sedimenting sist after a few days of liquid storage, the more sensitive ones,
agent to the donor’s blood to effect adequate separation of such as the ability to migrate, deteriorate more quickly.41,42 In
BLOOD BANKING
vivo studies have shown that blood recovery and survival are with chronic myelocytic leukemia by a manual leukapher-
adversely affected by as little as 24 hours of storage, and the esis technique whereby leukocyte-rich plasma was prepared
ability of the transfused cells to localize to areas of inflamma- from individual units of donor blood by sedimenting the
tion is decreased as much as 75% after 8 to 24 hours of stor- red blood cells. Schwarzenberg and associates52 treated 33
age.27,43 As a result of these observations, granulocytes should patients with various malignant diseases and reported that
be administered as soon as possible after collection. In the event approximately half showed a favorable response. In several
that this is not possible, the cells should be stored without agi- of the patients, the postinfusion rise in the patient’s neu-
tation for no more than 24 hours.36,44 The preponderance of trophil count persisted, and the Philadelphia chromosome
evidence to date indicates that the cells should be maintained could be demonstrated in the patient’s marrow cells, find-
at room temperature for any period of storage.41,45 ings suggesting that a temporary graft had occurred. Morse
and associates53 used a similar technique and transfused 40
patients with leukopenia, most of whom had acute leuke-
Transfusion of Granulocytes
mia. Following 81 transfusions given to patients who were
Once the decision is made to initiate granulocyte support, the febrile before the transfusion, more than half of the patients
physician should strive to provide daily granulocyte transfu- were afebrile by 36 hours after the transfusion. Lowenthal
sions until the patient’s infection clears or until the patient’s and colleagues54 collected granulocytes from normal donors
neutrophil count has returned to at least 500/μL. Data are and those with chronic myelocytic leukemia by continu-
insufficient to determine whether higher neutrophil counts ous-flow centrifugation and treated 41 febrile patients with
would be more appropriate as stopping points for certain acute leukemia or aplastic anemia. Two thirds of the patients
clinical conditions. Outside these parameters, it is generally responded by defervescence. Response was more likely in
not appropriate to stop and start transfusion support on the those patients with proven or probable bacterial or fungal
basis of the patient’s daily clinical status. infection than in those with fever of unknown origin.
Granulocyte concentrates usually contain 10 to 30 mL red With the development of the automated cell separator,
blood cells, enough to cause a hemolytic transfusion reaction it became possible to collect large numbers of granulocytes
if the donor and patient are incompatible. Therefore, AABB from normal donors, and the use of donors with chronic
standards require that an RBC crossmatch be performed myelocytic leukemia fell into disuse. A flurry of reports of
before transfusion.36 granulocyte transfusion therapy in patients with neutrope-
Granulocyte preparations contain viable lymphocytes, nia then followed. The aggregate experience was reviewed by
and graft-versus-host disease has been reported after granu- Strauss55 and is summarized in Table 24–1. In this review,
locyte transfusion.46 Although this complication can easily patients were categorized by the infection for which the
be prevented by gamma-irradiation of the component, the transfusions were begun and were counted only once. All
decision to irradiate granulocyte concentrates routinely has patients with documented fungal infection were categorized
been debated in the past. Those who opposed routine irradia- as such, whether or not they fit into other categories. After
II tion noted that graft-versus-host disease is an extremely rare these patients were excluded, all patients with sepsis were
complication, expressed concern that irradiation may com- listed only in the sepsis section. The number of patients indi-
344 promise the integrity of the cells, and argued that, as with any cated in Table 24–1 represents the number of patients who
other blood product, the decision to irradiate the cells should could actually be evaluated for treatment efficacy, sometimes
be based on clinical evaluation of the patient.47 However, in a much smaller number than actually treated. Therapy was
regard to the last concern, most studies suggest that stan- considered successful if so indicated by the authors of the
dard irradiation does not impair neutrophil function.48–51 study. Although these results indicate the general experience
Since most patients receiving granulocyte transfusions are with granulocyte transfusion therapy, care must be taken to
to some degree immunosuppressed and since transfusion- not overinterpret the data. The number of patients often was
related graft-versus-host disease is almost uniformly fatal, small, and the studies represented were heterogeneous, with
it is recommended that granulocyte concentrates always be different inclusion criteria, granulocyte preparations, and
irradiated before transfusion. criteria for success.
Granulocytes should be administered through a standard
Controlled Trials
blood administration set filter (170 μm) and infused over 1
to 2 hours. As mentioned above, the use of a leukoreduction Included in the foregoing aggregate experience were seven
filter is contraindicated. Administration of antipyretics or controlled trials, reported between 1972 and 1982, designed
corticocosteroids is appropriate for patients who experience
symptoms such as chills and fever; routine prophylaxis with
these agents is not recommended except for patients who Table 24–1 Treatment of Neutropenic Patients
have previously experienced such symptoms. with Granulocyte Transfusions
GRANULOCYTES
uncontrolled trials of granulocyte transfusion therapy, in
which large doses of cells could be obtained from donors such as CD64, CD35, CD14, and CD11/18,38,98,99 proteins that
with marked leukocytosis, clinical responses were reported are important in cell adhesion processes. G-CSF also inhibits
to be associated with the dose delivered. Morse and associ- neutrophil apoptosis,39,101 a finding that may explain, in part,
ates53 observed that the increase in the patient’s neutrophil the prolonged blood survival of neutrophils from subjects
count was directly related to the dose of cells provided and given the drug.96
was detectable only if the dose exceeded 1010/m2. The clinical
response, as defined by defervescence, was also proportional Stimulation of Donors with G-CSF
to the dose, the fraction of patients responding ranging from
30% to 100% with mean doses of 2.6 × 1010/m2 and 15.6 The use of G-CSF to stimulate normal granulocyte donors
× 1010/m2, respectively. Lowenthal and colleagues54 reported has been reported by numerous investigators (Table 24–2).
that patients with clinical responses received, on average, four In most of these studies, G-CSF was administered repeat-
times as many cells as patients without responses. Second, edly to donors who were family members or friends of
retrospective analysis of the controlled trials of therapeutic the patient who received the transfusions, although in the
granulocyte transfusion therapy suggested that higher doses study of Price and associates,91 the donors were community
of cells were provided in the studies that showed efficacy.63,64 apheresis donors who were recruited to donate for patients
Third, the experience with the provision of granulocytes to whom they did not know. The dose of G-CSF in these stud-
neonates, in whom the relative dose is much higher because of ies ranged from 5 to 10 μg/kg, and it resulted in granulo-
the size of these patients, suggests that efficacy is determined cyte concentrates containing an average of 40 to 60 × 109
by dose. Finally, studies in animals indicated the importance cells. Leitman and associates102 and Price and colleagues91
of dose. Appelbaum and associates93 examined the clinical achieved substantially higher average yields (80 × 109 cells)
effect of granulocyte support in dogs with Pseudomonas sep- by administering both G-CSF and dexamethasone (8 mg) to
sis and showed that dogs receiving 108 cells/kg did not survive normal subjects. Liles and coworkers103 determined that the
the infection, whereas 100% (5 of 5) survived when they were addition of corticosteroids resulted in higher donor blood
given 2 × 108 cells/kg. Epstein and Chow94 provided granulo- neutrophil counts, irrespective of the dose of G-CSF, and
cytes to dogs with Candida albicans meningitis and showed the maximum neutrophil counts occurred approximately
a direct relationship between the dose of cells administered, 12 hours after stimulation. Administration of G-CSF, with
the blood granulocyte increments, and the number of granu- or without corticosteroids, is well tolerated by donors.104,105
locytes migrating to the cerebrospinal fluid. Most donors experience mild to moderate bone aching,
headache, or insomnia. In one study, 98% of donors were
willing to undergo future G-CSF stimulation.91
Actions of G-CSF
Because G-CSF has been shown to inhibit neutrophil
With the availability of recombinant G-CSF, the possibility apoptosis,39,101 this cytokine may be useful in lengthening the
of greatly increasing the number of granulocytes for collec- acceptable storage time for neutrophils and thereby improv- 24
tion and thus for transfusion raised the hope that the efficacy ing the logistics of granulocyte therapy programs. Whether
of granulocyte transfusion therapy could be improved.3,13 this will turn out to be true awaits further studies. 347
When administered to normal subjects, G-CSF causes a rapid
dose-dependent increase in the neutrophil count, beginning Hematologic Effect in Recipients
within 2 hours and peaking at approximately 12 hours.95 This
phenomenon is the result of the rapid release of neutrophils In marked contrast to the situation in traditional granulocyte
from the marrow storage pool into the blood. When given transfusion therapy, the postinfusion neutrophil increments
daily, G-CSF also stimulates the proliferation of granulocyte seen in patients receiving these large doses of cells are quite
precursors and accelerates the transit time of the developing substantial. Hester and associates106 reported a mean post-
cells through the maturation pool into the blood.96 G-CSF transfusion neutrophil increment of 0.6 × 103/μL after infu-
also affects neutrophil function. It increases phagocytosis as sion of 40 × 109 granulocytes, and the value remained higher
well as bactericidal and fungicidal activity.39,97,98 It primes than the baseline value for 24 hours. In the study of Adkins
neutrophils and enhances their metabolic responses to sec- and colleagues,107 patients received a mean granulocyte dose
This chapter introduces the different types of coagulation tic factor. However, there are a number of distinct disad-
factor concentrates that are available, the advantages and vantages to purifying coagulation factor concentrates from
disadvantages of utilizing different products, and the indi- human donors. There is always the concern of transmitting
cations for their use. Detailed information on dosing and infectious agents from the donor into the recipient. Indeed,
administration of a given product should be obtained from before HIV was identified as the causative agent of AIDS,
the published guidelines for that product. a large number of hemophilia patients were infected from
HIV-tainted purified factor VIII. Current purified human
factor VIII is obtained only from donors who screen negative
REPLACEMENT THERAPY FOR for the known transmissible pathogens. However, this will
COAGULATION FACTOR DEFICIENCIES not protect factor VIII recipients from unidentified emerg-
ing pathogens that may enter the blood supply in the future.
Since the initial descriptions of hemophilia and von For this reason, there are ongoing efforts to perfect steriliza-
Willebrand disease, it has been appreciated that the coagu- tion techniques that would potentially inactivate all patho-
lation system is made up of discrete factors, the absence of gens. Since coagulation factors are proteins, these techniques
which can result in life-threatening bleeding. Early attempts focus on methodologies that do not damage proteins, such as
to replace such factors consisted of transfusing plasma. cross-linking of nucleic acids (DNA or RNA) or disrupting
However, the low concentrations of coagulation factors in membranes of lipid-enveloped pathogens. However, such
normal plasma are a limiting factor to this approach, since techniques will not inactivate prions, since, like coagulation
there is a maximal rate of plasma infusion and total volume factors, prions are also proteins. Thus, transmission of spon-
II that a patient can tolerate. Treatment of specific factor defi- giform encephalopathies remains a concern.7,8
ciencies was improved with the introduction of relatively The advent of recombinant DNA technology has allowed
352 crude fractions of plasma that effectively concentrated the the cloning of cDNAs and the ectopic expression of human
desired factors (i.e., cryoprecipitate). However, such prod- proteins in cell culture. This process allows for the expression
ucts contain only certain factors (predominantly vWF, VIII, and purification of large quantities of the protein products of
XIII, and fibrinogen). In addition, multiple pooled units human genetic sequences and essentially eliminates the risk
of cryoprecipitate carry the risk of transmitting infectious of pathogen transmission from human donors. A residual
disease from donors.1 Further advances in factor purifica- risk of infection from pathogens of animal origin remains,
tion provided concentrates of factors VIII and IX for the since animal serum typically is required to culture the cell
treatment of hemophilia A and hemophilia B, respectively. lines producing the factors. In addition, albumin of human
Modern concentrates are now subjected to pathogen inacti- or animal origin has been required to stabilize recombi-
vation techniques, which minimizes but does not eliminate nant factors. These additives provide a potential residual
the risk of infection.2,3 More recently, recombinant DNA source for introducing infectious material. However, recent
technology has allowed the expression of human coagulation advances have circumvented these technical limitations. For
factors in vitro, which essentially eliminates the risk of trans- example, in 2003 the U.S. Food and Drug Administration
mitting human pathogens that could co-purify from donor (FDA) approved a recombinant factor VIII product expressed
plasma. However, since some products are stabilized with in Chinese hamster ovary cells, which has never come into
human albumin, it has been suggested that a risk of trans- contact with products from human or animals. Although it
mission of selected pathogens, such as parvovirus B19 and is still theoretically possible that undetected pathogens har-
prions remains a possibility.2 The current move to albumin bored in the cell lines themselves may infect human recipi-
free stabilizers may address this issue.4–6 ents, recombinant factors expressed in serum-free systems
without albumin additives are as close to being pathogen free
as is possible in culture-based systems.
PURIFIED VERSUS RECOMBINANT Despite the above attributes, there are several disad-
COAGULATION FACTORS vantages to utilizing recombinant factors. These products
are typically expressed from nonhuman cell lines, such as
Prior to the advent of recombinant DNA technology, the Chinese hamster ovary cells. Thus, in patients who have
only available source of human coagulation factors was hypersensitivity to mice or hamsters, allergic side effects
purification from plasma donors. The advantage to this can theoretically be observed. Although it is rare to observe
approach is that the factors are derived from their natural allergic reactions to recombinant products,9 anaphylaxis
biological source, which results in a biochemically authen- after administration of recombinant factor VIII has been
produced from cell lines, they required that the cells be cul-
II
358
E. Special Processes and Products
Chapter 26
Leukocyte-Reduced Products
Walter H. Dzik ● Zbigniew M. Szczepiorkowski
Recipient exposure to allogeneic donor leukocytes can result contact of blood leukocytes with the medium. This require-
in several complications of blood transfusion. To reduce the ment for surface area can be met through the use of filter media
likelihood of these complications, leukoreduction of red composed of fibers with a very small diameter (microfibers) or
blood cells (RBCs) and platelets is widely practiced. This open-cell media that are geometrically like a sponge.
chapter addresses the basis for the adverse effects of recipient Two different approaches to manufacturing such non-
exposure to donor leukocytes and presents the evidence sup- woven media have been commercially viable. The first is a
porting the scientific foundation for leukoreduction of cellu- dry formation process in which a polymer (typically a poly-
lar blood products. In particular, we address the biophysical ester or polyolefin) is melted and extruded through very
mechanisms involved in the preparation of leukoreduced fine nozzles into a turbulent gas stream at high velocity to
RBCs and platelets and the biologic mechanisms account- produce a microfiber. The synthetic microfiber is simultane-
ing for febrile nonhemolytic reactions, primary human leu- ously stretched and cooled to form thin strands of fibers in a
kocyte antigen (HLA) alloimmunization, transmission and process akin to the making of cotton candy. The microfibers
reactivation of cytomegalovirus (CMV), transfusion-asso- are then collected, matted, and heat compressed to a con-
ciated immunosuppression, and the adverse effects from trolled density. Some variation of this basic technique is used
transfusion of leukoreduced blood. The reader is referred to in products from Pall Corporation, Asahi Medical Company,
other sources for a discussion of additional aspects of leuko- Fresenius AG, and MacoPharma, among others.
reduction, including device evaluations, quality control and A second approach is to prepare superfine glass fiber mem-
cell-counting methods, cost effectiveness, and operational branes.5 Foam-like structures with open-cell geometry contain 26
issues.1–3 interconnecting voids that allow a circuitous flow passage. The
manufacture of such media derives from principles used in 359
the manufacture of sponges. For example, filters from Terumo
TECHNOLOGIES FOR THE PREPARATION Corporation are made with the use of a porous polyurethane
OF LEUKOREDUCED BLOOD medium that exhibits an open-cell architecture.
Because they offer an effective pore size that is extremely
Achieving a 10,000-fold reduction in the leukocyte content small, filter media must be designed to have a hydrophilic
of blood requires specialized technologies of leukofiltration surface. Otherwise, the material will fail to “wet” as blood
or apheresis collection. Numerous factors affect the perfor- encounters the medium owing to the inherent surface ten-
mance of leukofiltration, including the temperature of filtra- sion of blood. The synthetic materials used in leukoreduction
tion, the speed of blood flow through the filter, the number filters—polyester and polyolefin microfibers, porous polyure-
of leukocytes presented to the filter, the protein content of thane, and glass microfibers—are all naturally hydrophobic.
the suspending medium, the platelet content of blood, the Manufacturers modify the surface chemistry of these mate-
use of a rinse step after filtration, the storage age of the rials to increase their ability to “wet.” Although prefiltration
blood, and the presence of hemoglobin S in the RBCs to be rinsing with a more hydrophilic liquid (e.g., saline rinse) can
filtered. Standards in the United States require that leuko- wet the filter medium, prerinsing introduces an inconvenience
reduced components be prepared by a method known to that manufacturers of filters generally wish to avoid.
reduce donor leukocytes to residual levels of <5 × 106 white The volume of blood left inside the filter after filtration is
blood cells (WBCs) per unit for RBCs and apheresis platelets referred to as the “hold-up volume.” Filters that have a hard
and <8.3 × 105 WBCs/unit for whole blood–derived plate- external housing often include a venting step at the end of fil-
lets. Methods to produce leukoreduced apheresis platelets tration, which allows sterile air to enter the filter and displace
should produce products with >3 × 1011 platelets. blood that would otherwise have been held up in the device.
One manufacturer (MacoPharma, Tourcoing, France) pack-
ages the filter medium in a flexible plastic housing that col-
Leukofilter Device Design
lapses under atmospheric pressure as blood drains out of the
Removal of leukocytes from whole blood, packed RBCs, or filter. U.S. Food and Drug Administration (FDA) guidelines
platelets depends on a combination of barrier filtration and cell require that the filtration process result in loss of no more than
adsorption to the filter material. Certain principles of design 15% of the original amount of therapeutic blood elements.6
are common to leukofiltration devices.4 A large surface area of The filtration medium is packaged in an external hous-
filter medium is required to allow sufficient opportunity for ing, and it is essential that the medium completely fill the
BLOOD BANKING
housing, allowing no opportunity for the path of blood Biophysical Reasons for Reduced
to flow around the filter (bypass) and thereby circumvent Performance of Leukofiltration
the leukodepletion medium. To appreciate the stringency
required, consider that the leukocyte content of an entire The reduced efficiency of leukofiltration applied to warmed
unit of leukoreduced RBC is equal to the leukocyte content RBCs compared with cold RBCs has repeatedly been dem-
of only one drop (100 μL) of nonleukoreduced blood. onstrated. For example, Beaujean and coworkers10 split RBCs
(storage age, 2 to 10 days) into two equal aliquots, which
were then filtered at either 4°C or 37°C. The mean postfiltra-
Mechanisms of Leukoreduction tion leukocyte content was 10-fold higher for units filtered
at 37°C. Ledent and Berlin8 found leukocyte content to be
Barrier Filtration 100-fold higher among units filtered at 37°C compared with
Simple barrier filtration, in which the effective pore size 4°C. Sirchia and colleagues11 studied filtration under condi-
of the filter medium is smaller than the size of the leuko- tions designed to mimic bedside use. They found that cold
cyte, is a major mechanism used by leukoreduction filters. RBCs warmed to room temperature within 90 minutes after
Modern leukoreduction filters have an effective pore size removal of the units from the refrigerator. They also docu-
on the order of 4 μm. This space is sufficient for passage mented that as the blood warmed, filtration performance
of platelets and deformable erythrocytes but is able to declined, and units failed to meet minimum standards for
retain leukocytes. Because effective leukoreduction by bar- leukoreduced blood (<5 × 106 WBCs/unit). The failure was
rier-based filtration is tightly linked to the deformability striking with units containing as much as 100-fold more leu-
of blood cells, factors affecting cell deformability have an kocytes than the upper threshold level for leukoreduction.
impact on filter performance. Increased leukocyte deform- The performance failure associated with bedside leukoreduc-
ability at higher temperatures probably accounts for the tion is very important to the interpretation of major clinical
poorer performance of leukofiltration when it is applied trials of leukoreduction technology. For example, both the
to room temperature RBCs compared with refrigerated CMV prevention trial of Bowden and associates12 and the
RBCs (see Biophysical Reasons for Reduced Performance HLA alloimmunization trial of Williamson and coworkers13
of Leukofiltration, below). used bedside filtration with filters subsequently determined
to have reduced performance under warm conditions. The
Cell Adhesion performance of newer versions of leukoreduction filters may
Contact-mediated adhesion between leukocytes and the be less sensitive to temperature. Van der Meer and colleagues
filter medium also contributes to the performance of leu- demonstrated enhanced performance at 4°C, but the differ-
kofiltration devices. For contact to occur, there must be ence between the results at 4°C and at 22°C was less than that
a sufficient dwell time of the blood with the medium. For seen in earlier studies.14
the cells not to detach from the medium, shear forces of Several other factors may reduce the effectiveness of leu-
II the flowing blood must not be too strong. High flow rates kofiltration. Excess shear force and resulting cell detachment
can result in insufficient contact and excess detachment of from the filter medium may be an important consideration
leukocytes from the medium. Because cell adhesion to fil- when bedside leukofiltration is combined with mechanical
360
ter media involves a complex surface chemistry that is not blood delivery systems that may “pull” blood through the fil-
fully understood, filter media development has been largely ter at excessive rates. Postfiltration rinsing of filters may also
determined by experimentation. The nature of the fluid in occur during bedside transfusion and has been documented
which cells are suspended, including the plasma protein con- to result in leukocyte detachment. Filtration of excessive vol-
tent and the platelet content, affects the adhesion of leuko- umes of blood or of blood containing excessive numbers of
cytes to synthetic media. Steneker7 demonstrated that during leukocytes can overwhelm the capacity of the filter medium,
leukofiltration of fresh RBCs the presence of viable platelets leading to ineffective leukoreduction.
improves the performance of leukofiltration. This has been Leukofiltration of RBCs has been shown to be less effective
attributed to sticking of platelets to protein-coated fibers in if the donor blood carries hemoglobin AS (sickle trait).15,16
the filter medium, with subsequent binding of leukocytes to Poor leukofiltration is believed to result from reduced red
these activated platelets. Ledent and Berlin8 demonstrated cell deformability among hemoglobin AS erythrocytes under
that replacement of plasma with crystalloid can decrease the conditions of reduced oxygen tension and pH, such as occur
performance of leukofiltration, presumably by decreasing within the filter. Stroncek and colleagues showed that increas-
the concentration of adhesive plasma proteins that participate ing the oxygen tension of hemoglobin AS blood prevented
in leukocyte adhesion to the medium. the poor filtration performance.16 Decreased deformability
Leukofiltration of platelet concentrates presents the of red cells containing hemoglobin S at low oxygen tensions
added complexity that platelets are naturally adhesive under is presumed either to directly “clog” the filter or to reduce
shear conditions in a plasma environment, as a result of von the effective area of filter medium available for retention of
Willebrand factor–mediated platelet adhesion. Therefore, leukocytes. Because sickle trait is relatively common in some
manufacturers of leukoreduction filters designed for plate- donor populations, the frequency of failure of leukoreduc-
lets have further modified the surface chemistry of the fil- tion by filtration may be higher than expected.
ter media to decrease binding of platelets to the filter. For
example, Nishimura and colleagues9 documented that by Leukoreduced Components
adjusting the molar ratio of positively charged diethyl-amino
ethyl methacrylate to negatively charged hydroxyethyl meth-
Prepared by Apheresis
acrylate, the net surface charge on the filter medium could Modern apheresis devices are able to collect concentrates of
be adjusted to optimize leukocyte adhesion but minimize platelets or RBCs that are leukoreduced during collection;
platelet adhesion to the medium. this is referred to as process leukoreduction. Several devices
LEUKOCYTE-REDUCED PRODUCTS
are licensed for collection of one or more components (for In addition to the reduced performance of leukofiltration
review see ref. 17). resulting from filtration at higher temperature or in the pres-
The Amicus apheresis system (Baxter Healthcare, Round ence of hemoglobin S blood, quality control studies of leuko-
Lake, Ill.) is a widely used apheresis system for the prepara- reduction have documented occasional process problems with
tion of leukoreduced platelets. The device incorporates three both leukofiltration and apheresis technologies. For example,
design features to achieve leukoreduction: active interface Kao and colleagues23 reported quality control data from the
control, autoelutriation, and fluid flow dynamics. An optical Trial to Reduce Alloimmunization to Platelets (TRAP). Using
interface detector is positioned within the separation cham- a propidium iodide stain and microscopic chamber counting,
ber to monitor changes in the platelet interface. The system they found that 7% of apheresis platelets and 5% of pooled
recirculates some of the donor plasma into the interface in platelets contained more than 5 × 106 residual donor leuko-
order to separate platelets from leukocytes by elutriation. Just cytes. For RBCs, 0.3% to 2.7% of units failed to meet leuko-
as a wind blows lightweight leaves but not heavy objects across reduction standards depending on the particular leukofilter
a street, plasma pumped through the interface dislodges the used. In addition, they reported substantial losses of platelets
platelets but not the leukocytes into the collection path. The and RBCs as a result of the process of leukofiltration. Although
collection path for platelet collection is in the opposite direc- quality control monitoring of apheresis platelets documents
tion to the flow path for the return of RBCs and plasma to the reliable performance of collection systems,17 as with leukofil-
donor, promoting further separation between donor platelets tration, the apheresis process can fail. Sudden changes in the
and donor leukocytes. rate of blood entering the machine or pauses in blood flow
The Trima Accel automated system separates platelets from during collection will disturb the centrifugal separation of
donor whole blood using a single-stage channel design. Whole blood in the chambers and may interrupt the controlled flow
blood enters the channel and separates under centrifugation paths upon which the cells travel during separation. These
into component layers. Platelets are drawn out through a spe- interruptions can lead to “spillover” of leukocytes into the
cially designed conical-shaped chamber (LRS chamber) while platelet collection stream. Manufacturers have attempted to
plasma and RBCs are returned to the donor. The LRS chamber, engineer alarms into the software that will alert operators to
originally developed for the COBE Spectra System, combines conditions that might result in the failure to collect a leuko-
centrifugal separation technology with saturated fluidized par- reduced product. The leukocyte content of these collections
ticle-bed dynamics to achieve a natural separation of WBCs can then be tested to determine whether the product is leu-
from platelets. White cells are separated from the platelets koreduced. Although refinements in filter design and apher-
according to the fluid dynamics of the separation chamber, the esis technology have improved leukoreduction, there remains
sedimentation velocity of the platelet and white cell compo- value in continuous quality monitoring of blood component
nents, and the plasma flow rate through the chamber. These preparation and modification.
dynamics result in a flow geometry that traps donor leukocytes
in the lower levels of the LRS chamber while allowing platelets
in the upper levels of the chamber to exit toward the platelet CLINICAL INDICATIONS FOR 26
collection/storage bag. Under some donor collection condi- LEUKOREDUCED BLOOD COMPONENTS
tions, leukocytes can spill out of the chamber into the collec- 361
tion product. The Trima System uses a computerized on-line Three prominent clinical indications for leukoreduced cel-
process that monitors for conditions known to increase the lular blood components are (1) to reduce the frequency of
risk of WBC contamination in the platelet product. febrile nonhemolytic transfusion reactions among patients
with a prior history of such reactions, (2) to decrease the
Quality Control of Leukoreduction incidence of HLA sensitization and platelet refractoriness
among patients with hematologic malignancy, and (3) to
Because any leukoreduction process may fail, standards for reduce the risk of transfusion-transmitted CMV infection
quality control testing exist. No common standard exists for among susceptible recipients24 (Box 26–1). Leukocyte-
all nations, however. For example, in the United States the reduction technology has also been applied to other clini-
American Association of Blood Banks standards require that cal situations in the absence of strong evidence but with an
95% of the units sampled meet threshold requirements for expectation of benefit.
residual leukocyte content.18 In Europe, the threshold level of
residual leukocytes is <1 × 106/unit.19 Validation of new leu- Universal Leukoreduction
koreduction methods or devices and ongoing quality control
evaluation requires counting methods specifically designed Use of leukocyte reduction for all cellular components is
for the extremely low concentration of white cells found in controversial25 (Box 26–2). Several countries adopted policies
leukoreduced blood components. A variety of techniques are
available based on microscopy, flow cytometry, and the poly-
merase chain reaction.20 No standards address the relative
proportion of residual leukocyte subpopulations in leuko- BOX 26–1 Indications for Leukoreduced
reduced blood. Although different leukocyte subpopulations Blood Components
are relevant for different biologic effects attributed to donor • Reduce rate of recurrent febrile nonhemolytic transfusion
leukocytes, there is no convincing clinical outcomes evidence reactions (FNHTRs)
that any particular leukoreduction technique is superior • Reduce rate of HLA alloimmunization among hematology-
to any other for removal of specific leukocyte subpopula- oncology patients
tions.21,22 Currently, all methods of leukoreduction that meet • Reduce rate of cytomegalovirus transmission to susceptible
the threshold standard for total number of residual donor recipients
leukocytes are considered equivalent.
BLOOD BANKING
BOX 26–2 Pros and Cons of Universal with the cohort receiving control RBCs. Finally, it was noted
that unadjusted mortality rates also declined among non-
Leukoreduction transfused patients during the two periods of observation.
Pro: Major secondary outcomes included infections, length of
• Patients with selected indications to receive leukoreduced
stay in intensive care, or total hospital length of stay—none
blood will be more likely to receive leukoreduced units of which were statistically different in the two cohorts of
• Inventory management is streamlined patients. In particular, the authors noted that the lack of
Con:
effect on infectious complications argued against a transfu-
sion-related immunosuppression effect. No difference was
• Large clinical studies fail to demonstrate clear or consistent
benefit of leukoreduction on proposed transfusion-associated
seen in the two groups for the proportion of patients requir-
immunosuppression ing ventilation support, hemodynamic support, or renal
• Adds >$500 million each year to health care costs dialysis. In addition, there was no observed beneficial effect
• May adversely affect adequacy of blood supply in setting of of leukoreduction among the subgroups of patients receiv-
recalls of leukoreduction devices; may interfere with supply of ing larger doses of blood (>5 units per patient) compared
blood donors of African ancestry (sickle trait blood) with those receiving fewer transfusions.
• Removes physician choice from product selection One large prospective randomized clinical trial spe-
cifically examined the potential benefit of conversion from
selective leukoreduction to universal leukoreduction.27 This
single center study randomized 2780 patients to receive
either prestorage leukoreduced RBCs and prestorage leuko-
in favor of universal leukoreduction for different reasons. reduced apheresis platelets (for patients needing platelets)
The decision was linked to a program of hemovigilance in versus nonleukoreduced RBCs (buffy coat not removed) and
France, to concern over the potential for spread of trans- nonleukoreduced pooled whole blood–derived platelets. All
missible spongiform encephalopathies by donor leukocytes patients (adult, pediatric, medical, and surgical) not meeting
in the United Kingdom, and to a general reorganization of standard criteria for leukoreduced blood (see Box 26–1) were
blood services in Canada. In the United States, the Blood eligible. Because individual patient consent was not required,
Product Advisory Committee of the FDA voted in 1998 in the study had the advantage that all eligible candidates were
favor of universal leukoreduction in the absence of any con- automatically enrolled (no exclusions). The primary out-
sideration of cost to the health care system and sponsored a come measures were in-hospital mortality, length-of-stay
conference in 1999 that was geared toward implementation in hospital following transfusion, and total hospital costs.
of universal leukoreduction. However, such implementation The study found no difference among primary outcomes for
was never mandated, and studies estimated that universal patients assigned to leukoreduced versus nonleukoreduced
leukoreduction would cost more than $500 million per year blood components. This large trial was powered to have an
II in the United States with uncertain benefit. Nevertheless, 85% chance to detect a 15% difference in the primary out-
despite the absence of any regulatory mandate, some major come with a 95% confidence. Several secondary outcomes
362 blood suppliers in the United States elected to sell only leu- were examined all of which also showed no benefit to leu-
koreduced products, and thus the proportion of leukore- koreduction. These included antibiotic use to treat infection
duced blood components has gradually increased to more following transfusion (all patients) or to treat postoperative
than 50% in the United States. infection (surgical subset), length of stay in intensive care,
The potential benefit of universal leukoreduction among postoperative length of stay, or readmission rates to hos-
surgical patients was investigated in a retrospective multi- pital. There was a statistically nonsignificant trend toward
center study conducted in Canada.26 The study was not a fewer febrile nonhemolytic reactions in the group assigned
prospective randomized trial, but rather compared outcomes to receive leukoreduced blood. Subset analysis restricted
in two cohorts of patients—one cohort transfused prior to to heavily transfused patients failed to show any advantage
the implementation of universal leukoreduction (N = 6982) for leukoreduced blood. The study provided randomized,
and the other cohort transfused after universal leukore- controlled trial evidence for a lack of measurable benefit
duction (N = 7804). The study did not examine outcomes resulting from conversion to universal leukoreduction.
among all transfusion recipients (universal impact) but
rather focused on three groups of more critically ill patients: Prevention of Febrile Nonhemolytic
those with trauma, those undergoing cardiac surgery, and
Transfusion Reactions by Leukoreduction
those undergoing major orthopedic surgery. Because these
patient groups receive large numbers of transfusions, they One of the first indications for the use of leukocyte filters was
were expected to be most likely to demonstrate any benefi- prevention of febrile nonhemolytic transfusion reactions
cial effect of leukoreduction. Prior to adjustment for poten- (FNHTRs). Original studies suggested that the frequency of
tial confounders the authors found a slightly lower risk of FNHTRs to packed RBCs is reduced when the residual leu-
overall mortality (unadjusted odds ratio = 0.87, 95% CI, kocyte content is less than 5 × 108 WBCs/unit.28–32 Although
0.76–0.99, P = 0.04) among cohort 2 (leukoreduced group). FNHTRs are one of the most common transfusion reactions
However, when the results were adjusted for the effects of experienced by recipients of blood components, they are also
cardiac medications, the adjusted odds ratio failed to show a relatively easy to manage. Many, but not all, FNHTRs can be
statistically significant benefit for leukoreduction. Moreover, prevented by leukoreduction of blood components.
within each of the three major patient categories studied Fever is a hallmark and part of the clinical definition of
(trauma, cardiac surgery, orthopedic surgery), the adjusted FNHTR. In addition to fever, patients may experience chills,
odds of death were not statistically different in the cohort rigors, cold, a sense of discomfort, headache, and nausea.
of patients receiving leukoreduced components compared Because the pathophysiology of fever involves a rigor/chill
to 30%.39–41 This significant difference between RBCs and
LEUKOCYTE-REDUCED PRODUCTS
response before the temperature rises, the initial signs of
an FNHTR may be only rigors without temperature eleva- platelet concentrates can be attributed to different mecha-
tion. Therefore, the presence of early-onset fever should nisms involved in development of febrile reactions.
not be required to diagnose an FNHTR.33,34 FNHTRs tend
FNHTR CAUSED BY DESTRUCTION OF
to develop toward the end of a transfusion, and in 10% to
TRANSFUSED DONOR LEUKOCYTES
20% of cases reactions are noted after the transfusion has
been discontinued. Such clinical observations underline First proposed in the 1960s, the original hypothesized
the fact that reactions are dose related and require time to mechanism suggests that donor leukocytes react with
develop. recipient antileukocyte antibodies, causing the donor cells
to release endogenous pyrogens (cytokines) (Fig. 26–1).
Pathophysiology of Fever This mechanism is consistent with the prevention of
Fever is one of the oldest signs and symptoms recognized in FNHTRs by leukoreduction of donor blood, and it is sup-
medicine. Fever results from cytokine generation by activated ported by studies of de Rie and colleagues,42 Decary and
monocytes, macrophages, and Kupffer cells. The involved colleagues,43 Perkins and colleagues,44 Brubaker,45 and oth-
cytokines, including interleukin-1β (IL-1β), IL-6, inter- ers46 that documented the high prevalence of antilympho-
feron-β (IFN-β), and tumor necrosis factor-α (TNF-α), are cyte or antigranulocyte antibodies in the sera of patients
polypeptides that act on the organum vasculosum of lamina who experienced FNHTRs. However, the mechanism fails
terminalis (OVLT).35 The OVLT interacts with the preoptic to account for the fact that FNHTRs are more common
area of the hypothalamus. In the case of TNFα, the induction among recipients of platelets compared with packed RBCs,
of cyclooxygenase-2 in brain blood vessels leads to increased and it failed to explain the observation that FNHTRs may
production of prostaglandins (mainly prostaglandin E2)36 occur among men with no history of prior transfusions. In
and subsequently to fever.37 Engel and associates38 reported addition, the mechanism can be challenged because leu-
the peripheral blood levels of cytokines measured in patients kocytes are not known to store IL-1 and TNF47 suitable
with fever and neutropenia. The median peak concentra- for “immediate release” upon destruction by recipient
tion for IL-6 was 400 pg/mL (range, 100 to 41,000 pg/mL); antibody. Finally, the original mechanism fails to explain
for IL-8, it was 1025 pg/mL (range, 600 to 26,000 pg/mL); for why patients with anti-HLA antibodies have febrile reac-
TNFα, less than 10 pg/mL; and for IL-1β, 17 pg/mL (range, tions when transfused with leukocyte-reduced platelet
<10 to 36 pg/mL). concentrates.
Incidence and Mechanisms of FNHTR FNHTRS CAUSED BY PASSIVE TRANSFER OF CYTOKINES
The reported incidence of FNHTR depends on the trans- In a series of simple and well-designed experiments, Heddle
fused component. Transfusions of nonleukoreduced RBCs and associates48 focused attention on the plasma constitu-
are associated with 0.5% to 6.0% risk, whereas transfu- ent of stored platelet concentrates, rather than the platelets
sions of platelet concentrates carry a risk as high as 20% themselves, as the source of febrile reactions after transfu- 26
363
Platelets
Recipient Mφ
Figure 26–1 Mechanisms underlying febrile nonhemolytic transfusion reactions. (Adapted from Klein HG, Dzik WH, Slichter SJ, Hillyer CD, Silberstein
LE. Leukocyte-reduced Blood Components: Current Status. Educational Program, American Society of Hematology, 1998, pp 154–177.)
BLOOD BANKING
sion of platelet concentrates. The authors selected 4- and dent of leukocyte concentration but correlated with platelet
5-day-old platelet concentrates. Using centrifugation, the concentration.
platelet-poor plasma was separated from the platelet pellet,
which was resuspended in fresh plasma. Patients were then Monocyte activation. Monocytes are a major constituent
transfused with both components in random sequence, with of the leukocyte population in platelet concentrates. They
a 2-hour washout period between transfusions. Signs and are also capable of secreting cytokines, namely IL-1β, IL-
symptoms suggestive of FNHTR were assessed. Transfusions 6, and TNFα. Several authors have investigated monocyte
of the platelet-poor plasma obtained from stored platelet activation in platelet concentrates as a potential explana-
concentrates were associated with a significantly higher rate tion for cytokine accumulation. Muller-Steinhardt and
of FNHTR, compared with the resuspended platelet pel- associates56 studied the influence of storage time, tem-
lets. The authors concluded that soluble substances that had perature, and type of anticoagulant on the capability of
accumulated in the plasma during platelet storage were pri- mononuclear cells to secrete cytokines. Mononuclear cells
marily responsible for febrile reactions in the older platelet were harvested on days 1, 3, and 5 from platelet concen-
concentrates. Increased levels of two cytokines, IL-1β and trates stored under routine conditions, and the response
IL-6, were present in stored platelets and correlated with the to mitogenic stimulants, such as lipopolysaccharide (LPS),
frequency of observed reactions. phytohemagglutinin, and staphylococcal enterotoxin B, was
evaluated by measuring secretion of IL-1β, IL-2, IL-6, and
Mechanism of cytokine accumulation in platelet concen- IFNα. The ability of monocytes to secrete cytokines did not
trates. Cytokines include a large family of molecules change significantly during the first 3 days of storage and
involved in innate immunity and cell signaling. After release decreased to 25% to 50% of original levels by day 5 of stor-
by mononuclear phagocytes, T lymphocytes, and several age. Mitogenic response of mononuclear cells was signifi-
other cell types, cytokines bind to specific receptors on tar- cantly higher at 37°C than at 22°C. These findings showed
get cells. Cytokine receptor chains cluster, and their intra- that under normal storage conditions mononuclear cells in
cellular portions are phosphorylated by Janus kinases (Jaks). platelet concentrates preserve the ability to synthesize and
The phosphorylation step is required for the Src homol- secrete cytokines for at least 5 days. Heddle and coworkers57
ogy–2 (SH-2) portion of the Jak to bind to the receptor. SH-2 demonstrated that cytokine accumulation during storage is
facilitates binding of a signal transducer and activator of temperature dependent. Platelet concentrates were split into
transcription (STAT) protein. After phosphorylation, these two identical portions, one of which was stored for 5 days at
STAT proteins form homodimers, which migrate to the 22°C and the other at 4°C. Cytokines failed to accumulate in
nucleus and bind to nuclear factors such as nuclear factor-κB the aliquots stored at 4°C.
(NF-κB) or activation protein–1 (AP-1) transcription fac- Grey and associates58 reported in vivo activation of
tors. The transcription factors then activate genes involved monocytes in platelet concentrates. They analyzed plate-
in innate immunity.49 let concentrates for leukocyte and monocyte total count,
II A number of reports showed increased accumulation of CD14 and CD16 monocyte-associated antigen expression,
cytokines, especially IL-1β, IL-6, IL-8, TNF-α, and RANTES and IL-1β and IL-6 concentrations on days 1, 2, 3, 4, and
364 (regulated on activation, normal T expressed and secreted), 5. The analyzed monocytes expressed increased levels of
with prolonged storage of platelet concentrates.50–52 Increased CD14 (LPS receptor) and CD16 (FcRIII), which are both
cytokine levels were observed especially on the fourth and markers of activation, starting on the first day of storage.
fifth day of storage. Stack and Snyder51 assayed 2-, 3-, 4-, On day 3, more than 50% of platelet concentrates had an
and 5-day-old platelet concentrates for IL-1β, IL-6, IL-8, increased IL-6 concentration. The elevation of IL-6 and IL-
and TNFα. Although IL-8 was the cytokine most frequently 1β correlated with the number of monocytes in the unit on
detected, IL-8 is not regarded as a fever-producing cytokine. day 1. However, increased IL-6 levels occurred earlier dur-
Increased IL-1β was observed in the units with elevated IL-8. ing storage. Several hypotheses have sought to explain the
In general, the highest levels of IL-8 were found in the units delay between monocyte activation and detectable elevation
with the longest storage times and highest leukocyte counts. of interleukins. TNFα and IL-1 are known to induce IL-1
Only 8% and 10% of tested units showed detectable levels of synthesis and together with PDGF can stimulate IL-6 syn-
IL-6 and TNFα, respectively. Leukoreduction before storage thesis.59 Both IL-1α and IL-1β were found in the cytoplasm
prevented the accumulation of IL-8 and IL-1β to day 5 of of resting and thrombin-activated platelets.60 It is possible
storage. This study underscored the importance of storage that the cumulative effect of PDGF and IL-1 derived from
time and initial number of leukocytes before storage in the platelets reaches a threshold concentration necessary to trig-
generation of cytokines. ger monocytes to generate additional cytokines such as IL-8
Palmer and coworkers53,54 combined results of their two and IL-6. This hypothesis is supported by the findings by Aye
studies and showed that cytokine accumulation during stor- and colleagues53 indicating that, despite uniform release of
age of platelet concentrate correlated with the number of PDGF from platelets during preparation and storage, only
leukocytes present before storage. Units with a leukocyte components with high leukocyte content had a detectable
concentration greater than 100/μL resulted in detectable level of cytokines.
levels of IL-8 and IL-1β. Although prestorage leukoreduc- Mononuclear cells in stored platelet concentrates may be
tion prevented cytokine accumulation, leukoreduction had activated by other mechanisms. Activation of monocytes by
no beneficial effect on markers of platelet activation such as the plastic used in storage bags was studied by El-Kattan and
P-selectin, transforming growth factor-β1 (TGF-β1), plate- coworkers.61 Whole blood–derived platelets were stored in
let-derived growth factor AB (PDGF-AB), von Willebrand plastic bags made of polyolefin (POF), and apheresis platelets
factor, and serotonin. These results were confirmed in a study were stored in the bags made of polyvinyl chloride (PVC).
by Fujihara and colleagues,55 who showed that accumula- Mononuclear cells showed preferential adherence to POF
tion of platelet-derived RANTES and TGF-β1 was indepen- compared with PVC. This adherence was associated with
LEUKOCYTE-REDUCED PRODUCTS
increased mean cytokine levels (IL-1β, TNF, IL-6) that were from the baseline of 0 ng/L to 5 to 13 ng/L. Thus, a 42-day-old
11- to 48-fold higher in POF bags compared with PVC bags. RBC with 125 mL of supernatant would passively transfer 0.6
to 1.6 ng of IL-1. These values are substantially below the 700
FNHTRS CAUSED BY CYTOKINE PRODUCTION
to 7000 ng required to achieve fever levels (see above).
BY RECIPIENT MACROPHAGES
SUMMARY OF THE MECHANISMS OF FNHTRS
A third mechanism for FNHTRs depends on cytokine release
by the recipients’ own macrophages (see Fig. 26–1). It has been FNHTRs to RBCs are best explained by release of recipient
proposed that recipient antibody bound to donor-cell antigen cytokines (immunocomplex mechanism). In particular,
may form an immunocomplex that serves to activate recipi- cytokines do not accumulate during refrigerated storage
ent macrophages to release inflammatory cytokines.62 This of red cells to levels considered sufficient to induce a clini-
mechanism does not depend on release of preformed stored cal febrile response.69,70 In contrast, FNHTRs to platelets
cytokines contained within donor leukocytes but acknowl- that have not undergone prestorage leukocyte reduction
edges that immunocomplexes are a known stimulus for can be attributed to either accumulation of cytokines
macrophage activation. Indirect support for this hypothesis during storage or to release of recipient cytokines (immu-
comes from a number of observations, including the develop- nocomplex mechanism). When prestorage leukoreduced
ment of fever among alloimmunized recipients of prestorage platelets are implicated in FNHTRs, the most likely mech-
leukoreduced platelets; fever after transfusion of platelets to anism is the immunocomplex in which donor platelets are
patients with immune thrombocytopenia or drug-induced the target of recipient antibodies. Such reactions may be
thrombocytopenia; fever among recipients of incompatible a clinical tip-off to the presence of HLA antibodies and
RBC transfusions; and the fever reactions to the cellular por- refractoriness.
tion of platelet concentrates seen in the study by Heddle and
Clinical Evaluation of FNHTRs
associates48 cited earlier.
Studies30,31 in chronically transfused patients with thalasse-
CYTOKINE CONCENTRATIONS IN VIVO AND IN VITRO
mia have documented that leukoreduction is a highly effec-
Sacher63 attempted to establish the causality between cyto- tive means to prevent FNHTRs to RBCs. Current methods
kines and FNHTR. IL-6 levels were measured in vivo before of leukoreduction—whether done before storage, just before
and after transfusion in a group of 42 patients. Acute blood issue, or at the bedside—are capable of reducing the
transfusion reactions occurred in 26 patients. The mean residual donor leukocyte concentration to levels well below
post-transfusion IL-6 level was 3.7-fold higher than in pre- those that result in FNHTRs. Moreover, owing to refriger-
transfusion specimens. Patients without acute reactions had ated blood storage (and in contrast to platelet concentrates),
insignificant increases in IL-6 levels. Unfortunately, the con- RBCs do not accumulate clinically important levels of cyto-
centration of IL-6 in the transfused components was not kines during storage. Indeed, fever reactions to properly
measured, leaving uncertainty as to the origin of the cyto- leukoreduced RBCs are sufficiently rare that patients who
kine elevation (i.e., a blood component or the recipient). receive leukoreduced RBCs and experience fever should be 26
In the absence of conclusive clinical trials, one may ques- evaluated for the presence of hemolytic reactions due to RBC
tion which cytokines found in platelet concentrates may cause blood group incompatibility and for the presence of bacte- 365
FNHTRs. The amount of TNF-α known to cause fever and rial contamination of the transfused product. In contrast to
chills is approximately 5 to 10 μg/m,2 or 8500 to 17,000 ng RBCs and as described earlier, FNHTRs to platelet concen-
of TNF-α for a 70-kg person.64,65 In order to achieve such trates are much more common and are not completely elimi-
an amount after transfusion of a blood component, the nated by leukoreduction. Patients experiencing FNHTRs to
amount of TNF-α present should be approximately 28,000 platelet concentrates should be evaluated for platelet incre-
to 56,000 ng/L in a 300-mL unit, or 170,000 to 340,000 ng/L ments, for evidence of alloimmunization, for drug-induced
in a 50-mL unit. However, the measured levels of TNF-α in platelet refractoriness, for bacterial contamination of plate-
platelet concentrates were far below this level, with the high- lets, and for passive transfer of antibodies to ABO antigens
est reported concentration being 1890 ng/L and the median from donor to recipient.
concentration in the range of 42 to 571 ng/L.51–53 The data
suggest that passive transfer of TNF-α cannot be solely Prevention of HLA Alloimmunization
responsible for symptoms observed in patients with FNHTR.
The situation is slightly different for the other endogenous Alloimmunization to HLA donor antigens is a well-recog-
pyrogen, IL-1. This cytokine can cause clinical symptoms nized complication of blood transfusion. Clinical conse-
at concentrations as low as 10 to 100 ng/kg, corresponding quences of HLA alloimmunization include FNHTRs, renal
to 700 to 7000 ng in a 70-kg recipient.66–68 The maximum allograft rejection, and platelet refractoriness. There are
measured levels of accumulated IL-1 in platelet concentrates many nonimmune causes of platelet refractoriness, includ-
vary from 143 to 26,000 ng/L, with medians ranging from 14 ing fever, use of amphotericin B, drug-related antiplatelet
to 5250 ng/L.51–53 These median values correspond to 0.7 to antibodies, hypersplenism, consumptive coagulopathy, and
260 ng in a 50-mL unit or 4.2 to 1560 ng in a 300-mL unit. idiopathic thrombocytopenic purpura. Because these con-
Passive transfer of such quantities of IL-1 might be able to ditions contribute to platelet refractoriness, no studies have
cause fever in a 70-kg person. documented that prevention of alloimmunization by leu-
Cytokine accumulation during storage of RBCs is not likely koreduction prevents bleeding complications due to platelet
to contribute to FNHTRs. The measured levels of cytokines are refractoriness. Nevertheless, prevention of HLA alloimmu-
substantially lower than found in the platelet concentrates.69 nization is generally regarded as an important benefit of
Stack and coworkers70 analyzed levels of IL-1β, IL-6, and IL-8 leukoreduction.
in leukoreduced and nonleukoreduced RBCs. Over 42 days After the demonstration in rodents by Claas and colleagues
of storage, IL-1β accumulated in the nonleukoreduced units that leukocytes and not platelets are responsible for primary
BLOOD BANKING
alloimmunization,71 leukoreduction of blood components CD4+ T cells were shown in a rodent transfusion model to
was the subject of numerous clinical trials assessing alloim- serve as direct APCs to recipient CD8+ cells.86
munization to donor HLA antigens. The reported rate of Kao and del Rosario provided additional direct experi-
anti-HLA alloimmunization due to unmodified components mental evidence for the importance of HLA class II cells
in randomized controlled trials varied from 20% to 50% in in transfusion-induced alloimmunization to class I MHC
the control arm, with median incidence of 42%.72 Meta- antigens.87 Using a rodent transfusion system, they com-
analysis of eight controlled, randomized trials demonstrated pared alloantigen response to transfusion of unmodified
a 70% reduction in the incidence of HLA alloimmunization donor mononuclear cells versus transfusion of mononuclear
in the group of patients who received leukoreduced blood cells that were first depleted of cells bearing class II MHC
components.13,72–79 The same report identified a correspond- antigens. Alloantibodies against class I MHC antigens were
ing reduction in platelet refractoriness. The currently recom- generated in 100% of mice infused with unmodified mono-
mended level of leukoreduction to prevent allosensitization nuclear cells, whereas only 25% of mice transfused with the
is less than 5 × 106 WBCs/unit (3 to 17 WBCs/μL).80,81 modified components became alloimmunized. This study
Immune recognition of foreign donor cells requires at confirmed the validity of direct immunization mediated by
least three fundamental elements: binding of the antigen donor APCs and also showed variability in response among
to the antigen receptor, binding of costimulatory molecules different strains.
mediating cell–cell contact, and local elaboration of cyto- Clinical studies of transfusion-induced HLA alloim-
kines and appropriate cytokine receptors. In whole blood, munization suggest that donor class I peptides presented
the majority (70%) of class I HLA antigen is found on the by class II cells are more immunogenic than intact class I
surface of platelets, which express 50,000 to 100,000 cop- molecules found on donor platelets. The failure of plate-
ies at their surface.82,83 The rest of the HLA molecules are lets to directly provoke immunization presumably reflects
distributed among RBCs (3%), granulocytes (2%), lympho- the absence of critical costimulatory molecules on plate-
cytes (2%), and plasma (23%). Platelet concentrates contain lets. For example, Gouttefangeas and coworkers showed
approximately 3.4 mg of HLA molecules, of which 3.2 mg that HLA class I molecules from platelets cannot directly
(94%) are associated with platelets, 0.2 mg is in plasma, and induce allogeneic CD8+ cytotoxic T-cell response in vitro.88
17 μg are present on leukocytes.82 Therefore, most of the Moreover, pure platelet suspensions are unable to stimu-
HLA antigens in nonleukoreduced platelet concentrates are late cells in a mixed lymphocyte reaction. Because leukore-
found on the platelets themselves. For this reason, it may duction depletes blood of cells bearing the combination of
seem counterintuitive that leukoreduction decreases the risk HLA antigens and costimulatory molecules, and because
of alloimmunization. neither RBCs nor platelets display the combination of class
The recipient of platelet concentrates is exposed to dif- I antigen, costimulatory molecule, and relevant cytokine,
ferent forms of HLA antigens: soluble class I/II antigens in leukoreduction of RBCs or platelets prevents direct HLA
the plasma, class I/II antigens on cell fragments shed from alloimmunization.
II leukocytes and platelets during processing and storage, class
Indirect Allorecognition Pathway
I antigens present on intact platelets, and class I/II antigens
366 expressed on leukocytes. The route by which the antigen pre- Indirect allorecognition refers to the process by which
sentation occurs influences the likelihood of alloimmuniza- recipient APCs first engulf donor cells and then process
tion. Indeed, an important feature of transfusion-induced donor antigen for redisplay to the recipient immune
alloimmunization is the availability of two different sets of system. Donor cells, cell fragments, and soluble donor
antigen-presenting cells (APCs): those of donor origin and antigens undergo endocytosis by phagocytosis, macropi-
those of recipient origin. Stimulation of recipient T or B cells nocytosis, or a clathrin-mediated process.89 The proteins
by donor APCs has been called “direct” alloimmunization. In are degraded to small peptides within a specialized lyso-
contrast, stimulation by recipient APCs, presenting peptides somal compartment termed the MHC class II compartment
of donor origin, has been called “indirect” alloimmunization. of the recipient APCs. The peptides are then loaded onto
The data derived from clinical trials of leukoreduction and class II molecules as follows. The MHC invariant chain (Ii),
from basic research suggest that the direct alloimmunization a chaperone molecule that guides α and β chains of class
pathway plays a more important role in transfusion-induced II molecules from the endoplasmic reticulum, is digested
alloimmunization.82 proteolytically. A fragment of Ii, the class II–associated Ii
peptide (CLIP), remains associated with αβ dimers and
Direct Alloimmunization Pathway occupies the peptide-binding groove. In addition, HLA-
Direct alloimmunization refers to the process by which recip- DM stabilizes the αβ complex and facilitates binding of
ient immune cells respond directly to donor HLA antigens the peptides to the peptide-binding groove. The allogeneic
without the processing of donor antigens by recipient APCs. peptides are then loaded into the MHC class II groove,
The mixed lymphocyte reaction is an in vitro example of replacing CLIP or HLA-DM.90 Empty class II molecules—
direct T-cell recognition. Direct sensitization to donor class those that contain neither peptide, invariant chain, nor
I antigens results when the recipient is exposed to donor cells HLA-DM—are unstable and are degraded in the low
bearing class II structures. The route by which donor peptides pH compartment of the lysosomes. Because HLA-DM is
derived from class I antigens are directly presented has not found at a fivefold lower concentration than HLA class II
been precisely delineated. Class II–positive donor cells might in the late endosomal compartments, excess self-HLA class
carry within the peptide-binding groove oligopeptides rep- II molecules are presumably degraded and their peptide
resenting the cell’s own HLA class I antigen. It is known, for fragments recycled.
example, that a proportion of the endogenous peptides eluted Since transfusion typically results in alloimmunization to
from major histocompatibility class (MHC) molecules are class I HLA antigens, trafficking of class I molecules in donor
from degraded self-MHC molecules.84,85 In addition, donor APCs is of special interest to transfusion science. Although
LEUKOCYTE-REDUCED PRODUCTS
no evidence currently documents that class I peptides are the classic endosome-dependent processing of exogenous
concentrated in the MHC class II compartment, current antigen and resulted in production of IgG1 antibodies. The
technical limitations make it hard to determine the propor- second pattern was insensitive to both chloroquine and pH,
tion of peptides bound to class II that originated from class suggesting a nonendosomal pathway, and led to an IgG2a
II compared with class I molecules.91 However, Turley and alloimmune response.
colleagues reported that class I HLA molecules can accom-
Clinical Studies of Leukoreduced Components
pany class II HLA molecules from the endoplasmic reticu-
to Prevent Alloimmunization
lum, where the class I molecules may be potentially degraded
and their peptides loaded onto class II structures ultimately There have been eight prospective, randomized, clinical
expressed on the surface of the cell.92 trials of filter-leukoreduced blood components to prevent
Expression of costimulatory molecules by APCs is platelet alloimmunization. The populations studied were
presumably required for transfusion alloimmunization. patients with chronic thrombocytopenia and, in most of the
Both CD80 and CD86 can bind to CD28 found on T-helper investigations, acute myelogenous leukemia. As reviewed
(Th) lymphocytes, thereby promoting T-cell activation. by Heddle94 and by Vamvakas,72 these trials varied greatly
Activated helper T cells in turn increase their expression in experimental design. For example, they differed in such
of CD40 ligand (CD40L) and the IL-2 receptor. CD40L fundamental issues as exclusion criteria, definition of allo-
interacts with CD40 present on B cells to induce activa- immunization and platelet refractoriness, methods of leuko-
tion of B cells and expression of B7-2, and later B7-1, on reduction, consistency of leukoreduction, and numbers and
the surface. Resting B cells do not express CD80/CD86 on types of patients enrolled. The majority of studies showed
their membrane, but on activation, either by cytokines or that fewer patients in the study arm receiving leukoreduced
by activated T lymphocytes, B cells are able to fully interact components developed lymphocytotoxic antibodies (Table
with T cells and follicular dendritic cells in germinal cen- 26–1). However, this difference between groups was less pro-
ters. In addition, a profile of cytokines secreted by T cells is nounced when the researchers looked at clinically significant
needed for B-cell activation and antibody synthesis. A Th1 platelet refractoriness. Not surprisingly, these studies docu-
response is associated with increased secretion of IL-2 and mented that alloimmunization to platelet-specific antigens
IFN-γ whereas a Th2 response is characterized by secretion was not affected by leukoreduction.
of IL-4, -5, -10, and -13. The cytokines generated by T cells The largest and the most authoritative prospective ran-
allow for B-cell proliferation, maturation, and antibody domized trial was the TRAP study, which enrolled 268
production. patients with acute myelogenous leukemia.79 The results
Investigators have attempted to dissect the pathway of demonstrated conclusively that leukoreduction of blood
antigen processing by recipient APCs in response to trans- components reduces the rate of alloimmunization among
fusion. Bang and colleagues93 developed a system in which patients with leukemia. The study also demonstrated that
allogeneic platelets were incubated with the recipient’s APCs leukoreduced, pooled, whole blood–derived platelets are as
in the presence of various compounds that potentially affect effective as apheresis platelets for the prevention of alloim- 26
intracellular peptide processing, such as IFN-γ, aminogua- munization. Although the rate of HLA alloimmunization
nidine, l-arginine, colchicine, ammonium chloride, chlo- was higher among patients with a history of prior pregnancy, 367
roquine, brefeldin A, and a cytosolic proteasome inhibitor compared with never-pregnant patients, the alloimmuniza-
(MG115). The pulsed APCs (enriched spleen macrophages) tion rate among previously pregnant patients was lower in
were then injected into recipients, and antidonor IgG pro- the leukoreduced arms than in the control arm. The finding
duction was evaluated. Forty-five percent of recipients devel- that patients who were previously pregnant might benefit
oped alloantibodies after two infusions, and all recipients from leukoreduction differed from the results of other stud-
were alloimmunized by the sixth transfusion. Two patterns ies that showed lack of efficacy of leukoreduction for such
of response were observed. The first was consistent with patients.78 Despite the size of the TRAP trial, the measured
Adapted from Klein HG, Dzik WH, Slichter S, et al. Leukocyte-reduced Blood Components: Current Status. Educational Program, American
Society of Hematology, 1998, pp 154–177.
BLOOD BANKING
rate of platelet refractoriness was low: 16% in the control The authors noted lack of statistically significant difference
arm and 7% in the study arm (P = 0.03). Platelet refractori- between the groups with the rate of alloimmunization of
ness was strictly defined as corrected count increments of 17% and 5%, respectively. This somewhat counterintuitive
less than 5000/μL on two sequential transfusions. The study result may reflect the competence of the immune system
defined the primary outcome as concurrent development of in the study groups. These two reports raise an important
both antibodies and platelet refractoriness. Platelet refracto- question of generalizability of leukoreduction as a means
riness (defined as described) that was present within 2 weeks to prevent alloimmunization in both immunosuppressed
after the development of antibodies constituted alloimmune and immunocompetent individuals. Additional studies are
platelet refractoriness. By these criteria, 13% of patients in clearly needed to answer this important question.
the control arm and 3% of those in the treatment arm had
alloimmune platelet refractoriness (P = 0.004). Cytomegalovirus: Transmission
A recent study on the value of prestorage leukoreduc-
and Reactivation
tion on alloimmunization comes from Canada.95 In this one
center study a retrospective analysis was performed on 617 CMV is a member of the herpes family of DNA viruses and is
patients with hematologic malignancies who were treated a significant pathogen for immunocompromised individuals.
with chemotherapy and/or stem cell transplantation. The Approximately 2 decades of clinical and basic research pro-
authors note that the level of alloimmunization strongly vide evidence that CMV can be transmitted by donor leuko-
correlates with history of previous transfusion of nonleuko- cytes and that transmission can be reduced by leukoreduction.
reduced components, pregnancy, and transfusion of more Clinical trials documenting the effectiveness of leukoreduction
than 13 doses of platelets. This study carries a number of as a means to reduce the risk of transmission of CMV have
caveats including its retrospective design; multi–year span, been reviewed elsewhere.98,99 Testing of donors for evidence
which may have affected treatment outcomes; and popula- of antibodies to CMV is another widely practiced method for
tion of patients who routinely receive leukoreduced compo- reducing the risk of CMV transmission by blood components.
nents in the United States. It is then difficult to interpret the Please see Chapter 46 for a detailed discussion.
observed decrease of alloimmunization. Though the number
Clinical Trials of Transfusion-Transmitted
of patients with pre-existing antibodies was small, there was
CMV and Leukoreduction
no evidence that universal leukoreduction affected their level
of alloimmunization. From among the many studies on transfusion-transmitted
The collective results from existing clinical trials support CMV, two clinical trials stand out. Both were large studies
the use of leukoreduced components to prevent primary and involved bone marrow transplant patients.
alloimmunization in patients with acute myelogenous leu- Bowden and associates performed a prospective, random-
kemia who are receiving induction chemotherapy. However, ized trial of leukoreduction versus CMV serologic screening
there have been few studies of leukoreduction and alloim- for prevention of transfusion-transmitted CMV.12 Patients
II munization in other patient groups. Because treatment of undergoing bone marrow transplantation (BMT) who were
leukemia is itself immunosuppressive, the impact of leuko- CMV seronegative were randomly assigned to receive either
368 reduction may be different in other patient groups, and the CMV-seronegative nonleukoreduced blood components (N
efficacy of leukoreduction for prevention of HLA alloimmu- = 252) versus CMV-untested leukoreduced blood products
nization among nonleukemic patients has not been formally (N = 250). The primary outcome was CMV infections occur-
demonstrated. ring after day 21 from transplant. This outcome was selected
The prevention of alloimmunization in immunocom- because infectious outcomes before day 21 of transplant were
petent patients has not been studied extensively. We can get considered to be the result of exposure prior to the study
a glimpse of the potentially complicated issues from two period. The primary analysis found no difference in the rate
recently published reports. The group from Leiden studied of CMV infection (CMV− = 0 1.3% vs LR = 2.4%, P = 1.00)
the influence of buffy coat depletion with additional pre- or or CMV disease (CMV− = 0% vs LR = 2.4%, P = 1.00). The
poststorage leukoreduction on formation of new anti-WBC study was designed to have 80% power to detect a 5% dif-
and/or anti-RBC alloantibodies in patients undergoing car- ference between the two study arms with 95% confidence.
diac surgery.96 Approximately 75% of patients in this study In a secondary analysis that included all events during days
had no detectable antibodies against WBC or RBC prior to 0 to 100, there was still no statistical difference in the rate of
transfusion. After a single transfusion episode with multiple CMV infection, but the rate of CMV disease was lower in
units of either buffy coat–depleted RBCs, prestorage leuko- the CMV-seronegative arm (CMV− = 0% vs LR = 2.4%, P
reduced RBCs, or poststorage leukoreduced RBCs, approxi- = 0.03). The authors concluded that leukofiltration was an
mately 10% of patients in each group developed anti-WBC effective alternative to the use of seronegative blood compo-
antibodies. In addition, the patients who were alloimmu- nents for prevention of transfusion-transmitted CMV. The
nized prior to transfusion increased their panel reactivity study sparked lively controversy on the difference between
in almost 30% of cases, again without significant difference the primary and the secondary analysis of the data.100 It
between study arms. This study raises a question of a fail- was also noted that patients could receive up to 6 units of
ure of leukoreduction to protect nonimmunosuppressed “nonstudy” blood and remain in their original assigned arm
patients from developing both anti-WBC and anti-RBC although these violations appeared to account for only one
antibodies. An interesting observation was published by infection in each arm.
Ohto and colleagues.97 The group performed a multicenter The interpretation of the Bowden paper involves two
randomized trial of bedside leukoreduced (average content important details. Firstly, RBCs in the leukoreduced arm
of 0.3 × 106 WBC/unit) versus buffy coat–depleted (average were leukofiltered at the bedside. It was not known at the
content 1234 × 106 WBC/unit) RBCs administered to non- time of the study that bedside leukoreduction of RBCs has
immunocompromised patients undergoing cardiac surgery. a very high failure rate.11 Thus, patients assigned to the
LEUKOCYTE-REDUCED PRODUCTS
leukoreduction arm of the study received products that had blood support between cohort 1 and cohort 2 did not involve
far inferior leukoreduction compared with today’s technol- RBCs but rather was a change from leukofiltered platelets to
ogy. On the other hand, the sensitivity of assays for CMV process leukoreduced apheresis platelets. One possible expla-
serologic screening has remained stable since the Bowden nation is that the use of leukofiltered RBCs from CMV-sero-
trial. Thus, if the trial were repeated today, it would be rea- positive donors may simply have been a marker (correlate)
sonable to expect an even lower rate of CMV transmission for sicker patients whose increased transfusion demands
in the leukoreduction arm. The Bowden study also provides resulted in the blood bank selecting CMV-seropositive
some insight into the issue of HLA refractoriness and fatal donors. Indeed, patients acquiring CMV from transfusion
hemorrhagic outcomes. While higher rates of HLA alloim- had a statistically higher overall blood requirement than
munization and platelet refractoriness would be expected to those not acquiring CMV infection.
have occurred in the patients randomized to receive nonleu- It is difficult to make firm policy decisions based on the
koreduced blood, there was no observed impact of study arm Nichols study. Although the authors’ conclusions were that
assignment to overall survival. Thus, any degree of benefit serologic screening might be superior to leukofiltration,
from leukoreduction regarding prevention of fatal bleeding other interpretations of the findings are possible: leukofil-
due to platelet refractoriness was so small as to be undetect- tration of platelets might be superior to process leukoreduc-
able in a trial of 500 BMT patients. Although the Bowden tion by apheresis, the use of CMV-seropositive leukofiltered
study remains the best clinical trial to date on the topic of RBCs might have been a correlate for sicker patients, patients
CMV transfusion-transmitted CMV, its findings need to from cohort 1 may have had by chance an unusually low
be interpreted in light of the technology used to prepare background incidence of CMV, or the results may have been
leukoreduced blood. confounded by other influences.
A second major study among BMT patients was the Comparative trials of leukoreduction versus CMV sero-
report by Nichols and colleagues, who took an entirely dif- testing have not shown a clear benefit of one method over
ferent approach and reached opposite conclusions.101 They the other. Although a meta-analysis concluded that serologic
compared CMV infection in two cohorts of patients. One testing was superior,102 that analysis was dominated by results
cohort was treated from 1994 to 1996 and received mostly of the Bowden trial and the Nichols trial. As noted above,
CMV-seronegative blood components. When these were the Bowden study used bedside leukofiltration prior to the
not available for individual patients, the patients received discovery that bedside leukofiltration has a high failure rate
leukofiltered blood from CMV-positive donors. The second and the Nichols trial was a “before-versus-after” study design
cohort was patients treated from 1996 to 2000 who received with several methodological weaknesses.
similar blood components except that platelet support was
CMV Is Present in Blood Donors as a
provided with apheresis platelets collected on machines
Latent Infection in Mononuclear Cells
designed to produce leukoreduced platelets (process leuko-
reduced apheresis platelets). The study design was, there- Almost all studies of the epidemiology of CMV infection
fore, a comparison of “before versus after” introduction of identify that the virus is common in healthy individuals and 26
apheresis leukoreduced platelets. The principal finding of that the prevalence of serologic markers for prior infection
this study was that the incidence of transfusion-transmitted increases with age. Although carriers of CMV may intermit- 369
CMV was 1.7% in the first cohort versus 4.0% in the second tently shed virus in their saliva, they do not have continuous
cohort (P = 0.05). Because the two cohorts were sequential in viremia. Rather, CMV is present in latent form—the viral
time, the results are subject to other confounding factors that genome is present, but gene expression is limited and infec-
differed during the two periods. Furthermore, the incidence tious virus is not produced. Blood cells, endothelial cells, tis-
of CMV in the “before” cohort (1.7%) may have been unusu- sue macrophages, stromal cells, and neural cells are all sites
ally low that year by chance. This is suggested by the authors’ of CMV latency.
observation that the usual background incidence of transfu- Leukoreduction reduces the likelihood of transfusion-
sion-transmitted CMV in their population is 2% for patients transmitted CMV because the virus is tropic for leuko-
transfused exclusively with CMV-seronegative donor blood. cytes and is not found in erythrocytes, platelets, or plasma
The authors did find that patient exposure to leukofiltered of healthy blood donors. CMV DNA is difficult to recover
blood from CMV-seropositive donors was marginally higher from leukocyte DNA of healthy carriers. Whereas CMV is
during period two. For example, the mean number of leuko- found in polymorphonuclear leukocytes of patients with
filtered RBCs from CMV-seropositive donors was 0.08 units active infection, viral DNA is found in mononuclear cells of
in period one and 0.15 in period two (P = NS). healthy donors.103,104 Polymerase chain reaction (PCR) has
Analysis of the data provided complex results. Using uni- been used to demonstrate that blood monocytes and mac-
variate analysis for the cohort 2 patients, the odds of CMV rophages (rather than T and B cells) are the principal sites
infection by transfusion rose weakly in association with of latent CMV infection. In particular, it appears that CMV
receiving CMV seronegative RBCs (OR = 1.00–1.02, P = 0.05). is tropic for peripheral blood cells expressing either CD13
A positive odds ratio was also found for cohort 2 patients antigen105 or CD14 antigen.106
who were transfused with process leukoreduced apheresis Using reverse transcriptase PCR, Taylor-Wiedeman and
platelets from CMV seropositive donors, but the association colleagues107 demonstrated that monocytes from latently
was also rather weak (OR = 1.00–1.06, P = 0.05). However, infected, healthy, seropositive individuals failed to transcribe
in multivariate analysis of the same cohort 2 data, the above- CMV genes. However, when the cells were cultured in vitro
mentioned associations with higher odds of CMV infection after exposure either to granulocyte-monocyte colony-
were no longer significant, but a third factor was associated stimulating factor plus hydrocortisone or to phorbol 12-
with higher odds of CMV infection—namely, transfusion myristate 13-acetate plus hydrocortisone, they differentiated
with leukofiltered RBCs from a CMV-positive donor. This into macrophages and expressed messenger RNA for the CMV
latter finding is unexpected because the principal change in early-intermediate gene. However, late CMV gene transcripts
BLOOD BANKING
were not produced, and the cells failed to shed complete to clinical isolates of CMV prepared from infected patients.
virus, suggesting an arrest in productive viral transcription On the other hand, Krajden and colleagues116 found that
at the early gene phase. the method of DNA extraction affected the performance of
CMV DNA tests. Among 101 samples obtained from CMV-
Few Mononuclear Cells in Healthy seropositive blood donors, the frequency of a positive PCR
Donors Harbor CMV result was 0%, 1%, or 8% depending on which of three
Among healthy blood donors, the proportion of mononu- different kits were used to isolate DNA for testing.
clear cells infected with CMV bears directly on the likelihood
Infection with Multiple CMV Strains Has Not
that 3- to 4-log leukoreduction will prevent CMV trans-
Been Shown to Result from Transfusion
mission by transfusion. Earlier studies reported that CMV
immediate-early gene transcripts were present in 0.03% to Molecular typing methods have demonstrated that some
2% of peripheral blood mononuclear cells from healthy patients are infected with more than one strain of CMV.
seropositive individuals.108 Subsequently, Slobedman and Using restriction enzyme analysis, Chou117 compared the
Mocarski109 analyzed the proportion of infected cells by PCR patterns of isolates among 36 pairs of recipients of CMV-
and in situ hybridization and by quantitative competitive seropositive renal allografts. Although material from the
PCR. Using normal donors undergoing granulocyte colony- kidney donors was not analyzed, they found paired recipi-
stimulating factor mobilization of hematopoietic progeni- ents who shared the same kidney donor also demonstrated a
tors, they found that 0.004% to 0.01% of mononuclear common CMV strain that was not present in the recipients
cells contained viral genomes at a copy number of 2 to 13 before the transplant. Follow-up studies confirmed that solid
genomes per infected cell. Among healthy blood donors who organ transplants were capable of infecting recipients with a
are not undergoing growth factor–mediated mobilization, second strain of CMV.118 Multiple-strain infection was also
the proportion of infected cells presumably would be lower. reported by Chandler and colleagues,119 who found molecu-
These findings generate a plausible explanation for the abil- lar evidence for multiple-strain infection among four of eight
ity of leukoreduction to reduce the transmission of CMV. If women attending a clinic for sexually transmitted diseases.
a unit of blood contains approximately 107 monocytes and Spector and associates120 reported multiple-strain infection
if 1 in 10,000 to 1 in 100,000 monocytes are infected, then in two persons with serologic evidence of infection with the
the unit contains 103 to 102 latently infected cells. Therefore, human immunodeficiency virus (HIV) who were diagnosed
a 3- to 4-log leukoreduction may be expected to render the with acquired immunodeficiency syndrome. However, the
unit noninfectious. However, in vitro experiments in which interpretation of studies of second-strain infection has been
units were spiked with CMV-infected fibroblasts failed to confounded by the fact that strain mutations develop under
show complete clearance of CMV transcripts when the leu- the selective pressure of antiviral therapy.121 These mutations
kofiltered blood was tested by PCR.110 On the other hand, lead to a different pattern of restriction digest, which can be
it is unlikely that clinical infections in humans result from misinterpreted as a second independent strain.
II exposure to a single latently infected cell, even though the Recognition of second-strain infection via organ trans-
threshold level required to acquire infection is not known. plantation or multiple sexual contacts raises the possibility
370 For example, quantitative studies of the level of viremia in of transfusion-transmitted second-strain infection and the
patients after liver transplantation have documented that question of whether CMV-seropositive transfusion recipi-
levels greater than 104 genomes per milliliter are required for ents should receive CMV–reduced risk blood components.
infections to become symptomatic.111 The risk of second-strain infection by transfusion remains
only theoretical to date, because no such cases have been
Seronegative Healthy Subjects May Test reported. Using restriction endonuclease analysis, Winston
Positive for CMV DNA and coworkers122 observed no second-strain infections in a
Although serologic testing has proved of great practical study of 18 allogeneic bone marrow transplant recipients
value for reducing transfusion-related CMV, some investiga- who developed CMV during the course of their treatment.
tors have reported that a minority of serologically negative
Reactivation of Latent CMV by Transfusion
samples may be positive for CMV DNA when tested by PCR.
of Allogeneic Donor Leukocytes
CMV-seronegative, PCR-positive blood donors could account
for the finding that patients supported with CMV-seronega- Decades ago, the hypothesis was put forward that alloge-
tive units experience a 1% to 4% rate of CMV transmission, neic transfusion would result in an in vivo mixed lympho-
as measured by CMV seroconversion, viremia, or viruria.112 cyte reaction and that this immunologic stimulus might
However, the actual incidence of CMV DNA-positive serone- result in reactivation of latent CMV infection.123 The
gative donors remains uncertain. In the largest study to date, hypothesis that allogeneic donor leukocytes would result
Roback and colleagues113 used two previously validated PCR in viral reactivation in the recipient following transfusion
assays and found that none of 514 CMV seronegative donor was definitively tested in the Viral Activation Transfusion
samples were DNA positive. This result is in sharp contrast Study (VATS).124 The study was a multicenter, random-
to Bevan and associates114 who studied 312 CMV-serone- ized, double-blind trial comparing the effects of allogeneic
gative samples and found that 25% were positive for CMV blood with and without leukoreduction on the outcomes of
DNA. There is no clear explanation for these different results 531 patients already infected with HIV-1 and CMV. A total
although numerous methodological details sharply affect the of 3864 units of RBCs were transfused. The primary study
performance of the tests. For example, Rahbar and co-work- outcome measures were mortality and plasma levels of HIV
ers115 reported that CMV-seronegative blood donors lacked (RNA assay) measured 7 days after transfusion. The results
antibodies to the laboratory strain AD 169—a strain com- showed no survival advantage with the use of leukoreduced
monly used as the antigen target in serologic assays, but that blood, with 151 deaths in the leukoreduced group and 138
36% of these CMV “seronegative” individuals had antibodies deaths in the control group. Median survival was 13 months
LEUKOCYTE-REDUCED PRODUCTS
(leukoreduced) versus 20.5 months (control), P = NS. When appear more likely to provoke alloimmunization than
adjusted for baseline prognostic factors (CD4 count and immunosuppression. Immune tolerance may depend on
plasma HIV RNA level) survival was statistically worse in partial histocompatibility matching of donor and recipient.
the leukoreduced group (relative hazard for death 1.35 com- However, the details of the MHC relationship required for
pared with nonleukoreduced, 95% CI, 1.06–1.72). The use of the induction of tolerance are poorly defined and appear to
leukoreduced blood did not delay the time to onset of new vary among species and with the experimental conditions.
opportunistic infections and had no effect on the frequency In human studies, HLA class II antigen matching appears
of transfusion reactions. No changes were seen in plasma to be particularly relevant. For example, van Twuyver
HIV RNA levels after transfusion. In addition, there was no and associates128 studied 23 untransfused, first-time renal
advantage to leukoreduction for a variety of secondary out- allograft recipients, each of whom was deliberately trans-
come measures including CMV DNA levels, CD4 cell counts, fused before transplantation with donor fresh blood con-
cell activation assays, or plasma cytokine levels. This study taining approximately 7 × 108 leukocytes. After transfusion,
provided substantial evidence against the hypothesis that cytotoxic T-cell precursors directed against donor antigen
donor leukocytes cause clinical viral reactivation. Because targets were measured. Ten patients demonstrated a signifi-
HIV-infected individuals are at high risk for immune break- cant decline in the level of antidonor T-cell response at 1
down, the failure to observe increased infections in the con- month after transfusion. Nine of them were found to share
trol group relative to the leukoreduced group also cast doubt one HLA haplotype (HLA-B and HLA-DR match) with
on the role for leukoreduction as a means to abrogate any their donor. In contrast, those patients transfused with mis-
proposed immunosuppressive effect of transfusion. In sum- matched blood maintained strong anti–T cell responsiveness
mary, the available clinical evidence fails to support a role for after transfusion.
leukoreduction in the prevention of viral reactivation. Third, the dose and molecular presentation of trans-
fused antigens may affect whether the recipient response is
Donor Leukocytes and Immunosuppression directed toward alloimmunization or tolerance. Evidence for
the induction of tolerance by infusion of large intravenous
Allogeneic blood transfusion has been suggested to induce a doses of antigen (high-zone tolerance) has existed for years
mild state of immunosuppression. The original suggestion in experimental immunology. The molecular presentation of
for this effect arose from the observation in 1973 by Opelz donor antigen may also play a critical role in the immune
and associates125 that allogeneic transfusions given before response of the transfusion recipient, and there is evidence
renal transplantation resulted in improved allograft survival. supporting a role for both cell-associated antigen and soluble
Although the allograft effect was confirmed by numerous antigen as agents for the induction of experimental tolerance.
subsequent reports, including studies conducted in the era Experimental work from several laboratories has demon-
of modern antirejection therapy,126 the exact mechanism strated that soluble HLA antigens can downregulate in vitro
has never been conclusively explained. In 1981 Gantt127 immune responses by several different mechanisms.129 For
questioned whether transfusions might also down-regulate example, peptides from the nonpolymorphic 3 region of the 26
host antitumor immunity, thereby resulting in either an class I HLA molecule were able to inhibit the differentiation
increased rate of tumor relapse or a shorter interval from of cytotoxic T cells in response to an alloantigen stimulus,130 371
primary tumor resection to time of relapse. Subsequently, and soluble HLA peptides can inhibit NK cells.130 Soluble
the hypothesis of transfusion-related immunosuppression HLA antigen or other factors released from leukocytes form
was extended to include concern that transfusion might the basis for the hypothesis that prestorage leukoreduction
increase the frequency of postoperative bacterial infec- would reduce an immunosuppressive effect of transfusion.
tions. A detailed account of the possible immunosuppres- However, Dzik and colleagues131 found no evidence that
sive effects of transfusion was published by Vamvakas and the measured soluble HLA antigens differ in units of RBCs
Blajchman.2 stored with or without prestorage leukoreduction. In con-
trast, Ghio and associates132 measured increased concentra-
Common Themes in Experimental Studies of tions of soluble class I HLA antigens and soluble Fas ligand
Transplant Tolerance and Transfusion in units of RBCs and platelets during storage. They found
Four themes regularly appear in experimental studies of that concentrations of soluble antigens and Fas ligand were
transfusion and immune tolerance to organ transplantation: proportionate to the number of residual donor leukocytes in
recipient immune conditioning, the histocompatibility rela- the blood components.
tionship between donor and recipient, the presentation of Finally, the persistence of donor antigen in the recipient is
antigen, and the persistence of donor antigen. First, experi- a common finding in experimental models for the induc-
mental induction of transplant tolerance appears to require tion of transplant tolerance.133 Studies have focused on
proper “conditioning” of the recipient. Conditioning regimens “microchimerism”—the persistence of low levels of donor
usually consist of mild immunosuppression with anti–T cell cells in the graft recipient. There appears to be little doubt
globulins, chemotherapy, or corticosteroids. Presumably that microchimerism accompanies tolerance. However, it is
these agents weaken the immune response in such a way as uncertain whether the persistence of donor cells contributes
to prevent the transfused cells or proteins from provoking to the induction of tolerance or is merely a consequence of
alloimmunization and to allow the transfused cells or pro- tolerance. Microchimerism after clinical solid organ trans-
teins to persist within the recipient long enough to induce plantation is unquestionably present and has been verified
tolerance. by numerous assays. Microchimerism with fetal cells may
Second, antigen-specific tolerance induction appears to also exist in women after normal pregnancy and delivery.134
depend on the proper relationship between the MHC anti- Whether microchimerism develops after blood transfusion
gens of the donor and those of the recipient. Transfusions is controversial. Lee and colleagues135 reported evidence for
from donors who are highly mismatched at MHC loci extended microchimerism after blood transfusion. Using
BLOOD BANKING
a PCR assay directed against donor-type HLA genes, they INDUCTION OF SUPPRESSOR CELLS
detected donor signal in one recipient for up to 1.5 years after
One important hypothesis for the immunosuppressive effect
transfusion. Of note, the transfusions occurred in the setting
of blood transfusion suggests that transfusion may drive the
of trauma and included infusions of fresh blood. In follow-
recipient toward T-cell suppression. In a series of experi-
up studies, Lee and colleagues reported that recipients of leu-
ments in mice, Chen and associates140 showed evidence for
koreduced blood also demonstrated microchimerism.136
tolerance induced by a suppressor T cell found in the spleen.
Immunologic Mechanisms Suggested for BALB/c mice (H-2d) were donors of cardiac allografts to
Transfusion-related Immunosuppression MHC-disparate CBA/Ca (H-2k) recipients. In the absence of
any conditioning, the recipients promptly rejected the cardiac
At various times, one or more of the following four immu-
allografts. Transplant tolerance was induced by pretransplan-
nologic mechanisms has been suggested as mechanisms for
tation blood transfusion from the donor after conditioning
a proposed immunosuppressive effect of blood transfusion:
with nonlytic antilymphocyte globulin. Under these condi-
clonal deletion, anergy, active suppression, and indirect effect
tions, the animals retained the cardiac allografts for more
via viral transmission.
than 100 days. Splenic lymphocytes taken from these toler-
CLONAL DELETION ant animals were then transferred to naive CBA/Ca animals,
which were then tolerant of BALB/c heart allografts in the
Central thymic tolerance to self-antigens develops when T absence of any immunosuppression. In fact, tolerance could
cells are deleted during thymic education. Thymic toler- be transferred successfully through nine passages of splenic
ance to foreign antigens can be induced by the inoculation lymphocytes into new animals. The suppressor cell was
of MHC peptide into the recipient thymus. These peptides localized to the CD4+ subset of T cells. Studies, such as this
are presumably taken up by recipient APCs and displayed one, that demonstrate adoptive transfer of tolerance argue
as if they were self-antigens. For example, Chowdhury and strongly in favor of a suppressor cell mechanism. Similar
colleagues137 were able to induce prolonged cardiac allograft findings were reported by Yang and colleagues,141 who sug-
acceptance in a rat model by intrathymic injection of syn- gested than the CD45RC+ subset of CD4+ T cells accounted
thetic class I allopeptides. Whether central tolerance is rel- for the immunosuppression.
evant for blood transfusion is unknown. It is conceivable More direct experimental evidence that allogeneic blood
that donor HLA peptide present in blood becomes localized transfusion can induce suppressor T cells in recipients
in the recipient thymus gland. Because the thymus involutes comes from the studies of Blajchman and colleagues,142,143
with age, the role of central thymic tolerance may be more who examined the potential of transfusions to promote
relevant for pediatric transfusions than for those given to tumor growth in animals. In one series of experiments,
adults. animals were first transfused with either unmodified allo-
T-cell clones may also undergo deletion outside the envi- geneic blood, leukoreduced allogeneic blood, or syngeneic
ronment of the thymus gland. Activation-induced apoptosis blood (as a control). After the transfusions, the animals were
II occurs when T cells respond to antigen via their T-cell recep- challenged with an injection of tumor cells. After a waiting
tor but are not sustained by costimulatory molecules or by period, the animals were sacrificed and the number of pul-
372 proper cytokine support. Normally, for any given environ- monary metastases were counted. The investigators found
mental antigen, only approximately 1 in 105 to 1 in 106 T that transfusion with unmodified blood promoted increased
cells have a matching T-cell receptor and are able to respond numbers of pulmonary metastases, compared with leukore-
to the antigen. However, in the setting of an allogeneic HLA duced or syngeneic blood. Moreover, spleen cells transferred
stimulus, such as occurs with organ transplantation or blood to naive animals from animals that had received unmodified
transfusion, approximately 1% to 10% of T cells are able to allogeneic blood transfusions promoted greater numbers of
respond. This remarkable finding is the basis of the strong pulmonary metastases than did spleen cells transferred from
blastogenic response of the routine mixed lymphocyte cul- animals that had not been transfused or had received trans-
ture. It is possible that this large activation signal leads to fusion of leukoreduced blood. These findings argue for the
considerable activation-induced apoptosis of recipient cells induction of suppressor spleen cells in the animals who had
and subsequent peripheral clonal deletion. received unmodified donor blood.
Further evidence for splenic suppressor cells induced
DEVELOPMENT OF CELLULAR ANERGY
by blood transfusion was presented by Kao.144 He was able
Multiple lines of experimental evidence demonstrate that to induce humoral immune nonresponsiveness in CBA
immune activation of T cells depends on the cell’s receiv- mice (H-2k) using transfusions of ultraviolet B–irradiated
ing not only a primary signal (the antigen signal) but also a leukocytes from BALB/c (H-2d) donor mice. Ultraviolet
secondary signal known as a costimulatory signal. Certain B irradiation is known to interfere with the expression of
costimulatory molecules send an activation signal, whereas costimulatory molecules (see earlier discussion). When
others send an anergy signal. One hypothesis concerning spleen cells from the tolerant animals were transferred to
the immunomodulatory effect of blood transfusion suggests naive CBA recipients, the recipients also became tolerant
that during refrigerated blood storage donor leukocytes to BALB/c donor antigens. Presumably, the transferred cells
undergo alterations that result in anergy signals to recipient suppressed the ability of CBA recipients to form a humoral
T cells. There is little direct experimental evidence to support immune response.
this contention, although Minchef and coworkers,138 using a In humans, direct evidence for the induction of Th2-type
rodent transfusion model, reported that refrigerated storage suppressor cells by blood transfusion is lacking. However,
resulted in both necrosis and apoptosis of donor cells. Other Kirkley and associates145 reported in vitro cytokine release
experimental evidence has also suggested that apoptotic cells in 43 patients transfused with either allogeneic or autolo-
are directly immunosuppressive.139 gous blood at the time of hip surgery. Mean levels of IL-10
LEUKOCYTE-REDUCED PRODUCTS
and IL-4 released in vitro were slightly higher among untransfused patients. However, as with cancer recurrence,
recipients of allogeneic blood, suggesting polarization by such studies may have only identified that transfusion was a
allogeneic transfusion toward a Th2 phenotype. However, marker of more serious illness.
other studies of transfused patients have not found statisti- Because earlier studies suggested that patients undergo-
cally significant differences in cytokine profiles as a result of ing colorectal surgery were more likely to benefit from leu-
transfusion.146,147 koreduction technology, Titlestad and colleagues performed
a prospective randomized trial in 279 such patients who were
Clinical Studies of Immunosuppression randomized prior to transfusion to receive either leukore-
and Leukocyte Depletion duced or nonleukoreduced RBCs.149 Because only a minor-
Whether blood transfusion results in a clinically measur- ity of patients required transfusion, the number of transfused
able increase in tumor recurrence or postoperative bacte- patients receiving leukoreduced RBCs (N = 48) or nonleuko-
rial infection has never been adequately resolved. The issue reduced RBCs (N = 64) was modest. The authors found that
is difficult to address experimentally in human subjects. A the rate of postoperative infection was not different in the two
large number of observational studies have documented that groups. In addition, there was no effect of leukoreduction on
patients who receive transfusions are more likely than their mortality or hospital length of stay.
untransfused counterparts to develop tumor recurrence or Baron and colleagues150 recorded infection rates following
bacterial infection. This observation, however, represents a abdominal aortic surgery in a cohort of patients (N = 192)
correlation and does not imply that the transfusions resulted transfused before implementation of leukoreduction and
in these adverse effects. Rather, blood transfusion may compared them with infection rates of a subsequent cohort
simply be a marker that correlates with severity of disease. of patients (N = 195) transfused after implementation of leu-
Moreover, the finding that transfusion correlates with tumor koreduction. The two cohorts showed no difference in rates of
recurrence or bacterial infection does not imply that leuko- postoperative infection (leukoreduced 27% [95% confidence
reduction would have a beneficial effect on the frequency of of 21–33%] versus control 31% [95% confidence of 25–38%]);
these complications. The effect of leukoreduction has been no difference in the proportion with severe infection or pneu-
addressed by randomized, prospective, blinded clinical trials monia; no differences in either total hospital length of stay or
in which patients are assigned to receive either leukoreduced intensive care stay, and no difference in overall mortality.
or nonleukoreduced blood components. In addition, trials Van Hilten and colleagues151 reported a large prospec-
that examine outcomes in cohorts of patients transfused tive randomized clinical study that examined infection rates
before versus after implementation of leukoreduction have among patients undergoing aortic aneurysm or gastrointes-
been reported. Summarized below and in Table 26–2 are tinal cancer surgery. Among 1200 patients randomized prior
large clinical studies published after the year 2000. to transfusion, 545 were transfused. Among those transfused,
the two groups were balanced between those randomized to
Transfusion and Cancer Recurrence receive leukofiltered products (N = 237) or nonfiltered prod-
It was proposed in the 1990s that transfusion would increase ucts (N = 257). Among transfused patients, there was no 26
the rate of cancer recurrence. This hypothesis was supported statistical difference between leukoreduced versus nonleuko-
by preclinical animal experiments that involved deliberate reduced groups for postoperative infection. There was also no 373
inoculation of tumor cells142,143 and by retrospective reports effect of leukoreduction on the incidence of multiorgan failure
in humans showing an association between transfusion at the or death. Median length of stay in intensive care was lower in
time of cancer surgery and the subsequent development of the patients assigned to receive leukoreduced blood, but mean
recurrent cancer. However, well-designed clinical trials have overall hospital length of stay was not different between the
not supported the hypothesis that leukoreduction would two groups.
affect tumor biology. For example, van de Watering and In the largest prospective randomized trial, Dzik and co-
colleagues provided results on long-term cancer recurrence workers27 randomly assigned all transfused patients who did
among 697 patients with colorectal cancer who were ran- not meet standard criteria for leukoreduced blood to receive
domized to receive either leukoreduced or nonleukoreduced either leukofiltered components (N = 1355 patients) or non-
(buffy coat–depleted) blood components.148 They found that leukoreduced components (N = 1425 patients). Patients
the 5-year survival rates (65% leukoreduced; 64% control) included both surgical and medical patients with no exclu-
and 5-year cancer recurrence rates (28% leukoreduced; 28% sions. Any infection sufficiently severe to cause treating phy-
control) were nearly identical in the two groups. The study sicians to begin antibiotic therapy was recorded. The results
provided randomized controlled trial evidence that leuko- showed no difference between the two groups for the propor-
cyte reduction at the time of cancer surgery did not affect the tion of patients who were treated with antibiotics after trans-
rate of cancer recurrence. fusion (leukoreduced = 66%; control 68%, P = 0.42) and
showed no difference for the duration of antibiotic therapy.
Transfusion and Postoperative Infection There was also no difference when antibiotic use was ana-
lyzed for the postoperative period among surgical patients.
NONCARDIAC SURGERY
Subgroup analysis also found no effect of leukoreduction on
The hypothesis that recipient exposure to donor leukocytes the proportion of patients receiving antibiotics (or the dura-
induced some form of immunosuppression was also exten- tion of antibiotic therapy) among specific patient groups,
sively investigated as a potential cause for post-transfusion including colorectal surgery (N = 110), noncardiac surgery
infection in surgical patients. Initially, retrospective stud- (N = 1087), and nonsurgical patients (N = 1077). There was
ies found an association between higher rates of postop- also no effect of leukoreduction on mortality rates or various
erative infection among transfused patients compared with measures of hospital length of stay.
II
BLOOD BANKING
374
Table 26–2 Clinical Trials of Leukoreduction (LR) Published after the Year 2000 That Investigate Clinical Evidence
for Immunosuppression and Infection
Author Year Design Patient Population Non-LR (N)* LR (N)* Infection Mortality Length of Stay (LOS)
*
Number of patients shows transfused patients. Several studies randomized more patients not all of whom were transfused.
†
LR group had lower in-hospital rate of nonserious infections, but LR group had higher rate of infections postdischarge.
‡
Infection rates declined (p < 0.05) when switched from LR to non-LR.
§
Unadjusted odds ratio for mortality significantly lower for LR group, but significance not found when odds ratio adjusted for effect of cardiac medications.
The prospective trial of Dzik and colleagues27 included all
LEUKOCYTE-REDUCED PRODUCTS
CARDIAC SURGERY patients undergoing cardiac surgery during the study period.
Patients (N = 246) received transfusion with leukoreduced
Interest in a possible beneficial effect of leukocyte reduc- blood compared with patients (N = 260) who received non-
tion more specific for cardiac surgery patients originated leukoreduced components. Infection rates, measured by the
from the early report of van de Watering and colleagues.152 clinical decision to administer antibiotic therapy, were not
They randomly assigned more than 900 patients undergo- different in the two groups.
ing cardiac surgery to three groups. Among those transfused, In addition to the above randomized trials, four studies
patients received either prestorage leukofiltered RBCs (N in cardiac surgery patients examined outcomes among
= 283), poststorage leukofiltered RBCs (N = 280), or buffy a cohort of patients transfused before implementation
coat–depleted but not extensively leukoreduced RBCs (N = of leukoreduction compared with outcomes in a subse-
303). The primary outcome of the study was infection com- quent cohort of patients transfused after implementation
plications, and the authors found no effect of leukoreduc- of leukoreduction.
tion on the frequency of infection outcomes (P = 0.15). An Volkova and associates156 reported a study based on three
unexpected finding was a lower 60-day mortality (3.5%) in sequential cohorts of patients: those receiving nonleukore-
either of the leukofiltered arms compared with the nonleu- duced blood (N = 416), then a cohort receiving transfusion
koreduced blood arm (7.9%); P = 0.025. Although van de following implementation of leukocyte-reduced blood (N =
Watering found no evidence that transfusion increased the 484), and finally a cohort transfused following return to the
rate of postoperative infection, the authors noted that the use of nonleukocyte-reduced blood (N = 317). They noted
difference in mortality was only apparent among patients that mean hospital length of stay decreased progressively
receiving more than 3 units of RBCs and appeared to be due with each subsequent time interval (15.9 days, 14.1 days,
to different rates of wound dehiscence or multiorgan failure. and 12.1 days). Thus, the implementation of leukoreduc-
Three prospective randomized trials published since 2000 tion in 1992 was associated with a reduced length of stay,
examined infection rates and mortality among cardiac sur- but the discontinuation of leukoreduction and a return to
gery patients. Wallis and associates153 reported results among nonleukoreduced blood was also associated with a further
510 transfused patients assigned to receive control RBCs (N decreased length of stay. This finding suggests that hospital
= 163); buffy coat–depleted RBCs (N = 173), or leukofiltered length of stay was simply decreasing independently of blood
RBCs (N = 174). Although patients were randomized prior policy and demonstrates that “before-versus-after” study
to transfusion, significantly fewer patients assigned to receive designs can produce misleading conclusions. Observed rates
control RBCs were transfused. Among transfused patients, of infection were not reduced with implementation of leu-
they found that use of control RBCs was associated with more kocyte reduction but declined in the third cohort after the
events coded as infection (P = 0.02). However, when events switch from leukoreduced to nonleukoreduced blood.
coded as urinary tract infection were excluded, there was In the study by Fung and colleagues157 501 patients
no significant difference among the three groups; P = 0.25. received unfiltered blood products.158 The following year, 26
Following discharge from hospital, a significantly higher rate 645 patients received leukofiltered products. No significant
of infection requiring antibiotic treatment was found among changes were seen in the rate of mediastinitis, operative 375
patients assigned to receive leukoreduced blood (47%) com- mortality, or stay in intensive care. There was a statistically
pared with control RBCs (34%); P = 0.01. The authors con- significant decrease in postoperative hospital length of stay.
cluded that no evidence was found that leukocyte reduction This study also did not identify an immunosuppressive effect
reduced postsurgical infection rates. of donor leukocytes as measured by infection rates.
Bilgin and colleagues154 conducted a second clinical trial Llewelyn and colleagues159 reported outcomes among
in cardiac surgery focusing on patients undergoing valve- cardiac surgery and orthopedic surgery patients in 11 hos-
replacement surgery (with or without bypass grafting). pitals before implementation of leukoreduction (N = 997)
Patients were randomly assigned to receive either buffy coat– and compared these to outcomes following implementation
depleted RBCs (N = 216) or leukofiltered RBCs (N = 216). (N = 1098). The proportion of all transfused patients with
Multiple outcomes were compared with no statistical correc- suspected or proved infection was unchanged before and
tion for multiple comparisons. Among transfused patients, after implementation of leukoreduction (odds ratio 0.83;
infection rates were higher in patients assigned to buffy coat– 95% CI, 0.77–1.02). Among cardiac patients, use of leukore-
depleted blood (mean odds ratio = 1.64; 95% CI, 1.08–2.49). duced blood was statistically associated with a higher rate of
However, in-hospital mortality, frequency of multiorgan fail- proved infections (P = 0.004). Overall postoperative length
ure, length of stay in intensive care, overall length of stay in of stay was also not affected by leukoreduction. Subgroup
hospital, or 90-day mortality were all not significantly differ- analysis showed no effect related to dose of blood transfused.
ent between the two groups. Mortality rates among cardiac surgery patients were increased
Bracey and co-workers conducted a prospective ran- following implementation of leukocyte reduction (P = 0.031)
domized trial in cardiac surgery patients at the Texas Heart but were decreased among orthopedic patients. The authors
Institute.155 Patients were randomized prior to transfusion. concluded that leukoreduction did not improve outcomes.
Transfused patients assigned to receive control blood compo- The large Canadian study that examined outcomes
nents (N = 159) or leukoreduced RBCs (N = 136) were bal- before versus after implementation of leukoreduction
anced for preoperative risk factors, surgical procedures, and included cardiac patients.26 Patients in cohort 1 (N = 4475)
dose of transfusion. The authors found no difference in the were transfused with nonleukoreduced blood, and those in
two groups for in-hospital infections or infections during the cohort 2 (N = 5050) received leukoreduced blood. There
first 3 months after discharge. There was also no difference in was no effect of leukoreduction on the adjusted odds of
the two groups for mortality, intensive care length of stay, or death and no observed effect of leukoreduction on infec-
overall hospital length of stay. tious complications.
BLOOD BANKING
Taken together, the findings of clinical studies in surgery lyzed the influence of ambient temperature at which filters
patients since 2000 fail to demonstrate that leukocyte reduc- are used on the degree of hemolysis. The results supported
tion has any significant or consistent effect on the incidence a notion that when the leukoreduction is performed at
of postoperative infection. Whether allogeneic transfu- 4° C in a strictly controlled environment, hemolysis can be
sion by itself (with or without donor leukocytes) causes an avoided. Janatpour and co-workers158 studied three differ-
immunosuppression not seen in untransfused individuals ent methods for determination of hemolysis in segments
is not resolved by these studies. However, the finding that and the unit. They split the units into two aliquots, one
leukoreduction fails to change infection outcomes argues of which was leukoreduced, and measured plasma hemo-
against a role for donor leukocytes in any proposed immu- globin using (1) a HemoCue Plasma/Low Hb Photometer
nosuppressive effect on bacterial defenses. The studies pro- system, (2) a tetramethyl-benzidine (TMB) chemical
vide conflicting results regarding overall hospital length of method, and (3) a free Hb visual comparator. They con-
stay among cardiac patients assigned to receive leukoreduced cluded that visual criteria overestimate the degree of
blood components, and the extent to which any findings hemolysis. Other methods could be more objective, raising
can be attributed to leukoreduction (especially in before- the question whether a visual inspection of hemolysis can
versus-after study designs) is difficult to determine. The early be used as the part of the quality control of the unit prior
observation by van de Watering of decreased mortality in a to distribution.
subgroup of cardiac valve patients receiving leukoreduced Although there were no reported clinical side effects of the
blood compared with nonleukoreduced blood was not sub- increased hemolysis postfiltration, this scenario underscores
stantiated in any subsequent study. In summary, the clinical the need for postmarketing surveillance of WBC filters and
trials data since the year 2000 suggest that transfusion-related the quality of components. It remains to be seen if increased
immunosuppression, if it does exist, may be mediated by hemolysis translates into clinically detectable side effects.
factors other than donor leukocytes. However, even without such adverse events, transfusion ser-
vices should remain vigilant in detecting potential problems
with any blood components. This proactive approach should
ADVERSE REACTION TO FILTRATION be able to identify manufacturing problems early on and
AND LEUKOCYTE REDUCED eliminate them in a timely manner.
COMPONENTS
Hypotensive Reactions to Bedside
Administration of leukocyte-reduced components is gener- Leukoreduction
ally safe with a few adverse effects. Now, when the process
of leukoreduction occurs mostly during, or shortly after, There have been several case reports of severe hypotension,
component collection, the profile of adverse reactions has occasionally accompanied by skin flushing and loss of con-
changed. When bedside leukoreduction was commonplace, sciousness, developing in patients who received bedside filtered
II the main observed adverse reaction was hypotensive epi- blood components.163–170 Although the reported incidence
sodes due to bradykinin production upon activation of the of these reactions is relatively low and is decreasing as fewer
376 contact phase. With the increased use of prestorage leuko- products are transfused using bedside filtration, some reac-
reduced components, this type of reaction is less likely to tions have been severe and illustrate an interesting phenome-
occur; however, the manufacturing process may now play a non. Four clinical features intersect to result in these reactions:
more important role as a cause of observed reactions. This transfusion of plasma-containing blood components; use of
section discusses three types of reported adverse reactions: a bedside leukoreduction filter; use of a filter whose medium
(1) hemolysis due to leukoreduction, (2) hypotensive reac- carries a net negative charge; and, most importantly, the con-
tions due to contact system activation, and (3) hypotensive current administration of angiotensin-converting enzyme
reactions to prestorage leukoreduced components. (ACE) inhibitors to the recipient. The sudden elaboration of
bradykinin was a prime suspect as the cause of these reactions
because bradykinin had previously been implicated in ana-
Hemolysis Due to Prestorage Filtration
phylactic reactions observed among patients receiving ACE
There has been a concern that prestorage leukoreduction inhibitors and undergoing low-density lipoprotein apheresis
may lead to an increased level of hemolysis in stored RBC or hemodialysis.171–173
components. First reports posted on transfusion medicine
Implication of Kinins and Activation
discussion group160 in 2001 described increased hemolysis
of the Contact Pathway in the Pathogenesis
in segments of leukoreduced units versus nonleukoreduced
of Hypotensive Reactions
units. The ensuing discussion did not identify a cause for
hemolysis but raised awareness of this potential manufac- Two related vasodilator peptides—the nonapeptide bradyki-
turing problem. However, no adverse clinical sequelae were nin and the decapeptide lysyl-bradykinin (kallidin)—exert
reported related to hemolysis. After this initial report the strong hypotensive effects in humans. Both peptides are
FDA continued to receive reports of hemolysis in RBC and metabolized by kininase I, a carboxypeptidase that removes
WBC components following leukocyte reduction using one amino acid (arginine) from the carboxyterminal end.
a high-efficiency filter. Because of this information, Pall Kininase II, also known as ACE, removes two amino acids
Medical decided to perform a voluntary market recall of (phenylalanine and arginine) from the carboxyterminal end,
their Leukotrap SC RC.161 Additional units were recalled rendering the peptides inactive. Bradykinin and lysyl-brady-
in February 2005. The hemolysis has been reported both kinin are formed from two precursor proteins: high-molecu-
immediately following filtration as well as 24 to 48 hours lar-weight kininogen, or HMWK (approximately 110,000 D),
after leukocyte removal. The cause of this phenomenon is and low-molecular-weight kininogen, or LMWK (approxi-
unclear at this time. Gyongyossy-Issa and associates162 ana- mately 68,000 D) (Fig. 26–2). The physiologic function of both
LEUKOCYTE-REDUCED PRODUCTS
Contact system
H-Arg-Pro-Pro-Gly-Phe-Ser-Pro-Phe-Arg-OH
(Bradykinin; BK; T1/2 30 sec)
H-Pro-Pro-Gly-Phe-Ser-Pro-Phe-OH H-Arg-Pro-Pro-Gly-Phe-OH
(BK2–8) (BK1–5)
B1 receptor activation
Figure 26–2 Simplified diagram representing synthesis and catabolism of bradykinin. The contact system upon activation through contact with a
negatively charged surface (e.g., glass, kaolin, dextran sulfate, blood filter, dialysis membrane) leads to activation of the intrinsic coagulation cascade
and to bradykinin (BK) generation. The active components of this pathway include kallikrein, bradykinin (acts through B2 receptor), and des-Arg9
bradykinin (acts through B1 receptor), all labeled in boldface. The other metabolites of bradykinin are inactive. The inhibition of the major pathway
of catabolism by ACE inhibitors results in prolongation of the half-life (T1/2) of active components and clinical signs and symptoms. ACE, angiotensin-
converting enzyme; APP,. aminopeptidase; HMW, high molecular weight; LMW, low molecular weight.
kinins is exerted through their action on two receptors, B1 bradykinin is significantly diminished among patients with
and B2. These receptors were cloned and identified as serpen- genetically low activity of pulmonary kininases and patients 26
tine receptors coupled to G proteins. Because bradykinin and receiving ACE inhibitor medications. In addition, patients
lysyl-bradykinin are primarily tissue hormones, their concen- undergoing cardiopulmonary bypass are at increased risk 377
tration in the circulation is usually low. Reported physiologic for bradykinin-mediated hypotensive reactions, because
levels vary from less than 3 to 55 pg/mL.174,175 Both peptides during the bypass venous blood flow is diverted around
are responsible for contraction of visceral smooth muscles, the pulmonary circulation. The largest series of severe
but they relax vascular smooth muscles through the action of hypotensive reactions accompanying bedside leukoreduc-
nitric oxide. In addition to the effect on smooth muscle cells, tion filtration was reported among patients undergoing
the kinins cause increased capillary permeability, accumula- cardiopulmonary bypass.165
tion of leukocytes, and pain after injection under the skin. The vasodilatory effect of bradykinin was studied in sev-
Bradykinin acting on the B2 receptor triggers nitric oxide for- eral experimental settings. Forearm blood flow increased
mation, which activates guanylate cyclase in smooth muscle after doses of 10 and 100 ng/min.180,181 Icatibant, a B2-kinin
cells, leading to increased concentrations of cyclic guanosine receptor antagonist, was used to demonstrate dose-depen-
monophosphate (cGMP) and smooth muscle relaxation.176 dent vasodilation in response to bradykinin. With increased
Previous studies have shown that bradykinin also activates dosage of icatibant, blood flow diminished significantly.
phospholipase A2, phospholipase C, protein kinases, and pros- N(G)-monomethyl-l-arginine, a specific inhibitor of nitric
taglandins, thereby resulting in the accumulation of cGMP oxide synthase, was used to demonstrate that bradykinin-
and cyclic adenosine monophosphate in cells.177 Although induced vasodilation is mediated in part by nitric oxide.180
bradykinin can activate tissue mast cells, leading to release of Bonner and colleagues182 reported on the hemodynamic
histamine,178 histamine release does not account for the full effects of bradykinin on the systemic and pulmonary
vasodilatory effects of bradykinin. Dachman and colleagues179 circulation in normotensive and hypertensive subjects.
demonstrated a residual vasodilatory effect even in the pres- Bradykinin was injected intravenously and intra-arteri-
ence of a histamine1 receptor antagonist (brompheniramine) ally at doses ranging from 42.4 to 6413 ng/kg. Bradykinin
and a histamine2 receptor antagonist (cimetidine). lowered blood pressure by decreasing systemic vascular
In healthy individuals, approximately 95% of injected resistance. ACE inhibitors potentiated this effect by approx-
bradykinin is metabolized during the first passage through imately 20- to 50-fold. Systolic blood pressure declined by
the lungs. Pulmonary kininases rapidly degrade bradyki- more than 20 mm Hg when the arterial bradykinin concen-
nin, thereby preventing its effect on the arterial circulation. tration reached at least 100 pg/mL. The study demonstrated
Indeed, these pulmonary kininases may normally provide that the physiologic effect of bradykinin is very rapid, with
transfusion recipients with protection against bradykinin- a hypotensive effect demonstrable within seconds after
induced vasodilation. However, pulmonary breakdown of administration.
BLOOD BANKING
Activation of Contact System filters. The interpretation of this study is difficult because,
During Passage of Blood through on a molar basis, kininogen is a very abundant molecule.
Leukoreduction Filters Therefore, conversion of a minor fraction of kininogen could
result in very significant levels of bradykinin.
Because bradykinin elaboration would occur if the contact
Why is it that not all patients react uniformly when given
system were activated, investigators have examined whether
bedside transfusions while receiving ACE inhibitors? Cyr
leukofiltration could induce contact activation of blood.
and colleagues187 studied the influence of ACE inhibitor
Studies have measured either an increase in bradykinin or
medication on the in vitro generation of bradykinin and its
its stable metabolite 1,5-bradykinin or a decrease in the
active metabolite des-Arg9-bradykinin. The in vitro half-life
substrates HMWK or LMWK. Although direct measure-
of bradykinin and of des-Arg9-bradykinin was measured
ment of bradykinin is more appealing, bradykinin is tech-
in the presence of an ACE inhibitor in serum from four
nically difficult to assay owing to ex vivo activation of the
patients who had experienced hypotensive reactions during
contact system. Shiba and coworkers183 measured bradyki-
blood transfusions of leukoreduced products. Although the
nin during leukofiltration of platelet concentrates, using a
half-life of bradykinin did not differ between the patients
negatively charged filter and a positively charged filter. After
and controls, the degradation of des-Arg9-bradykinin was
filtration through the negatively charged filter, decreased lev-
significantly slower in the patient samples (1549 vs 661 sec-
els of prekallikrein and increased levels of bradykinin were
onds). The authors proposed the hypothesis that an anoma-
observed. The bradykinin level was inversely related to the
lous metabolism of des-Arg9-bradykinin might contribute
activity of ACE in the platelet concentrates. The same group
to selection of patients who experience clinical reactions to
studied the effects of storage time, plasma dilution, and filtra-
bedside leukodepletion.
tion on contact-system activation in packed RBCs preserved
On the basis of the published studies, bradykinin pro-
in mannitol-adenine phosphate solution.184 The authors
duced during bedside filtration of platelets seems to be
noted an increase in the bradykinin level, up to 500 pg/mL
responsible for reactions during bedside leukofiltration in
on the 10th day of storage. The level decreased to 200 pg/mL
patients taking ACE inhibitor medications. Reactions to
during the subsequent 5 days and remained at this low level
RBC components may be less likely, because cold storage of
until the end of the storage time. A significantly decreased
RBCs inhibits contact activation enzymes and because RBCs
level of ACE activity was noted in the packed RBCs stored in
contain less plasma and kininogens. Concern regarding
solutions containing mannitol. Filtration using two different
hypotensive reactions among patients receiving blood com-
negatively charged filters generated bradykinin levels up to
ponents filtered at the bedside prompted the FDA to issue a
6000 pg/mL. The authors concluded that mannitol may act
letter to physicians in May 1999. The FDA recommended use
as an ACE inhibitor, slowing down catabolism of bradyki-
of blood products leukoreduced at the time of collection or
nin and leading to its accumulation in stored RBCs. Hild and
during laboratory storage whenever available.
colleagues185 studied generation of bradykinin during leuko-
filtration of platelets using three different filters—negatively Hypotensive Reactions to Prestorage
II
charged, positively charged, and neutral. Only the negatively Leukoreduced Components
charged filter contributed significantly to bradykinin pro-
378 duction. The levels of bradykinin detected in the eluate var- Arnold and colleagues188 report two patients who developed
ied from less than 200 pg/mL to as much as 10,000 pg/mL in hypotensive reactions after receiving prestorage leukoreduced
samples collected after the processing of 50 and 100 mL of components. Both patients were taking ACE inhibitors. This
platelet concentrates. The final concentration in units ranged observation puts in question the assumption (as discussed
from 200 to 2500 pg/mL. Interestingly, bradykinin present above) that only bedside filters can activate the contact sys-
after filtration was rapidly metabolized; after 60 minutes of tem to a sufficient degree to cause bradykinin generation with
storage, the bradykinin level was below the limit of detection. subsequent development of hypotension. A hypothesis was
However, when an ACE inhibitor was added to the storage put forward that the ACE inhibitors present during the col-
bag, bradykinin levels remained elevated for as long as 90 lection of the autologous (or allogeneic) unit were responsi-
minutes.185 Significant differences in bradykinin production ble for bradykinin accumulation during leukodepletion. This
were observed among the donors, so that some components hypothesis would require significant contact phase activa-
generated measurable levels of bradykinin but others did tion with significant accumulation of bradykinin. Additional
not. The highest bradykinin levels (>20,000 pg/mL) were studies are needed to answer this question.
observed after leukofiltration of apheresis platelets. Scott and Similar reactions have been observed and reported in the
associates186 investigated bradykinin generation by negatively patients undergoing therapeutic plasmapheresis, even if the
charged filters by measuring substrates of the contact system replacement fluid consisted only of 5% albumin, though it
during leukoreduction of apheresis platelets. The number of was more common to observe such reaction with fresh fro-
WBCs before leukoreduction varied from 0.5 to 1000 cells/ zen plasma. Owen and Brecher189 performed a retrospective
μL. Two different leukocyte filters were used. Measurements study to identify all patients who developed “atypical reac-
of the cleavage products of HMWK and LMWK were used tions” while undergoing therapeutic plasma exchange over
as markers of contact-phase activation and the potential for the course of 12 years. The authors concluded that the use
bradykinin production. Although no significant changes of ACE inhibitors was associated with symptomatic hypo-
were detected in kininogen cleavage products, the assay tension during the procedure and recommended that ACE
could not exclude conversion of small amounts (<5%) of inhibitors be discontinued at least 24 hours prior to apher-
kininogen to kinin. The authors concluded that clinically esis procedure. Thus, it seems plausible that the patients
significant activation of the contact system did not occur as undergoing therapeutic plasmapheresis with albumin as
a result of leukofiltration. However, they did observe a tem- a replacement fluid are at increased risk for hypotension if
porary decrease in kininogen levels with one of the studied concomitantly on ACE inhibitors.189–192
REFERENCES
LEUKOCYTE-REDUCED PRODUCTS
27. Dzik WH, Anderson JK, O’Neill EM, Assmann SF, Kalish LA, Stowell
CP. A prospective, randomized clinical trial of universal WBC reduc-
1. Dzik S, Aubuchon J, Jeffries L, et al. Leukocyte reduction of blood tion. Transfusion 2002;42:1114–1122.
components: public policy and new technology. Transfus Med Rev 28. Mintz PD. Febrile reactions to platelet transfusions. Am J Clin Pathol
2000;14:34–52. 1991;95:609–612.
2. Vamvakas EC, Blajchman MA (eds). Immunomodulatory Effects of 29. Dzieczkowski JS, Barrett BB, Nester D, et al. Characterization of reac-
Blood Transfusion. Bethesda, Md., American Association of Blood tions after exclusive transfusion of white cell-reduced cellular blood
Banks Press, 1999. components. Transfusion 1995;35:10–25.
3. Miller JP, Aubuchon JP. Leukocyte-reduced and cytomegalovirus- 30. Sirchia G, Rebulla P, Parravicini A, Carnelli V, Gianotti GA, Bertolini
reduced risk blood components. In Mintz PD (ed). Practice Guide- F. Leukocyte depletion of red cell units at the bedside by transfusion
lines for Transfusion Therapy. Bethesda, Md., American Association of through a new filter. Transfusion 1987;27:402–405.
Blood Banks Press, 1998, pp 163–189. 31. Sirchia G, Wenz B, Rebulla P, Parravicini A, Carnelli V, Bertolini F.
4. Dzik S. Leukodepletion blood filters: filter design and mechanisms of Removal of white cells from red cells by transfusion through a new
leukocyte removal. Transfus Med Rev 1993;7:65–77. filter. Transfusion 1990;30:30–33.
5. Zou Y, Sun Q, Li A, et al. Efficiency of leukocyte removal by filters 32. Wenz B. Microaggregate blood filtration and the febrile transfusion
made of superfine glass fiber membranes. Vox Sang 1999;76:22–26. reaction: a comparative study. Transfusion 1983;23:95–98.
6. US Food and Drug Administration, Center for Biologics Evaluation 33. Heddle NM, Kelton JG. Febrile nonhemolytic transfusion reactions. In
and Research. Available at http://www.fda.gov/cber/gdlns/preleuk. Popovsky MA (ed). Transfusion Reactions. Bethesda, Md., American
htm. Accessed March 28, 2006. Association of Blood Banks Press, 1996, pp 45–80.
7. Steneker I. Leukocyte Depletion from Fresh Red Cell Concentrates 34. Brecher ME (ed). Technical Manual,15th ed. Bethesda, Md., American
by Fiber Filtration: Filtration Mechanisms. Amsterdam, VU University Association of Blood Bank Press, 2005.
Press, 1992. 35. Luheshi GN. Cytokines and fever. Mechanisms and sites of action. Ann
8. Ledent E, Berlin G. Factors influencing white cell removal from red cell NY Acad Sci 1998;856:83–89.
concentrates by filtration. Transfusion 1996;36:714–718. 36. Cao C, Matsumura K, Yamagata K, Watanabe Y. Cyclooxygenase-2 is
9. Nishimura T, Kuroda T, Mizoguchi Y, et al. Advanced methods for induced in brain blood vessels during fever evoked by peripheral or
leukocyte removal by blood filtration. In Brozovic B (ed). The Role of central administration of tumor necrosis factor. Brain Res Mol Brain
Leukocyte Depletion in Blood Transfusion Practice: Proceedings of the Res 1998;56:45–56.
International Workshop. Oxford, Blackwell Scientific, 1989, pp 35–40. 37. Ganong WF. Central regulation of visceral function. In Ganong WF.
10. Beaujean F, Segier JM, le Forestier C, Duedari N. Leukocyte depletion Review of Medical Physiology, 19th ed. Stamford, Conn., Appleton
of red cell concentrates by filtration: influence of blood product tem- and Lange, 1999, p 242.
perature. Vox Sang 1992;62:242–243. 38. Engel A, Kern WV, Murdter G, Kern P. Kinetics and correlation with
11. Sirchia G, Rebulla P, Sabbioneda L, Garcea F, Greppi N. Optimal con- body temperature of circulating interleukin-6, interleukin-8, tumor
ditions for white cell reduction in red cells by filtration at the patient’s necrosis factor alpha and interleukin-1 beta in patients with fever and
bedside. Transfusion 1996;36:322–327. neutropenia. Infection 1994;22:160–164.
12. Bowden RA, Slichter SJ, Sayers M, et al. A compar ison of filtered 39. Menitove JE, McElligott MC, Aster RH. Febrile transfusion reaction:
leukocyte-reduced and cytomegalovirus (CMV) seronegative blood what blood component should be given next? Vox Sang 1982;42:
products for the prevention of transfusion-associated CMV infection 318–321.
after marrow transplant. Blood 1995;86:3598–3603. 40. Chambers LA, Kruskall MS, Pacini DG, Donovan LM. Febrile reac-
13. Williamson LM, Wimperis JZ, Williamson P, et al. Bedside filtration tions after platelet transfusion: the effect of single versus multiple
of blood products in the prevention of HLA alloimmunization: a pro- donors. Transfusion 1990;30:219–221.
spective randomized study. Alloimmunisation Study Group. Blood 41. Heddle NM, Klama LN, Griffith L, Roberts R, Shukla G, Kelton JG.
1994;83:3028–3035. A prospective study to identify the risk factors associated with acute
26
14. van der Meer PF, Pietersz RN, Nelis JT, Hinloopen B, Dekker WJ, reactions to platelet and red cell transfusions. Transfusion 1993;33:
Reesink HW. Six filters for the removal of white cells from red cell con- 794–797. 379
centrates, evaluated at 4°C and/or at room temperature. Transfusion 42. de Rie MA, van der Plas-van Dalen CM, Engelfriet CP, von dem
1999;39:265–270. Borne AE. The serology of febrile transfusion reactions. Vox Sang
15. Schuetz AN, Hillyer KL, Roback JD, Hillyer CD. Leukoreduction filtra- 1985;49:126–134.
tion of blood with sickle cell trait. Transfus Med Rev 2004;18:168–176. 43. Decary F, Ferner P, Giavedoni L, et al. An investigation of nonhemo-
16. Stroncek DF, Byrne KM, Noguchi CT, Schechter AN, Leitman SF. Increas- lytic transfusion reactions. Vox Sang 1984;46:277–285.
ing hemoglobin oxygen saturation levels in sickle trait donor whole 44. Perkins HA, Payne R, Ferguson J, Wood M. Nonhemolytic febrile
blood prevents hemoglobin S polymerization and allows effective white transfusion reactions. Quantitative effects of blood components with
blood cell reduction by filtration. Transfusion 2004;44:1293–1299. emphasis on isoantigenic incompatibility of leukocytes. Vox Sang
17. Burgstaler EA. Blood component collection by apheresis. J Clin Apher- 1966;11:578–600.
esis 2006, in press. 45. Brubaker DB. Clinical significance of white cell antibodies in febrile
18. Silva MA (ed). Standards for Blood Banks and Transfusion Services, nonhemolytic transfusion reactions. Transfusion 1990;30:733–737.
23rd ed. Bethesda, Md., American Association of Blood Banks, 2005. 46. Thulstrup H. The influence of leukocyte and thrombocyte incompati-
19. Beckman N, Sher G, Masse M, et al. Review of the quality monitoring bility on nonhaemolytic transfusion reactions. I. A retrospective study.
methods used by countries using or implementing universal leukore- Vox Sang 1971;21:233–250.
duction. Transfus Med Rev 2004;18:25–35. 47. Arend WP, Joslin FG, Massoni RJ. Effects of immune complexes on
20. van der Meer PF, Gratama JW, van Delden CJ, et al. Comparison of production by human monocytes of interleukin 1 or an interleukin 1
five platforms for enumeration of residual leucocytes in leucoreduced inhibitor. J Immunol 1985;134:3868–3875.
blood components. Br J Haematol 2001;115:953–962. 48. Heddle NM, Klama L, Singer J, et al. The role of the plasma from platelet
21. Triulzi DJ, Meyer EM, Donnenberg AD. WBC subset analysis of WBC- concentrates in transfusion reactions. N Engl J Med 1994;331:625–628.
reduced platelet components. Transfusion 2000;40:771–780. 49. Abbas AK, Lichtman AN, Pober JS. Cytokines in Cellular and Molecu-
22. Roback JD, Bray RA, Hillyer CD. Longitudinal monitoring of WBC lar Immunology, 3rd ed. Philadelphia, WB Saunders, 1997.
subsets in packed RBC units after filtration: implications for transfu- 50. Muylle L, Joos M, Wouters E, De Bock R, Peetermans ME. Increased
sion transmission of infections. Transfusion 2000;40:500–506. tumor necrosis factor alpha (TNF alpha), interleukin 1, and interleu-
23. Kao KJ, Mickel M, Braine HG, et al. White cell reduction in platelet kin 6 (IL-6) levels in the plasma of stored platelet concentrates: rela-
concentrates and packed red cells by filtration: a multicenter clinical tionship between TNF alpha and IL-6 levels and febrile transfusion
trial. The Trap Study Group. Transfusion 1995;35:13–19. reactions. Transfusion 1993;33:195–199.
24. Ratko TA, Cummings JP, Oberman HA, et al. Evidence-based recom- 51. Stack G, Snyder EL. Cytokine generation in stored platelet concen-
mendations for the use of WBC-reduced cellular blood components. trates. Transfusion 1994;34:10–25.
Transfusion 2001;41:1310–1319. 52. Muylle L. Ex vivo production in blood components: relevant and irrel-
25. Vamvakas EC, Blajchman MA. Universal WBC reduction: the case for evant. Paper presented at the 21st International Symposium on Blood
and against. Transfusion 2001;41:691–712. Transfusion, Gronigen, 1996.
26. Hebert PC, Fergusson D, Blajchman MA, et al. Clinical outcomes fol- 53. Aye MT, Palmer DS, Giulivi A, Hashemi S. Effect of filtration of plate-
lowing institution of the Canadian universal leukoreduction program let concentrates on the accumulation of cytokines and platelet release
for red blood cell transfusions. JAMA 2003;289:1941–1949. factors during storage. Transfusion 1995;35:117–124.
BLOOD BANKING 54. Palmer DS, Aye MT, Dumont L, et al. Prevention of cytokine accumu- 79. The Trial to Reduce Alloimmunization to Platelets Study Group. Leu-
lation in platelets obtained with the COBE spectra apheresis system. kocyte reduction and ultraviolet B irradiation of platelets to prevent
Vox Sang 1998;75:115–123. alloimmunization and refractoriness to platelet transfusions. N Engl J
55. Fujihara M, Ikebuchi K, Wakamoto S, Sekiguchi S. Effects of filtration Med 1997;337:1861–1869.
and gamma radiation on the accumulation of RANTES and trans- 80. Fisher M, Chapman JR, Ting A, Morris PJ. Alloimmunisation to HLA
forming growth factor-beta 1 in apheresis platelet concentrates during antigens following transfusion with leucocyte-poor and purified plate-
storage. Transfusion 1999;39:498–505. let suspensions. Vox Sang 1985;49:331–335.
56. Muller-Steinhardt M, Kirchner H, Kluter H. Impact of storage at 22°C 81. Petranyi GG, Padanyi A, Horuzsko A, Rethy M, Gyodi E, Perner F.
and citrate anticoagulation on the cytokine secretion of mononuclear Mixed lymphocyte culture: evidence that pretransplant transfu-
leukocytes. Vox Sang 1998;75:12–17. sion with platelets induces FcR and blocking antibody production
57. Heddle NM, Tan M, Klama L, Schroeder J. Factors affecting cyto- similar to that induced by leukocyte transfusion. Transplantation
kine production in platelet concentrates [abstract]. Transfusion 1988;45:823–824.
1994;34(Suppl):67S. 82. Kao KJ, Luz de Rosario M. Platelet alloimmunization. In Anderson KC,
58. Grey D, Erber WN, Saunders KM, Lown JA. Monocyte activation in Ness P (eds). Scientific Basis of Transfusion Medicine: Implications for
platelet concentrates. Vox Sang 1998;75:110–114. Clinical Practice. Philadelphia, WB Saunders, 2000, pp 409–419.
59. Molloy RG, Mannick JA, Rodrick ML. Cytokines, sepsis, and immuno- 83. Kao KJ, Cook DJ, Scornik JC. Quantitative analysis of platelet sur-
modulation. Br J Surg 1993;80:289–297. face HLA by W6/32 anti-HLA monoclonal antibody. Blood 1986;68:
60. Hawrylowicz CM, Howells GL, Feldmann M. Platelet-derived inter- 627–632.
leukin 1 induces human endothelial adhesion molecule expression 84. Rudensky A, Preston-Hurlburt P, Hong SC, Barlow A, Janeway CA
and cytokine production. J Exp Med 1991;174:785–790. Jr. Sequence analysis of peptides bound to MHC class II molecules.
61. El-Kattan I, Anderson J, Yun JK, et al. Mononuclear cell (MC) adhe- Nature 1991;353:622–627.
sion to platelet storage bag plastic polymers correlates with cytokine 85. Chicz RM, Urban RG, Gorga JC, Vignali DA, Lane WS, Strominger
levels [abstract]. Transfusion 1995;35(Suppl):44S. JL. Specificity and promiscuity among naturally processed peptides
62. Dzik WH. Is the febrile response to transfusion due to donor or recipi- bound to HLA-DR alleles. J Exp Med 1993;178:27–47.
ent cytokine? [letter] Transfusion 1992;32:594. 86. Fast LD. Recipient elimination of allogeneic lymphoid cells: donor CD4+
63. Sacher RA, Boyle L, Freter CE. High circulating interleukin 6 levels cells are effective alloantigen-presenting cells. Blood 2000;96:1144–1149.
associated with acute transfusion reaction: cause or effect? Transfusion 87. Kao KJ, del Rosario ML. Role of class-II major histocompatibility complex
1993;33:962–963. (MHC)-antigen-positive donor leukocytes in transfusion-induced allo-
64. Schiller JH, Storer BE, Witt PL, et al. Biological and clinical effects of immunization to donor class-I MHC antigens. Blood 1998;92:690–694.
intravenous tumor necrosis factor-alpha administered three times 88. Gouttefangeas C, Diehl M, Keilholz W, Hornlein RF, Stevanovic S,
weekly. Cancer Res 1991;51:1651–1658. Rammensee HG. Thrombocyte HLA molecules retain nonrenewable
65. Agosti JM, Coombs RW, Collier AC, et al. A randomized, double- endogenous peptides of megakaryocyte lineage and do not stimulate
blind, phase I/II trial of tumor necrosis factor and interferon-gamma direct allocytotoxicity in vitro. Blood 2000;95:3168–3175.
for treatment of AIDS-related complex (Protocol 025 from the AIDS 89. Mellman I, Turley SJ, Steinman RM. Antigen processing for amateurs
Clinical Trials Group). AIDS Res Hum Retroviruses 1992;8:581–587. and professionals. Trends Cell Biol 1998;8:231–237.
66. Tewari A, Buhles WC Jr, Starnes HF Jr. Preliminary report: effects of 90. Vogt AB, Kropshofer H. HLA-DM—an endosomal and lysosomal chap-
interleukin-1 on platelet counts. Lancet 1990;336:712–714. erone for the immune system. Trends Biochem Sci 1999;24:150–154.
67. Crown J, Jakubowski A, Kemeny N, et al. A phase I trial of recombinant 91. Gould DS, Auchincloss H Jr. Direct and indirect recognition: the role
human interleukin-1 beta alone and in combination with myelosup- of MHC antigens in graft rejection. Immunol Today 1999;20:77–82.
pressive doses of 5-fluorouracil in patients with gastrointestinal can- 92. Turley SJ, Inaba K, Garrett WS, et al. Transport of peptide-MHC class
cer. Blood 1991;78:1420–1427. II complexes in developing dendritic cells. Science 2000;288:522–527.
II 68. Dinarello CA. Interleukin-1 and interleukin-1 antagonism. Blood 93. Bang KW, Speck ER, Blanchette VS, Freedman J, Semple JW. Unique
1991;77:1627–1652. processing pathways within recipient antigen-presenting cells deter-
380 69. Jacobi KE, Wanke C, Jacobi A, Weisbach V, Hemmerling TM. Determina- mine IgG immunity against donor platelet MHC antigens. Blood
tion of eicosanoid and cytokine production in salvaged blood, stored red 2000;95:1735–1742.
blood cell concentrates, and whole blood. J Clin Anesth 2000;12:94–99. 94. Heddle NM. The efficacy of leukodepletion to improve platelet trans-
70. Stack G, Baril L, Napychank P, Snyder EL. Cytokine generation in fusion response: a critical appraisal of clinical studies. Transfus Med
stored, white cell-reduced, and bacterially contaminated units of red Rev 1994;8:15–28.
cells. Transfusion 1995;35:199–203. 95. Seftel MD, Growe GH, Petraszko T, et al. Universal prestorage leukore-
71. Claas FH, Smeenk RJ, Schmidt R, van Steenbrugge GJ, Eernisse JG. duction in Canada decreases platelet alloimmunization and refractori-
Alloimmunization against the MHC antigens after platelet transfu- ness. Blood 2004;103:333–339.
sions is due to contaminating leukocytes in the platelet suspension. 96. van de Watering L, Hermans J, Witvliet M, Versteegh M, Brand A. HLA
Exp Hematol 1981;9:84–89. and RBC immunization after filtered and buffy coat-depleted blood
72. Vamvakas EC. Meta-analysis of randomized controlled trials of the transfusion in cardiac surgery: a randomized controlled trial. Transfu-
efficacy of white cell reduction in preventing HLA-alloimmunization sion 2003;43:765–771.
and refractoriness to random-donor platelet transfusions. Transfus 97. Ohto H, Nomizu T, Kuroda F, Hoshi T, Rokkaku Y. HLA alloimmuni-
Med Rev 1998;12:258–270. zation of surgical patients by transfusion with bedside leukoreduced
73. Sniecinski I, O’Donnell MR, Nowicki B, Hill LR. Prevention of refrac- blood components. Fukushima J Med Sci 2003;49:45–54.
toriness and HLA-alloimmunization using filtered blood products. 98. Hillyer CD, Emmens RK, Zago-Novaretti M, Berkman EM. Methods
Blood 1988;71:1402–1407. for the reduction of transfusion-transmitted cytomegalovirus infec-
74. Schiffer CA, Dutcher JP, Aisner J, Hogge D, Wiernik PH, Reilly JP. A tion: filtration versus the use of seronegative donor units. Transfusion
randomized trial of leukocyte-depleted platelet transfusion to modify 1994;34:929–934.
alloimmunization in patients with leukemia. Blood 1983;62:815–820. 99. Smith DM, Shoos-Lipton K. Leukocyte Reduction for the Prevention
75. Andreu G, Dewailly J, Leberre C, et al. Prevention of HLA immuni- of Transfusion-Transmitted Cytomegalovirus. Bethesda, Md., Ameri-
zation with leukocyte-poor packed red cells and platelet concentrates can Association of Blood Banks Press, 1997.
obtained by filtration. Blood 1988;72:964–969. 100. Landaw EM, Kanter M, Petz LD. Safety of filtered leukocyte-reduced
76. van Marwijk Kooy M, van Prooijen HC, Moes M, Bosma-Stants I, blood products for prevention of transfusion-associated cytomegalo-
Akkerman JW. Use of leukocyte-depleted platelet concentrates for the virus infection [letter]. Blood 1996;87:4910.
prevention of refractoriness and primary HLA alloimmunization: a 101. Nichols WG, Price TH, Gooley T, Corey L, Boeckh M. Transfusion-
prospective, randomized trial. Blood 1991;77:201–205. transmitted cytomegalovirus infection after receipt of leukoreduced
77. Oksanen K, Kekomaki R, Ruutu T, Koskimies S, Myllyla G. Prevention blood products. Blood 2003;101:4195–4200.
of alloimmunization in patients with acute leukemia by use of white 102. Vamvakas EC. Is white blood cell reduction equivalent to antibody
cell-reduced blood components—a randomized trial. Transfusion screening in preventing transmission of cytomegalovirus by transfu-
1991;31:588–594. sion? A review of the literature and meta-analysis. Transfus Med Rev
78. Sintnicolaas K, van Marwijk Kooij M, van Prooijen HC, et al. Leukocyte 2005;19:181–199.
depletion of random single-donor platelet transfusions does not prevent 103. Taylor-Wiedeman J, Hayhurst GP, Sissons JG, Sinclair JH. Polymor-
secondary human leukocyte antigen-alloimmunization and refractori- phonuclear cells are not sites of persistence of human cytomegalovirus
ness: a randomized prospective study. Blood 1995;85:824–828. in healthy individuals. J Gen Virol 1993;74(pt 2):265–268.
LEUKOCYTE-REDUCED PRODUCTS
104. Taylor-Wiedeman J, Sissons JG, Borysiewicz LK, Sinclair JH. Monocytes 129. Murphy B, Krensky AM. HLA-derived peptides as novel immunomod-
are a major site of persistence of human cytomegalovirus in peripheral ulatory therapeutics. J Am Soc Nephrol 1999;10:1346–1355.
blood mononuclear cells. J Gen Virol 1991;72(pt 9):2059–2064. 130. Clayberger C, Lyu SC, DeKruyff R, Parham P, Krensky AM. Peptides
105. Larsson S, Soderberg-Naucler C, Moller E. Productive cytomegalovi- corresponding to the CD8 and CD4 binding domains of HLA mol-
rus (CMV) infection exclusively in CD13-positive peripheral blood ecules block T lymphocyte immune responses in vitro. J Immunol
mononuclear cells from CMV-infected individuals: implications for 1994;153:946–951.
prevention of CMV transmission. Transplantation 1998;65:411–415. 131. Dzik S, Szuflad P, Eaves S. HLA antigens on leukocyte fragments and
106. Bolovan-Fritts CA, Mocarski ES, Wiedeman JA. Peripheral blood plasma proteins: prestorage leukoreduction by filtration. Vox Sang
CD14+ cells from healthy subjects carry a circular conformation of 1994;66:104–111.
latent cytomegalovirus genome. Blood 1999;93:394–398. 132. Ghio M, Contini P, Mazzei C, et al. Soluble HLA class I, HLA class
107. Taylor-Wiedeman J, Sissons P, Sinclair J. Induction of endogenous II, and Fas ligand in blood components: a possible key to explain the
human cytomegalovirus gene expression after differentiation of immunomodulatory effects of allogeneic blood transfusions. Blood
monocytes from healthy carriers. J Virol 1994;68:1597–1604. 1999;93:1770–1777.
108. Stanier P, Taylor DL, Kitchen AD, Wales N, Tryhorn Y, Tyms AS. Per- 133. Starzl TE, Zinkernagel RM. Antigen localization and migration in
sistence of cytomegalovirus in mononuclear cells in peripheral blood immunity and tolerance. N Engl J Med 1998;339:1905–1913.
from blood donors. BMJ 1989;299:897–898. 134. Bianchi DW, Zickwolf GK, Weil GJ, Sylvester S, DeMaria MA. Male
109. Slobedman B, Mocarski ES. Quantitative analysis of latent human fetal progenitor cells persist in maternal blood for as long as 27 years
cytomegalovirus. J Virol 1999;73:4806–4812. postpartum. Proc Natl Acad Sci USA 1996;93:705–708.
110. Visconti MR, Pennington J, Garner SF, Allain JP, Williamson LM. 135. Lee TH, Paglieroni T, Ohto H, Holland PV, Busch MP. Survival of
Assessment of removal of human cytomegalovirus from blood com- donor leukocyte subpopulations in immunocompetent transfusion
ponents by leukocyte depletion filters using real-time quantitative recipients: frequent long-term microchimerism in severe trauma
PCR. Blood 2004;103:1137–1139. patients. Blood 1999;93:3127–3139.
111. Cope AV, Sabin C, Burroughs A, Rolles K, Griffiths PD, Emery VC. 136. Lee TH, Paglieroni T, Utter GH, et al. High-level long-term white blood
Interrelationships among quantity of human cytomegalovirus cell microchimerism after transfusion of leukoreduced blood compo-
(HCMV) DNA in blood, donor-recipient serostatus, and administra- nents to patients resuscitated after severe traumatic injury. Transfusion
tion of methylprednisolone as risk factors for HCMV disease follow- 2005;45:1280–1290.
ing liver transplantation. J Infect Dis 1997;176:1484–1490. 137. Chowdhury NC, Murphy B, Sayegh MH, et al. Acquired systemic tol-
112. Miller WJ, McCullough J, Balfour HH Jr, et al. Prevention of cyto- erance to rat cardiac allografts induced by intrathymic inoculation
megalovirus infection following bone marrow transplantation: a ran- of synthetic polymorphic MHC class I allopeptides. Transplantation
domized trial of blood product screening. Bone Marrow Transplant 1996;62:1878–1882.
1991;7:227–234. 138. Mincheff MS, Getsov SI, Meryman HT. Mechanisms of alloimmuni-
113. Roback JD, Drew WL, Laycock ME, Todd D, Hillyer CD, Busch MP. zation and immunosuppression by blood transfusions in an inbred
CMV DNA is rarely detected in healthy blood donors using validated rodent model. Transplantation 1995;60:815–821.
PCR assays. Transfusion 2003;43:314–321. 139. Dzik S, Mincheff M, Puppo F. Apoptosis, transforming growth factor-
114. Bevan IS, Daw RA, Day PJ, Ala FA, Walker MR. Polymerase chain reac- beta, and the immunosuppressive effect of transfusion. Transfusion
tion for detection of human cytomegalovirus infection in a blood 2002;42:1221–1223.
donor population. Br J Haematol 1991;78:94–99. 140. Chen ZK, Cobbold SP, Waldmann H, Metcalfe S. Amplification of
115. Rahbar AR, Sundqvist VA, Wirgart BZ, Grillner L, Soderberg-Naucler natural regulatory immune mechanisms for transplantation tolerance.
C. Recognition of cytomegalovirus clinical isolate antigens by sera Transplantation 1996;62:1200–1206.
from cytomegalovirus-negative blood donors. Transfusion 2004;44: 141. Yang CP, McDonagh M, Bell EB. CD45RC+ CD4 T cell subsets are
1059–1066. maintained in an unresponsive state by the persistence of transfusion-
116. Krajden M, Shankaran P, Bourke C, Lau W. Detection of cytomegalo- derived alloantigen. Transplantation 1995;60:192–199.
26
virus in blood donors by PCR using the digene SHARP signal system 142. Blajchman MA, Bardossy L, Carmen R, Sastry A, Singal DP. Alloge-
assay: effects of sample preparation and detection methodology. J Clin neic blood transfusion-induced enhancement of tumor growth: two 381
Microbiol 1996;34:29–33. animal models showing amelioration by leukodepletion and passive
117. Chou SW. Acquisition of donor strains of cytomegalovirus by renal- transfer using spleen cells. Blood 1993;81:1880–1882.
transplant recipients. N Engl J Med 1986;314:1418–1423. 143. Bordin JO, Bardossy L, Blajchman MA. Growth enhancement of estab-
118. Chou SW. Reactivation and recombination of multiple cytomegalovi- lished tumors by allogeneic blood transfusion in experimental animals
rus strains from individual organ donors. J Infect Dis 1989;160:11–15. and its amelioration by leukodepletion: the importance of the timing
119. Chandler SH, Handsfield HH, McDougall JK. Isolation of multiple of the leukodepletion. Blood 1994;84:344–348.
strains of cytomegalovirus from women attending a clinic for sexually 144. Kao KJ. Induction of humoral immune tolerance to major histocom-
transmitted disease. J Infect Dis 1987;155:655–660. patibility complex antigens by transfusions of UVB-irradiated leuko-
120. Spector SA, Hirata KK, Newman TR. Identification of multiple cyto- cytes. Blood 1996;88:4375–4382.
megalovirus strains in homosexual men with acquired immunodefi- 145. Kirkley SA, Cowles J, Pellegrini VD Jr, Harris CM, Boyd AD, Blumberg
ciency syndrome. J Infect Dis 1984;150:953–956. N. Cytokine secretion after allogeneic or autologous blood transfusion
121. Baldanti F, Simoncini L, Sarasini A, et al. Ganciclovir resistance as a result [letter]. Lancet 1995;345:527.
of oral ganciclovir in a heart transplant recipient with multiple human 146. Quintiliani L, Iudicone P, Di Girolamo M, et al. Immunoresponsiveness
cytomegalovirus strains in blood. Transplantation 1998;66:324–329. of cancer patients: effect of blood transfusion and immune reactivity of
122. Winston DJ, Huang ES, Miller MJ, et al. Molecular epidemiology of tumor infiltrating lymphocytes. Cancer Detect Prev 1995;19:518–526.
cytomegalovirus infections associated with bone marrow transplanta- 147. Tietze M, Kluter H, Troch M, Kirchner H. Immune responsiveness in
tion. Ann Intern Med 1985;102:16–20. orthopedic surgery patients after transfusion of autologous or alloge-
123. Lang DJ. Cytomegalovirus infections in organ transplantation and neic blood. Transfusion 1995;35:378–383.
post transfusion: an hypothesis. Arch Gesamte Virusforsch 1972;37: 148. van de Watering LM, Brand A, Houbiers JG, Klein Kranenbarg WM,
365–377. Hermans J, van de Velde C. Perioperative blood transfusions, with or
124. Collier AC, Kalish LA, Busch MP, et al. Leukocyte-reduced red blood without allogeneic leucocytes, relate to survival, not to cancer recur-
cell transfusions in patients with anemia and human immunodefi- rence. Br J Surg 2001;88:267–272.
ciency virus infection: the Viral Activation Transfusion Study, a ran- 149. Titlestad IL, Ebbesen LS, Ainsworth AP, Lillevang ST, Qvist N, Georg-
domized controlled trial. JAMA 2001;285:1592–1601. sen J. Leukocyte-depletion of blood compon ents does not significantly
125. Opelz G, Sengar DP, Mickey MR, Terasaki PI. Effect of blood trans- reduce the risk of infectious complications. Results of a double-
fusions on subsequent kidney transplants. Transplant Proc 1973;5: blinded, randomized study. Int J Colorectal Dis 2001;16:147–153.
253–259. 150. Baron JF, Gourdin M, Bertrand M, et al. The effect of universal leu-
126. Opelz G, Vanrenterghem Y, Kirste G, et al. Prospective evaluation of kodepletion of packed red blood cells on postoperative infections in
pretransplant blood transfusions in cadaver kidney recipients. Trans- high-risk patients undergoing abdominal aortic surgery. Anesth Analg
plantation 1997;63:964–967. 2002;94:529–537.
127. Gantt CL. Red blood cells for cancer patients [letter]. Lancet 151. van Hilten JA, van de Watering LM, van Bockel JH, et al. Effects
1981;2:363. of transfusion with red cells filtered to remove leucocytes: ran-
128. van Twuyver E, Mooijaart RJ, ten Berge IJ, et al. Pretransplantation domised controlled trial in patients undergoing major surgery. BMJ
blood transfusion revisited. N Engl J Med 1991;325:1210–1213. 2004;328:1281–1291.
BLOOD BANKING 152. van de Watering LM, Hermans J, Houbiers JG, et al. Beneficial effects deficiency in plasma prekallikrein and kininogen. J Lab Clin Med
of leukocyte depletion of transfused blood on postoperative compli- 1982;100:81–93.
cations in patients undergoing cardiac surgery: a randomized clinical 176. Berridge MJ. Inositol trisphosphate and diacylglycerol: two interacting
trial. Circulation 1998;97:562–568. second messengers. Annu Rev Biochem 1987;56:159–193.
153. Wallis JP, Chapman CE, Orr KE, Clark SC, Forty JR. Effect of WBC 177. Sung CP, Arleth AJ, Shikano K, Berkowitz BA. Characterization and
reduction of transfused RBCs on postoperative infection rates in function of bradykinin receptors in vascular endothelial cells. J Phar-
cardiac surgery. Transfusion 2002;42:1127–1134. macol Exp Ther 1988;247:8–13.
154. Bilgin YM, van de Watering LM, Eijsman L, et al. Double-blind, ran- 178. Mousli M, Bueb JL, Bronner C, Rouot B, Landry Y. G protein acti-
domized controlled trial on the effect of leukocyte-depleted erythrocyte vation: a receptor-independent mode of action for cationic amphi-
transfusions in cardiac valve surgery. Circulation 2004;109:2755–2760. philic neuropeptides and venom peptides. Trends Pharmacol Sci
155. Bracey AW, Radovancevic R, Nussmeier NA, et al. Leukocyte-reduced 1990;11:358–362.
blood in open heart surgery patients: effects on outcome [abstract]. 179. Dachman WD, Ford GA, Blaschke TF, Hoffman BB. Mechanism of
Transfusion 2002;42(Suppl):5S. bradykinin-induced venodilation in humans. J Cardiovasc Pharmacol
156. Volkova N, Klapper E, Pepkowitz SH, Denton T, Gillaspie G, Goldfin- 1993;21:241–248.
ger D. A case-control study of the impact of WBC reduction on the 180. O’Kane KP, Webb DJ, Collier JG, Vallance PJ. Local L-NG-mono-
cost of hospital care for patients undergoing coronary artery bypass methyl-arginine attenuates the vasodilator action of bradykinin in the
graft surgery. Transfusion 2002;42:1123–1126. human forearm. Br J Clin Pharmacol 1994;38:311–315.
157. Fung MK, Rao N, Rice J, Ridenour M, Mook W, Triulzi DJ. Leuko- 181. Cockcroft JR, Chowienczyk PJ, Brett SE, Bender N, Ritter JM. Inhibi-
reduction in the setting of open heart surgery: a prospective cohort- tion of bradykinin-induced vasodilation in human forearm vascula-
controlled study. Transfusion 2004;44:30–35. ture by icatibant, a potent B2-receptor antagonist. Br J Clin Pharmacol
158. Janatpour KA, Paglieroni TG, Crocker VL, DuBois DJ, Holland PV. 1994;38:317–321.
Visual assessment of hemolysis in red blood cell units and segments 182. Bonner G, Preis S, Schunk U, Toussaint C, Kaufmann W. Hemody-
can be deceptive. Transfusion 2004;44:984–989. namic effects of bradykinin on systemic and pulmonary circula-
159. Llewelyn CA, Taylor RS, Todd AA, Stevens W, Murphy MF, Williamson tion in healthy and hypertensive humans. J Cardiovasc Pharmacol
LM. The effect of universal leukoreduction on postoperative infections 1990;15(Suppl 6):S46–S56.
and length of hospital stay in elective orthopedic and cardiac surgery. 183. Shiba M, Tadokoro K, Sawanobori M, Nakajima K, Suzuki K, Juji T.
Transfusion 2004;44:489–500. Activation of the contact system by filtration of platelet concentrates
160. California Blood Bank Society. Available at http://www.cbbsweb.org/ with a negatively charged white cell-removal filter and measurement
enf/enetcomm2001.html. Accessed March 28, 2006. of venous blood bradykinin level in patients who received filtered
161. US Food and Drug Administration, Center for Biologics Evaluation platelets. Transfusion 1997;37:457–462.
and Research. Available at http://www.fda.gov/cber/recalls/leuk- 184. Shiba M, Tadokoro K, Nakajima K, Juji T. Bradykinin generation in
pal022305.htm. Accessed March 28, 2006. RC-MAP during storage at 4°C and leukocyte removal filtration.
162. Gyongyossy-Issa MI, Weiss SL, Sowemimo-Coker SO, Garcez RB, Thromb Res 1997;87:511–520.
Devine DV. Prestorage leukoreduction and low-temperature filtra- 185. Hild M, Soderstrom T, Egberg N, Lundahl J. Kinetics of bradykinin
tion reduce hemolysis of stored red cell concentrates. Transfusion levels during and after leucocyte filtration of platelet concentrates. Vox
2005;45:90–96. Sang 1998;75:18–25.
163. Sano H, Koga Y, Hamasaki K, Furuyama H, Itami N. Anaphylaxis asso- 186. Scott CF, Brandwein H, Whitbread J, Colman RW. Lack of clinically
ciated with white-cell reduction filter [letter]. Lancet 1996;347:1053. significant contact system activation during platelet concentrate filtra-
164. Fried MR, Eastlund T, Christie B, Mullin GT, Key NS. Hypotensive reac- tion by leukocyte removal filters. Blood 1998;92:616–622.
tions to white cell-reduced plasma in a patient undergoing angiotensin- 187. Cyr M, Hume HA, Champagne M, et al. Anomaly of the des-Arg9-
converting enzyme inhibitor therapy. Transfusion 1996;36:900–903. bradykinin metabolism associated with severe hypotensive reac-
II 165. Mair B, Leparc GF. Hypotensive reactions associated with platelet tions during blood transfusions: a preliminary study. Transfusion
transfusions and angiotensin-converting enzyme inhibitors. Vox Sang 1999;39:1084–1088.
382 1998;74:27–30. 188. Arnold DM, Molinaro G, Warkentin TE, et al. Hypotensive transfusion
166. Yenicesu I, Tezcan I, Tuncer AM. Hypotensive reactions during platelet reactions can occur with blood products that are leukoreduced before
transfusions. Transfusion 1998;38:410; author reply 413–415. storage. Transfusion 2004;44:1361–1366.
167. Sweeney JD, Dupuis M, Mega AP. Hypotensive reactions to red cells 189. Owen HG, Brecher ME. Atypical reactions associated with use of
filtered at the bedside, but not to those filtered before storage, in angiotensin-converting enzyme inhibitors and apheresis. Transfusion
patients taking ACE inhibitors. Transfusion 1998;38:410–411; author 1994;34:891–894.
reply 413–415. 190. Reutter JC, Sanders KF, Brecher ME, Jones HG, Bandarenko N. Inci-
168. Abe H, Ikebuchi K, Shimbo M, Sekiguchi S. Hypotensive reactions dence of allergic reactions with fresh frozen plasma or cryo-super-
with a white cell-reduction filter: activation of kallikrein-kinin cascade natant plasma in the treatment of thrombotic thrombocytopenic
in a patient. Transfusion 1998;38:411–412; author reply 413–415. purpura. J Clin Apher 2001;16:134–138.
169. Belloni M, Alghisi A, Bettini C, Soli M, Zampieri L. Hypotensive reac- 191. Perseghin P, Capra M, Baldini V, Sciorelli G. Bradykinin production
tions associated with white cell-reduced apheresis platelet concentrates during donor plasmapheresis procedures. Vox Sang 2001;81:24–28.
in patients not receiving ACE Inhibitors. Transfusion 1998;38:412–415. 192. Molinaro G, Adam A, Lepage Y, Hammerschmidt D, Koenigbauer U,
170. Myers T, Uhl L, Kruskall MS. Association between angiotensin-con- Eastlund T. Hypotensive reaction during staphylococcal protein A
verting enzyme (ACE) inhibitors and hypotensive transfusion reac- column therapy in a patient with anomalous degradation of brady-
tions [abstract]. Transfusion 1996;36(Suppl):60S. kinin and Des-Arg9-bradykinin after contact activation. Transfusion
171. Verresen L, Waer M, Vanrenterghem Y, Michielsen P. Angiotensin-con- 2002;42:1458–1465.
verting-enzyme inhibitors and anaphylactoid reactions to high-flux 193. Elghouzzi MH, Vedrenne JB, Jullien AM, Delcey D, Nadal M, Habibi B.
membrane dialysis. Lancet 1990;336:1360–1362. Etude technique immunologique et clinique des performances de fil-
172. Olbricht CJ, Schaumann D, Fischer D. Anaphylactoid reactions, tration du sang a la l’aide de l’appareill Erypur. Rev Fr Transfus Immu-
LDL apheresis with dextran sulphate, and ACE inhibitors. Lancet nohematol 1981;24:579–595.
1992;340:908–909. 194. Lane TA, Myllyla G. Leukocyte-depleted blood products. In Leikola J,
173. Davidson DC, Peart I, Turner S, Sangster M. Prevention with icatibant Lundgaard-Hansen P (eds). Current Studies in Hematology and Blood
of anaphylactoid reactions to ACE inhibitor during LDL apheresis [let- Transfusion. Basel, Karger, 1994;60:1–145.
ter]. Lancet 1994;343:1575. 195. Murphy MF, Metcalfe P, Thomas H, et al. Use of leucocyte-poor blood
174. Nielsen F, Damkjaer Nielsen M, Rasmussen S, Kappelgaard AM, Giese components and HLA-matched platelet donors to prevent HLA allo-
J. Bradykinin in blood and plasma: facts and fallacies. Acta Med Scand immunization. Br J Haematal 1986;62:529–534.
Suppl 1983;677:54–59.
175. Scicli AG, Mindroiu T, Scicli G, Carretero OA. Blood kinins, their
concentration in normal subjects and in patients with congenital
Chapter 27
Virus-Safe Products: Pathogen
Reduction and Inactivation
Laurence Corash
Cell- Integrated
associated proviral DNA
virus in nucleus
Cell-free virus
Figure 27–1 Infectious pathogens (e.g., viruses) may be present in the plasma as cell-free virus, associated with cell membranes, in the cell
cytoplasm, or in the nucleus. Moreover, some viruses (e.g., retroviruses) may integrate nucleic acid sequences into host genomic DNA. A robust
pathogen inactivation technology should be effective in each of these compartments.
inactivation of infectious pathogens in blood components. two basic groups: nucleic acid targeted and photodynamic
Treatment of plasma fractions with the solvent detergent methods. Both methods utilize photochemical reactions in
process has demonstrated the benefits of this approach.34 an ex vivo treatment process combining a photo reactive
II A robust inactivation technology that is compatible with compound and ultraviolet (UV) light. Nucleic acid targeting
current blood component processing procedures offers the using psoralen compounds has been extensively investigated
384 potential for significantly improving transfusion safety. To and is discussed in detail below, since a method using the
be highly effective, a successful technology must inactivate novel psoralen amotosalen has undergone clinical trials and
pathogens in extracellular, intracellular, and nuclear com- been commercialized. The photodynamic methods generally
partments (Fig. 27–1). In the last case, inactivation also result in lower levels of pathogen inactivation and are associ-
must be effective against pathogen nucleic acid sequences ated with more platelet injury.38 These technologies will not
integrated into donor leukocytes. be reviewed further (see Table 27–2), with the exception of
Furthermore, a technology capable of inactivating resid- the riboflavin method, which has advanced to clinical trials.
ual leukocytes may confer additional benefits owing to inhi- The psoralen-based methods generally rely on nucleic
bition of critical leukocyte functions, including cytokine acid–specific adduct formation,39 in contrast to photo-
synthesis, lymphocyte proliferation, and antigen presenta- dynamic processes that tend to utilize the production of
tion. Donor leukocytes may be associated with a variety active oxygen species in addition to nucleic acid reactions
of adverse immune events ranging in severity from febrile as the primary mechanism for pathogen inactivation.40
transfusion reactions to alloimmunization and graft-versus- Psoralens are low-molecular-weight, planar furocoumarins
host disease.35,36 Recently, studies have shown that massively (Fig. 27–2). In the absence of UV light, psoralens revers-
transfused patients may develop stable microchimerism, the ibly intercalate into helical regions of DNA and RNA,
significance of which is unclear.37 Although a number of under equilibrium kinetics. Upon illumination with UVA
measures have been implemented to reduce the likelihood (320 to 400 nm), psoralens react with pyrimidine bases to
of these adverse immune reactions, a robust nucleic acid tar- form covalent mono-adducts and cross-links with nucleic
geted pathogen inactivation process offers the potential to acids (Fig. 27–3). Bacteria,41 protozoa,42,43 viruses,44 and
inactivate leukocytes as well as infectious pathogens. nucleated cells45 with genomes that have been modified by
psoralens are unable to replicate.
Early investigations with psoralen-mediated pathogen
SYSTEMS FOR INACTIVATION inactivation were conducted with 8-methxyopsoralen (8-
OF PATHOGENS IN PLATELET MOP)46 based on the history of prior human use to treat pso-
CONCENTRATES riasis and cutaneous T-cell lymphoma. These initial studies
by Lin and co-workers established the principle of psoralen-
Considerable effort has been devoted to developing methods mediated pathogen inactivation, but 8-MOP photochemical
for pathogen inactivation in platelet concentrates (Tables treatment was not a sufficiently rapid process for treatment
27–1 and 27–2). The potential processes are divided into of platelet concentrates in clinical use.46,47
VIRUS-SAFE PRODUCTS
Table 27–1 Psoralen Methods Used to Inactivate Pathogens and Leukocytes in Platelet Concentrates
8-MOP, 8-methoxypsoralen; AMT, aminomethyltrimethyl psoralen; PSR-Br, brominated psoralens; fd, bacteriophage; R17, bacteriophage;
FeLV, feline leukemia virus; MCMV, murine cytomegalovirus; FeRTV, feline rhinotracheitis virus; HIV, human immunodeficiency virus; HSV,
herpes simplex virus; DHBV, duck hepatitis B virus; VSV, vesicular stomatitis virus; Sindbis, Sindbis virus; BVDV, bovine viral diarrhea virus; CMV,
cytomegalovirus.
Two laboratories investigated the use of aminomethyltri- A new amino psoralen, amotosalen, was synthesized and
methylpsoralen (AMT), a synthetic psoralen with enhanced shown to be highly effective for inactivation of pathogenic
nucleic acid–binding efficiency (see Table 27–2). Although viruses, bacteria, and leukocytes in platelet concentrates with
AMT has increased nucleic acid–binding affinity compared preservation of in vitro platelet function properties (Table 27–
with 8-MOP, it exhibits mutagenicity in the absence of light 3). Lin and co-workers reported that human platelet concen-
and thus has an unfavorable toxicology profile. Several classes trates (300 mL) contaminated with high titers of HCV(104.5)
of new psoralens have been synthesized that offer potential and HBV(105.5) and treated with amotosalen did not transmit
advantages over AMT and 8-MOP. The halogenated pso- hepatitis after transfusion into naive chimpanzees. Jordan and
ralens do not appear to be sufficiently effective for viral inac- colleagues have shown that amotosalen treatment effectively
tivation and in preliminary studies demonstrated adverse prevents transfusion-transmitted CMV infection in a sensi-
effects on platelet viability.48,49 tive murine model.50 Other studies demonstrated that amo-
tosalen inactivates high levels of T cells, inhibits leukocyte 27
cytokine synthesis during platelet storage, and inhibits nucleic
acid amplification.45,51 More importantly, treatment of T cells
385
with the amotosalen process prevented transfusion-associated
Table 27–2 Photodynamic Methods Used to
graft-versus-host disease (TA-GVHD) in both immunocom-
Inactivate Pathogens in Platelet Concentrates
petent and immunocompromised murine bone marrow trans-
Photoreactive Agent Target Reference plant models.52 Amotosalen treatment of platelet concentrates
NH2
NH2 O
+ UVA
HIV, human immunodeficiency virus; HBV, hepatitis B virus; HCV, hepatitis C virus; HCMV, human cytomegalovirus, DHBV, duck hepatitis
B virus; BVDV, bovine viral diarrhea virus, MCMV, murine cytomegalovirus; pfu, plaque-forming units; ID50 infectious dose which is measured
from an endpoint dilution that causes infection in 50% of inoculated animals; cfu, colony-forming units; TCID50, tissue culture infectious dose,
which is measured from an endpoint dilution that causes infection in 50% of inoculated samples; CID50, chimpanzee infectious dose, which is
measured from an endpoint dilution that causes infection in 50% of inoculated chimpanzees.
*
Three studies are indicated: the HIV-1 provirus inactivation was performed in cell culture medium and the other two studies were
performed with platelet sample size less than 300 mL.
†
Highest titers possible.
‡
The infectivity of the MS-2 strain of HBV and the Hutchinson strain of HCV was measured in susceptible chimpanzees.
II
386
effectively blocks PCR mediated amplification of mitochon- of platelets in 100% donor plasma.56 This process has dem-
drial DNA sequences without impairment of ATP levels.53,54 onstrated activity against enveloped viruses, a nonenvel-
A photochemical treatment process using riboflavin (Fig. oped model virus, and two species of bacteria.55 Earlier
27–4) and a combination of UVB and UVA light (5 J/cm2: studies showed capacity for the inactivation of protozoa.57
265–370 nm) has been developed for platelet components In vitro aspects of platelet function and hemostatic func-
suspended in donor plasma without the use of a platelet tion assessed in an ex vivo system have shown conserva-
additive solution.55 Photodynamic reactions involving the tion of function after riboflavin treatment.58 The process
generation of active oxygen species are a critical aspect appears to produce comparable effects on in vitro plate-
of riboflavin pathogen inactivation mechanism.40 Earlier let properties for apheresis and pooled buffy coat platelet
studies had used this process with additive solutions, but components.59 Recently, preliminary data were reported
the system evaluated in clinical trials requires suspension describing the inactivation of lymphocytes using a murine
immunodeficiency model.60
HIV, human immunodeficiency virus; HSV, herpes simplex virus; CMV, cytomegalovirus; SIV, simian immunodeficiency virus; VSV, vesicular
stomatitis virus; FeLV, feline leukemia virus; Friend LV, Friend erythroleukemia virus; Φ6, bacteriophage; Sindbis, Sindbis virus; PSR-Br,
brominated psoralen; PRV, pseudorabies virus; BVDV, bovine viral diarrhea virus; EMC, encephalo-myocarditis virus; R17, bacteriophage; DMMB,
dimethylmethylene blue.
BLOOD BANKING
R1 R2
N [R5−N+(R6,R7)]nR8+Xn− Reaction
R3 R4 X
Figure 27–6 Structure of INACTINE compounds. PEN 110 is a mem-
ber of the class.
Breakdown X−
S-303
Cl
Treated Treated
N Red Red red red
O blood blood blood blood
cells cells cells cells
O Cl
S-303 CAD
NH
Figure 27–9 The prototype FRALE S-303 process for treatment of
red blood cell concentrates. The treatment is conducted in a closed sys-
tem. S-303 is maintained at low pH and is activated by addition to red
cell concentrates at neutral pH. The inactivation process is conducted in
N full-sized units ranging in hematocrit from 60% to 80%. After addition
of S-303 to the red cells, the plastic container is incubated at room tem-
Figure 27–7 Structure of the frangible anchor linker effector perature for 6 to 8 hours to complete pathogen inactivation. Following
(FRALE) compound S-303. The molecule consist of three parts: an pathogen inactivation, the red cells are transferred to a plastic con-
anchor for nucleic acid intercalation, an effector region for covalent tainer with a compound absorption device (CAD) to reduce the levels
addition to nucleic acid, and a labile, “frangible” linker to facilitate of S-300. The red cells remain in contact with the CAD for up to 35
S-303 degradation. days of storage at 4° C.
VIRUS-SAFE PRODUCTS
35-day-old S-303 red cells was statistically lower than that of studies until the mechanism of the immunologic response
control red cells (78.7 ± 5.7% vs 83.9 ± 6.1%, P = 0.002), but could be elucidated.
recovery for both types exceeded 75%, the generally accepted Evaluation of sera from two patients with consistently
threshold for viability of stored red cells. No adverse events reactive crossmatches to S-303–treated red cells confirmed
were observed following transfusion of S-303 red cells. that the reactivity was directed against the acridine (anchor
A second phase 1 study was conducted in which the moiety) portion of S-303.99 The antibody was low titer and
viability and potential immune response of healthy sub- did not promote red cell phagocytosis in a monocyte-macro-
jects to multiple exposures to autologous S-303–treated red phage ingestion assay.100 No patients in the acute anemia trial
cells were evaluated.95 Subjects from the first study, of either developed crossmatch reactivity with S-303–treated red cells.
treatment group, were invited to participate. The second Prior to detection of positive crossmatch reactivity in the two
study was initiated approximately 6 months after comple- chronic study patients, 148 patients completed the acute ane-
tion of the first study; thus, there was an interval of 6 to 9 mia study. Analysis of results from the study demonstrated
months before the next transfusion exposure. In the second that the primary end point was met, as well as all secondary
study, each subject donated a unit of blood that was pro- end points.97 The treated red cells were well tolerated and
cessed into packed red cells. All the units were treated with the safety profile was comparable to that of conventional red
S-303 and stored for 35 days. At four times, approximately 7 cells. In the chronic trial, no patients completed the study,
days apart, during the 35-day storage period, subjects were but an interim analysis of data suggested that S-303–treated
transfused with an aliquot of autologous S-303–treated red red cells were similar to conventional red cells for transfusion
cells. The final aliquot on day 35 was radiolabeled with 51- support of chronic anemia.98 Further investigation of the
Cr to measure post-transfusion viability. The red cell recov- S-303 process resulted in development of a modified process
ery 24 hours after transfusion of S-303–treated red cells was using an increased concentration of glutathione. Additional
compared with each subject’s red cell recovery in the first clinical trials are required to demonstrate that this modi-
study. Sixteen subjects received control red cells in the first fied process avoids cross-reactivity and alloimmunization in
study and S-303 red cells in the second study. The average chronically transfused patients.
red cell recovery in the first study with control cells was
82.6 ± 6.4%, which was not statistically different from the
average recovery of 84.3 ± 6.4% after transfusion of S-303 SYSTEMS FOR INACTIVATION OF
cells in the second study (P = 0.30). Twelve subjects received PATHOGENS IN PREPARATION
S-303 red cells in the first study, with an average recovery OF FRESH FROZEN PLASMA
of 77.5 ± 6.6%, and after transfusion in the second study,
these subjects demonstrated an average recovery of 79.2 ± The solvent detergent (SD) process for inactivation of envel-
6.6%. After exposure to either four or five aliquots of S-303– oped viruses in plasma used for transfusion was introduced
treated red cells, no subjects developed antibodies directed into clinical practice.101 The SD process is generally highly
against S-303 red cells.95 effective against enveloped viruses, although very recently 27
A third phase 1 study was conducted in 29 healthy new evidence has been reported to suggest that some
subjects using a randomized crossover design to measure enveloped viruses, such as vaccinia, may be resistant to SD 391
both post-transfusion recovery and life span of autologous treatment.102 Moreover, since SD treatment is not effective
S-303–treated red cells compared with control red cells pro- against nonenveloped viruses, concern has been expressed
cessed with conventional methods.96 Both types of red cells over the potential for transmission of these pathogens as a
were stored for 35 days. The study was designed to detect result of pooling with failure to inactivate resistant viruses.103
small differences in red cell viability. A small difference in Despite these issues, solvent detergent fresh frozen plasma
24-hour post-transfusion recovery of S-303–treated red cells (SD-FFP) has demonstrated therapeutic efficacy for replace-
was detected compared with conventional red cells (81.7 ± ment of congenital104 and acquired coagulation factor defi-
6.3% vs 84.5 ± 6.2%, P = 0.048). However, mean half-life was ciencies.105,106 Furthermore, since SD-FFP does not contain
identical (37.4 ± 8.9 vs 37.4 ± 6.7 days, P = 1.0). high-molecular-weight multimers of von Willebrand fac-
On the basis of the three phase 1 red cell clinical trials, tor (vWF), it has been advocated for use during therapeutic
phase 3 clinical trials were initiated to evaluate the therapeu- plasma exchange (TPE) therapy of thrombotic thrombocy-
tic efficacy and safety of S-303–treated red cells in patients for topenic purpura (TTP).104 Transfusion of SD-FFP has been
the two major indications for red cell transfusion: acute and well tolerated with a low incidence of transfusion reactions
chronic anemia. For the indication of acute anemia, a ran- and adverse events.107
domized, two-arm controlled study to evaluate the response With increasing experience, other studies with SD-FFP
to transfusion in 200 patients undergoing cardiovascular sur- documented mild to marked decreased levels of antithrom-
gical procedures was initiated.97 The primary end point of this botic proteins: protein S and plasmin inhibitor.108 Mast and
study was the composite incidence of renal failure, myocardial co-workers confirmed that SD-FFP has markedly reduced lev-
infarction, and death as a reflection of tissue oxygenation effi- els of the antithrombotic proteins antiplasmin and antitrypsin
cacy during correction of acute anemia. For the therapeutic and that these reduced levels are due to conformational changes
indication of chronic anemia, a randomized, two-period cross- induced by detergent treatment.109 The clinical significance of
over study enrolling patients with thalassemia and sickle cell these observations was unclear when initially reported, but
anemia requiring chronic transfusion support was initiated. subsequent reports indicated that there was an increased inci-
The primary end point of this study was the consumption of dence of thrombotic events after large volume exposure dur-
hemoglobin (g/kg body weight per day) over approximately 6 ing liver transplantation110 and TPE for TTP.111,112
months.98 Both studies were under way when two patients in The SD process utilizes pooled plasma from 100 to 2500
the chronic trial developed positive crossmatch reactivity to donors and is not amenable to a single-unit viral inactivation
S-303–treated red cells.99 The sponsor elected to suspend both process. In Europe, a single-unit treatment process using the
BLOOD BANKING
phenothiazine dye methylene blue (MB) and long wavelength tion, chemistry, or hematology profiles.125,126 Peak post-trans-
visible light has been used in clinical practice for preparation of fusion levels (9 ± 1 ng/mL) of amotosalen were determined
fresh frozen plasma (MB-FFP).113 MB exhibits limited binding immediately after transfusion of 1000 mL of treated plasma
affinity for nucleic acid; rather, the predominant mechanism of and 18 to 24 hours later (0.52 ± 0.1 ng/mL).
action is photodynamic via production of active oxygen spe- In a second study, using a randomized crossover design,
cies. More importantly, methylene blue is not effective against Hambleton and colleagues compared the pharmacokinetics
intracellular viruses owing to conversion to the inactive leuko– of factor VII and the post-transfusion recoveries of factors
methylene blue species. Certain coagulation proteins, primarily II, VII, IX, and X in 27 healthy subjects after transfusion of
factor VIII and fibrinogen, are sensitive to MB treatment and conventional and treated FFP. Subjects received a 4-day regi-
undergo a decrease in functional activity during treatment men (7.5 mg/day) of warfarin prior to each study transfusion
ranging from 30% to 40% of the levels in untreated plasma.113 to suppress endogenous factor VII levels.127 At the start of the
More recent studies with a commercial system using both fresh study, subjects donated 2 L of FFP by apheresis collection.
and frozen plasma with leukocyte reduction and a device to One liter of plasma was treated with amotosalen and frozen
reduce the residual levels of MB demonstrated decreases in at −18 °C; the other liter was prepared as standard FFP and
factor VIII activity ranging from 24% to 29%.114 MB-FFP did stored at −18 °C. Each subject received both types of FFP in
provide adequate levels of fibrinogen, factor VIII, factor XIII, random order. Prior to each transfusion, subjects were given
and vWF in cryoprecipitate, and the corresponding cryosu- 4 days of warfarin therapy followed by transfusion of 1 L of
pernatant fraction was depleted of the high-molecular-weight amotosalen FFP or untreated FFP. Warfarin therapy resulted
vWF multimers.115 in reduction of factor VII levels to approximately 30% of nor-
MB-FFP has been well tolerated when transfused into mal. No statistical differences (Wilcoxon signed-rank test) of
healthy subjects.116 Reports of randomized, controlled clini- clearance, recovery, half-life, or mean residence time (MRT)
cal trials for MB-FFP for the major therapeutic indications of for factor VII were observed between amotosalen and con-
FFP use are limited, despite transfusion of more than 1 mil- trol FFP.127 No differences in recoveries of other factors (II,
lion units in Europe.117 A single study compared the usage IX, and X) were observed. Transfusion of amotosalen FFP
of SD and MB-FFP in cardiac surgery patients and reported provided acceptable coagulation factor preservation without
improved levels of protein S and α2-antiplasmin with MB- adverse effects. The anticoagulant challenge crossover trial
FFP.118 Several reports have indicated that MB-FFP use for demonstrated that transfusion of amotosalen FFP yielded
TPE to treat TTP resulted in reduced efficacy.119,120 A more therapeutic coagulation factor increments similar to those of
generalized experience with MB-FFP was reported based on standard FFP.
a retrospective analysis of three 1-year periods before and A phase 2 randomized, controlled, pilot study of amo-
after adoption of MB-FFP.121 This study indicated that use of tosalen FFP was completed in 13 patients with acquired
MB-FFP resulted in increased use of FFP and cryoprecipitate coagulopathy to evaluate the response of prolonged pro-
owing to reduced hemostatic efficacy of the treated plasma. thrombin (PT) and partial thromboplastin times (PTT) to
II These observations have yet to be confirmed by additional transfusion with amotosalen FFP. Patients with a diagnosis
studies. of acquired coagulopathy undergoing minor invasive pro-
392 A photochemical method using amotosalen and UVA cedures were randomized to transfusion with either amoto-
light has been developed to inactivate viruses in single units salen FFP or conventional FFP.128 The average PT and PTT
of plasma prepared as FFP. Using a prototype configuration, were prolonged before FFP transfusion and responded simi-
with 3 J/cm2 of UVA delivered in approximately 3 minutes, larly to transfusion with either amotosalen or conventional
the average log reduction of virus was cell associated–HIV FFP. The response lasted 8 to 12 hours, and amotosalen FFP
6.4 ± 0.2, cell-free HIV >5.9 (95% CI), DHBV 5.4 ± 0.4, and demonstrated acceptable control of bleeding in patients
BVDV 6.7 ± 0.4 after only 0.5 J/cm.2 Because the amotosalen undergoing invasive procedures, such as liver biopsy. Peak
process is nucleic acid–targeted, it has demonstrated some post-transfusion plasma levels of amotosalen were 3.10 ±
inactivation efficacy of several nonenveloped viruses (e.g., 1.65 ng/mL.
rotavirus, calicivirus, and blue tongue virus). After PCT, FFP A phase 3 clinical trial program for amotosalen FFP was
units were treated for 1 hour with a CAD to reduce residual conducted to assess therapeutic efficacy and safety for the
amotosalen concentration followed by freezing (−20 °C). major indications for FFP transfusion: congenital coagu-
Coagulation activity of thawed amotosalen FFP units was lopathy, acquired coagulopathy, and therapeutic plasma
compared with matched-control FFP units without PCT or exchange. The first study evaluated the post-transfusion
CAD. Retained factor activities (% of control ± SD) were recovery and clearance of specific coagulation factors in
clottable fibrinogen, 87 ± 3%; factor V, 98 ± 2%; factor VII, 34 patients with congenital coagulopathies requiring FFP
86 ± 2%; factor VIII, 73 ± 4%; factor IX, 95 ± 4%; factor X, 98 transfusion.129 This open-label, single arm study enrolled
± 4%; and factor XI, 91 ± 5%.122 This system has been modi- patients with deficiencies of factors I, II, V, VII, X, XI, and
fied to treat up to 650 mL of plasma in a single process using protein C. Enrolled patients received at least one transfusion
a flow instead of a batch CAD to shorten processing time.123 of amotosalen FFP (suggested dose of 15 mL/kg) to measure
This system is compatible with either apheresis plasma col- the recovery and clearance of the specified deficient clotting
lections124 or pools of 2 or 3 units of whole blood–derived factor. Patients were eligible to receive additional amoto-
plasma. salen FFP transfusions as required to manage active bleed-
Plasma prepared with the prototype system has been eval- ing or for prophylaxis during surgical procedures. This study
uated in a series of clinical trials. A single blinded, crossover, demonstrated that the post-transfusion clearance of factors
stepwise ascending dose clinical trial (100–1000 mL) was V, VII, X, XI, and protein C was comparable to reported val-
conducted in 15 healthy subjects. No adverse events were ues.129 The clearance of factors I, II, and XIII were shorter
attributed to transfusion of amotosalen FFP at any dose and than the reported values but provided acceptable post- trans-
no clinically significant changes in post-transfusion coagula- fusion recoveries for support of patients.129 Pretransfusion
VIRUS-SAFE PRODUCTS
prolonged prothrombin times (PT) and PTT corrected after approach to transfusion safety by using pathogen inactiva-
transfusion of amotosalen FFP.129 Seventy-seven transfu- tion and a modified testing strategy to achieve higher levels
sions were administered for management of active bleeding of blood component transfusion safety. As new pathogens of
or for prophylaxis against bleeding during surgical proce- clinical importance are identified in the donor population,
dures. In each case, acceptable hemostasis was achieved after both pathogen inactivation and testing strategies will require
transfusion.129 Some patients received repeated transfusions further modification to meet new challenges.
of amotosalen FFP over 2 years. Peak plasma levels of amoto-
salen immediately after transfusion were 8.42 ± 2.72 ng/mL,
similar to those observed in healthy subjects. The treated REFERENCES
plasma was well tolerated, and the adverse event profile was
similar to that reported for conventional plasma. 1. Dodd RY. Will blood products be free of infectious agents? In Nance
A controlled, double-blinded, randomized study of 121 SJ (ed). Transfusion Medicine in the 1990s. Arlington, Va., American
Association of Blood Banks, 1990, pp 223–251.
patients with acquired coagulopathy, largely due to chronic liver 2. Laperche S. Blood safety and nucleic acid testing in Europe. Euro Sur-
disease, was conducted to evaluate therapeutic efficacy and safety veill 2005;10:3–4.
for this indication.130 Prior to elective surgical procedures, includ- 3. Alvarez do Barrio M, Gonzalez Diez R, Sanchez Hernandez JM, Oyon-
ing liver transplantation, patients were randomized to receive arte Gomez S. Residual risk of transfusion-transmitted viral infections
either conventional FFP or amotosalen FFP for up to 7 days of in Spain, 1997–2002, and impact of nucleic acid testing. Euro Surveill
2005;10:20–22.
support. The primary end point of this study was the response 4. Offergeld R, Faensen D, Ritter D, Hamouda O. Human immunode-
of PT and PTT to FFP transfusion. This study demonstrated ficiency virus, hepatitis C and hepatitis B infections among blood
correction of the PT and PTT responses on a dose- and patient donors in Germany, 2000–2002: risk of virus transmission and the
weight–adjusted basis, hemostasis, and use of blood products impact of nucleic acid testing. Euro Surveill 2005;10:8–11.
5. Busch MP, Kleinman SH, Nemo GJ. Current and emerging infectious
comparable to that of patients supported with conventional FFP. risks of blood transfusions. JAMA 2003;289:959–962.
The mean incremental recovery for factor VII was not different 6. Satake M. Lookback study for transfusion-related HBV infection in
between the treatment groups (0.69 ± 1.80 vs 0.48 ± 0.76 IU/dL/ Japan. Transfusion 2005;45(Suppl 3):9A–10A.
IU/kg). After large doses of amotosalen FFP, peak post-transfu- 7. McQuiston JH, Childs JE, Chamberland ME, Tabor E. Transmission of
sion plasma levels of amotosalen (5.4 ± 3.5 ng/mL) were lower tick-borne agents of disease by blood transfusion: a review of known
and potential risks in the United States. Transfusion 2000;40:274–284.
than those of nonbleeding healthy subjects. Transfusion of amo- 8. Schmunis GA. Trypanosoma cruzi, the etiologic agent of Chagas dis-
tosalen FFP was well tolerated in this patient group. ease: status in the blood supply in endemic and nonendemic countries.
A double-blinded randomized study of patients with TTP Transfusion 1991;31:547–557.
was subsequently carried out.131 In this study, 35 patients 9. Grant IH, Gold JM, Wittner M, et al. Transfusion-associated acute Cha-
gas disease acquired in the United States. Ann Intern Med 1989;111:
were assigned to receive up to 30 days of TPE with either 849–851.
conventional FFP or amotosalen FFP. The primary end point 10. Nickerson P, Orr P, Schroeder ML, Sekla L, Johnston JB. Transfusion-
was the proportion of patients in remission in each treat- associated Trypanosoma cruzi infection in a nonendemic area. Ann
ment group after 30 days of TPE. Patients who failed to reach Intern Med 1989;111:851–853. 27
remission after 30 days of TPE were eligible to continue TPE 11. Riera C, Muncunill J, Fisa R, et al. Asymptomatic infection by Leishma-
nia infantum in blood donors from the Balearic Islands, Spain. Trans-
for an additional 30 days. The rates of remission were not dif- fusion 2005;45(Suppl 3):18A. 393
ferent between the treatment groups (82 vs 89%, P = 0.66). In 12. Leiby DA, Gill JE. Transfusion-transmitted tick-borne infections: a
addition, time to remission, relapse rates, and time to relapse cornucopia of threats. Transfus Med Rev 2004;18:293–306.
were not different for patients treated with amotosalen FFP. 13. Linden JV, Wong SJ, Chu FK, Schmidt GB, Bianco C. Transfusion-
associated transmission of babesiosis in New York State. Transfusion
Despite the exposure to large volumes of amotosalen FFP, 2000;40:285–289.
peak plasma levels were within the same range as the other 14. Petersen LR, Epstein JS. Problem solved? West Nile virus and transfu-
patient groups (5.4 ± 3.6 ng/mL). As in the other studies, sion safety. N Engl J Med 2005;353:516–517.
amotosalen FFP was well tolerated and the safety profile was 15. Iwamoto M, Jernigan DB, Guasch A, et al. Transmission of West Nile
similar to that for patients supported with conventional FFP. virus from an organ donor to four transplant recipients. N Engl J Med
2003;101:2196–2203.
16. Busch MP, Cagliotti S, Robertson EF, et al. Screening the blood supply
for West Nile virus RNA by nucleic acid amplification testing. N Engl J
CONCLUSIONS Med 2005;353:460–467.
17. Stramer SL, Fang CT, Foster GA, Wagner AG, Brodsky JP, Dodd RY.
West Nile virus among blood donors in the United States, 2003 and
New threats to the safety and availability of labile blood com- 2004. N Engl J Med 2005;353:451–459.
ponents continue to evolve. To meet these challenges, new tests 18. Fiebig EW, Hildebrant CM, Smith RI, Conra s EA, Vos MC, Vermeij
and increased donor restrictions have been implemented, but H, van Rhenen DJ. Clinical significance of bacteriologic screening in
platelet concentrates. Transfusion 2005;45:514–519.
transfusion-transmitted infections persist. Robust pathogen 19. Visconti Mr, Pennington J, Garner SF, Allain JP, Williamson LM. Assess-
inactivation systems for treatment of each blood component ment of removal of human cytomegalovirus from blood components
have been developed and are in various stages of clinical tri- by leucocyte depletion filters using real-time quantitative PCR. Blood
als in the United States and Europe. However, progress has 2004;103:1137–1139.
20. Pennington J, Taylor GP, Sutherland J, et al. Persistence of HTLV-I in
been slow, and problems have been encountered during blood components after leukocyte depletion. Blood 2002;100:677–681.
development. Despite these hurdles, systems for treatment 21. Nichols WG, Price TH, Gooley T, Corey L, Boeckh M. Transfusion-
of platelets and plasma are in clinical use, and as experience transmitted cytomegalovirus infection after receipt of leukoreduced
is gained these technologies may result in a paradigm shift blood products. Blood 2003;101:4195–4200.
for prevention of transfusion-transmitted infections. The 22. Blajchman MA. Bacterial contamination of cellular components: risks,
sources, and control. Vox Sang 2004;87(Suppl 1):S98–S103.
availability of pathogen inactivation systems for all blood 23. Te Boekhorst PA, Beckers EA, Vos MC, Vermeij H, van Rhenen DJ.
components, ultimately, may permit modification of current Clinical significance of bacteriologic screening in platelet concentrates.
testing strategies to take advantage of a stratified, integrated Transfusion 2005;45:514–519.
BLOOD BANKING 24. Munksgaard L, Albjerg L, Lillevang ST, Gahrn-Hansen B, Georgsen 49. Goodrich RP, Yerram NR, Crandall SL, Sowemimo-Coker SO. In vivo
J. Detection of bacterial contamination of platelet components: six survival of platelets subjected to virus inactivation protocols using pso-
years’ experience with the BactT/ALERT system. Transfusion 2004;44: ralen and coumarin photosensitizers. Blood 1995;86(Suppl 1):354A.
1166–1173. 50. Jordan CT, Saakadze N, Newman JL, et al. Photochemical treatment of
25. Claeys H, Logghe F, Vandekerchove B, et al. Four-year experience with platelet concentrates with amotosalen hydrochloride and ultraviolet A
routine bacterial screening of platelet concentrates.Transfus Med light inactivates free and latent cytomegalovirus in a murine transfu-
2003;1(Suppl 1):S326. sion model. Transfusion 2004;44:1159–1165.
26. Larsen CP, Ezligini F, Hermansen NO, Kjeldsen-Kragh J. Six years’ experi- 51. Hei DJ, Grass J, Lin L, Corash L, Cimino G. Elimination of cytokine
ence of using the BactT/ALERT system to screen all platelet concentrates, production in stored platelet concentrate aliquots by photochemical
and additional testing of outdated platelet concentrates to estimate the treatment with psoralen plus ultraviolet A light. Transfusion 1999;
frequency of false-negative results. Vox Sang 2005;88:93–97. 39:239–248.
27. Sapatnekar S, Wood EM, Miller JP, et al. Methicillin-resistant Staphy- 52. Grass JA, Wafa T, Reames A, et al. Prevention of transfusion-associated
lococcus aureus sepsis associated with the transfusion of contaminated graft-versus-host disease by photochemical treatment. Blood 1999;93:
platelets: a case report. Transfusion 2001;41:1426–1430. 3140–3147.
28. Chiu EK, Yuen KY, Lie AK, et al. A prospective study of symptomatic 53. Bruchmiller I, Janetzko K, Bugert P, et al. Polymerase chain reaction
bacteremia following platelet transfusion and of its management. inhibition assay documenting the amotosalen-based photochemical
Transfusion 1994;34:950–954. pathogen inactivation process of platelet concentrates. Transfusion
29. Sullivan MT, McCullough J, Schreiber GB, Wallace EL. Blood col- 2005;45:1464–1472.
lection and transfusion in the United States in 1997. Transfusion 54. Van Rhenen DJ, Vermeij J, Mayaudon V, Hind C, Lin L, Corash L.
2002;42:1253–1260. Functional characteristics of S-59 photochemically treated platelet
30. Srinivasan A. Septic transfusion reactions despite implementation of concentrates derived from buffy coats. Vox Sang 2000;79:206–214.
methods to reduce bacterial contamination. Paper presented at the Advi- 55. Ruane PH, Edrich R, Gampp D, Keil SD, Leonard L, Goodrich RP.
sory Committee on Blood Safety and Availability, Bethesda, Md., 2005. Photochemical inactivation of selected viruses and bacteria in platelet
31. Laperche S, Pillonel J. Residual risk of transfusion-transmitted HIV, concentrates using riboflavin and light. Transfusion 2004;44:877–885.
HCV, and HBV infections in France and impact of NAT. Vox Sang 56. Goodrich RP. The use of riboflavin for the inactivation of pathogens in
2004;87(Suppl 3):S23. blood products. Vox Sang 2000;78(Suppl 2):211–215.
32. Blajchman MA, Beckers EA, Dickmeiss E, Lin L, Moore G, Muylle L. 57. Lippert L, Watson R, Doane S, Reddy H, Goodrich R. Inactivation
Bacterial detection of platelets: current problems and possible resolu- of Plasmodium falciparum by riboflavin and light. Vox Sang 2002;83
tions. Transfus Med Rev 2005;19:259–272. (Suppl 2):163.
33. Benjamin RJ. Bacterial detection in platelet components and the ratio- 58. Perez-Pujol S, Tonda R, Lozano M, et al. Effects of a new pathogen-
nale for pathogen inactivation: a blood center perspective. J Clin Apher reduction technology (Mirasol PRT) on functional aspects of platelet
2005;20:117–122. concentrates. Transfusion 2005;45:911–919.
34. Horowitz B, Wiebe ME, Lippin A, Stryker MH. Inactivation of viruses 59. Li J, de Korte D, Woolum MD, et al. Pathogen reduction of buffy coat
in labile blood derivatives. I. Disruption of lipid-enveloped viruses platelet concentrates using riboflavin and light: comparisons with
by tri(n-butyl)phosphate detergent combinations. Transfusion pathogen-reduction technology-treated apheresis platelet products.
1985;25:516–522. Vox Sang 2004;87:82–90.
35. Heddle NM, Kalma L, Singer J, et al. The role of plasma from platelet 60. Fast LD, DiLeone GR, Li J, Goodrich RP. Mirasol PRT treatment of
concentrates in transfusion reactions. N Engl J Med 1994;331:625–628. human white blood cells prevents the development of xenogeneic graft
36. Ohto H, Anderson KC. Survey of transfusion-associated graft- versus host disease. Transfusion 2005;45(3 Suppl):28A–29A.
versus-host disease in immunocompetent recipients. Transfus Med 61. McCullough J, Vesole DH, Benjamin RJ, et al. Therapeutic efficacy and
Rev 1996;10:31–43. safety of platelets treated with a photochemical process for pathogen
II 37. Lee TH, Parlieroni T, Utter GH, et al. High-level long-term white blood inactivation: the SPRINT trial. Blood 2004;104:1534–1541.
cell microchimerism after transfusion of leukoreduced blood compo- 62. Van Rhenen D, Gulliksson H, Cazenave JP, et al. Transfusion of
394 nents to patients resuscitated after severe traumatic injury. Transfusion pooled buffy coat platelet components prepared with photochemical
2005;45:1280–1290. pathogen inactivation treatment: the euroSPRITE trial. Blood 2003;
38. Ben-Hur E, Moor ACE, Margolis-Nunno H, et al. The photodecontami- 101:2426–2433.
nation of cellular blood components: mechanisms and use of photosen- 63. Snyder E, Raife T, Lin L, et al. Recovery and lifespan of 111 indium-radio-
sitization in transfusion medicine. Transfus Med Rev 1996;10:15–22. labeled platelets treated with pathogen inactivation with amotosalen
39. Wollowitz S. Targeting DNA and RNA in pathogens: mode of action of HCl (S-59) and ultraviolet A light. Transfusion 2004;44:1732–1740.
amotoslaen HCl. Transfus Med Hemother 2004;31(Suppl 1):11–16. 64. Snyder E, McCullough J, Slichter SJ, et al. Clinical safety of plate-
40. Kumar V, Lockerbie O, Keil SD, et al. Riboflavin and UV-light based lets photochemically treated with amotosalen HCl and ultravio-
pathogen reduction: extent and consequence of DNA damage at the let A light for pathogen inactivation: the SPRINT trial. Transfusion
molecular level. Photochem Photobiol 2004;80:15–21. 2005;45:1864–1875.
41. Lin L, Dikeman R, Molini B, et al. Photochemical treatment of platelet 65. Osselaer JC, Doyen C, Sonet A, et al. Routine use of platelet compo-
concentrates with amotosalen and UVA inactivates a broad spectrum nents prepared with photochemical treatment (INTERCEPT plate-
of pathogenic bacteria. Transfusion 2004;44:1496–1504. lets): impact on clinical outcomes and costs. Blood 2004;104:3629.
42. Van Voorhis WC, Barrett LK, Eastman RT, Alfonso R, Depuis K. 66. Slichter SJ, Raife TJ, Davis K, et al. Platelets photochemically treated
Trypanosoma cruzi inactivation in human platelet concentrates and with amotosalen HCl and ultraviolet A light correct prolonged
plasma by a psoralen (amotosalen HCl) and long-wavelength UV. bleeding times in thrombocytopenic patients. Transfusion 2005;45:
Antimicrob Agents Chemother 2003;47:475–479. 1864–1875.
43. Eastman RT, Barrett LK, Dupuis K, Buckner FS, Van Voorhis WC. 67. Lin L, Cook DN, Wiesehahn GP, et al. Photochemical inactivation of
Leishmania inactivation in human pheresis platelets by a psoralen viruses and bacteria in platelet concentrates by use of a novel psoralen
(amotosalen HCl) and long-wavelength ultraviolet irradiation. Trans- and long-wavelength ultraviolet light. Transfusion 1997;37:423–435.
fusion 2005;45:1459–1463. 68. Miller AB, Hoogstraten B, Staquet M, Winkler A. Reporting results of
44. Lin L, Hanson CV, Alter HJ, et al. Inactivation of viruses in platelet cancer treatment. Cancer 1981;47:207–214.
concentrates by photochemical treatment with amotosalen and long- 69. Rebulla P, Finazzi G, Marangoni F, et al. The threshold for prophylac-
wavelength ultraviolet light. Transfusion 2005;45:580–590. tic platelet transfusions in adults with acute myelogenous leukemia. N
45. Grass JA, Hei DJ, Metchette K, et al. Inactivation of leukocytes in plate- Engl J Med 1997;337:1870–1875.
let concentrates by psoralen plus UVA. Blood 1998;91:2180–2188. 70. Heddle NM, Cook RJ, Webert KE, Sigouin C, Rebulla P. Methodologi-
46. Lin L, Wiesehahn GP, Morel PA, Corash L. Use of 8-methoxypsoralen cal issues in the use of bleeding as an outcome in transfusion medicine
and long wavelength ultraviolet radiation for decontamination of studies. Transfusion 2003;43:742–752.
platelet concentrates. Blood 1989;74:517–525. 71. Corash L. How much do we know about the platelet transfusion
47. Alter HJ, Morel PA, Dorman BP, et al. Photochemical decontamination threshold? Transfusion 2003;43:691–693.
of blood components containing hepatitis B and non-A, non-B virus. 72. Heddle NM, Cook RJ, Rebulla P, Slichter SJ, Sigouin C, Murphy MF.
Lancet 1988;2:1446–1450. Bleeding in patients with acute leukemia. Vox Sang 2002;83(Suppl 2):8.
48. Goodrich RP, Yerram NR, Tay GB, et al. Selective inactivation of viruses 73. Slichter SJ, Davis K, Enright H, et al. Factors affecting posttransfusion
in the presence of human platelets: UV sensitization with psoralen platelet increments, platelet refractoriness, and platelet transfusion
derivatives. Proc Natl Acad Sci USA 1994;91:5552–5556. intervals in thrombocytopenic patients. Blood 2005;105:4106–4114.
VIRUS-SAFE PRODUCTS
74. Murphy S, Snyder E, Cable R, et al. Platelet dose consistency and its 96. Wages D, Hambleton J, Viele M, et al. RBCs treated with Helinx
effect on the number of platelet transfusions for support of throm- pathogen inactivation have recovery and half-life comparable to con-
bocytopenia: an analysis of the SPRINT trial of platelets photochemi- ventional RBCS in a randomized crossover trial. Hematol J 2002;3
cally treated with amotosalen HCl and ultraviolet A light. Transfusion (Suppl 1):171.
2006;46:24–33. 97. Benjamin RJ, McCullough J, Mintz PD, et al. Therapeutic efficacy
75. Janetzko K, Cazenave JP, Kluter H, et al. Therapeutic efficacy and safety and safety of red blood cells treated with a chemical process (S-303)
of photochemically treated apheresis platelets processed with an opti- for pathogen inactivation: a phase III clinical trial in cardiac surgery
mized integrated set. Transfusion 2005;45:1443–1452. patients. Transfusion 2005;45:1739–1749.
76. Snyder E, McCullough J, Slichter SJ, et al. Clinical safety of plate- 98. Conlan MG, Vichinsky E, Snyder E, et al. S-303 pathogen inactivated
lets photochemically treated with amotosalen HCl and ultravio- red blood cells in patients with hemoglobinopathies participating in
let A light for pathogen inactivation: the SPRINT trial. Transfusion chronic RBC programs: preliminary safety and efficacy results. Vox
2005;45:1864–1875. Sang 2005;89(Suppl 1):121.
77. Lin L, Conlan MG, Tessman J, Cimino G, Porter S. Amotosalen inter- 99. Conlan MG, Lin J, Stassinopoulos A. Investigation of immunoreac-
actions with platelet and plasma components: absence of neoanti- tivity observed after transfusion of S-303 RBCs in 2 phase III clini-
gen formation after photochemical treatment. Transfusion 2005;45: cal trials in support of acute or chronic anemia. Transfusion 2005;45
1610–1620. (Suppl 3):29A.
78. Osselaer JC, Bueno JL, Messe N, Jacquet M, Castro E, Flament J. 100. Arndt PA, Garratty G. A retrospective analysis of the value of mono-
Prospective active hemovigilance plan for INTERCEPT platelets in cyte monolayer assay results for predicting the clinical significance of
Europe: a status report. Vox Sang 2005;89:137. blood group alloantibodies. Transfusion 2004;44:1273–1281.
79. AuBuchon JP, Herschel L, Roger J, et al. Efficacy of apheresis platelets 101. Horowitz B, Bonomo R, Prince AM, Chin SN, Brotman B, Shulman
treated with riboflavin and ultraviolet light for pathogen reduction. RW. Solvent/detergent treated plasma: a virus-inactivated substitute
Transfusion 2005;45:1335–1341. for fresh frozen plasma. Blood 1992;79:826–831.
80. Rywkin S, Ben-Hur E, Reid ME, Oyen R, Ralph H, Horowitz B. Selective 102. Roberts P. Resistance of vaccinia virus to inactivation by solvent/deter-
protection against IgG binding to red cells treated with phthalocyanines gent treatment of blood products. Biologicals 2000;28:29–32.
and red light for virus inactivation. Transfusion 1995;35:414–420. 103. Luban NL. Human parvovirus: implications for transfusion medicine.
81. Ben-Hur E, Rywkin S, Rosenthal I, Geacintov NE, Horowitz B. Virus Transfusion 1994;34:821–827.
inactivation in red cell concentrates by photosensitization with phthalo- 104. Horowitz MS, Pehta JC. SD plasma in TTP and coagulation fac-
cyanines: protection of red cells but not of vesicular stomatitis virus with tor deficiencies for which no concentrates are available. Vox Sang
a water-soluble analogue of vitamin E. Transfusion 1995; 35:401–406. 1998;74(Suppl 1):231–235.
82. Wagner SJ, Skirpchenko A, Robinette D, Mallory DA, Cincotta L. Pres- 105. Haubelt H, Blome M, Kiessling AH, et al. Effects of solvent/detergent-
ervation of red cell properties after virucidal phototreatment with treated plasma and fresh frozen plasma on haemostasis and fibrinoly-
dimethylmethylene blue. Transfusion 1998;38:729–737. sis in complex coagulopathy following open-heart surgery. Vox Sang
83. Besselink GA, van Engelenburg FA, Korsten HG, et al. The band III 2002;82:9–14.
ligand dipyridamole protects human RBCs during photodynamic 106. Williamson LM, Llewelyn CA, Fisher NC, et al. A randomized trial
treatment while extracellular virus inactivation is not affected. Trans- of solvent/detergent-treated and standard fresh-frozen plasma in the
fusion 2002;42:728–733. coagulopathy of liver disease and liver transplantation. Transfusion
84. Wagner SJ, Skripchenko A, Cincotta L, Thompson-Montgomery 1999;39:1227–1234.
D, Awatefe H. Use of a flexible thiopyrylium photosensitizer and 107. Baudoux E, Margraff U, Coenen A, et al. Hemovigilance: clinical tol-
competitive inhibitor for pathogen reduction of viruses and bacte- erance of solvent-detergent treated plasma. Vox Sang 1998;74(Suppl
ria with retention of red cell storage properties. Transfusion 2005; 1):237–239.
45:752–760. 108. Beeck H, Hellstern O. In vitro characterization of solvent/detergent-
85. Fast LD, DiLeone G, Edson CM, Purmal A. PEN 110 treatment treated human plasma and quarantine fresh frozen plasma. Vox Sang
27
functionally inactivates the PBMNCs present in RBC units: com- 1998;74(Suppl 1):219–223.
parison to the effects of exposure to gamma irradiation. Transfusion 109. Mast AE, Stadanlick JE, Lockett JM, Dietzen DJ. Solvent/detergent- 395
2002;42:1318–1325. treated plasma has decreased antitrypsin activity and absent antiplas-
86. Jayarama V, Lazo A, Marcello J, et al. Inactine PEN 110 inactivates min activity. Blood 1999;94:3922–3927.
cell-free and cell-associated CMV in red cell concentrates. Transfusion 110. Beach KJ. Reported Serious Adverse Events in Liver Transplant
2003;43(Suppl 9):8A. Patients. V.I. Technologies Inc. Available at http://www.fda.gov/med-
87. Ohagen A, Gibaja V, Aytay S, Horrigan J, Lunderville D, Lazo A. Inacti- watch/safety/2000/plassd.pdf. Accessed March 28, 2006.
vation of HIV in blood. Transfusion 2002;42:1308–1317. 111. Yarranton H, Cohen H, Pavord SR, Benjamin S, Hagger D, Machin SJ.
88. Edson CM, Purmal A, Brown F, Valeri CR, Budowsky E, Chapman JR. Venous thromboembolism associated with management of acute throm-
Viral inactivation of red blood cell concentrates by Inactine: mecha- botic thrombocytopenic purpura. Br J Haematol 2003;121:778–785.
nism of action and lack of effect on red cell physiology. Transfusion 112. Flamholz R, Jeon HR, Baron JM, Baron BW. Study of three patients
1999;39(Suppl 1):108S. with thrombotic thrombocytopenic purpura exchanged with solvent/
89. AuBuchon JP, Pickard CA, Herschel LH, et al. Production of pathogen- detergent-treated plasma: is its decreased protein S activity clinically
inactivated red cell concentrates using PEN 110 chemistry: a phase I related to their development of deep venous thrombosis? J Clin Apher
clinical study. Transfusion 2002;42:146–152. 2000;15:169–172.
90. Snyder E, Mintz P, Burks S, et al. Pathogen inactivated red blood cells 113. Mohr H, Lambrecht B, Knueyer-Hopf J. Virus inactivated single-donor
using Inactine technology demonstrate 24-hour posttransfusion fresh plasma preparations. Infusionsther Transfusionsmed 1992;19:79–83.
recovery equal to untreated red cells after 42 days of storage. Blood 114. Garwood M, Cardigan RA, Drummond O, et al. The effect of methy-
2001;98(Suppl 1):109A. lene blue photoinactivation and methylene blue removal on the qual-
91. Cook D, Stassinopoulos A, Merritt J, et al. Inactivation of pathogens ity of fresh-frozen plasma. Transfusion 2003;43:1238–1247.
in packed red blood cell (PRBC) concentrates using S-303. Blood 115. Aznar JA, Bonanad S, Montoro JM, et al. Influence of methylene blue pho-
1997;90(Suppl 1):409A. toinactivation treatment on coagulation factors from fresh frozen plasma,
92. Cook D, Stassinopoulos A, Wollowitz S, et al. In vivo analysis of packed cryoprecipitates, and cryosupernatants. Vox Sang 2000; 79:156–160.
red blood cells treated with S-303 to inactivate pathogens. Blood 116. Simonsen AC, Sorensen H. Clinical tolerance of methylene blue virus-
1998;92(Suppl 1):503A. inactivated plasma. A randomized crossover trial in 12 healthy volun-
93. Ciaravino V, McCullough T, Woods N, Sullivan T. Absence of carci- teers. Vox Sang 1999;77:210–217.
nogenicity in a 26-week intravenous study with S-303-treated mouse 117. Williamson LM, Cardigan R, Prowse CV. Methylene blue-treated
red blood cells in C57BL/6TAC-TRP53TML heterozygote mice. Blood fresh-frozen plasma: what is the contribution to blood safety? Transfu-
Bank Transfus Med 2003;1(Suppl 1):S365. sion 2003;43:1322–1329.
94. Greenwalt TJ, Hambleton J, Wages D, et al. Viability of red blood 118. Wieding JU, Rathberger J, Zenjer D. Prospective, randomized trial
cells treated with a novel pathogen inactivation system. Transfusion and controlled study on solvent detergent versus methylene blue virus
1999;39(Suppl 1):109S. inactivated plasma. Transfusion 1999;39:23S.
95. Hambleton J, Greenwalt T, Viele M, et al. Posttransfusion recovery 119. Alvarez-Larran A, Del Rio J, Ramirez C, et al. Methylene blue-pho-
after multiple exposures to red blood cell concentrates (RBCS) treated toinactivated plasma vs fresh frozen plasma as replacement fluid for
with a novel pathogen inactivation (PI) process. Blood 1999;94(Suppl plasma exchange in thrombotic thrombocytopenic purpura. Vox Sang
1):376A. 2004;86:246–251.
BLOOD BANKING 120. de la Rubria J, Arriaga F, Linares D, et al. Role of methylene blue- ultraviolet A light in the presence of quenchers. Transfusion 1992;32:
treated or fresh-frozen plasma in the response to plasma exchange in 541–547.
patients with thrombotic thrombocytopenic purpura. Br J Haematol 138. Benade LE, Shumaker J, Xu Y, Chen X, Dodd RY. Inactivation of free
2001;114:721–723. and cell-associated human immunodeficiency virus in platelet suspen-
121. Atance R, Pereira A, Ramirez B. Transfusing methylene blue-photoin- sions by aminomethyltrimethylpsoralen and ultraviolet light. Transfu-
activated plasma instead of FFP is associated with an increased demand sion 1994;34:680–684.
for plasma and cryoprecipitate. Transfusion 2001;41:1548–1552. 139. Margolis-Nunno M, Williams B, Rywkin S, Horowitz B. Photochemi-
122. Alfonso R, Lin C, Dupuis K, et al. Inactivation of viruses with pres- cal virus sterilization in platelet concentrates with psoralen derivatives.
ervation of coagulation function in fresh frozen plasma. Blood Thromb Haemost 1991;65:1162.
1996;88(Suppl 1):526A. 140. Yerram N, Forster P, Goodrich T, et al. Comparison of virucidal prop-
123. Singh Y, Sawyer L, Pinkoski L, et al. Photochemical treatment of plasma erties of brominated psoralen with 8-methoxy psoralen (8-MOP)
with amotosalen and UVA light inactivates pathogens while retaining and aminomethyl trimethyl psoralen (AMT) in platelet concentrates.
coagulation function. Transfusion 2006;46:1168–1177. Blood 1993;82(Suppl 1):402A.
124. Hervig T, Cazenave JP, Schlenke P, et al. INTERCEPT plasma: process 141. Rai S, Kasturi C, Grayzar J, et al. Dramatic improvements in viral inac-
validation studies in three European blood centers. Hematologica tivation with brominated psoralens, napthalenes, and anthracenes.
2005;90:246. Photochem Photobiol 1993;58:59–65.
125. Wages D, Smith D, Walsh J, et al. Transfusion of therapeutic doses 142. Prodouz KN, Fratantoni JC, Boone EJ, Bonner RF. Use of laser-UV for
of virally inactivated fresh frozen plasma in healthy subjects. Blood inactivation of virus in blood products. Blood 1987;70:589–592.
1997;90(Suppl 1):409A. 143. Prodouz KN, Lytle CD, Keville EA, Budacz AP, Vargo S, Fratantoni JC.
126. Wages D, Radu-Radurescu L, Adams M, et al. Quantitative analysis of Inhibition by albumin of merocyanine 540-mediated photosensitiza-
coagulation factors in response to transfusion of S-59 photochemi- tion of platelets and viruses. Transfusion 1991;31:415–422.
cally treated fresh frozen plasma (S-59 FFP) and standard FFP. Blood 144. Klein-Struckmeier A, Mohr H. Virus inactivation by methylene blue
1998;92(Suppl 1):503A. light in thrombocyte concentrates. Vox Sang 1994;67(Suppl 2):36.
127. Hambleton J, Wages D, Radu-Radulescu L, et al. Pharmacokinetic 145. Horowitz B, Rywkin S, Margolis-Nunno H, et al. Inactivation of viruses
study of FFP photochemically treated with amotosalen (S-59) and UV in red cell and platelet concentrates with aluminum phthalocyanine
light compared to FFP in healthy volunteers anticoagulated with war- (AlPc) sulfonates. Blood Cells 1992;18:141–150.
farin. Transfusion 2002;42:1302–1307. 146. Matthews JL, Sogandres-Bernal F, Judy M, et al. Inactivation of viruses
128. Wages D, Bass N, Keefe E, et al. Treatment of acquired coagulopathy by with photoactive compounds. Blood Cells 1992;18:75–89.
transfusion of fresh frozen plasma (FFP) prepared using a novel, single 147. North J, Neyndorff H, King D, Levy JG. Viral inactivation in blood
unit photochemical. Blood 1999;94(Suppl 1):247A. and red cell concentrates with benzoporphyrin derivative. Blood Cells
129. de Alarcon P, Benjamin R, Dugdale M, et al. Fresh frozen plasma pre- 1992;18:129–140.
pared with amotosalen HCl (S-59) photochemical pathogen inac- 148. Sieber F, Krueger GJ, O’Brien JM, Schober SL, Sensenbrenner LL,
tivation (PCT-FFP): transfusion of patients with congenital factor Sharkis SJ. Inactivation of friend erythroleukemia virus and friend
deficiencies. Transfusion 2005;45:1362–1372. virus-transformed cells by merocyanine 540-mediated photosensitiza-
130. Mintz PD, Bass NM, Petz LD, et al. Photochemically treated fresh fro- tion. Blood 1989;73:345–350.
zen plasma for transfusion of patients with acquired coagulopathy of 149. O’Brien JM, Gaffney DK, Wang TP, Sieber F. Merocyanine 540-sensi-
liver disease. Blood 2006;107:3753–3760. tized photoinactivation of enveloped viruses in blood products: site
131. Mintz PD, Neff A, MacKenzie M, et al. Therapeutic plasma exchange and mechanism of phototoxicity. Blood 1992;80:277–285.
(TPE) for thrombotic thrombocytopenic purpura (TTP) using plasma 150. Smith OM, Dolan SA, Dvorak JA, Wellems TE, Sieber F. Merocyanine
prepared with photochemical treatment (INTERCEPT plasma). Blood 540-sensitized photoinactivation of human erythrocytes parasitized
2004;104:239A. by Plasmodium falciparum. Blood 1992;80:21–24.
II 132. Londe H, Damonte P, Corash L, Lin L. Inactivation of human cyto- 151. Wagner SJ, Storry JR, Mallory DA, Stromberg RR, Benade LE, Fried-
megalovirus with psoralen and UVA in human platelet concentrates. man LI. Red cell alterations associated with virucidal methylene blue
396 Blood 1995;86(Suppl 1):544A. phototreatment. Transfusion 1993;33:30–36.
133. Lin L, Londe H, Hanson CV, et al. Photochemical inactivation of cell- 152. Wagner SJ, Robinette D, Storry J, Chen XY, Shumaker J, Benade L. Dif-
associated human immunodeficiency virus in platelet concentrates. ferential sensitivities of viruses in red cell suspensions to methylene
Blood 1993;82:292–297. blue photosensitization. Transfusion 1994;34:521–526.
134. Eble BE, Corash L. Photochemical inactivation of duck hepatitis B 153. Horowitz B, Williams B, Rywkin S, et al. Inactivation of viruses in
virus in human platelet concentrates: a model of surrogate human blood with aluminum phthalocyanine derivatives. Transfusion 1991;
hepatitis B virus infectivity. Transfusion 1996;36:406–418. 31:102–108.
135. Lin L, Londe H, Janda JM, et al. Photochemical inactivation of pathogenic 154. Yerram N, Platz MS, Forster P, Goodrich T, Goodrich R. Selective viral
bacteria in human platelet concentrates. Blood 1994;83:2698–2706. inactivation in RBC, platelets, and plasma using a novel psoralen deriva-
136. Dodd RY, Moroff G, Wagner S, et al. Inactivation of viruses in plate- tive plus ultraviolet A (UVA) light. Transfusion 1993;33(Suppl 9):50S.
let suspensions that retain their in vitro characteristics: comparison 155. Lavie G, Mazur Y, Lavie D, et al. Hypericin as an inactivator of infec-
of psoralen-ultraviolet A and merocyanine 540-visible light methods. tious viruses in blood components. Transfusion 1995;35:392–400.
Transfusion 1991;31:483–490. 156. Zhang QX, Edson C, Budowsky E, Purmal A. Inactine: a method
137. Margolis-Nunno H, Williams B, Rywkin S, Geacintov N, Horow- for viral inactivation in red blood cell concentrates. Transfusion
itz B. Virus sterilization in platelet concentrates with psoralen and 1998;38(Suppl 10):75S.
Chapter 28
Irradiated Products
Naomi L. C. Luban ● Edward C. C. Wong
Transfusion-associated graft-versus-host disease (TA- Three prerequisites for the development of GVHD in the
GVHD), an often fatal alloimmune complication mediated transplant setting have been proposed: (1) differences in
by donor T cells in the blood component, was first reported histocompatibility between recipient and donor, (2) pres-
in the 1960s in individuals with hematologic malignancies ence of immunocompetent cells in the graft, and (3) inability
and in infants with congenital immunodeficiencies who of the host to reject the immunocompetent cells. A similar
developed “runting disease” after blood transfusion.1 set of circumstances underlie TA-GVHD. First, in almost
Since these early reports, the spectrum of individuals all cases, leukocytes in homologous blood components are
at risk has expanded, the pathogenesis has been partially mismatched with the transfusion recipient at HLA and
elucidated, and preventive strategies have been established.2–4 minor histocompatibility loci. Second, transfused donor
Despite these advances, many unanswered questions remain. leukocytes are immunologically functional, except in cir-
For example, there are no adequate estimates of prevalence. cumstances in which they have been specifically inactivated
In Japan, an estimated annual incidence of one TA-GVHD (e.g., following gamma-irradiation). And third, the com-
case per 212 transfusions has been calculated based on monality between many of the patients at risk for TA-GVHD
homozygosity for one-way human leukocyte antigen (HLA) is their inability to reject transfused leukocytes, either due
haplotype sharing, the use of familial donors, and the use of to immune immaturity (low-birth-weight neonates) or iat-
fresh rather than stored red blood cells (RBCs).5 In the United rogenic immunosuppression (transplant recipients). The
Kingdom and Canada, in contrast, risk estimates are much development of TA-GVHD in additional clinical settings
lower as determined by adverse transfusion outcome report- illustrates the importance of these factors. For example, most 28
ing. The Serious Hazards of Transfusion (SHOT) study in immunocompetent recipients destroy transfused donor T
the United Kingdom has identified only 13 cases from 1996 cells through lymphocytolysis and therefore do not develop 397
to present, although all were fatal.6 In Canada, mathematical TA-GVHD. However, transfusion from an HLA homozy-
modeling by Kleinman and colleagues estimated the magni- gous donor to an HLA heterozygous, but immunocompe-
tude of risk to be between 1:12,893 and 1:21,157, but based tent, recipient who shares one HLA haplotype may result in
on cases reported and other factors a true risk of less than 1 failure of recognition of the donor cells as foreign.9,10 This
per million units transfused is cited.7 No incidence figures so-called “one-way HLA match” is often responsible for the
are available for the United States, and it is likely that under- development of TA-GVHD in immunocompetent recipients,
recognition and under-reporting are common. Additionally, such as might occur in familial (directed) blood transfusions,
while the pathogenesis of TA-GVHD partially resembles that in populations with limited HLA diversity, and when HLA-
of GVHD in the setting of hematopoietic stem cell transplan- matched platelet transfusions are administered. In these cases,
tation, further, more detailed investigations of TA-GVHD the non-self HLA or minor histocompatibility antigens of the
have been hindered by the lack of well-established animal host, or both, stimulate clonal expansion of donor T cells and
models. Instead, investigators typically utilize the classic par- the induction of an inflammatory response that is ultimately
ent-to-F1 hybrid mouse model of GVHD, which may or may responsible for clinical manifestations of TA-GVHD.
not provide the correct tool for TA-GVHD investigations.8 Studies of the pathogenesis of TA-GVHD, both in vitro
Given the almost uniform lethality of TA-GVHD, experiments as well as animal investigations, have begun
extensive clinical investigations have fortunately led to the to define roles for donor leukocyte subsets in the process.
development of pretransfusion irradiation methods that Among leukocytes, donor T cells play the most prominent
uniformly prevent TA-GVHD. The technical and regula- role in disease pathogenesis. CD4+ T cells (sometimes known
tory aspects of these methodologies are addressed in detail as T-helper cells) can be functionally divided into Th1 and
in this chapter. Since irradiation devices are not available Th2 subsets. Th1 CD4 T cells secrete interleukin-2 (IL-2),
in all transfusion settings, attention has recently focused while Th2 CD4 T cells produce IL-4, IL-5, IL-6, IL-10, IL-13,
on alternative approaches to prevention. Data are accumu- and lesser amounts of tumor necrosis factor–α (TNFα). Th1
lating to suggest that filtration and nucleic acid–targeted and Th2 both produce IL-3 and granulocyte-macrophage
pathogen reduction technologies may remove or inactivate colony-stimulating factor. The type 1 cell is proinflammatory
donor leukocytes, respectively, to the extent that TA-GVHD and induces cell-mediated immunity, whereas the type 2 cell
is also prevented, although further study and validation are is considered anti-inflammatory. Differentiation toward Th1
required. or Th2 is a complicated process that involves early exposure
BLOOD BANKING
to IL-4 or IL-12, the type of antigen-presenting cells, costim- stimulates gut-associated macrophages and lymphocytes,
ulating molecules, and the presence of macrophages and stimulates keratocytes and dermal fibroblasts, and further
their unique cytokines. promotes the inflammatory response and end-organ damage
On the basis of their work in mice, Ferrera and colleagues that are classic hallmarks of the disorder.
proposed a three-step process for the development of acute Whether the development of TA-GVHD is dependent on
GVHD in the transplant setting, including the involvement of similar mechanisms remains an open question. To begin to
T-cell subsets. In this model (Fig. 28–1), host tissue damaged address this question, the importance of CD4 and CD8 T
through irradiation or chemotherapy secretes TNFα and cells in the pathogenesis of TA-GVHD has been studied by
IL-1, which enhance recognition of host histocompatibility Fast and coworkers12 in a mouse model and by Nishimura
antigens by donor T cells. Donor T-cell activation results in and associates13 in a patient with TA-GVHD. Their find-
proliferation of Th1 T cells and secretion of IL-2 and TNFα, ings are further supported by clinical correlation in patients
which in turn further activate T cells and induce cytotoxic with human immunodeficiency virus (HIV) infection and
T- lymphocyte (CTL) and natural killer (NK) responses. acquired immunodeficiency syndrome (AIDS).14 In the
Additional studies have shown that donor and residual mouse, depletion of CD4+ cells increases the number of
host phagocytes are stimulated to produce IL-1, TNFα, and donor cells needed to induce TA-GVHD, whereas depletion
the free radical nitric oxide (NO), which has further delete- of CD8+ or NK cells, or both, decreases the number of donor
rious effects on host tissues. In addition, NO up-regulates cells needed to induce the disorder. In HIV and AIDS, there
alloreactivity and mediates the cytotoxic function of macro- has been only one report of TA-GVHD15 despite widespread
phages.11 A secondary triggering signal initiates a subsequent use of supportive transfusion and profound immunosup-
stage in the disease process wherein lipopolysaccharide (LPS) pression. This observation is consistent with the concept that
Radiation,
chemotherapy
Step 1
Tissue damage
Inflammation IL-1, TNF-␣
CTL
NK
Donor/host
Mononuclear
phagocytes
Endotoxin
(LPS)
Step 3 sTNFR
IL-1ra
TNF-␣
Inflammatory cytokines
Target organ damage IL-1
Nitric
GVHD target oxide
tissues
Figure 28–1 Proposed interactions between T-cell cytokines and mononuclear phagocyte–derived cytokines during graft-versus-host disease
(GVHD). IFN, interferon; IL-1, interleukin-1; LPS, lipopolysaccharide; sTNFR, soluble tumor necrosis factor receptor; TNFα, tumor necrosis factor–α.
IRRADIATED PRODUCTS
early depletion of CD4 may well protect against establish- cell surface molecules like CD31, CD47, and CD200, a phe-
ment of TA-GVHD. Additionally, expansion and activation nomenon yet to be confirmed in human studies.
of NK and CD8+ lymphocytes against HIV-infected CD4 T
cells may limit the development of the GVHD process.
Since donor leukocytes must persist to cause TA-GVHD, CLINICAL MANIFESTATIONS OF TA-GVHD
other investigators have evaluated clearance of leukocytes
after transfusion.16 In 1996, Busch and colleagues17 demon- Fever, anorexia, nausea, vomiting, and diarrhea typically
strated a thousand-fold expansion of donor lymphocytes in develop 7 to 10 days post-transfusion. Skin manifestations of
the circulation of otherwise healthy recipients 3 to 5 days TA-GVHD are variably severe and begin as an erythematous
after transfusion for elective orthopedic procedures. Within 2 maculopapular eruption that may proceed to erythroderma
weeks, the allogeneic cells were cleared. In contrast, in a study with bullae and frank desquamation. Gastrointestinal bleed-
of adult trauma victims who received large numbers (4 to 18 ing is commonly seen, usually as bloody diarrhea. Hepatic
units) of fresh-packed RBCs, 8 of 10 had long-term persis- dysfunction with transaminitis and hyperbilirubinemia,
tence of donor leukocytes with confirmed microchimerism including a progressively increasing direct fraction, is also
(MC). Two of the eight had persistence of MC when studied seen. In contrast to acute GVHD in the setting of allogeneic
as long as 1.5 years after transfusion.18 These studies have transplantation, TA-GVHD leads to severe pancytopenia
been further expanded in additional patients transfused for because donor T cells attack marrow hematopoietic cells.
acute trauma who received nonleukoreduced17 or leukore- This often results in the death of the patient. The diagnosis
duced19 products. That MC in some cases persisted for years18 typically is made at postmortem examination and is based
recently has been confirmed and attributed to a single donor on pathognomonic histopathologic findings of donor lym-
source from which leukocytes expanded over time.18,20 These phocyte infiltration and expansion in skin, lymph nodes,
studies suggest that transfusion during a period of diminished liver, and the gastrointestinal tract.28
lymphocyte responsiveness may result in a microchimeric
state with persistence of functional donor leukocytes, pos-
sibly with long-term consequences. DIAGNOSIS OF TA-GVHD
Wang-Rodriquez and associates21 studied post-transfusion
immune modulation in 14 premature infants and identified Clinical suspicion may warrant a skin biopsy, which often
two of six female infants, transfused with nonleukodepleted reveals vacuolization of the epidermal basal cell layer,
RBCs, who experienced transient MC detected by Y chromo- dermal-epithelial layer separation, and formation of bullae.
some PCR amplification. In both, these cells were cleared by Other findings include mononuclear cell migration into
2 weeks after transfusion. An additional three infants who the epidermis, hyperkeratosis, and dyskeratosis. Liver biop-
received leukodepleted RBCs also had transient MC. sies may reveal eosinophilic infiltration and degeneration of
Vietor and coworkers22 studied 9 surviving recipients of small bile ducts, peripheral inflammation, and lymphocyte
intrauterine transfusion whose donors were still available infiltration. The bone marrow findings are classically those 28
for testing. Using fluorescence in situ hybridization, PCR of “empty marrow,” with pancytopenia, fibrosis, and some
of Y chromosome–specific sequences, and assays for the lymphocytic infiltration. 399
frequencies of CTL and T-helper lymphocyte precursors, Definitive confirmation of TA-GVHD is more com-
they detected true MC in 6 of 7 young adults studied 20 plicated. Several methods have been utilized to identify
years after transfusion. Reed and colleagues23 have developed lymphocytes of foreign origin in the patient’s circulation or
sequence-specific amplification of DRB1, which permitted in affected tissues. Serologic HLA typing, DNA-based HLA
identification of minor chimeric populations at the 0.01% class II typing, karyotype analysis, restriction fragment length
level. The establishment of stable MC and identification of polymorphism analysis using probes from both HLA and
its biological consequences are critical for pediatric patients non-HLA regions, and genetic fingerprinting have all been
who are expected to live to adulthood and may well be stable used.29–33 Fibroblast and buccal mucosal cells of the recipi-
transfusion-induced chimeras, an intriguing and at the same ent are often needed, as the lymphocytolysis accompanying
time potentially worrisome occurrence. The persistence of the disorder prohibits standardized serologic HLA typing
MC may predispose to autoimmune disease,24–26 chronic using recipient lymphocytes. Alternatively, parental or famil-
GVHD, and recurrent abortion27 and may serve as an allogeneic ial specimens may be necessary to deduce a recipient’s HLA
stimulus of latent viral reactivation in the recipient. type.34 Donor lymphocytes in attached remaining segments
In a mouse system, allogeneic male donor WBCs derived from suspected blood products often require PCR amplifi-
from C57BL/6 mice persisted in female BALB/c recipi- cation and sequence-specific oligonucleotide probe method-
ent mice and their survival was unaffected by irradiation. ologies to provide confirmation of donor cell origin.35
Although the experimental model could not differentiate
nonproliferative from proliferative WBC, these data set
the stage for additional studies on MC and post-transfu- GROUPS AT RISK FOR TA-GVHD
sion WBC kinetics. Fast and colleagues studied the effects
of different forms of irradiation on murine splenocytes to Patients in whom TA-GVHD may develop have been described
better understand inhibition of both in vitro and in vivo in a number of reviews,2–4,10,11,36,37 and this remains an area
proliferative responses. Among their findings was that the of active discussion. For example, on the basis of two reports
genotype of the donor-recipient pair regulates recipient of TA-GVHD in older infants with severe combined immu-
alloantibody formation, a conclusion suggested by studies nodeficiency,18,35 it has been recommended that infants well
of prolonged MC in humans.19 They also showed that the past the neonatal age group be considered at risk.38,39 Because
accelerated elimination of donor cells from recipient lym- of the lack of adequate animal models and laboratory tests to
phoid compartments might be due to altered expression of identify individual TA-GVHD risk, many reports stratify the
BLOOD BANKING
need for irradiation using such terms as clearly indicated or damage to nuclear DNA either directly or through genera-
probably indicated.3,36,37 In reality, the spectrum of individu- tion of ions and free radicals that have damaging biological
als at risk is likely to grow as intensive immunomodulatory actions. Irradiation prevents post-transfusion donor T-cell
therapies expand beyond oncologic disease and transplanta- proliferation in response to host antigen-presenting cells,
tion (Table 28–1). which, in turn, abrogates GVHD.43–46 Two types of ionizing
radiation—gamma rays and x-rays—are equivalent in inacti-
vating T cells in blood components at a given absorbed dose.
THERAPY FOR TA-GVHD Gamma rays originate from the radioactive decay process
within the atomic nucleus of cesium 137 (137Cs) or cobalt 60
The rapid progression of clinical manifestations of the dis- (60Co). Freestanding blood bank gamma-irradiators, which
order and similarity of clinical presentation to that of viral, are the predominant instruments for blood component
drug-related, or disease-related enteropathy and hepatocellu- irradiation, use either of these two isotopes as the irradia-
lar damage contribute to the high mortality rates. Anecdotal tion source.
success has been reported with immunosuppressive treat- In contrast, x-rays are generated from the interac-
ment including cytoxan, antithymocyte globulin, and high- tion of a beam of electrons with a metallic surface. Linear
dose corticosteroids. Recently, the serum protease inhibitor accelerators typically used to generate x-rays for clinical
nafamostat mesylate has been used with some success.40,41 radiation therapy (teletherapy) may serve as an irradiation
The effectiveness of hematopoietic stem cell transplantation source for blood and blood components. The FDA has also
is mixed.35,42 approved the use of a freestanding x-ray machine (Rad-
Source RS3000, Coral Springs, Fla.) for irradiation of blood
components.
PREVENTION OF TA-GVHD Instrumentation for Irradiation
The basic operating principles and configurations of a free-
Irradiation
standing irradiator with either a 137Cs source or a linear
Among methodologies (e.g., irradiation, photoinactivation, accelerator are shown schematically in Figure 28–2. With a
pegylation, and ultraviolet [UV] light) that can be used to pre- freestanding 137Cs irradiator, blood components are contained
vent TA-GVHD, only irradiation of whole blood and cellular within a metal canister that is positioned on a rotating turn-
components is currently accepted by the U.S. Food and Drug table. Continuous rotation allows the gamma rays, originating
Administration (FDA). Exposure of cellular components to from one to four closely positioned pencil sources, to penetrate
ionizing radiation results in the inactivation of T cells by all portions of the blood component. The number of sources
and their placement depend on the instrument and model.
The speed of rotation of the turntable is also specific to each
II model. A lead shield encloses the radiation chamber, protecting
Table 28–1 Clinical Indications for Irradiated the operator from radiation exposure. Freestanding irradiators
400 Products employing 60Co as the source of gamma rays are comparable
except that the canister containing the blood component does
In Fetus/Infant not rotate during the irradiation process; rather, tubes of 60Co
Intrauterine transfusion are placed in a circular array around the entire canister within
Prematurity
Congenital immunodeficiency
the lead chamber. When freestanding irradiators are used, the
Exchange transfusion for erythroblastosis gamma rays are attenuated as they pass through air and blood
In Child/Adult but at different rates.45 The magnitude of attenuation is greater
Congenital immunodeficiency with 137Cs than with 60Co sources.
Hematologic malignancy or solid tumor (neuroblastoma, Linear accelerators generate a beam of x-rays over a field of
sarcoma, Hodgkin’s disease where ablative chemotherapy given dimensions. Routinely, the field is projected on a table-
and/or radiotherapy is administered) top structure. The blood component is placed (flat) between
Peripheral blood stem cell or marrow transplant
Recipient of familial blood donation
two sheets of biocompatible plastic several centimeters thick.
Human leukocyte antigen–matched products The plastic on the top of the blood component (i.e., nearer
Lupus or any other condition requiring fludarabine to the radiation source) generates electronic equilibrium
Potential Indications of the secondary electrons at the point where they pass
Recipient and donor pair from a genetically homogeneous through the component container. The plastic sheet on the
population bottom of the blood component provides radiation backscat-
Other patients with hematologic malignancy or solid tumor tering that helps ensure homogeneous delivery of the x-rays.
receiving immunosuppressive agents
Infant/child with congenital heart disease with 22 qll
The blood component is usually left stationary during deliv-
deletions, other than DiGeorge syndrome ery of the entire x-ray dose. Alternatively, it may be inverted
Recipient and donor pair from genetically less homogeneous when half of the dose has been delivered, although additional
populations data on this practice are needed. Guidance is available on the
Those receiving “less intensive” immunosuppressive FDA website (http://www.fda.gov/cder/guidelines.htm).47 In
regimens
June 1999, the FDA licensed the first x-ray irradiator based
No/Limited Indications
on principles utilized in standard x-ray machines. This irra-
Patients infected with HIV
Term infants
diator does not require federal or nuclear regulatory licens-
Obstetric, surgical, and general medicine patients ing or reporting, a shielded room for operation, or special
floor reinforcement.
IRRADIATED PRODUCTS
FREE-STANDING IRRADIATOR LINEAR ACCELERATOR
Lead-enclosed chamber
Turntable
Table 28–2 White Blood Cell (WBC) Content of Different Blood Components
Volume
Component (approximate)* (mL) Average WBC Content
*
Less with new modified chambers.
Data from Brechner ME (ed). Technical Manual. Bethesda, Md., American Association of Blood Banks, 2005; Dzik
WH. Leukoreduced blood components: Laboratory and clinical aspects. In Slonim TL, Dzik WH, Snyder EL, Stowell CP,
Strauss RG (eds). Rossi’s Principles of Transfusion Medicine, 3rd ed. Philadelphia, Lippincott, 2002, pp 270–287.
BLOOD BANKING
Table 28–3 Blood Components Requiring presence of a sizable number (approximately 1 × 107) of via-
Irradiation for Patients at Risk of Graft- ble lymphocytes.
versus-Host Disease Blood components containing lymphocytes that are
homozygous for an HLA haplotype that is shared with the
Products Known to Contain Viable T Cells recipient, whether that recipient is immunocompromised or
Whole blood immunocompetent, pose a specific risk for TA-GVHD. This
Packed red blood cells (pRBCs) circumstance occurs when first- and second-degree relatives
Frozen/deglycerolized RBCs serve as directed donors10,11,37,39 and when HLA-matched
Leukoreduced pRBCs
Platelet concentrates, pooled
platelet components donated by related or unrelated indi-
Platelets, pheresis viduals are transfused.63,64 Irradiation of blood components
Granulocytes must be performed in these situations.
Nonfrozen plasma (fresh plasma) One group has evaluated lymphocyte cell surface activa-
Products That May Contain Viable T Cells tion markers over time to determine whether older red blood
Fresh frozen plasma cells (4 days or more) had alterations that could account for
Frozen plasma (e.g., FP24) the reduced frequency of TA-GVHD as compared with “fresh”
Products Unlikely to Contain Viable T Cells blood transfusion. They demonstrated reduced expression
Cryoprecipitate of CD3, CD4, CD28, CD2, and CD45 and less responsiveness
SD-plasma in a mixed lymphocyte culture (MLC) by day 4 of storage.65
“Pathogen-reduced” products (in development)
Additional studies are warranted, however, to confirm these
findings using techniques more sensitive than MLC. In addi-
tion, the number and specific subtype of T cells present in a
product that induce TA-GVHD may depend on the patient’s
immunocompetence at the time of transfusion. It is likely
through filtration may decrease the potential for GVHD that the greater the degree of immunosuppression, the fewer
and serve as an alternative to irradiation in the future when viable T cells are required to produce GVHD in susceptible
filtration technology is improved and questions about the patients. It has also been suggested that CTLs or IL-2–secret-
minimum level of viable T cells that can lead to GVHD are ing precursors of helper T cells may be more predictive of
resolved. However, there have been reports of TA-GVHD in GVHD than the number of proliferating T cells alone.
patients who received leukoreduced but not irradiated RBCs,
although the extent of leukoreduction was not uniformly Storage of Red Blood Cells and Platelets
quantified in such reports.51–54 after Irradiation
Irradiated RBCs undergo enhanced efflux of potassium
RED BLOOD CELLS
during storage at 1°C to 6°C,55,56 which is not affected by
II prestorage leukoreduction.57 Washing of RBC units before Irradiation of RBCs is not a benign process. The viability in
transfusion to reduce the supernatant potassium load is not vivo of irradiated RBCs, evaluated as the 24-hour recovery,
402 typically warranted in most cases, because post-transfu- is reduced during storage compared with that of nonirradi-
sion dilution prevents increases in plasma potassium.58 On ated RBCs.66–70 This reduced viability has raised questions
the other hand, when irradiated RBCs are used for neona- concerning the maximum storage time for RBCs after irradi-
tal exchange transfusion or the equivalent of a whole blood ation. Davey and colleagues66 found that the 24-hour recov-
exchange is anticipated, RBC washing should be considered ery for RBCs preserved in a solution containing adenine,
to prevent the possible adverse cardiotoxicity caused by sodium chloride, dextrose, and mannitol (Adsol) and treated
hyperkalemia associated with irradiation and storage.59 In with 3000 cGy on day 0 was 68.5% ± 8.1% (mean ± stan-
this regard, recent studies have provided guidance on storage dard deviation [SD]) after 42 days of storage compared with
of irradiated and washed packed RBCs: storage at 4° C for 3 78.4% ± 7.1% for control, nontreated RBCs. Subsequent
hours postirradiation and washing will produce units with studies employed total storage periods of 21 to 35 days after
less than 5 mEq of potassium per liter.60 irradiation on day 0 or day 1. After storage for 35 days, the
Platelet components that have low levels of leukocytes mean (± SD) 24-hour recovery for irradiated (3000 cGy) and
because of apheresis collection or leukofiltration, or both, control Adsol-preserved RBCs was 78.0% ± 6.8% and 81.8% ±
should also be irradiated if intended for transfusion to sus- 4.4%, respectively. In studies with a 28-day storage period, the
ceptible patients because, as with RBC transfusions, the min- values for irradiated (2500 cGy) and control Adsol-preserved
imum number of T cells that induces TA-GVHD has not yet RBCs were 78.6% ± 5.9% and 84.2% ± 5.l%, respectively.68
been delineated. In a study which compared x-ray to gamma-irradiation with
In contrast, there is controversy about irradiation of CPDA1 red cells, slight increases in extracellular potassium
fresh frozen plasma. It is generally accepted that the freez- were noted in gamma-irradiated cells with no differences in
ing and thawing processes destroy the T cells that are present lymphocyte inhibition as measured by MLC; RBC survival
in such plasma. However, two brief articles have suggested and viability studies were not included in this report.69
that immunocompetent progenitor cells may be present in Moroff and colleagues70 evaluated the effect of irradiation
frozen/thawed plasma, and the authors recommended that on Adsol-preserved RBCs stored from day 1 to 28 (irradiated
frozen/thawed plasma be irradiated.61,62 Further studies are day 1, protocol 1), day 14 to 28 (irradiated day 14, protocol
needed to validate these findings and to assess whether the 2), day 14 to 42 (irradiated day 14, protocol 3), and day 26 to
number of immunocompetent cells present in thawed fresh 28 (irradiated day 26, protocol 4). In comparison with pre-
frozen plasma is sufficient to induce TA-GVHD. In the rare vious work, these investigations were unique because RBCs
instances in which nonfrozen plasma (termed fresh plasma) were studied after being irradiated for various times in stor-
is transfused, it should certainly be irradiated because of the age and then examined after further storage. For protocol 1,
IRRADIATED PRODUCTS
the mean ± SD recovery was 84.2% ± 5.1% for control RBCs Later studies from our laboratory with a more sensi-
and 78.6% ± 5.9% for irradiated RBCs (n =16; P 0.01). With tive limiting dilution assay (LDA) indicated that 2500 cGy
protocol 3, the recoveries were 76.3% ± 7.0% for control (measured at the internal midplane of a component) is the
RBCs and 69.5% ± 8.6% for irradiated RBCs (n =16; P <0.01). most appropriate dose.86,87 In these experiments, RBC and
Protocols 2 and 4 demonstrated comparable 24-hour recov- platelet components were irradiated in their original plastic
eries for control and irradiated RBCs. Long-term RBC sur- containers (blood bags) with increasing radiation doses. The
vival was comparable for control and irradiated RBCs in all author’s laboratory selected LDA as a tool to study the effect
protocols, confirming previous data showing that long-term of gamma-irradiation for several reasons. LDA measures the
survival of RBCs is minimally influenced by irradiation. On clonogenic potential of both CD4+ and CD8+ T cells in a
the basis of multiple linear regression analysis, only the length functional assay. It provides a quantification at low T-cell
of storage after irradiation had a significant effect on the numbers and has been used to determine residual, functional
24-hour recovery. No effect was related to day of irradiation T cells in bone marrow purged of T cells, thus providing a
or total storage time. clinical correlate of prevention of GVHD.88 Assays of T-cell
Most RBC properties such as adenosine triphosphate proliferation using MLC, mitogens, or detection of T cells by
(ATP) levels and the amount of hemolysis were altered to flow cytometry can detect up to a 2 logarithm (log) reduction
only a small extent compared with control values with in viable cells. PCR techniques are capable of detecting up to
extended storage after irradiation, but potassium leakage a 6-log reduction but cannot distinguish between viable and
from the RBCs during storage was substantially enhanced nonviable cells and hence are not informative if T cells are
by irradiation.70 Despite elevated potassium levels, the irra- inactivated but remain in the sample. While the LDA assay
diation-induced changes in RBC viability and potassium may fail to detect an as yet undescribed human T-cell subset
leakage were not complementary. On the basis of analysis that contributes to GVHD, it is nonetheless an appropriate
of these studies, FDA guidelines call for a 28-day maximum assay for selection of radiation doses for platelet pheresis and
storage period for RBCs after irradiation, irrespective of the RBC components to abrogate TA-GVHD.
day of storage on which the treatment was performed, with In the authors’ studies, after each radiation dose, samples
the provision that the total storage time not exceed that for were removed and the clonogenic proliferation of T cells was
nonirradiated RBCs.47 measured by LDA. With RBC units, 500 cGy had a minimal
The etiology of the RBC irradiation lesion has never been influence, whereas 1500 cGy inactivated T-lymphocyte pro-
completely elucidated. Lipid peroxidation and RBC mem- liferation by approximately 4-log; however, some growth
brane protein array appear unaffected71 whereas purine was still observed in each experiment. Increasing the dose
nucleotides decrease over time.72 While the actual structural to 2000 cGy resulted in no T-lymphocyte proliferation in
changes that make RBCs sensitive to irradiation-induced all but one experiment. No proliferation was observed after
oxidative damage and result in potassium leakage are not 2500 cGy in any experiment.86 In a subsequent study using
yet clarified, addition of antioxidants such as dipyridamole, platelet pheresis components with sufficient T cells to per-
Trolox, and mannitol to RBCs prior to irradiation is under form the LDA, the effects of 1500 and 2500 cGy were again 28
consideration in Japan.73 evaluated.87 With 1500 cGy, substantial inactivation was
measured; however, some growth was still observed in all 403
PLATELETS
experiments. As noted with the RBC experiments, 2500 cGy
In contrast to RBCs, platelets appear to be relatively unaf- resulted in complete abrogation of clonogenic T-lymphocyte
fected by irradiation. The storage period at 20°C to 24°C for proliferation. Other laboratories used more traditional assay
irradiated platelet components does not need to be modified. methods to assess T-lymphocyte inactivation and suggested
Neither in vitro nor in vivo platelet properties are influenced a radiation dose of between 2800 to 3000 cGy. Based on
to any extent by irradiation. Many studies have confirmed that their review, the FDA has recommended that the irradiation
platelet properties are retained immediately after conventional process deliver 2500 cGy to the internal midplane of a free-
levels of irradiation and at the conclusion of a 5-day storage standing irradiation instrument canister, with a minimum of
period, whether irradiation is performed before storage or in 1500 cGy at any other point within the canister.47
the middle of storage.74–81 One report, however, has indicated
Quality Assurance Measures
some differences in selected in vitro parameters between irra-
diated and control platelets after storage.80 The instrument being used for irradiation must be docu-
mented to be operating appropriately and it must be con-
Selection of Radiation Dose firmed that blood components have been irradiated. To
In the past, there were no standards pertaining to the radia- ensure that the irradiation process is being conducted cor-
tion dose that should be used. A survey in 1989 indicated rectly, specific procedures are recommended for freestanding
that a range of radiation dose levels between 1500 and irradiators and linear accelerators, as summarized in Table
5000 cGy (1 rad = 1 cGy) were being used, with the majority 28-4 and discussed in detail in a published review.89
of facilities employing 1500 cGy.82 These reported radiation Dose mapping measures the delivery of irradiation
doses were typically estimates and retrospective calcula- within a simulated blood component or over an area in
tions because most facilities were not performing any type which a blood component is placed. This applies to a
of dosimetry measurements. Furthermore, these values may radiation field when a linear accelerator is used or to the
be different from current values determined through careful canister of a freestanding irradiator. Dose mapping is the
dose mapping because there was previously no standardized primary means of ensuring that the irradiation process is
way of calculating or reporting radiation dose. The selection being conducted correctly. It documents that the intended
of 1500 cGy as the target radiation dose was based on stud- dose of irradiation is being delivered at a specific loca-
ies in the 1970s showing that 500 cGy abrogated the MLC tion (such as the central midplane of a canister), and it
response of isolated lymphocytes.83–85 describes the variation of the delivered radiation dose
BLOOD BANKING
Table 28–4 Quality Assurance Guidelines for Irradiating Blood Components
Dose
2500 cGy to the central midplane of a canister (freestanding irradiator) or to the center of an irradiation field
(linear accelerator) with a minimum of 1500 cGy
Dose Mapping (Freestanding Irradiators)
Routinely, once a year (137Cs) or twice a year (60Co) and after major repairs, using a fully filled canister
(water/plastic) with a dosimetry system to map the distribution of the absorbed dose
Dose Mapping (Linear Accelerators)
Yearly dose mapping with an ionization chamber and a water phantom recommended; more frequent
evaluation of instrument conditions to ensure consistency of x-rays
Correction for Radioisotopic Decay
With 137Cs as the source, annually
With 60Co as the source, quarterly
Turntable Rotation (Freestanding 137Cs Irradiators)
Daily verification
Storage Time for Red Blood Cells after Irradiation
For up to 28 days; total storage time cannot exceed maximum storage time for unirradiated
red blood cells
Storage Time for Platelets after Irradiation
No change related to irradiation procedure
Adapted from Moroff G, Luban NLC. The irradiation of blood and blood components to prevent graft-versus-host disease: Technical issues and
guidelines. Transfus Med Rev 1997;11:15–26.
within a simulated component or over a given area. This Other studies have shown that the variability in the
allows conclusions to be drawn about the maximum and dose delivered to the interior of simulated blood phantoms
minimum doses being delivered. Dose mapping should be depended on the model of the 137Cs freestanding irradia-
performed with sensitive dosimetry techniques. Several tor.93,94 An immobilized grid of thermoluminescent dosi-
commercially available systems have been developed (see meters in a plastic sheet was placed within the simulated
Dosimetry Systems in Use, below). blood units to measure dose delivery. A spacer in the bottom
Other quality assurance measures include routine con- of the canister increased the minimum level of irradiation
II firmation that the turntable is operating correctly (for 137Cs within the simulated blood units, as expected from the
irradiators), measurements to ensure that the timing device results of full-canister dose mapping involving a phantom.94
404 is accurate, and periodic lengthening of the radiation time to The variability with 137Cs irradiation is influenced by a num-
correct for source decay. With linear accelerators, it is neces- ber of factors, including the blood-compatible environment.
sary to measure the characteristics of the x-ray beam to ensure Cumulatively, these studies underscored the need for consis-
consistency of delivery. Confirming that a blood component tency in loading the canister.
has, in actuality, been irradiated is also an important part of When an irradiator is purchased, the distributor provides
a quality assurance program. Several firms have developed a central dose level that is determined in a blood-compat-
indicator labels for this purpose. ible environment. In the 1970s and 1980s, manufacturers
provided a central dose that was determined in air, resulting
Dose Mapping with Freestanding Irradiators in the use of timer settings that provided a dose level some-
For freestanding irradiators, a dose-mapping procedure what less than was expected. Since the issuance of the FDA
measures the dose delivered throughout the circular canister guidelines in July 1993 and the use of dose mapping, it has
in which the blood component is placed. To establish a two- been necessary to readjust irradiation times with some instru-
dimensional map, a dosimetry system is placed in a canis- ments because the attenuation effect had not been considered
ter that is completely filled with a blood/tissue-compatible previously.
phantom composed of water or an appropriate plastic such The dose map can also be used to assess whether the
as polystyrene.90,91 The dosimetry material is placed within turntable of a 137Cs irradiator is rotating in an appropriate
the phantom in a predetermined way. This approach pro- manner. The occurrence of comparable readings at the two
vides data showing the minimum levels of irradiation that edges of the two-dimensional map, as depicted in the theo-
would be absorbed by a blood component placed in the retical dose map, indicates that the canister is rotating evenly
canister and recognizes that maximum attenuation occurs in front of the 137Cs source. If the turntable were not rotating,
when the canister is completely filled with a blood-compat- the dose levels at the edge of the map closest to the source
ible material. The absorbed dose at the central midplane of would be much higher than those on the opposite edge (i.e.,
a canister (i.e., at the center point) may be decreased by 25% the side farthest from the source).
(from 3100 to 2500 cGy) in a 137Cs irradiator (JL Shepherd
Dosimetry Systems in Use
and Associates, San Francisco, Calif.) when the loading of
the canister is changed from 0% (air) to 100% (with blood The radiation dose delivered can be measured by a variety
components).92 An irradiation-sensitive film dosimetry of dosimetry systems. Several commercial systems have been
system (International Specialty Products) was used in this introduced to the market; each system consists of a phan-
study. tom that fills the canister and a sensitive dosimeter. These
IRRADIATED PRODUCTS
dosimeters are referred to as routine dosimeters. They are TOP 2400 2000 2450
calibrated against standard systems, usually at national refer-
ence laboratories such as the National Institute of Standards
and Technology in the United States. The routine dosimeter 2700 2340 2730
measurement systems were initially developed for use with
137
Cs irradiators because this is the predominant irradiation
CENTER 2950 2560 2980
source for blood. Subsequently, they were also developed for
use with 60Co irradiators.
Thermoluminescent dosimeters (TLD chips) are one type 2810 2400 2780
of routine dosimeter. TLD chips are small plastic chips with
millimeter dimensions having a crystal lattice that absorbs
ionizing radiation. Specialized equipment is used to release BOTTOM 2380 1750 2420
and measure the energy absorbed by the TLD chip at the EDGE MIDPLANE EDGE
time of the test irradiation. In one commercially available
Figure 28–3 Two-dimensional dose map showing the irradiation
system, chips are placed at nine different locations within a dose distribution through a fully filled canister of a freestanding cesium
polystyrene phantom that fits into the canister for the IBL 137 irradiator.
437C irradiator (CIS US, Inc., Bedford, Mass.). The timer
setting routinely used with the instrument is also used in the
test procedure.
Two alternative systems use radiochromic film. On expo- All dosimetry measurements are associated with a degree of
sure to irradiation, the film darkens, resulting in an increase uncertainty or possible error. The magnitude of the uncer-
in optical density. The optical density, determined at various tainty depends on the kind of dosimeter used. For most
locations on the film, is linearly proportional to the absorbed dosimeters, the level is ±5% of the measured value. For a cen-
radiation dose. Standard films that are irradiated at a given tral absorbed dose level of 2560 cGy (see theoretical dose map
dose level with a source calibrated at a national reference in Fig. 28–3), the value could be as high as 2788 cGy or as low
laboratory provide the means to assess the absolute level as 2432 cGy. Correspondingly, a measured value of 2400 cGy
of absorbed irradiation. This type of dosimeter is basically could be as high as 2520 cGy or as low as 2380 cGy. Because
an x-ray film comparable with that used in clinical prac- the measured value could in actuality meet the 2500 cGy
tice. With this device, the map that is developed identifies standard, it is appropriate to accept a value of 2400 cGy as
the absorbed radiation dose measured at a large number meeting the current standard. The same approach should be
of locations. used when evaluating the minimum value on a dose map.
In one such system, a film contained in a thin watertight Albeit arbitrary and cautious, the actual minimum on an
casement is placed into the canister (International Specialty irradiation dose map should not be below 1500 cGy.
Products, Wayne, N.J.). This approach is being used with a Other Measures with Freestanding Irradiators 28
variety of irradiators. The canister is filled completely with
CORRECTION FOR ISOTOPIC DECAY
water before the irradiation procedure. This system provides 405
a direct readout of the dose that is delivered throughout It is important to lengthen the time of irradiation periodi-
the canister. The timer setting used routinely is employed cally to correct for decay of the isotopic source that emits
for the test procedure. In a second system, a film having the gamma-irradiation. Previously, this was the only major
different irradiation-sensitive characteristics is embedded quality assurance measure that was performed routinely.
between two halves of a circular-fitting polystyrene plastic With the 30-year half-life of 137Cs, annual lengthening of the
phantom (Nordion International, Kanata, Ontario, Canada). timer setting is appropriate. On the other hand, since the
Irradiation of specialized films is performed with a number half-life of 60Co is only 5 years, the time of irradiation should
of timer settings, each longer than that used routinely. The be increased on a quarterly basis. The additional seconds of
map produced is normalized for a central midplane dose irradiation that are needed can be calculated using formu-
of 2500 cGy. The time to produce the 2500 cGy is predeter- las that can be found in a physics textbook. Alternatively,
mined with a different dosimeter system, the Fricke system, distributors of irradiators provide a chart that specifies the
in which absorbed radiation causes a change in the state of appropriate setting as a function of calendar time.
an iron salt that can be assessed spectrophotometrically.
TURNTABLE ROTATION
Another approach to irradiation dose mapping employs
a solid-state electronic dosimeter that is technically referred For 137Cs irradiators, it is essential that the turntable oper-
to as a metal oxide semiconductor field effect transistor ate at a constant speed in a circular pattern to ensure that
(MOSFET). A board contains a number of small transistors all parts of a blood component are exposed equally to the
in an arrangement that provides data for a dose map. The source. Daily verification of turntable rotation is an appro-
board is placed between two halves of a circular polystyrene priate quality assurance measure. With some freestanding
phantom that fits into the canister. This dosimeter absorbs irradiator models, rotation of the turntable can be observed
and stores the radiation dose imparted to it electronically. before the door of the compartment in which the canister
Irradiation causes the formation of holes in the metal oxide is positioned is closed. In other models, it can be observed
layer that become trapped within the transistor. The magni- only indirectly by ensuring that an indicator light is oper-
tude of the holes is evaluated by measuring the voltage across ating appropriately. With some older models, there have
the transistor with a voltmeter. The voltages measured are been occasional reports that the turntable failed to rotate
converted to absorbed dose. because of mechanical problems. Such problems should not
With each dosimetry system, measurements are used to be encountered with the newer models because of changes in
express the absorbed radiation dose in grays (or centigrays). the turntable mechanisms.
BLOOD BANKING
ASSESSING RADIOACTIVITY LEAKAGE
array, which can be expressed in terms of the number of
photons delivered per square centimeter. These parameters
Irradiators are constructed so that the isotopic sources should be assessed routinely as part of quality control pro-
are contained in a chamber heavily lined with a protective grams used by radiation physicists. A code of blood practice
lead shield to prevent leakage of radioactivity. Accordingly, was published in 1994 by the Radiation Therapy Committee
gamma-irradiators are considered to be safe instruments. of the American Association of Physicists in Medicine for the
Although there have been no reports of source leakage quality control of radiotherapy accelerators.96,97
of radioactivity, periodic measurements are warranted.
Attaching a film badge to the outside of the irradiator, using Unusual Geometries
a Geiger counter periodically, and performing a wipe test of Dose delivery can be influenced by several physical and
the inside of the chamber where the canister is positioned at geometrical factors. The greatest influence on delivery is
least semiannually are typical measures. the geometry of the sample being irradiated. The authors’
laboratory used gel dosimetry to measure three-dimensional
Dose Mapping with Linear Accelerators radiation dose distributions in blood containers with irregular
Linear accelerators that are used to provide radiation therapy geometries, such as syringes, platelet bags, and small-volume
are carefully monitored to ensure appropriateness of dose to red cell bags.98 This approach, while technically complex,
an irradiation field. When blood components are treated permits “imaging” of the actual dose delivered.
with x-rays, the instrument settings are different from those
Confirmation That Irradiation Occurred
used to treat oncology patients. Hence, additional periodic
quality control measures, primarily to assess the dose deliv- It is important to have positive confirmation that the irradia-
ered to blood components, are needed to ensure that linear tion process has taken place. For example, irradiation would
accelerators are being operated appropriately when used for not occur if an operator failed to initiate the electronically
blood irradiation. controlled irradiation process or there was an instrumenta-
Currently, there are no commercially available systems for tion malfunction. A radiation-sensitive indicator label has
assessing the dose delivered throughout the area of an irra- been developed specifically for this purpose (International
diation field in which blood components are placed for treat- Specialty Products, Wayne, N.J.). The label containing a radia-
ment with x-rays. An ideal dosimeter for this purpose would tion-sensitive filmstrip is placed on the external surface of the
be made of a tissue-compatible plastic phantom containing blood component. Irradiation causes distinct visually observ-
appropriate dosimeter material that could be placed at the able changes in the label that can be readily assessed by the
correct distance from the source. An alternative approach operator, which represents a permanent visual record that the
might involve use of a blood bag filled with water (simulating irradiation process has taken place. The reliability of this type
a blood unit) containing TLD chips, as described earlier. In of indicator has been documented in a multisite study.99
comparative studies using such simulated blood units, it was Two versions of the indicator label have been manufac-
II determined that radiation delivery was more uniform with tured. They differ in the range of irradiation needed to cause
linear accelerators than with 137Cs freestanding irradiators.93 a change in the radiation-sensitive film. The ratings for these
406 This uniformity reflects the relative homogeneity of x-ray indicators are 1500 and 2500 cGy. The ratings serve as an
beams. One group of investigators has developed a phantom approximate guideline for the amount of absorbed radiation
with TLD chips that can serve as both a dose monitor and a needed to change the window completely from reddish to
temperature monitor, a critical issue when multiple units are opaque. Because the indicator labels are designed and used to
held out of refrigeration during irradiation.95 confirm that the irradiation process has occurred, the authors’
In the absence of an available system modified for laboratory utilizes the 1500 cGy label as the most appropriate
irradiation of blood bags, the dose delivered throughout tool to perform this quality control measure. This is based
an irradiation field should be mapped with the dosimetric on the routinely observed pattern of dose distribution to a
measuring system known as an ionization chamber. The ioniza- blood component in a canister of a freestanding irradiator.
tion chamber is used to calibrate linear accelerators for use with Despite a targeted central dose of 2500 cGy, there are spots
patients. In addition, on a yearly basis, dose mapping should at which the dose is less. If the theoretical dose map pre-
be performed using a tissue-compatible phantom. In view of sented in Figure 28–3 is used as an example, there is a spot
the widely divergent conditions that are employed during the that receives only 1800 cGy. If the label rated 2500 cGy was
operation of linear accelerators, other parameters pertaining located on the external surface of a component, there might
to the x-ray beam should be evaluated on at least a quarterly be minimal changes in the appearance of the radiation-
basis to provide assurance that the instrument is being used sensitive film window. This would result in a judgment that
appropriately for the irradiation of blood components. the blood component was not irradiated when in actuality it
When setting a linear accelerator for blood component was treated satisfactorily.
irradiation, the following should be measured: (1) the distance
between the x-ray source and the position where the blood New Methods
components are to be placed, (2) consistency of the strength
of the x-ray beam, and (3) intensity of the x-ray beam. The Photochemical treatment (PCT) using psoralens and long-
distance between the source and position on the table where wavelength UV irradiation (UVA) have been developed to
blood components will be placed (referred to as the target) reduce the risks of bacterial and viral contaminants of platelet
can be evaluated easily with a calibrated measuring device. transfusions (see Chapter 27 for further discussion of pathogen
This is a simple task that can be performed routinely. The con- inactivation technologies). Psoralens bind reversibly to nucleic
sistency of beam output can be evaluated by measuring the acids by intercalation and, after UVA illumination, form cova-
beam current. Beam intensity can be evaluated by measuring lent monoadducts and crosslinks with RNA and DNA. The
the ionization current in a monitoring ionization chamber process modifies bacterial and viral genomes sufficiently to
IRRADIATED PRODUCTS
prevent replication. Among a broad group of compounds, the 13. Nishimura M, Uchida S, Mitsunaga S, et al. Characterization of T-cell
psoralen S-59 has been shown to be particularly effective in clones derived from peripheral blood lymphocytes of a patient with
transfusion-associated graft-versus-host disease: FAS-mediated killing
inactivating bacteria and viruses without adversely affecting by CD4+ and CD8+ cytotoxic T-cell clones and tumor necrosis factor
platelet function in vitro and in vivo.100 Clinical trials confirm beta production by CD4+ T-cell clones. Blood 1997;89:1440–1445.
adequate in vivo survival, lack of adverse effects, and equiva- 14. Ammann AJ. Hypothesis: absence of graft-versus-host disease in AIDS
lence to standard platelet units.101 is a consequence of HIV-1 infection of CD4+ T cells. J Acquir Immune
Defic Syndr 1993;6:1224–1227.
The use of S-59 and PCT has been studied for possible 15. Klein C, Fraitag S, Foulon E, et al. Moderate and transient transfusion-
inactivation of leukocytes in platelet concentrates.102,103 PCT associated cutaneous graft versus host disease in a child infected by
inactivation of T cells was evaluated with four assay systems. human immunodeficiency virus. Am J Med 1996;101:445–446.
These included T-cell quantitation, inhibition of cytokine 16. Lee TH, Reed W, Mangawang-Montalvo L, et al. Donor WBCs can
synthesis, modification of leukocyte genomic DNA by quan- persist and transiently mediate immunologic function in a murine
transfusion model: effects of irradiation, storage, and histocompatibil-
tification of psoralen-DNA adducts, and inhibition of repli- ity. Transfusion 2001;41:637–642.
cation of T cells using PCR amplification of genomic DNA 17. Lee TH, Donegan E, Slichter S, Busch MP. Transient increase in
sequences. This work built on previous studies in which IL-8, circulating donor leukocytes after allogeneic transfusions in immuno-
a marker of cytokine generation in platelet concentrates, was competent recipients compatible with donor cell proliferation. Blood
1996;85:1207–1214.
significantly reduced in pools treated with PCT compared 18. Lee TH, Paglieroni T, Ohto H, et al. Survival of donor leuko-
with those treated with irradiation.102,103 cyte subpopulations in immunocompetent transfusion recipients:
A more detailed study by Grass and coworkers103 com- frequent long term microchimerism in severe trauma patients. Blood
pared PCT with irradiation at 2500 cGy; cytokine synthesis 1999;93:3127–3139.
was not inhibited, and induction of DNA strand breaking 19. Lee TH, Paglieroni T, Utter GH, et al. High-level long-term white blood
cell microchimerism after transfusion of leukoreduced blood compo-
was inhibited less than with S-59 and PCT. LDA was used to nents to patients resuscitated after severe traumatic injury. Transfusion
confirm inactivation of T cells in the platelet concentrates. 2005;45:1280–1290.
The efficacy of S-59 and PCT was further supported by a 20. Utter GH, Owings JT, Lee TH, et al. Microchimerism in transfused
study of transfusion-induced GVHD in a murine F1 hybrid trauma patients is associated with diminished donor-specific lympho-
cyte response. J Trauma 2005;58:925–931; discussion 931–932.
model.104,105 No GVHD was noted in mice receiving splenocytes 21. Wang-Rodriguez J, Fry E, Fiebig E, et al. Immune response to blood
treated with either 2500 cGy or S-59 at 150 μmol/L and UVA transfusion in very-low-birthweight infants. Transfusion 2000;40:
at 2.1 J/cm. Another set of experiments was performed to 25–34.
study the use of PCT to prevent GVHD in an immunocom- 22. Vietor HE, Hallensleben E, van Bree SP, et al. Survival of donor cells 25
promised mouse model.106 Taken together, these studies suggest years after intrauterine transfusion. Blood 2000;95:2709–2714.
23. Reed WF, Lee TH, Trachlenberg E, et al. Detection of microchime-
that PCT may well be an alternative to irradiation and may rism by PCR as a function of amplification strategy. Transfusion
provide a mechanism to prevent an increase in cytokine con- 2001;41:39–44.
centration in platelet concentrates. The limitation of PCT 24. Nelson J, Furst D, Maloney S, et al. Microchimerism and HLA com-
methodology is the need for UVA penetration, which is not patible relationships of pregnancy in scleroderma. Lancet 1998;351:
559–562.
currently possible with RBC products. 25. Arlett CM, Smith JB, Jimenez SA. Identification of fetal DNA and cells
28
in skin lesions from women with systemic sclerosis. N Engl J Med
1998;333:1186–1191. 407
REFERENCES 26. Evans PC, Lambert N, Maloney S, et al. Long-term fetal microchime-
rism in peripheral blood mononuclear cell subsets in healthy women
1. von Fliedner V, Higby DJ, Kim U. Graft-versus-host reaction following and women with scleroderma. Blood 1999;93:2033–2037.
blood product transfusion. Am J Med 1972;72:951–961. 27. Daya S, Gunby J, Clark DA. Intravenous immunoglobulin therapy for
2. Anderson KC, Weinstein HJ. Transfusion-associated graft-versus-host recurrent spontaneous abortion: a meta-analysis. Am J Reprod Immunol
disease. N Engl J Med 1990;323:315–321. 1998;39:69–76.
3. Webb IJ, Anderson KC. Transfusion-associated graft-versus-host dis- 28. Brubaker DB. Immunopathogenic mechanisms of posttransfusion
ease. In Anderson KC, Ness PM (eds). Scientific Basis of Transfusion graft versus host disease. Proc Soc Exp Biol Med 1993;202:122–147.
Medicine. Philadelphia, WB Saunders, 2000, pp 420–425. 29. Blundell EL, Pamphilon DH, Anderson NA, et al. Transfusion-
4. Linden JV, Pisciotto PT. Transfusion-associated graft-versus-host associated graft-versus-host disease, monoclonal gammopathy and
disease and blood irradiation. Transfus Med Rev 1992;6:116–123. PCR. Br J Haematol 1992;82:622–623.
5. Yasuura K, Okamoto H, Matsuura A. Transfusion-associated graft- 30. Capon SM, DePond WD, Tyan DB, et al. Transfusion associated graft-
versus-host disease with transfusion practice in cardiac surgery. versus-host disease in an immunocompetent patient. Ann Intern Med
J Cardiovasc Surg (Torino) 2000;41:377–380. 1991;114:1025–1026.
6. Serious Hazards of Transfusion. Available at http://www.shotuk.org. 31. Drobyski W, Thibodeau S, Truitt RL, et al. Third-party-mediated graft
Accessed March 29, 2006. rejection and graft-versus-host disease after T-cell-depleted bone
7. Kleinman S, Chan P, Robillard P. Risks associated with transfusion of cel- marrow transplantation, as demonstrated by hypervariable DNA
lular blood components in Canada. Transfus Med Rev 2003;17:120–162. probes and HLA-DR polymorphism. Blood 1989;74:2285–2294.
8. Ferrera JM, Krenger W. Graft-versus-host disease: the influence of type 32. Kunstmann E, Bocker T, Roewer L, et al. Diagnosis of transfusion-
1 and type 2 T cell cytokines. Transfus Med Rev 1998;12:1–17. associated graft-versus-host disease by genetic fingerprinting and
9. McMilan KD, Johnson RL. HLA-homozygosity and the risk of related- polymerase chain reaction. Transfusion 1992;32:766–770.
donor transfusion-associated graft-versus-host disease. Transfus Med 33. Depalma L, Bahrami KR, Kapur S, et al. Amplified fragment length
Rev 1993;7:37–41. polymorphism analysis in the evaluation of posttransfusion graft
10. Petz LD, Calhoun L, Yam P, et al. Transfusion-associated graft-versus-host versus host disease. J Thorac Cardiovasc Surg 1994;108:182–184.
disease in immunocompetent patients: report of a fatal case associated 34. Wang L, Juji T, Tokunaga K, et al. Polymorphic microsatellite mark-
with transfusion of blood from a second-degree relative, and a survey ers for the diagnosis of graft-versus-host disease. N Engl J Med
of predisposing factors. Transfusion 1993;33:742–750. 1994;330:398–401.
11. Worrall NK, Lazenby WD, Misko TP, et al. Modulation of in vivo 35. Friedman DF, Kwittken P, Cizman B, et al. DNA-based HLA typing of
alloreactivity by inhibition of inducible nitric oxide synthetase. J Exp nonhematopoietic tissue used to select the marrow transplant donor
Med 1995;181:63–70. for successful treatment of transfusion-associated graft-versus-host
12. Fast LD, DiLeone G, Edson CM, et al. PEN110 treatment function- disease. Clin Diagn Lab Immunol 1994;1:590–596.
ally inactivates the PBMNCs present in RBC units: comparison to 36. Higgins MJ, Blackall DP. Transfusion-associated graft-versus-host
the effects of exposure to gamma irradiation. Transfusion 2002;42: disease: a serious residual risk of blood transfusion. Curr Hematol
1318–1325. Reports 2005;4:470–476.
BLOOD BANKING 37. Kanter MH. Transfusion-associated graft-versus-host disease: do 65. Chang H, Voralia M, Bali M, et al. Irreversible loss of donor blood
transfusions from second-degree relatives pose a greater risk than leucocyte activation may explain a paucity of transfusion-associ-
those from first-degree relatives? Transfusion 1992;32:323–327. ated graft-versus-host disease from stored blood. Br J Haematol
38. Luban NL, DePalma L. Transfusion-associated graft-versus-host 2000;111:146–156.
disease in the neonate—expanding the spectrum of disease [editorial]. 66. Davey RJ, McCoy NC, Yu M, et al. The effect of pre-storage irradiation
Transfusion 1996;36:101–103. on post-transfusion red cell survival. Transfusion 1992;32:525–528.
39. Ohto H, Anderson KC. Post-transfusion graft-versus-host disease in 67. Mintz PD, Anderson G. Effect of gamma irradiation on the in vivo
Japanese newborns. Transfusion 1996;36:117–123. recovery of stored red blood cells. Ann Clin Lab Sci 1993;23:216–220.
40. Van Royen-Kerkhof A, Wulffraat NM, Kamphuis SS, et al. Nonlethal 68. Moroff G, Holme S, Heaton A, et al. Effect of gamma irradiation on
transfusion associated graft-versus-host disease in a severe com- viability of AS-1 red cells [abstract]. Transfusion 1992;32(Suppl):70S.
bined immunodeficient patient. Bone Marrow Transplant 2003;32: 69. Janatpour K, Denning L, Nelson K, et al. Comparison of x-ray vs.
1027–1030. gamma irradiation of CPDA-1 red cells. Vox Sang 2005;89:215–219.
41. Juji T, Nishimura M, Tadokoro K. Treatment of posttransfusion graft- 70. Moroff G, Holme S, AuBuchon JP, et al. Viability and in vitro properties of
versus-host disease. Vox Sang 2000;78(Suppl 2):277–279. AS-1 red cells after gamma irradiation. Transfusion 1999;39:128–134.
42. Hutchinson K, Kopko PM, Muto KN, et al. Early diagnosis and 71. Cicha I, Suzuki Y, Tateishi N, et al. Gamma-ray-irradiated red blood
successful treatment of a patient with transfusion-associated GVHD cells stored in mannitol-adenine-phosphate medium: rheological eval-
in autologous peripheral blood progenitor cell transplantation. Trans- uation and susceptibility to oxidative stress. Vox Sang 2000;79:75–82.
fusion 2002;42:1567–1572. 72. Leitner GC, Neuhauser M, Weigel G, et al. Altered intracellular purine
43. Shlomchik WD, Couzens MS, Tang CB, et al. Prevention of graft versus nucleotides in gamma-irradiated red blood cell concentrates. Vox Sang
host disease by inactivation of host antigen-presenting cells. Science 2001;81:113–118.
1999;285:412–415. 73. Hirayama J, Abe H, Azuma H, et al. Leakage of potassium from
44. Gorlin JB, Mintz PD. Transfusion-associated graft-vs-host-disease. In red blood cells following gamma ray irradiation in the presence of
Mintz PD (ed). Transfusion Therapy: Clinical Principles and Practice, dipyridamole, trolox, human plasma, or mannitol. Biol Pharm Bull
Bethesda, Md., American Association of Blood Banks, 2005, pp 579–591. 2005;28:1318–1320.
45. Fearon TC, Luban NL. Practical dosimetric aspects of blood and blood 74. Moroff G, George VM, Siegl AM, et al. The influence of irradiation on
product irradiation. Transfusion 1986;26:457–459. stored platelets. Transfusion 1986;26:453–456.
46. Korngold R. Biology of graft-versus-host disease. Am J Pediatr Hema- 75. Espersen GT, Ernst E, Christiansen OB, et al. Irradiated blood platelet
tol Oncol 1993;15:18–37. concentrates stored for five days: evaluation by in vitro tests. Vox Sang
47. U.S. Food and Drug Administration, Center for Biologics Evaluation 1988;55:218–221.
and Research. Available at http://www.fda.gov/cber/guidelines.htm. 76. Duguid JK, Carr R, Jenkins JA, et al. Clinical evaluation of the effects
Accessed March 29, 2006. of storage time and irradiation on transfused platelets. Vox Sang
48. Suda BA, Leitman SF, Davey RJ. Characteristics of red cells irradiated 1991;60:151–154.
and subsequently frozen for long term storage. Transfusion 1993; 77. Read EJ, Kodis C, Carter CS, et al. Viability of platelets following
33:389–392. storage in the irradiated state. A paired-controlled study. Transfusion
49. Miraglia CC, Anderson G, Mintz PD. Effect of freezing on the in vivo 1988;28:446–450.
recovery of irradiated red cells. Transfusion 1994;34:775–778. 78. Rock G, Adams GA, Labow RS. The effects of irradiation on platelet
50. Crowley JP, Skrabut EM, Valeri CR. Immunocompetent lymphocytes function. Transfusion 1988;28:451–455.
in previously frozen washed red cells. Vox Sang 1974;26:513–517. 79. Sweeney JD, Holme S, Moroff G. Storage of apheresis platelets after
51. Akahoshi M, Takanashi M, Masuda M, et al. A case of transfusion- gamma irradiation. Transfusion 1994;34:779–783.
associated graft-versus-host disease not prevented by white cell- 80. Seghatchian MJ, Stivala JF. Effect of 25 Gy gamma irradiation on
reduction filters. Transfusion 1992;32:169–172. storage stability of three types of platelet concentrates: a comparative
II 52. Heim MU, Munker R, Sauer H, et al. Graft versus host Krankheit: analysis with paired controls and random preparation. Transfus Sci
GVH mit letalem Ausgang nach der Gabe von gefilterten Erythrozy- 1995;16:121–129.
408 tenkonzentraten [Graft versus host disease with fatal outcome after 81. Bessos H, Atkinson A, Murphy WG, et al. A comparison of in vitro
administration of filtered erythrocyte concentrates]. Beitr Infusion- storage markers between gamma-irradiated and non-irradiated apher-
sther 1992;30:178–181. esis platelet concentrates. Transfus Sci 1995;16:131–134.
53. Hayashi H, Nishiuchi T, Tamura H, et al. Transfusion associated 82. Anderson KC, Goodnough LT, Sayers M, et al. Variation in blood
graft-versus-host disease caused by leukocyte filtered stored blood. component irradiation practice: implications for prevention of transfu-
Anesthesiology 1993;79:1419–1421. sion-associated graft-versus-host disease. Blood 1991;77:2096–2102.
54. Anderson KC. Leukodepleted cellular blood components for preven- 83. Sprent J, Anderson RE, Miller JF. Radiosensitivity of T and B lympho-
tion of transfusion-associated graft-versus-host disease. Transfus Sci cytes. II. Effect of irradiation on response of T cells to alloantigens. Eur
1995;16:265–268. J Immunol 1974;4:204–210.
55. Ramirez AM, Woodfield DG, Scott R, et al. High potassium levels in 84. Valerius NH, Johansen KS, Nielsen OS, et al. Effect of in vitro x-irradia-
stored irradiated blood. Transfusion 1987;27:444–445. tion on lymphocyte and granulocyte function. Scand J Hematol 1981;
56. Rivet C, Baxter A, Rock G. Potassium levels in irradiated blood [letter]. 27:9–18.
Transfusion 1989;29:185. 85. Rosen NR, Weidner JG, Bold HD, et al. Prevention of transfusion-
57. Swann ID, Williamson LM. Potassium loss from leukodepleted red associated graft-versus-host disease: selection of an adequate dose of
cells following gamma-irradiation. Vox Sang 1996;70:117–118. gamma irradiation. Transfusion 1993;33:125–127.
58. Strauss RG. Routine washing of irradiated red cells before transfusion 86. Pelszynski MM, Moroff G, Luban NL, et al. Effect of gamma-irradia-
seems unwarranted. Transfusion 1990;30:675–677. tion of red blood cell units on T-cell inactivation as assessed by limiting
59. Luban NL, Strauss RG, Hume HA. Commentary on the safety of red dilution analysis: implications for preventing transfusion-associated
cells preserved in extended-storage media for neonatal transfusion. graft-versus-host disease. Blood 1994;83:1683–1689.
Transfusion 1991;31:229–235. 87. Luban NL, Drothler D, Moroff G, et al. Irradiation of platelet com-
60. Weiskopf RB, Schnapp S, Rouine-Rapp K, et al. Extracellular potas- ponents: inhibition of lymphocyte proliferation assessed by limiting
sium concentrations in red blood cell suspensions after irradiation and dilution analysis. Transfusion 2000;40:348–352.
washing. Transfusion 2005;45:1295–1301. 88. Quinones RR, Gutierrez RH, Dinndorf PA, et al. Extended cycle
61. Wielding JU, Vehmeyer K, Dittman J, et al. Contamination of fresh- elutriation to adjust T-cell content in HLA-disparate bone marrow
frozen plasma with viable white cells and proliferable stem cells transplantation. Blood 1993;82:307–317.
[letter]. Transfusion 1994;34:185–186. 89. Moroff G, Luban NL. The irradiation of blood and blood components
62. Bernvill SS, Abdulatiff M, Al-Sedairy S, et al. Fresh frozen plasma to prevent graft-versus-host disease: technical issues and guidelines.
contains viable progenitor cells—should we irradiate [letter]? Vox Transfus Med Rev 1997;11:15–26.
Sang 1994;67:405. 90. Masterson ME, Febo R. Pre-transfusion blood irradiation: clinical
63. Benson K, Marks AR, Marshall MJ, et al. Fatal graft-versus-host disease rationale and dosimetric considerations. Med Phys 1992;19:649–657.
associated with transfusions of HLA-matched, HLA-homozygous 91. Leitman SF. Dose, dosimetry and quality improvements of irradiated
platelets from unrelated donors. Transfusion 1994;34:432–437. blood components [editorial]. Transfusion 1993;33:447–449.
64. Grishaber JE, Birney SM, Strauss RG. Potential for host transfusion- 92. Perkins JT, Papoulias SA. The effect of loading conditions on dose
associated graft-versus-host disease due to apheresis platelets matched distribution within a blood irradiator [abstract]. Transfusion 1994;
for HLA class I antigens. Transfusion 1993;33:910–914. 34:75S.
IRRADIATED PRODUCTS
93. Moroff G, Luban NLC, Wolf L, et al. Dosimetry measurements after 102. Hei DJ, Grass J, Lin L, et al. Elimination of cytokine production in
gamma irradiation with cesium-137 and linear acceleration sources stored platelet concentrate aliquots by photochemical treatment with
[abstract]. Transfusion 1993;33:52S. psoralen plus ultraviolet A light. Transfusion 1999;39:239–248.
94. Luban NL, Fearon T, Leitman SF, et al. Absorption of gamma irra- 103. Grass JA, Hei DJ, Metchette K, et al. Inactivation of leukocytes in platelet
diation in simulated blood components using cesium irradiators concentrates by psoralen plus UVA. Blood 1998;91:2180–2188.
[abstract]. Transfusion 1995;35:63S. 104. Grass JA, Wafa T, Reames A, et al. Prevention of transfusion-associated
95. Goes EG, Covas DT, Haddad R, et al. Quality control system for blood graft-versus-host disease by photochemical treatment. Blood 1999;93:
irradiation using a teletherapy unit. Vox Sang 2004;86:105–110. 3140–3147.
96. Kutcher GJ, Coia L, Gillin M, et al. Comprehensive QA for radiation 105. Fast LD. The effect of exposing murine splenocytes to UVB light,
oncology: report of AAPM Radiation Therapy Committee Task Group psoralen plus UVA light, or gamma-irradiation on in vitro and in vivo
40. Med Phys 1994;21:581–618. immune responses. Transfusion 2003;43:576–583.
97. Nath R, Biggs PJ, Bova FJ, et al. AAPM code of practice for radiother- 106. Grass J, Delmonte J, Wages D, et al. Prevention of transfusion-
apy accelerators: report of AAPM Radiation Therapy Task Group 45. associated graft vs. host disease (TA-GVHD) in immunocompromised
Med Phys 1994;21:1093–1121. mice by photochemical treatment (PCT) of donor T cells. Blood
98. Fearon T, Criss VR, Luban NLC. Blood irradiator dosimetry with 1997;90(Suppl 1):207A.
BANG polymer gels. Transfusion 2005;45:1658–1662. 107. Brechner ME (ed). Technical Manual. Bethesda, Md., American
99. Leitman SF, Silberstein L, Fairman RM, et al. Use of a radiation- Association of Blood Banks, 2005.
sensitive film label in the quality control of irradiated blood compo- 108. Dzik WH. Leukoreduced blood components: laboratory and clinical
nents [abstract]. Transfusion 1992;32:4S. aspects. In Slonim TL, Dzik WH, Snyder EL, Stowell CP, Strauss RG
100. Lin L, Cook DN, Wiesehahn GP, et al. Photochemical inactivation of (eds). Rossi’s Principles of Transfusion Medicine, 3rd ed. Philadelphia,
viruses and bacteria in human platelet concentrates using a novel psoralen Lippincott, 2002, pp 270–287.
and long wavelength UV light. Transfusion 1997;37:423–435.
101. Corash L, Lin L. Novel processes for inactivation of leukocytes to
prevent transfusion-associated graft-versus-host disease. Bone Marrow
Transplant 2004;33:1–7.
28
409
Chapter 29
Washed and Volume-Reduced
Blood Components
S. Gerald Sandler ● Viviana V. Johnson ●
Jayashree Ramasethu
Large-volume or rapid transfusion Particularly in newborns and small children, decreases risk of
hyperkalemia and cardiac arrhythmias
Following gamma irradiation and storage Decreases potassium in plasma or additive solution
Allergic or anaphylactic reaction Removes allergenic plasma proteins, whether or not they are
specifically identified
Intraoperative salvaged autologous blood Removes clots, cellular debris, and heparin and suspends shed
RBCs in saline solution
T-activation During ongoing hemolysis, reduces further hemolysis by
eliminating IgM anti-T present in normal donors’ plasma
Paroxysmal nocturnal hemoglobinuria No longer recommended
The potassium concentration in plasma of CPDA-1 RBC com- in the supernatant, it does not arrest the process. From the
ponents increases from 4.2 mmol/L to nearly 80 mmol/L during time irradiated RBC components are stored following wash-
the 35-day storage period.6 For a 1-kg newborn, a conventional ing, potassium concentration increases in a time-dependent
transfusion (15 mL/kg) of 42-day stored AS-1 RBCs (hematocrit fashion, which is more rapid than for nonirradiated RBC
of 80%) may have a potassium concentration as high as 50 mEq/ components. The potassium concentration in irradiated and
L, but the total potassium content is only 0.15 mEq. While that washed RBC components may reach a critical concentration
relatively high concentration of potassium is tolerable and safe of 5 mEq/L after only 3 hours of storage at 4°C, compared
if the component is transfused slowly (over 2 to 4 hours), the with 6 hours for nonirradiated washed RBC components.17
same component could present a potentially fatal acute potas-
Allergic or Anaphylactic Reactions
sium load if transfused rapidly in a small child or susceptible
adult. Rapid transfusions of stored RBC components in such Mild, first-time allergic transfusion reactions (hives, flush-
patients may increase potassium acutely and, therefore, require ing, pruritus) often seem to be product-related, rather than
washed or freshly collected RBC components. Typical situa- patient-related. These reactions typically respond to antihis-
tions include exchange transfusions, extracorporeal membrane tamines and do not recur when additional plasma-containing
oxygenation (ECMO), cardiopulmonary bypass, or solid organ blood components are transfused. In contrast, some chroni-
transplantation.8–10 Washed RBC components should also be cally transfused persons develop more serious and generalized
considered for large-volume (>25 mL/kg) or rapid transfusions allergic reactions (bronchospasm, rash, nausea, vomiting, or
in newborns and other small children with renal failure, hyper- diarrhea) that are not prevented by, or respond to, treatment
kalemia, or severe acidosis. with antihistamines. Transfusing washed RBC components 29
Some physicians are also concerned that the doses of usually prevents allergic reactions in such patients.
certain additives in extended-storage solutions (AS-1, AS-3, If acute allergic reactions progress to life-threatening 411
AS-5) may exceed known limits for safety for large-volume anaphylaxis, bronchospasm, and hypotension, a diagno-
transfusions in small children.11 These concerns derive from sis of an immunoglobulin A (IGA) anaphylactic reaction
the theoretical possibility that constituents in the storage should be considered.18 Conventionally, that diagnosis is
media could contribute to hyperosmolality, hyperglycemia, made by demonstrating IGA deficiency (by a highly sen-
hypernatremia, or hyperphosphatemia. For that reason, some sitive passive hemagglutination inhibition assay) and the
transfusion services routinely wash or volume-reduce RBC presence of anti-IgA (by passive hemagglutination assay).19
components stored in additive solutions for large-volume or However, we advise caution when interpreting results of
rapid transfusions in newborns or other small children. IgA hemagglutination assays because they may yield non-
specific results.20 The same hemagglutination assay that is
used to diagnose patients with IgA anaphylactic reactions
Transfusion of RBC Components following detected IgA deficiency and the presence of anti-IgA in
Gamma Irradiation 1:1200 asymptomatic healthy blood donors.19 Clearly, this
To prevent transfusion-associated graft-versus-host disease number of healthy persons at risk for an IgA anaphylactic
(TA-GVHD) in immunocompromised or other patients reaction greatly exceeds the incidence of IgA anaphylactic
who are at risk, transfusion services (or blood centers) rou- transfusion reactions in clinical practice. Nonetheless, if
tinely gamma-irradiate blood components with a dose of the diagnosis of IgA deficiency is made, RBCs and platelets
2000 to 3000 cGy. Although gamma-irradiation is highly can be washed with high volumes of saline (4–6 L) to effec-
effective for preventing TA-GVHD, irradiation has adverse tively remove plasma antibodies. Washed RBC and platelet
effects on RBC survival, plasma hemoglobin concentration, components have been transfused successfully in persons
and RBC ATP.12–15 Plasma potassium concentration nearly with IgA deficiency and anti-IgA who had a history of ana-
doubles during the 48 hours after RBC components are irra- phylactic transfusion reactions when blood components
diated with 3000 cGy.12 Gamma-irradiation increases pas- from IgA-deficient donors were not available.21,22 Other
sive permeability of RBC membrane lipid bilayers, resulting severe generalized reactions to plasma-containing blood
in a progressive, reciprocal increase in intracellular sodium components that may be abrogated by washing compo-
and decrease in intracellular potassium.13 AABB Standards nents prior to transfusion include transfusion-related
requires RBC components to outdate no longer than 28 acute lung injury (TRALI), leukocyte-mediated cytokine
days from the date of irradiation.16 While washing irradi- reactions, and anaphylaxis in persons with ahaptoglo-
ated RBC components reduces potassium concentration binemia and antihaptoglobin.23
BLOOD BANKING
with plasma-reduced components may be necessary for
Intraoperative Salvage of Shed RBCs
infants with T-activation and ongoing hemolysis.39
Intraoperative salvage of shed RBCs has been used for more In patients with T-activation, some physicians recom-
than four decades to reduce the number of allogeneic blood mend completely avoiding transfusions of fresh frozen
transfusions in trauma and surgery associated with large- plasma, platelets, or other plasma-containing blood products.
volume blood loss.24–26 Typically, shed blood is aspirated Others question the clinical importance of T-activation and
from the sterile surgical field, anticoagulated using a hepa- recommend that if plasma-containing components are indi-
rin-saline solution, and washed with 0.9% sodium chloride cated, they should not be withheld.41 Recommendations,
in an automated cell salvage machine. The effluent contain- which are based on case reports, are conflicting and there are
ing clots, leukocytes, platelets, cellular debris, and heparin neither evidenced-based guidelines nor results of a random-
is discarded. Saline-washed concentrated autologous RBCs ized, controlled clinical trial to direct practice. The authors
are returned to the patient. When suitable wash volumes are concur with the preponderance of clinical evidence, which
used, activated coagulation factors, cytokines, and heparin favors not withholding plasma-containing blood compo-
are substantially reduced.27–32 nents when they are indicated. The authors do not recommend
An alternative method of salvaging shed autologous RBCs that transfusion services, which routinely use murine or other
involves collecting heparinized sanguineous drainage from monoclonal blood typing reagents, consider adding human
the chest, large joints, or other sites that is returned directly plasma–derived typing reagents as a screening procedure to
(unwashed) from collection canisters.33–36 Postoperative detect T-activation.
infusion of unwashed filtered blood has been reported to
be effective and without clinically relevant complications Paroxysmal Nocturnal Hemoglobinuria
in orthopedic patients.33,34 However, this method is contro- PNH is an uncommon stem cell disorder manifested by
versial, because such drainage may contain procoagulant complement-mediated hemolytic anemia, thrombophilia,
material, variable amounts of anticoagulant, and unsterile and marrow failure.42–44 The mainstay of managing the
debris.35 Even saline-washed salvaged RBCs collected under typical Coombs-negative, treatment-resistant, intravascular
optimal conditions may contain residual biologically active hemolytic anemia is RBC transfusions.45 Promising results
materials capable of causing increased vascular permeability, have been reported for managing hemolysis using eculi-
acute respiratory distress syndrome, or disseminated intra- zumab, a humanized monoclonal antibody to complement
vascular coagulation.36,37 These complications, which have C5.46 However, the role for this new treatment has yet to be
been described as the “salvaged blood syndrome,” can be defined.
averted by standardizing aspiration methods, controlling In 1948, Dacie reported that blood transfusions exac-
saline wash volumes, and monitoring for an abnormal accu- erbated complement-mediated intravascular hemolysis in
mulation of debris on the inner wall of the rotating bowl.36 patients with PNH, increasing hemoglobinuria and hemo-
globinemia.47 As a consequence, some hematologists request
T-Activation of RBCs
II washed RBC components for routine transfusions in patients
Immune-mediated hemolysis has been reported follow- with PNH. Admittedly, post-transfusion hemolysis is rare in
412 ing transfusion of plasma-containing blood components patients with PNH when washed RBC components are trans-
to patients whose RBC T-crypt antigens have been exposed fused,47 although it has been reported.48,49 However, it should
by bacterial infection.38–40 T-activation occurs when bac- be noted that most PNH patients do not experience hemoly-
terial neuraminidase removes N-acetyl neuraminic acid sis after RBC transfusions, even when whole blood is trans-
and exposes RBC T-crypt antigens. Exposed T-crypt anti- fused.50 A retrospective review of 23 PNH patients who were
gens bind with IgM anti-T, a normal constituent of adult transfused with a total of 556 units of whole blood, packed
plasma, resulting in RBC agglutination (polyagglutination) RBCs, leukocyte-poor RBCs, washed RBCs, frozen RBCs,
and hemolysis. T-activation has been reported to occur in and intraoperatively salvaged RBCs, identified only one case
a wide range of bacterial infections, including necrotizing of post-transfusion hemolysis.51 This case occurred after
enterocolitis, septicemia, hemolytic uremic syndrome, and transfusion of group O whole blood to a group AB-positive
Streptococcus pneumoniae infection. T-activation should be recipient. This specific serologic incompatibility is similar to
suspected in children who have the onset of intravascular that in the 1948 case reported by Dacie, which is the origin
hemolysis following transfusion of plasma-containing blood of the practice of washing RBC components before transfu-
components. sion to patients with PNH.47 Based on this large-scale study,
T-activation has been diagnosed less frequently in recent a supporting opinion from the PNH Interest Group,52 as well
years because many hospital transfusion services now use as the authors’ own uneventful personal experiences trans-
non-plasma-containing monoclonal typing reagents instead fusing patients with PNH with unwashed AS-1 RBCs, we
of traditional plasma-derived blood typing reagents (anti-A, believe that transfusing washed RBC components to PNH
-B, and -A,B). Previously, some cases of clinically unrecog- patients is unwarranted.
nized polyagglutination due to T-activation were detected
during routine compatibility testing using plasma-derived Methods for Preparing Washed Red
ABO typing reagents. Discrepancies in forward and reverse
Blood Cells
ABO typing results caused by polyagglutination or in vitro
hemolysis suggested T-activation. T-activation is confirmed RBC components may be washed manually using a refrig-
by specific agglutination tests using Arachis hypogaea and erated centrifuge or, more commonly, an automated cell
Glycine soja lectins. To prevent further hemolysis in patients washer. Washing an RBC component using 1 to 2 L of 0.9%
who have active hemolysis and polyhemagglutination, RBC sodium chloride removes approximately 99% of plasma pro-
(or platelet) components should be washed using 0.9% teins, electrolytes, and antibodies but may result in loss of
sodium chloride before transfusion. Exchange transfusion up to 20% of the RBCs depending on the protocol.53 Rarely,
Methods for Preparing Volume-Reduced
Table 29–2 Clinical Indications and Rationale for Washed Platelet Components
Allergic or anaphylactic reactions Removes allergenic plasma proteins, whether or not they have been
specifically identified
Febrile nonhemolytic transfusion reactions May be effective in some adults but unproved in children
Neonatal alloimmune thrombocytopenia Removes maternal alloantibodies from platelet components
collected from the mother, whether or not serologic
specificity has been identified
Out-of-group platelets Eliminates risk of hemolysis due to anti-A or -B when compatible
platelet components are not available for susceptible progenitor
cell transplantation or newborn recipients
T-activation During ongoing hemolysis, reduces further hemolysis by eliminating
IgM anti-T present in normal donors’ plasma
Paroxysmal nocturnal hemoglobinuria No longer recommended
WASHED AND VOLUME-REDUCED BLOOD COMPONENTS
acute intravascular hemolysis may occur (e.g., when ABO- gation of stored platelet components (secondary volume
incompatible apheresis platelet components from donors reduction).79 Additional concentration of standard platelet
who have high-titer anti-A/A,B have been transfused to components is controversial.
A1 recipients).71,72 As clinical practice shifts from transfus-
ing pools of random donor platelet concentrates (where a Hypervolemia
high-titer unit is diluted) to transfusing more single-donor A typical argument to support the availability of such
(apheresis) platelet components,71 more cases of hemolysis “super- or hyper-concentrated” platelet components for
may be identified. Some transfusion services perform anti-A/ newborns is that a standard 50-mL unit of random donor
anti-A,B titers on group O single-donor platelet components platelet components represents more than half the blood
prior to an “out-of-group” transfusion.73 However, there is volume of a 1-kg infant and may precipitate circulatory
lack of agreement as to what titer is clinically relevant and overload.80 In vitro studies of recentrifuged, volume-reduced
whether IgM or IgG antibody is more significant.73 Although platelet components demonstrated satisfactory viability, and
washing of out-of-group platelet components removes in vivo studies have demonstrated satisfactory survival of
incompatible plasma, the process may decrease platelet func- 51r-labeled concentrated platelets and post-transfusion
tion,74 requires considerable skill and experience,59 and is not platelet count increments.81,82 Arguments against routinely
practiced routinely. reducing the volume of standard platelet components focus
In some hospitals, the practice of washing (or volume- on the reliability of transfusing 5 to 10 mL/kg of standard
reducing) platelet components to prevent anti-A/A,B or platelet concentrates or apheresis platelets to increase the
anti-B hemolysis is limited to transfusions in progenitor platelet count to >100 × 109 L.83 A secondary concern is that
cell transplant (PCT) recipients and newborns. In PCT while satisfactory concentration of platelets may be achieved
recipients, transfusion support between myeloablation and by experienced research technologists in a limited study,
engraftment may require washed or volume-reduced plate- technologists working in routine transfusion service opera-
let components to avoid anti-A/A,B– or anti-B–associated tions may not be able to achieve that level of quality control.
hemolysis. Plasma in platelet concentrates may represent a A transfusion service undertaking recentrifugation of plate-
significant percentage of a newborn’s blood volume, plac- lets must take special precautions to control for platelet loss,
ing newborns at higher risk than adults for ABO-related clumping, and dysfunction caused by additional handling.78
hemolysis. Some hospitals routinely wash or volume-reduce The AABB Pediatric Hemotherapy Committee pub-
out-of-group platelet components for newborns. lished the results of a national survey of neonatal transfusion
practices, noting that “because of the potential for harm, insti-
tutions transfusing volume-reduced platelets should moni-
Methods for Washing Platelet Components
tor both the quality of the final product (i.e., the number of
Methods have been developed for manual and automated platelets, degree of clumping, and function) and in vivo effects
washing using blood cell processors.75–77 Washed plate- such as post-transfusion increment in platelet count and
let components have a threefold increase in spontaneous adverse reactions, including altered vital signs and pulmonary 29
activation, as well as impaired ADP-induced aggregation, distress.”84 The Committee made the observation that 61% of
compared with unwashed platelet components.74 Washing respondents reported a final desired volume of 10 to 15 mL, 415
platelets requires skill and experience to minimize platelet and an additional 30% desired 18 to 25 mL, both volumes
loss, activation, and clumping.78 Automated cell processors being within the range likely to achieve the targeted platelet
are more likely to deliver a consistent result and are recom- count increase using an unmodified platelet concentrate.
mended. Kalman and Brown developed a protocol using the
Recurrent Febrile Nonhemolytic Transfusion
IBM 2991 Blood Cell Processor, which removed a mean of
Reactions
99.6% of plasma proteins following a 1500-mL wash using
0.9% sodium chloride. That cell processor is currently In adults, the incidence of febrile nonhemolytic transfu-
marketed as the COBE 2991 Blood Cell Processor (COBE sion reactions (FNHTRs) is decreased when plasma-reduced
Laboratories, Lakewood, Colo.). platelet components are transfused.85,86 This effect, similar to
Washed platelet components should be used immediately that of prestorage leukoreduction, is attributed to decreased
after washing owing to lack of substrates to support platelet leukocyte-derived proinflammatory cytokines, which accu-
metabolism during storage. In vitro studies of new additive mulate in plasma of stored platelet components.86,87 In chil-
solutions for washing platelets suggest that a postwash dren, a randomized, prospective, crossover study compared
storage duration of up to 48 hours may be feasible.79 the frequency of acute reactions to post storage plasma-
removed platelet components and to standard platelets.88
Study platelet components were prepared by removing
VOLUME-REDUCED PLATELET plasma from stored components and replacing it with an
COMPONENTS equal volume of ABO-compatible fresh frozen plasma (FFP).
While there was a trend toward lower frequency of FNHTRs
Indications for Volume-Reduced Platelet with post storage plasma removal, the results were not statis-
tically significant.88 The investigators could not explain why
Components
FNHTRs occurred less frequently in children than in adults.
There are two indications for transfusing volume-reduced On the basis of these findings, the authors do not recom-
platelets: to prevent circulatory overload and to lessen the mend removing plasma (with or without replacement using
volume of any potentially adverse constituents in the plasma FFP) as a method to reduce the incidence of FNHTRs in
of platelet components. Volume-reduction can be achieved children.
either during collection and processing of collections (pri- We believe that routine recentrifugation of platelet com-
mary volume reduction) or subsequently by recentrifu- ponents is neither necessary nor prudent. As stated by the
BLOOD BANKING
Table 29–3 Methods for Preparation of Secondary Volume-Reduced Platelet Components
In Vitro In Vivo
Platelet Platelet
Volume of Platelet Count Function Efficacy
Starting Final Platelet of Final Platelet Studied Studied after
Platelet Component Component after Volume Volume
Method Component N (mL) (×109/mL) Reduction Reduction
Modified from Schoenfeld H, Spies C, Jakob C. Volume-reduced platelet concentrates. Current Hematol Rep 2006;5:82–88.
II
418
Chapter 30
Blood Management: Conservation, Salvage,
and Alternatives to Allogeneic Transfusion
Beth Shaz
Figure 30–1 Example of guidelines for red cell transfusion. (From Garrioch M, Sandbach J, Pirie E, Morrison A, Todd A, Green R. Reducing red cell
transfusion by audit, education and a new guideline in a large teaching hospital. Transfus Med 2004;14:27.)
Aprotinin
>200 >50
XIIa FDPs
KIU/ml KIU/ml
Fibrin
Vasodilatation Fibrinogen
Figure 30–2 Aprotinin inhibition. (From Porte RJ, Hendriks HG, Slooff MJ. Blood conservation in liver transplantation: The role of aprotinin. J Car-
diothorac Vasc Anesth 2004;18(4 Suppl):31S–37S.)
Aprotinin has been used in other surgeries including
BLOOD MANAGEMENT
Most of the data on antifibrinolytics is in cardiac surgery
orthopedic surgery, liver transplantation, liver resection, patients. There is a debate about its use in primary surgeries,
and vascular surgery.8 In liver transplantation, studies have where blood loss should not be substantial and the drug may
demonstrated a decrease in transfusion with no difference therefore be of limited benefit. The second most common
in mortality.11,12 In major orthopedic surgery, there are use in the literature is in orthopedic surgery, where it can be
conflicting data about the usefulness of aprotinin, and the beneficial in surgeries with substantial blood loss. Few trials
studies consist of small numbers of patients with varying in orthotopic liver transplantation have shown a decrease in
procedures.13–15 Small trials have shown decreased transfu- transfusion, but there are multiple case reports of thrombotic
sion rates with aprotinin versus placebo in orthognathic sur- complications with its use.28
gery16 and major thoracic surgery.17,18 In summary, aprotinin
has not been extensively studied in noncardiac surgery but
may be beneficial if substantial blood loss is predicted.
Recombinant Factor VIIa (rFVIIa)
Factor VIIa binds to tissue factor to activate factors IX and X,
e-Aminocaproic Acid (EACA) and Tranexamic
thereby activating factor V, which converts prothrombin to
Acid (TXA)
thrombin. Thrombin activates platelets and factors VII, V, and
TXA and EACA are synthetic lysine analogs that inhibit XI. This creates a thrombin burst, which converts fibrinogen
fibrinolysis. These drugs block the lysine-binding site on to fibrin for clot formation (Fig. 30–3).29 Recombinant acti-
the plasminogen molecule, which inhibits the formation of vated factor VII (rFVIIa) is licensed for use to control bleeding
plasmin and therefore inhibits fibrinolysis. In the Cochrane in hemophilia with inhibitors, but there are case series demon-
database of systematic reviews, TXA, but not EACA, statisti- strating its success in controlling severe or refractory bleeding
cally significantly reduces the rate of allogeneic transfusion.8 in nonhemophilic patients as well. One randomized controlled
There are only a handful of trials with small numbers of study in patients undergoing retropubic prostatectomy demon-
patients using EACA to decrease the rate of transfusion, and strated a reduction in blood loss, number of units transfused,
a few more trials using TXA. A small study demonstrated a and likelihood of transfusion.30 rFVIIa did not reduce blood
small but statically significant decrease in transfusion rate loss or transfusion in trauma patients undergoing pelvic recon-
in scoliosis repair in EACA versus placebo.19 A similar study struction.31 A third trial of rFVIIa versus placebo in patients
with TXA showed decreased blood loss, with no difference in receiving liver resection demonstrated no change in transfusion
transfusion.20 The use of TXA versus placebo reduced the total or blood loss.32 In a recent trial of its use in intracranial hemor-
number of packed red cell units transfused in patients receiv- rhage, there was an increased rate of thrombosis, especially as
ing coronary artery bypass grafting on and off pump.21 No dif- the dose was escalated.33 rFVIIa may decrease blood loss and
ference in transfusion rates in total knee replacement was seen decrease transfusion needs in cases of large blood loss but car-
with the use of TXA or aprotinin.22,23 Multiple small trials with ries a risk of thrombosis and a high cost.
either TXA or EACA for a variety of surgeries have demon-
strated conflicting data on their ability to decrease allogeneic 30
red cell transfusion, but in general more benefit has been seen
Fibrin Sealant
in procedures associated with greater blood loss. Fibrin sealant is the combination of thrombin and fibrinogen 421
mixed with calcium to form fibrin, which is used as a topical
hemostatic agent. Products may contain antifibrinolytics (apro-
Comparison of Antifibrinolytics
tinin) to reduce fibrinolysis or factor XIII to increase strength of
Antifibrinolytics decrease the need for allogeneic transfu- the clot (Fig. 30–4).34 A variety of commercially and individually
sion. Aprotinin is the most expensive but the most frequently produced fibrin sealants are available. Bovine thrombin prod-
studied of the antifibrinolytics. The Ischemia Research and ucts are commonly used but have a risk for allergic reactions
Education Foundation recently published an observational and antibody formation. Antibodies to bovine thrombin cross-
study involving 4374 patients undergoing cardiac revascular- react with human factor V, leading to factor V deficiency and
ization. They compared aprotinin, aminocaproic acid, and risk for hemorrhage.35 In contrast, human thrombin, though
tranexamic acid with no agent and discovered that aprotinin virally inactivated, has a small risk for transfusion-transmit-
was associated with serious end-organ damage involving the ted disease. Currently, two fibrin sealant products consisting
heart, brain, and kidney. There was a dose-response relation- of human fibrinogen, human thrombin, calcium chloride, and
ship for aprotinin with increased death, renal dysfunction, aprotinin (bovine) are commercially available in the United
and cardiovascular events at a higher dose. All three antifi- States. Alternatively, automated devices exist to produce fibrin
brinolytics reduced blood loss equally, but aminocaproic sealant from autologous plasma.36 Another option is autolo-
acid and tranexamic acid were not associated with serious gous fibrin glue prepared from the cryoprecipitated portion of
end-organ damage.24 In a recent study in cardiac surgery autologous plasma; after thawing, this material is mixed with
patients of aprotinin versus tranexamic acid versus control, bovine thrombin immediately before application to the surgi-
aprotinin significantly decreased the likelihood of alloge- cal field. A disadvantage of fibrin sealant is the time it takes to
neic transfusion.25 A similar study in coronary artery bypass prepare, especially autologous products, and also the time for
patients receiving aspirin showed no significant difference the clot to form.34 Fibrin sealants have been studied in a number
between aprotinin, TXA, and EACA on blood loss or trans- of surgeries, including prostatectomy, lung resection, liver resec-
fusion.26 In the Cochrane review, no clinical significant dif- tion, carotid endarterectomy, cardiac surgery, and orthopedic
ference between TXA and aprotinin on transfusion could be surgery.34 In the Cochrane database of systematic reviews, fibrin
found, but there is significant heterogeneity in the trials.8 In sealants reduced allogeneic transfusion and decreased intraop-
orthotopic liver transplantation, use of TXA versus aprotinin erative and postoperative blood loss, but most trials were small
showed no difference in transfusion or mortality.27 and unblinded, resulting in less reliable data.37
BLOOD BANKING
Figure 30–3 Role of FVIIa in coagulation. A, Coagulation is initiated when coagulation proteins and platelets come into contact with the extravasculature.
Factor VII binds to tissue factor, is activated, and activates both factor IX and factor X. The factor Xa forms a complex with factor Va on the tissue fac-
tor–bearing cell and activates a small amount of thrombin. This thrombin acts to amplify the initial coagulation signal by activating platelets, causing release
of factor V, activating factor V, cleaving factor VIII and releasing it from vWF, and activating factor XI. In the propagation phase, factor IXa, formed by factor
VIIa/tissue factor or generated on the platelet surface by factor XIa, forms a complex with factor VIIIa to activate factor X on the platelet surface, where,
in complex with factor Va and in the presence of prothrombin, it is protected from inhibition. Formation of the factor Xa/Va complex results in a burst of
thrombin generation. B, In hemophilia, the initiation and amplification phases proceed normally. The propagation phase is absent or significantly decreased
because factor Xa cannot be generated on the platelet surface. High-dose factor VIIa acts to partially restore platelet surface factor Xa generation so that
factor Xa/Va complex formation proceeds and the propagation phase is improved relative to the hemophilic state. (From Roberts HR, Monroe DM, White
GC. The use of recombinant factor VIIa in the treatment of bleeding disorders. Blood 2004;104:3858–3864.)
Preoperative Erythropoietin
II BLOOD CONSERVATION
Erythropoietin is a 165-amino-acid glycoprotein hormone,
422 which is synthesized in fetal liver and adult kidney as a response
Autologous Blood
to hypoxia to stimulate erythropoiesis. Erythropoietin recep-
Autologous blood can be collected from a patient in advance tors have been identified on both hematopoietic and other
of anticipated blood loss (preoperative donation) or at the (brain, heart, liver, and retina, vascular endothelium, gastro-
start of the procedure (acute normovolemic hemodilu- intestinal tract, and reproductive tract) tissues. Therefore,
tion); in addition, shed blood can be salvaged for reinfusion erythropoietin likely mediates additional activities, such as
both during surgery and in the postoperative period (Table having a protective antiapoptotic effect.45 Erythropoietin
30–1). As early as the late 19th century, the concept of using corrects anemia caused by renal failure, cancer, cancer ther-
a patient’s own blood as a source of transfusable RBCs apy, and HIV.46 Erythropoietin is contraindicated in patients
was born of necessity because allogeneic transfusions were with uncontrolled hypertension. Adverse events associated
fraught with immunologic risks. Early authors proposed the with its use include thrombotic events, hypertension, sei-
recovery of blood shed during childbirth, ectopic pregnancy, zures, and rare cases of pure red cell aplasia.6 Erythropoietin
and splenectomy.38,39 With the discovery of the ABO blood can be used preoperatively, with or without preoperative
group system, the development of approaches for storage of blood donation, or with acute normovolemic hemodilution
blood and preparation of components, and the need during in elective surgery.46 Two to four weeks is necessary for ade-
two world wars for a large and immediately available blood quate erythropoietin-stimulated erythropoiesis to occur.46
supply, interest in autologous blood waned. The appearance Depending on the estimated blood loss, patient’s blood vol-
in the early 1980s of the human immunodeficiency virus ume, and amount of time prior to surgery, erythropoietin
(HIV) as a transfusion-transmissible complication of alloge- may be useful.
neic transfusion fostered a recurrence of interest in autolo- Because erythropoietin increases the preoperative hemo-
gous blood techniques. As the risk of transfusion decreased globin, it can be used in mildly anemic patients preopera-
and understanding of the risks and costs of autologous dona- tively to decrease the risk of transfusion in elective surgeries
tion increased, there has been a decline in the utilization of with moderate blood loss.47 Erythropoietin given periopera-
preoperative blood donation.40 Blood collection statistics tively decreases the number of intraoperative and postop-
bear out this trend: donations of blood for autologous use erative transfusions and results in improved 1-year survival
increased from 28,000 units in 1982 to a peak of 1,117,000 rate.48 One study demonstrated that patients receiving daily
units in 1992 and have currently decreased to 651,000 units erythropoietin (300 IU/kg or 100 IU/kg subcutaneously
in 1999 (Fig. 30–5).41–44 beginning 10 days prior to orthopedic surgery for a total of
BLOOD MANAGEMENT
Figure 30–4 Schematic drawing of the coagulation cascade and composition of components of the fibrin tissue adhesive (FTA). (From Levy O,
Martinowitz U, Oran A, Tauber C, Horoszowski H. The use of fibrin tissue adhesive to reduce blood loss and the need for blood transfusion after 30
total knee arthroplasty. A prospective, randomized, multicenter study. J Bone Joint Surg Am 1999;81:1580–1588.)
423
15 doses) versus placebo perioperatively had significantly Preoperative erythropoietin has also been shown to decrease
decreased transfusion rates. This improvement was signifi- allogeneic transfusion in anemic patients receiving elective
cant even in patients with a preoperative hemoglobin level open heart surgery or gastrointestinal cancer resection.51 In
above 13 g/dL.49 An earlier study with erythropoietin given anemic patients who are not eligible to donate autologous
daily perioperatively to patients undergoing elective hip blood preoperatively, erythropoietin decreases the likelihood
replacement showed a decreased transfusion rate with eryth- of allogeneic blood transfusion.
ropoietin compared with placebo, but the patients who ben- Iron supplementation maximizes the benefit of erythro-
efited most had a hemoglobin below 13.5 g/dL at baseline.50 poietin used either orally or intravenously.52 Other major
*
Need not conclusively established.
†
Intraoperative salvage is unnecessary when a tourniquet is used; postoperative salvage may be of value in cementless procedures.
‡
Hypothetical risk of cancer spread after transfusion of intraoperatively salvaged blood.
§The safety of blood containing amniotic fluid has not been conclusively established.
BLOOD BANKING
16000 by human and animal data, some authors also believe that
allogeneic transfusions induce immunosuppression in the
14000
recipient and that autologous blood may not.56 Only prospec-
Units collected (x1000) 12000 tive observational studies have been performed to answer the
question of allogeneic versus autologous blood transfusion
10000 and the risk of postoperative infection.57–59 Other advantages
8000 include the stimulation of erythropoiesis in the repeatedly
bled autologous donor, which could speed recovery from
6000 postoperative anemia.60 Intraoperatively salvaged RBCs are
4000 typically transfused back to the patient within a few hours
after collection, thereby precluding the acquired membrane
2000 defects and enzyme deficiencies (2,3-diphosphoglycerate
0
and adenosine triphosphate) that develop during refrigera-
tion (the “storage lesion”).61
1980
1982
1984
1986
1987
1989
1992
1994
1997
1999
Disadvantages
Year
Autologous techniques are not without drawbacks. Although
Total collections Autologous collections an ideal preoperative donation schedule should allow suffi-
cient time for compensatory erythropoiesis to occur, some
Figure 30–5 Units of total whole blood or red cell collections compared individuals develop donation-induced anemia at the time
with units of autologous whole blood or red cell collections over time. of admission for surgery. Concerns regarding complications
of donation in patients with underlying cardiac disease have
been raised62 although it can be difficult in such settings to
causes of hyporesponsiveness to erythropoietin include folic distinguish between an untoward effect of blood letting and
acid, vitamin B12, vitamin C deficiency, infection, inflamma- the natural history of the baseline condition. Clerical errors,
tory states, and chronic blood loss.53 including release of the wrong unit of blood, still occur.63
Erythropoietin has also been used in combination with Finally, although it is rare, autologous blood may itself pose
preoperative autologous blood donation to increase the infectious risks. Yersinia enterocolitica, a common cause of
number of units donated or with other blood conservation community-acquired diarrhea, persists in the bloodstream
strategies, as discussed in the sections on Indications for for many weeks after infection and can grow at refrigerator
Autologous Whole Blood Collection in Combination with temperatures; an individual who donates blood during this
Erythropoietin and Acute Normovolemic Hemodilation in time period can be made severely ill by later transfusion of
Combination with Erythropoietin. the component.64
II
Preoperative Iron Supplementation Costs of Autologous Blood Techniques
424
Correcting iron deficiency anemia prior to surgery should As the allogeneic blood supply has become safer, more
decrease the likelihood of the need for allogeneic transfu- attention has been focused on the costs associated with
sion.54 Iron therapy is not beneficial in non–iron-deficient autologous transfusion techniques. Costs come from
patients without concomitant erythropoietin.52 A prelimi- unused autologous collections (e.g., when a patient has
nary study shows a decrease in allogeneic blood transfusion donated enough blood to match the mean number of
with an accompanied reduction in morbidity-mortality rate components used by others undergoing the procedure but
and length of hospital stay with the use of intravenous iron requires less). This problem is magnified by overcollection
preoperatively in patients undergoing displaced subcapital and unnecessary utilization (e.g., in plastic surgery) and by
hip fracture repair.55 the extra work involved in deviation from routine, large-
scale allogeneic collection practices.65 In one hypothetical
cost-utility analysis of patients undergoing primary elective
hip replacement, the cost effectiveness of autologous trans-
Advantages and Disadvantages of fusion per quality-adjusted life-year (QALY) was estimated
Autologous Blood Transfusion at an extremely high 3.4 million. However, if allogeneic
transfusions were assumed to increase the risk of postop-
Advantages erative bacterial infection, a possibility suggested by some
In the past, autologous donations have been viewed as a workers,66–68 the cost of using autologous blood fell to less
strategy for obtaining blood to augment short allogeneic than $50,000 per QALY, and the procedure became domi-
supplies. More relevant in current times is the use of autolo- nant (cheaper to use than allogeneic blood) as the infection
gous blood to eliminate the risk of transfusion-transmitted risk rose.69 Some authors have suggested that autologous
microorganisms, in particular HIV and hepatitis viruses. blood could be kept cost effective by streamlining collec-
The institution of increasingly sensitive tests for such infec- tion and processing, including forgoing infectious disease
tions, including nucleic acid testing, has softened the impact testing of autologous donors.70,71 Reducing costs through
of this argument; however, the risk of future incursions into the intentional use of autologous blood by a recipient other
the blood supply by new infectious agents remains a threat. than the donor (“crossover”) is not recommended. Only
Immunologic complications, such as antibody-mediated 30% of collections are typically eligible for allogeneic use,
hemolysis and leukocyte-associated febrile reactions, are and the costs, complexity, and risk of error of the transition
eliminated with the use of autologous blood. Supported all serve to negate the value.72
Preoperative Autologous Donation
BLOOD MANAGEMENT
etin to preoperative autologous donation was demonstrated
in a three-arm study of preoperative erythropoietin alone,
Indications for Autologous Whole Blood erythropoietin plus preoperative autologous donation, ver-
Collection sus preoperative autologous donation alone in patients
The decision to use preoperative autologous blood donations undergoing total joint arthroplasty with an initial hemoglo-
should be predicated on the type of surgery, the amount of bin of ≤14.0 g/dL. Allogeneic transfusion rates were lowest
time available for donation and hematopoietic reconstitu- in patients who had received erythropoietin and predonated
tion, the patient’s hematocrit, and the predicted vigor of autologous blood (11% vs 28% or 33%).93
the erythropoietic response to donation. Patients planning When large blood loss is expected and multiple units of
to undergo elective orthopedic surgery are ideal candidates autologous blood are needed, erythropoietin increases the
for autologous transfusion, because they require moderate number of preoperative autologous blood units that can be
amounts of blood during and immediately after surgery, and donated.47 In one study, orthopedic patients with a baseline
they typically have sufficient time prior to surgery to make hematocrit of ≤39% were requested to donate six weekly
multiple donations.73–75 Open heart surgery and vascular autologous units. Erythropoietin (six weekly doses) versus
surgery are other procedures in which autologous blood col- placebo significantly increased the number of units donated
lections have led to reduction or elimination of allogeneic (4.5 vs 3.0), resulting in fewer patients who received erythro-
blood use.6 The use of preoperative autologous blood dona- poietin being transfused allogeneic blood (26% vs 38%).94 An
tion has also been reported for a variety of other surgical pro- earlier small study in women undergoing total hip replace-
cedures, including radical prostatectomies, hysterectomies, ment with hematocrit <40% showed donation of more units
and other gynecologic procedures, gastrointestinal surgery, with erythropoietin versus placebo (4.1 vs 2.8 units) resulted
and neurosurgery.76,77 Donors of bone marrow for trans- in less transfused allogeneic blood (0.4 vs 1.2 units). This study
plantation undergo multiple iliac crest aspirations and may had an aggressive donation schedule, with one donation every
develop a moderate anemia; advance donation of autologous 3 to 4 days for 3 weeks, and a high transfusion rate of 4 units
blood forestalls the need for allogeneic transfusions.78 As on average.95 A similar study in spine surgery showed benefit
surgical techniques change, the need for blood transfusions in erythropoietin with autologous donation, evidenced by a
may be affected, and in such instances the role of autologous decrease in allogeneic blood and hospital stay.96 A single small
blood should be reexamined. Radical prostatectomy is a case study in patients with iron deficiency anemia (hematocrit ≤
in point: ten years ago many hospitals encouraged patients 33%) and gastrointestinal cancer showed an increase in autol-
undergoing the procedure to donate autologous blood,79,80 ogous unit donation and a trend toward decrease of alloge-
but today in some hands fewer than 2% of patients require neic blood transfusion with the use of intravenous iron and
blood, and autologous blood donations may be superflu- erythropoietin preoperatively versus intravenous iron alone.97
ous.81,82 Blood transfusion is also rarely necessary in plastic A small study in patients undergoing coronary artery bypass
surgical procedures, and the collection of autologous blood surgery demonstrated that erythropoietin increased the num-
in such cases is often frivolous.83 ber of autologous units collected and decreased the rate of 30
Autologous blood can be collected from a pregnant woman allogeneic blood transfusion.98
for use during childbirth. The expanded maternal blood vol- Other studies have demonstrated an increase in autolo- 425
ume contributes a substantial safety cushion, and the donation gous units donated preoperatively with the use of eryth-
process appears safe for both mother and fetus.84 However, the ropoietin but no significant decrease in allogeneic blood
use of blood transfusions in uncomplicated pregnancies is low transfusion. In patients undergoing hysterectomy, erythro-
(1%) even in patients undergoing cesarean section (3.3%).85–87 poietin increased the number patients able to donate 3 units
A role for autologous collections may exist for patients with in 2 weeks but did not decrease the transfusion of allogeneic
placenta previa, in which the likelihood of transfusion may blood.99 Patients undergoing resection of rectal cancer were
exceed 25%, and the condition is often identified with suffi- able to donate more red cells (4 units in 2 weeks) if treated
cient time for advance blood donations.88 In addition, autol- with erythropoietin, but there was no difference in autolo-
ogous blood has been collected from patients with unusual gous or allogeneic transfusions.100 Multiple studies show no
antibodies discovered during the pregnancy.89 use for erythropoietin in combination with preoperative
Long-term frozen storage of autologous RBCs in the autologous donation in nonanemic patients undergoing
absence of an anticipated transfusion episode is both inef- surgery, even with large blood needs.101–103
fective and expensive. An exception is the storage of blood Clearly determining which patients are most likely to ben-
by individuals with high-frequency or complex alloantibod- efit from erythropoietin with or without preoperative autol-
ies, for whom stockpiling of rare autologous units may be ogous donation depends on the estimated blood loss for a
beneficial. Even here, however, the likelihood that such blood surgical procedure, the patient’s ability to donate autologous
would be helpful is slim. To be of value, sufficient autologous blood, the amount of time available prior to surgery, the base-
blood would have to be available to meet the needs of an line hematocrit, and the patient’s blood volume. A single insti-
unexpected emergency; furthermore, delays in sending the tution’s experience with a tailored program of no preoperative
blood expeditiously to the hospital where it is needed and autologous donation in patients with hematocrits greater
in preparing units (thawed and washed free of the glycerol than 39%, autologous blood donation (2 weekly donations)
cryoprotectant) would make its use unwieldy. in patients with hematocrits between 37% and 39%, and
erythropoietin if the hematocrit was less than 37% (3 weekly
Indications for Autologous Whole Blood doses preoperatively) showed a decrease in transfusion (88%
Collection in Combination with Erythropoietin vs 57%), especially in patients with hematocrits greater than
Preoperative erythropoietin and preoperative autologous 39%, (majority of the patients).104 This study proved the con-
donation have the equivalent likelihood of allogeneic red cept of tailoring preoperative treatment based on a patient’s
cell transfusion.90–92 The superiority of adding erythropoi- baseline hematocrit.
BLOOD BANKING Indications for Other Autologous Components ion, blood may be comfortably collected from an autologous
Although autologous plasma is easily prepared from whole donor on a weekly schedule. Oral ferrous sulfate is com-
blood, little need exists for plasma in most elective surgery.105 monly prescribed (325 mg three times daily) although the
Autologous platelet-rich plasma can be prepared at the start amount absorbed may not be sufficient to counter the iron
of open heart surgery, using apheresis equipment before lost with the donations.128,129 In non–iron-deficient patients
bypass, to be returned to the patient after heparin rever- undergoing preoperative autologous blood donation, oral or
sal.106 Because thrombocytopenia or an acquired platelet intravenous iron versus no iron did not increase the number
defect can occur after blood passes through the membrane of units donated or decrease the need for allogeneic transfu-
oxygenator,107 the theoretical advantages of transfusing sion.130 The shelf life of refrigerated whole blood is limited
platelet-rich plasma should include improvement in hemo- to 42 days with current formulations of anticoagulant-
stasis and reduced transfusion requirements. Although initial preservative solutions, and a schedule for multiple donations
studies of this approach provided supportive data,106,108–110 is usually fits into this 6-week window. Alternatively, some
later prospective, blinded protocols were not able to dem- or all of the units can be frozen at −65°C, with glycerol as a
onstrate a reduction in blood use in either primary heart cryopreservative, for up to 10 years. Although frozen units
surgery or reoperations.111,112 In addition, the harvesting allow collections to occur over a longer period, the flexibil-
of platelet-rich plasma has been followed by intraoperative ity of utility at the time of surgery is affected: thawing and
heparin resistance, possibly owing to release of platelet factor deglycerolizing takes a few hours, and the thawed units have
4 and other procoagulants from platelets damaged during an outdate of 24 hours.
the collection.113 The Cochrane Database of systematic
reviews identified 19 trials of platelet-rich plasmapheresis in Intraoperative Autologous Transfusion:
mostly cardiac surgery patients for which allogeneic transfu- Blood Salvage
sion data were available. They concluded that platelet-rich
plasmapheresis reduces allogeneic transfusion, but there was Several techniques have been developed for the salvage and
heterogeneity in the studies and the majority was unblinded; reinfusion of blood lost during an operative procedure.
therefore, the procedure is not justified at this time.114 Interest in intraoperative salvage has been spurred by the
introduction of pumps, separation chambers for washing
RBCs, and increasing automation of the collection pro-
Collecting Autologous Blood
cess.131,132 The simplest approach—direct reinfusion without
Autologous blood donations are well tolerated by a variety washing—involves collection of blood under low vacuum
of ostensibly high-risk donors, including the elderly,76,115 pressure in a plastic bag seated within a hard outer canis-
children,116 pregnant women,117,118 and patients with ath- ter. An anticoagulant, usually citrate, is added. As soon as
erosclerotic coronary artery disease.119 One group reported the bag is full, or within 4 hours after the start of the collec-
an increased frequency of serious reactions among autolo- tion (to prevent bacterial growth), the contents of the bag
II gous donors at blood collection facilities, although this may are reinfused through a standard blood filter to the patient
reflect an intentionally conservative approach to patients (Fig. 30–6). RBCs shed into a surgical field, already poten-
426 (compared with the “normal” volunteers to which donor
centers are accustomed).120 A weekly phlebotomy schedule
fosters some degree of RBC regeneration before surgery (in
one study, a mean of 522 mL of RBCs donated over 3 weeks
resulted in a mean RBC production of 351 mL).121 However,
the most important medical problem associated with autol-
ogous donation is anemia developing during the collection
interval. When this occurs, it is typically as a result of mar-
ginal iron stores and insufficient erythropoietic response
(with little or no increase in serum erythropoietin levels),
probably because the hematocrit of most donors is not
allowed to fall below 30%.122 This situation may be improved
by the administration of recombinant human erythropoi-
etin.123 Many variables affect the response of blood donors
to this drug, including route of administration, adequacy of
iron stores, and method of iron supplementation (oral vs
parenteral).124,125 Especially in the United States, the expense
of recombinant human erythropoietin has limited its use
to situations in which autologous blood donation might
otherwise be difficult or impossible (e.g., in a patient who
is already anemic).94,95 An alternative approach employs
RBC apheresis—collection of two units of RBCs (without
plasma)—because each collection enhances the rate of com-
pensatory erythropoiesis.126 Other advantages of RBC apher- Figure 30–6 Equipment for direct reinfusion of perioperatively sal-
esis include increased time between donation and surgery, vaged blood without washing. A 600-mL plastic bag is seated within
which allows more time for compensation for the red cell the rigid plastic outer shell. A suction aspirator wand and filter are con-
nected to the bottom port; vacuum suction is connected to the top left
loss, and savings in time and cost.127 port. The top right port is connected to a blood filter for transfusion
Provided that the donor has satisfactory iron stores and to the autologous recipient. (From Solco Basle, Rockland, Mass., and
that bone marrow erythropoiesis can occur in a timely fash- Williams & Wilkins, with permission.)
BLOOD MANAGEMENT
tially damaged by their travail, are accompanied by activated the recipient this can result in hemoglobinemia and hemo-
coagulation factors and platelets, cellular debris and soluble globinuria (Table 30–2). Nevertheless, renal sequelae are
factors released from injured tissue cells, pharmaceuticals uncommon.136 The survival of chromium 51–labeled sal-
applied to the field, and irrigant solutions. Despite this sce- vaged cells is normal in most patients, presumably because
nario, salvaged blood has been reinfused directly into patients damaged cells are cleared during processing.137,138
with few untoward consequences. Coagulation abnormalities are often observed in recipi-
Alternatively, the contents of the bag can be washed with ents of large volumes of salvaged blood and include hypo-
saline. Devices that include a reservoir for collecting the sal- fibrinogenemia, elevated fibrin degradation products,
vaged blood and a centrifuge for washing are available (Fig. thrombocytopenia, and prolonged prothrombin and partial
30–7) to collect and process large volumes (e.g., 225 mL of thromboplastin times.139,140 In general, this clinical picture
RBCs with a final hematocrit of 50% in less than 3 min- is related to a combination of the characteristics of the sal-
utes).133 With these techniques, intraoperative blood salvage vaged blood and hemodilution in the recipient. After expo-
has become practical in situations in which blood loss may sure to serosal surfaces in the operative field, blood becomes
be extremely rapid, such as trauma or liver transplantation. depleted of coagulation factors and platelets; in the case of
Approximately half of the blood lost during a surgical pro- unwashed autologous blood, fibrin degradation products
cedure can be recovered; the rest is irretrievably absorbed by accumulate.141 Although the clinical picture can resemble
drapes and sponges or damaged during collection.134 that of disseminated intravascular coagulation, which in
Complications of intraoperative salvage are surprisingly theory might be initiated by phospholipids and other mate-
infrequent. The hematocrit of salvaged unprocessed blood rials released from damaged blood cells, no evidence exists to
is typically low because of a combination of dilution from support a cause-and-effect relationship.
irrigation fluids and some degree of mechanical hemoly- Other substances in salvaged blood include fat, fibrin,
sis.135 Free hemoglobin levels may exceed 1000 mg/dL, and in and microaggregates. Infusion of unprocessed blood has not
been shown to be harmful in either animals or humans, pos-
sibly because of the removal of most particulate matter by
standard blood filters.142,143 Pharmaceutical contaminants,
2
such as heparin, topical antibiotics, hemostatic agents, and
3 biologic substances such as tissue enzymes and hormones,
can be removed, but usually not completely, by wash-
ing.144,145 Bacterial contamination during the collection and
processing of autologous blood is inevitable owing to envi-
1
ronmental organisms such as coagulase-positive and -nega-
tive staphylococci, propionibacteria, and Corynebacterium
species. Administration of antibiotics to the patient reduces
the microorganisms but does not eliminate them,146 and 30
complete removal of bacteria after collection also is not pos-
sible, even when the washing solution includes antibiotics.147 427
8 7 4 There is no apparent clinical significance to such low levels of
contamination. Larger bacterial counts are of more concern.
Collection of blood from a contaminated site, such as that
associated with spilled intestinal contents, is probably con-
traindicated, although some authors have argued that, if no
other blood is available, such transfusions may be lifesaving
and worth the risk.148,149 Tumor cells have also been found
in salvaged blood during cancer surgery; their malignant
potential is unknown, and many consider cancer another
contraindication.150,151
5 Finally, although it is uncommon, the collection process
can be associated with fatal air embolism; such events were
originally reported in association with a device that allowed
reservoir contents to be pumped directly into a venous cathe-
ter, without an air detection system.152 Although instruments
with this design are no longer marketed, rare fatalities are
still reported. A 1997 report cautioned that external pressure
devices magnify the risk of air embolism and should never be
6
used with perioperatively salvaged blood except if absolutely
necessary and under close supervision.153
Figure 30–7 Schematic of an instrument used for collection and
washing of perioperatively salvaged blood. Shed blood is suctioned
The collection and transfusion of intraoperatively sal-
from the operative field (1), an anticoagulant is added (2), and the vaged blood has been associated with substantial reductions
blood moves past a mesh filter into a reservoir (3). A pump (4) forces in allogeneic transfusions (>50%), particularly in spine sur-
the blood into a spinning plastic centrifuge bowl (5). With separation, gery,154,155 hip replacement,156 and vascular procedures such
plasma flows into a waste bag (6); saline (7) is continuously pumped as aortic reconstruction.157 During cardiac surgery, the larg-
into the bowl to wash the packed red blood cells. At the completion of
washing, the red blood cells are moved into a reinfusion bag (8) for re- est volume of blood that can be processed for return to the
turn to the patient. (From Haemonetics, Braintree, Mass., and Williams patient comes from the membrane oxygenator. Although this
& Wilkins, with permission.) blood technically is not shed, in that it is removed from the
BLOOD BANKING
Table 30–2 Characteristics of Perioperatively Salvaged Blood Compared with Banked Blood and
Normal Patient Values*
Fibrin
Free Hemoglobin Platelet Count Coagulation Degradation
Component Hematocrit (mg/dL) (per mm3) Factors Products
Salvaged blood, Low (25%) Very high (≥200) Low (100,000) Low (35%– High (300
unwashed 75%) mg/dL)
Salvaged blood, High (60%) Low (<50) Very low (<10,000) Absent Absent
washed
Allogeneic blood High (60%) Variable with age Low and Low (35%– Increased
(packed red of component dysfunctional 75%)
blood cells) (100,000)
Normal patient Normal (40%) <5 300,000 100% <10 mg/dL
*
Typical results of laboratory tests are given. The transfusion of large volumes of salvaged blood could result in similar alterations in the
recipient.
From Noon GP. Intraoperative autotransfusion. Surgery 1978;84:719–721, and Silva R, Moore EE, Bar-Or D, et al. The risk:benefit ratio of
autotransfusion: Comparison to banked blood in a canine model. J Trauma 1984;24:557–564. Used with permission.
extracorporeal circuit at the end of surgery, the processing recipients after transfusion.178 Occasional complications do
is helpful in concentrating RBCs and removing cardioplegia occur, however, including respiratory distress,179 hypoten-
solution.158,159 In liver transplantation, volumes as large as 25 sion with anaphylaxis,180 and fever.181 The last more likely to
units have been salvaged,160–162 and salvage during trauma occur when the product is collected over a long time interval
is also feasible.147,163,164 The collection of autologous blood (6 to 12 hours). The pathophysiology of these events remains
during cesarean section carries theoretical risks associated unclear.
with transfusing amniotic fluid; however, an analysis of 139 After open heart surgery, mediastinal blood may contain
women who received processed salvaged blood identified no very high levels of cardiac muscle enzymes, especially cre-
increased incidence of obstetric complications.165 Blood has atine phosphokinase, as well as lactate dehydrogenase from
also been recovered from the hemoperitoneum in association hemolyzed RBCs.141,182 The reinfusion of shed mediastinal
with ectopic pregnancy,166 during radical prostatectomy,167 blood can result in increased levels of these enzymes and can
and during splenectomy.168 Intraoperatively salvaged blood confound the diagnosis of myocardial infarction in the post-
has been a useful adjunct in the treatment of some Jehovah’s operative period.183,184 The volume of RBCs actually salvaged
Witnesses, whose literal acceptance of the Bible includes is often small and the effect on reducing transfusions debat-
II abstention from routine allogeneic blood transfusions. In able. Infusion of shed mediastinal blood after cardiac opera-
this situation, an uninterrupted circuit between the salvaged tions appears to have the potential to reduce the volume of
428 blood processor and the patient facilitates acceptance.169 allogeneic blood required (by 1.4 units in one study).185 In
other situations, the benefit is less clear. Although the vol-
Postoperative Autologous Transfusion: ume of postoperative drainage is often substantial in ortho-
Blood Salvage pedic procedures,186,187 much of the collection is plasma and
other serosanguineous fluids rather than RBCs. One study
Both canister systems and RBC processors can be used to reported a mean total collection of only 55 ± 29 mL of RBCs
collect postoperative blood drainage, such as that from the in drains after hip surgery.188 Orthoplasty procedures per-
mediastinum after heart surgery,170 from the peritoneal cav- formed without cement are associated with larger periop-
ity after hepatic injury,171 or from the knee or hip site after erative blood losses, and postoperative salvage may be more
orthopedic procedures.172 Blood salvaged from a serosal cav- effectively used in such cases.189
ity has little residual fibrinogen and few platelets, and clotting
is usually not a problem; therefore, the addition of anticoagu-
Acute Normovolemic Hemodilution
lants to the collection is usually not necessary.173 Despite the
substantial levels of free hemoglobin in the salvaged blood, The collection of autologous blood at the start of surgery,
RBCs survive normally, as documented by studies involving for return to the patient at the end of the procedure, had its
radiolabeled markers.174 origins in open-heart surgery. The original goal was preven-
In addition to free hemoglobin, the salvaged blood may tion of postoperative coagulopathies through ex vivo main-
be contaminated with tissue exudate, bone, bone marrow, tenance of a supply of platelets undamaged by exposure to
and other biologic and surgical materials; nevertheless, most the membrane oxygenator.190,191 However, additional advan-
patients tolerate the infusions well. Bioactive substances tages to the intentionally created anemia were also identi-
measured in the unwashed drainage include histamine, fied. Hemodilution can contribute to a reduction in RBC
interleukin-6 and other cytokines, prostaglandins, and acti- loss. In simplest terms, a patient with a hematocrit of 45%
vated complement components; however, these levels have and a 2 L blood loss during surgery loses roughly 900 mL of
not been associated with transfusion reactions, and levels RBCs, whereas a similar patient with a hematocrit of 20%
in patients after infusion are not significantly altered.175–177 loses only 400 mL of RBCs. More elaborate mathematical
Similarly, methyl methacrylate (used as a cement in ortho- modeling studies have been published that take into account
pedic surgery) and its breakdown product, methanol, can be the dynamic nature of the patient’s RBC mass as it is affected
measured in blood salvaged postoperatively from the surgi- by blood loss, fluid replacement, and blood transfusions
cal site; however, these materials have not been detected in (Fig. 30–8).192,193 Hemodilution is probably less expensive to
BLOOD MANAGEMENT
12000 12000
8000 8000
6000 6000
15%
accomplish than preoperative autologous blood donation, in a study of elective open-heart surgery patients receiving
and it may be the only option available when surgery is per- erythropoietin or placebo preoperatively with intraoperative
formed in other than elective settings.194 isovolemic hemodilution. In the erythropoietin group, more
The technique involves removal of blood into standard patients were able to have intraoperative isovolemic hemo-
collection bags with citrate anticoagulation (unless the dilution, fewer patients received allogeneic blood transfu-
patient is already heparinized) and replacement of lost vol- sion, and the total mean transfusion per patient was lower.
ume with either crystalloids or colloids. Close monitoring Further analysis showed blood loss, age, baseline hematocrit,
of the patient’s cardiovascular status is necessary during the as well as preoperative treatment with erythropoietin, to
hemodilution process. Units are stored in the operating room be independent predictors of the need for allogeneic blood
during surgery and reinfused as needed, in reverse order of transfusion.205
collection, reserving the bags with the highest concentration
of RBCs for the end of the procedure, after blood loss has
been controlled. MULTIDISCIPLINARY APPROACH 30
In orthopedic and cardiovascular surgery, reductions in TO BLOOD MANAGEMENT
allogeneic blood use have been reported after extreme hemo- 429
dilution (reduction of the patient’s RBC mass by as much as To decrease allogeneic transfusion, communication between
50%).193,195 More modest hemodilution (e.g., removal of 2 all health care providers must occur. The likelihood of blood
units of blood at the beginning of surgery) may also be benefi- transfusion is decreased if the care of the patient is optimized
cial, according to some workers,196,197 but this is not accepted before, during, and after surgery.
by others.192,198,199 The severity of the anemia could affect oxy-
gen transport, although the concomitant drop in blood viscos-
ity, and compensatory cardiac output increases, could restore
Preoperative Factors
oxygen delivery. However, one group has provided evidence The strongest predictor of a patient needing blood during an
that hemodilution may jeopardize patients at risk for myocar- elective surgery is the baseline hematocrit, with other signifi-
dial infarction.200 Further clinical studies appear necessary to cant contributing factors being the patient’s blood volume
resolve the continued controversy over the value of hemodilu- and red cell loss during the procedure.181,206,207 Optimization
tion in contemporary transfusion practice.201,202 of the patient’s hemoglobin prior to surgery will decrease
the chance of transfusion. Iron or erythropoietin may be
Acute Normovolemic Hemodilution indicated, depending on the cause of the patient’s anemia.
Preoperative anemia additionally increases perioperative
in Combination with Erythropoietin
infection and mortality, which may be a result of increase
Erythropoietin can be used preoperatively in conjunction risk of allogeneic blood transfusion.208
with acute normovolemic hemodilution. A three-armed The correction of impaired hemostasis should decrease
study looked at preoperative autologous donation (3 units) blood loss and, therefore, decrease the need for transfusion.
versus erythropoietin plus acute normovolemic hemo- Impaired hemostasis can be corrected by discontinuing anti-
dilution versus acute normovolemic hemodilution alone coagulants and antiplatelet agents in a timely manner or
in patients undergoing prostatectomy. The investigation treating the cause of the impaired hemostasis. In a retrospec-
showed similar allogeneic transfusion rates with the lowest tive study, the frequency of blood transfusion was higher in
cost in the acute normovolemic hemodilution alone arm.203 patients without preoperative correction of primary hemo-
Mathematical modeling predicts preoperative erythropoietin stasis than in patients with correction.209 In another retro-
will be most beneficial when used in conjunction with acute spective study in cardiac surgery patients, preoperative use
normovolemic hemodilution, especially in patients with of antithrombotics (enoxaparin, clopidogrel, or GP IIb/IIIa
small blood volumes and mild anemia.204 This is confirmed inhibitors) increased transfusion, but uniquely preoperative
BLOOD BANKING 5. Oliver WC Jr, Santrach PJ, Danielson GK, Nuttall GA, Schroeder
enoxaparin increased postoperative bleeding and the need
for reexploration.210 DR, Ereth MH. Desmopressin does not reduce bleeding and transfu-
sion requirements in congenital heart operations. Ann Thorac Surg
2000;70:1923–1930.
Intraoperative Factors 6. Physicians’ Desk Reference, 59th ed. Montvale, NJ, Thomson PDR,
2005.
Intraoperative factors to reduce allogeneic transfusion 7. Carless PA, Henry DA, Moxey AJ, et al. Desmopressin for minimising
include preventing hypothermia, optimizing surgical tech- perioperative allogeneic blood transfusion. Cochrane Database Syst
Rev 2004;CD001884.
nique, and controlling hemostasis. Surgical techniques that 8. Henry DA, Moxey AJ, Carless PA, et al. Anti-fibrinolytic use for mini-
reduce bleeding include laparoscopic, robotic, or endovas- mising perioperative allogeneic blood transfusion. Cochrane Database
cular approaches.211 Surgical instruments that maximize Syst Rev 2001;CD001886.
coagulation include the ultrasonic scalpel and argon beam 9. Lemmer JH Jr, Dilling EW, Morton JR, et al. Aprotinin for primary
coronary artery bypass grafting: a multicenter trial of three dose regi-
coagulator. mens. Ann Thorac Surg 1996;62:1659–1667.
10. Smith PK, Datta SK, Muhlbaier LH, Samsa G, Nadel A, Lipscomb
J. Cost analysis of aprotinin for coronary artery bypass patients: analysis
Controlled Hypotension of the randomized trials. Ann Thorac Surg 2004;77:635–642.
Controlled hypotension has been shown in some small 11. Findlay JY, Rettke SR, Ereth MH, Plevak DJ, Krom RA, Kufner RP.
Aprotinin reduces red blood cell transfusion in orthotopic liver trans-
studies to decrease blood loss and transfusion in orthope- plantation: a prospective, randomized, double-blind study. Liver
dic surgery.212,213 Combining controlled hypotension with Transpl 2001;7:802–807.
other banked blood saving techniques, such as hemodilu- 12. Porte RJ, Molenaar IQ, Begliomini B, et al. Aprotinin and transfu-
tion, may increase the likelihood of avoiding transfusion.214 sion requirements in orthotopic liver transplantation: a multicen-
tre randomised double-blind study. EMSALT Study Group. Lancet
Other studies have not shown a benefit with controlled 2000;355:1303–1309.
hypotension alone.215 The concern with controlled hypoten- 13. Amar D, Grant FM, Zhang H, Boland PJ, Leung DH, Healey JA. Anti-
sion is that it may produce end-organ ischemia, especially in fibrinolytic therapy and perioperative blood loss in cancer patients
patients with atherosclerosis. The decrease in blood loss as a undergoing major orthopedic surgery. Anesthesiology 2003;98:
result of controlled hypotension may not outweigh its risk 337–342.
14. Samama CM, Langeron O, Rosencher N, et al. Aprotinin versus pla-
of ischemia.211 cebo in major orthopedic surgery: a randomized, double-blinded,
dose-ranging study. Anesth Analg 2002;95:287–293.
15. Jeserschek R, Clar H, Aigner C, Rehak P, Primus B, Windhager
Normothermia R. Reduction of blood loss using high-dose aprotinin in major ortho-
paedic surgery: a prospective, double-blind, randomised and placebo-
Central body temperature decreases after induction of gen- controlled study. J Bone Joint Surg Br 2003;85:174–177.
eral anesthesia, which may increase blood loss and transfu- 16. Stewart A, Newman L, Sneddon K, Harris M. Aprotinin reduces blood
sion requirements. In a small study of patients undergoing loss and the need for transfusion in orthognathic surgery. Br J Oral
major abdominal surgery, active warming with forced air Maxillofac Surg 2001;39:365–370.
showed reduced blood loss and blood product use.216 Two 17. Bedirhan MA, Turna A, Yagan N, Tasci O. Aprotinin reduces postop-
II erative bleeding and the need for blood products in thoracic surgery:
studies show conflicting results in mild hypothermia affect- results of a randomized double-blind study. Eur J Cardiothorac Surg
430 ing blood loss in total hip arthroplasty.217,218 Studies demon- 2001;20:1122–1127.
strate conflicting results regarding blood loss and transfusion 18. Kyriss T, Wurst H, Friedel G, Jaki R, Toomes H. Reduced blood loss
requirements with normothermia versus hypothermia car- by aprotinin in thoracic surgical operations associated with high risk
of bleeding: a placebo-controlled, randomized phase IV study. Eur
diopulmonary bypass.219,220 In summary, no definitive J Cardiothorac Surg 2001;20:38–41.
answer is available about maintaining normothermia to 19. Florentino-Pineda I, Thompson GH, Poe-Kochert C, Huang RP, Haber
decrease blood loss. LL, Blakemore LC. The effect of amicar on perioperative blood loss in
idiopathic scoliosis: the results of a prospective, randomized double-
blind study. Spine 2004;29:233–238.
Postoperative Factors 20. Sethna NF, Zurakowski D, Brustowicz RM, Bacsik J, Sullivan LJ, Shapiro
F. Tranexamic acid reduces intraoperative blood loss in pediatric patients
Phlebotomy undergoing scoliosis surgery. Anesthesiology 2005;102:727–732.
21. Casati V, Della VP, Benussi S, et al. Effects of tranexamic acid on post-
Patients in the intensive care unit have a mean blood draw operative bleeding and related hematochemical variables in coronary
of approximately 40 mL a day with a total volume of 760 mL. surgery: comparison between on-pump and off-pump techniques.
Patients with arterial lines have a total blood loss of approxi- J Thorac Cardiovasc Surg 2004;128:83–91.
mately 900 mL. These large losses from phlebotomy contribute 22. Veien M, Sorensen JV, Madsen F, Juelsgaard P. Tranexamic acid given intra-
operatively reduces blood loss after total knee replacement: a randomized,
to increased transfusion requirements.221 controlled study. Acta Anaesthesiol Scand 2002;46:1206–1211.
23. Engel JM, Hohaus T, Ruwoldt R, Menges T, Jurgensen I, Hempelmann
G. Regional hemostatic status and blood requirements after total knee
REFERENCES arthroplasty with and without tranexamic acid or aprotinin. Anesth
Analg 2001;92:775–780.
1. Carson JL, Noveck H, Berlin JA, Gould SA. Mortality and morbidity 24. Mangano DT, Tudor IC, Dietzel C; the Multicenter Study of Periopera-
in patients with very low postoperative Hb levels who decline blood tive Ischemia Research Group and the Ischemia Research and Educa-
transfusion. Transfusion 2002;42:812–818. tion Foundation. The risk associated with aprotinin in cardiac surgery.
2. Hill SR, Carless PA, Henry DA, et al. Transfusion thresholds and other N Engl J Med 2006;354:353–365.
strategies for guiding allogeneic red blood cell transfusion. Cochrane 25. Diprose P, Herbertson MJ, O’Shaughnessy D, Deakin CD, Gill RS.
Database Syst Rev 2002;CD002042. Reducing allogeneic transfusion in cardiac surgery: a randomized dou-
3. Mallett SV, Peachey TD, Sanehi O, Hazlehurst G, Mehta A. Reducing ble-blind placebo-controlled trial of antifibrinolytic therapies used in
red blood cell transfusion in elective surgical patients: the role of audit addition to intra-operative cell salvage. Br J Anaesth 2005;94:271–278.
and practice guidelines. Anaesthesia 2000;55:1013–1019. 26. Landymore RW, Murphy JT, Lummis H, Carter C. The use of low-dose
4. Garrioch M, Sandbach J, Pirie E, Morrison A, Todd A, Green R. Reduc- aprotinin, epsilon-aminocaproic acid, or tranexamic acid for prevention
ing red cell transfusion by audit, education, and a new guideline in a of mediastinal bleeding in patients receiving aspirin before coronary
large teaching hospital. Transfus Med 2004;14:25–31. artery bypass operations. Eur J Cardiothorac Surg 1997;11:798–800.
BLOOD MANAGEMENT
27. Dalmau A, Sabate A, Koo M, et al. The prophylactic use of tranexamic 54. Okuyama M, Ikeda K, Shibata T, Tsukahara Y, Kitada M, Shimano
acid and aprotinin in orthotopic liver transplantation: a comparative T. Preoperative iron supplementation and intraoperative transfusion
study. Liver Transpl 2004;10:279–284. during colorectal cancer surgery. Surg Today 2005;35:36–40.
28. Xia VW, Steadman RH. Antifibrinolytics in orthotopic liver transplan- 55. Cuenca J, Garcia-Erce JA, Martinez AA, et al. Role of parenteral iron
tation: current status and controversies. Liver Transpl 2005;11:10–18. in the management of anaemia in the elderly patient undergoing dis-
29. Spahn DR, Tucci MA, Makris M. Is recombinant FVIIa the magic bul- placed subcapital hip fracture repair: preliminary data. Arch Orthop
let in the treatment of major bleeding? Br J Anaesth 2005;94:553–555. Trauma Surg 2005;125:342–347.
30. Friederich PW, Henny CP, Messelink EJ, et al. Effect of recombinant 56. Blajchman MA. Transfusion-associated immunomodulation and uni-
activated factor VII on perioperative blood loss in patients undergo- versal white cell reduction: are we putting the cart before the horse?
ing retropubic prostatectomy: a double-blind placebo-controlled ran- Transfusion 1999;39:665–670.
domised trial. Lancet 2003;361:201–205. 57. Innerhofer P, Walleczek C, Luz G, et al. Transfusion of buffy coat-
31. Raobaikady R, Redman J, Ball JA, Maloney G, Grounds RM. Use of depleted blood components and risk of postoperative infection in
activated recombinant coagulation factor VII in patients undergoing orthopedic patients. Transfusion 1999;39:625–632.
reconstruction surgery for traumatic fracture of pelvis or pelvis and 58. Innerhofer P, Klingler A, Klimmer C, Fries D, Nussbaumer W. Risk for
acetabulum: a double-blind, randomized, placebo-controlled trial. Br postoperative infection after transfusion of white blood cell-filtered
J Anaesth 2005;94:586–591. allogeneic or autologous blood components in orthopedic patients
32. Lodge JP, Jonas S, Oussoultzoglou E, et al. Recombinant coagulation undergoing primary arthroplasty. Transfusion 2005;45:103–110.
factor VIIa in major liver resection: a randomized, placebo-controlled, 59. Westphal RG. Transfusion of buffy coat-depleted blood components
double-blind clinical trial. Anesthesiology 2005;102:269–275. and risk of postoperative infection in orthopedic patients. Transfusion
33. Wernet D, Mayer G. Isoagglutinins following ABO-incompatible bone 2000;40:381–383.
marrow transplantation. Vox Sang 1992;62:176–179. 60. Owings DV, Kruskall MS, Thurer RL, Donovan LM. Autologous blood
34. Milne AA. Clinical impact of fibrin sealants. Vox Sang 2004;87(Suppl donations prior to elective cardiac surgery: safety and effect on subse-
2):29–30. quent blood use. JAMA 1989;262:1963–1968.
35. Neschis DG, Heyman MR, Cheanvechai V, Benjamin ME, Flinn WR. 61. Wolfe LC, Byrne AM, Lux SE. Molecular defects in the membrane
Coagulopathy as a result of factor V inhibitor after exposure to bovine skeleton of blood bank-stored red cells: abnormal spectrin-protein
topical thrombin. J Vasc Surg 2002;35:400–402. 4.1-actin complex formation. J Clin Invest 1986;78:1681–1686.
36. Buchta C, Dettke M, Funovics PT, et al. Fibrin sealant produced by 62. Spiess BD, Sassetti RJ, McCarthy RJ, Narbone RF, Tuman KJ, Ivankov-
the CryoSeal FS System: product chemistry, material properties, and ich AD. Autologous blood donation: hemodynamics in a high-risk
possible preparation in the autologous preoperative setting. Vox Sang patient population. Transfusion 1992;32:17–22.
2004;86:257–262. 63. Linden JV. Autologous blood errors and incidents [abstract]. Transfu-
37. Carless PA, Henry DA, Anthony DM. Fibrin sealant use for minimising sion 1994;34:28S.
perioperative allogeneic blood transfusion. Cochrane Database Syst 64. Haditsch M, Binder L, Garbriel C, Muller-Uri P, Wartschinger R, Mit-
Rev 2003;CD004171. termayer H. Yersinia enterocolitica septicemia in autologous blood
38. Highmore W. Practical remarks on an overlooked source of blood sup- transfusion. Transfusion 1994;34:907–909.
ply for transfusion in postpartum hemorrhage, suggested by a recent 65. Etchason J, Petz L, Keeler E, et al. The cost-effectiveness of preoperative
fatal case. Lancet 1874;1:89. autologous blood donations. N Engl J Med 1995;332:719–724.
39. Theis HJ. Zur Behandlung der Estrauteringraviditar. Zentralbl Gyn- 66. Heiss MM, Mempel W, Jauch KW, et al. Beneficial effect of autologous
aekol 1914;38:1191–1193. blood transfusion on infectious complications after colorectal cancer
40. Brecher ME, Goodnough LT. The rise and fall of preoperative autolo- surgery. Lancet 1993;342:1328–1333.
gous blood donation. Transfusion 2001;41:1459–1462. 67. Busch OR, Hop WC, van Papendrecht MA, Marquet RL, Jeekel
41. Surgenor DM, Wallace EL, Hao SL, Chapman RH. Collection and J. Blood transfusion and prognosis in colorectal cancer. N Engl J Med
transfusion of blood in the United States, 1982–1988. N Engl J Med 1993;328:1372–1376.
30
1990;322:1646–1651. 68. Jensen LS, Kissmeyer-Nielsen P, Wolff B, Qvist N. Randomised
42. Wallace EL, Churchill WH, Surgenor DM, et al. Collection and trans- comparison of leucocyte-depleted versus buffy-coat-poor blood 431
fusion of blood and blood components in the United States, 1992. transfusion and complications after colorectal surgery. Lancet
Transfusion 1995;35:802–812. 1996;348:841–845.
43. Sullivan MT, Wallace EL. Blood collection and transfusion in the 69. Sonnenberg FA, Gregory P, Yomtovian R, et al. The cost-effective-
United States in 1999. Transfusion 2005;45:141–148. ness of autologous transfusion revisited: implications of an increased
44. Sullivan MT, McCullough J, Schreiber GB, Wallace EL. Blood col- risk of bacterial infection with allogeneic transfusion. Transfusion
lection and transfusion in the United States in 1997. Transfusion 1999;39:808–817.
2002;42:1253–1260. 70. Kruskall MS, Yomtovian R, Dzik WH, Friedman KD, Umlas J. On
45. Lewis LD. Preclinical and clinical studies: a preview of potential future improving the cost-effectiveness of autologous blood transfusion
applications of erythropoietic agents. Semin Hematol 2004;41(4 Suppl practices. Transfusion 1994;34:259–264.
7):17–25. 71. Kruskall MS. Cost effectiveness of autologous blood donation. N Engl
46. Goodnough LT, Monk TG, Andriole GL. Erythropoietin therapy. J Med 1995;333:461–462.
N Engl J Med 1997;336:933–938. 72. Blum LN, Allen JR, Genel M, Howe JP. Crosover use of donated blood
47. Monk TG. Preoperative recombinant human erythropoietin in anemic for autologous transfusion: report of the Council on Scientific Affairs,
surgical patients. Crit Care 2004;8(Suppl 2):S45–S48. American Medical Association. Transfusion 1998;38:891–895.
48. Kosmadakis N, Messaris E, Maris A, et al. Perioperative erythropoi- 73. Haugen RK, Hill GE. A large-scale autologous blood program in a
etin administration in patients with gastrointestinal tract cancer: community hospital: a contribution to the community’s blood supply.
prospective randomized double-blind study. Ann Surg 2003;237: JAMA 1987;257:1211–1214.
417–421. 74. Woolson ST, Marsh JS, Tanner JB. Transfusion of previously deposited
49. Faris PM, Ritter MA, Abels RI. The effects of recombinant human autologous blood for patients undergoing hip-replacement surgery.
erythropoietin on perioperative transfusion requirements in patients J Bone Joint Surg Am 1987;69:325–328.
having a major orthopaedic operation. The American Erythropoietin 75. Woolson ST, Pottorff G. Use of preoperatively deposited autologous
Study Group. J Bone Joint Surg Am 1996;78:62–72. blood for total knee replacement. Orthopedics 1993;16:137–141.
50. Canadian Orthopedic Perioperative Erythropoietin Study Group. 76. Kruskall MS, Glazer EE, Leonard SS, et al. Utilization and effectiveness
Effectiveness of perioperative recombinant human erythropoietin in of a hospital autologous preoperative blood donor program. Transfu-
elective hip replacement. Lancet 1993;341:1227–1232. sion 1986;26:335–340.
51. Podesta A, Carmagnini E, Parodi E, et al. Elective coronary and valve 77. Toy PT, Strauss RG, Stehling LC, et al. Predeposited autologous blood
surgery without blood transfusion in patients treated with recombi- for elective surgery. N Engl J Med 1987;316:517–520.
nant human erythropoietin (epoetin-alpha). Minerva Cardioangiol 78. Thompson HW, McCullough J. Use of blood components containing red
2000;48:341–347. cells by donors of allogeneic bone marrow. Transfusion 1986;26:98–100.
52. Olijhoek G, Megens JG, Musto P, et al. Role of oral versus IV iron sup- 79. Peters CA, Walsh PC. Blood transfusion and anesthetic practices in
plementation in the erythropoietic response to rHuEPO: a random- radical retropubic prostatectomy. J Urol 1985;134:81–83.
ized, placebo-controlled trial. Transfusion 2001;41:957–963. 80. Toy PT, Menozzi D, Strauss RG, Stehling LC, Kruskall M, Ahn DK. Effi-
53. Drueke T. Hyporesponsiveness to recombinant human erythropoietin. cacy of preoperative donation of blood for autologous use in radical
Nephrol Dial Transplant 2001;16(Suppl 7):25–28. prostatectomy. Transfusion 1993;33:721–724.
BLOOD BANKING 81. Koch MO, Smith JA Jr. Blood loss during radical retropubic prosta- 106. Giordano GF, Rivers SL, Chung GK, et al. Autologous platelet-rich
tectomy: is preoperative autologous blood donation indicated? J Urol plasma in cardiac surgery: effect on intraoperative and postoperative
1996;156:1077–1080. transfusion requirements. Ann Thorac Surg 1988;46:416–419.
82. Goh M, Kleer CG, Kielczewski P, Wojno KJ, Kyungmann K, Oesterling 107. Harker LA, Malpass TW, Branson HE, Hessel EA, Slichter SJ. Mecha-
JE. Autologous blood donation prior to anatomical radical retropubic nism of abnormal bleeding in patients undergoing cardiopulmonary
prostatectomy: is it necessary? Urology 1996;49:569–574. bypass: acquired transient platelet dysfunction associated with selec-
83. Kruskall MS. Autologous blood transfusions and plastic surgery [edi- tive alpha-granule release. Blood 1980;56:824–834.
torial]. Plast Reconstr Surg 1989;84:662–664. 108. Boldt J, von Bormann B, Kling D, Jacobi M, Moosdorf R, Hempel-
84. Kruskall MS. Controversies in transfusion medicine: the safety and mann G. Preoperative plasmapheresis in patients undergoing cardiac
utility of autologous donations by pregnant patients. Transfusion surgery procedures. Anesthesiology 1990;72:282–288.
1990;30:168–171. 109. Davies GG, Wells DG, Mabee TM, Sadler R, Melling NJ. Platelet-leuko-
85. Larsen R, Titlestad K, Lillevang ST, Thomsen SG, Kidholm K, Georg- cyte plasmapheresis attenuates the deleterious effects of cardiopulmo-
sen J. Cesarean section: is pretransfusion testing for red cell alloanti- nary bypass. Ann Thorac Surg 1992;53:274–277.
bodies necessary? Acta Obstet Gynecol Scand 2005;84:448–455. 110. DelRossi AJ, Cernaianu AC, Vertrees RA, et al. Platelet-rich plasma
86. Combs CA, Murphy EL, Laros RK. Cost-benefit analysis of autologous reduces postoperative blood loss after cardiopulmonary bypass.
blood donation in obstetrics. Obstet Gynecol 1992;80:621–625. J Thorac Cardiovasc Surg 1990;100:281–286.
87. Andres RL, Piacquadio KM, Resnik R. A reappraisal of the need for 111. Tobe CE, Vocelka C, Sepulvada R, et al. Infusion of autologous platelet
autologous blood donation in the obstetric patient. Am J Obstet Gyne- rich plasma does not reduce blood loss and product use after coronary
col 1990;163:1551–1553. artery bypass. J Thorac Cardiovasc Surg 1993;105:1007–1014.
88. Klapholz H. Blood transfusion in contemporary obstetric practice. 112. Ereth MH, Oliver WC, Beynen FMK, et al. Autologous platelet-rich
Obstet Gynecol 1990;75:940–943. plasma does not reduce transfusion of homologous blood prod-
89. Katz AR, Ali V, Ross PJ, Gammon E. Management of a rare blood type: ucts in patients undergoing repeat valvular surgery. Anesthesiology
O “Bombay” in pregnancy. Obstet Gynecol 1981;57:16S–17S. 1993;79:540–547.
90. Hardwick ME, Morris BM, Colwell CW Jr. Two-dose epoetin alfa 113. Wickey GS, Keifer JC, Larach DR, Diaz MR, Williams DR. Hepa-
reduces blood transfusions compared with autologous donation. Clin rin resistance after intraoperative platelet-rich plasma harvesting.
Orthop Relat Res 2004;23:240–244. J Thorac Cardiovasc Surg 1992;103:1172–1176.
91. Stowell CP, Chandler H, Jove M, Guilfoyle M, Wacholtz MC. An open- 114. Carless PA, Rubens FD, Anthony DM, O’Connell D, Henry DA. Plate-
label, randomized study to compare the safety and efficacy of periop- let-rich plasmapheresis for minimising perioperative allogeneic blood
erative epoetin alfa with preoperative autologous blood donation in transfusion. Cochrane Database Syst Rev 2003;CD004172.
total joint arthroplasty. Orthopedics 1999;22(Suppl 1):S105–S112. 115. Greenwalt TJ. Autologous and aged blood donors. JAMA 1987;
92. Chun TY, Martin S, Lepor H. Preoperative recombinant human eryth- 257:1220–1221.
ropoietin injection versus preoperative autologous blood donation 116. Silvergleid AJ. Safety and effectiveness of predeposit autologous
in patients undergoing radical retropubic prostatectomy. Urology transfusions in preteen and adolescent children. JAMA 1987;257:
1997;50:727–732. 3403–3404.
93. Bezwada HP, Nazarian DG, Henry DH, Booth RE Jr. Preoperative use 117. Kruskall MS, Leonard S, Klapholz H. Autologous blood donation dur-
of recombinant human erythropoietin before total joint arthroplasty. ing pregnancy: analysis of safety and blood utilization. Obstet Gynecol
J Bone Joint Surg Am 2003;85:1795–1800. 1987;70:938–941.
94. Price TH, Goodnough LT, Vogler WR, et al. Improving the efficacy of 118. McVay PA, Hoag RW, Hoag MS, Toy PT. Safety and use of autologous
preoperative autologous blood donation in patients with low hema- blood donation during the third trimester of pregnancy. Am J Obstet
tocrit: a randomized, double-blind, controlled trial of recombinant Gynecol 1989;160:1479–1486.
human erythropoietin. Am J Med 1996;101:22S–27S. 119. Goldfinger D, Capon S, Czer L, et al. Safety and efficacy of preoperative
II 95. Mercuriali F, Zanella A, Barosi G, et al. Use of erythropoietin to donation of blood for autologous use by patients with end-stage heart
increase the volume of autologous blood donated by orthopedic or lung disease who are awaiting organ transplantation. Transfusion
432 patients. Transfusion 1993;33:55–60. 1993;33:336–340.
96. Shapiro GS, Boachie-Adjei O, Dhawlikar SH, Maier LS. The use of epo- 120. Popovsky MA, Whitaker B, Arnold NL. Severe outcomes of alloge-
etin alfa in complex spine deformity surgery. Spine 2002;27:2067–2071. neic and autologous blood donation: frequency and characterization.
97. Braga M, Gianotti L, Vignali A, et al. Evaluation of recombinant human Transfusion 1995;35:734–737.
erythropoietin to facilitate autologous blood donation before surgery 121. Kasper SM, Gerlich W, Buzello W. Preoperative red cell production in
in anaemic patients with cancer of the gastrointestinal tract. Br J Surg patients undergoing weekly autologous blood donation. Transfusion
1995;82:1637–1640. 1997;37:1058–1062.
98. Kulier AH, Gombotz H, Fuchs G, Vuckovic U, Metzler H. Subcutane- 122. Kickler TS, Spivak JL. Effect of repeated whole blood donations on
ous recombinant human erythropoietin and autologous blood dona- serum immunoreactive erythropoietin levels in autologous donors.
tion before coronary artery bypass surgery. Anesth Analg 1993;76: JAMA 1988;260:65–67.
102–106. 123. Goodnough LT, Price TH, Rudnick S, Soegiarso RW. Preoperative red
99. Hyllner M, Avall A, Swolin B, Bengtson JP, Bengtsson A. Autologous cell production in patients undergoing aggressive autologous blood
blood transfusion in radical hysterectomy with and without erythro- phlebotomy with and without erythropoietin therapy. Transfusion
poietin therapy. Obstet Gynecol 2002;99:757–762. 1992;32:441–445.
100. Rau B, Schlag PM, Willeke F, Herfarth C, Stephan P, Franke W. 124. Brugnara C, Chambers LA, Malynn E, Goldberg MA, Kruskall MS. Red
Increased autologous blood donation in rectal cancer by recombinant blood cell regeneration induced by subcutaneous recombinant eryth-
human erythropoietin (rhEPO). Eur J Cancer 1998;34:992–998. ropoietin: iron-deficient erythropoiesis in iron-replete subjects. Blood
101. Avall A, Hyllner M, Bengtson JP, Carlsson L, Bengtsson A. Recombi- 1993;81:956–964.
nant human erythropoietin in preoperative autologous blood dona- 125. Rutherford CJ, Schneider TJ, Dempsey H, Kirn DH, Brugnara C,
tion did not influence the haemoglobin recovery after surgery. Acta Goldberg MA. Efficacy of different dosing regimens for recombi-
Anaesthesiol Scand 2003;47:687–692. nant human erythropoietin in a simulated perisurgical setting: the
102. De Pree C, Mermillod B, Hoffmeyer P, Beris P. Recombinant human importance of iron availability in optimizing response. Am J Med
erythropoietin as adjuvant treatment for autologous blood donation 1994;96:139–145.
in elective surgery with large blood needs (≥ 5 units): a randomized 126. Smith KJ, James DS, Hunt WC, McDonough W, Quintana R. A ran-
study. Transfusion 1997;37:708–714. domized, double-blind comparison of donor tolerance of 400 mL,
103. Cazenave JP, Irrmann C, Waller C, et al. Epoetin alfa facilitates pre- 200 mL, and sham red cell donation. Transfusion 1996;36:674–680.
surgical autologous blood donation in non-anaemic patients sched- 127. Hocker P. Red cell apheresis in autologous preoperative blood dona-
uled for orthopaedic or cardiovascular surgery. Eur J Anaesthesiol tion. Transfus Apher Sci 2001;24:75–78.
1997;14:432–442. 128. Monsen ER, Critchlow CW, Finch CA, Donohue DM. Iron balance in
104. Couvret C, Laffon M, Baud A, Payen V, Burdin P, Fusciardi J. A restric- super donors. Transfusion 1983;23:221–225.
tive use of both autologous donation and recombinant human erythro- 129. Lieden G, Hoglund S, Ehn L. Iron supplement to blood donors. II:
poietin is an efficient policy for primary total hip or knee arthroplasty. Effect of continuous iron supply. Acta Med Scand 1975;197:37–41.
Anesth Analg 2004;99:262–271. 130. Weisbach V, Skoda P, Rippel R, et al. Oral or intravenous iron as an
105. Consensus conference. Fresh-frozen plasma: indications and risks. adjuvant to autologous blood donation in elective surgery: a random-
JAMA 1985;253:551–553. ized, controlled study. Transfusion 1999;39:465–472.
BLOOD MANAGEMENT
131. Wilson JD, Utz DC, Taswell HF. Autotransfusion during transurethal 158. Keeling MM, Gray LA, Brink MA, Hillerich VK, Bland KI. Intraop-
resection of the prostate: technique and preliminary clinical evalua- erative autotransfusion: experience in 725 consecutive cases. Ann Surg
tion. Mayo Clin Proc 1969;44:374–386. 1983;197:536–541.
132. Long GW, Glover JL, Bendick PJ, et al. Cell washing versus immedi- 159. McCarthy PM, Popovsky MA, Schaff HV, et al. Effect of blood conser-
ate reinfusion of intraoperatively shed blood during abdominal aortic vation efforts in cardiac operations at the Mayo Clinic. Mayo Clin Proc
aneurysm repair. Am J Surg 1993;166:97–102. 1988;63:225–229.
133. Williamson KR, Taswell HF. Intraoperative blood salvage: a review. 160. Dzik WH, Jenkins R. Use of intraoperative blood salvage during ortho-
Transfusion 1991;31:662–675. topic liver transplantation. Arch Surg 1985;120:946–948.
134. O’Hara PJ, Hertzer NR, Santilli PH, Beven EG. Intraoperative auto- 161. Van Voorst SJ, Peters TG, Williams JW, Vera SR, Britt LG. Autotransfu-
transfusion during abdominal aortic reconstruction. Am J Surg sion in hepatic transplantation. Am Surg 1985;51:623–626.
1983;145:215–220. 162. Williamson KR, Taswell HF, Rettke SR, Krom RAF. Intraoperative
135. Aaron RK, Beazley RM, Riggle GC. Hematologic integrity after intra- autologous transfusion: its role in orthotopic liver transplantation.
operative allotransfusion: comparison with bank blood. Arch Surg Mayo Clin Proc 1989;64:340–345.
1974;108:831–837. 163. Reul GJ Jr, Solis RT, Greenberg SD, Mattox KL, Whisennand HH. Expe-
136. Brener BJ, Raines JK, Darling RC. Intraoperative autotransfusion in rience with autotransfusion in the surgical management of trauma.
abdominal aortic resections. Arch Surg 1973;107:78–84. Surgery 1974;76:546–555.
137. Ansell J, Parrilla N, King M, et al. Survival of autotransfused red blood 164. Smith RS, Meister RK, Tsoi EK, Bohman HR. Laparoscopically guided
cells recovered from the surgical field during cardiovascular opera- blood salvage and autotransfusion in splenic trauma: a case report.
tions. J Thorac Cardiovasc Surg 1982;84:387–391. J Trauma 1993;34:313–314.
138. Buth J, Raines JK, Kolodny GM, Darling RC. Effect of intraoperative 165. Rebarber A, Lonser R, Jackson S, Copel JA, Sipes S. The safety of intra-
autotransfusion on red cell mass and red cell survival. Surg Forum operative autologous blood collection and autotransfusion during
1975;26:276–278. cesarean section. Am J Obstet Gynecol 1998;179:715–720.
139. Stillman RM, Wrezlewicz WW, Stanczewski BS, Chapa L, Fox MJ, 166. Merrill BS, Mitts DL, Rogers W, Weinberg PC. Autotransfusion:
Sawyer PN. The haematological hazards of autotransfusion. Br J Surg intraoperative use in ruptured ectopic pregnancy. J Reprod Med
1976;63:651–654. 1980;24:14–16.
140. Moore EE, Dunn EL, Breslich DJ, Galloway WB. Platelet abnormalities 167. Klimberg I, Sirois R, Wajsman Z, Baker J. Intraoperative autotransfu-
associated with massive autotransfusion. J Trauma 1980;20:1052–1056. sion in urologic oncology. Arch Surg 1986;121:1326–1329.
141. Griffith LD, Billman GF, Daily PO, Lane TA. Apparent coagulopathy 168. Witte CL, Esser MJ, Rappaport WD. Updating the management of sal-
caused by infusion of shed mediastinal blood and its prevention by vageable splenic injury. Ann Surg 1992;215:261–265.
washing of the infusate. Ann Thorac Surg 1989;47:400–406. 169. Spence RK, Alexander JB, DelRossi AJ, et al. Transfusion guidelines for
142. Dorang LA, Klebanoff G, Kemmerer WT. Autotransfusion in long- cardiovascular surgery: lessons learned from operations in Jehovah’s
segment spinal fusion: an experimental model to demonstrate the Witnesses. J Vasc Surg 1992;16:825–831.
efficacy of salvaging blood contaminated with bone fragments and 170. Johnson RG, Rosenkrantz KR, Preston RA, Hopkins C, Daggett WM.
marrow. Am J Surg 1972;123:686–688. The efficacy of postoperative autotransfusion in patients undergoing
143. Bennett SH, Geelhoed GW, Terrill RE, Hoye RC. Pulmonary effects of cardiac operations. Ann Thorac Surg 1983;36:173–179.
autotransfused blood: a comparison of fresh autologous and stored 171. Semkiw LB, Schurman DJ, Goodman SB, Woolson ST. Postoperative
blood with blood retrieved from the pleural cavity in an in situ lung blood salvage using the Cell Saver after total joint arthroplasty. J Bone
perfusion mode. Am J Surg 1973;125:696–702. Joint Surg Am 1989;71:823–827.
144. Umlas J, O’Neill TP. Heparin removal in an autotransfusor device. 172. Reiner DS, Tortolani AJ. Postoperative peritoneal blood salvage
Transfusion 1981;21:70–73. with autotransfusion after hepatic trauma. Surg Gynecol Obstet
145. Paravicini D, Thys J, Hein H. Use of neomycin-bacitracin irrigating 1991;173:501–504.
solution with intraoperative autotransfusions during orthopedic 173. Glover JL, Broadie TA. Intraoperative autotransfusion. World J Surg
30
operations. Arzneimittelforschung 1983;33:997–999. 1987;11:60–64.
146. Wollinsky KH, Oethinger M, Buchele M, Kluger P, Puhl W, Mehrkens 174. Schmidt H, Lund JO, Nielsen SL. Autotransfused shed mediastinal 433
HH. Autotransfusion bacterial contamination during hip arthroplasty blood has normal erythrocyte survival. Ann Thorac Surg 1996;62:
and efficacy of cefuroxime prophylaxis: a randomized controlled study 105–108.
of 40 patients. Acta Orthop Scand 1997;68:225–230. 175. Jensen CM, Pilegaard R, Hviid K, Nielsen JD, Nielsen HJ. Quality of
147. Rumisek JD, Weddle RL. Autotransfusion in penetrating abdominal reinfused drainage blood after total knee arthroplasty. J Arthroplasty
trauma. In Hauer JM, Thurer RL, Dawson RB (eds). Autotransfusion. 1999;14:312–318.
New York, Elsevier, 1981, pp 105–113. 176. Schmidt H, Bendtzen K, Mortensen PE. The inflammatory cytokine
148. Timberlake GA, McSwain NE. Autotransfusion of blood contaminated response after autotransfusion of shed mediastinal blood. Acta Anaes-
by enteric contents: a potentially life-saving measure in the massively thesiol Scand 1998;42:558–564.
hemorrhaging trauma patient? J Trauma 1988;28:855–857. 177. Mottl-Link S, Russlies M, Klinger M, Seyfarth M, Ascherl R, Gradinger
149. Ozmen V, McSwain NE Jr, Nichols RL, Smith J, Flint LM. Autotransfu- R. Erythrocytes and proinflammatory mediators in wound drainage.
sion of potentially culture-positive blood (CPB) in abdominal trauma: Vox Sang 1998;75:205–211.
preliminary data from a prospective study. J Trauma 1992;32:36–39. 178. Hand GC, Henderson M, Mace P, Sherif N, Newman JH, Goldie DJ.
150. Yaw PB, Sentany M, Link WJ, Wahle WM, Glover JL. Tumor cells Methyl methacrylate levels in unwashed salvage blood following
carried through autotransfusion: contraindication to intraoperative unilaterial total knee arthroplasty. J Arthroplasty 1998;13:576–579.
blood recovery? JAMA 1975;231:490–491. 179. Woda R, Tetzlaff JE. Upper airway oedema following autologous
151. Lane TA. The effect of storage on the metastatic potential of tumor blood transfusion from a wound drainage system. Can J Anaesth
cells collected in autologous blood: an animal model. Transfusion 1992;390:290–292.
1989;29:418–420. 180. Dich-Nielsen JO, Rajan RM, Jensen JJ. An anaphylactoid reaction following
152. Duncan SE, Klebanoff G, Rogers W. A clinical experience with intra infusion of salvaged unwashed drain blood. Can J Anaesth 1998;45:189.
operative autotransfusion. Ann Surg 1974;180:296–304. 181. Faris PM, Ritter MA, Keating EM, Valeri CR. Unwashed filtered shed
153. Linden JV, Kaplan HS, Murphy MT. Fatal air embolism due to periop- blood collected after knee and hip arthroplasties: a source of autolo-
erative blood recovery. Anesth Analg 1997;84:422–426. gous red blood cells. J Bone Joint Surg Am 1991;73:1169–1178.
154. Lennon RL, Hosking MP, Gray JR, Klassen RA, Popovsky MA, Warner 182. Klebanoff G. Early clinical experience with a disposable unit for the
MA. The effects of intraoperative blood salvage and induced hypoten- intraoperative salvage and reinfusion of blood loss (intraoperative
sion on transfusion requirements during spinal surgical procedures. autotransfusion). Am J Surg 1970;120:718–722.
Mayo Clin Proc 1987;62:1090–1094. 183. Schmidt H, Mortensen PE, Folsgaard SL, Jensen EA. Cardiac enzymes
155. Kruger LM, Colbert JM. Intraoperative autologous transfusion in chil- and autotransfusion of shed mediastinal blood after mycoardial revas-
dren undergoing spinal surgery. J Pediatr Orthop 1985;5:330–332. cularization. Ann Thorac Surg 1997;63:1288–1292.
156. Bovill DF, Moulton CW, Jackson WST, Hensen JK, Barcellos RW. The 184. Nguyen DM, Gilfix BM, Dennis F, et al. Impact of transfusion of medi-
efficacy of intraoperative autologous transfusion in major orthopedic astinal shed blood on serum levels of cardiac enzymes. Ann Thorac
surgery: a regression analysis. Orthopedics 1986;9:1403–1407. Surg 1996;62:109–114.
157. Hallett JW Jr, Popovsky M, Ilstrup D. Minimizing blood transfusions 185. Kilgore ML, Pacifico AD. Shed mediastinal blood transfusion after
during abdominal aortic surgery: recent advances in rapid autotrans- cardiac operations: a cost-effectiveness analysis. Ann Thorac Surg
fusion. J Vasc Surg 1987;5:601–606. 1998;65:1248–1254.
BLOOD BANKING 186. Gannon DM, Lombardi AV, Mallory TH, Vaughn BK, Finney CR, 205. Sowade O, Warnke H, Scigalla P, et al. Avoidance of allogeneic blood
Niemcryk S. An evaluation of the efficacy of postoperative blood transfusions by treatment with epoetin beta (recombinant human
salvage after total joint arthroplasty. J Arthroplasty 1991;1:109–114. erythropoietin) in patients undergoing open-heart surgery. Blood
187. Majkowski RS, Currie IC, Newman JH. Postoperative collection and 1997;89:411–418.
reinfusion of autologous blood in total knee arthroplasty. Ann R Coll 206. de Andrade JR, Jove M, Landon G, Frei D, Guifoyle M, Young DC.
Surg Engl 1991;73:381–384. Baseline hemoglobin as a predictor of risk of transfusion and
188. Umlas J, Foster RR, Dalal SA, O’Leary SM, Garcia L, Kruskall MS. Red response to epoetin alfa in orthopedic surgery patients. Am J Orthop
cell loss following orthopedic surgery: the case against postoperative 1996;25:533–542.
blood salvage. Transfusion 1994;34:402–406. 207. Faris PM, Spence RK, Larholt KM, Sampson AR, Frei D. The predictive
189. Martin JW, Whiteside LA, Milliano MT, Reedy ME. Postoperative power of baseline hemoglobin for transfusion risk in surgery patients.
blood retrieval and transfusion in cementless total knee arthroplasty. Orthopedics 1999;22(Suppl 1):S135–S140.
J Arthroplasty 1992;7:205–210. 208. Dunne JR, Malone D, Tracy JK, Gannon C, Napolitano LM. Periopera-
190. Cooley DA, Beall AC Jr, Grondin P. Open-heart operations with dispo- tive anemia: an independent risk factor for infection, mortality, and
sable oxygenators, 5 per cent dextrose prime, and normothermia. Sur- resource utilization in surgery. J Surg Res 2002;102:237–244.
gery 1962;52:713–719. 209. Koscielny J, von Tempelhoff GF, Ziemer S, et al. A practical concept for
191. Petry AF, Jost T, Sievers H. Reduction of homologous blood require- preoperative management of patients with impaired primary hemo-
ments by blood-pooling at the onset of cardiopulmonary bypass. stasis. Clin Appl Thromb Hemost 2004;10:155–166.
J Thorac Cardiovasc Surg 1994;107:1210–1214. 210. McDonald SB, Renna M, Spitznagel EL, et al. Preoperative use of enoxa-
192. Brecher ME, Rosenfeld M. Mathematical and computer modeling of parin increases the risk of postoperative bleeding and re-exploration
acute normovolemic hemodilution. Transfusion 1994;34:176–179. in cardiac surgery patients. J Cardiothorac Vasc Anesth 2005;19:4–10.
193. Weiskopf RB. Mathematical analysis of isovolemic hemodilution indi- 211. Shander A. Surgery without blood. Crit Care Med 2003;31(Suppl 12):
cates that it can decrease the need for allogeneic blood transfusion. S708–S714.
Transfusion 1995;35:37–41. 212. Thompson GE, Miller RD, Stevens WC, Murray WR. Hypoten-
194. Monk TG, Goodnough LT, Brecher ME, et al. Acute normovolemic sive anesthesia for total hip arthroplasty: a study of blood loss and
hemodilution can replace preoperative autologous blood donation as organ function (brain, heart, liver, and kidney). Anesthesiology 1978;
a standard of care for autologous blood procurement in radical pros- 48:91–96.
tatectomy. Anesth Analg 1997;85:953–958. 213. Niemi TT, Pitkanen M, Syrjala M, Rosenberg PH. Comparison of
195. Milam JD, Austin SF, Nihill MR, Keats AS, Cooley DA. Use of sufficient hypotensive epidural anaesthesia and spinal anaesthesia on blood loss
hemodilution to prevent coagulopathies following surgical correction and coagulation during and after total hip arthroplasty. Acta Anaesthe-
of cyanotic heart disease. J Thorac Cardiovasc Surg 1985;89:623–629. siol Scand 2000;44:457–464.
196. Ness PM, Bourke DL, Walsh PC. A randomized trial of perioperative 214. Shapira Y, Gurman G, Artru AA, et al. Combined hemodilution and
hemodilution versus transfusion of preoperatively deposited autolo- hypotension monitored with jugular bulb oxygen saturation, EEG, and
gous blood in elective surgery. Transfusion 1992;32:226–230. ECG decreases transfusion volume and length of ICU stay for major
197. Johnson LB, Plotkin JS, Kuo PC. Reduced transfusion requirements orthopedic surgery. J Clin Anesth 1997;9:643–649.
during major hepatic resection with use of intraoperative isovolemic 215. Karakaya D, Ustun E, Tur A, et al. Acute normovolemic hemodilution
hemodilution. Am J Surg 1998;176:608–611. and nitroglycerin-induced hypotension: comparative effects on tissue
198. Pliam MB, McGoon DC, Tarhan S. Failure of transfusion of autolo- oxygenation and allogeneic blood transfusion requirement in total hip
gous whole blood to reduce banked-blood requirements in open-heart arthroplasty. J Clin Anesth 1999;11:368–374.
surgical patients. J Thorac Cardiovasc Surg 1975;70:338–343. 216. Bock M, Muller J, Bach A, Bohrer H, Martin E, Motsch J. Effects of
199. Sherman MM, Dobnik DB, Dennis RC, Berger RL. Autologous blood preinduction and intraoperative warming during major laparotomy.
transfusion during cardiopulmonary bypass. Chest 1976;70:592–595. Br J Anaesth 1998;80:159–163.
II 200. Weisel RD, Charlesworth DC, Mickleborough LL, et al. Limitations of 217. Johansson T, Lisander B, Ivarsson I. Mild hypothermia does not
blood conservation. J Thorac Cardiovasc Surg 1984;88:26–38. increase blood loss during total hip arthroplasty. Acta Anaesthesiol
434 201. Goodnough LT, Monk TG, Brecher ME. Acute normovolemic hemo- Scand 1999;43:1005–1010.
dilution should replace the preoperative donation of autologous 218. Schmied H, Kurz A, Sessler DI, Kozek S, Reiter A. Mild hypothermia
blood as a method of autologous-blood procurement. Transfusion increases blood loss and transfusion requirements during total hip
1998;38:473–476. arthroplasty. Lancet 1996;347:289–292.
202. Rottman G, Ness PM. Acute normovolemic hemodilution is a 219. Stensrud PE, Nuttall GA, de Castro MA, et al. A prospective, random-
legitimate alternative to allogeneic blood transfusion. Transfusion ized study of cardiopulmonary bypass temperature and blood transfu-
1998;38:477–480. sion. Ann Thorac Surg 1999;67:711–715.
203. Monk TG, Goodnough LT, Brecher ME, Colberg JW, Andriole GL, 220. Birdi I, Regragui IA, Izzat MB, Bryan AJ, Angelini GD. Effects of car-
Catalona WJ. A prospective randomized comparison of three blood diopulmonary perfusion temperature: a randomized, controlled trial.
conservation strategies for radical prostatectomy. Anesthesiology 1999; Ann Thorac Surg 1995;60:747.
91:24–33. 221. Smoller BR, Kruskall MS. Phlebotomy for diagnostic laboratory tests
204. Brecher ME, Goodnough LT, Monk T. Where does preoperative eryth- in adults: pattern of use and effect on transfusion requirements. N Engl
ropoietin therapy count? A mathematical perspective. Transfusion J Med 1986;314:1233–1235.
1999;39:392–395.
Chapter 31
Blood Substitutes: Basic Principles
and Practical Aspects
Robert M. Winslow
Examples Intravascular
(see also Molecular Persistence Oncotic
Class Table 31-3) Radius* (nm) (h) Pressure Viscosity Vasoactivity
*Data from Vandegriff K, McCarthy M, Rohlfs R. Winslow R. Colloid osmotic properties of modified hemoglobins: chemically cross-linked
versus polyethylene glycol surface–conjugated. Biophys Chem 1997;69:23–30.
and physiologic effects have been reported in the open, peer- Based on its oxygen equilibrium curve, this product should
reviewed literature.9 It is therefore a useful model to under- have in vivo oxygen transport properties similar to those of
stand the complexities of the research and development that whole blood; its oxygen equilibrium curve is closer to that of
has taken place over the past 2 decades. red blood cells than that of other types of red blood cell sub-
Production of αα-Hb is complex.10 Stroma-free hemoglo- stitutes (Fig. 31–2). The intravascular persistence is markedly
bin is separated from red blood cell membranes by osmotic extended in the rat: uncrosslinked hemoglobin has a half-life
lysis. It is then deoxygenated to achieve the proper molecular of about 1.2 hours, and crosslinked hemoglobin has a half-
conformation for crosslinking, and the 2,3-diphosphoglyc- life of about 4.3 hours. In the rabbit, the persistence is longer,
erate pocket is blocked reversibly with an allosteric effector. approximately 16 hours for crosslinked hemoglobin, for the
After crosslinking with DBBF, it is heated to pasteurize it and monkey about 14 hours, and for the pig about 7 hours.13 No 31
also to remove unreacted hemoglobin, then passed through a doubt exists that cell-free hemoglobin transports oxygen.
series of cross-flow filters. Finally, it is sterilized by filtration Many studies in the literature have demonstrated the abil- 437
through a 0.2-μm filter. ity of hemoglobin solutions to resuscitate animals in lethal
The product has an oxygen equilibrium curve with an hemorrhagic shock.4
oxygen half-saturation pressure under physiologic condi- DCLHb was studied extensively in clinical trials. However,
tions (i.e., P50) of 28 mm Hg. The degree of cooperativity the product was found to be severely vasoactive; it raised
(the slope of the dissociation curve) is similar to that of blood pressure and increased systemic and pulmonary vas-
blood (Hill coefficient 2.62 for blood, 2.31 for crosslinked cular resistance.14,15 Phase III clinical trials were disappoint-
hemoglobin), and the Bohr11 and carbon dioxide12 effects are ing; trials in trauma16 and in stroke17 both showed reduced
essentially intact. survival in patients who received the product.
1
Br O H O Br PEG-Hb
0.9
O C C C C O 0.8 Human red
Saturation, fraction
Br C O O C Br 0.6
αα-hemoglobin
A OH HO 0.5
0.4 Polymerized
0.3 hemoglobin
O H O
0.2
Lys α199 C C C C Lys α299
0.1
H 0
B
0 20 40 60 80 100 120 140
Figure 31–1 Structure of the crosslinker, bis(2,3-dibromosalicyl)
pO 2, Torr
fumarate (DBBF; A) and crosslinked hemoglobin (B). The brominated
acetyl groups of the DBBF are leaving groups in the reaction, which Figure 31–2 Oxygen equilibrium curves for some representative red
results in a 4-carbon bridge between the two α-chain polypeptide sub- cell substitutes. Note the large difference between the curves for PEG-
units of hemoglobin. Hb, O-raffinose polymerized hemoglobin, and red blood cells.
BLOOD BANKING
Closely related to αα-hemoglobin is a recombinant prod- Other potential approaches to the solution of these prob-
uct (rHb1.1, Optro) produced in Escherichia coli.18 Starting lems could include polymerizable phospholipids37 and other
material for hemoglobin modification can also be produced polymers.38
in transgenic animals.19 Hemoglobin additionally can be
polymerized by means of various polyfunctional reagents to Perfluorocarbon-Based Products
yield molecules with markedly increased molecular weights.
One example is human hemoglobin, reacted first with Fluosol-DA (Fluosol, Green Cross Corp., Osaka, Japan) was
pyridoxal 5′-phosphate and then polymerized with glutar- approved for marketing by the FDA for use in coronary
aldehyde.20 This product has a reduced colloidal osmotic angioplasty in 1990, but manufacture was discontinued in
(oncotic) pressure and longer intravascular persistence 1995 and the product was withdrawn for multiple reasons:
compared with smaller molecules, but the polymerization efficacy was marginal, the product was technically difficult
reaction is notoriously difficult to control.21,22 to use, and new developments in coronary angioplasty (the
Large molecules, such as hemoglobin coupled to polyeth- indication for which it was approved) made its use obsolete.
ylene glycol,23,24 starch,25 or dextran26 may have prolonged Newer perfluorocarbon emulsions were developed by indus-
plasma retention times, may have reduced interactions with try that carried more oxygen than did previous products39;
the reticuloendothelial systems and may extravasate less read- however, regardless of oxygen capacity, a fundamental dif-
ily than smaller molecules. The same result can be achieved ference exists between perfluorocarbon- and hemoglobin-
by directly linking hemoglobin molecules to create large poly- based red blood cell substitutes: oxygen is transported by
mers27 or by use of naturally occurring large polymers such perfluorocarbons as dissolved gas, whereas hemoglobin car-
as found in the earthworm.28 In addition to reduced extrava- ries oxygen chemically bound to the protein itself. Because of
sation, these large molecules have reduced diffusivity in the the nature of the binding of oxygen by hemoglobin, oxygen
plasma, which in turn reduces vasoconstriction by limiting is not released until hemoglobin reaches regions of the circu-
oxygen supply to arterial walls. lation where pO2 is low (i.e., ischemic tissue). In contrast, the
release of oxygen from perfluorocarbon emulsions is linear
with pO2, which means that the bulk of the dissolved oxygen
Liposome-Encapsulated Hemoglobin
will be lost from the circulation before the product reaches
Because hemoglobin is normally packaged inside a mem- ischemic or hypoxic tissue. One product, Oxygent (Alliance
brane, it seems intuitively correct that encapsulated hemo- Pharmaceutical Corp.) is an emulsion of perflubron and
globin would be the ultimate solution to the red blood cell egg yolk phospholipid that contains about five times more
substitute problem. In 1957, Thomas Chang reported the fluorocarbon than Fluosol-DA does and therefore five times
use of microencapsulated hemoglobin as artificial red blood more dissolved oxygen. Oxygent underwent extensive phase
cells.29 Since that time, dramatic results have been reported 2 clinical testing but in critical phase 3 trials of hemodilution
in the complete exchange transfusion of laboratory ani- and perioperative autologous blood collection, there was a
II mals,30,31 but progress toward development of an artificial higher incidence of stroke in treated patients compared with
red blood cell for human use has been slow because of dif- controls.40 Development of Oxygent was subsequently scaled
438 ficulties with reticuloendothelial and other macrophage back, and it is questionable whether trials will continue.
stimulation.32 Other problems include maintaining steril-
ity, limiting hemoglobin oxidation, and the economics of
large-scale production. SAFETY
In the years that have followed Chang’s initial descriptions
of encapsulated hemoglobin, much work with lipid vesicles Demonstration of safety of red blood cell substitutes is a
(liposomes) has been done. Liposomes have served as mod- critical issue, because the risks of transfusion of allogeneic
els for understanding natural cell membranes. They also blood are well known (see Table 31–1). To be used, a sub-
have been used investigationally as vehicles for gene transfer, stitute should be at least as safe as red blood cells, unless a
as targeted carriers, for pharmacologic agents, and even as decisive therapeutic advantage can be demonstrated.
lubricants for degenerated joint surfaces. The most exten- In a review of almost a century of clinical trials with red
sively studied liposomes used to encapsulate hemoglobin are blood cell substitutes, reported side effects involved renal
composed of phospholipid in combination with cholesterol dysfunction and systemic symptoms (fever, chills, nausea,
and other lipids that confer flexibility and stability, such as headache, flushing, vomiting, allergic reactions, tachycar-
ganglioside GM1 or cholesterol.33 When injected into animals, dia, bradycardia, hypertension, rigors, low back pain, chest
such liposomes are rapidly coated with immunoglobulin G, pain, abdominal pain, decreased platelets, and increased par-
albumin, and other opsonins.34 Newer formulations include tial thromboplastin time).4 Many of these effects could be
the use of surface components such as polyethylene glycol or explained by the depletion of nitric oxide, in the case of hemo-
dextran, which can stabilize the liposomes in the circulation.35 globin-based products, or by stimulation of macrophages, in
Hemoglobin vesicles (HbV) are being developed in Japan and the case of liposomes or perfluorocarbon emulsions. Many
may enter clinical trials soon.36 are smooth muscle effects, and some involve macrophages
The limitations to the development of liposome-encap- and platelets. Preclinical animal studies with hemoglobin-
sulated hemoglobin as a red blood cell substitute are diffi- based solutions clearly have not been completely successful
culties in stabilizing the final product and the massive scale in predicting human reactions to the products.41
that would be required to produce a commercial product. Cell-free hemoglobin is widely distributed in the tissues
The size of most liposome particles is approximately 0.2 to after administration. Studies of the distribution of cross-
1.0 μm, too large to be filter-sterilized. In addition, neither linked hemoglobin in the intact animal show that significant
the liposome nor its hemoglobin contents can withstand amounts of hemoglobin are retained in the kidney, spleen,
pasteurization temperature without some type of stabilizer. liver, adrenal gland, lung, heart, brain, and muscle well after
any hemoglobin is detected in plasma.42–44 Thus, cell-free
BLOOD SUBSTITUTES
products tend to be removed from the circulation by the
hemoglobin is distributed in almost every tissue of the body, phagocytic cells of the reticuloendothelial system.54 This situ-
and there could be unpredictable or unknown toxic effects. ation leads to significant enlargement of the liver and spleen.
Extensive histologic studies were carried out in animals after Current research is aimed at prolongation of the intravascu-
exchange transfusion and summarized.4 lar persistence to minimize this problem.55
The effect of cell-free hemoglobin of most concern is its
known ability to cause vasoconstriction and hypertension.
This vasoconstriction can be mediated in part by the reaction EFFICACY
of hemoglobin with nitric oxide, an endothelium-derived
relaxing factor.45 Nitric oxide is synthesized from arginine in It seems intuitively obvious that a plasma expander that
endothelial (and other) cells by an enzyme, nitric oxide syn- carries oxygen would be superior to one that does not, and
thase, which produces nitric oxide and citrulline. It binds to experimental proof of this concept should be relatively
a heme group in guanylate cyclase that activates cyclic gua- straightforward. However, the problem of efficacy can be
nosine monophosphate. Nitric oxide diffuses rapidly out of appreciated by considering the difficulties in showing effi-
endothelial cells into the vessel lumen, the interstitial space, cacy for red blood cell transfusions. The problem is a lack
and smooth muscle cells, where it binds to a heme group in of clear end points: no single measure of oxygen transport
guanylate cyclase, activating cyclic guanosine monophosphate is accurate and easily obtainable. It may be possible to show
and moving calcium from the unbound to bound state. The improved clinical outcome after transfusion of red blood
result is smooth muscle relaxation. Nitric oxide also stimu- cells to patients with extremely low hematocrits, but the bulk
lates platelets and polymorphonuclear leukocytes and mac- of allogeneic blood is given intraoperatively in response to
rophages. Hemoglobin binds nitric oxide very tightly, more blood loss and hemodynamic instability, not severe anemia.
tightly in fact than it binds oxygen, whether hemoglobin is in Most demonstrations of efficacy in animals have been
the red blood cell or free in solution.46 The reaction is virtu- either by exchange transfusions with test material or by
ally irreversible. Whether this interaction of hemoglobin with resuscitation from shock. Resuscitation from shock is exceed-
nitric oxide will limit clinical usefulness of hemoglobin-based ingly complex, however, and the most urgent requirement is
red blood cell substitutes remains to be determined. for volume replacement.56 Clinical trials involving trauma
Nitric oxide binding by cell-free hemoglobin is not the patients are particularly difficult to design because of the
sole explanation for its vasoconstrictor activity, however.47 problems of controls and informed consent. Future clinical
In addition to its effect as a scavenger of nitric oxide, cell- trials will most likely be aimed at, for example, reduced use
free hemoglobin may induce vasoconstriction by disrupting of allogeneic blood, rather than at specific oxygen transport
normal autoregulation of vascular tone. In other words, it parameters, which may be controversial, at best. For exam-
may make oxygen so readily available to regulatory arteri- ple, one trial with Fluosol-DA during surgical procedures
oles that reflexive vasoconstriction could occur that para- showed that its use did not reduce the need for allogeneic
doxically limits blood flow.48,49 This concept is suggested by blood transfusions in the postoperative period.57 31
direct observation in the microcirculation50 and has led to
new design strategies for cell-free red blood cell substitutes.51 439
Perfluorocarbon emulsions have the most extensive history CLINICAL TRIALS
of use in humans. Fluosol-DA was approved by the FDA
for use in coronary angioplasty and therefore was given to Early trials with various cell-free hemoglobin solutions
many human patients. Similar formulations have been used were reviewed and showed an array of side effects involving
on the battlefield in China and Afghanistan, although data every organ of the body.4 However, most of these are mild or
are generally not available. Perfluorocarbon emulsions have reversible, and only one death in more than 211 patients was
also been tested in humans as imaging agents. The principal reported in the early literature; this patient was terminally ill
toxicity of perfluorocarbon emulsions appears to be in their and would likely have died even without the administration
stimulation of macrophages.52,53 This can result in pulmo- of hemoglobin.58
nary hypertension and elaboration of thromboxane in swine Certain red blood cell substitute products are in various
and could lead to nonspecific symptoms such as fever, chills, stages of advanced clinical trials (Table 31–3). The greatest
and flulike symptoms in humans. concern for hemoglobin-based products is that the known
Biocompatibility studies with liposome-encapsulated vasoactivity of the solutions could lead to hypertension or
hemoglobin have been generally favorable,31 but such underperfusion of ischemic tissue. A major concern with
SVI, mL/m2
hypovolemia and anemia often occur as a result of blood 80
loss. Tissue hypoxia or anoxia may occur as a result of either SVI 60
decreased blood flow (stagnant hypoxia) or decreased O2- 70
carrying capacity (anemic hypoxia).21–24 The body primar-
ily attempts to preserve O2 delivery to vital organs through 60
increased myocardial contractility and heart rate as well as
increased arterial and venous vascular tone mediated through 50
50
increased sympathetic activity. In addition, central and
regional reflexes redistribute organ blood flow. The adrener- 6
gic system plays an important role in altering blood flow to
2
1·m
and within specific organs. The renin-angiotensin-aldoste- 5
rone hormone system is stimulated to retain both water and
10
is quite possible that the complex interrelationship between
disease severity, the number of transfusions, and the degree of
7 anemia may have resulted in a spurious association between
a cardiovascular diagnosis and the reported mortality risk
4 with anemia.
Wu and associates105 retrospectively studied Medicare
records of 78,974 patients older than age 65 who were hospi-
1 talized with a primary diagnosis of acute myocardial infarc-
6 7 8 9 10 11 12+ tion. The authors then categorized patients according to their
Preoperative Hemoglobin (g/Dl) hematocrit on admission. Although anemia, defined in the
study as a hematocrit less than 39%, was present in nearly half
the patients, only 3680 patients received an RBC transfusion.
No cardiac disease Cardiac disease Lower admission hematocrit values were associated with
Figure 32–2 Adjusted odds ratio for mortality by cardiovascular dis- increased 30-day mortality, with mortality rate approach-
ease and preoperative hemoglobin. (Adapted from Carson JL, Duff A, ing 50% among patients with a hematocrit of 27% or lower
Poses RM, et al. Effect of anemia and cardiovascular disease on surgical who did not receive an RBC transfusion. Unfortunately, this
mortality and morbidity. Lancet 1996;348:1055–1060.) study did not have any data on nadir hemoglobins and their
RED CELL TRANSFUSION IN PERIOPERATIVE AND CRITICALLY ILL PATIENTS
relationship to mortality. Interestingly, whereas RBC trans- pretransfusion hemoglobin concentration was 84 ± 13 g/L. In
fusion was associated with a reduction in 30-day mortality an effort to control for confounding created by illness sever-
for patients who received at least one RBC transfusion if ity and the need for transfusion, these investigators employed
their admitting hematocrit was less than 33%, RBC trans- a strategy of matching transfused and nontransfused patients
fusion was associated with increased 30-day mortality for based on their propensity to receive a transfusion, thereby
patients whose admitting hematocrit values were 36.1% or defining two well-balanced groups (516 patients in each group)
higher. In the analysis, these associations were present even to determine the influence of RBC transfusions on mortality.
when adjustments were made for clinical patient factors Using this approach, the associated risk of death was increased
including APACHE II scores, location of myocardial infarc- instead of decreased by 33% for patients who received a trans-
tion and presence of congestive heart failure, and treatment fusion compared with similar patients who did not receive
factors including use of reperfusion therapies, aspirin, and blood. However, as pointed out in the accompanying edito-
β-adrenergic blockade. rial,108 the results may have differed if the propensity scores
In the only study exclusively focusing on the perioperative were derived separately for categories of pretransfusion hemo-
period, Carson and colleagues106 attempted to determine the globin concentrations (e.g., <80, 80–100, and >100 g/L) instead
effect of perioperative transfusion on 30- and 90-day postop- of hemoglobin concentrations at ICU admission. For example,
erative mortality with a retrospective cohort study involving if one were to consider groups of patients with pretransfusion
8787 patients with hip fractures undergoing repair between hemoglobin concentrations less than 60 g/L, it is unlikely that
1983 and 1993 in 20 different U.S. hospitals. This was a large, the observed 33% increase in mortality would hold true, or
high-risk, elderly (median age, 80.3 years) population with blood transfusion would never be recommended.
extensive coexisting disease and with an overall 30-day mor- Unfortunately, as evidenced by a recent systematic review,
tality rate of 4.6%. A total of 3699 patients (42%) received a there is a paucity of clinical trials comparing restrictive
perioperative transfusion within 7 days of the surgical repair. and liberal transfusion policies to examine efficacy of RBC
After controlling for hemoglobin concentrations, cardio- transfusion. Carson and coworkers109 were able to identify
vascular disease, and other risk factors for death, the results only 10 randomized clinical trials of adequate methodologi-
suggested that patients who had hemoglobin concentra- cal quality in which different RBC transfusion triggers were
tions as low as 80 g/L and did not receive transfusion were evaluated. Included were a total of 1780 surgery, trauma,
no more likely to die than those with similar hemoglobin and ICU patients enrolled in trials conducted over the past
concentration levels who received a transfusion. With hemo- 40 years. The transfusion triggers evaluated in these trials
globin concentrations < 80 g/L, nearly all patients received a varied between 70 and 100 g/L. Data on mortality or hospi-
transfusion, which did not allow investigators to draw con- tal length of stay were available in only six trials (Fig. 32–3).
clusions about the effect of transfusion at these lower hemo- Conservative transfusion triggers were not associated with
globin concentration levels. However, as the authors point an increase in mortality; on average, mortality was one fifth
out, despite the large sample size, inadequate power may still lower (RR 0.80; 95% CI 0.63, 1.02) with conservative com-
explain the inability to detect a reduction in mortality related pared with liberal transfusion triggers. Likewise, cardiac 32
to transfusion; they estimated that the study would need to be morbidity and length of hospital stay did not appear to be
10 times larger to detect a 10% difference in 30-day mortality adversely affected by the lower use of red cell transfusions. 449
with 80% power. There were insufficient data on potentially relevant clinical
More recently, Vincent and colleagues107 completed a pro- outcomes such as stroke, thromboembolism, multiorgan fail-
spective observational cross-sectional study involving 3534 ure, delirium, and infection or delayed wound healing to per-
patients admitted to 146 Western European ICUs during a 2- form any pooled analysis. Carson and colleagues109 stated that
week period in November 1999. Thirty-seven percent of these the data were insufficient to address the full range of risks and
patients received RBC transfusions during their ICU admission benefits associated with different transfusion thresholds, par-
with the overall transfusion rate increasing to 41.6% over a 28- ticularly in patients with coexisting disease. They also noted
day period. For those patients who were transfused, the mean that their meta-analysis was dominated by a single trial: the
1 5 10
Favours restrictive Favours liberal
Figure 32–3 Effect of restrictive transfusion triggers on 30-day, all-cause mortality. (Adapted from Carson JL, Hill S, Carless P, et al. Transfusion
triggers: a systematic review of the literature. Transfus Med Rev 2002;16:87–199.)
TRANSFUSION MEDICINE
Transfusion Requirements in Critical Care (TRICC) trial,110 The two groups were fairly equally balanced with regard to
which enrolled 838 patients and was the only individual baseline characteristics and concurrent therapies with a few
trial identified that was adequately powered to evaluate the exceptions: there was less frequent diuretic use in the restric-
impact of different transfusion strategies on mortality and tive group (43% vs 58%, P <0.01) and the use of epidural
morbidity. anesthetics was greater in the restrictive group (8% vs 2%, P <
The TRICC study110 documented an overall nonsignifi- 0.01). Overall, in this subgroup analysis, there was no signifi-
cant trend toward decreased 30-day mortality (18.7% vs cant difference in mortality rate between the two treatment
23.3%, P = 0.11) and significant decreases in mortality among groups. However, there was a nonsignificant (P = 0.3) decrease
patients who were less acutely ill (8.7% vs 16.1%, P = 0.03) in overall survival rate in the restrictive group for patients with
in the group treated using a hemoglobin transfusion trigger confirmed ischemic heart disease, severe peripheral vascular
of 70 g/L compared with a more liberally transfused group disease, or severe comorbid cardiac disease.
that received 54% more red cell transfusions. The investiga- The subgroup analysis of patients receiving mechanical
tors also noted that the 30-day mortality rates were signifi- ventilation was limited to 713 (85% of the 838 patients in
cantly lower with the restrictive transfusion strategy among the TRICC trial who required invasive mechanical venti-
patients who were less acutely ill (APACHE II scores less than latory support).112 Of these, 357 had been in the restric-
20) and among patients who were less than 55 years of age tive RBC transfusion group and 356 in the liberal group.
(Fig. 32–4). The mean duration of mechanical ventilation was 8.3 ± 8.1
A number of additional questions arose from the TRICC days in the restrictive group and 8.8 ± 8.7 days in the lib-
trial. The investigators were particularly interested in the risks eral group (P = 0.48). Ventilator-free days were 17.5 ± 10.9
and benefits of anemia and transfusion in patients with car- and 16.1 ± 11.4 in the restrictive and liberal RBC transfu-
diovascular disease and in patients attempting to wean from sion groups, respectively (P = 0.09). Eighty-two percent of
mechanical ventilation. In the first of these subgroup analy- the patients in the restrictive transfusion group were con-
ses,111 357 patients (43%) were identified with cardiovascular sidered successfully weaned and extubated for at least 24
disease. Of these, 160 had been in the restrictive RBC transfu- hours, compared with 78% for the liberal group (P = 0.19).
sion group and 197 in the liberal transfusion strategy group. Among the 219 patients who required mechanical ventila-
100 100
Restrictive strategy
Restrictive
90 strategy 90
Survival (%)
Survival (%)
80 80 Liberal strategy
III Liberal
strategy
70 70
450
60 P=0.10
60 P=0.020
50 50
0 5 10 15 20 25 30 0 5 10 15 20 25 30
A Time (days) B Time (days)
100
Restrictive strategy
90
Liberal strategy
Survival (%)
80
70
60
P < 0.02
50
0 5 10 15 20 25 30
C Time (days)
Figure 32–4 ICU survival over 30 days in study patients in the restrictive and liberal allogeneic red blood cell transfusion strategy groups. Graph
A illustrates Kaplan-Meier survival curves for all patients in both study groups. There is a trend toward lower mortality in patients in the restrictive
group (dotted line) compared with the liberal group (solid line) (P = 0.10). In the subgroup with an APACHE II score less than 20 (graph B), fewer
patients died in the restrictive group than in the liberal group (P = 0.02). There were also significant differences in survival between groups in the
subgroup with ages less than 55 years (P = 0.02), as depicted in graph C. (Adapted from Hébert PC, Wells G, Blajchman M, et al. A multicenter,
randomized, controlled clinical trial of transfusion requirements in critical care. N Engl J Med 1999;340:409–417.)
RED CELL TRANSFUSION IN PERIOPERATIVE AND CRITICALLY ILL PATIENTS
tion for more than 7 days, there were no differences in the ALTERNATIVES TO TRANSFUSION
time to successful weaning (Fig. 32–5). The independent
effects of RBC transfusions and hemoglobin concentration Numerous strategies have been explored and recommended
were also examined. Each additional transfusion was asso- to decrease or to eliminate the need for blood transfusions
ciated with an increased duration of mechanical ventilation during major surgery and critical illness. Some are relatively
(RR = 1.10; 95% CI 1.14 to 1.06, P < 0.01) adjusting for benign, but others carry risks that must be weighed against
the effect of age, APACHE II score, and comorbid illnesses. the benefit of red cell administration. Alternatives include
Hemoglobin concentrations did not influence the duration decreasing the use of medications that result in perioperative
of mechanical ventilation (RR = 0.99; 95% CI 1.01 to 0.98, bleeding such as nonsteroidal anti-inflammatory drugs and
P = 0.45). Complications including pulmonary edema acetylsalicylic acid (ASA); avoidance of unnecessary phle-
and ARDS were increased in patients in the liberal strategy botomies and the use of blood conservation strategies such
group. as pediatric test tubes and arterial line reinfusion set-ups;
Even though a large RCT has been completed, a number of medications to decrease blood loss, such as antifibrinolytic
questions remain. One of the most important is why the lib- agents; and medications to increase hemoglobin production.
eral RBC transfusion strategy failed to improve 30-day mor- In addition to a restrictive transfusion strategy, the two most
tality and rates of organ failure in critically ill patients. It is useful approaches to decrease red cell transfusions in criti-
conceivable that the greater number of allogeneic RBC units cally ill patients appear to be blood conservation techniques
in the liberal group significantly depressed host immune such as decreased phlebotomies and erythropoietin therapy.
responses112,113 or resulted in altered microcirculatory flow Other therapeutic strategies are better suited to patients
as a consequence of prolonged RBC storage times. undergoing high-risk surgical procedures.
Following the publication of the TRICC trial, a study Decreased red cell production is one of the causes of
published by Rivers and colleagues114 documented that anemia in the critically ill. Indeed, critical illness is char-
early use of goal-directed care based on a mixed central acterized by a blunted erythropoietin production and
venous saturation decreased mortality from 46.5% in the response.115 This blunted response appears to result from
control group to 30.5% in the goal-directed therapy group inhibition of the erythropoietin gene by inflammatory
(P = 0.009). As one of the many interventions in patients mediators.116,117 It has also been shown that these same
with early septic shock, hematocrit concentrations were inflammatory cytokines directly inhibit red cell produc-
increased beyond 30% if the central venous saturations tion by the bone marrow and may produce distinct abnor-
fell below 70%. As a consequence of goal-directed therapy, malities of iron metabolism.118,119 In patients with multiple
64% of patients compared with 18.5% of the control group organ failure, recombinant human erythropoietin therapy
received RBC transfusions (P < 0.0001). There are signifi- (600 units/kg) has been shown to stimulate erythropoie-
cant differences in patient populations between the study sis.120 Similarly, in a small, randomized, placebo-controlled
conducted by Rivers and colleagues and the TRICC trial. trial (160 patients), therapy with recombinant human
The early goal-directed therapy study does highlight the erythropoietin resulted in an almost 50% reduction in RBC 32
need to perform additional studies in subpopulations of transfusions versus placebo.121 Erythropoietin was given at
critically ill patients. a dose of 300 units/kg daily for 5 days followed by every 451
other day therapy until ICU discharge. Despite receiv-
ing fewer RBC transfusions, patients in the recombinant
human erythropoietin group had a significantly greater
increase in hematocrit.
100
Recently, the efficacy of recombinant human erythropoi-
Patients remaining ventilated (%)
III
454
Chapter 33
Post-Transfusion Red Blood Cell
and Platelet Survival and Kinetics:
Basic Principles and Practical Aspects
Richard J. Davey ● James P. AuBuchon
Investigating the fate of transfused blood cells can offer gamma-counting instrument (gamma counter) is 100 to
important information for the manufacturers of blood 300 keV. The “yield” of gamma-photon emissions is the num-
collection and storage systems, the transfusion specialists ber of photons emitted per 100 radioactive decays. A high
who collect and prepare the units, and patients with disor- yield means that a lower radiation dose can be used to achieve
ders involving reduced survival or abnormal trafficking of adequate detection levels.
these cells. Tracking the recovery and survival of blood cells
also allows for validation of systems’ capabilities (and their
Chromium-51
improvement with innovations) as well as diagnostic proce-
51
dures that can lead to therapeutic interventions for a variety Cr is a useful red cell label and also has utility as a platelet
of disorders. label. Advantages of this radionuclide include ease of red cell
The clinical and research applications of these techniques labeling, excellent red cell uptake, low toxicity, and low and
have led to their adoption by many medical centers. However, stable elution rate. 51Cr is produced in a reactor by neutron
the technical complexity and regulatory requirements for activation. It decays by electron capture, with a radioactive
handling radionuclides have limited their availability. Newer half-life of 27.7 days. The principal gamma-photon emission
techniques that avoid the use of radioactive substances are occurs by electron capture at 320 keV, slightly above the opti- 33
under study and may eventually play a larger role in clini- mum detection range of standard gamma counters. The high
cal applications as well as in trials of new blood collection, energy and low yield (9.8%) of gamma-photon emissions
processing, and storage systems. from decay of this radionuclide necessitate a larger (typi- 455
cally 3-inch) NaI crystal for efficient detection in a gamma
counter. Otherwise, a larger dose of this radionuclide may be
RADIONUCLIDES IN TRANSFUSION necessary to achieve adequate counts.
MEDICINE The quite long radioactive half-life of 51Cr makes it suit-
able for most red cell survival and recovery studies. However,
Gray and Sterling1 described the first clinically useful red cell this long half-life poses difficulties when multiple studies
radiolabel, chromium-51, in 1950. Fisher and colleagues2 need to be performed in a short amount of time, since the
introduced technetium-99m as a cell label in 1967, and the overlapping survival curves make separation of the individ-
importance of indium-111 was recognized after an evalua- ual studies difficult. This problem can be addressed by cor-
tion of leukocyte radiolabels by McAfee and Thakur3 in 1976. recting for residual radiation, by separating studies until the
Other cell radiolabels have also been evaluated and used, nuclide decays, or by using alternative radionuclides.4
with varying degrees of success. However, the vast major-
ity of blood cell survival and imaging studies are performed
with either chromium-51 (51Cr), technetium-99m (99mTc),
or indium-111 (111In).
Table 33–1 Desirable Characteristics of Blood
Each of the three commonly used radionuclides has advan-
Cell Radiolabels
tages and disadvantages that affect its applicability in clinical
and research settings. Investigators should have an under- Minimal radiation dose to the recipient
standing of fundamentals of nuclear medicine and radiation Nontoxic to the recipient
biophysics to use these materials appropriately. Desirable Nontoxic to the cell
characteristics of a blood cell radiolabel include factors that Specific for the cell
No metabolism of the label by the cell
relate to recipient safety as well as ease and reproducibility of Radioactive half-life appropriate for the study
use (Table 33–1). Radioactive emissions suitable for efficient detection or
The selection of an appropriate radionuclide requires imaging
familiarity with the radioactive half-life, elution characteris- Minimal manipulation of the cell required
tics, and gamma-photon energy (Table 33–2). The optimal No elution of the label
No relabeling of other cells in vivo
range for detection of photoelectric events in a standard
TRANSFUSION MEDICINE
Table 33–2 Major Features of Radionuclides Used for Blood Cell Labeling
51 99m 111
Cr Tc In
*
Target organ is dependent on cell type and clinical or experimental situation as well.
51
Cr is supplied as sodium radiochromate (Na251CrO4) for 99m
Tc, as pertechnetate, diffuses across the red cell mem-
cell labeling purposes. Hexavalent chromate labels red cells by brane into the cell, where it binds to hemoglobin and other
first binding rapidly and reversibly to the red cell membrane. intracellular ligands with a labeling efficiency of about 90%.
After reduction to the trivalent state, the chromic ion binds The short half-life of 99mTc precludes its use for long-term
more slowly and firmly to the β-globin chain of intracellular red cell survival studies. In fact, groups of samples must
hemoglobin (and probably to other intracellular ligands as be counted rapidly, and, even with that, a correction must be
well). Labeling efficiency is about 90%. Younger red cells take made for radioactive decay that occurs during the counting
up slightly more label than do older cells. procedure. However, because of its short half-life, 99mTc is
Chromium elutes from red cells in two phases. There is useful as a red cell label when a series of survival studies must
an early loss of 1% to 4% of the label within 24 hours, which be performed for transfusion compatibility. This radionu-
may represent elution of a loosely bound fraction of the clide is also useful for the accurate determination of red cell
label.5 Garby and Mollison6 found that subsequent elution volume as part of red cell recovery studies (using 51Cr) or in
varies from 0.56% to 2.04% per day, similar to the 0.70% to patients being investigated for possible pathologic expansion
1.55% reported by Bentley and associates.7 An estimated elu- of their red cell mass (e.g., polycythemia).
tion rate of 1% per day is satisfactory for most investigational This nuclide also has a high and variable rate of elution.
and clinical purposes.8 Trivalent chromium that elutes from The use of stannous compounds as intracellular reducing
the labeled cells is rapidly excreted in the urine and does not agents in the labeling process has permitted firmer binding
relabel other cells in vivo. of 99mTc, but elution rates of 4% per hour with considerable
Moroff and colleagues9 have described the technical steps variation may be expected.16,17 Heaton and colleagues18 have
for labeling red cells with 51Cr for the evaluation of blood stor- described the technical steps necessary to minimize vari-
III age and preservation systems. The International Committee ability in labeling and elution with this nuclide. Its use as
for Standardization in Hematology10 has described a slightly a platelet label has received little attention, however, primar-
456 different technique suitable for labeling red cells for the ily because its short half-life would prevent determination
determination of transfusion compatibility. of postinfusion survival over the following 5 to 10 days, a
Baldini and colleagues11 first described the use of 51Cr parameter that is essential in the assessment of the effects of
as a platelet radiolabel. The platelet-labeling efficiency of platelet collection and storage conditions.
chromium-51 is low (9%) compared with its red cell label-
ing efficiency (90%). Therefore, it is important that platelet
Indium-111
preparations be free of red cells before being labeled.12 A pro-
cedure for labeling platelets with 51Cr, which is more diffi- 111
In can be used as a label for red cells, platelets, and leuko-
cult than that for red cells, has been described by Snyder and cytes. It is prepared in a cyclotron by proton bombardment
colleagues.13 Further useful refinements were developed by of a cadmium target. It has two major gamma-photon emis-
Heaton, Holme, and colleagues.14,15 Unfortunately, the dose sions—173 keV (90.5% yield) and 247 keV (94% yield)—
of 51Cr required for imaging studies, which require a high which are excellent for counting in a gamma counter and
yield of gamma-photon emissions, is typically unacceptably for external imaging. The radioactive half-life of the nuclide
high for most other applications in transfusion medicine. is 2.83 days, making it appropriate for leukocyte imaging
studies that extend over several days.
111
Technetium-99m In must be complexed with a lipophilic chelating agent
if it is to traverse cell membranes and label intracellular pro-
99m
Tc, which is widely employed as an imaging agent in teins. Although several lipophilic chelates have been stud-
nuclear medicine, is another useful red cell label. This meta- ied, 8-hydroxyquinoline (oxine) is the only agent currently
stable radionuclide is produced in a molybdenum genera- licensed for this purpose in the United States. Technical pro-
tor as a product of molybdenum-99 decay. 99mTc decays, in cedures have been published for labeling red cells,19 platelets,13
turn, to 99Tc, with a half-life of 6.02 hours. In doing so, 99mTc and leukocytes20 with 111In.
emits a gamma-photon at 140 keV, which falls within the Although 111In is a nonselective label, it is preferentially
optimal range of gamma counters. The high yield (90%) of taken up by platelets. Reduced uptake is observed by leuko-
this radionuclide is useful for external imaging. Because of cytes, and still less by red cells.21 Cell preparations therefore
the frequent use of this radionuclide in imaging procedures, must be free of extraneous elements before being labeled.
many hospitals maintain a generator on site, and thus there Like 99mTc,111In has a high and variable rate of elution from
is a readily available and inexpensive source of 99mTc for cell red cells. AuBuchon and Brightman19 have shown that 8% to
labeling applications. 10% of the label elutes within 24 to 48 hours, with a 4% per
POST-TRANSFUSION RED BLOOD CELL AND PLATELET SURVIVAL AND KINETICS
day loss after that time. This problem with elution renders
111
In unsuitable for red cell studies without a specific cor-
rection factor,22 but 111In is useful for determining red cell
volume when precision is not critical.
Red Cells
The red cell life span (110 to 120 days) was first approxi-
mated by differential absorption techniques23 and subse-
quently confirmed by radioisotopic labeling studies using
cohort (marrow) labels24 and random (peripheral blood)
labels.6 When chromium-51 is used as the label, the time at
which the recovered counts are 50% of the originally injected
counts is the T50Cr, normally 31 ± 6 days. Correction for elu-
tion will yield the true red cell life span. The T50Cr is also
useful for evaluating the extent of hemolysis in chronic hemo- Figure 33–1 Red cell survival patterns representing different mecha-
lytic anemia. Modest reductions in the T50Cr can indicate a nisms of immune red cell destruction. (1), Normal survival; (2), two-
component curve with partial complement activation to C3b→C3d,g;
substantial increase in the rate of red cell destruction.25 (3), single exponential curve characteristic of IgG non-complement-
Red cell life span studies also have been conducted to activating antibodies; (4), complete complement activation to C5b-C9
assess the therapeutic benefit of transfused units that are complex with intravascular hemolysis.
enriched in young red cells (“neocytes”). The transfusion of
such units has the theoretical benefit of reducing the iron
burden in patients who undergo frequent transfusions by to 40 IgM molecules per red cell being required to initiate
increasing the number of long-lived red cells in the circu- red cell clearance. IgG antibodies can activate complement if
lation.26 Chromium-51 life span studies have demonstrated two or more molecules are physically adjacent on the red cell
that neocytes survive 30% to 60% longer in the circulation membrane37; thus, many more IgG molecules are required
than do red cells derived from standard red cell units.27,28 for complement activation to occur. In most cases, the kinet-
However, because neocyte units contain only about half the ics of complement activation result in generation of the
hemoglobin of standard units, their use results in only a mod- membrane-bound C3b fragment but not the final C5b-C9
est reduction in transfusion requirements.29,30 Additionally, “membrane attack” complex. Plasma factor I, with plasma 33
since they are costly to prepare, neocyte transfusions have factor H as a cofactor, cleaves C3b to C3bi. Red cell–bound
not achieved wide clinical acceptance. C3b and C3bi adhere to the complement receptors CR1 and 457
In contrast to long-term life span studies, short-term CR3 on reticuloendothelial system (RES) macrophages in
red cell survival studies are useful in identifying patterns the liver and spleen. These red cells may thus be removed
of immune red cell destruction. These studies are clinically from the circulation by either antibody-dependent cell-
important when standard pretransfusion serologic testing mediated cytotoxicity or complete phagocytosis, or they may
cannot clearly identify or characterize red cell alloantibod- sustain membrane damage through partial phagocytosis.
ies or autoantibodies in a transfusion recipient. The various Alternatively, C3bi may be degraded before red cell damage
factors that contribute to immune red cell destruction have occurs.38,39
been discussed in detail elsewhere.31–33 Of primary impor- Plasma factors I and H further cleave C3bi to C3c, which
tance are the immunoglobulin class and subclass of the detaches from the cell, and C3d,g, which remains attached
recipient red cell alloantibody34 and the extent to which that to the cell membrane. Because C3d,g binds weakly to CR3
antibody activates complement.35,36 receptors, most red cells coated with C3d,g detach from RES
Four characteristic survival patterns can be recognized macrophages and survive relatively normally.40 The inac-
in a 24-hour red cell survival study conducted to determine tive C3d,g fragments may protect the red cell from further
transfusion compatibility (Fig. 33–1). They are (1) normal complement-mediated damage by occupying complement-
survival, (2) extravascular destruction characterized by a binding sites. A “two-component” recovery curve, therefore,
“two-component” survival curve, (3) extravascular destruc- represents the early loss of C3b-coated red cells and more
tion characterized by a “single exponential” survival curve, normal survival of those cells on whose membranes C3b
and (4) intravascular destruction. has degraded to C3d,g. A “two-component” survival curve
with more than 70% recovery at 24 hours indicates that the
Normal Survival patient can safely receive larger quantities of similar red cells
The normal range for recovery of fresh, compatible red cells with little risk of rapid immune hemolysis.
is ≥97% at 60 minutes and 95% to 100% at 24 hours.
Extravascular Destruction Characterized
Extravascular Destruction Characterized by a “Single Exponential” Survival Curve
by a “Two-Component” Survival Curve Many IgG red cell alloantibodies (e.g., Rh antibodies) do not
Most immunoglobulin M (IgM) red cell alloantibodies and activate complement. Instead, there is progressive removal of
some IgG alloantibodies (e.g., anti-Jka) can activate comple- sensitized red cells with a survival curve described by a sin-
ment. IgM antibodies directly activate complement, with 20 gle exponential. Factors such as the concentration and IgG
TRANSFUSION MEDICINE
subclass of the antibody determine the rate of removal of the form lipophilic complexes with 111In, notably, acetylacetone,48
cells. The efficiency of the four IgG subclasses in binding Fc tropolone,49 and 2-mercaptopyridine-N-oxide (merc),50 have
receptors on RES macrophages is IgG3 > IgG1 > IgG2 > IgG4. been extensively studied as possible alternative agents to over-
IgG1 is present in higher titer than the other subclasses and is come this problem. Tropolone is widely used in Europe, but
most often involved in IgG-mediated red cell destruction.34 oxine remains the only licensed agent in the United States.
Red cell survival studies that demonstrate an IgG-medi- Heaton and colleagues contributed to the advancement
ated “single exponential” pattern of cell destruction indicate of platelet labeling accuracy by developing a technique for
that the transfusion of similar cells may result in a delayed determining the amount of radionuclide that elutes from
hemolytic transfusion reaction. platelets shortly after labeling and infusion.15 With this “elu-
tion correction,” the recovery curves for platelets labeled
Intravascular Destruction with 111In are coincident with those labeled with 51Cr. This
IgM antibodies that are efficient activators of complement, procedure is the basis of the standard platelet radiolabel-
such as anti-A and anti-B, can drive the kinetics of the reac- ing protocol adopted by the Biomedical Excellence for Safer
tion to the formation of the terminal C5b-C9 “membrane Transfusion (BEST) Collaborative.51
attack” complex. Insertion of this complex into the red cell There are four methods for the calculation of platelet life
membrane results in loss of membrane integrity and osmotic span—linear, exponential, weighted mean, and γ-function
lysis of the cell. Transfused cells that generate complement (multiple hit)—each of which yields slightly different results.
activation of this magnitude are usually hemolyzed within The γ-function analysis is the recommended method13,43 and
minutes. the most frequently used, but investigators often report their
results using one or more additional statistical methods. An
Synergistic Effect of Complement and IgG easy-to-use computer program that models platelet survival
Complement and IgG appear to act synergistically in pro- by a variety of techniques performs these calculations in a
moting red cell opsonization. Studies have demonstrated standardized, validated manner.52,53 These calculations can
that the addition of complement to red cells sensitized with also be performed on statistical analysis software to achieve
IgG alone will enhance phagocytosis.41,42 A hybrid red cell the same results. The area under the survival curve has also
survival curve may occur when both IgM and IgG antibod- been reported, but some believe this measure places an
ies are active. There is a rapid loss of a minor population of inordinate weight on the survival parameter.54
the cells from IgM-mediated complement activation, and a Because no radioactive tracer selectively labels newly
slower phase of cell destruction mediated primarily by the formed platelets, cohort studies have not been practical to
IgG antibody. measure platelet production. Instead, the kinetics of normal
platelet production has been measured indirectly by deter-
mining the turnover rate of circulating platelets (platelet
Platelets
count divided by platelet survival corrected for recovery).
III Most studies of platelet life span and kinetics have been Estimates of daily platelet production in the steady state
performed with chromium-51 and/or indium-111 using using this method have ranged from 35,000–66,000/μL of
458 labeling procedures defined by the ICSH43 and Heaton and blood.55 The life span of human platelets ranges between 8
Holme.15 The low labeling efficiency, long half-life, and poor and 12 days; Harker and Finch56 found it to be 9.5 ± 0.6 days.
imaging characteristics of 51Cr have led to preferential use Senescent platelets are removed from the circulation by RES
of 111In when a single platelet label is required.44 However, macrophages in the liver and spleen and, to a lesser extent, by
serial studies in a subject using a single label may not be pos- bone marrow and lungs.57,58
sible, since platelet kinetics vary over time. By using both There is also evidence of a fixed platelet requirement nec-
chromium and indium labels for simultaneous infusion, a essary to maintain vascular integrity. Hanson and Slichter59
reduction in this variability is seen, and a more accurate esti- have proposed that 82% of platelet turnover in normal
mation of the difference between platelets stored in a stan- persons is due to senescence, and 18% (≈7100 platelets/
dard (or licensed) system and those in a test system can be μL/day) is due to the fixed requirement. The fixed require-
achieved.45 Using “dual-label” approaches in a protocol can ment becomes an increasingly large component of platelet
allow simultaneous storage and labeling of test and control removal as thrombocytopenia from bone marrow hypoplasia
platelets (e.g., in different bag types). This technique has also worsens. Thus, platelet life span correlates directly with the
been used to create novel approaches to the timing of collec- platelet count; an accelerated reduction in life span is noted
tions, such as in a study where whole blood–derived platelets as platelet counts decrease below 50,000/μL. The shortened
stored for 5 days were compared with those stored for 7 days. platelet life span observed in thrombocytopenic patients,
The test (7-day) platelets were derived from a unit collected therefore, does not necessarily indicate an increased platelet
on day 0 (with reinfusion of the red cells), and the control destruction rate.
(5-day) platelets were derived from a collection 2 days later. Patients with increased peripheral platelet destruction,
Platelets from both study arms were labeled and infused such as those with idiopathic thrombocytopenic purpura,60
simultaneously.46 have platelet life spans that are shorter than predicted by
Standard technical methods for preparation of radio- Hanson and Slichter’s model.59 Thus, platelet life span stud-
labeled platelets have been published by the ICSH and oth- ies using 111In are useful in discriminating between thrombo-
ers.13,15,47 The major drawback of 111In-oxine is its high affinity cytopenia caused by decreased platelet production and that
for red cells, leukocytes, and plasma proteins, primarily trans- caused by increased platelet destruction (Fig. 33–2). Platelet
ferrin. The washing and centrifugation steps necessary to life span studies are not generally useful in the alloimmu-
prepare a “clean” platelet preparation cause a substantial loss nized patient, however. Post-transfusion platelet counts usu-
of platelets (≈40%) and a “collection injury” that can distort ally provide the clinician with the information necessary for
test results if not performed carefully. Other compounds that the management of these patients.
Lymphocytes
Granulocytes
Studies done with granulocytes labeled with diisopropyl- PROTOCOLS TO EVALUATE COMPONENT
fluorophosphate demonstrated that the circulating cells COLLECTION, PROCESSING, AND
had a half-life of 3.8 to 6.7 hours. About half of the injected STORAGE SYSTEMS
cells were recoverable in the circulation, with the remainder
constituting a “marginating pool” of cells.63,64 More recent The two critical advances that led to modern blood pres-
studies done with 111In-labeled granulocytes have shown ervation techniques were the introduction of acid-citrate-
a mean recovery of 30% of the labeled cells and a mean dextrose in 194377 and the development of plastic blood
circulating half-life of 5.0 hours.65 Imaging studies have containers in the early 1950s. Now, blood components are
shown that labeled and transfused granulocytes transiently stored in various anticoagulant-preservative or additive
sequester in the lungs. The extent and duration of the pul- solutions, in plastic containers with differing characteristics,
monary sequestration appear to be more severe when the and at nonphysiologic temperatures for extended periods.
labeled cells are suspended in saline rather than plasma.66 The advent of pathogen inactivation systems offers great
Alloimmunized patients demonstrate a more prolonged potential for improving transfusion safety but at potential
and intense pulmonary sequestration phase than do nor- risk of injury to the cells during the process. For this reason,
mal subjects.67 In addition, HLA (human leukocyte anti- the safety and efficacy of new or modified storage conditions
gen)– and granulocyte-specific antibodies appear to reduce must be determined by both in vitro and in vivo evaluation
the intravascular half-life of transfused granulocytes68 of cell function and viability.
and to impair their ability to migrate to a site of active A variety of biochemical, hematologic, and functional
infection.69 parameters are often measured across the storage period in
TRANSFUSION MEDICINE
to receive sufficient therapeutic benefit from the component
under study.
Major
changes
Patient Red Cells
clinical
trial The 24-hour post-transfusion recovery of red cells stored
In vivo radio- under experimental conditions is the standard measure of
labeled Significant
autologous acceptability used in the United States. The FDA requires that
changes
recovery and a mean of 75% or more of the transfused red cells be recov-
survival ered 24 hours after infusion. The lower bound of the 95%
In vitro biochemical, confidence interval is expected to be at least 70%. Although
Minor
hematologic and
functional tests changes these criteria have never been encoded in federal regula-
Increasing Complexity tions or statutes, they have become the accepted measures
risk of patient of to obtain licensure. Long-term survival is not a parameter
impact testing Frequency of application
usually measured, since, at least before the advent of patho-
Figure 33–3 Scheme for progressive analysis of new blood com- gen inactivation technologies, postinfusion longevity was
ponent. A representation of the progressive analysis of a new blood assumed to be normal providing the red cell showed accept-
component illustrating (1) the stepwise nature of investigation (begin-
ning with in vitro testing and culminating with clinical trials) and (2) able 24-hour survival.80 Other measures, such as poststorage
application of different degrees of scrutiny based on the extent to ATP and supernatant hemoglobin levels, contribute to the
which the new component differed from previously investigated ones. assessment of storage condition acceptability. Red cell recov-
Note that major changes would involve more extensive investigation ery data have been important in studies evaluating the effect
using more complex testing techniques in response to an increase in
the risk of (adverse) patient impact. More commonly encountered, sim-
of gamma-irradiation,81,82 blood bag plasticizing agents,83
pler changes to a collection or storage system would more likely require and prestorage leukocyte reduction84 on stored red cells.
only in vitro testing. Successful testing (indicative of safety and efficacy) The technique for performing post-transfusion recovery
at one level of the pyramid would be used to justify moving to the next studies to assess new storage conditions differs in impor-
“higher” level of testing, were that required. (Adapted from a presen- tant ways from red cell recovery studies to assess transfusion
tation by Dr. Jaro Vostal.79 Reprinted with permission from AuBuchon
JP. Radioisotopic reflections. Transfusion 2005;45:28S–32S.) compatibility. As noted in Table 33–3, autologous red cells
are chosen for storage studies, thereby avoiding the risks
of transmissible disease and transfusion reactions inher-
ent in allogeneic transfusions. The shape of the recovery
an attempt to predict the in vivo outcome after transfusion. It curve is most important in transfusion compatibility stud-
is known, for example, that retention of at least half the orig- ies. However, the recovery percentage at 24 hours is of prime
inal adenosine triphosphate (ATP) concentration in red cells importance for studies evaluating storage systems.
III is necessary for adequate recovery and that a pH22°C ≥ 6.2 for Because an accurate determination of 24-hour recovery
platelets is required in order to retain viability. Beyond these is necessary, the method of establishing the zero-time point
460 notable exceptions, however, few strong correlations between (100% recovery) is of critical importance. A direct measure-
any in vitro parameters and life span in circulation after stor- ment of the zero-time recovery value is not possible, how-
age have been identified.78 Therefore, documentation of the ever, because an infused blood sample is not completely
in vivo recovery (and, for platelets, survival) remains neces- mixed until approximately 3 minutes after transfusion in
sary whenever a substantial change in collection, processing, a normal person.85 Senescent infused cells and cells dam-
or storage is being proposed. The FDA has described the role aged during storage are removed from the circulation by
of in vitro and in vivo testing as part of a “pyramid” of trials the spleen, with loss of these cells beginning immediately
leading to approval of a new device or protocol79 (Fig. 33–3). after infusion. By 3 minutes, a significant loss of the infused
Radiolabeled kinetic studies thus remain a critical feature of cells may have occurred that is not detected if the 3-min-
this approach, which allows definitive documentation in a ute sample is regarded as representing 100% survival. In this
relevant (human) system and ensures that patients are likely case, the radioactive counts in the 3-minute sample would be
Table 33–3 Features of Red Cell Survival Studies to Evaluate Transfusion Compatibility
and Red Cell Storage Systems
Platelets
The evaluation of platelet storage and preservation has been
complex and challenging. Investigators generally conduct
studies using normal volunteer subjects, with each subject
serving as his or her own control. Data are usually reported
as platelet recovery at 1 to 3 hours and as platelet survival
or half-life over several days. Both 51Cr and 111In are suit-
able radiolabels for platelet storage studies,89–91 and double
labeling with 111In and 51Cr is well established.15,92 Platelet
recovery and survival studies have been critical to the devel-
opment of plastics that permit 5-day platelet storage,93,94 to
the identification of the optimal temperature for platelet
storage,95,96 to the determination of proper agitation of the
platelet concentrates,97 and to the comparison of manual
Figure 33–4 Two methods to calculate the counts per minute (CPM) platelet preparation with automated plateletpheresis tech-
at zero-time (T0) for determination of red cell mass. With the single- niques.98 These studies have traditionally compared a new
label technique (solid dots), the 51Cr-labeled test cells are infused and
samples are obtained at 5, 7.5, 10, 12.5, and 15 minutes after infu-
approach for collection, processing, or storage with an estab-
sion. A regression line is plotted, and the y-intercept (T0) is determined. lished one.
In the double-label technique, an independent, or “true,” measure of Unlike red cell studies, however, there is no established
T0 is determined (arrow) with fresh autologous red cells labeled with criterion of acceptable performance. The test system has usu-
99m
Tc. The single-label regression technique underestimates CPM at T0 ally been accepted if (1) the statistical analysis failed to show
and, therefore, overestimates 24-hour red cell recovery, with the error
increasing as red cell storage time increases. (Adapted from Beutler E, a statistical difference in comparison to an established (i.e.,
West C. Measurement of the viability of stored red cells by the single- previously licensed) method, or (2) a difference was docu-
isotope technique using 51Cr. Transfusion 1984;24:100.) mented but it was not regarded as being clinically significant.
TRANSFUSION MEDICINE
There are obvious limitations to this approach. Trials involv- is suspected. For these needs, red cells damaged by heat and
ing radionuclide injections in normal subjects are usually labeled with 99mTc localize in the spleen without significant
small (N < 25) to limit exposure to radioactivity and costs hepatic uptake.105
and, as a result, have limited power to detect real differences The blood pool can be imaged, and related red cell kinet-
that may be present. When differences are detected, the defi- ics determined, with red cells labeled with 111In. Quantitative
nition of what is a clinically relevant difference has remained sequential counting and imaging are possible, focusing on a
subjective. Repetitive comparisons—as today’s “test” system selected target organ such as the spleen. This procedure can
becomes tomorrow’s “control system”—may lead inevita- determine the rate of splenic sequestration when increased
bly to a “creeping inferiority”54 of platelet components that peripheral destruction of red cells is suspected (Fig. 33–5).
could negatively affect patient care. Arbitrary standards for However, this technique is infrequently utilized.
recovery and survival of platelets could be developed, but A common use of radionuclide labeling of autologous red
one would still need to validate a laboratory’s techniques. cells is for the determination of red cell mass (or volume)
An alternative standard based on fresh platelets has been in the diagnosis of polycythemic states. 99mTc provides the
proposed by Murphy.54 The control arm for trials would requisite information with a relatively small dose of radioac-
always be the infusion of fresh (autologous) platelets. Test tivity. Samples need only be acquired over the first 15 to 30
platelets would be expected to demonstrate at least two minutes after reinfusion.
thirds the recovery and half the survival of the standard-
setting fresh platelets. The actual protocol design for this
kind of trial has received much discussion.99 Currently
utilized apheresis platelets meet the criterion proposed by
Murphy,100 as do both apheresis101 and whole blood–derived
platelets46 stored for 7 days. This approach needs to be tested
through additional studies, but it has promise to permit clear
regulatory decisions as well as to increase the assurance of
platelet transfusion efficacy for patients.
Granulocytes
Granulocytes collected by a variety of techniques and labeled
with 111In migrate to sites of infection. They demonstrate
a similar kinetic pattern in the lung, liver, and spleen.102,103
McCullough and colleagues104 demonstrated that post-
transfusion granulocyte migration to skin window chambers
III was impaired when the cells were stored for 8 hours at 1° C to
6° C or for 24 hours at 22° C. Granulocytes stored for 8 hours
462 at 22° C, however, migrated as well as unstored controls.
Red Cells
The diagnostic uses of imaged radiolabeled red cells include
the localization and sizing of the spleen or accessory spleen,
blood pool scanning, and the detection of gastrointestinal
bleeding and vascular anomalies such as hemangiomas.
Colloids such as rhenium sulfur labeled with 99mTc are often
used to determine spleen size and location. Radiolabeled
colloids also image the liver. On occasion, it is desirable to Figure 33–5 111In-labeled red cell blood pool scan. The heart, great
image the spleen alone, especially when an enlarged liver vessels, liver, and enlarged spleen are clearly imaged. Differential count-
interferes with splenic images or when an accessory spleen ing over selected organs can identify sites of red cell sequestration.
Platelets
III
466
Chapter 34
Transfusion of the Patient with
Congenital Coagulation Defects
Suzanne Shusterman ● Catherine S. Manno
Method of Viral
Inactivation or Specific Activity of
Product Manufacturer Depletion Stabilizer Final Product*
IAC, immunoaffinity chromatography, SD1, solvent–detergent (TNBP and Triniton X-100); P, pasteurization (60°C 10 h); DH, dry heat (72°C,
72 h); SD2 (TNBP and polysorbate-80).
*
IU factor VIII/mg total protein, including stabilizer.
†
Distributed by Aventis Behring.
‡
Distributed by Genetics Institute.
§Manufactured from American Red Cross–collected plasma, distributed by American Red Cross.
||Contain von Willebrand factor.
¶For use in patients with inhibitors to factor VIII.
a rFVIII with structure homologous to the previous prod- The gene for FIX was cloned in 1982 and was success-
uct Recombinate but made without addition of exogenous fully transfected into Chinese hamster ovary cells, a process 34
human or animal proteins. Efficacy and safety have been leading to rFIX production in 1985 by Genetics Institute.36–39
demonstrated in patients for the treatment of hemophilia Initially, rFIX was tested in a canine model and in previ- 469
A without increased immunogenicity.32 ously treated patients and was shown to have clinical efficacy
comparable to that of plasma-derived products, with a low
Factor IX Concentrates thrombogenic potential even at high doses.37,40 Inhibitor for-
Three types of factor concentrates are available for treatment mation was also low; only 1 of 44 previously treated patients
of patients with FIX deficiency: PCCs, coagulation FIX con- developed a low-titer inhibitor that was detectable for 11
centrates, and recombinant FIX (rFIX) (Table 34–2). months.41
Prothrombin complex concentrates, also known as FIX In October 1995, a trial of rFIX in previously untreated
complex concentrates, contain FIX as well as prothrom- patients was initiated. Although the final results of this trial
bin, FVII, and FX, some of which becomes activated during are not yet published, preliminary reports show good clinical
preparation.17 PCCs are low-purity products with a specific efficacy with no adverse effects, including no viral transmis-
activity less than 50 IU/mg total protein.1 When PCCs are sion. The only difference detected between plasma-derived
used at frequent intervals or for prolonged periods, they and rFIX in these initial studies was that recovery of rFIX
have been associated with paradoxic thrombotic complica- is approximately 20% less than in the plasma-derived prod-
tions, such as myocardial infarction, venous thromboembo- ucts. Poorer than anticipated recovery has been attributed
lism, and disseminated intravascular coagulation.16,33 These to minor differences in post-translational modifications
problems are caused either by the presence of activated fac- between rFIX and plasma-derived products.40 One rFIX
tors that serve to trigger coagulation or by accumulation of concentrate, BeneFix, is currently available in the United
high levels of the factors.16 High-purity FIX concentrates, States. The specific activity of BeneFix is greater than 200
also known as coagulation FIX concentrates, were first to 360 IU/mg protein, and the final preparation does not
licensed in 1992.16 They are purified by immunoaffinity or contain human albumin.28
gel chromatography, contain only FIX, and have little or no
thrombotic potential.16,17 High-purity concentrates are pre-
ferred over PCCs, particularly when frequent replacement Nonconcentrate Therapeutic
is required, such as in surgical patients or in clinical situ- Options for Hemophilia
ations associated with an increased risk of thrombosis, as
Desmopressin
in patients with advanced liver disease.12 FIX concentrates
currently available in the United States include Alphanine Desmopressin (DDAVP, 1-deamino-8-d-arginine vasopres-
SD and Mononine.16,34,35 sin) is a synthetic analog of vasopressin that stimulates
TRANSFUSION MEDICINE
Table 34–2 Factor IX Concentrates Available in the United States in 2005
Method of Viral
Inactivation or Specific Activity
Product Manufacturer Depletion Stabilizer of Final Product*
*
IU factor VIII/mg total protein, including stabilizer.
†
Distributed by Nabi.
DAC, dual-affinity chromatography; DH; dry heat (72°C, 72 h); IAC, immunoaffinity chromatography; NF, nanofiltration; SD, solvent-
detergent (TNBP and polysorbate-80); ST, sodium thiocyanate; UF, ultrafiltration; VH, vapor heat (10h, 60°C 1190 mbar pressure plus 1h, 80°C
1375 mbar).
Table 34–3 Factor Replacement Guidelines for Common Bleeding Problems in Hemophilia
Duration of Factor
Type of Bleeding Target Factor Level (IU/dL) Replacement Comments
Titer at Time of
Type of Patient Hemorrhage Agent Dosage Agent/Method Dosage
High responder <10BU PCC 75 IU/kg q12h Human FVIII 100–200 IU/kg, q12h*
aPCC 75 IU/kg q12h Porcine FVIII 50–200 IU/kg*†
aPCC 75–100 IU/kg q12h
rFVIIa 90–120 μg/kg q2h
>10BU PCC 75 IU/kg q12h Porcine FVIII 50–200 IU/kg*†
aPCC 75 IU/kg q12h aPCC 75–100 IU/kg q12h
rFVIIa 90–120 μg/kg q2h
Plasmapheresis 100–200 IU/kg q12h
and high-dose
human FVIII
Low responder <5 BU Human FVIII 50–75 IU/kg q 8–12 h Human FVIII 100–200 IU/kg*
<10 BU PCC 75 IU/kg q12h Human FVIII 100–200 IU/kg q12h*
aPCC 75 IU/kg q12h Porcine FVIII 20–50 IU/kg*†
aPCC 50–100 IU/kg q12h
rFVIIa 60–120 μg/kg q2h
when immune tolerance therapy is initiated when the inhibi- VON WILLEBRAND DISEASE
tor titer is low.77 Other data show improved outcomes when
the interval between inhibitor detection and starting immune Von Willebrand disease is the most common congenital coag-
tolerance therapy is short.77 It is therefore important to start ulation disorder, occurring in approximately 1% to 2% of the
immune tolerance therapy as soon as a patient’s high-titer, population.82,83 It is actually a collection of disorders caused
high-response status is confirmed. Once immune tolerance by either quantitative or qualitative abnormalities of vWF, a
has been achieved, management of bleeding episodes can be glycoprotein that is important in both primary and second-
III the same as before inhibitor development. An international ary hemostasis.84 In contrast to hemophilia, mucosal bleed-
immune tolerance study is currently assessing optimal dose ing, especially epistaxis and menorrhagia, and easy bruising
474 and interval for patients with high-titer inhibitors.78 are the most frequent clinical signs, although more significant
bleeding can occur with trauma or surgery. Bleeding symp-
Comprehensive Hemophilia Care
toms are often most severe in children and adolescents.85
Patients with hemophilia have lifelong disease punctuated by vWD occurs at the same rate in male and female patients and
hospitalization and the need for expensive intravenous medi- is generally inherited as an autosomal trait with variable pen-
cation. The management of such complicated patients is best etrance.86 In contrast to hemophilia, bleeding symptoms may
supervised by a multidisciplinary expert team that includes a vary among members of affected families.85
hematologist, a hemophilia nurse specialist, a physical thera-
pist, a social worker, an orthopedist, and a dentist. Regular Structure and Function of
clinic visits for evaluation by these health care providers give
von Willebrand Factor
the patient with hemophilia a solid foundation for manag-
ing the long-term medical and psychosocial issues associated The gene for vWF is located on chromosome 12 and is
with chronic disease. 178 kb in length with 52 exons.87 There is also a pseudogene
on chromosome 22 with 97% homology.88 vWF is initially
Gene Therapy
transcribed as a large, single-chain protein with 2813 amino
Hemophilia is an attractive disease for a gene therapy acids (prepro VFW) that undergoes a series of modifications
approach for several reasons.79 First, small increases in cir- to become the mature multimeric vWF.85 First, a small sig-
culating levels of FVIII or FIX through expression of a suc- nal peptide is removed to form pro-vWF. Next, the remain-
cessfully transferred gene would likely result in significant ing peptide undergoes glycosylation, sulfation, and dimer
reduction of bleeding episodes. Second, tight regulation of formation, leading to the formation of vWF multimers
transgene expression is not necessary because normal levels of various sizes ranging in molecular weight from 500 to
of FVIII and FIX extend to 150%. Gene therapy trials using 20,000 kD.89 vWF is the largest soluble protein in humans. Its
numerous different viral and nonviral approaches to trans- plasma concentration and functional activity are generally
fer the recombinant gene to human cells, including liver expressed either as a percentage of normal or as units per
and muscle cells, have been completed, demonstrating no milliliter related to a normal plasma pool calibrated against
major safety problems but no long-term efficacy either.79,80 an international plasma standard for vWF.89
Whether a gene therapy approach will be safe or effective has Von Willebrand factor is synthesized in endothelial cells
yet to be determined.81 and megakaryocytes. In endothelial cells, vWF is stored in
TRANSFUSION OF THE PATIENT WITH CONGENITAL COAGULATION DEFECTS
Weibel-Palade bodies and is secreted into the plasma or genetic implications can be predicted by type. Mixed pheno-
subendothelial matrix, predominantly as high-molecu- types can exist, resulting from compound heterozygosity.91
lar-weight multimers.90 Platelet vWF, consisting of mostly
Type 1
low-molecular-weight multimers, is synthesized in the
megakaryocyte and is stored in platelet α granules.89 vWF Type 1 disease is most common and accounts for 70% to
multimers with the highest molecular weight are most effec- 80% of patients diagnosed with vWD.89,93 Patients with type
tive in hemostasis because they have a large surface area with 1 disease have a mild to moderate decrease in plasma lev-
a high concentration of binding sites for various ligands and els of vWF with a normal vWF multimer pattern. Type 1
receptors.89 disease is usually inherited in an autosomal dominant pat-
Von Willebrand factor is important in both primary and tern. Clinically, patients with type 1 disease usually have only
secondary hemostasis. It is a cofactor in platelet adhesion mild symptoms and are often only identified during routine
and a participant in platelet aggregation. In response to vas- preoperative screening tests.85,89
cular injury, a conformational change occurs in plasma vWF
Type 2 (2A, 2B, 2M, and 2N)
that allows it to bind to platelet glycoprotein Ib.85,90 Platelets
then adhere to the site of endothelial damage and become Type 2 vWD disease is diagnosed in 15% to 20% of affected
activated after interactions with subendothelial components patients and is characterized by a qualitative defect in vWF
such as collagen. Platelet activation causes release of stored caused by missense mutations and small inframe deletions
platelet components, including vWF from α granules, and or insertions.84,89 Typically, patients have more significant
expression of activated glycoprotein IIb/IIIa, the receptor that bleeding symptoms than are seen in type 1 disease. Family
is involved in platelet aggregation.90 Platelet vWF can bind history is often positive with this diagnosis, with both auto-
to glycoprotein IIb/IIIa to facilitate aggregation, although somal dominant and recessive patterns of inheritance seen.89
fibrinogen is the main cofactor in this reaction.85 vWF also Type 2 disease can be further classified into four variants
plays an important role in the coagulation cascade, by acting according to the particular vWF abnormality.
as a carrier protein for FVIII so that it is protected from pro- Type 2A vWD is the most frequent subtype and is char-
teolytic cleavage in the plasma.90 Deficiency of vWF causes acterized by decreased vWF platelet-dependent function
FVIII to have a reduced half-life and leads to a deficiency in and an absence of high- and intermediate-molecular-weight
plasma FVIII. In addition, vWF plays a role in stimulating vWF multimers as demonstrated on agarose gel electro-
FVIII release into plasma. This is shown in vitro, where cells phoresis.84,91 This multimer abnormality can result from
transfected with the FVIII gene express rFVIII more effi- either abnormal synthesis of vWF or increased breakdown
ciently when vWF is present in the media or is cotransfected of circulating vWF. Type 2A vWD is usually inherited as a
along with FVIII. In addition, in vivo, when plasma from dominant trait.90,91
FVIII-deficient patients is infused into patients with vWD, Type 2B is a rarer form of vWD in which an abnor-
an increase in FVIII in the patients with vWD is seen.85 mal vWF has excessive affinity for platelet glycoprotein
Ib. This causes excessive platelet activation and removal 34
from the circulation and leads to variable thrombocytope-
Classification
nia as well as decreased plasma concentrations of vWF.84,86 475
As is evident from the previous discussion, vWF plays several Thrombocytopenia may be intermittent and is often exac-
critical roles in platelet function and thrombus formation. erbated by stress such as infection.86 Type 2B is inher-
For vWF to function normally, several requirements must ited as a dominant trait and is associated with absence of
be met, including the presence of high-molecular-weight high-molecular-weight multimers.84,93
multimer forms, the appropriate release of activated vWF, Type 2M disease describes patients with decreased plate-
and the presence of binding sites for platelet glycoproteins, let-dependent function that is similar to type 2A disease,
FVIII, and constituents of the subendothelial cell matrix. except high-molecular-weight multimers are not absent.
Abnormalities of any of these components can lead to vWD, Decreased platelet function is instead caused by structural
with particular deficiencies leading to different clinical phe- or functional defects in binding regions of the vWF protein.
notypes. In addition, because the vWF locus is autosomal Like type 2A and 2B disease, this type is usually inherited as
and the protein is polymeric, the phenotype of a particular a dominant trait.84,93
patient with vWD is the result of interactions between two Type 2N disease (also known as vWD Normandy) refers
different alleles and protein products that lead to the further to patients with vWF variants with decreased binding affinity
clinical heterogeneity of this disorder.91 More than 20 sub- for FVIII. This leads to decreased survival of FVIII in the cir-
types of vWD have been identified so far.92 The current clas- culation and a moderate to severe reduction of FVIII plasma
sification system for vWD, approved by the Subcommittee levels with normal vWF levels. Affected persons therefore
on vWF at the 39th Annual Meeting of the Scientific and have a mild hemophilia phenotype. This type usually has an
Standardization Committee of the International Society of autosomal recessive inheritance.84–86
Thrombosis and Haemostasis in 1993, attempts to simplify
Type 3
this clinical heterogeneity by classifying patients based on
pathophysiology, clinical behavior, and genetic abnormali- Type 3 vWD is characterized by near-complete deficiency
ties.91 This classification scheme is based on the principle of vWF (less than 10%) as well as a secondary deficiency of
that all vWD is the result of mutations of the vWF locus. In FVIII, resulting in a severe bleeding disorder with abnor-
the scheme are three major categories of vWD: type 1 dis- malities in both primary and secondary hemostasis.93 This
ease, a partial quantitative deficiency of vWF; type 2 disease, type of vWD is very rare, with a prevalence of 1 to 3 cases
a qualitative deficiency of vWF; and type 3 disease, a com- per 1 million, and is inherited as an autosomal recessive
plete quantitative deficiency of vWF. The clinical course and disorder.94
TRANSFUSION MEDICINE
1 vWD.82,95 Finally, the vWF can be separated by agarose gel
Diagnosis
electrophoresis into high-, intermediate-, and low-molecu-
Von Willebrand disease should be suspected in patients with lar-weight multimers. All the multimers are present in type
a history of increased bruising, prolonged bleeding after 1 disease and are absent in type 3 disease. In type 2 vWD,
minor trauma, or mucosal bleeding, particularly epistaxis gel electrophoresis can vary with disease type, but gener-
and menorrhagia, and is supported by a family history of ally shows a reduction in high- and intermediate-weight
abnormal bleeding. Initial screening tests include prothrom- multimer.86 Table 34–5 summarizes the laboratory findings
bin time (PT), PTT, platelet count, and bleeding time. The according to vWD type.
PT is usually normal, whereas the PTT may be prolonged by
a decrease in FVIII.93 The platelet count is also usually nor- Therapeutic Products
mal, although, as described earlier, thrombocytopenia can be
present in patients with type 2B disease.86 The bleeding time, Superficial bleeding, particularly in type 1 vWD, can usually
which is usually prolonged secondary to abnormal platelet be managed by applying local pressure, ice, or a local topi-
function, may be normal in patients with type 1 disease with cal hemostatic agent. Systemic therapy is generally reserved
normal platelet vWF content, but it is rarely used today in the for bleeding at sites not controllable by local measures or
diagnostic workup.93 All these screening tests may be normal as prophylaxis before and after surgical treatment. The two
in persons with vWD, so further testing may be needed when main approaches to systemic therapy of vWD are increasing
suspicion of disease is significant.95 the release of endogenous vWF and exogenous replacement
Specific tests used to evaluate patients for vWD include of vWF. Selecting the appropriate therapy for an individual
vWF antigen, FVIII assay, vWF activity (also known as ristoce- patient depends on the specific type of vWD and clinical
tin cofactor activity), and vWF multimer analysis. Because nor- treatment goal.44
mal physiologic variation can occur in plasma levels of vWF,
Desmopressin
FVIII, and vWF activity, repeated plasma measurements over
time may be necessary to establish a diagnosis of disease.86,95 As previously discussed, DDAVP, a synthetic analog of
Results of these tests vary according to type of vWD. vasopressin, causes immediate endothelial cell release of
Von Willebrand factor antigen is decreased in type 1 dis- vWF, FVIII, and plasminogen activator.43 It is effective in
ease, decreased or normal in type 2 disease, and undetect- patients with vWD who have adequate stores of functional
able in type 3 disease.93 vWF is an acute phase reactant, and vWF. Therefore, it is useful in type 1 disease but ineffective
factors such as pregnancy, exercise, infection, and cigarette in type 3 disease. DDAVP can be used in some patients with
smoking, as well as medications, including corticosteroids, type 2A vWD, but it is not recommended in patients with type
birth control pills, and DDAVP, can increase plasma con- 2B disease because it can exacerbate thrombocytopenia.90,97
centrations of vWF and need to be taken into consideration Patients with type 2N disease generally have high FVIII levels
when one evaluates test results.86 In addition, vWF antigen in response to DDAVP, but the released FVIII has a shorter
III varies with ABO blood types and therefore should be inter- than normal half-life that limits its therapeutic utility in
preted in reference to values specific for the patient’s blood major bleeding episodes.97 DDAVP is given at the same dose
476 type; persons with blood group O have the lowest mean vWF used in mild FVIII deficiency, either intravenously or as a
antigen level.96 FVIII levels are normal or mildly decreased concentrated nasal spray (Stimate). Infusion of 0.3 μg/kg
in patients with types 1 and 2 vWD.93 In contrast, patients body weight results in an average three- to fivefold increase
with type 2N and type 3 vWD can have an extremely low in vWF and FVIII; nasal dosing is slightly less effective.43,90
FVIII level.86 The ristocetin cofactor activity assay measures Before therapeutic use, patients should have a trial to mea-
the vWF function, specifically the interaction of vWF with sure individual response to DDAVP.98 Response to DDAVP is
platelet glycoprotein Ib. Patients with type 1 and type 3 vWD generally consistent over time, and family members usually
tend to have a decrease in vWF activity that is proportional have similar responses.99 DDAVP is used preferentially over
to their decrease in vWF antigen, whereas those with type 2 plasma-derived products in patients who have an adequate
disease have a more substantial decrease in vWF activity than response, because it does not carry risks of viral transmis-
vWF antigen.86,93 Studies have shown that measurement of sion. In addition, DDAVP is substantially less expensive than
vWF activity is the most sensitive test for diagnosing type exogenous vWF replacement.43
Adapted from Montgomery RR, Gill JC, Scott JP. Hemophilia and von Willebrand disease. In Nathan DH, Orkin SH (eds). Naton and OSKI’s
Hematology of Infancy and Childhood, vol 2, 5th ed. Philadelphia, WB Saunders, 1998, p 1646.
TRANSFUSION OF THE PATIENT WITH CONGENITAL COAGULATION DEFECTS
Desmopressin may not be useful when prolonged hemo- RARE INHERITED CONGENITAL
stasis is required. When it is administered more often than CLOTTING DISORDERS REQUIRING
once every 24 to 48 hours, decreased effectiveness may be TRANSFUSION THERAPY
observed because of depletion of storage pools (tachyphy-
laxis).99 Therefore, when DDAVP is used for several days,
Fibrinogen Deficiencies
coagulation parameters should be followed closely (vWF,
FVIII, and ristocetin cofactor) to monitor the need for alter- Congenital afibrinogenemia is a rare disorder, with an esti-
native replacement therapy. mated incidence of 1 to 2 per 1 million that is inherited as an
autosomal recessive trait with the gene located on chromo-
Exogenous von Willebrand Factor Replacement some 4.102,103 It is characterized by virtual absence of fibrin-
Exogenous replacement of vWF is used in patients in ogen secondary to deficient liver cell synthesis. Symptoms
whom DDAVP is not effective, including most patients range from minimal bleeding to life-threatening hemor-
with type 2B, type 2M, and type 3 vWD, as well as those with rhage, and are commonly seen in the newborn period and
type 1 and 2A disease who do not show adequate response include hematomas or intracranial hemorrhage from birth
in a DDAVP trial. Fresh frozen plasma contains both FVIII trauma, bleeding from the umbilicus, and excessive bleed-
and vWF; however, the large volumes required to achieve ing after circumcision. Similar to hemophilia, spontaneous
hemostasis limit its clinical use.98 Cryoprecipitate, contain- hemorrhage and excessive post-traumatic and postsurgical
ing vWF, FVIII, and fibrinogen, was the treatment of choice bleeding are seen and can result in excessive ecchymoses,
for patients with vWD from the early 1960s until the early hemarthroses, gastrointestinal bleeding, and intracranial
1980s.44,100 The development of products with higher purity hemorrhage.102,104 In addition, menstrual bleeding can be
that are more convenient to store, have decreased volumes of severe, and first-trimester spontaneous abortion is com-
infusion, and have undergone viral inactivation subsumed mon.102,105 Laboratory screening tests that use clot forma-
the need for cryoprecipitate. However, if other alternatives tion as an endpoint, including thrombin time, PT, and PTT,
are not available, transfusion of cryoprecipitate will pro- are markedly prolonged. Definitive diagnosis is made by
duce hemostasis.90 Each bag of cryoprecipitate contains 80 measurement of plasma fibrinogen, which is undetectable
to 100 IU of vWF, and the usual starting dose is 1 bag/10 kg by both functional and immunologic assays.
body weight every 12 to 24 hours.90,98 Cryoprecipitate from a Dysfibrinogenemia is characterized by structural fibrino-
single donor who has been pretreated with DDAVP contains gen defects causing alterations in the conversion of fibrinogen
supraphysiologic amounts of vWF and is another treatment to fibrin. Approximately 250 cases have been reported, and
option.85 the disorder is inherited as an autosomal dominant trait.106
The early plasma-derived FVIII concentrates were not Dysfibrinogenemias can be associated with either hemor-
effective in treating vWD because the vWF multimers were rhage or thromboses, or they can be asymptomatic.103,106 The
partially proteolyzed; this resulted in a loss of functional bleeding symptoms most commonly seen are ecchymoses,
high-molecular-weight multimers.44 In addition, high- epistaxis, menorrhagia, and mild to moderate postoperative 34
purity plasma-derived (monoclonal antibody-derived) or post-traumatic bleeding. Symptoms are generally mild,
FVIII as well as rFVIII cannot be used to treat vWD because although they can be more severe in patients with homozy- 477
they do not contain appreciable amounts of vWF.90 Two gous gene mutations. Results of screening tests are variable;
plasma-derived, virus-inactivated FVIII concentrates avail- the PT and PTT can be normal or prolonged. The throm-
able in the United States contain sufficient amounts of bin time is prolonged in dysfibrinogenemias associated with
functional vWF: Humate-P and Alphanate.90,100,101 The lat- hemorrhage, whereas it can be shortened or prolonged in
ter product is not yet licensed by the FDA for use in vWD. those conditions associated with thromboses. More defini-
In general, dosing for vWD is in units of ristocetin cofactor. tive diagnosis is made by measurement of fibrinogen; levels
In addition, a recombinant vWF concentrate is currently are normal or increased by immunologic assay, but they are
under development.94 reduced by functional assays.104,106,107
For both quantitative and qualitative fibrinogen defi-
Adjunctive Treatment ciencies, cryoprecipitate can be used as replacement therapy
As previously described for hemophilia treatment, antifibri- for significant episodic bleeding. Cryoprecipitate should
nolytic therapy, which inhibits lysis of newly formed clots, be administered at a dose to raise the fibrinogen level to
is also useful in the treatment of vWD. Aminocaproic acid between 50 and 100 mg/dL.102 A typical bag of cryoprecipi-
and tranexamic acid, the agents most commonly used, are tate contains 200 to 300 mg of fibrinogen in a volume of 10 to
given at the same doses used in hemophilia. Both agents 20 mL.3 Because the half-life of fibrinogen is 80 hours, dos-
may be useful alone in mild type 1 disease or as adjuncts in ing is required only every other day.104 Prophylactic replace-
the treatment of oral hemorrhage, epistaxis, gastrointestinal ment of fibrinogen, practical because of the long half-life
bleeding, and menorrhagia. These agents may increase the of fibrinogen, should be considered for patients with severe
risk of thrombosis and are contraindicated in patients with symptoms.108
a known pre-existing prothrombotic state and in the treat-
ment of genitourinary bleeding.90,98 Prothrombin Deficiency
Estrogen causes an increase in plasma vWF. This effect is
variable and is not dose related; therefore, it is not widely Prothrombin deficiency is a very rare autosomal recessive dis-
clinically applicable. However, estrogen therapy, particularly order that results from either an absolute protein deficiency
in the form of oral contraceptives, is useful in reducing the or production of an abnormal protein.109 It is characterized
severity of menorrhagia, a common problem in women with by mild symptoms, including mucocutaneous bleeding and
vWD.98 As previously discussed, patients with vWD should hemorrhage after surgery or trauma. The PT is moderately
avoid drugs that interfere with platelet function. prolonged, the PTT is normal or mildly prolonged, and the
TRANSFUSION MEDICINE
thrombin time is normal. Definitive diagnosis is made by and among different lots produced by a single manufacturer,
immunologic and functional prothrombin assays.104,109 so it is important to monitor the patient’s PT and FVII level
Therapy is usually not required. Fresh frozen plasma carefully.112 As discussed previously, patients receiving PCCs
or PCCs can be used to treat clinically significant bleeding. need to be monitored for an increased risk of developing
The half-life of prothrombin is 48 to 120 hours.103 Plasma thromboses. Recombinant FVIIa has more recently been
prothrombin levels greater than 20 to 30 U/dL are generally shown to be effective as a more specific treatment without
sufficient to stop bleeding.109 an accompanying risk of thrombosis.115 Studies have shown
efficacy of rFVIIa as treatment for intracranial hemorrhage
and as prophylaxis for surgical procedures, and these fea-
Factor V Deficiency
tures make rFVIIa the likely treatment of choice for FVII
Factor V deficiency occurs in less than 1 per 1 million per- deficiency in the future.116,117
sons and is associated with a mild to moderate bleeding
tendency.104 It is inherited as an autosomal recessive dis- Factor X Deficiency
ease. Symptoms, generally only seen in homozygotes, consist
mainly of mucocutaneous bleeding, including menorrhagia, Factor X deficiency results from either a protein deficiency
and ecchymoses. Both the prothrombin and PTT are pro- or production of a functionally abnormal protein.109 It is
longed, with a normal thrombin time. Definitive diagnosis inherited as an autosomal recessive trait, although some
is made by a specific FV assay, with deficiency defined as FV heterozygotes can have bleeding with severe trauma or
levels lower than 20 U/dL.110 major surgery. Symptoms generally correlate with the
Fresh frozen plasma should be used to treat severe bleed- severity of the deficiency and consist mainly of mucocu-
ing episodes and trauma and as prophylaxis for major sur- taneous and post-traumatic bleeding. Hemarthrosis and
gery. For treatment, the FV level should be initially raised to intracranial hemorrhage have been reported in severely
25% to 30% of normal using a 20 mL/kg dose of plasma fol- affected patients. Laboratory evaluation reveals prolonged
lowed by infusions of 6 mL/kg every 12 hours.109 FV is more PT and PTT. Diagnosis is confirmed with FX immunologic
labile in frozen plasma than other hemostatic factors, and and functional assays.
therefore fresh frozen plasma that is less than 1 to 2 months Factor X replacement should be given as needed for
old should be used.109 severe bleeding and trauma as well as before major surgi-
Several families with combined deficiency of FV and FVIII cal procedures. The half-life of transfused FX is between 24
have been reported.111 Symptoms are usually milder than and 48 hours. FX is found in both fresh frozen plasma and
in hemophilia A because of higher baseline levels of FVIII. PCCs.104 Fresh frozen plasma is given at an initial dose of
Treatment should include fresh frozen plasma in addition to 10 to 20 mL/kg, followed every 12 hours by a dose of 3 to
FVIII concentrates. 6 mL/kg to reach a target plasma FX level of 20 to 30 IU/
dL.109 PCCs contain a variable concentration of FX that can
III be measured in a specific preparation before elective use.109
Factor VII Deficiency
PCCs are more practical to use in situations such as major
478 Factor VII deficiency, resulting from either a protein defi- surgery, when sustained hemostasis is necessary and would
ciency or production of a nonfunctional protein, occurs in 1 require large volumes of plasma infusions. As previously
in 1 million persons and is inherited in an autosomal reces- mentioned, when using PCCs, the risk of thrombosis should
sive pattern with high penetrance and variable expressiv- be considered and monitored.
ity.112 Laboratory evaluation of patients with FVII deficiency
reveals a prolonged PT with a normal PTT and thrombin Factor XI Deficiency
time, a pattern that distinguishes FVII deficiency from other
inherited clotting disorders. The diagnosis is confirmed by Factor XI deficiency, sometimes called hemophilia C, is a
measuring FVII activity using a tissue factor-dependent relatively rare disorder distinguished by a variable bleeding
one-stage technique. The clinical expression of the disease tendency in affected individuals. It is inherited as an autosomal
is generally correlated with the degree of factor deficiency, trait and is seen most commonly among persons of Ashkenazi
although reports exist of patients with extremely low FVII Jewish descent, with an 8% frequency of heterozygosity.118
levels who present without a history of bleeding.113 Patients Three different point mutations in the FXI gene account for
with FVII levels greater than 10% to 15% rarely have signifi- nearly all cases of FXI deficiency among Ashkenazi Jews.119
cant bleeding problems, whereas patients with levels lower FXI deficiency is most often detected in association with a
than 1% can have severe bleeding symptoms.112 The most positive family history or secondary to a prolonged PTT
serious complication of FVII deficiency is intracranial hem- found during a routine presurgical evaluation. The diagnosis
orrhage, which occurs most frequently during the first year is confirmed with a plasma FXI assay.
of life, particularly in the first postnatal week.114 Mucous Factor XI levels correlate with the genotype. Severe defi-
membrane bleeding, including excessive bruising, epistaxis, ciency (FXI levels lower than 15 to 20 IU/dL) is seen in
gastrointestinal bleeding, and menorrhagia, is the most homozygotes or compound heterozygotes, and partial defi-
common manifestation of FVII deficiency. Hemarthroses, in ciency (FXI levels from 20 to 60 IU/dL) is seen in heterozy-
a pattern similar to what is seen in hemophilia, also occur gotes. Bleeding tendency, however, does not correlate with
occasionally.114 genotype.120 Excessive bleeding is most commonly seen in
Treatment is required for severe bleeding episodes and association with severe deficiency, but it can also be seen
before surgical procedures. PCCs have previously been the in patients with partial deficiency. A compilation of studies
mainstays of therapy for FVII deficiency. Replacement doses shows abnormal bleeding in 30% to 50% of patients with
are calculated to raise the FVII level to more than 15%. FVII partial deficiency.121 In general, bleeding symptoms are
concentrations vary among commercially available PCCs mild or absent, only occurring in association with trauma
TRANSFUSION OF THE PATIENT WITH CONGENITAL COAGULATION DEFECTS
or surgery, particularly tonsillectomy. Epistaxis, soft tissue CONCLUSION
hemorrhage, bleeding after dental extractions, and menor-
rhagia can also be seen in affected persons. Accurate and early diagnosis, comprehensive clinical care,
Fresh frozen plasma is most commonly given as treat- and the availability of factor concentrates that are manu-
ment for severe bleeding episodes or as prophylaxis for factured with limited risk of transmitting blood-borne
major surgery. The half-life of FXI in plasma is approxi- viruses and minimal risk of thrombosis have all contributed
mately 45 hours. As an alternative treatment, two specific to improved outcomes for patients with congenital bleed-
FXI concentrates are available in Europe.118 Both have been ing disorders. Despite these advances, patients with severe
shown to be hemostatically effective and virally safe, but hemophilia require tremendous dedication and support to
they have been associated with an increased occurrence of successfully manage their disease. The promise of gene trans-
thrombotic events, particularly in older patients with pre- fer, which could allow for amelioration or elimination of the
existing vascular disease and when these concentrates are phenotypic abnormalities seen in severe hemophilia, is still
given at very high doses.122,123 These products may be useful in the future.
in selected patients without pre-existing hypercoagulable
states at doses less than 30 IU/kg.118 Bleeding from minor
procedures such as dental extraction can be controlled with
local measures or antifibrinolytic agents (aminocaproic REFERENCES
acid and tranexamic acid) at the same doses that are used 1. Soucie JM, Evatt B, Jackson D. Occurrence of hemophilia in the United
for hemophilia and vWD.118 States: The Hemophilia Surveillance System Project Investigators. Am
J Hematol 1998;59:288–294.
2. DiMichele D. Hemophilia 1996. New approach to an old disease. Pedi-
Contact Factor Deficiencies atr Clin North Am 1996;43:709–736.
3. Kasper CK. Hereditary plasma clotting factor disorders and their man-
Deficiencies of the contact factors—FXII, prekallikrein, and agement. Haemophilia 2000;6:13–27.
high-molecular-weight kininogen—are not associated with 4. Goodeve AC. Advances in carrier detection in haemophilia. Haemo-
bleeding symptoms.104 Because these factors are necessary for philia 1998;4:358–364.
5. Lillicrap D. Molecular diagnosis of inherited bleeding disorders and
the initiation of the intrinsic pathway, the PTT is markedly thrombophilia. Semin Hematol 1999;36:340–351.
prolonged, frequently beyond what is seen in hemophilia, 6. Bagnall RD, Waseem N, Green PM, Giannelli F. Recurrent inversion
whereas the PT is normal. Both FXII and prekallikrein defi- breaking intron 1 of the factor VIII gene is a frequent cause of severe
ciencies are inherited as autosomal recessive traits.104 Patients hemophilia A. Blood 2002;99168–174.
7. Tedgard U. Carrier testing and prenatal diagnosis of haemophilia: utili-
are usually identified during routine preoperative screening, sation and psychological consequences. Haemophilia 1998;4:365–369.
and definitive diagnosis is made with specific contact factor 8. Andrew M, Brooker LA. Blood component therapy in neonatal hemo-
assays. Treatment is not necessary. static disorders. Transfus Med Rev 1995;9:231–250.
9. DiMichele D, Neufeld EJ. Hemophilia: a new approach to an old dis-
ease. Hematol Oncol Clin North Am 1998;12:1315–1344. 34
Factor XIII Deficiency 10. Hoyer LW. Hemophilia A. NEJM 1994;330:38–47.
11. Hoag MS, Johnson FF, Robinson JA, Aggeler PM. Treatment of hemo-
Factor XIII deficiency is an autosomal recessive disorder. philia B with a new clotting-factor concentrate. NEJM 1969;280:581–586. 479
FXIII is also known as fibrin-stabilizing factor and is respon- 12. Mannucci PM. Modern treatment of hemophilia: from the shadows
sible for crosslinking of the fibrin polymer. Deficiency of towards the light. Thromb Haemost 1993;70:17–23.
13. Centers for Disease Control and Prevention. Update on acquired
FXIII is associated with reduced clot stability, and therefore immune deficiency syndrome (AIDS) among patients with hemo-
ecchymoses or hematomas are usually seen 24 to 36 hours philia A. MMWR Morb Mortal Wkly Rep 1982;31:644–646, 652,.
after trauma. FXIII deficiency is associated with severe 14. Berntorp E. Impact of replacement therapy on the evolution of HIV
bleeding, particularly delayed intracranial hemorrhage infection in hemophiliacs. Thromb Haemost 1994;71:678–683.
with minimal trauma. Other symptoms of FXIII deficiency 15. Brettler DB. Comments on the development of inhibitor antibodies in
patients using recombinant factor VIII concentrates. Semin Hematol
include delayed separation of the umbilical stump in the 1991;28:45–46.
neonatal period (at more than 3.5 to 4 weeks), poor wound 16. Kasper CK, Lusher JM. Recent evolution of clotting factor concentrates
healing, and recurrent spontaneous abortions in women for hemophilia A and B: Transfusion Practices Committee. Transfu-
secondary to severe decidual bleeding. Screening tests are sion 1993;33:422–434.
17. Lusher JM. Transfusion therapy in congenital coagulopathies. Hema-
normal, including bleeding time, PT, PTT, and thrombin tol Oncol Clin North Am 1994;8:1167–1180.
time. Clots from affected persons appear more friable than 18. Mannucci PM. Clinical evaluation of viral safety of coagulation factor
normal. When the condition is suspected clinically, diagno- VIII and IX concentrates. Vox Sang 1993;64:197–203.
sis of homozygous patients is made using a urea clot solu- 19. Horowitz MS, Rooks C, Horowitz B, Hilgartner MW. Virus safety of
bility assay or by measuring plasma levels of FXIII subunit solvent/detergent-treated antihaemophilic factor concentrate. Lancet
1988;2:186–189.
proteins.124 20. Ludlam CA. Viral safety of plasma-derived factor VIII and IX concen-
Fresh frozen plasma is usually used for hemostasis in FXIII trates. Blood Coagul Fibrinolysis 1997;8(Suppl):S19–S23.
deficiency.109 It is given at a dose of 5 to 10 mL/kg. Because the 21. Toole JJ, Knopf JL, Wozney JM, et al. Molecular cloning of a cDNA
half-life of FXIII is very long (3 to 6 days), one dose is gener- encoding human antihaemophilic factor. Nature 1984;312:342–347.
22. Wood WI, Capon DJ, Simonsen CC, et al. Expression of active human
ally sufficient. Cryoprecipitate also contains FXIII and can factor VIII from recombinant DNA clones. Nature 1984;312:330–337.
also be used to treat acute bleeding episodes.109 Patients with 23. Schwartz RS, Abildgaard CF, Aledort LM, et al. Human recombi-
severe bleeding symptoms can receive either product as pro- nant DNA-derived antihemophilic factor (factor VIII) in the treat-
phylaxis every 2 to 4 weeks. In addition, two plasma-derived ment of hemophilia A: recombinant Factor VIII Study Group. NEJM
FXIII concentrates are currently produced (by Centeon in 1990;323:1800–1805.
24. Lusher JM, Arkin S, Abildgaard CF, Schwartz RS. Recombinant factor
the United States and Bio Products Laboratory in the UK).125 VIII for the treatment of previously untreated patients with hemophilia
These are not yet available for general distribution in the A: safety, efficacy, and development of inhibitors. Kogenate Previously
United States. Untreated Patient Study Group. NEJM 1993;328: 453–459.
TRANSFUSION MEDICINE 25. Bray GL, Gomperts ED, Courter S, et al. A multicenter study of recom- acs: The Orthopaedic Outcome Study Group. J Intern Med 1994;236:
binant factor VIII (Recombinate): safety, efficacy, and inhibitor risk 391–399.
in previously untreated patients with hemophilia A. The Recombinate 55. Ljung RC. Prophylactic infusion regimens in the management of
Study Group. Blood 1994;83:2428–2435. hemophilia. Thromb Haemost 1999;82:525–530.
26. Manno CS. Treatment options for bleeding episodes in patients under- 56. Manco-Johnson MJ, Nuss R, Geraghty S, et al. Results of second-
going immune tolerance therapy. Haemophilia 1999;5:33–41. ary prophylaxis in children with severe hemophilia. Am J Hematol
27. Mannucci PM, Tuddenbam EG. The hemophilias: progress and prob- 1994;47:113–117.
lems. Semin Hematol 1999;36:104–117. 57. Aledort L. Inhibitors in hemophilia patients: current status and man-
28. Hilgartner MW, Manno CS, Nuss R, Di Michele DM. Pediatric issues agement. Am J Hematol 1994;47:208–217.
in hemophilia. In Proceedings of the American Society of Hematology. 58. Manco-Johnson MJ, Nuss R, Geraghty S, Funk S. A prophylactic pro-
San Diego, American Society of Hematology, 1997, pp 46–60. gram in the United States: experience and issues. Semin Hematol
29. Abshire TC, Brackmann HH, Scharrer I, et al. Sucrose formulated 1994;31:10–12.
recombinant human antihemophilic factor VIII is safe and efficacious 59. Smith PS, Teutsch SM, Shaffer PA, et al. Episodic versus prophylac-
for treatment of hemophilia A in home therapy: International Koge- tic infusions for hemophilia A: a cost-effectiveness analysis. J Pediatr
nate-FS Study Group. Thromb Haemost 2000;83:811–816. 1996;129:424–431.
30. Berntorp E. Second generation, B-domain deleted recombinant factor 60. Carlsson M, Berntorp E, Bjorkman S, Lindvall K. Pharmacokinetic
VIII. Thromb Haemost 1997;78:256–260. dosing in prophylactic treatment of hemophilia A. Eur J Haematol
31. Sandberg H, Almstedt A, Brandt J, et al. Structural and functional 1993;51:247–252.
characteristics of the B-domain-deleted recombinant factor VIII pro- 61. Blanchette VS, Al-Musa A, Stain AM, et al. Central venous access
tein, r-VIII SQ. Thromb Haemost 2001;85:93–100. devices in children with hemophilia: an update. Blood Coagul Fibri-
32. Tarantino MD, Collins PW, Hay CR, et al. Clinical evaluation of an nolysis 1997;8(Suppl):S11–S14.
advanced category antihaemophilic factor prepared using a plasma/ 62. Bollard CM, Teague LR, Berry EW, Ockelford PA. The use of central
albumin-free method: pharmacokinetics, efficacy, and safety in pre- venous catheters (Portacaths) in children with haemophilia. Haemo-
viously treated patients with haemophilia A. Haemophilia 2004;10: philia 2000;6:66–70.
428–437. 63. Ragni MJ, Hord JD, Blatt J. Central venous catheter infection in hae-
33. Kohler M, Hellstern P, Lechler E, et al. Thromboembolic complications mophiliacs undergoing prophylaxis or immune tolerance with clotting
associated with the use of prothrombin complex and factor IX concen- factor concentrate. Haemophilia 1997;3:90–95.
trates. Thromb Haemost 1998;80:399–402. 64. Conway JH, Hilgartner MW. Initial presentations of pediatric hemo-
34. Goldsmith JC, Kasper CK, Blatt PM, et al. Coagulation factor IX: Suc- philiacs. Arch Pediatr Adolesc Med 1994;148:589–594.
cessful surgical experience with a purified factor IX concentrate. Am J 65. Baehner RL, Strauss HS. Hemophilia in the first year of life. NEJM
Hematol 1992;40:210–215. 1966;275:524–528.
35. Shapiro AD, Ragni MV, Lusher JM, et al. Safety and efficacy of mono- 66. Buchanan GR. Factor concentrate prophylaxis for neonates with
clonal antibody purified factor IX concentrate in previously untreated hemophilia. J Pediatr Hematol Oncol 1999;21:254–256.
patients with hemophilia B. Thromb Haemost 1996;75:30–35. 67. Kulkarni R, Lusher JM, Henry RC, Kallen DJ. Current practices regard-
36. Bond M, Jankowski M, Patel H, et al. Biochemical characterization of ing newborn intracranial haemorrhage and obstetrical care and mode
recombinant factor IX. Semin Hematol 1998;35:11–17. of delivery of pregnant haemophilia carriers: a survey of obstetricians,
37. Brinkhous KM, Sigman JL, Read MS, et al. Recombinant human factor neonatologists and haematologists in the United States, on behalf of
IX: replacement therapy, prophylaxis, and pharmacokinetics in canine the National Hemophilia Foundation’s Medical and Scientific Advi-
hemophilia B. Blood 1996;88:2603–2610. sory Council. Haemophilia 1999;5:410–415.
38. Choo KH, Gould KG, Rees DJ, et al. Molecular cloning of the gene for 68. Hoyer LW. Why do so many haemophilia A patients develop an inhibi-
human anti-haemophilic factor IX. Nature 1982;299:178–180. tor? Br J Haematol 1995;90:498–501.
III 39. Kurachi K, Davie EW. Isolation and characterization of a cDNA coding 69. Hoyer LW. Factor VIII inhibitors. Curr Opin Hematol 1995;2:365–371.
for human factor IX. Proc Natl Acad Sci USA 1982;79:6461–6464. 70. Hay CR. Why do inhibitors arise in patients with haemophilia A? Br J
480 40. White G, Shapiro A, Ragni M, et al. Clinical evaluation of recombinant Haematol 1999;105:584–590.
factor IX. Semin Hematol 1998;35:33–38. 71. Tuddenham EG, McVey JH. The genetic basis of inhibitor develop-
41. White GC 2nd, Beebe A, Nielsen B. Recombinant factor IX. Thromb ment in haemophilia A. Haemophilia 1998;4:543–545.
Haemost 1997;78:261–265. 72. Kasper CK, Aledort L, Aronson D, et al. Proceedings: A more uniform
42. Furie B, Limentani SA, Rosenfield CG. A practical guide to the evalua- measurement of factor VIII inhibitors. Thromb Diath Haemorrh
tion and treatment of hemophilia. Blood 1994;84:3–9. 1975;34:612.
43. Mannucci PM. Desmopressin (DDAVP) in the treatment of bleeding 73. Leissinger CA. Use of prothrombin complex concentrates and acti-
disorders: the first 20 years. Blood 1997;90:2515–2521. vated prothrombin complex concentrates as prophylactic therapy in
44. Logan LJ. Treatment of von Willebrand’s disease. Hematol Oncol Clin haemophilia patients with inhibitors. Haemophilia 1999;5:25–32.
North Am 1992;6:1079–1094. 74. Green D. Complications associated with the treatment of haemophili-
45. Manno CS, Butler RB, Cohen AR. Low recovery in vivo of highly purified acs with inhibitors. Haemophilia 1999;5:11–17.
factor VIII in patients with hemophilia. J Pediatr 1992;121:814–818. 75. Santagostino E, Gringeri A, Mannucci PM. Home treatment with
46. Kelly KM, Butler RB, Farace L, et al. Superior in vivo response of recombinant activated factor VII in patients with factor VIII inhibitors:
recombinant factor VIII concentrate in children with hemophilia A. J the advantages of early intervention. Br J Haematol 1999;104:22–26.
Pediatr 1997;130:537–540. 76. Teitel JM. Recombinant factor VIIa versus aPCCs in haemophiliacs
47. Lusher JM. Prophylaxis in children with hemophilia: is it the optimal with inhibitors: treatment and cost considerations. Haemophilia 1999;
treatment? Thromb Haemost 1997;78:726–729. 5:43–49.
48. Ahlberg A. Haemophilia in Sweden. VII. Incidence, treatment and pro- 77. DiMichele DM. Immune tolerance: a synopsis of the international
phylaxis of arthropathy and other musculo-skeletal manifestations of experience. Haemophilia 1998;4:568–573.
haemophilia A and B. Acta Orthop Scand Suppl 1965;3:132. 78. Conversation with Dr. CRM Hay and Dr. DM DiMichele on December
49. Aronstam A, Kirk PJ, McHardy J, et al. Twice weekly prophylactic ther- 21, 2005. Untitled abstract: “International Randomized Immune Tol-
apy in haemophilia A. J Clin Pathol 1977;30:65–67. erance (ITI) Study: Progress Report.”
50. Gruppo RA. Prophylaxis for hemophilia: state of the art or state of 79. Kay MA, High K. Gene therapy for the hemophilias. Proc Natl Acad Sci
confusion [editorial]? J Pediatr 1998;132:915–917. USA 1999;96:9973–9975.
51. Nilsson IM, Berntorp E, Lofqvist T, Pettersson H. Twenty-five years’ 80. Herzog RW, High KA. Adeno-associated virus-mediated gene transfer
experience of prophylactic treatment in severe haemophilia A and B. J of factor IX for treatment of hemophilia B by gene therapy. Thromb
Intern Med 1992;232:25–32. Haemost 1999;82:540–546.
52. Liesner RJ, Khair K, Hann IM. The impact of prophylactic treatment 81. Kay MA, Manno CS, Ragni MV, et al. Evidence for gene transfer and
on children with severe haemophilia. Br J Haematol 1996;92:973–978. expression of factor IX in haemophilia B patients treated with an AAV
53. Lofqvist T, Nilsson IM, Berntorp E, Pettersson H. Haemophilia pro- vector. Nat Genet 2000;24:257–261.
phylaxis in young patients: a long-term follow-up. J Intern Med 82. Rodeghiero F, Castaman G, Dini E. Epidemiological investigation of
1997;241:395–400. the prevalence of von Willebrand’s disease. Blood 1987;69:454–459.
54. Aledort LM, Haschmeyer RH, Pettersson H. A longitudinal study of 83. Werner EJ, Broxson EH, Tucker EL, et al. Prevalence of von Willebrand
orthopaedic outcomes for severe factor-VIII-deficient haemophili- disease in children: a multiethnic study. J Pediatr 1993;123:893–898.
TRANSFUSION OF THE PATIENT WITH CONGENITAL COAGULATION DEFECTS
84. Ginsburg D. Molecular genetics of von Willebrand disease. Thromb 105. Kobayashi T, Kanayama N, Tokunaga N, et al. Prenatal and peripartum
Haemost 1999;82:585–591. management of congenital afibrinogenaemia. Br J Haematol 2000;
85. Werner EJ. von Willebrand disease in children and adolescents. Pediatr 109:364–366.
Clin North Am 1996;43:683–707. 106. Haverkate F, Samama M. Familial dysfibrinogenemia and thrombo-
86. Batlle J, Torea J, Rendal E, Fernandez MF: The problem of diagnosing philia: report on a study of the SSC Subcommittee on Fibrinogen.
von Willebrand’s disease. J Intern Med Suppl 1997;740:121–128. Thromb Haemost 1995;73:151–161.
87. Mancuso DJ, Tuley EA, Westfield LA, et al. Structure of the gene for 107. Martinez J. Congenital dysfibrinogenemia. Curr Opin Hematol 1997;
human von Willebrand factor. J Biol Chem 1989;264:19514–19527. 4:357–365.
88. Mancuso DJ, Tuley EA, Westfield LA, et al. Human von Willebrand 108. Rodriguez RC, Buchanan GR, Clanton MS. Prophylactic cryoprecipi-
factor gene and pseudogene: structural analysis and differentiation by tate in congenital afibrinogenemia. Clin Pediatr 1988;27:543–545.
polymerase chain reaction. Biochemistry 1991;30:253–269. 109. Bauer KA. Rare hereditary coagulation factor abnormalities. In
89. Schneppenheim R, Thomas KB, Sutor AH. Von Willebrand disease in Nathan DG, Orkin SH (eds). Nathan and Oski’s Hematology of
childhood. Semin Thromb Hemost 1995;21:261–275. Infancy and Childhood, vol 2, 5th ed. Philadelphia, WB Saunders,
90. Phillips MD, Santhouse A. von Willebrand disease: Recent advances in 1998, pp 1660–1675.
pathophysiology and treatment. Am J Med Sci 1998;316:77–86. 110. Mammen EF. Factor V deficiency. Semin Thromb Hemost 1983;9:
91. Sadler JE. A revised classification of von Willebrand disease: for the 17–18.
Subcommittee on von Willebrand Factor of the Scientific and Stan- 111. Seligsohn U, Ramot B. Combined factor-V and factor-VIII deficiency:
dardization Committee of the International Society on Thrombosis report of four cases. Br J Haematol 1969;16:475–486.
and Haemostasis. Thromb Haemost 1994;71:520–525. 112. Ingerslev J, Kristensen HL. Clinical picture and treatment strategies in
92. Sadler JE, Matsushita T, Dong Z, et al. Molecular mechanism and factor VII deficiency. Haemophilia 1998;4:689–696.
classification of von Willebrand disease. Thromb Haemost 1995;74: 113. Triplett DA, Brandt JT, Batard MA, et al. Hereditary factor VII defi-
161–166. ciency: heterogeneity defined by combined functional and immuno-
93. Federici AB. Diagnosis of von Willebrand disease. Haemophilia 1998;4: chemical analysis. Blood 1985;66:1284–1287.
654–660. 114. Ragni MV, Lewis JH, Spero JA, Hasiba U. Factor VII deficiency. Am J
94. Schwarz HP, Turecek PL, Pichler L, et al. Recombinant von Willebrand Hematol 1981;10:79–88.
factor. Thromb Haemost 1997;78:571–576. 115. Bauer KA. Treatment of factor VII deficiency with recombinant factor
95. Werner EJ, Abshire TC, Giroux DS, et al. Relative value of diagnostic VIIa. Haemostasis 1996;26:155–158.
studies for von Willebrand disease. J Pediatr 1992;121:34–38. 116. Mariani G, Testa MG, Di Paolantonio T, et al. Use of recombinant, acti-
96. Gill JC, Endres-Brooks J, Bauer PJ, et al. The effect of ABO blood group vated factor VII in the treatment of congenital factor VII deficiencies.
on the diagnosis of von Willebrand disease. Blood 1987;69:1691–1695. Vox Sang 1999;77:131–136.
97. Mannucci PM. Treatment of von Willebrand’s disease. J Intern Med 117. Wong WY, Huang WC, Miller R, et al. Clinical efficacy and recovery
Suppl 1997;740:129–132. levels of recombinant FVIIa (NovoSeven) in the treatment of intra-
98. Mannucci PM. Treatment of von Willebrand disease. Int J Clin Lab Res cranial haemorrhage in severe neonatal FVII deficiency. Haemophilia
1998;28:211–214. 2000;6:50–54.
99. Mannucci PM. Desmopressin: a nontransfusional form of treat- 118. Bolton-Maggs PH. The management of factor XI deficiency. Haemo-
ment for congenital and acquired bleeding disorders. Blood 1988;72: philia 1998;4:683–688.
1449–1455. 119. Asakai R, Chung DW, Ratnoff OD, Davie EW. Factor XI (plasma
100. Menache D, Aronson DL. New treatments of von Willebrand disease: thromboplastin antecedent) deficiency in Ashkenazi Jews is a bleeding
plasma derived von Willebrand factor concentrates. Thromb Haemost disorder that can result from three types of point mutations. Proc Natl
1997;78:566–570. Acad Sci USA 1989;86:7667–7671.
101. Lubetsky A, Schulman S, Varon D, et al. Safety and efficacy of continu- 120. Asakai R, Chung DW, Davie EW, Seligsohn U. Factor XI deficiency in
ous infusion of a combined factor VIII–von Willebrand factor (vWF) Ashkenazi Jews in Israel. NEJM 1991;325:153–158.
34
concentrate (Haemate-P) in patients with von Willebrand disease. 121. Bolton-Maggs PH, Young Wan-Yin B, McCraw AH, et al. Inheritance
Thromb Haemost 1999;81:229–233. and bleeding in factor XI deficiency. Br J Haematol 1988;69:521–528. 481
102. al-Mondhiry H, Ehmann WC. Congenital afibrinogenemia. Am 122. Bolton-Maggs PH, Colvin BT, Satchi BT, et al. Thrombogenic potential
J Hematol 1994;46:343–347. of factor XI concentrate. Lancet 1994;344:748–749.
103. Hilgartner M, Corrigan JJ. Coagulation disorders. In Miller DR, Baeh- 123. Mannucci PM, Bauer KA, Santagostino E, et al. Activation of the coag-
ner RL (eds). Blood Diseases of Infancy and Childhood, 7th ed. Phila- ulation cascade after infusion of a factor XI concentrate in congenitally
delphia, Mosby, 1995, pp 924–986. deficient patients. Blood 1994;84:1314–1319.
104. Blanchette VS, Dean J, Lillicrap D. Rare congenital hemorrhagic 124. Anwar R, Miloszewski KJ. Factor XIII deficiency. Br J Haematol
disorders. In Lilleyman JS, Hann IM, Blanchette VS (eds). Pediat- 1999;107:468–484.
ric Hematology, 2nd ed. Philadelphia, Churchill Livingstone, 1999, 125. Gootenberg JE. Factor concentrates for the treatment of factor XIII
pp 611–628. deficiency. Curr Opin Hematol 1998;5:372–375.
Chapter 35
Transfusion of the Patient with
Acquired Coagulation Defects
Barbara Alving*
Patients who develop bleeding disorders because of under- The complete blood count, along with evaluation of the
lying systemic disease, autoantibody formation, or as a peripheral smear, may reinforce the possibility of sepsis (e.g.,
result of complications of anticoagulant use can be effec- toxic granulations and Döhle bodies in the neutrophils) and
tively treated with strategies that may include, in addition also help to determine whether or not an apparent thrombo-
to blood products, one or more hemostatic agents as well cytopenia is real or due to platelet clumping in ethylenedi-
as newer immunosuppressive agents. This chapter discusses aminetetraacetic acid (EDTA).
the diagnosis and treatment of the following acquired coag-
ulopathies: those related to underlying systemic disorders Hemostatic Support, Including
such as liver disease and renal failure; disseminated intra- Blood Products
vascular coagulation (DIC); formation of autoantibodies to
coagulation factors VIII, V, II, or XIII; and coagulopathies In the surgical setting, a realistic goal is to provide sufficient
associated with anticoagulants and the newer antiplatelet hemostatic support to allow the patient to undergo surgery
agents. or other invasive procedures without undue risk. Thus, com-
plete correction of coagulation laboratory abnormalities
may not be possible or even desirable. For example, replace-
APPROACH TO THE PATIENT ment of factor VII, which has a half-life of 2 to 6 hours, by
WITH A COAGULOPATHY infusion of plasma could lead to excessive volume expansion
III and pulmonary edema.
Patient Evaluation A critical factor for patients who require invasive proce-
482 dures such as line placement is the skill of the operator. In
The evaluation of the patient with an apparent or possible one retrospective review of patients who had received arte-
coagulopathy includes a review of the past and present illness rial, pulmonary artery, or central venous lines, the authors
as well as use of medications and a physical examination. found that hemostatic complications were rare and were
Patients who have developed autoantibodies to a clotting related more to the experience of the operator than to the
factor such as factor VIII may have no complaints other than underlying hemostatic defect.1 They concluded that only
easy bruising or unexpected excessive bleeding with minor patients with severe hemostatic defects require correction of
trauma. For these patients, laboratory data may be the first the abnormalities before line placement.
indication of the etiology of a bleeding disorder. For major surgery in patients with coagulopathies,
Screening tests are an integral part of the initial evalu- hemostasis is best managed if there is close communication
ation (Table 35–1). If the activated partial thromboplastin among the consultants, the surgeons, the personnel of the
time (aPTT) and prothrombin time (PT) are prolonged, coagulation laboratory, and the staff of the blood bank (or
then performing mixing studies to assess for the presence pharmacy) who are responsible for supplying the products.2
of an inhibitor and measuring coagulation factor levels are Examples of two newer products that are being used to pro-
appropriate. If DIC is suspected, the levels of d-dimer, a late vide hemostasis in a wide variety of clinical settings are fibrin
plasmin digest of crosslinked fibrin, should also be measured. sealant and recombinant factor VIIa (rFVIIa).
Increased levels indicate that thrombin has been generated Products, such as fibrin sealant, that provide localized
in sufficient quantities to activate factor XIII, which induces hemostasis can be used as adjuvants for systemic therapy
crosslinking of fibrin, and that plasmin has also been gen- in patients with coagulopathies. In the United States, one
erated to degrade the fibrin. This process is consistent with commercial product is now available (Tisseel). It is indi-
the generalized activation of the coagulation and fibrinolytic cated for patients who are undergoing reoperative coro-
system that occurs in DIC. nary artery bypass surgery or who have bleeding from sites,
such as the spleen, where bleeding cannot be controlled by
sutures. The product is composed of human fibrinogen
(75 to 115 mg/mL) and human thrombin (500 IU/mL), both
*This chapter was written by Dr. Alving in her private capacity. The of which are virally inactivated.3 The thrombin is solubi-
views expressed in this article do not necessarily represent the views of the
National Institutes of Health, Department of Health and Human Services, lized in calcium chloride (40 mM), which stimulates cross-
or the United States. All material in this chapter is in the public domain, linking of the clot by factor XIII in the product (or patient).
with the exception of any borrowed figures or tables. The product also contains aprotinin, a bovine-derived
TRANSFUSION OF THE PATIENT WITH ACQUIRED COAGULATION DEFECTS
Table 35–1 Laboratory Testing for Patients with an Acquired Coagulopathy
Perform complete blood count, aPTT, PT, and thrombin time; measurement of fibrinogen concentration; and test for D-dimer
as appropriate.
If thrombin time is prolonged, evaluate for heparin (perform reptilase time, add heparinase to sample). If heparin is present,
check to be sure that the patient is not excessively anticoagulated or has developed a heparin-like inhibitor.
If aPTT is prolonged, perform “correction” studies. Mix equal volumes of patient and normal plasma and repeat the aPTT
immediately and after 1 hour’s incubation at 37° C. The values for the mixing study should be within 5 seconds of the control
value of the normal plasma aPTT.
An aPTT that becomes more prolonged with incubation is suggestive of an inhibitor to a specific factor. An aPTT that is initially
prolonged and remains prolonged to the same degree at 1 hour is suggestive of a lupus anticoagulant.
Measure factor levels to be sure that a true deficiency of a coagulation factor is recognized, even if the diagnosis is “lupus
anticoagulant.”
Consider the presence of an inhibitor to factor XIII if bleeding persists and the screening studies are normal. This requires a special
test for the solubility of the clot in urea and monochloroacetic acid.
protein that directly inhibits plasmin, which is included to Fibrin sealants produced by several manufacturers (in
increase clot stability. both the United States and Europe) are undergoing or will
A much less expensive form of this product can be be undergoing clinical trials in the United States. Differences
made by mixing cryoprecipitate in one syringe (fibrino- may include methods of viral inactivation and presence or
gen concentration, 10 to 15 mg/mL) and bovine thrombin absence of a fibrinolytic inhibitor in the product. In addi-
(1000 IU/mL) in the second syringe. However, this prepara- tion, companies are developing other forms of sealants or
tion contains a lower concentration of fibrinogen and has adhesives composed of collagen and thrombin and devices
not undergone viral inactivation. Furthermore, the bovine that allow production of autologous fibrin sealant in the
thrombin may be contaminated with bovine factor V, which operating room.
can stimulate production of antibodies in the recipient that Recombinant activated factor VII (rFVIIa, NovoSeven) is
crossreact with human factor V, inducing a potentially severe approved by the U.S. Food and Drug Administration (FDA)
bleeding disorder several days after surgery.4–6 Currently, the for use in individuals with hemophilia (factor VIII or IX
bovine thrombin preparation with the lowest degree of con- deficiency) who have inhibitors and are actively bleeding or
tamination with factor V is that produced for Jones Medical at high risk to bleed. During the past few years it has gained
Industries, St. Louis.7 widespread use as a “universal hemostatic agent,” although
Efficacy data for the commercial fibrin sealant were clinical trials to define the dose and efficacy in multiple situ-
derived in part from a study in which 333 patients from 11 ations have not yet been performed. 35
centers in the United States who were undergoing reoperative The “off-label” applications that are reported in the litera-
cardiac surgery or emergency resternotomy were randomly ture include treatment of bleeding episodes and/or preven- 483
assigned to receive fibrin sealant or conventional hemostatic tion of bleeding related to surgery in patients with platelet
agents; the end point for efficacy was the number of bleed- disorders (qualitative and/or quantitative), liver disease (cir-
ing episodes controlled at 5 minutes.8 The success rate for rhosis, transplantation), surgery and trauma, and reversal of
fibrin sealant was 92.6% compared with 12.4% for conven- warfarin therapy.12–16 For patients with hemophilia who have
tional topical agents (p < 0.001). Fibrin sealant also rapidly inhibitors, the dose of rFVIIa is usually 90 μg/kg given every
controlled bleeding episodes that did not initially respond to 2 hours (half-life, approximately 3 hours). The best doses to
conventional hemostatic agents in 82% of patients. be used for continuous infusion are still being determined.
In addition to use in cardiac surgery and trauma, fibrin For patients with factor VII deficiency, the most effective
sealants have been used as adhesives to seal dural leaks in doses are 20 to 25 μg/kg. An international registry is being
neurosurgery, to promote union of middle ear bones in used to record experience on the doses used in patients with
otolaryngology, and as a matrix to repair bone defects.3 platelet disorders and hepatic failure as well as orthotopic
Although randomized, blinded studies have not been done, liver transplantation.
fibrin sealant has been applied to sites of dental extractions In trauma and surgery, antecodotal reports suggest that
in individuals with hemophilia; this treatment, combined in patients with excessive uncontrolled bleeding, one or two
with the use of an antifibrinolytic agent, has reduced or infusions of rFVIIa at doses ranging from 20 to 120 μg/kg
eliminated the need for systemic factor VIII replacement.9,10 can have a significant hemostatic effect.12–16 The incidence of
This combined therapy would potentially be efficacious in thrombosis has been estimated to be 1% to 2%.14
patients with coagulation factor inhibitors as well as those
with severe thrombocytopenia who require dental extrac-
tion. Outside the operating room, these sealants may also be
applicable to providing local hemostasis in the critical care COAGULOPATHIES IN SYSTEMIC
setting in patients with coagulopathies who have a local- DISORDERS
ized site of bleeding. In a randomized, prospective evalua-
tion of fibrin sealant in neonates undergoing extracorporeal Liver Disease
membrane oxygenation, application of fibrin sealant at the
Pathophysiology and Laboratory Evaluation
cannulation sites reduced the risk for any bleeding and was
associated with a shorter duration of hemorrhage with less Patients with liver disease may have a complex coagulopa-
blood loss.11 thy consisting of impaired coagulation factor synthesis,
TRANSFUSION MEDICINE
increased fibrinolytic activity, and thrombocytopenia.17 In fibrinogen level, and d-dimer. Although measurement of
patients with splenomegaly related to cirrhosis and portal t-PA or the euglobulin clot lysis time is desirable, these tests
hypertension, 90% of the circulating platelets can be seques- are usually not available in a hospital setting. The PT is also
tered in the spleen, with platelet counts decreasing to as low reported as an International Normalized Ratio (INR), which
as 30,000 to 40,000/μL. The thrombocytopenia may be due was originally developed to standardize anticoagulation
in part to decreased platelet production because of reduced among patients taking warfarin. A study of 29 patients with
levels of circulating thrombopoietin, which is synthesized liver impairment (INR, 1.5 to 3.5) was conducted with three
in the liver.18 In one study of 44 patients with cirrhosis and different thromboplastin reagents (International Sensitivity
thrombocytopenia, thrombopoietin levels were undetect- Index, 0.86 to 2.53).24 The mean INR for the patient plas-
able in 89%; however, after undergoing liver transplantation, mas as determined with the three different reagents ranged
94% had detectable levels of thrombopoietin and all had from 1.88 to 2.63, depending on the reagent. In contrast, the
resolution of the thrombocytopenia.18 In patients undergo- three reagents gave the same INR in plasmas from patients
ing portal decompression or splenectomy, thrombocytope- taking warfarin. Thus, INR values are not well standardized
nia often persists, indicating that a defect in thrombopoietin for patients with liver disease and can vary depending on the
production is present in addition to splenic sequestration. reagent. However, this is also true for the measurement of the
A further reduction in platelet count is usually due to addi- PT with the different reagents. Patients with liver disease may
tional factors such as coexisting immune thrombocytope- have only a prolonged PT (normal aPTT and thrombin time),
nia.17 Patients may also have platelet dysfunction due to the reflecting a decrease in factor VII, which is the first factor to be
effects of increased plasmin on platelet receptors. Increased reduced in liver disorders because of its short half-life.
levels of nitric oxide may also induce some degree of platelet The bleeding time does not predict gastrointestinal bleed-
dysfunction.19 ing in patients with cirrhosis.25 Furthermore, mild to moder-
The liver is the site of production for almost all coagula- ate coagulopathy is not associated with prolonged bleeding
tion factors, including the inhibitors protein C, protein S, and after liver biopsy, as assessed by measurement of the bleeding
antithrombin. Von Willebrand factor (vWF) and tissue plas- time at the biopsy site; even if the studies are normal, patients
minogen activator (t-PA) are produced by endothelial cells. may still bleed at the biopsy site.26 Although controlled stud-
The factor VIII activity is usually increased in liver disease, ies in 21 patients with cirrhosis (platelet counts, 45,000 to
which may reflect activation of the molecule or increased 286,000/μL) have shown that desmopressin (DDAVP) at a
synthesis at sites other than the liver.17 Factor VII, which has dose of 0.3 μg/kg significantly shortens the prolonged bleed-
the shortest half-life of the coagulation factors (2 to 6 hours), ing time for up to 4 hours, its role in prophylaxis has not
is the first to be decreased in liver disease. To be functional, been well defined.27 Nonetheless, DDAVP would be a rea-
factors II, VII, IX, and X as well as protein S and protein C sonable choice for a hemostatic agent if there were concerns
must undergo γ-carboxylation, which is mediated by vitamin about platelet function.27,28
K. However, liver disease causes impairment of γ-carboxyl-
Product Support
III ation that cannot be reversed with vitamin K1.20
Patients with liver disease, including those with hepa- Guidelines for prophylaxis in patients with liver disease
484 tomas, may have an acquired dysfibrinogenemia caused by who are undergoing invasive procedures are summarized in
synthesis of a fibrinogen molecule that has impaired poly- Table 35–2. These suggestions are empirical and are based
merization because of an increased sialic acid content.21 The on many factors. For example, in a retrospective study of
fibrinogen levels are usually normal; however, PT, aPTT, and patients undergoing percutaneous liver biopsy, McVay and
thrombin time are prolonged. One case has been reported Toy29 found that mild elevations of PT or aPTT were not
in which acquired dysfibrinogenemia was part of a paraneo- associated with bleeding. Risk factors for bleeding were
plastic manifestation of renal cell carcinoma and resolved malignancy (hepatoma) and multiple passes. These authors
after the tumor was removed.22 Acquired dysfibrinogenemia concluded that PT and aPTT prolonged to less than 1.5 times
is not associated with a bleeding diathesis, and patients with the midnormal range do not require treatment and that a
this abnormality do not need any treatment before invasive platelet count above 50,000/μL is satisfactory.
procedures. For patients in whom standard percutaneous biopsy
Although inhibitors of the coagulation system are is contraindicated, such as those who have ascites, portal
decreased in patients with hepatic disease, a thrombotic hypertension, and coagulopathy, laparoscopic liver biopsy
state does not usually occur. Liver disease is associated with can be performed successfully using direct pressure and topi-
hyperfibrinolysis, as defined by increased activity of t-PA and cal Gelfoam and thrombin to achieve hemostasis.30 In these
elevation of d-dimer. In one study of cirrhotic patients, those patients there appears to be no correlation between the risk
with elevated t-PA and d-dimer had a significantly higher of bleeding and the prophylactic administration of fresh fro-
rate of gastrointestinal bleeding than those who did not have zen plasma (FFP) or platelets. The advantage of laparoscopic
these laboratory findings.23 Patients with ascites and hyper- liver biopsy is the ability to place direct pressure along with
fibrinolysis were also at higher risk for bleeding than those hemostatic agents.
without ascites. The authors postulated that hyperfibrinoly- There is usually little or no need to infuse FFP in an indi-
sis results in part from activation of the coagulation system, vidual who has only a prolonged PT and a normal aPTT
resulting in release of t-PA in the presence of fibrin. Decreased because factor VII levels of 10% or greater are sufficient for
clearance of t-PA as well as reduced levels of α2-antiplasmin, hemostasis. If PT and aPTT are prolonged and the patient has
which is synthesized in the liver and is the major inhibitor of not responded to empirical treatment with vitamin K1 at a
plasmin, could result in continued plasmin generation and a dose of 10 mg/day subcutaneously for 3 days,17 measurement
bleeding diathesis. of levels of factors IX, X, and V (a non-vitamin K–depen-
The assessment of bleeding risk in a patient with liver dis- dent factor) provides a good assessment of liver function
ease includes measurement of the platelet count, PT, aPTT, with respect to synthesis of coagulation factors. On occasion,
TRANSFUSION OF THE PATIENT WITH ACQUIRED COAGULATION DEFECTS
Table 35–2 Management of Patients with Liver Disease
aPTT, activated thromboplastin time; FFP, fresh frozen plasma; PT, prothrombin time.
Maintain hematocrit 27%–32% Erythropoietin Monitor for hypertension; maintain adequate iron
stores; check iron stores if patient resistant to
erythropoietin.
Acutely enhance platelet function DDAVP, 0.3 μg/kg IV Administer no more frequently than once in 24 hours.
Monitor for hyponatremia to avoid seizures.
Chronic enhancement of platelet Conjugated estrogens, Monitor for adverse effects: fluid retention, hot flashes.
function 0.6 mg/kg/day IV for
4–5 days; or transdermal
17β-estradiol,
50–100 μg/24 hours
applied as patch every
3–5 days for 2 months
DDAVP, desmopressin.
times found that the bleeding time was partially corrected products such as platelets, FFP, and cryoprecipitate may need
by 6 to 48 hours after the first dose, with a peak effect at 5 to be administered. There is no evidence that administration
to 7 days and a duration as long as 14 days.43 Transdermal of blood components increases the severity of the DIC.47
estrogen was efficacious in reducing or stopping bleeding in
a series of six patients with renal failure, prolonged bleed- Sepsis
ing times, and excessive bleeding from telangiectasia of the
gastrointestinal tract.44 The estrogen was administered as Thrombocytopenia may be a prominent aspect of sepsis and
17β-estradiol (Estraderm) in a skin patch that delivered 50 can be due to DIC or to hemophagocytosis.48,49 Patients with
or 100 μg/day. In these patients, the mean bleeding time hemophagocytosis also have elevated levels of ferritin and
decreased from 14 to 7 minutes. The effects were noted at 24 lactate dehydrogenase in addition to thrombocytopenia. The
hours and persisted at 17 days (total duration of administra- process resolves with treatment of the underlying condition.
tion was 2 months). No adverse effects were noted.44 Vigano Disseminated intravascular coagulation in sepsis is due
and colleagues postulate that the prolonged effect of estro- in part to interleukin-6-induced expression of tissue fac-
gens is due to their ability to enter endothelial cells, which are tor, which generates procoagulant activity. The fibrinolytic
known to have estrogen receptors, and alter their function response to fibrin formation is reduced because interleukin-6
III
so that improved hemostasis occurs.45 Other studies have also increases the expression of plasminogen activator inhib-
shown that administration of estrogen inhibits endothelial itor-1, the major inhibitor of t-PA. Thus, the major manifes-
486 tation in sepsis may be microvascular fibrin deposition. The
nitric oxide production.36
In renal dialysis patients, graft thrombosis is a frequent efficacy of heparin in this setting has not been demonstrated.
event; secondary prevention includes administration of aspi- Although administration of antithrombin III concentrates
rin, warfarin, or heparin depending on the clinical situation; initially suggested promise in the treatment of sepsis, a phase
low-molecular-weight (LMW) heparin is not used because III international trial of antithrombin III showed no effect
of its prolonged half-life in patients with renal failure.46 on 28-day all-cause mortality in adult patients with severe
sepsis and septic shock who received the concentrate within
6 hours of symptoms.50
Disseminated Intravascular Coagulation Activated protein C (drotrecogin alfa [activated]) is
Disseminated intravascular coagulation is associated with approved for patients with severe sepsis; however, it does not
clinical conditions, such as sepsis, acute brain injury, or appear to be beneficial in patients with severe sepsis who are
abruptio placentae, that induce expression of tissue factor at low risk for death.51,52 In this population, it is associated
and overwhelming activation of the coagulation system, with an increase in the rate of serious bleeding compared to
with a resulting fibrinolytic response. The clinical manifes- placebo (2.4% vs 1.2%, p = 0.02).51
tations of DIC are variable and depend on whether micro-
Acute Leukemia and DIC
vascular thrombosis with fibrin formation is the prominent
feature or the major component is the fibrinolytic response Although DIC can be associated with any type of leukemia,
to fibrin formation.47 The clinical manifestations determine it is most frequently associated with acute promyelocytic
which hemostatic therapy, if any, is needed. In general, the leukemia because the promyelocytes are rich in releasable
most critical factor in the resolution of DIC is the ability to tissue factor. The plasmin which is generated in DIC is ini-
control the condition responsible for the initiation of the tially inhibited by α2-antiplasmin, its major inactivator.47
process. However, with time, the α2-antiplasmin is depleted, allowing
The laboratory evaluation of DIC includes measurement plasmin activity to remain unchecked. This results in ongo-
of the platelet count, PT, aPTT, and d-dimer. An abnormality ing fibrinolysis and excessive bleeding.
in any one of these tests is not specific for DIC. Rather, the The laboratory diagnosis of α2-antiplasmin deficiency can
test results are combined with the clinical setting to make be confirmed by a specific assay for α2-antiplasmin that con-
a diagnosis. Serial coagulation tests are especially useful in sists of measuring the rate at which plasmin added to patient
establishing a diagnosis. If a patient is actively bleeding, blood plasma undergoes inhibition. Treatment is usually empiri-
TRANSFUSION OF THE PATIENT WITH ACQUIRED COAGULATION DEFECTS
cal and consists of administration of an antifibrinolytic patient with or without hemophilia A. Postoperative patients
agent such as oral aminocaproic acid or tranexamic acid.53,54 may also develop an antibody against factor V, which may be
Although the antifibrinolytic agents can be used alone, due to exposure to bovine thrombin preparations that con-
heparin is generally administered as an adjunctive agent at tain factor V. Patients may also develop lupus anticoagulants;
a dose of 300 to 500 U/hr to prevent continued generation of in the critical care setting, these are usually IgM antibod-
thrombin. A critical aspect of management of patients with ies that are directed against phospholipid-binding proteins
acute promyelocytic leukemia is to maintain a platelet count such as prothrombin. They are not associated with excessive
of 50,000/μL or greater. bleeding unless the patient also has a true deficiency of pro-
A combination of heparin and antifibrinolytic agents is thrombin in association with the lupus anticoagulants. The
also useful in patients with DIC related to solid tumors, such essential goals for the successful treatment of patients with
as prostatic carcinoma. Efficacy of treatment can be moni- inhibitors are to establish the initial diagnosis and then plan
tored clinically and by the increase in fibrinogen levels fol- for appropriate factor replacement as well as for immuno-
lowed by an increase in platelets in 24 to 48 hours. suppression to eliminate the inhibitor.
For patients who are not known to have hemophilia but
Obstetrical Conditions
who have a progressive prolongation of aPTT and continued
Disseminated intravascular coagulation can occur with or excessive bleeding with surgery, further evaluation should
abruptio placentae, amniotic fluid embolism, retained pla- include a mixing study to test for the presence of an inhibi-
centa, preeclampsia, acute fatty liver, and in utero fetal tor and measurement of coagulation factors XII, XI, IX, and
death.55 These processes are all associated with the release of VIII. These tests can be followed by a Bethesda assay, which
tissue factor from the dead fetus or necrotic placenta. DIC is will provide an estimate of the inhibitor titer. The results are
clinically noted by excessive or spontaneous bleeding com- less accurate for those patients who have autoantibodies to
bined with decreases in fibrinogen levels and platelet counts factor VIII than for those who have alloantibodies (congeni-
and increases in d-dimer. tal hemophilia). One Bethesda unit (BU) is the reciprocal
The best treatment is to accomplish delivery as soon as of the dilution of patient’s plasma that destroys 50% of the
possible. Blood products are given as needed until this can be factor VIII in normal plasma after incubation for 2 hours at
achieved. The contraction of the myometrium and removal 37° C. For example, a titer of 40 BU indicates that at a 1:40
of the source of tissue factor are the two most important dilution the patient’s plasma can inhibit 50% of the factor
factors in controlling the DIC.55 With proper treatment, the VIII in normal plasma.
coagulopathy reverses in hours and no further treatment
is required. The only obstetrical condition associated with Factor VIII Deficiency
DIC for which heparin has been efficacious is the rare condi-
tion in which fetal death has occurred in one of two or more For patients who have hemophilia and known alloantibod-
fetuses carried by the mother. Full-dose intravenous heparin ies, exposure to factor VIII during the surgical procedure
has been administered to allow the maturation and delivery increases the antibody titer, resulting in a poor response to 35
of the other viable fetuses.56,57 factor VIII by the fourth or fifth postoperative day. In these
patients, serial aPTT and factor VIII measurements can be 487
performed so that appropriate changes in therapy can be
AUTOANTIBODIES AGAINST made if necessary. Depending on the clinical setting and the
COAGULATION FACTORS inhibitor titer, treatment choices include human factor VIII,
rFVIIa, or activated PCCs.
Acquired antibodies to specific blood coagulation factors Autoantibodies to factor VIII generally occur in individu-
have been reported in association with a variety of condi- als after age 50 and are first suspected when the patient pre-
tions, including infections, malignancy, pregnancy, and auto- sents with a complaint of new-onset bruising or bleeding into
immune disorders.58,59 The antibodies, which are usually of the soft tissues or from the gastrointestinal tract.58 Associated
the immunoglobulin G (IgG) isotype, may not be detected conditions include rheumatoid arthritis, the postpartum
until the patient undergoes a hemostatic challenge such as period, allergies, asthma, autoimmune disorders, and malig-
surgery, or they may develop in the postoperative period, nancy; however, in approximately 50% of patients, no other
causing increased morbidity and even mortality. Inhibitors medical disorders are identified.58,59 In 1981, Green and
should be suspected in patients who have sustained trauma Lechner58 reported that major bleeding occurred in as many
or who have undergone a surgical procedure and yet con- as 87% of patients with an acquired factor VIII antibody and
tinue to bleed excessively for no apparent reason. The first resulted in death in 22%. With the advent of improved rec-
manifestation of an inhibitor may not be apparent for 4 to 5 ognition and treatment of inhibitors, in terms of both treat-
days after surgery. A review of serial coagulation studies may ment of bleeding episodes and use of immunosuppression,
show an increasingly prolonged aPTT or both PT and aPTT these data should now be improved.
with no apparent explanation (i.e., not corrected by empiri- Although antibodies to factor VIII can disappear sponta-
cal use of vitamin K1 or FFP or, in the case of individuals neously, the general approach is to begin immunosuppres-
with known hemophilia, decreased recovery immediately sive therapy to increase the factor VIII level to normal levels.
after infusion and shortened half-life of infused factor VIII). Several therapies have been used with different degrees of suc-
Evaluation of coagulation screening studies and measure- cess. Green and coworkers60 reported a response rate of 30%
ment of appropriate factor levels can establish the diagnosis to oral prednisone alone at a dose of 1 mg/kg of body weight
so that appropriate blood product replacement can be pro- administered daily for 3 to 6 weeks. The responders tended to
vided and immunosuppressive therapy initiated. have lower Bethesda titers than nonresponders (median, 3 BU
The antibodies most commonly associated with exces- vs 50 BU), although responses did occur in three patients
sive bleeding are those that develop against factor VIII in a with titers of 33, 52, and 240 BU. They also concluded that
TRANSFUSION MEDICINE
cyclophosphamide at an oral dose of 2 mg/kg daily is an effi- In patients with high-titer inhibitors, rFVIIa has been
cacious therapy for patients who cannot tolerate prednisone.60 efficacious at a dose of 90 μg/kg administered intravenously
In a follow-up publication, Green61 reported an 80% response every 2 hours during surgery and for 48 hours postopera-
rate in 10 patients treated only with prednisone; their initial tively, followed by every 2 to 6 hours for the next 3 days.67
titers ranged from less than 1 BU to 27 BU. Patients who did Factor VIIa has also been tested for home use in the treatment
not respond to steroids or had adverse effects all responded to of mild to moderately severe bleeding episodes in the joints,
single-agent cyclophosphamide. muscles, or mucocutaneous tissues in patients with hemo-
In a report of nine consecutive patients (ages 50 to 79) philia A or B with inhibitors.68 In this study, patients received
with inhibitor titers ranging from 2.5 to 1040 BU, Shaffer rFVIIa (90 μg/kg) intravenously at 3-hour intervals within 8
and Phillips62 described complete remission in 2 to 10 weeks hours of the onset of the episode. If the dose was considered
with combined treatment with oral cyclophosphamide (100 effective after one to three applications, one additional injec-
to 200 mg/day) and prednisone (50 to 80 mg/day) with slow tion was provided. rFVIIa was considered effective in 92%
tapering of these doses when the inhibitor titer was no longer of bleeding episodes after a mean of 2.2 injections. The time
detectable. Cyclosporine has also been used successfully in a from onset of bleeding to the first injection in the successfully
limited number of patients at a dose that provides therapeutic treated episodes was 1.1±2.0 hours (standard deviation).
serum levels (150 to 350 ng/mL).63 Several papers have described the use of sequential or
The most promising treatment to eliminate factor VIII alternative doses of activated PCCs and rFVIIa in patients
autoantibodies appears to be rituximab, a monoclonal anti- with hemophilia and inhibitors who were actively bleed-
body against CD20, which removes CD20+ B lymphocytes ing.69,70 Activated PCCs have been given in doses ranging
from the circulation. The dose generally used is 375 mg/m2 from 35 to 80 U/kg alternating every 6 hours with rFVIIa
IV every week for 4 weeks with or without prednisone.64 This (80 to 225 μg/kg).69 If sequential therapy is to be undertaken,
treatment may reduce the duration of exposure to cytotoxic the patients should undergo serial laboratory monitoring
treatment for these patients. Based on their experience in a for DIC. Additional studies need to be done to confirm the
limited case series, Aggarwal and colleagues have suggested safety and efficacy of this treatment.
a treatment algorithm for older patients with autoimmune
hemophilia.64 They recommend prednisone alone for those Factor V Deficiency
with a Bethesda titer of <5BU, and addition of rituximab to
prednisone for those with low to intermediate titers (<30 Antibodies to factor V are rare and are first manifested clini-
BU). For those with higher titers, a combination of predni- cally as unexpected bleeding or bruising in a patient with
sone, cyclophosphamide, and rituximab may be required. a prolonged PT and aPTT. Multiple reports have described
Patients with very high titers may need to be treated with postoperative patients who have experienced significant
regimens that are used for chronic lymphocytic leukemia.64 bleeding related to factor V antibodies.4–7 The coagulopa-
A general approach to the use of coagulation factor compo- thy develops as a result of exposure of the patient to bovine
III nents in patients with acquired factor VIII inhibitors is sum- thrombin during surgery. Bovine thrombin, which is a com-
marized in Table 35–4. For patients with an inhibitor titer that monly used hemostatic agent, contains multiple impurities,
488 is less than 5 BU, factor VIII concentrates can still be used to including bovine factor V. Patients develop antibodies to the
neutralize the antibody and achieve a hemostatic level of fac- bovine factor V, which then crossreact with their endogenous
tor VIII.65 For those with a higher titer, other options exist, factor V, causing a marked inhibition of factor V activity.5
such as activated PCCs (Feiba and Autoplex T) and rFVIIa. In Patients may also develop antibodies to the bovine throm-
the United States, rFVIIa has largely replaced the use of PCCs. bin. However, these antibodies do not crossreact with human
In a retrospective French study of use of Feiba in 60 thrombin and are of no clinical significance.
patients with inhibitors (most of whom had congenital Patients who have antibodies to factor V have a pro-
hemophilia), efficacy was rated as excellent in 81% of bleed- longed PT and aPTT that are not corrected when the tests
ing episodes, which included surgical procedures, and poor are repeated with a 1:1 mix of normal and patient’s plasma.
in 17%.66 The treatment was well tolerated in 98% of epi- The prolongation of the PT and aPTT becomes even greater
sodes; the doses infused were 65 to 100 U/kg at 6- to 12-hour when the tests are repeated after incubation of patient’s and
intervals (total, 65 to 510 U/kg/day). normal plasma for 1 or 2 hours at 37°C. Measurement of
Table 35–4 Treatment of Bleeding Episodes in Patients with Inhibitors to Factor VIII
If Bethesda titer is < 5 BU, use recombinant factor VIII concentrates; monitor factor VIII activity. Because Bethesda titers are not
highly accurate in nonhemophiliac patients with an inhibitor, recombinant factor VIII may also be effective if the titer is > 5 BU.
The factor VIII activity is the best guide for treatment. An initial high dose may be required to neutralize the inhibitor in vivo,
and then lesser doses may provide satisfactory activity. If possible, factor VIII should be the first choice for treatment.
Recombinant factor VIIa can be used at a dose of 90 μg/kg every 2 hours for the duration of bleeding and for several days
thereafter. Choice of recombinant factor VIIa over activated prothrombin complex concentrates may be based on experience
of the physician, cost, and availability. The two treatments have not been directly compared. Disadvantage of the activated
prothrombin complex concentrates is potential for inducing thrombosis.
Activated prothrombin complex concentrates (Feiba 75–100 U/kg every 8–12 hours or Autoplex 50–75 U/kg every 8–12 hours)
can be used for soft tissue bleeds and other bleeding episodes. Efficacy is judged by patient’s response. For soft tissue bleeds,
patients can be treated every 8 hours for 48 hours followed by every 12 hours for the next 48 hours.
TRANSFUSION OF THE PATIENT WITH ACQUIRED COAGULATION DEFECTS
coagulation factors shows a decrease in factor V. An antibody Lupus anticoagulants and clinically significant hypopro-
to bovine thrombin is detected by a prolongation of the thrombinemia have been associated with viral illnesses in
thrombin time if bovine thrombin is used in the test system. children.76 In these cases, the antibody disappeared sponta-
The test remains abnormal when repeated with a 1:1 mix of neously. The antibody production can usually be suppressed
normal and patient’s plasma. The thrombin time is normal, by administration of corticosteroids with or without adjunc-
however, if human thrombin is used in the test system. tive agents such as azathioprine; with this treatment, the PT
The majority of patients in whom an antibody to factor can become normal as soon as 7 days after initiation of treat-
V develops do not have clinical symptoms and the PT and ment.73,75 For example, a 3-year-old girl with a viral illness
aPTT become normal within 3 to 6 weeks after exposure to and severe hypoprothrombinemia that resulted in epistaxis
thrombin. For patients who are actively bleeding, FFP is the and gastrointestinal hemorrhage was treated successfully
first choice for replacement of factor V. Infusion of platelets with intravenous methylprednisolone daily for 3 days fol-
is a second option because they contain approximately 20% lowed by oral prednisone.77 For a patient who is actively
of the body stores of factor V, although they may not provide bleeding and has a severe prothrombin deficiency, treatment
hemostasis in all situations.71 Treatment with rFVIIa may options are FFP or PCCs. Although not reported, rFVIIa
also be an option for these patients. could also be used.
Corticosteroids, cyclophosphamide, and vincristine have
been used alone or in combination for suppression of anti- Factor XIII Deficiency
bodies. Corticosteroids have mostly been used alone as first-
line therapy with success. Methylprednisolone at an initial Factor XIII deficiency may be due to the formation of auto-
dose of 1 to 1.5 mg/kg is usually given for 2 to 3 weeks and then antibodies that occur spontaneously or in association with
tapered slowly over the next several weeks.7 Cyclophosphamide drugs such as penicillin, isoniazid, or diphenylhydantoin
and vincristine have been added to prednisone to suppress the or with autoimmune disorders.78,79 The inhibitor is not
antibody production successfully in cases with acquired fac- detected on routine screening tests and can be confirmed
tor VIII inhibitors; similar success has been achieved in some only by finding rapid lysis of a clot that has been prepared
cases with factor V inhibitors. The use of rituximab has not from recalcified plasma and placed in 1% monochloroace-
been described in this clinical setting. tic acid or urea. Treatment consists of immunosuppressive
agents and administration of cryoprecipitate if the patient is
actively bleeding.
Factor II Deficiency
Laboratory monitoring of therapy can be done serially
Autoantibodies to prothrombin usually occur in asso- and qualitatively by assessing the time required for solu-
ciation with lupus anticoagulants, which are antibodies to bilization of a clot derived from patient plasma in urea or
phospholipid-binding proteins such as prothrombin or β2- monochloroacetic acid.80,81 Successful immunosuppression
glycoprotein I.72 Lupus anticoagulants are associated with with rituximab has been reported in one patient with an
the antiphospholipid syndrome, systemic lupus erythema- apparent ciprofloxacin-induced antibody.82 35
tosus (SLE), and infections as well as use of medications,
such as procainamide (Pronestyl). In some patients with 489
lupus anticoagulants, a true deficiency of prothrombin has COAGULOPATHIES RELATED TO
also been recognized; this appears to be due to IgG antibod- ANTICOAGULANT ADMINISTRATION
ies that selectively bind prothrombin without neutralizing
its activity. The antibodies do not interfere with the proteo- Warfarin
lytic cleavage of prothrombin or with the activity of throm-
bin. They appear to induce true prothrombin deficiency The frequency of bleeding in patients who are receiving
by binding to prothrombin in vivo and causing increased warfarin ranges from 8 to 16 per hundred patient-years,
clearance.73–75 with major bleeding occurring in 1.3 to 2.7 patients per
The majority of patients with lupus anticoagulants and hundred years.83 Patients at increased risk for bleeding are
SLE or other autoimmune connective tissue disorders may the elderly; those who have comorbid conditions such as
have IgG autoantibodies to prothrombin, although only 30% cardiac, cerebrovascular, renal or hepatic disease; and those
of patients with antibodies have a detectable prothrombin who have just recently been started on warfarin therapy
deficiency.72 Antiprothrombin antibodies are more likely or who have increased or fluctuating INRs.84 One of the
to occur in patients with an aPTT that is significantly pro- most important factors that predicts the risk for major
longed (at least 50 seconds; normal range, 22 to 30 sec- bleed is a history of bleeding, especially if the site is in the
onds).74 Bleeding symptoms are correlated with the level of gastrointestinal tract.85
functional prothrombin and occur in patients with a pro- Maintenance daily doses of warfarin can range from 4 mg
thrombin activity of 10% or lower.74 In some cases, the first up to 90 mg or higher in rare circumstances. Overall, more
manifestation of lupus is the development of a lupus anti- than 40% of the variability in warfarin dose requirements
coagulant with true prothrombin deficiency. A significant may be explained by genetic factors, use of concomitant spe-
decrease in the prothrombin activity related to antibody for- cific medications such as amiodarone, age, and body surface
mation can occur at any time in a patient who has a lupus area.86 Two examples of the influence of genetic variation
anticoagulants and underlying SLE or primary antiphos- affecting dose requirements are related to the polymor-
pholipid syndrome. For patients with a lupus anticoagulants phisms in the gene that codes for the cytochrome P450 com-
who are receiving warfarin, the first manifestation of an anti- plex (CYP), which is responsible for metabolizing the potent
body to prothrombin that induces a significant decrease in “S” form of warfarin. Individuals who have polymorphisms
the prothrombin activity may be a gradually increasing INR in the CYP2C9 gene, which produces the protein that metab-
with no other apparent etiology. olizes “S” warfarin, may have reduced catabolism of the “S”
TRANSFUSION MEDICINE
form and therefore have greatly reduced dose requirements.86 bleeding) who needed rapid reversal of excessive warfarin
Another genetic determinant of warfarin requirement is anticoagulation.90 rFVIIa was effective in doses ranging from
due to polymorphisms in the vitamin K epoxide reductase 15 to 20 μg/kg. The cost of the product was approximately
complex subunit 1 (VKORC1) gene, which affects the activ- $1.40/μg at the time of this study. The product initially was
ity of vitamin K epoxide reductase complex. This complex, administered at a dose as high as 90 μg/kg; however, lower
which is the target of warfarin, is responsible for regenerat- doses were found to be effective. None of the patients sus-
ing reduced vitamin K so that it can serve as a cofactor for tained any adverse side effects from rVIIa. The INR value
γ-carboxylation of coagulation factors. Polymorphisms in does not reflect the efficacy of the treatment; rather, patients
the gene can confer marked resistance or increased sensitivity who receive rFVIIa need to be clinically evaluated.
to warfarin.87 For patients who have ingested rat poison containing a
The main strategies for treating warfarin overdose, depend- superwarfarin such as brodifacoum, prolonged treatment
ing on the degree of overanticoagulation and the clinical site with daily oral vitamin K is needed because the half-life of
of bleeding, include administration of vitamin K1, rFVIIa, the superwarfarin is 16 to 36 days.91 Oral vitamin K has a
PCCs, or FFP if the other products are not available (Table bioavailability of 10% to 60% and is therefore given at a dose
35–5).85 PCCs, which contain factors II, VII, and X in addi- three to five times greater than the parenteral dose. Initially,
tion to factor IX and are virally inactivated, were originally oral vitamin K may have to be taken every 6 to 8 hours and
developed for treatment of individuals with hemophilia B. then tapered to a daily dose.
Although PCCs are no longer used for this purpose, they are
effective for use in the emergency correction of coagulation Heparin Preparations (Unfractionated
factor deficiencies in individuals who have major bleeding Heparin, LMW Heparin,
and are anticoagulated with warfarin.88
and Pentasaccharide)
Approximately 50% to 90% of intracranial hemorrhages
in patients receiving warfarin anticoagulation occur at a According to one study, risk factors for bleeding are the
time when the INR is within the target range.84 The mortal- same for patients receiving unfractionated heparin or
ity associated with intracranial bleeding is 16% to 68% and LMW heparin and include the performance status of the
is influenced by the rapidity with which normalization of the patient as determined by World Health Organization cri-
coagulation factor deficiencies occurs. teria, history of a bleeding tendency, cardiopulmonary
The preferred replacement therapy in intracranial hem- resuscitation, recent trauma or surgery, body surface area,
orrhage is rFVIIa or PCCs (25 to 50 U/kg). When PCCs and total dose of anticoagulant administered in 24 hours.92
were compared with infusion of 800 mL of FFP in a small The bleeding risk may also be increased in older patients
study of patients with intracranial hemorrhage who were (>70 years).93,94
also anticoagulated with warfarin, the coagulation factor For unfractionated heparin, reversal of anticoagulant
levels were raised into the normal range with the PCCs but activity can be achieved by slow infusion (over 10 min-
III not with FFP.84 The INR can be corrected four to five times utes) of protamine sulfate, which neutralizes 100 units of
more quickly with the PCCs than with FFP. The elimina- heparin for every milligram of protamine administered.
490 tion half-lives of the coagulation factors are as follows: fac- Assuming that heparin has a half-life of 60 minutes, the
tor II (58 hours), factor VII (5 hours), factor IX (19 hours), protamine dose can be calculated in a patient receiving a
and factor X (35 hours); thus, repeated infusion may be constant IV infusion by adding the heparin dose infused
necessary if the factor deficiency has not been reversed in during each of the past 3 hours.95 The risk for anaphy-
24 hours by the administration of vitamin K. Factor VII is laxis with protamine sulfate is approximately 1% and may
in low concentration in some PCCs relative to the other be higher in diabetics who have taken NPH insulin or in
coagulation factors, and it has a short half-life. The impor- patients who have allergies to fish or have undergone a
tance of replacing the factor VII level to values greater than vasectomy. Patients at risk for anaphylaxis should be pre-
10% is unknown. FFP can be used to achieve minimal levels treated with corticosteroids and with antihistamines.96 For
of 10% if this is deemed necessary. For some patients, espe- LMW heparin, protamine neutralizes only about 60% of
cially those with mechanical heart valves, resuming antico- the anti-factor Xa activity. If LMW heparin is administered
agulant therapy is critical; one report of 15 patients showed within 2 hours of the bleeding episode, then protamine can
that this could be done safely.89 be given based on a calculated dose of 1 mg for every 100
In a prospective case series, Deveras and Kessler admin- anti-factor Xa units given. This can be repeated at half the
istered rFVIIa to 13 patients (some of whom were actively dose if bleeding continues.
*
Waterstone M, Bewley S, Wolfe C. Incidence and predictors of severe obstetric morbidity: case-control study. BMJ 2001;322:1089–1094.
Clinically significant
Paternal Phenotype or
Rh Neg Infant not
Fetal Genotype
at risk
(where available)
Rh Pos
28 weeks RhIG 300 g
May elect out of 28 week dose
if father confirmed Infant at risk or
Rh negative by testing unknown status
At Delivery At Delivery 36
ABO/RhD/antibody screen if ABO/RhD/antibody screen if
1) Transfusion likely 1) Transfusion likely High Titer (Critical) Titer is £8 (except Anti-K)
(Mom or baby) (Mom or baby) Anti-K: any level Titer and Middle cerebral 499
2) Only 1 historic ABO/Rh 2) Only 1 historic ABO/Rh All others: 16 artery peak systolic velocity
3) No ABO/Rh in this 3) No ABO/Rh in this OR previous affected infant every 2-4 weeks
pregnancy pregnancy
Rh D Positive Baby Rh D Negative Baby Intensive fetal monitoring Titer becomes 16
Test for FMH No RhIG intrauterine transfusions OR fetal monitoring
Give 300 g RhIG No FMH test No further titers confirms HDN
If FMH test positive,
additional RhIG as
required
At Delivery
ABO/Rh and antibody screen to
exclude other antibodies (or if
newborn requires transfusion)
No titres required
Figure 36–1 Sample algorithm for antenatal and peripartum blood bank testing.
have delivered infants with hemolytic disease of the fetus sequelae resulted from the newly formed antibodies.74,75 Thus,
and newborn (HDFN).72 antibody screening can be safely restricted to those patients at
The risk of alloimmunization to the D antigen between delivery requiring pretransfusion testing or investigation for
the first trimester and 28 weeks’ gestation is exceptionally low HDFN. Titration studies after RhIG administration to ensure
(0.18%)73; therefore, the cost effectiveness of repeat antibody adequate coverage for prevention of alloimmunization or to
screening before administration of RhIG at 28 weeks has been differentiate between active and passive anti-D are not useful-
questioned. Two retrospective series found that routine anti- nor recommended.66 There is no data to support the need for
body screening in the third trimester or at delivery was also an additional dose of RhIG in the last few weeks of pregnancy,
of low yield (0.06% to 0.24%) and no significant neonatal even if the antibody screen is found to be negative.
TRANSFUSION MEDICINE
Some clinicians recommend routine pretransfusion test- were no significant differences in the prevalence of hepati-
ing at delivery on all patients, in case there is a need for an tis C virus and human immunodeficiency virus between
emergency transfusion and the patient has an antibody that Rh-negative and Rh-positive female donors.85 Adverse drug
was not previously detected. A large retrospective study does reactions are uncommon and may include fever and pain at
not support this opinion.76 This group estimated that only 1 the injection site. Hypersensitivity reactions are rare.86
in 11,050 patients undergoing cesarean sections would have
both a new clinically significant antibody and the need for a Fetomaternal Hemorrhage Testing
red cell transfusion. The investigators performed the same
analysis for those with vaginal deliveries and found this dual Fetomaternal hemorrhage quantitation is usually performed
occurrence to be 1 in 61,343 deliveries. Their data supports by acid elution test (Kleihauer-Betke test) or hemoglobin
the practice of withholding pretransfusion testing in patients F quantitation by flow cytometry, although other methods
undergoing vaginal delivery or cesarean section, unless sig- are available. The test is performed to determine either the
nificant risk factors for hemorrhage are present. In the unfor- appropriate dosage of RhIG in an RhD-negative mother
tunate event of an unexpected obstetrical hemorrhage, group or to diagnose and quantitate a fetomaternal hemorrhage.
O, RhD-negative blood should be provided while expediting The Kleihauer-Betke test has numerous limitations, includ-
blood grouping and antibody screening. ing low sensitivity, poor reproducibility, and a tendency to
overestimate the volume of hemorrhage. An important limi-
tation of the Kleihauer-Betke test is the inability to differen-
Rh Immune Globulin Administration
tiate between maternal and fetal F cells. This is particularly a
A combination of an antepartum and a postpartum dose problem in the second trimester, when maternal F cells may
of RhIG reduces the risk of sensitization to 0.1% in clinical occasionally reach 5% to 10%. If a Kleihauer-Betke test is
studies.77 All unsensitized RhD-negative women are candi- performed during this time, this physiologic change may be
dates for RhIG at 28 weeks, at delivery of an RhD-positive mistaken for a fetomaternal hemorrhage.
fetus, and in the case of spontaneous abortion, ectopic preg- Hemoglobin F quantitation by flow cytometry has been
nancy, therapeutic abortion, amniocentesis, chorionic vil- found to be simple, reliable, and more precise than the
lus sampling, external cephalic version, abdominal trauma, Kleihauer-Betke test.82,87 In the 2001 CAP survey, samples
antepartum hemorrhage, and other inciting events.78 If the representing a 20-mL and a 40-mL bleed were sent as exter-
father of the fetus has been tested and has been shown to be nal proficiency testing. Beyond a 30-mL fetomaternal hemor-
RhD-negative, the woman may elect not to receive RhIG. The rhage, a single 300-μg dose of RhIG would be insufficient to
paternal sample should be obtained during the current preg- prevent alloimmunization. For the 20-mL hemorrhage, 88%
nancy, and the result should be placed in the medical records of laboratories utilizing hemoglobin F quantitation and 50%
of the pregnant partner. Beyond 20 weeks’ gestation, quanti- of the laboratories using the Kleihauer-Betke test correctly
fication of fetomaternal hemorrhage should be performed to reported the result as less than 30 mL. For the 40 mL hem-
III determine if a single dose of RhIG is adequate for prevention orrhage, 100% of laboratories utilizing hemoglobin F quan-
of immunization. Such testing must be performed at each titation and 89% of the laboratories using Kleihauer-Betke
500 inciting event where RhIG administration is required. After correctly reported the result as greater than 30 mL. This data
20 weeks, the fetal blood volume exceeds 30 mL79; thereafter, suggests that hemoglobin F quantitation may be a superior
a single dose of 300 μg may be insufficient and testing for test method. No clinical management decisions should be
fetomaternal hemorrhage is recommended.68,80 Massive feto- based solely on the results of fetomaternal hemorrhage test-
maternal hemorrhage may occur without any foreseen cause; ing given the inaccuracies found in the studies and surveys
therefore, the need for testing should not be dictated by described above. See Figure 36–2 for sample histograms from
clinical judgment of the risk of fetomaternal hemorrhage.81 control and patient samples tested with the hemoglobin F
Massive fetomaternal hemorrhage in excess of 30 mL is rare. quantitation method.
In one series of 1131 patients tested for fetomaternal hem-
orrhage by hemoglobin F quantitation, 1.3% were found to
have a hemorrhage of greater than 30 mL.82 If a dose of RhIG MONITORING OF ALLOIMMUNIZATION
is inadvertently not administered within the recommended DURING PREGNANCY
72 hours of an inciting event, the dosage should be given as
soon as possible and up to 28 days postpartum.83 Incidence of Alloimmunization
and Hemolytic Disease of the Fetus
Risks of Rh Immune Globulin and Newborn
Rh immune globulin is manufactured with multiple viral Overall, approximately 6.8 per 1000 pregnancies are com-
inactivation strategies to minimize the risk of transmitting plicated by RhD alloimmunization.88 In one large series,
blood-borne diseases. The safety strategies include donor 1% of pregnancies were shown to be complicated by
selection, donor transmissible disease testing, solvent-deter- alloimmunization to antigens implicated in HDFN.89 The
gent viral inactivation, and nanofiltration. Failures in the implicated antibody in this series was anti-D in 41%, other
manufacturing process or donor testing could result in infec- Rh antibodies in 30%, and non-Rh antibodies in 29%. In
tious complications. RhIG has been implicated in the trans- 244 alloimmunized pregnancies, there were 3 intrauter-
mission of hepatitis C virus in Ireland, where inadequate ine deaths, 3 fetuses requiring intrauterine transfusion,
viral safety measures were in place.84 In North America, there 10 infants requiring allogeneic transfusion after delivery,
has been no evidence of viral transmission from this pooled and 27 infants requiring phototherapy. Half of Rh-posi-
human blood product. Of 624,939 female donors included tive infants born to women alloimmunized to the D anti-
in a U.S. study to determine the viral safety of RhIG, there gen require no treatment for HDFN.90 In the presence of
100
1000
quantitation performed by flow
cytometry with anti–hemogobin
F in a nonpregnant control and
three patients. For each sample,
a quantitative histogram of fluo-
rescence intensity (left) and a dot
plot representation of side scatter
SS LOG
Count
.1
J K
0
1000
fetomaternal hemorrhage (gate C,
50
C
D
.1
J I
0
.1 1000 .1 1000
B FL2 LOG FL2 LOG
1000
50
36
501
SS LOG
Count
C
.1
J I
0
.1 1000 .1 1000
C FL2 LOG FL2 LOG
1000
50
L K
SS LOG
Count
D
C
.1
J I
0
.1 1000 .1 1000
D FL2 LOG FL2 LOG
TRANSFUSION MEDICINE
multiple antibodies, the presence of an anti-D is the single pregnancy. The risk of complications from such interven-
most significant factor predicting the risk of HDFN, with the tions must be weighed against the perceived benefits.
presence of anti-D plus another antibody further increasing
Antibody Titers and Cellular Assays
that risk.91 Table 36–2 details the clinical significance of red
cell alloantibodies in HDFN. The role of blood bank testing is to identify the alloantibod-
ies present, to determine their clinical significance, and to
perform titrations to assist the clinicians as to when more
Assessing the Severity of Hemolytic Disease invasive fetal monitoring is required. An excellent review
of the Fetus and Newborn on the role of titrations has been published and should be
reviewed prior to setting testing algorithms.66 Titers should
Introduction be performed by the saline tube technique, with previous
The role of laboratory testing and fetal imaging is to deter- samples run in parallel. The use of gel technology should
mine whether intrauterine transfusion or early delivery be avoided unless validation studies are performed within
is warranted in the management of an alloimmunized the laboratory to establish a new critical titer. The use of
Table 36–2 Potential of Blood Group Alloantibodies to Cause Hemolytic Disease of the Fetus
and Newborn (HDFN)*
*
Guidelines for Prenatal and Perinatal Immunohematology. Bethesda, Md., AABB Press, 2005.
OBSTETRIC AND INTRAUTERINE TRANSFUSION
heterozygous cells and enhancement strategies, such as gel Middle Cerebral Artery Peak Systolic Velocity
technology, likely explains the wide disparity in titer results and Fetal Ultrasonography
seen in external proficiency testing programs. A sample
Middle cerebral artery peak systolic velocity, measured by
algorithm is shown in Figure 36–1. The critical titer var-
Doppler ultrasonography, is an accurate test for detecting
ies between laboratories and is approximately 8 to 32 for
fetal anemia.101,102 This technique is noninvasive and there-
anti-D.66,92 The critical titer is thought only to be of value in
fore presents no risk of miscarriage or preterm labor, and
evaluating the first affected pregnancy.93 The lack of correla-
thus is a preferable method of screening for fetal anemia
tion of titers with fetal outcomes has caused some to ques-
when compared to invasive alternatives. There is a recipro-
tion whether titers have any role in the antenatal assessment
cal relationship between the fetal hemoglobin concentra-
of alloimmunized pregnancies.94 A critical titer for non-D
tion and the velocity of cerebral blood flow. Middle cerebral
antibodies has not yet been defined, and the role of titers
artery peak systolic velocity can also be utilized to predict
for non-D alloimmunization pregnancies remains unclear.
fetal anemia in non-Rh alloimmunized pregnancies, includ-
Anti-Kell is a clear example of an instance where titers of 4
ing anti-K alloimmunized pregnancies.103 Intrahepatic
or less can be associated with HDFN.95,96 Given the above
umbilical venous maximum velocity may also correlate
limitations of titrations, once the critical titer is reached,
with middle cerebral artery peak systolic velocity and fetal
more invasive testing such as ultrasonography, Doppler
anemia.103 A threshold value for the peak systolic veloc-
assessment of the middle cerebral artery peak systolic veloc-
ity of 1.50 multiples of the median for gestational age has
ity, amniocentesis, and cordocentesis is warranted. After the
a 100% sensitivity (95% CI, 86%–100%) for the detection
titer has reached a critical level, no further titers should be
of moderate or severe anemia in fetuses without hydrops.102
performed. Titration studies are of no value after delivery.
Compared with conventional management, middle cerebral
It is recommended that titrations be performed on homo-
artery peak systolic velocity may have a better predictive
zygous cells for the offending antigen and when anti-D is
accuracy for moderate to severe fetal anemia.104 Management
present on R2R2 cells, where the expression of the D antigen
with middle cerebral artery peak systolic velocity will likely
is more uniform donor to donor.66 In addition, where avail-
eliminate the need for amniocentesis and reduce the num-
able, the current sample should be run in parallel with the
ber of periumbilical cord blood samplings performed in
previous sample.97 Titrations should be performed every 2
red blood cell alloimmunized pregnancies. As ultrasonog-
to 4 weeks after 18 weeks’ gestation. If there is a previously
raphy evolves and improves, it is extremely likely that titers
affected infant with HDFN, titers should not be performed
will prove to be obsolete for predicting which fetuses have
and more aggressive monitoring should be adopted early in
significant HDFN.
pregnancy.98
Cellular assays, such as the monocyte monolayer assay, the
Invasive Monitoring with Amniocentesis
chemiluminescence test, and antibody-dependent cell-medi-
and Periumbilical Cord Blood Sampling
ated cytotoxicity assay, which are sensitive to factors affect-
ing antibody function, have been developed to improve the Serial amniocentesis with measurement of the OD450 36
prediction of disease severity.99 These assays are cumber- allowed Liley to divide infants into three risk zones, see
some and not broadly available. There are data to suggest that Figure 36–3.105 The timing of repeat amniocentesis or cord 503
these cellular assays provide clinically useful information to blood sampling depends on the level of amniotic fluid biliru-
complement serologic testing.100 bin. Unfortunately, prior to 27 weeks’ gestation its reliability
0.2
OD
0.1
0.09 ZONE 2
0.08
0.07
0.06
0.05
0.04
0.03
ZONE 1
0.02
0.01
27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42
Weeks
TRANSFUSION MEDICINE
is questionable106; the procedure itself may cause fetomater- blood is an acceptable source for intrauterine transfusion
nal hemorrhage, further stimulating alloimmunization107; where antigen-compatible allogeneic blood is unavailable.116
it carries the risk of procedure-related pregnancy loss of The red cell unit should be plasma-reduced to a final
0.6%.108 Thus, serial middle cerebral artery peak systolic hematocrit of 75% to 80%.93 The red cells should be washed
velocity, instead of amniotic fluid OD450 measurement, is before transfusion if they are older than 7 days or of maternal
used to identify fetuses in whom periumbilical cord blood origin.117 Intrauterine transfusions become feasible around
sampling is required. Once an at-risk infant is identified by 20 weeks’ gestation when safe cannulation of the umbilical
either middle cerebral artery peak systolic velocity or amnio- cord can be performed.118 Intrauterine red cell transfusions
centesis, cordocentesis can be used for: measurement of the may be indicated for reasons other than red cell alloimmu-
fetal hemoglobin, reticulocyte count, and bilirubin; deter- nization, including pure red cell aplasia from parvovirus B19
mination of blood group; and for performance of the direct infection, fetomaternal hemorrhage, twin-to-twin transfu-
antiglobulin test and, where necessary, intrauterine transfu- sion, and hemoglobin Bart’s (α-thalassemia with complete
sion. Cordocentesis is associated with a risk of fetal loss of absence of all four α-globin genes).
approximately 1.0% per procedure.109,110 Cordocentesis can The infant is paralyzed with a short-acting paralytic
be complicated by infection, bleeding, fetal bradycardia, agent and then the needle is inserted into the umbilical vein
intrauterine growth retardation, and premature rupture of under continuous ultrasound guidance. Blood samples are
the membranes. 109,110 drawn for hematocrit and bilirubin. The amount of blood
infused is based on a formula that takes into consideration
Fetal Blood Typing the fetoplacental volume estimated by fetal weight using
The first step in fetal blood typing is to perform paternal ultrasound, the fetal hematocrit, and the unit hematocrit
Rh phenotyping for Rh DCE or the offending antigen to (Fig. 36–4).114 The final target hematocrit is usually 40%.93
determine whether the father is homozygous or heterozy- For severely anemic fetuses, the hematocrit should not be
gous. If the father is homozygous, all infants will express the increased by more than fourfold at the first infusion.119
offending antigen; and if the father is heterozygous, half of Samples are also obtained at the completion of the transfu-
the infants are at risk. Paternal zygosity can be assessed by sion for final hematocrit and for hemoglobin F quantitation
serologic analysis of the major Rh antigens, and the most to determine the ratio of transfused red blood cells to fetal red
probable genotype can be determined by haplotype tables.111 blood cells. Transfusions are repeated every 14 days or less,
Alternatively, quantitative polymerase chain reaction analysis with the interval based on the rate of decline in the hemato-
can be performed to evaluate genomic DNA for the presence crit from the previous transfusion. Severely affected fetuses
of one or two copies of the RhD gene.112 may require transfusions as frequently as every 7 days. The
If the fetus is at risk of expressing the antigen, fetal geno- final transfusion is performed at or before 35 weeks, with
typing can be performed on cells obtained from either chori- delivery planned at 37 weeks’ gestation.
onic villus sampling or amniocentesis. Molecular genotyping Fetal outcome is superior when intrauterine transfu-
III is available for most clinically significant antigens; therefore, sions are begun before the onset of hydrops (mortality is
cordocentesis for fetal red cell phenotyping is not required. 22% to 30% for hydropic and 2% to 8% for nonhydropic
504 Noninvasive methods have been developed to obtain fetal fetuses).120,121 The vast majority of infants require only top-
DNA from maternal plasma. This testing is performed simul- up transfusions in the neonatal period for management of
taneously with maternal and paternal genotyping. Such test- their HDFN.122 Intrauterine transfusions suppress erythro-
ing is complicated by the fact that the D-negative status can poiesis and can cause hypoproliferative anemia in the neona-
arise from multiple genetic variations.113 tal period, especially if the last transfusion occurs close to the
time of delivery.123 It is critical that infants born to alloim-
munized mothers receive close follow-up so that late anemia
MANAGEMENT OF ALLOIMMUNIZATION is detected and managed appropriately.
DURING PREGNANCY A cohort study of 254 fetuses treated with 740 intrauter-
ine transfusions for red cell alloimmunization between 1988
and 2001 documented the risk of fetal death from procedure-
Intrauterine Red Cell Transfusions
related complications at 1.6%.124 In this series, the overall
Intrauterine transfusions are performed by direct infusion procedure-related complication rate was 3.1% (0.1% prema-
of allogeneic blood into the umbilical cord, the intrahepatic ture rupture of membranes, 0.3% infection, 2.0% emergency
portion of the hepatic vein, or intraperitoneally. Red blood cesarean section, 0.9% fetal death, 0.7% neonatal death). Other
cells are selected to be group O, RhD-negative, Kell-negative, known complications of intrauterine transfusion include cir-
cytomegalovirus-seronegative, leukoreduced, irradiated, culatory overload, splenic rupture, and iron overload.125–127
hemoglobin S–negative, and less than 72 hours old.114,115 Bleeding from the puncture site is thought to occur more
The unit must also be negative for the offending antigen
and crossmatch-compatible with the maternal plasma. Kell-
negative red cells are selected to eliminate the risk of immu- Desired PCV − Fetal PCV
nizing the mother against an additional antigen that has Fetoplacental blood volume
Donor PCV − Desired PCV
been implicated in severe HDFN. Irradiated red blood cells
are warranted to eliminate the risk of transfusion-transmit-
PCV = Packed cell volume
ted graft-versus-host disease (GVHD), especially if mater-
nal red cells are selected for transfusion. In the absence of Figure 36–4 Formula for the calculation of the volume of red
blood cells required for an intrauterine transfusion. (From Rodeck CH,
compatible allogeneic blood, such as alloantibodies to high- Dean A. Red cell allo-immunization. In Rodeck CJ, Whittle MJ (eds).
frequency antigens, maternal blood is usually selected for Fetal Medicine: Basic Science and Clinical Practice. London, Churchill
transfusion. In rare situations, maternal ABO-mismatched Livingstone, 1999, pp 785–804.)
OBSTETRIC AND INTRAUTERINE TRANSFUSION
frequently when thrombocytopenia and severe hydrops are Premature Delivery
present, with some physicians recommending routine plate-
let transfusions for infants with hydrops.128 Transamniotic Early delivery is performed in all cases of severe HDFN,
cord needling or arterial puncture is thought to increase the usually between 35 and 37 weeks’ gestation. No data exist
incidence of bleeding and both should be avoided. Infants to guide the exact timing of delivery in HDFN; however,
should be closely monitored during and after the procedure. once fetal lung maturity has been achieved and hemolysis is
The procedure should be performed in close proximity to ongoing, delivery should be expedited.
the operating room, so that if emergency cesarean section is
required it can be performed without delay.
Despite severe fetal hemolytic disease, normal develop- NEONATAL ALLOIMMUNE
mental outcome can be expected for children treated with THROMBOCYTOPENIA
intrauterine transfusions based on a study with long-term
follow-up.129 In this report, 40 children who survived severe
Incidence
HDFN were followed until age 62 months, and no difference
was found in the global developmental quotient (101.9 +/− Neonatal alloimmune thrombocytopenia is a serious dis-
9.5, compared to the mean for the normal population of 100). order resulting from platelet–antigen incompatibility
There was no correlation between the global developmental between the mother and fetus. It is the most common
quotient and the severity of the HDFN (fetal hematocrit, fetal cause of severe thrombocytopenia in an otherwise healthy
bilirubin, presence of hydrops fetalis, total number of intra- infant.137 Overall, 2.5% of women are HPA-1a negative,
uterine transfusions, duration of neonatal phototherapy, and with 12% of these mothers having detectable anti-HPA-
number of neonatal exchange transfusions). 1a.138 Thus, 1 of 350 pregnancies is alloimmunized with
anti-HPA-1a. The risk of alloimmunization is highest in
women who are HLA class II DRB3*0101 (DR52a).139 The
Intravenous Immunoglobulin
incidence of severe thrombocytopenia secondary to anti-
Intravenous immunoglobulin (IVIG) is effective in the HPA-1a is estimated at 1 in 1000 neonates.140 The diagnosis
postnatal treatment of infants affected with HDFN. In this is usually made after the discovery of unexpected neonatal
setting, IVIG reduces the length of hospital stay, the dura- thrombocytopenia or when an intracranial hemorrhage is
tion of phototherapy, and the need for exchange transfusion visualized on a routine prenatal ultrasound. Approximately
(odds ratio, 0.28; CI, 0.17–0.47).130 Prenatally, the reports 10% to 20% of affected fetuses have intracranial hemor-
of IVIG are limited to encouraging case reports131 and case rhages, one quarter to one half of which occur in utero.141
series132,133; thus, the efficacy of this therapy prenatally is The initial platelet count is <20,000/μL in 50% of affected
unknown. Encouraging case reports have also been published fetuses, including 46% of fetuses sampled before 24 weeks’
on the use of plasma exchange and IVIG in the treatment of gestation.142 Fetuses with HPA-1a incompatibility have
severe HDFN presenting before intrauterine transfusions are more severe thrombocytopenia than fetuses with other 36
feasible.134 platelet–antigen incompatibilities. A history of antena-
tal intracranial hemorrhage in a sibling predicts a greater 505
severity of thrombocytopenia in a subsequent fetus.142 One
Phenobarbital Administration
report estimated a 79% risk of intracranial hemorrhage
The rate-limiting enzyme in bilirubin metabolism is biliru- after a diagnosis of intracranial hemorrhage in a previ-
bin-UDP-glucuronosyltransferase (UGT1A1). By increas- ous infant.143 The literature strongly suggests that neonatal
ing the expression of UGT1A1, phenobarbital enhances the alloimmune thrombocytopenia is currently underdiag-
capacity of the neonatal liver to conjugate and eliminate nosed.144 In comparison to HDFN, where first children are
bilirubin. Phenobarbital binds to a 290–base pair enhancer rarely affected, 20% to 59% of NAIT cases occur during the
sequence on the gene for UGT1A1, leading to an increase first pregnancy.141
in the production of this enzyme. A retrospective case-
control study found prenatal administration to be effec- Etiology
tive in decreasing the need for exchange transfusion.135
Intrauterine transfusions were continued until 35 weeks’ HPA-1a and HPA-5b antibodies are implicated in approxi-
gestation. After the last intrauterine transfusion, women mately 80% and 10% of cases of neonatal alloimmune
were offered phenobarbital 30 mg orally three times a day thrombocytopenia, respectively. In one series of 1162
for 10 days. Delivery was planned after the 10-day treatment. serologically confirmed cases in the United States, the
Of 71 women identified with HDFN, 33 were administered implicated antigens were HPA-1a (79%), HPA-5b (9%),
phenobarbital. Both the exchange transfusion rate (9% vs HPA-1b (4%), HPA-3a (2%), and others (6%).145 Antibody
52%, p < 0.01) and neonatal mortality rate (0% vs 24%) identification is performed using an enzyme-linked
were reduced by the administration of phenobarbital. After immunosorbent assay or monoclonal antibody immobi-
adjusting for gestational age at delivery and neonatal peak lization of platelet antigens assay. PCR sequence-specific
total bilirubin, administration of phenobarbital decreased primer assays capable of testing for the 15 HPA allelic vari-
the relative risk of exchange transfusion (risk ratio, 0.23; ants has been developed and allows for rapid and accu-
95% CI, 0.06–0.76). Halpin and colleagues also found prom- rate HPA genotyping.146 Alternatively, phenotyping can
ising results in an earlier report on phenobarbital.136 These be performed using HPA monoclonal antibodies. Paternal
preliminary studies suggest that predelivery phenobarbital genotyping is performed to determine if the chance of
is an effective method to prevent exchange transfusion in recurrence in a subsequent pregnancy is 100% (HPA-1a/1a)
documented HDFN. or 50% (HPA-1a/1b).
TRANSFUSION MEDICINE
Management maternal plasma must be tested for transmissible diseases,
as required for all allogeneic units. The platelets are almost
Maternal Administration of Intravenous always required for emergency transfusion before the results
Immunoglobulin and Corticosteroids of the transmissible disease testing are available. Where
The role of maternal IVIG, compared to IVIG plus dexameth- maternal platelets are unavailable or unsuitable, matched
asone, in the antenatal management of neonatal alloimmune allogeneic platelets should be obtained. This can be compli-
thrombocytopenia was evaluated in a controlled trial of 54 cated by the delay in obtaining compatible platelets due to
affected pregnancies. No additional benefit was seen with the the time required for laboratory confirmation of the diag-
combination over IVIG alone. Overall, there was a mean plate- nosis, the identification of compatible donors, the platelet
let increase from the first to the second fetal blood sampling apheresis collection, the donor transmissible disease testing,
of 36,000/μL and from the first fetal blood sampling to birth and the transport of the product to the hospital transfusion
of 69,000/μL. A total of 62% to 85% of fetuses responded.147 service. Several blood suppliers have evaluated the effec-
In another report, 18 women who had previously delivered tiveness of platelet concentrates collected from a panel of
infants with severe alloimmune thrombocytopenia were donors to maintain an “off-the-shelf ” supply of HPA-1a/5b
treated with weekly 1 g/kg infusions of IVIG from the diagno- negative platelets, which would be appropriate for 90% of
sis of fetal thrombocytopenia until birth. Only three treated affected neonates.152 The estimated cost of identifying one
fetuses had platelet counts of less than 30,000/μL, compared eligible HPA-1a/5b negative donor was £8,000, because sev-
with 16 of 20 untreated siblings.148 If IVIG treatment fails to eral thousand donors must be screened to find an eligible
give a response, high-dose steroids (60 mg prednisone/day) donor. It was also suggested that HPA-1a/5b negative plate-
are often added, although their efficacy is uncertain. Although lets be utilized for all neonates, thereby ensuring that most
the above data are limited, antenatal IVIG appears to be a infants with neonatal alloimmune thrombocytopenia would
promising therapy. Maternal IVIG is usually commenced be managed correctly from a platelet transfusion perspective
when the fetal platelet count by cordocentesis is found to be and fewer units would be outdated. The ability to genotype
less than 100,000/μL. Infants are followed for response by donors on a large scale for HPA-1a/5b may be one means
both cordocentesis and ultrasonography. of generating an extensive list of eligible donors to maintain
in-date stock of antigen-negative platelets by targeting these
Intrauterine Platelet Transfusions
donors to enroll as apheresis platelet donors.
In the presence of persistent fetal thrombocytopenia, despite
maternal IVIG therapy, frequent fetal blood sampling is Neonatal Intravenous Immunoglobulin
required to monitor the severity of thrombocytopenia and to The role of IVIG to treat thrombocytopenia after delivery has
provide intrauterine platelet transfusion support for platelet been less well-characterized but is often employed.153 IVIG
counts of less than 50,0000 to 100,000/μL. Where required is usually administered in combination with HPA-matched
for severe thrombocytopenia, the first cordocentesis is usu- platelets.
III ally performed at 22 to 24 weeks’ gestation and is repeated
at weekly intervals.149 Antigen-negative, irradiated maternal
Prevention
506 platelets are recommended. The platelets should be irradi-
ated to eliminate the risk of transfusion-transmitted GVHD. Testing all pregnancies to identify women at risk of deliver-
The platelets should be plasma-reduced to minimize the ing children with neonatal alloimmune thrombocytopenia
passive transfer of the maternal alloantibodies and resus- secondary to anti-HPA-1a has been studied.154 The low rate
pended in saline or compatible plasma. Pretransfusion and of intracranial hemorrhage in affected pregnancies, the high
post-transfusion platelets counts are performed. The median complication rate from cordocentesis, the lack of certainty
drop in the platelet count after transfusion is approximately on the best course of treatment, and our inability to screen all
24,000/μL per day; thus, the post-transfusion platelet count pregnancies for all platelet alloantibodies have prevented the
can be used to estimate the required interval between proce- implementation of widespread population screening for this
dures.150 The incidence of fetal loss at the time of each cordo- condition.155 It is critical that all mothers delivering neonates
centesis and intrauterine platelet transfusion is estimated at with unexplained thrombocytopenia should be screened for
0.6% to 1.2%.149,151 Early delivery after achievement of fetal antibodies implicated in neonatal alloimmune thrombocy-
lung maturity is recommended. topenia to ensure that future pregnancies are appropriately
managed.
Neonatal Platelet Transfusions
Often when a neonate with unexpected thrombocytopenia
is suspected of having neonatal alloimmune thrombocyto- REFERENCES
penia, the management is problematic due to difficulty in
obtaining compatible platelets for infusion. Where possible, 1. Waterstone M, Bewley S, Wolfe C. Incidence and predictors of severe
plasma-depleted, irradiated maternal platelets are recom- obstetric morbidity: case-control study. BMJ 2001;322:1089–1094.
2. Andolina K, Daly S, Roberts N, et al. Objective measurement of blood
mended because compatibility with maternal antibodies is loss at delivery: Is it more than a guess? Am J Obstet Gynecol 1999;
assured. Maternal platelets can be collected from a unit of 180:S69.
whole blood or by platelet apheresis. Where apheresis collec- 3. Ueland K. Maternal cardiovascular dynamics. VII. Intrapartum blood
tion is not immediately available for the collection of mater- volume changes. Am J Obstet Gynecol 1976;126:671–677.
nal platelets, a unit of whole blood can be collected. Once 4. Panchal S, Arria A, Labhsetwar S. Maternal morbidity during hospital
admission for delivery: a retrospective analysis using a state-main-
obtained, the unit can be separated into its components tained database. Anesth Analg 2001;93:134–141.
in the hospital transfusion service and the autologous red 5. Mousa HA, Walkinshaw S. Major postpartum haemorrhage. Curr
cells can be transfused back to the mother. In addition, the Opin Obstet Gynecol 2001;13:595–603.
6. Goffinet F. Hémorragies de la délivrance. Gynecol Obstet Fertil 2000;
The principles of transfusion support for older children have greater abilities to compensate for anemia and may be
and adolescents are similar to those for adults, but infants transfused at lower Hb or hematocrit (Hct) levels.
and younger children have many special needs. Therefore, In the perioperative period or after resuscitation for
administration of each component to each of these two age trauma, it is unnecessary for most children to maintain Hb
groups is discussed separately. Because many issues deal- levels greater than 8.0 g/dL or Hct greater than 24%, a level
ing with transfusion therapy in adults, as discussed in other frequently desired for adults. There should be a compelling
chapters, apply to older children, the bulk of information in reason to administer any postoperative RBC transfusion,
this chapter deals with transfusion in infants and younger because most children (without continued bleeding) can
children. The following types of components for transfu- quickly restore their RBC volume if given iron and adequate
sion will be discussed as they apply to neonates and pediatric nutritional therapy.3 As is true for adults, the most important
patients: red blood cells (RBCs), platelets (PLTs), and neu- measures in the treatment of acute hemorrhage occurring
trophils (granulocytes), fresh frozen plasma (FFP), and cryo- with surgery or injury are first to control the hemorrhage
precipitate. General guidelines and recommendations are and restore blood volume and tissue perfusion with crystal-
given for pediatric blood component transfusions. However, loid and/or colloid solutions. Then, if the estimated blood
it is important that they be adapted to fit local standards of loss is greater than 25% of the circulating blood volume (i.e.,
practice. In particular, terms used to describe clinical condi- >17 mL/kg body weight) and the patient’s condition remains
tions, such as severe and symptomatic, must be defined by unstable, RBC transfusions may be indicated.
III local physicians. In acutely ill children with severe cardiac or pulmonary
disease, particularly in those who require assisted ventilation,
510 it is common practice to maintain the Hb level close to the
RED BLOOD CELL TRANSFUSIONS normal range at a level of 100 to 110 g/L (10 to 11 g/dL) or
Hct at 30% to 33%. Although this practice seems logical, its
Older Children and Adolescents efficacy has not been documented by controlled scientific
studies in children, and it has been challenged.3 The vari-
Red blood cells, the most frequently transfused blood com- ability of practices in critical care settings was demonstrated
ponent, are given to increase the oxygen-carrying capacity of by a survey administered to Canadian, French, Belgium, and
the blood, the goal being to maintain satisfactory tissue oxy- Swiss pediatric intensivists regarding transfusion practices
genation. Guidelines for RBC transfusions in older children in a tertiary care pediatric setting. They were queried about
and adolescents are similar to those for adults (Table 37–1).1 patients with bronchiolitis, septic shock, trauma, and about
However, transfusions may be given more stringently to chil- postoperative care of a patient with tetralogy of Fallot. The
dren, because normal hemoglobin (Hb) levels are lower in pretransfusion Hb transfusion “trigger” was 7 to 13 g/dL for
healthy children than in adults, and most children do not the majority of scenarios.4
have the concomitant cardiorespiratory and vascular diseases Randomized prospective studies are needed to define
that develop with aging in adults, who then may require more optimal pediatric practices because liberal RBC transfusion
aggressive RBC transfusions.2 Therefore, children generally practices in critically ill adults have been reported to have
detrimental effects.5 A similar harmful effect has been pro-
posed in a retrospective cohort study, including 240 critically
*
This work was supported in part by National Institutes of ill children (131 transfused and 109 nontransfused; mean
Health Program Project Grant P01 HL46925. Transfusions of blood age, 5.5 years) with mean initial Hb levels of 7.4 ± 1.4 g/dL,
components are mandatory for modern management in many premature
infants, children with cancer and hematologic disorders, recipients of
where the number of days of oxygen use, mechanical ventila-
hematopoietic progenitor cell transplants and organ allografts, and tion, vasoactive agents used, and pediatric intensive care unit
children undergoing various surgical procedures. Although transfusions and hospital lengths of stay were all significantly increased
can be lifesaving, they are not without risks. Accordingly, they should in association with RBC transfusion.6 Hospital mortality
be given only when true benefits are likely—for example, to correct a in the transfused group was 6.9% versus 0.9% (p = 0.43) in
deficiency or functional defect of a blood component that has caused
or threatens to cause a clinically significant problem. Because of the the nontransfused group. However, mortality could not be
extended life span of children after transfusions, it is critical to avoid adequately evaluated due to the limited number of deaths
post-transfusion complications that may lead to progressive morbidity and observed in the study. Overall, the investigators concluded
mortality and to considerable expense over the years (e.g., hepatitis). that RBC transfusions are associated with an increase in
TRANSFUSION OF NEONATES AND PEDIATRIC PATIENTS
Table 37–1 Authors' Guidelines for Transfusing Children and Adolescents*
*
Words in italics must be defined according to local practices.
Hct, hematocrit.
resource utilization in this patient population. Importantly, birth weights lower than 1.0 kg.7–9 Because this postnatal
due to the study’s retrospective nature a true cause-and-effect drop in Hb level in full-term infants is well tolerated, it is
relationship between transfusions and adverse outcomes commonly referred to as the physiologic anemia of infancy.
could not be proven. Obviously, a randomized, controlled However, the pronounced decline in Hb concentration that
clinical trial is essential to help clinicians make more evi- occurs in many extremely preterm infants is associated with
dence-based decisions regarding when to transfuse RBCs to abnormal clinical signs and the need for RBC transfusions.
critically ill children.6 Therefore, the anemia of prematurity is not accepted to be a
With anemias that develop slowly, the decision to trans- normal, benign event.10,11
fuse RBCs should not be based solely on the blood Hb con- Many interacting physiologic factors are responsible for
centration, because children with chronic anemias may be the anemia of prematurity. A key reason that the Hb nadir
asymptomatic despite very low Hb levels. Children with iron is lower in preterm than in full-term infants is the former
deficiency anemia, for example, often are treated successfully group’s diminished plasma EPO level in response to ane- 37
with oral iron alone, even at Hb levels lower than 5.0 g/dL. mia.9,12–15 Although anemia provokes EPO production in pre-
Factors other than Hb concentration that must be considered mature infants, the plasma levels achieved in anemic infants, 511
in the decision to transfuse RBCs include the patient’s symp- at any given Hct, are lower than those observed in comparably
toms, signs, and functional capacities; the presence or absence anemic older persons.14 Erythroid progenitor cells in blood16
of cardiorespiratory and central nervous system disease; the and bone marrow17 of preterm infants are quite responsive to
cause and anticipated course of the underlying anemia; and EPO in vitro—a finding suggesting that inadequate produc-
use of alternative therapies such as iron or recombinant tion of EPO is the major cause of physiologic anemia, not
human erythropoietin (EPO) therapy, the latter of which has marrow unresponsiveness.
been demonstrated to reduce the need for RBC transfusions The mechanisms responsible for the diminished EPO
and to improve the overall condition of children with chronic output by preterm neonates are only partially defined. One
renal insufficiency. In anemias that are likely to be permanent factor is that the primary site of EPO production in preterm
(e.g., thalassemia, hemoglobinopathies), the effects of anemia infants is the liver rather than the kidneys.18,19 Because the
on growth and development (which might be ameliorated liver is less sensitive to anemia and tissue hypoxia, a relatively
by RBC transfusions) must be balanced against the potential sluggish EPO response to the falling Hct occurs. The timing
toxicities of repeated transfusions (e.g., iron overload). of the switch from liver to kidney is set at conception and is
not accelerated by preterm birth. Viewed from a teleologic
perspective, decreased hepatic production of EPO under
Infants and Younger Children in utero conditions of tissue hypoxia may be an advantage
for the fetus. If this were not the case, normal levels of fetal
Pathophysiology of the Anemia of Prematurity
hypoxia could trigger high levels of EPO and produce marked
All infants experience a decline in circulating RBC volume erythrocytosis and consequent hyperviscosity in utero. After
during the first weeks of life. This decline results both from birth, however, diminished EPO responsiveness to tissue
physiologic factors and, in sick preterm infants, from phle- hypoxia is disadvantageous and leads to anemia because of
botomy blood losses for laboratory monitoring. In healthy impaired compensation for the falling Hct.
full-term infants, the nadir Hb value rarely is lower than Diminished EPO production cannot entirely explain low
9.0 g/dL at age 10 to 12 weeks. The decline is more rapid plasma EPO levels in preterm infants, however. Extraordinarily
(i.e., nadir at age 4 to 6 weeks) and the blood Hb concen- high plasma levels of EPO were reported in some fetuses of
tration falls to lower levels in infants born prematurely— postconceptional age, comparable to those of neonates treated
to approximately 8.0 g/dL in infants with birth weights of in intensive care settings,20,21 and macrophages from human
1.0 to 1.5 kg and to approximately 7.0 g/dL in infants with cord blood were found to produce normal quantities of EPO
TRANSFUSION MEDICINE messenger RNA and protein.22 These studies documented Most RBC transfusions given to infants are small in vol-
intact synthetic capability, at least under some circumstances. ume (10 to 15 mL/kg) and are repeated frequently to replace
Therefore, additional mechanisms are likely to contribute to blood drawn for laboratory studies. There is no proven benefit
diminished plasma EPO levels. For example, plasma EPO lev- to routine replacement of phlebotomy blood losses “milliliter
els undoubtedly are influenced by metabolism (clearance) as for milliliter.” Instead, RBCs should be transfused to maintain
well as by production. Data obtained in human infants23,24 a Hct level deemed appropriate for the clinical condition of
and in neonatal monkeys25 demonstrate that low plasma EPO the infant. In neonates with severe respiratory disease, such
levels may result from increased plasma clearance and volume as those requiring high volumes of oxygen with ventilator
of distribution and from shorter fractional elimination and support, it is customary to maintain the Hct at greater than
mean residence times for EPO in neonates compared with 40% (Hb concentration > 13.0 g/dL), particularly if blood is
adults. Therefore, accelerated catabolism may contribute to being drawn frequently for testing. This practice is based on
the low plasma levels—with the low plasma EPO in infants the belief that transfused donor RBCs containing adult Hb
possibly representing the combined effects of decreased will provide optimal oxygen delivery throughout the period
synthesis and increased metabolism. of diminished pulmonary function that requires mechani-
Phlebotomy blood losses play a key role in the anemia of cal ventilation. Consistent with this rationale for ensuring
prematurity. The modern practice of neonatology requires optimal oxygen delivery in neonates with pulmonary failure,
critically ill neonates to be monitored closely with serial it seems logical—although unproven by controlled stud-
laboratory studies such as blood gases, electrolytes, blood ies—to maintain the Hct at greater than 40% in infants with
counts, and cultures. Small preterm infants are the most congenital heart disease that is severe enough to cause either
critically ill, require the most frequent blood sampling, and cyanosis or congestive heart failure.
suffer the greatest proportional loss of RBCs because their Definitive studies are not available to establish the opti-
circulating RBC volumes are the smallest. Promising “in- mal Hb level for infants facing major surgery. However, it
line” devices that withdraw blood, measure multiple analytes, seems reasonable to maintain the Hb at greater than 10.0 g/
and then reinfuse the sampled blood are being investigated.26 dL (Hct > 30%) because of the limited ability of the infant
However, until these devices are proven effective and safe for heart, lungs, and vasculature to compensate for anemia.
infants, the replacement of blood losses due to phlebotomy Additional factors include the inferior off-loading of oxygen
will remain a critical factor responsible for multiple RBC by neonatal RBCs that results from the diminished interac-
transfusions in critically ill neonates. tion between fetal Hb and 2,3-diphosphoglycerate and from
the developmental impairment of neonatal renal, hepatic,
Recommendations for RBC Transfusions
and neurologic function. This transfusion guideline is sim-
during Infancy
ply a recommendation for perioperative management, not
Guidelines for transfusing RBCs to neonates are contro- a firm indication, and it should be applied with flexibility
versial, and practices vary.7,27–29 The lack of a consistent to individual infants who are facing surgical procedures of
III approach stems from incomplete knowledge of the cellular varying complexity.
and molecular biology of erythropoiesis during the perinatal The clinical indications for RBC transfusions in preterm
512 period as well as incomplete understanding of the infant’s infants who are not critically ill but nonetheless develop
compensation for anemia and the physiologic response to moderate anemia (Hct < 24% or blood Hb concentration
RBC transfusions. Generally, RBC transfusions are given to < 8.0 g/dL) are extremely variable.7,29 In general, it was
maintain the level of Hb or Hct believed to be most desir- accepted that infants who are clinically stable despite mod-
able for each neonate’s clinical condition. Broad guidelines est anemia do not require RBC transfusions unless they
for RBC transfusions during early infancy are listed in Table exhibit significant problems that either are ascribed to the
37–2.30 These guidelines are very general, and it is important presence of anemia or are predicted to be corrected by RBC
that terms used to describe clinical conditions, such as severe transfusions. To illustrate, proponents of RBC transfusions
and symptomatic, be defined to fit local practices. to treat disturbances of cardiopulmonary rhythms believe
*
Words in italics must be defined according to local practices.
Hct, hematocrit.
TRANSFUSION OF NEONATES AND PEDIATRIC PATIENTS
that a low Hct contributes to tachypnea, dyspnea, apnea, Another physiologic factor to be considered in the trans-
and tachycardia or bradycardia because of decreased oxygen fusion decision is the use of circulating RBC volume rather
delivery to the respiratory center of the brain. If this theory than blood Hct or Hb level.11,34,37 Although circulating RBC
is true, transfusions of RBCs should decrease the number volume is a potentially useful index of the blood’s oxy-
of apneic spells by improving oxygen delivery to the cen- gen-carrying capacity, it cannot be predicted accurately by
tral nervous system. However, results of clinical studies have measurement of the Hct in infants.38 Low circulating RBC
been contradictory.7,29 volume identifies, better than Hb or Hct, those infants who
A recent study by Bell and colleagues provides evidence will respond to transfusion with a decrease in cardiac out-
to the contrary.31 In this study, 100 preterm infants, weighing put.39 At present, circulating RBC volume measurements are
from 500 to 1300 grams, were randomized to either a liberal not widely available. However, promising techniques using
or restrictive RBC transfusion regimen (i.e., a relatively low nonradioactive biotin to tag RBCs have been adapted for
or high pretransfusion Hct level, respectively). The investiga- infant studies. Strauss and coworkers were able to demon-
tors found that the mean number of RBC transfusions was strate with this technique that the post-transfusion recov-
higher in the liberal group than the restrictive group; how- ery and in vivo survival of donor RBCs, stored for up to 42
ever, because a “single donor/unit program” was used, the days, confirmed earlier studies that defined the efficacy and
number of donor exposures was not statistically significantly safety of stored allogeneic RBCs for small-volume transfu-
different. Furthermore, the most surprising finding was that sions to neonates. Specifically, no significant differences were
those infants in the more restrictive group were more likely observed between allogeneic RBCs stored on days 1 to 21
to have severe grades of intraparenchymal brain hemorrhage compared to days 22 to 42.39–41
or periventricular leukomalacia as well as having increased
Selecting an RBC Product to Transfuse an Infant
frequency of apnea. The authors suggest that more restrictive
transfusion practices may be detrimental to preterm infants The RBC products usually chosen for small-volume trans-
and hypothesize that decreased oxygen delivery to brain tissue fusions given to infants are RBCs suspended either in
may be the pathophysiology of the brain injury and increased citrate-phosphate-dextrose-adenine solution (CPDA) or in
frequency of apnea—a speculation supported by another extended storage media (AS-1, AS-3, AS-5) at a Hct ranging
study in which high cerebral fractional oxygen extraction was from 55% to 70%. Some centers prefer to centrifuge RBC
found in infants with hemorrhagic parenchymal infarction as aliquots before transfusion, to prepare packed RBCs at a Hct
a manifestation of possible low cerebral blood flow.31,32 Very of 80% to 90%. Most RBC transfusions are infused slowly
importantly, a multicenter study of similar design with only over 2 to 4 hours at a dose of about 15 mL/kg body weight.
preliminary reports to date33 found results contradictory to Because of the small quantity of RBC preservative fluid
Bell and colleagues. In the randomized study, 451 preterm infused and the slow rate of transfusion, the type of anti-
infants were transfused according to either relatively low or coagulant/preservative medium selected is believed not to
high pretransfusion blood Hb concentrations. In contrast to pose risks for the majority of premature infants given small-
the findings of Bell and colleagues, no difference in morbid- volume transfusions.42 Accordingly, the traditional use of rel- 37
ity and mortality were found—favoring acceptance of rela- atively fresh RBCs (<7 days of storage) has been challenged,
tively low pretransfusion blood Hb or Hct levels in clinical and it has been shown by many investigators43–45 that donor 513
practice. Because the experimental design and outcomes were exposure of multiply transfused infants can be diminished
not identical in the two studies, there is no definitive recom- safely by the exclusive use of a dedicated unit of stored RBCs
mendation possible for transfusion practices at this time. (i.e., 21 to 42 days after collection) for each infant.
However, the potential for harm due to “undertransfusion,” Neonatologists who object to stored RBCs and continue
if very low pretransfusion Hb or Hct levels are accepted, must to insist on transfusing infants with fresh RBCs generally
be considered when transfusion decisions are made. raise three objections: the rise in plasma potassium (K+) and
Generally in practice, the decision whether to transfuse the drop in RBC 2,3-diphosphoglycerate that occur dur-
RBCs is based on the desire to maintain the Hct at a level ing extended storage and the possible dangers of additives
judged to be most beneficial for the infant’s clinical condi- present in extended-storage media. After 42 days of stor-
tion. Investigators who believe that this “clinical” approach is age, plasma K+ levels in RBC units approximate 50 mEq/L
too imprecise have suggested the use of “physiologic” crite- (0.05 mEq/mL), a value that, at first glance, seems alarmingly
ria for transfusions, such as RBC mass,34 available oxygen,35 high. By simple calculations, however, the dose of bioavail-
mixed venous oxygen saturation, or measures of oxygen able K+ transfused (i.e., ionic K+ in the extracellular fluid) is
delivery and utilization,36 to develop guidelines for transfu- shown to be very small. An infant weighing 1 kg who is given
sion decisions. In one study of 10 infants with severe (oxygen- a 15 mL/kg transfusion of packed RBCs (Hct, 80%) will
dependent) bronchopulmonary dysplasia, improvement in receive 3 mL of extracellular fluid, containing only 0.15 mEq
physiologic end points (increased systemic oxygen transport of K+, which will be infused slowly. Even if RBCs are not
and decreased oxygen use) was shown to be a consequence of packed but are removed from the blood bag and directly
small-volume RBC transfusions.36 However, these promising infused at a Hct of 60%, the K+ dose will be only 0.3 mEq.
but technically demanding methods are, at present, difficult These doses are quite small compared with the usual daily
to apply in the day-to-day practice of neonatology, and stud- K+ requirement of 2 to 3 mEq/kg. However, this rationale
ies conducted directly in infants are needed. Application of does not apply to large-volume transfusions (>25 mL/kg), in
data obtained from studies of animals and adult humans that which larger doses of K+ may be harmful, especially if the
correlate tissue oxygenation with the clinical effects of ane- infusion is rapid. Cardiac surgeons and other pediatric sur-
mia and the need for RBC transfusions is confounded by the geons who may transfuse large RBC volumes expectedly or
differences between infants and adults in Hb oxygen affinity, unexpectedly need to be especially cognizant of this point.
ability to increase cardiac output, and regional patterns of As for the second objection, 2,3-diphosphoglycerate
blood flow. is totally depleted from RBCs by 21 days of storage, and
TRANSFUSION MEDICINE
Table 37–3 Formulation of Anticoagulant-Preservative Solutions Present in Blood Collection Sets
*
Approximately 450 mL of donor blood is drawn into 63 mL of CPDA. A unit of red blood cells (hematocrit, approximately 70%) is prepared
by centrifugation and removal of most plasma.
†
When AS-1 or AS-5 is used, 450 mL of donor blood is first drawn into 63 mL of CPD, which is identical to CPDA except that it contains
1610 mg of dextrose per 63 mL and has no adenine. When AS-3 is used, donor blood is drawn into CP2D, which is identical to CPD except that
it contains double the amount of dextrose. After centrifugation and removal of almost all plasma, red blood cells are resuspended in 100 mL of
the additive solution (AS-1, AS-3, or AS-5) at a hematocrit of approximately 55% to 60%.
AS-1, AS-3, and AS-5, extended storage media; CPDA, citrate-phosphate-dextrose-adenine solution.
Data from Luban NLC, Strauss RG, Hume HA. Commentary on the safety of red blood cells preserved in extended storage media for
neonatal transfusions. Transfusion 1990;30:229.
this is reflected by a decrease in the oxygen half-saturation transfusions cannot be overemphasized. Therefore, these
pressure (P50) from about 27 mmHg in fresh blood to 18- same rules do not apply to massive transfusion situations,
mmHg at the time of outdate. The last value of older trans- which many pediatric surgeons and neonatologists may
fused RBCs corresponds to the “physiologic” P50 obtained from encounter knowingly or unknowingly and for which they
the RBCs of many normal preterm infants at birth, reflecting need to be prepared accordingly. This information is espe-
the relatively high affinity for oxygen normally exhibited by cially important for surgeons and neonatologists who are
infant RBCs. Therefore, the P50 of older transfused RBCs is recommending that the parents, relatives, or friends of their
no worse than that of RBCs produced endogenously by the patients perform directed RBC donation. Communication
infant’s own bone marrow. Moreover, these older adult RBCs between the physicians and blood bank is encouraged for
provide a benefit to the infant because the 2,3-diphosphoglyc- many reasons: to confirm the child’s blood type so type-
erate and the P50 of transfused RBCs (but not of endogenous compatible RBCs can be donated; to help with the collec-
infant RBCs) increase rapidly after transfusion. tion center’s choice of anticoagulant-preservative solution
Regarding the third objection, the quantity of addi- into which the units will be drawn (i.e., CPDA and/or AS
III tives present in RBCs stored in extended-storage media units), which might differ if a massive transfusion situa-
is believed not to be dangerous to neonates given small- tion seems likely; and to coordinate timing of the surgery
514 volume transfusions (≤15 mL/kg).42 A comparison of CPDA so that tested units are in hospital inventory prior to the
with three types of extended-storage media is presented in surgical date and time.
Table 37–3. Regardless of the type of suspending solution,
Recombinant Erythropoietin in
the quantity of additives is quite small in the clinical set-
the Anemia of Prematurity
ting, in which infants are given small-volume transfusions
of RBCs transfused over 2 to 4 hours, and it is far lower Recognition of the low plasma EPO levels in preterm infants
than doses believed to be toxic (Table 37–4). Importantly, provides a rational basis for the use of recombinant human
the efficacy and safety of these theoretical calculations have EPO as therapy for the anemia of prematurity. More than
been confirmed by clinical experience. In addition, many 20 controlled trials have tested several doses and treatment
investigators have reported the successful transfusion of schedules in preterm infants, and results are mixed, making
stored, rather than fresh, RBCs for small-volume trans- consensus impossible on the optimal role of recombinant
fusions in infants.43–46 The small-volume nature of these human EPO treatment in the anemia of prematurity.7,47,48
Table 37–4 Constituents Infused (mg/kg) in 10 mL/kg Red Blood Cells (Hematocrit, 60%)
*
Actual toxic dose is difficult to predict accurately because the infusion rate usually is slow, permitting metabolism and distribution from
blood into extravascular sites, and dextrose, adenine, and phosphate enter red blood cells and are somewhat sequestered. Potential toxic doses
are based on Luban, Strauss, and Hume (1990).32
AS-1 and AS-3, extended storage media; CPDA, citrate-phosphate-dextrose-adenine solution.
Data from Luban NLC, Strauss RG, Hume HA. Commentary on the safety of red blood cells preserved in extended storage media for
neonatal transfusions. Transfusion 1990;30:229.
TRANSFUSION OF NEONATES AND PEDIATRIC PATIENTS
Unquestionably, proper doses of recombinant human should be given to patients with platelet counts lower than
EPO and iron effectively stimulate erythropoiesis in preterm 50 × 109/L due to marrow failure if they are bleeding or are
infants, as evidenced by increased marrow erythroid activ- scheduled for an invasive procedure. Studies of patients
ity and blood reticulocyte counts. However, the efficacy of with thrombocytopenia caused by poor marrow produc-
recombinant human EPO in substantially diminishing the tion indicate that spontaneous bleeding increases markedly
number of RBC transfusions—the major goal for which it when platelet levels fall to less than 20 × 109/L, particularly
is prescribed—has not been convincingly demonstrated for in patients who are ill with infection, anemia, or dysfunc-
all groups of preterm infants.48 In many trials, the subjects tion of the liver, kidneys, or lungs. For this reason, many
were relatively large preterm infants and those in stable clini- pediatricians recommend prophylactic platelet transfusions
cal condition; such infants currently receive few RBC trans- to maintain the platelet count at greater than 20 ×109/L in
fusions when given only standard supportive care (i.e., not children with thrombocytopenia due to bone marrow fail-
given recombinant human EPO).27,49 Currently, even with- ure. This threshold has been challenged, and some favor a
out use of recombinant human EPO, fewer than 50% of platelet transfusion trigger of 5 to 10 × 109/L for patients
infants with birth weights greater than 1.0 kg receive RBC with uncomplicated conditions. Many pediatric hematol-
transfusions. Almost all infants weighing less than 1.0 kg at ogy/oncology and stem cell transplantation physicians have
birth are given RBCs, and most transfusions are given during lowered their prophylactic platelet transfusion trigger to
the first 3 to 4 weeks of life.49 10 × 109/L, extrapolating from the adult acute myeloid leu-
To illustrate the difficulty of avoiding RBC transfusions, kemia studies by Rebulla and colleagues in 1997 and Wandt
a multicenter, randomized North American trial, in which and coworkers in 1998.53,54 These prospective clinical trials,
infants received either recombinant human EPO or placebo among others, demonstrated that nonbleeding stable throm-
during a 6-week study period, reported a statistically signifi- bocytopenic patients without coexisting symptoms can be
cant difference but only modest success.50 Although signifi- managed safely with a more restrictive platelet transfusion
cantly fewer RBC transfusions were given to infants treated trigger, 10 × 109/L, and fewer platelet transfusions were given.
with recombinant human EPO during the study phase (1.1 However, those with fever, active bleeding, and/or a coagula-
transfusion vs 1.6 for placebo), all infants required multiple tion disorder had a higher trigger of at least 20 × 109/L.
transfusions during the 3-week prestudy phase. Therefore, Qualitative platelet disorders may be inherited or acquired
recombinant human EPO exerted only a modest effect on (e.g., in advanced hepatic or renal insufficiency, after cardio-
total RBC transfusions given throughout the entire study pulmonary bypass procedures). In such patients, platelet
(4.4 for recombinant human EPO vs 5.3 for placebo) and transfusions are justified only if significant bleeding occurs.
did not resolve the problem of severe neonatal anemia.50 Because individuals with platelet dysfunction may have
Physicians wishing to prescribe recombinant human intermittent bleeding episodes throughout their life necessi-
EPO are faced with a dilemma. The relatively large or stable tating repeated transfusions, which may lead to alloimmuni-
preterm infants who respond best to recombinant human zation and refractoriness, prophylactic platelet transfusions
EPO plus iron are given relatively few RBC transfusions and, are rarely justified unless an invasive procedure is planned. 37
accordingly, have little need for recombinant human EPO In such cases, a bleeding time greater than twice the upper
to avoid transfusions. Extremely preterm infants, who are limit of laboratory normal may be taken as diagnostic evi- 515
sick and have the greatest need for RBC transfusions shortly dence that platelet dysfunction exists, but this test is poorly
after birth, have not consistently responded to recombi- predictive of hemorrhagic risk or the need to transfuse plate-
nant human EPO plus iron, again questioning the efficacy lets. The bleeding time has been supplanted by other platelet
of recombinant human EPO therapy.47 However, extremely function assays and is rarely performed anymore, especially
preterm infants are being evaluated in therapeutic trials of in children. In these patients, alternative therapies, par-
recombinant human EPO and iron, both given intravenously ticularly subcutaneous or intranasal desmopressin acetate,
shortly after birth.51,52 Although preliminary review of these should be considered to avoid platelet transfusions.
promising studies suggests success in avoiding transfusions
early in life, the data are limited and are insufficient to clearly
establish efficacy or to detect potential toxicity. Therefore, Infants and Younger Children
firm guidelines for the use of recombinant human EPO in
the treatment of the anemia of prematurity cannot be offered Pathophysiology of Neonatal
at this time. Thrombocytopenia
Blood platelet counts of 150 × 109/L or greater are present
in normal fetuses (=17 weeks of gestation) and neonates.
PLATELET TRANSFUSIONS Lower platelet counts indicate potential problems, and
preterm infants exhibit thrombocytopenia commonly.55,56
In one neonatal intensive care unit, 22% of infants had blood
Older Children and Adolescents
platelet counts lower than 150 × 109/L during hospitaliza-
Guidelines for platelet support of children and adolescents tion.55 Although multiple pathogenic mechanisms prob-
with quantitative and qualitative platelet disorders are simi- ably are involved in these sick neonates, a predominant one
lar to those for adults (see Table 37–1), in whom the risk of is accelerated platelet destruction, as shown by shortened
life-threatening bleeding that occurs after injury or sponta- platelet survival time, increased platelet-associated immu-
neously can be related to the severity of thrombocytopenia noglobulin G, increased platelet volume, a normal number
(when low blood platelet counts are caused by diminished of megakaryocytes, and an inadequate increment in blood
marrow production). Thrombocytopenia caused by acceler- platelet values after platelet transfusion.55,57 Another major
ated turnover (e.g., immune thrombocytopenia) usually is mechanism contributing to neonatal thrombocytopenia is
not treated with platelet transfusions. Platelet transfusions diminished platelet production, as evidenced by decreased
TRANSFUSION MEDICINE
numbers of clonogenic megakaryocyte progenitors58,59 Regarding the second circumstance, the need to main-
and relatively low levels of thrombopoietin59,60 in response tain a completely normal platelet count (≥150 × 109/L) or
to thrombocytopenia, when compared with children and even higher in preterm infants without bleeding is unproven.
adults. Similar to the situation with EPO and the anemia of Intracranial hemorrhage occurs commonly in sick preterm
prematurity, thrombopoietin is produced by thrombocyto- infants, and, although the etiologic role of thrombocyto-
penic preterm infants, but at relatively low levels. Controlled penia and the therapeutic benefit of platelet transfusions
clinical trials are needed to determine the possible role and have not been conclusively established in this disorder, it
potential toxicity of recombinant thrombopoietin therapy seems logical to presume that thrombocytopenia is a risk
in infants. factor.62 However, a randomized trial designed to address
Blood platelet counts lower than 100 × 109/L pose sig- this issue—in which transfusion of platelets whenever the
nificant clinical risks for premature neonates. In one study, platelet count fell to less than 150 × 109/L so as to main-
infants with birth weights lower than 1.5 kg and blood tain the average platelet count at greater than 200 × 109/L
platelet counts lower than 100 × 109/L were compared with was compared with transfusion of platelets only when the
nonthrombocytopenic control infants of similar size.56 The platelet count fell to less than 50 × 109/L—did not detect a
bleeding time was prolonged when platelet counts were difference in the incidence of intacranial hemorrhage (28%
lower than 100 × 109/L, and in many infants platelet dysfunc- vs 26%, respectively).61 Therefore, there is no documented
tion was suggested by bleeding times that were dispropor- benefit to transfusing “prophylactic platelets” to maintain a
tionately long for the degree of thrombocytopenia present. completely normal platelet count, compared with transfus-
Hemorrhage was more frequent in the thrombocytopenic ing “therapeutic platelets” in response to thrombocytopenia
infants, with the incidence of intracranial hemorrhage being when it actually occurs.
78% in those weighing less than 1.5 kg at birth, compared Currently, there are no alternatives to platelet transfu-
with 48% for nonthrombocytopenic infants of similar size. sions to treat thrombocytopenia in neonates. Recombinant
Moreover, the extent of hemorrhage and neurologic morbid- thrombopoietin (i.e., c-mpl ligand or megakaryocyte growth
ity was greater in the group of thrombocytopenic infants.56 and differentiation factor) and interleukin-11 are promis-
ing agents. However, neither is recommended for use during
Recommendations for Platelet
infancy, and both have potential toxicities that might pre-
Transfusions during Infancy
clude their use in sick preterm infants. Clearly, they must not
The use of prophylactic platelet transfusions in an attempt be prescribed at present, except in experimental settings.
to prevent bleeding in preterm neonates has been studied
systematically.61 However, no randomized clinical trials have Selecting a Platelet Product to Transfuse
been reported examining therapeutic platelet transfusions an Infant or Younger Child
in bleeding thrombocytopenic neonates. Therefore, basic The ideal goal of most platelet transfusions is to raise the
questions regarding the relative risks of different degrees of platelet count to greater than 50 × 109/L or, for sick pre-
III thrombocytopenia in various clinical settings during infancy term infants, to greater than 100 × 109/L. This goal can be
remain largely unanswered. However, it seems logical to achieved by the infusion of 5 to 10 mL/kg of standard (i.e.,
516 transfuse platelets to thrombocytopenic infants, and guide- unmodified) platelet concentrates, collected by centrifuga-
lines acceptable to many neonatologists are listed in Table tion of fresh units of whole blood or from a plateletpheresis
37–2. Two firm indications for neonatal platelet transfusions unit collected by automated plateletpheresis. The platelet
are to treat hemorrhage that has already occurred and to pre- dose should be transfused as rapidly as the overall condi-
vent hemorrhage from complicating an invasive procedure. tion permits, certainly within 2 hours. Routinely reducing
Little disagreement exists regarding the use of a blood plate- the volume of platelet concentrates for infants by addi-
let count lower than 50 × 109/L as a “transfusion trigger” in tional centrifugation steps is both unnecessary and unwise.
these instances. However, platelet transfusions are given to Transfusion of 10 mL/kg platelet concentrate provides
infants by some physicians to treat bleeding that occurs at approximately 10 × 109/L platelets. Assuming that the esti-
higher platelet counts (between 50 and 100 × 109/L) or to mated blood volume of an infant is 70 mL/kg body weight,
diminish the threat of intracranial hemorrhage in high-risk the platelet dose of 10 mL/kg will increase the platelet count
preterm infants whenever the platelet count is lower than by 143 × 109/L. This calculated increment is consistent with
100 × 109/L.56 the observed increment after this dose reported in clinical
Prophylactic platelet transfusions can be given under two studies.61 In general, 10 mL/kg is not an excessive transfusion
circumstances: to prevent bleeding when severe thrombocy- volume, provided that the intake of other intravenous flu-
topenia is present and poses a risk of spontaneous hemor- ids, medications, and nutrients is monitored and adjusted.
rhage and to maintain the presence of a normal platelet count It is desirable that the infant and the platelet donor be of
to prevent the infant from slipping into high-risk situations. the same ABO blood group, and it is important to minimize
Regarding the first circumstance, most agree that it is rea- repeated transfusions of group O platelets to group A or B
sonable to give platelets to any neonate whose blood platelet recipients, because large quantities of passive anti-A or anti-
count is lower than 20 × 109/L. There is broad acceptance that B can lead to hemolysis, resulting in severe morbidity and
spontaneous hemorrhage is a risk with platelet counts below mortality in children.63–66
this level. Also, severe thrombocytopenia occurs most com- Although proven methods exist to reduce the volume
monly in sick infants who, because of their illnesses, receive of platelet concentrates when truly warranted (i.e., when
medications that may further compromise platelet function. many transfusions are anticipated in which multiple doses
Because all of these factors are pronounced in extremely pre- of passive anti-A or anti-B might lead to hemolysis or when
term infants, some neonatologists favor prophylactic platelet there is failure to respond to 10 mL/kg of unmodified plate-
transfusion whenever the platelet count falls to less than 50 × let concentrate), additional processing (i.e., plasma volume
109/L, or even 100 × 109/L, in critically ill infants.56 reduction) should be performed with great care because of
TRANSFUSION OF NEONATES AND PEDIATRIC PATIENTS
probable platelet loss, clumping, and dysfunction caused by antibiotics, closer monitoring of antibiotic blood levels, and
the additional handling. use of intravenous immunoglobulin (IVIG), G-CSF, other
recombinant cytokines, and immune-modulating agents.
The role of granulocyte transfusion added to antibiotics
NEUTROPHIL TRANSFUSIONS for patients with severe neutropenia (<0.5 × 109/L) caused
by bone marrow failure is similar in adults and children
(see Table 37–1). Infected neutropenic patients usually
Older Children and Adolescents
respond to antibiotics alone, provided bone marrow func-
Several methodologic advances—in particular, the use of tion recovers early in infection. Because children with newly
recombinant granulocyte colony-stimulating factor (G- diagnosed leukemia respond rapidly to induction chemo-
CSF) to stimulate donors—have made it possible to collect therapy, only rarely are they candidates for granulocyte
extraordinarily large numbers of normal neutrophils (poly- transfusion. In contrast, infected children with sustained
morphonuclear neutrophils, PMNs) for transfusion into bone marrow failure (e.g., malignant neoplasms resistant to
neutropenic patients who have life-threatening infections. treatment, aplastic anemia, bone marrow transplantation)
Because larger doses of PMNs can be transfused, renewed may benefit from the addition of granulocyte transfusion
interest has arisen in the use of PMN (granulocyte) trans- to antibiotic therapy. The use of granulocyte transfusion for
fusions to treat adult oncology patients and hematopoietic bacterial sepsis that is unresponsive to antibiotics in patients
progenitor cell transplant recipients, in whom neutropenia with severe neutropenia (<0.5 × 109/L) is supported by most
complicated by severe infections persists as a significant controlled studies.67,70
problem despite combination antibiotic therapy, recombi- Children with qualitative neutrophil defects (neutrophil
nant cytokines, myeloid growth factors, and use of mobilized dysfunction) usually have adequate numbers of blood neu-
peripheral blood progenitor cells to minimize neutropenic trophils but are susceptible to serious infections because their
infections. If children are suffering significant morbidity cells kill pathogenic microorganisms inefficiently. Neutrophil
and mortality from neutropenic infections despite modern dysfunction syndromes are rare, and no definitive studies
supportive care, it is logical to explore the efficacy, poten- have established the efficacy of granulocyte transfusion in
tial toxicity, and cost effectiveness of tranulocyte transfu- these patients. However, several patients with progressive
sion therapy through properly designed, randomized clinical life-threatening infections have improved strikingly with the
trials performed in pediatric subjects.67 addition of granulocyte transfusion to antimicrobial ther-
Serious and repeated infections with bacteria, yeast, and apy.73 These disorders are chronic, and because of the risk of
fungi are a consequence of severe neutropenia and PMN dys- inducing alloimmunization, GTX is recommended only if the
function in some settings. In the multicenter Trial to Reduce infections are clearly unresponsive to antimicrobial drugs.
Alloimmunization to Platelets (TRAP) study, 7% of adult
patients with acute nonlymphocytic leukemia died from
infection during first-remission induction therapy, despite Infants and Younger Children 37
the use of modern antibiotic therapy.68 In another study of
Pathophysiology of Neonatal Neutropenia
patients given intensive chemotherapy, some of whom also 517
and Neutrophil Dysfunction
received transfusions of autologous hematopoietic stem
cells, 7.6% of patients experienced systemic fungal infec- Neonates are unusually susceptible to severe bacterial infec-
tion.69 Unless severe neutropenia is reversed fairly quickly tions, and several defects of neonatal body defenses have
in adult patients, the mortality of systemic fungal infections been reported as possible contributing factors. PMNs iso-
approaches 100%. Therefore, modern “high-dose” granu- lated from the blood of neonates exhibit both quantitative
locyte transfusion therapy is considered by some experts and qualitative abnormalities that may be related to the
to be very promising for adult oncology and transplanta- increased incidence, morbidity, and mortality of bacterial
tion patients.70,71 However, contrary data suggest that, with infections. Abnormalities of neonatal PMNs include absolute
appropriate anti-infective therapy, serious infections are rare and relative neutropenia, diminished chemotaxis, abnormal
in patients who are transplanted with adequate numbers of adhesion and aggregation, defective cellular orientation and
peripheral blood progenitor cells.72 receptor capping, decreased deformability, inability to alter
Because of these controversial views, pediatricians must membrane potential during stimulation, imbalances of oxi-
survey the outcome of life-threatening infections with bacte- dative metabolism, and a diminished ability to withstand
ria, yeast, and fungi in children who are undergoing intense oxidant stress.74 A complete discussion of neonatal PMN
chemotherapy or hematopoietic progenitor cell transplanta- physiology is beyond the scope of this chapter, and only
tion in their own institutions to determine whether there is aspects that are particularly relevant to PMN transfusions
a need for therapeutic granulocyte transfusion. If infections and alternative therapies are reviewed here.
in neutropenic children respond promptly to antibiotics plus Neutropenia can occur during neonatal bacterial infec-
standard supportive care and survival approaches 100%, tions, particularly with fulminant sepsis. Because a physiologic
granulocyte transfusion is unnecessary. Moreover, it should neutrophilia occurs in normal neonates, it is considered quite
not be used if there is no apparent need, because the lack abnormal for the absolute blood PMN count to fall below
of demonstrable benefit would not outweigh the potential 3.0 × 109/L during the first week of life. Although an abnor-
risks. However, if significant numbers of infected high-risk mally low PMN count can occur in neonates with disorders as
patients fail to respond to antibiotics alone, or if the intensity diverse as sepsis, asphyxia, and maternal hypertension, suspi-
of therapy is compromised because it is limited by fear of cion of severe bacterial infection must always be high when-
neutropenia, the addition of granulocyte transfusion should ever relative neutropenia (PMN count, <3.0 × 109/L) occurs.
be considered, along with other modifications of therapy The mechanisms responsible are only partially defined, but
intended to reduce infections, such as selection of different abnormalities of neonatal granulopoiesis frequently are
TRANSFUSION MEDICINE
involved. As one factor, the postmitotic marrow PMN stor- risks for both neonates and donors. Accordingly, alterna-
age pool (metamyelocytes and mature, segmented PMNs) is tive therapies have been suggested. However, their efficacy
small. The PMN storage pool accounts for 26% to 60% of has not been clearly established, their risks are only partially
all nucleated cells in the bone marrow of normal neonates. defined, and they require extensive study before they can be
Neonates with sepsis may exhibit a storage pool numbering widely accepted. Two modalities that have been suggested are
less than 10% of nucleated marrow cells and are considered IVIG and myeloid cytokines or growth factors.
to have severely diminished marrow PMN reserves.75 Second, Most studies evaluating IVIG to prevent infections have
storage pool PMNs are released at an excessively rapid and, found little or only modest benefit.90–101 However, results
apparently, poorly regulated rate from the marrow during are inconsistent. Only a few studies have suggested prophy-
stress. Third, PMN production in response to infection is lactic benefit.88–90 In contrast, several therapeutic studies
decreased. The number of committed (clonogenic) PMN pre- have demonstrated a benefit from the addition of IVIG to
cursors in neonatal marrow is lower in neonates than in older antibiotics in the treatment of neonatal infections.102–106 In
patients, and a high percentage of these cells are proliferating a meta-analysis, studies of prophylactic IVIG were found to
even when studied at an apparently basal state.75,76 Therefore, demonstrate only minimal benefit, whereas studies of thera-
neonatal marrow is functioning at capacity and is unable to peutic IVIG exhibited unequivocal benefit.107 Overall, the
rapidly expand production to meet the increased demands data are insufficient to justify the routine use of IVIG in all
of infection.77 For this reason, it is logical to consider PMN preterm neonates to prevent or treat sepsis. However, modest
transfusions until the marrow recovers. “physiologic” doses (0.3 to 0.4 mg/kg) may lessen the severity
of bacterial sepsis in newborns with very low birth weights,
Recommendations for Neutrophil who are likely to be hypogammaglobulinemic as a result of
Transfusions during Infancy extremely premature birth (i.e., before the major placental
Because both quantitative and qualitative abnormalities of transport of immunoglobulin G has taken place). However,
neonatal PMNs have been reported, PMN transfusions have caution must be used when prescribing IVIG therapy to pre-
been used to treat neonatal sepsis with or without neutrope- vent or treat neonatal sepsis. IVIG therapy, particularly at
nia. Neonates exhibiting fulminant sepsis, relative neutrope- high doses, can impair body defense mechanisms.108–110
nia (PMN count, <3.0 × 109/L during the first week of life To date, properly designed clinical studies of recombinant
or <1.0 × 109/L thereafter), and a severely diminished PMN myeloid growth factors given to human neonates are limited.
marrow storage pool (less than 10% of nucleated marrow In a controlled study,111 42 neonates with presumed bacterial
cells being postmitotic PMNs) are at increased risk of death sepsis, recognized within the first 3 days of life, were ran-
if treated only with antibiotics. Results of 11 studies78–88 on domly assigned to receive three doses of either G-CSF or a
the use of PMN transfusions to treat infected neonates, 6 placebo. Although the outcome of sepsis was not reported,
of which were designed as controlled studies,78–81,86,87 have G-CSF induced a significant increase in the blood PMN
been reported. The fact that four of the six controlled studies count, an increase in the marrow PMN storage pool, and an
III noted significant benefit from PMN transfusions is encour- increase in PMN membrane C3bi expression—the last being
aging.78–81 However, the controlled studies contained several an indication of enhanced functional capability.
518 experimental flaws.89 In a controlled study of granulocyte-macrophage colony-
Because of these scientific imperfections, firm recommen- stimulating factor (GM-CSF) in premature neonates,112 20
dations for the role of PMN transfusions in the treatment premature neonates were randomly assigned within 72 hours
of neonatal sepsis cannot be made at this time. Guidelines after birth to receive either GM-CSF or a placebo for 7 days.
for PMN transfusions are presented in Table 37–2. Although GM-CSF increased the blood PMN count, the marrow PMN
antibiotics are the key to successful treatment of neona- storage pool, and C3bi receptor expression. In addition,
tal sepsis, antibiotic therapy is not 100% successful, and neonates receiving GM-CSF exhibited an increase in blood
attempts to bolster body defenses are warranted. PMN trans- monocyte and platelet counts. The study was not designed to
fusions have not provided a complete answer; although they assess efficacy in the prevention or treatment of infections.
are efficacious for infants with neutropenia and fulminant Two additional randomized clinical trials have been con-
sepsis,89 only PMN transfusions obtained by automated leu- ducted to assess the efficacy of G-CSF and GM-CSF. Neither
kapheresis have demonstrated effectiveness.88,89 Moreover, demonstrated clear clinical benefit. In the G-CSF trial, 20
in many instances, standard supportive care with antibiotics infants with neutropenia and sepsis received either G-CSF
seems adequate. Each institution must assess its own experi- (10 mg/kg/day) or placebo for 3 days.113 Acknowledging that
ence with neonatal sepsis. If almost all infants survive with- the number of study subjects was too small for definitive
out apparent long-term morbidity when treated only with conclusions, the study authors noted that G-CSF did not sig-
antibiotics, PMN transfusions are unnecessary, and attention nificantly improve severity of illness, morbidity, or mortal-
should be focused on prompt diagnosis and optimal antibi- ity. In a preliminary report of the GM-CSF trial,114 preterm
otic therapy. If the outcome of standard therapy is not opti- infants received either GM-CSF (8 μg/kg/day) or placebo
mal, alternative therapies such as PMN transfusions must be for the first 28 days of life in an attempt to reduce the inci-
considered to improve the outlook. dence of infections. Although GM-CSF was well tolerated
and significantly increased blood leukocyte counts, it did not
Alternatives to Neonatal Neutrophil significantly decrease infection rates. Therefore, firm guide-
Transfusions lines cannot be made at this time regarding the proper role
Not all neonatologists prescribe PMN transfusions. Their of myeloid growth factors in the management of neonatal
proper role has not been irrefutably established by controlled neutropenia or sepsis.
clinical trials. Moreover, the preparation of PMN concen- Clearly, there is no universally accepted role for PMN
trates by leukapheresis can be cumbersome and expensive, transfusions, IVIG, or myeloid growth factors in the treat-
and the process of collecting and transfusing PMNs can pose ment of neonatal sepsis. However, it seems reasonable to
TRANSFUSION OF NEONATES AND PEDIATRIC PATIENTS
treat fulminant sepsis in neonates with neutropenia (blood neonates and pediatric patients as for adults; however, dos-
PMN counts <3 × 109/L during the first week of life or <1 × ing is in mL/kg rather than units. The dose of FFP in chil-
109/L thereafter) as follows. For infants born before 30 weeks’ dren is 10 to 20 mL/kg (Table 37–5). This dose will raise
gestation, give one dose of 500 mg/kg of IVIG plus 5 μg/kg of most coagulation factor levels by approximately 20%.117
G-CSF on 3 consecutive days. For infants born at 30 weeks’ However, plasma frozen within 24 hours has lower levels of
gestation or later, give 5 μg/kg of G-CSF on 3 consecutive factor VIII and factor V and thus might not increase those
days. This therapy should be adjunctive to optimal antibiotic factors as high as 20%.118 Generally, this is not of clinical
and supportive care. significance, and FFP and plasma frozen within 24 hours
are used interchangeably. On the other hand, cryoreduced
plasma (devoid of factor VIII, von Willebrand factor, fac-
PLASMA PRODUCT TRANFUSIONS tor XIII, and fibrinogen) was recently approved by the U.S.
Food and Drug Administration for use in refractory throm-
Recommendations for use of plasma product transfusions botic thrombocytopenic purpura (TTP), those patients who
do not need to be broken down into sections for older chil- are unresponsive to standard therapy with FFP, as its only
dren and adolescents and infants and younger children true indication. Some authorities advocate the use of cryo-
because product selection is similar in each group and in reduced plasma as a first-line therapy for TTP, an off-label
adults. However, due to smaller plasma volumes in many use. However, in 2001 the North American TTP Group pub-
of these individuals, certain aspects need to be emphasized. lished a multicenter prospective, randomized trial compar-
First, plasma (all types) should be ABO-compatible with ing exchange transfusion with FFP to cryoreduced plasma
the recipient’s RBCs; in other words, it should not contain for the initial treatment of TTP. They demonstrated that
donor isohemagglutinins that may react with the recipient survival was the same in both study groups and concluded
A and/or B red cell antigens. This is to prevent a “minor equal efficacy between FFP and cryoreduced plasma for ini-
side” ABO incompatibility that could result in an acute tial therapy in TTP.119,120
hemolytic transfusion reaction, which has a higher prob- Reference values of plasma concentrations for each coag-
ability of occurring in a smaller patient receiving “out-of- ulation factor for pediatric patients under age 6 months are
group” plasma.66,115 When considering the patient’s RhD lower than for children and adults with regard to vitamin
status, FFP is usually not matched because it contains very K–dependent factors (factors II, VII, IX, and X) in addition
low numbers of RBCs. However, when large volumes of to the contact factors and the vitamin K–dependent natural
RhD-positive FFP (>20 mL/kg) are being transfused to inhibitors of coagulation. As a result the prothrombin time
RhD-negative pediatric patients (or women of childbearing and activated partial thromboplastin time are prolonged.
age), RhD immunization prevention with RhD immune This distinct difference should be taken into account when
globulin should be considered. considering frozen plasma therapy for prolonged screening
tests in children under age 6 months who are septic with dis-
seminated intravascular coagulation and hemorrhage from 37
Frozen Plasma
surgery or trauma, because these already low factors may be
Recommendations with regard to blood component admin- depleted due to lower baseline levels initially and may war- 519
istration and guidelines for transfusion of FFP, plasma fro- rant earlier intervention with plasma transfusion.121 Another
zen within 24 hours F24, and cryoreduced plasma have been important point regarding frozen plasma is that it is not
recently published for neonates and pediatric patients.1,116 indicated for volume expansion nor for improving wound
The indications and contraindications are very similar for healing.1,122
Table 37-5 Guidelines for the Dosing and Transfusion of FFP/F24/CRP and Cryoprecipitate
Adapted from Roseff SD, Luban NLC, Manno CS. Guidelines for assessing appropriateness of pediatric transfusion. Transfusion 2002;42:1398.
TRANSFUSION MEDICINE
CRYOPRECIPITATE 23. Widness JA, Veng-Pedersen P, Peters C, et al. Erythropoietin phar-
macokinetics in premature infants: developmental, nonlinearity, and
treatment effects. J Appl Physiol 1996;80:140–148.
Just as with the various types of frozen plasmas available, the 24. Ruth V, Widness JA, Clemons G, Raivio KO. Postnatal changes in
indications for cryoprecipitate transfusion are the same in serum immunoreactive erythropoietin in relation to hypoxia before
children as in adults. Again, in children it is recommended to and after birth. J Pediatr 1990;116:950–954.
25. George JW, Bracco CA, Shannon KM, et al. Age related difference
give ABO-compatible units; however, the RhD group need in erythropoietic response to recombinant human erythropoietin:
not be honored. The dose is approximately 1 to 2 pooled comparison of adults and infants rhesus monkeys. Pediatr Res 1990;
units for every 10 kg of body weight. Table 37–5 outlines the 28:567–571.
dosing and indications for transfusion of cryoprecipitate. 26. Widness JA, Kulhavy JC, Johnson KJ, et al. Clinical Performance of
In neonates and pediatric patients cryoprecipitate is mostly an in-line point-of-care monitor in neonates. Pediatrics 2000;106:
497–504.
administered to increase fibrinogen. It is rarely transfused in 27. Ringer SA, Richardson DK, Sacher RA, et al. Variations in transfusion
the United States for von Willebrand’s disease or factor VIII practice in neonatal intensive care. Pediatrics 1998;101:194–200.
deficiency because safer products are recommended, such as 28. Bednarek FJ, Weisberger S, Richardson DK, et al, for the SNAP II Study
virally inactivated and recombinant factor VIII products.123 Group. Variations in blood transfusions among newborn intensive
care units. J Pediatr 1998;133:601–607.
29. Ramasethu J, Luban NL. Red blood cell transfusions in the newborn.
Semin Neonatol 1999;4:5–16.
REFERENCES 30. Blanchette VS, Hume HA, Levy GJ, et al. Guidelines for auditing pedi-
atric blood transfusion practices. Am J Dis Child 1991;145:787–796.
1. Roseff SD, Luban NLC, Manno CS. Guidelines for assessing appropri- 31. Bell EF, Strauss RG, Widness RG, et al. Randomized trial of liberal
ateness of pediatric transfusion. Transfusion 2002;42:1398–1413. versus restrictive guidelines for red blood cell transfusion in preterm
2. Wu WC, Rathmore SS, Wand Y, et al. Blood transfusion in elderly infants. Pediatrics 2005;115:1685–1691.
patients with acute myocardial infarction. NEJM 2001;345:1230–1236. 32. Kissack CM, Garr R, Wardle SP, Weindling AM. Postnatal changes in
3. Bratton SL, Annich GM. Packed red blood cell transfusions for criti- cerebral oxygen extraction in the preterm infant are associated with
cally ill pediatric patients: when and for what conditions? J Pediatr intraventricular hemorrhage and hemorrhagic parenchymal infarction
2003;142:95–97. but not periventricular leukomalacia. Pediatr Res 2004;56:111–116.
4. Laverdiere C, Gauvin F, Hebert PC, et al. Survey on transfusion prac- 33. Heddle NM, Kirpalani H. Identifying the optimal red cell transfusion
tices of pediatric intensivists. Pediatr Crit Care Med 2002;3:381–382. threshold for extremely low birth weight infants: the Premature In
5. Hebert PC, Wells G, Blajchman MA, et al, and the Transfusion Require- Need of Transfusion (PINT) study. Transfusion 2004;44:1A.
ments in Critical Care Investigators for the Canadian Critical Care 34. Phillips HM, Holland BM, Abdel-Moiz A, et al. Determination of red-
Trials Group. A multicenter, randomized, controlled clinical trial of cell mass in assessment and management of anaemia in babies needing
transfusion requirements in critical care. NEJM 1999;340:409–417. blood transfusion. Lancet 1986;1:882–884.
6. Goodman AM, Murray PM, Kantilal PM, Luban NLC. Pediatric red blood 35. Jones JG, Holland BM, Veale KE, Wardrop CA. “Available oxygen,” a
cell transfusions increase resource use. J Pediatr 2003;142:123–127. realistic expression of the ability of the blood to supply oxygen to tis-
7. Strauss RG. Red blood cell transfusion practices in the neonate. Clin sues. Scand J Haematol 1979;22:77–82.
Perinatol 1995;22:641–655. 36. Alverson DC, Isken VH, Cohen RS. Effect of booster blood transfu-
8. Strauss RG, Sacher RA, Blazina JF, et al. Commentary on small-volume sions on oxygen utilization in infants with bronchopulmonary dyspla-
red cell transfusions for neonatal patients. Transfusion 1990;30:565–570. sia. J Pediatr 1988;113:722–726.
III 9. Stockman JA. Anemia of prematurity: current concepts in the issue of 37. Hudson IR, Cavill IA, Cooke AD, et al. Biotin labeling of red cells in
when to transfuse. Pediatr Clin North Am 1986;33:111–128. the measurement of red cell volume in preterm infants. Pediatr Res
520 10. Wardrop CA, Holland BM, Veale KE, et al. Nonphysiological anaemia 1990;28:199–202.
of prematurity. Arch Dis Child 1978;53:855–860. 38. Hudson I, Cooke A, Holland B, et al. Red cell volume and cardiac out-
11. Holland BM, Jones JG, Wardrop CA. Lessons from the anemia of pre- put in anaemic preterm infants. Arch Dis Child 1990;65:672–675.
maturity. Hematol Oncol Clin North Am 1987;1:355–366. 39. Mock DM, Lankford GL, Widness JA, et al. Measurement of circulating
12. Stockman JA III, Garcia JF, Oski FA. The anemia of prematurity: fac- red blood cell volume using biotin labeled red cells: validation against
51
tors governing the erythropoietin response. NEJM 1977;296:647–650. Cr labeled red cells. Transfusion 1999;39:149–155.
13. Brown MS, Phibbs RH, Garcia JF, Dallman PR. Postnatal changes in 40. Mock DM, Lankford GL, Widness JA, et al. Measurement of red cell
erythropoietin levels in untransfused premature infants. J Pediatr 1983; survival using biotin labeled red cells: Validation against 51Cr labeled
103:612–617. red cells. Transfusion 1999;39:156–162.
14. Stockman JA III, Graeber JE, Clark DA, et al. Anemia of prematurity: 41. Strauss RG, Mock DM, Widness JA, et al. Posttransfusion 24-hour
Determinants of the erythropoietin response. J Pediatr 1984;105:786–792. recovery and subsequent survival of allogeneic red blood cells in the
15. Brown MS, Garcia JF, Phibbs RH, Dallman PR. Decreased response of bloodstream of newborn infants. Transfusion 2004;6:871–876.
plasma immunoreactive erythropoietin to “available oxygen” in ane- 42. Luban NLC, Strauss RG, Hume HA. Commentary on the safety of red
mia of prematurity. J Pediatr 1984;105:793–798. blood cells preserved in extended storage media for neonatal transfu-
16. Shannon KM, Naylor GS, Torkildson JC, et al. Circulating erythroid sions. Transfusion 1990;30:229–235.
progenitors in the anemia of prematurity. NEJM 1987;317:728–733. 43. Liu EA, Mannio FL, Lane TA. Prospective, randomized trial of the
17. Rhondeau SM, Christensen RD, Ross MP, et al. Responsiveness to safety and efficacy of a limited donor exposure transfusion program
recombinant human erythropoietin of marrow erythroid progeni- for premature neonates. J Pediatr 1994;125:92–96.
tors from infants with the “anemia of prematurity.” J Pediatr 1988;112: 44. Lee DA, Slagle TA, Jackson TM, Evans CS. Reducing blood donor
935–940. exposures in low birth weight infants by the use of older, unwashed
18. Zanjani ED, Ascensao JL, McGlave PB, et al. Studies on the liver to kid- packed red blood cells. J Pediatr 1995;126:280–286.
ney switch of erythropoietin production. J Clin Invest 1981;67:1183– 45. Wood A, Wilson N, Skacel P, et al. Reducing donor exposure in pre-
1188. term infants requiring multiple blood transfusions. Arch Dis Child
19. Dame C, Fahnenstich H, Freitag P, et al. Erythropoietin mRNA expres- Fetal Neonat 1995;72:F29–F33.
sion in human fetal and neonatal tissue. Blood 1998;92:3218–3225. 46. Strauss RG, Burmeister LF, Johnson K, et al. AS-1 red blood cells for
20. Widness JA, Susa JB, Garcia JF, et al. Increased erythropoiesis and ele- neonatal transfusions: a randomized trial assessing donor exposure
vated erythropoietin in infants born to diabetic mothers and in hyper- and safety. Transfusion 1996;36:873–878.
insulinemic rhesus fetuses. J Clin Invest 1981;67:637–642. 47. Strauss RG. Recombinant erythropoietin for the anemia of prematu-
21. Snijders RJ, Abbas A, Melby O, et al. Fetal plasma erythropoietin con- rity: still a promise, not a panacea. J Pediatr 1997;131:653–655.
centration in severe growth retardation. Am J Obstet Gynecol 1993; 48. Widness JA, Strauss RG. Recombinant erythropoietin in treatment of
168:615–619. the premature newborn. Semin Neonatol 1998;3:163.
22. Ohls RK, Li Y, Trautman MS, Christensen RD. Erythropoietin produc- 49. Widness JA, Seward VJ, Kromer IJ, et al. Changing patterns of red
tion by macrophages from preterm infants: implications regarding the blood cell transfusion in very low birthweight infants. J Pediatr 1996;
cause of the anemia in prematurity. Pediatr Res 1994;35:169–170. 129:680–687.
TRANSFUSION OF NEONATES AND PEDIATRIC PATIENTS
50. Shannon KM, Keith JF III, Mentzer WC, et al. Recombinant human eryth- 75. Christensen RD, MacFarlane JL, Taylor NL, et al. Blood and marrow
ropoietin stimulates erythropoiesis and reduces erythrocyte transfusions neutrophils during experimental group B streptococcal infection:
in very-low-birth-weight preterm infants. Pediatrics 1995;95:1–8. quantification of the stem cell, proliferative, storage and circulating
51. Ohls RK, Veerman MW, Christensen RD. Pharmacokinetics and pools. Pediatr Res 1982;16:549–553.
effectiveness of recombinant erythropoietin administered to preterm 76. Erdman SH, Christensen RD, Bradley PP, Rothstein G. Supply and
infants by continuous infusion in total parenteral nutrition solution. release of storage neutrophils: A developmental study. Biol Neonate
J Pediatr 1996;128:518–523. 1982;41:132–137.
52. Ohls RK, Harcum J, Schibler KR, Christensen RD. The effect of eryth- 77. Christensen RD, Harper TE, Rothstein G. Granulocyte-macrophage
ropoietin on the transfusion requirements of preterm infants weigh- progenitor cells in term and preterm neonates. J Pediatr 1986;109:
ing 750 grams or less: a randomized, double-blind, placebo-controlled 1047–1051.
study. J Pediatr 1997;131:661–665. 78. Christensen RD, Rothstein G, Anstall HB, Bybee B. Granulocyte trans-
53. Rebulla P, Finazzi G, Marangoni F, et al. The threshold for prophy- fusions in neonates with bacterial infection, neutropenia and deple-
lactic platelet transfusions in adults with acute myeloid leukemia. tion of mature marrow neutrophils. Pediatrics 1982;70:1–6.
Gruppo Italiano Malattie Ematologiche Maligne dell’Adulto. NEJM 79. Laurenti F, Ferro R, Isacchi G, et al. Polymorphonuclear leukocyte
1997;337:1870–1875. transfusion for the treatment of sepsis in the newborn infants. J Pedi-
54. Wandt H, Frank M, Ehninger G, et al. Safety and cost effectiveness of a atr 1981;98:118–123.
10 × 109/L trigger for prophylactic platelet transfusions compared with 80. Cairo MS, Rucker R, Bennetts GA, et al. Improved survival of new-
the traditional 20 × 109/L trigger: a prospective comparative trial in borns receiving leukocyte transfusions for sepsis. Pediatrics 1984;74:
105 patients with acute myeloid leukemia. Blood 1998;91:3601–3606. 887–892.
55. Castle V, Andrew M, Kelton J, et al. Frequency and mechanism of neo- 81. Cairo MS, Worcester C, Rucker R, et al. Role of circulating comple-
natal thrombocytopenia. J Pediatr 1986;108:749–755. ment and polymorphonuclear leukocyte transfusion in treatment
56. Andrew M, Castle V, Saigal S, et al. Clinical impact of neonatal throm- and outcome in critically ill neonates with sepsis. J Pediatr 1987;110:
bocytopenia. J Pediatr 1987;110:457. 935–941.
57. Castle V, Coates G, Kelton JG, Andrew M. 111In-oxine platelet survivals 82. DeCurtis M, Romano G, Scarpato N, et al. Transfusions of polymor-
in thrombocytopenic infants. Blood 1987;70:652–656. phonuclear leukocytes (PMN) in an infant with necrotizing enteroco-
58. Murray NA, Roberts IA. Circulating megakaryocytes and their pro- litis (NEC) and a defect of phagocytosis. J Pediatr 1981;99:665–666.
genitors in early thrombocytopenia in preterm neonates. Pediatr Res 83. Christensen RD, Anstall H, Rothstein G. Neutrophil transfusion in
1996;40:112–119. septic neutropenic neonates. Transfusion 1982;22:151–153.
59. Wolber E-M, Dame C, Fahnenstich H, et al. Expression of the throm- 84. Laing IA, Boulton FE, Hume R. Polymorphonuclear leukocyte transfu-
bopoietin gene in human fetal and neonatal tissues. Blood 1999;94:- sion in neonatal septicemia. Arch Dis Child 1983;58:1003–1005.
97–105. 85. Laurenti F, LaGreca G, Ferro R, Bucci G. Transfusion of polymorpho-
60. Murray NA, Watts TL, Roberts IA. Endogenous thrombopoietin levels nuclear neutrophils in a premature infant with Klebsiella sepsis. Lancet
and effect of recombinant human thrombopoietin on megakaryocyte 1978;2:111–112.
precursors in term and preterm babies. Pediatr Res 1998;43:148–151. 86. Baley JE, Stork EK, Warkentin PI, Shurin SB. Buffy coat transfusions
61. Andrew M, Vegh P, Caco C, et al. A randomized trial of platelet trans- in neutropenic neonates with presumed sepsis: a prospective, random-
fusions in thrombocytopenic premature infants. J Pediatr 1993;123: ized trial. Pediatrics 1987;80:712–720.
285–291. 87. Wheeler JC, Chauvenet AR, Johnson CA, et al. Buffy coat transfusions
62. Lupton BA, Hill A, Whitfield MF, et al. Reduced platelet count as a in neonates with sepsis and neutrophil storage pool depletion. Pediat-
risk factor for intraventricular hemorrhage. Am J Dis Child 1988;142: rics 1987;79:422–425.
1222–1224. 88. Newman RS, Waffarn F, Simmons GE, et al. Questionable value of
63. Conway LT, Scott EP. Acute hemolytic transfusion reaction due to ABO saline prepared granulocytes in the treatment of neonatal septicemia.
incompatible plasma in a plateletpheresis concentrate. Transfusion Transfusion 1988;28:196–197.
37
1984;24:413–414. 89. Strauss RG. Current status of granulocyte transfusions to treat neona-
64. Pierce RN, Reich LM, Mayer K. Hemolysis following platelet transfu- tal sepsis. J Clin Apheresis 1989;5:25–29. 521
sions from ABO-incompatible donors. Transfusion 1985;25:60–62. 90. Haque KN, Zaidi MN, Haque SK, et al. Intravenous immunoglobulin
65. Valbonesi M, Deluigi MC, Lercari G, et al. Acute intravascular hemoly- for prevention of sepsis in preterm and low birth weight infants. Pedi-
sis in two patients transfused with dry-platelet units obtained from the atr Infect Dis 1986;5:622–625.
same ABO incompatible donor. Int J Artif Organs 2000;23:642–646. 91. Chirico G, Rondini G, Plebani A, et al. Intravenous gammaglobulin
66. Angiolillo A, Luban NLC. Hemolysis following an out-of-group plate- therapy for prophylaxis of infection in high-risk neonates. J Pediatr
let transfusion in an 8-month-old with Langerhans cell histiocytosis. 1987;110:437–442.
J Pediatr Hematol Oncol 2004;26:267–269. 92. Clapp DW, Kliegman RM, Baley JE, et al. Use of intravenously admin-
67. Strauss RG. The rebirth of granulocyte transfusions: Should it involve istered immune globulin to prevent nosocomial sepsis to low birth
pediatric oncology and transplant patients? Am J Pediatr Hematol weight infants: report of a pilot study. J Pediatr 1989;115:973–978.
Oncol 1999;21:475–478. 93. Conway SP, Ng PC, Howel D, et al. Prophylactic intravenous immu-
68. The Trial to Reduce Alloimmunization to Platelets Study Group. Leu- noglobulin in preterm infants: a controlled trial. Vox Sang 1990;59:
kocyte reduction and ultraviolet B irradiation of platelets to prevent 6–11.
alloimmunization and refractoriness to platelet transfusions. NEJM 94. Stabile A, Miceli Sopo S, Romanelli V, et al. Intravenous immunoglob-
1997;337:1861–1870. ulin for prophylaxis of neonatal sepsis in premature infants. Arch Dis
69. Peters BG, Adkins DR, Harrison BR, et al. Antifungal effects of yeast- Child 1988;63:441–443.
derived rhu-GM-CSF in patients receiving high-dose chemotherapy 95. Magny JF, Bremard-Oury C, Brault D, et al. Intravenous immunoglob-
given with or without autologous stem cell transplantation: a retro- ulin therapy for prevention in high-risk premature infants: report of a
spective analysis. Bone Marrow Transplant 1996;18:93–102. multicenter, double-blind study. Pediatrics 1991;88:437–443.
70. Vamvakas EC, Pineda AA. Meta-analysis of clinical studies of efficacy 96. Baker CJ, Melish ME, Hall RT, et al. Intravenous immune globulin for
of granulocyte transfusions in the treatment of bacterial sepsis. J Clin the prevention of nosocomial infection in low-birth-weight neonates.
Apher 1996;11:1–9. The Multicenter Group for the Study of Immune Globulin in Neo-
71. Adkins D, Spitzer G, Johnson M, et al. Transfusions of granulocyte- nates. NEJM 1992;327:213–219.
colony-stimulating factor-mobilized granulocyte components to 97. Bussel JB. Intravenous gammaglobulin in the prophylaxis of late sepsis in
allogeneic transplant recipients: analysis of kinetics and factors deter- very-low-birth-weight infants: preliminary results of a randomized, dou-
mining posttransfusion neutrophil and platelet counts. Transfusion ble-blind, placebo-controlled trial. Rev Infect Dis 1990;12:S457–S461.
1997;37:737–748. 98. Fanaroff AA, Korones SB, Wright LL, et al, for the National Institute of
72. Kolbe K, Domkin D, Derigs HG, et al. Infectious complications during Child Health and Human Development Neonatal Research Network. A
neutropenia subsequent to peripheral blood stem cell transplantation. controlled trial of intravenous immune globulin to reduce nosocomial
Bone Marrow Transplant 1997;19:143–147. infections in very-low-birth-weight infants. NEJM 1994;330:1107–1113.
73. Ikinciogullari A, Dogu F, Solaz N, et al. Granulocyte transfusions in 99. van Overmeire B, Bleyaert S, van Reempts PJ, van Acker KJ: The use
children with chronic granulomatous disease and invasive aspergillo- of intravenously administered immunoglobulins in the prevention
sis. Ther Apher Dial 2005;2:137–141. of severe infection in very low birth weight neonates. Biol Neonate
74. Rosenthal J, Cairo MS. Neonatal myelopoiesis and immunomodula- 1993;64:110–115.
tion of host defenses. In Petz LD, Swisher SN, Kleinman S, et al (eds): 100. Weisman LE, Stoll BJ, Kueser TJ, et al. Intravenous immune globu-
clinical Practice of Transfusion Medicine, 3rd ed. New York, Churchill lin prophylaxis of late-onset sepsis in premature neonates. J Pediatr
Livingstone, 1995, p 685. 1994;125:922–930.
TRANSFUSION MEDICINE 101. Christensen RD, Hardman T, Thornton J, Hill HR. A randomized, factor in very low birthweight neonates: significant induction of circu-
double-blind, placebo-controlled investigation of the safety of intra- latory neutrophils, monocytes, platelets, and bone marrow neutrophils.
venous immune globulin administration to preterm neonates. J Peri- Blood 1995;86:2509–2515.
natol 1989;9:126–130. 113. Schibler KR, Osborne KA, Leung LY, et al. A randomized, placebo-con-
102. Sidiropoulos D, Boehme U, von Muralt G, et al. Immunoglobulin sup- trolled trial of granulocyte colony-stimulating factor administration
plementation in prevention or treatment of neonatal sepsis. Pediatr to newborn infants with neutropenia and clinical signs of early-onset
Infect Dis J 1986;5:S193–S194. sepsis. Pediatrics 1998;102:6–13.
103. Haque KN, Zaidi MH, Bahakim H. IgM-enriched intravenous immuno- 114. Cairo MS, Agosti J, Ellis R, et al. A randomized, double-blind, placebo-
globulin therapy in neonatal sepsis. Am J Dis Child 1988;142:1293–1296. controlled trial of prophylactic recombinant human granulocyte-mac-
104. Friedman CA, Wender DG, Temple DM, Rawson JE. Intravenous rophage colony-stimulating factor to reduce nosocomial infections in
gamma globulin as adjunct therapy for severe group B streptococcal very low birth weight neonates. J Pediatr 1999;134:64–70.
disease in the newborn. Am J Perinatol 1990;7:1–4. 115. Duguid JKM, Minards J, Bolton-Maggs PHB. Incompatible plasma
105. Weisman LE, Stoll BJ, Kueser TJ, et al. Intravenous immune globu- transfusions and haemolysis in children. BMJ 1999;31:176–177.
lin therapy for early-onset sepsis in premature neonates. J Pediatr 116. Pisciotto P (ed). Pediatric Hemotherapy Data Card. Bethesda, Md.,
1992;121:434–443. American Association of Blood Banks, 2002.
106. Haque KN, Remo C, Bahakim H. Comparison of two types of intrave- 117. Hume HA, Limoges P. Perioperative blood transfusion therapy in pedi-
nous immunoglobulins in the treatment of neonatal sepsis. Clin Exp atric patients. Amer J Therap 2002;9:396–403.
Immunol 1995;101:328–333. 118. Downes KA, Wilson E, Yomtovian R, Sarode R. Serial measurements
107. Jenson HB, Pollock BH. Meta-analyses of the effectiveness of intra- of clotting factors in thawed plasma stored for 5 days. Transfusion
venous immune globulin for prevention and treatment of neonatal 2001;41:570.
sepsis. Pediatrics 1997;99:E2. 119. Blackall DP, Uhl L, Spitalnik SL. Cryoprecipitate-reduced plasma:
108. Cross AS, Siegel G, Byrne WR, et al. Intravenous immune globulin Rationale for use and efficacy in the treatment of thrombotic thrombo-
impairs anti-bacterial defenses of a cyclophosphamide-treated host. cytopenic purpura. Transfusion 2001;41:840–844.
Clin Exp Immunol 1989;76:159–164. 120. Zeigler ZR, Gryn JF, Rintels PB, et al. Cryoprecipitate poor plasma does
109. Weisman LE, Lorenzetti PM. High intravenous doses of human not improve early response in primary adult thrombotic thrombocyto-
immune globulin suppress neonatal group B streptococcal immunity penic purpura (TTP). J Clin Apheresis 2001;16:19–22.
in rats. J Pediatr 1989;115:445–450. 121. Goodnight SH, Hathaway WE. Disorders of Hemostasis and Thrombo-
110. Cross AS, Alving BM, Sadoff JC, et al. Intravenous immune globulin: A sis. New York, McGraw-Hill, 2001, p 31.
cautionary note. Lancet 1984;1:912. 122. Luk C, Eckert KM, Barr RM, Chin-Yee IH. Prospective audit of the use
111. Gillan ER, Christensen RD, Suen Y, et al. A randomized, placebo-con- of fresh-frozen plasma, based on Canadian Medical Association trans-
trolled trial of recombinant human granulocyte colony-stimulating fusion guidelines. Can Med Assoc J 2002;166:1539–1540.
factor administration in newborn infants with presumed sepsis: sig- 123. Josephson CD, Abshire T. The new albumin-free recombinant factor
nificant induction of peripheral and bone marrow neutrophilia. Blood VIII concentrates for treatment of hemophilia: do they represent an
1994;84:1427–1433. actual incremental improvement? Clin Adv Hematal Oncol 2004;2:
112. Cairo MS, Christensen R, Sender LS, et al. Results of a phase I/II trial 441–446.
of recombinant human granulocyte-macrophage colony-stimulating
III
522
Chapter 38
Transfusion of the
Hemoglobinopathy Patient
Krista L. Hillyer ● James R. Eckman
The goals of RBC transfusion in SCD patients are: to Acute Simple Transfusion
improve oxygen-carrying capacity by increasing the total In simple transfusion, normal donor RBCs are infused into the
Hb concentration; to decrease blood viscosity and improve patient without removal of the patient’s own RBCs. In general,
blood flow by diluting RBCs containing sickle hemoglo- acute simple transfusion is indicated when the immediate
bin; and to suppress endogenous erythropoiesis by increas- need for oxygen-carrying capacity is increased but there is no
ing tissue oxygenation.6 The first goal is common to many need for a dramatic decrease in the percentage of HbS in the
clinical indications for RBC transfusion. The latter two goals patient’s blood.9 The volume of RBCs to be transfused can be
for RBC transfusion are required only in individuals with calculated by the formula shown in Figure 38–1.
TRANSFUSION MEDICINE
Table 38–1 Clinical Complications of Sickle Table 38–2 Methods of Transfusion for the
Cell Disease Patient with Sickle Cell Disease
Figure 38–1 Volume of red blood cells (RBCs) to be transfused to sickle cell patients using the acute simple transfusion method. Hct, hematocrit.
(From Wayne AS, Kevy SV, Nathan DG. Transfusion management of sickle cell disease. Blood 1993;81:1109–1123.)
TRANSFUSION OF THE HEMOGLOBINOPATHY PATIENT
Figure 38–2 Volume of red blood cells (RBCs) to be transfused to sickle cell patients using the chronic simple transfusion method. HbS, sickle
hemoglobin; Hct, hematocrit; TBV, total blood volume. (From Wayne AS, Kevy SV, Nathan DG. Transfusion management of sickle cell disease. Blood
1993;81:1109–1123.)
centrifugation, returns the plasma to the patient, and discards volume, which may lead to hypotension or a critically low
the patient’s sickled RBCs. At the same time, donor RBCs are Hct, or both, if the machine is primed with saline. Instead,
infused, maintaining a stable blood volume throughout the the automated apheresis instrument should be primed with
procedure. In the past, RBC exchange transfusion could be per- RBCs if the extracorporeal circuit represents 12% or more
formed only manually, and a variety of formulas were used to of the child’s blood volume, the child weighs less than 20 kg,
estimate the appropriate exchange volumes. Today, indications or the child is anemic or unstable.13
still exist for manual RBC exchange, but erythrocytapheresis is Occasionally, certain patients require manual RBC
currently the most common method of RBC exchange. exchange transfusions. Manual RBC exchange may be neces-
The automated apheresis instrument has an internal pro- sary in infants who have very small total blood volumes and
grammable computer that precisely calculates both the vol- in emergencies involving children or adults, when the addi-
ume of the patient’s RBCs to be removed and the volume tional time required to establish appropriate venous access
of the donor’s RBCs to be infused. First, the patient’s total or to mobilize the apheresis team would be deleterious to the
blood volume and RBC volume are calculated by entering patient’s health. A rapid manual partial RBC exchange may
into the computer the patient’s gender, height, weight, and be performed by withdrawing blood from a peripheral (usu-
current Hct. Next, the desired HbS percentage, the desired ally antecubital) vein and infusing RBCs via a stopcock into
final Hct, and the Hct of the donor RBC preparation are the same vein (or directly into a different peripheral vein),
entered, and the appropriate volumes for exchange are deter- using the methods outlined in Table 38–3. The requirements
mined. However, the formula shown in Figure 38–3 provides that must be honored to ensure a safe and effective manual
a general, practical estimate of the number of RBC units to exchange transfusion are the following: marked increases
be used in automated RBC exchange transfusion. Having an in blood viscosity should be avoided; blood volume should
estimate of the number of units to be exchanged before initi- be maintained throughout the exchange; and the exchange
ation of the procedure is important, because the blood bank should be completed in the shortest length of time pos-
may need extra time to procure the appropriate RBC units. sible.14 These three important factors must be evaluated and
The process of erythrocytapheresis is as follows. In most balanced, based on the unique clinical circumstances of the
cases, central venous access is established, typically in the sub- individual patient in need of exchange transfusion.
clavian vein or the internal jugular vein, using a hemodialy- 38
SELECTION OF ANTICOAGULANT/PRESERVATIVE SOLUTIONS
sis-grade, rigid-wall, large-lumen, double-bore catheter that
can withstand high flow rates. In adults and larger children In the past, when RBC units were collected primarily in 525
with easily accessible peripheral veins, the antecubital veins citrate-phosphate-dextrose (CPD) or citrate-phosphate-
may be acceptable for use, with 16- to 18-gauge needles used dextrose-adenine (CPDA-1) solutions, these units were
for blood removal and 18- to 20-gauge catheters for blood reported to have average hematocrits of 80% or more, and
return. The apheresis instrument, after being programmed it was recommended that RBCs be “reconstituted” to the
with the appropriate patient data and the desired end points volume and Hct of whole blood (30% to 40%) by adding
for therapy, is used by a trained operator to remove the albumin or saline to the blood bag or syringe before manual
patient’s RBCs and replace them with donor RBCs. Of note, exchange transfusion to the SCD patient.14
it is usually desirable to replace the patient’s sickled RBCs However, the current average Hct of an RBC unit is
with HbS-negative (sickle trait–negative) donor RBCs in an approximately 50% to 70%,11,15 whether collected in CDPA-1
exchange procedure, for the purposes of appropriately cal- solutions or in additive systems (e.g., AS-1, AS-3) that extend
culating and achieving the desired percentages of HbS and the shelf life of the product. The alternating administration
HbA.12 However, if HbS-negative units cannot be procured to an SCD patient of RBCs and saline in equal volumes (see
in a timely manner, it is at the clinician’s discretion (depen- Table 38–3) should theoretically deliver a product essentially
dent on the severity and acuteness of patient illness) whether identical to “reconstituted” whole blood.
or not to wait for HbS-negative units to be identified. Red blood cells containing additive solutions have more
In infants and small children, the amount of blood in the adenine and/or mannitol, in comparison with CPD or CPDA-
extracorporeal circuit of the automated apheresis machine 1–preserved RBC units, depending on the type of additive
represents a significant percentage of the child’s total blood solution used. Some clinicians hesitate to transfuse RBC units
with additional mannitol and adenine, in response to case
reports of renal toxicity and significant fluid shifts, in neonates
given large volumes of these substances. In a comprehensive
Exchange RBC volume (mL) desired % Hct TBV review of the existing literature, Strauss concluded that small-
volume transfusions of RBCs stored in additive solutions
administered to neonates do not lead to an increased risk of
Figure 38–3 Practical estimate of the volume of red blood cells adverse events.16 In the setting of large-volume transfusions,
(RBCs) to be transfused to sickle cell patients using the automated RBC
exchange transfusion method. Hct, hematocrit; TBV, total blood vol- such as RBC exchanges, existing literature does not support
ume. (From Wayne AS, Kevy SV, Nathan DG. Transfusion management or disavow the use of RBCs stored in additive solutions. Thus,
of sickle cell disease. Blood 1993;81:1109–1123.) many clinical protocols choose to utilize RBC units stored in
TRANSFUSION MEDICINE
Table 38-3 Method for Rapid Manual Red usually less than 30%, with desirable postpheresis hematocrits
Blood Cell (RBC) Exchange Transfusion in Sickle of 30% or less. Currently, chronic erythrocytapheresis is
Cell Patients, Using Practical Exchange Volume not universally used in the United States for SCD patients,
Estimates, for Children and Adults. although recent reports suggest that the potential benefits
of this method may outweigh its risks and costs. A detailed
Children review of the current literature regarding the role of chronic
1. Calculate exchange volume, using 60 mL/kg as a practical erythrocytapheresis is provided later in this chapter (see
estimate.
2. Divide the calculated exchange volume of RBCs into four
“Iron Overload”).
equal aliquots.
3. Withdraw blood from the patient equal to one exchange
aliquot.
Indications for Transfusion Therapy
4. Infuse saline equal to one exchange aliquot. RBCs may be administered to patients with SCD by simple
5. Withdraw blood from the patient equal to one exchange
aliquot. RBC infusion or by RBC exchange, either episodically, for
6. Transfuse a volume of RBCs equal to two exchange aliquots. the relief of acute symptoms, or chronically, for the preven-
7. Repeat steps 3-6. tion of long-term complications. A summary of indications
Adults for transfusion of RBCs in SCD patients is shown in Table
1. The exchange volume is 6–8 RBC units, depending on 38–4. These indications are characterized in the follow-
the size of the patient (an average 70-kg man requires ing sections as generally indicated or controversial, based on
approximately 6 RBC units). several recently published reviews of of the work of various
2. Withdraw 500 mL of blood from the patient.
3. Infuse 500 mL of saline. sickle cell experts.13,24
4. Withdraw 500 mL of blood from the patient.
5. Transfuse two units of RBCs. Indications for Episodic Transfusion
6. Repeat until 6–8 RBC units have been transfused.
ACUTE SYMPTOMATIC ANEMIA
RBC units and saline infused in this alternating manner
should theoretically deliver a blood product that is Because patients with SCD are chronically anemic, they are
essentially “reconstituted” (within the patient) to the often asymptomatic despite very low Hb levels. Biochemical
volume and hematocrit of whole blood (30% to 40%), and physiologic factors that decrease symptoms during chronic
because the current estimated average hematocrit of an RBC anemia include increased levels of 2,3-diphosphoglycerate,
unit is 50% to 70%.60,130 However, if a known exact percent
hematocrit of the blood product to be infused is desired,
decreased oxygen–Hb affinity, increased plasma volume, and
RBCs may be reconstituted to the volume and hematocrit of increased cardiac stroke volume and output.25 Patients may
whole blood (30% to 40%) within a blood bag or syringe, become acutely symptomatic, making simple transfusion nec-
by the addition of saline or other diluents.131 (From Reid essary, if they experience a rapid decrease in Hb, hypoxia, or
CD, Charache S, Lubin B (eds). Management and Therapy of acute cardiac decompensation. Acute anemia can result from
Sickle Cell Disease, 3rd ed. Bethesda, Md., National Institutes
III of Health Publication No.99-2117, 1999, p 62.) bleeding, suppression of erythropoiesis by infection, seques-
tration, or increased hemolysis. Pulmonary or cardiac disease
526 can cause decompensation, requiring acute transfusion to
CPDA-1 solution for large-volume transfusions in neonates, increase the Hb level above the patient’s stable baseline.
particularly if these children have liver or kidney dysfunction
APLASTIC CRISIS
that would put them at risk for adverse events.
In addition, the amount of potassium transfused to neo- Aplastic crisis, typically defined as a decrease in Hb of more
nates is of concern to clinicians, because mortality follow- than 3.0 g/dL with reticulocytopenia, is a relatively common
ing transfusion of RBCs with high potassium levels has been occurrence in SCD.26 Aplastic crisis occurs in SCD patients
reported.17 As the shelf life of RBC units is extended, potas- after marked suppression of erythropoiesis for 7 to 10 days,
sium is known to leak from the RBCs, although in amounts usually as a result of infection of RBC precursors in the bone
that are typically easily processed by individuals with nor- marrow by human parvovirus B19. In a sample group of 308
mal kidney function. However, in large-volume and RBC children with SCD, human parvovirus B19 accounted for all
exchange transfusions, many transfusion protocols for SCD cases of aplastic crisis that occurred during an 8-year obser-
patients recommend (but do not require) the use of RBC vation period.27 Because mean RBC survival time in many
units that are 7 days old or less. This general practice of trans- SCD patients is only 12 to 15 days, the acute life-threatening
fusing “young” RBC units to SCD patients is of practical use anemia caused by human parvovirus B19 infection necessi-
as well, in that younger RBCs will remain in the circulation tates immediate RBC transfusion to sustain oxygen delivery
of the patient for a longer period than will older RBCs. to the tissues until the bone marrow recovers. Simple RBC
transfusion is usually administered slowly (1 mL/kg/hr),
Chronic Erythrocytapheresis (Automated
because these patients have expanded plasma volumes due
Red Blood Cell Exchange)
to chronic anemia and care must be taken to avoid volume-
Iron overload is one of the serious long-term complications induced congestive heart failure.8,28 Partial exchange trans-
of chronic transfusion in SCD patients.18 As a result, certain fusion, performed by manually removing whole blood and
investigators19–23 have suggested that chronic erythrocytapher- returning RBCs to the patient without replacing the plasma,
esis should be used in place of chronic simple transfusion may be preferred if heart failure from acute volume overload
for the prevention of complications of SCD and to decrease is of significant concern.9
long-term iron accumulation in these patients.
ACUTE CHEST SYNDROME
The procedure for chronic erythrocytapheresis is the
same as that for acute erythrocytapheresis. Similarly, target According to a report by the Cooperative Study of Sickle
postpheresis HbS levels for chronic erythrocytapheresis are Cell Disease, acute chest syndrome occurs at least once
TRANSFUSION OF THE HEMOGLOBINOPATHY PATIENT
Table 38–4 Clinical Indications for Transfusion in Patients with Sickle Cell Disease
Type of
Transfusion Indication Controversial Indications Nonindications
Episodic Acute symptomatic anemia Acute painful episode (acute Normal pregnancy
pain crisis)
Transient red cell aplasia (“aplastic crisis”) Acute priapism
Acute chest syndrome Preparation for infusion of
contrast media
Hyperhemolysis
Acute splenic sequestration
Acute hepatic sequestration
Stroke
Prior to surgery requiring general
anesthesia
Prior to eye surgery
Acute multiorgan system failure
Severe infection with concurrent
severe anemia
Chronic Prevention of recurrent strokes Recurrent priapism Leg ulcers
in children
Prevention of first stroke in children “Silent” cerebral infarcts Growth and
developmental delays
Complicated pregnancy Neurocognitive defects Early retinopathy
Chronic hypoxic lung disease Recurrent acute chest syndrome Early renal disease
Frequent severe pain episodes Prevention of pulmonary hyper-
tension/cor pulmonale
Chronic heart failure
Chronic renal failure
Table 38–5 Common Immune and Nonimmune Adverse Effects of Transfusion Therapy in Patients
with Sickle Cell Disease and Strategies for Their Prevention and Management
Immune-Related
Febrile nonhemolytic Prevention: Transfuse leukoreduced blood products.
transfusion reactions Treatment: Administer antipyretic medication.
Alloimmunization to RBC Prevention: Transfuse RBC units matched for antigens most commonly associated with
antigens and delayed alloimmunization.
hemolytic transfusion Treatment: Transfuse RBC units matched for antigens to which antibodies have been made;
reactions provide supportive therapy for delayed reactions.
Autoimmunization to Prevention: None known.
RBC antigens Treatment: Administer corticosteroids with or without intravenous immune globulin.
Alloimmunization to Prevention: Transfuse leukoreduced blood products.
platelet- or HLA-specific Treatment: Transfuse crossmatched or HLA-matched platelet products, if proven indication.
antigens
Nonimmune-Related Prevention: Provide chronic erythrocytapheresis.
Iron overload Treatment: Administer deferoxamine or deferasirox.
Transfusion-transmitted Prevention: Use the most advanced screening tests for donated blood products; do not
infection transfuse unnecessarily.
TRANSFUSION OF THE HEMOGLOBINOPATHY PATIENT
Immune-Related Adverse Effects allogeneic donor RBCs and those of SCD patients, many
centers have suggested that SCD patients receive RBC units
FEBRILE NONHEMOLYTIC TRANSFUSION REACTIONS
matched for antigens most commonly associated with
White blood cells (WBCs) synthesize and release various alloimmunization (Table 38–6).83,85,87
cytokines during storage of cellular blood products that In Vichinsky’s study, comparable to reports by other
may cause fever and chills (febrile nonhemolytic transfusion researchers,83,86–89 66% of all alloantibodies that formed in
reactions) in the transfusion recipient.80 Current technolo- transfused SCD patients were directed against C, E, and K
gies in leukoreduction filtration reduce the leukocyte count RBC antigens.87 See Table 38–6 for a summary of the most
in RBCs to less than 5 × 106 WBC/unit, mitigating the devel- common antigens to which antibodies are made by SCD
opment of febrile nonhemolytic transfusion reactions in patients.
recipients of leukoreduced blood products.80 Because SCD Tahhan and coworkers90 reported that none (0%) of 40
patients typically receive large numbers of transfusions and patients studied who received matched transfusions for C, E,
febrile nonhemolytic transfusion reactions occur in associa- K, Fya, JKb, and S antigens developed alloantibodies, whereas
tion with 0.5% to 1% of transfusions,81 most experts recom- 16 (34.8%) of 46 patients who received both matched and
mend the use of leukoreduced blood products for all SCD nonmatched transfusions became alloimmunized against
patients. In addition, an acute infection or a pain crisis can one or more of these antigens.
manifest with the same symptoms (i.e., fever, chills, malaise) A College of American Pathologists (CAP) study, pub-
as a febrile nonhemolytic transfusion reaction, confounding lished in 2005, surveyed 1182 laboratories to determine the
the clinical picture and possibly delaying appropriate treat- extent of RBC antigen phenotyping and RBC antigen match-
ment of underlying disorders in SCD patients. ing used in the care of nonalloimmunized SCD patients
throughout North America.91 Of the laboratories surveyed
ALLOIMMUNIZATION TO RBC ANTIGENS AND DELAYED
by the CAP, 63% did not routinely phenotype SCD patients
HEMOLYTIC TRANSFUSION REACTIONS
any differently than other patients (ABO and RhD antigen
Alloimmunization to RBC antigens is a common problem phenotyping only). The remaining one third of the North
in transfused SCD patients, leading both to difficulty in American laboratories reported routinely performing addi-
obtaining compatible RBC units and to the development tional RBC phenotyping for SCD patients; 75% of these cen-
of delayed hemolytic transfusion reactions.82–86 Delayed ters gave units phenotypically matched for additional RBC
hemolytic transfusion reactions are especially problematic antigens C, E, and K to SCD patients.
in SCD patients, in that the symptoms of a delayed hemo- Protocols for phenotyping and limited RBC antigen
lytic transfusion reaction can mimic those of a pain crisis matching differ among sickle cell centers. In many academic
and can even lead to a pain crisis,85 complicating both the centers, all SCD patients undergo extensive RBC antigen phe-
clinical diagnosis and the subsequent appropriate treatment notyping before their first transfusion, a policy supported by
of these patients. most experts.86,89–91 Those SCD patients who have not yet
The most comprehensive study of the frequency of made RBC alloantibodies routinely receive RBC units phe- 38
and risk factors associated with alloimmunization in SCD notypically matched for the antigens responsible for most of
patients was performed by Vichinsky and colleagues.87 They the alloantibodies made by SCD patients to allogeneic donor 531
prospectively determined the transfusion history, RBC anti- RBCs, most often C, E, and K antigens.
gen phenotype, and alloantibody development of 107 trans- Once an SCD patient has made an RBC alloantibody, he
fused African American patients with SCD. These results were or she then typically receives RBC units matched for other
compared with those from similar studies in 51 nontrans- RBC antigens to which SCD patients are known to most
fused African American SCD patients and in 19 Caucasian frequently make antibodies. C, E, K, Fya, Jkb, and S are the
patients who had undergone multiple transfusions for other six most common and most clinically significant antibod-
forms of chronic anemia. The results showed that the average ies made by SCD patients (see Table 38–6).87 These concepts
alloimmunization rate for transfused SCD patients was 30%, are supported by most centers who do extended phenotype
compared with only 5% for the multiply transfused patients
with other forms of anemia (p < 0.001). None of the non-
transfused SCD patients developed alloantibodies, and the Table 38–6 Average Frequencies of the
alloimmunization rate in individual SCD patients increased Most Common Red Blood Cell Alloantibodies
exponentially with increasing numbers of transfusions. Of Made by Patients with Sickle Cell Disease
the 32 patients who developed alloantibodies, 17 developed
multiple antibodies, and 12 of these 17 patients had more Antibody Average Frequency (%)
than three different alloantibodies. E 21
After conducting an RBC phenotyping study of local K 18
blood bank donors and comparing these phenotypes with C 14
those of SCD patients and of white patients with other forms Lea 8
Fya 7
of chronic anemia, the authors suggested that the increased Jkb 7
alloimmunization rate in SCD patients most likely resulted D 7
from RBC antigenic differences between the SCD patients Leb 7
(African Americans) and the blood donors (the majority of S 6
whom were Caucasian). Such antigenic differences did not Fyb 5
M 4
exist between blood donors and the multiply transfused e 2
Caucasian patients who had other forms of chronic anemia. c 2
Because of this documented lack of phenotypic compat-
ibility in antigen profile between the majority of volunteer Data from references 74–77.
TRANSFUSION MEDICINE
antigen matching, and antigen-matched transfusion results patients enrolled in the antigen-matching, dedicated donor
in an alloimmunization rate of approximately 1% to 5%, PFL program is significantly lower than the published 30%
a significant decrease compared with the rates observed rate observed in SCD patients who routinely received non-
among SCD patients transfused with RBCs not matched for antigen-matched RBCs, but it is similar to rates observed
these common RBC antigens.9,83,86,87,90,92 in other antigen-matching programs without the dedicated
Limited donor pool programs offer a similar approach donor aspect.86,90,92,95
to preventing, or at least limiting, alloimmunization in SCD Based on Hillyer’s review and the fact that the program
patients. This strategy combats RBC alloimmunization by was expensive and labor intensive (with respect to recruit-
transfusing RBCs only from donors who are ethnically or ment, collection, inventory management, and distribution
antigenically closely matched with the SCD patient, most of dedicated RBC units), this limited donor pool program
often for the four RBC antigens (C, D, E, and K) to which was changed to an antigen-matching program only, using
antibodies are most commonly made. Sosler and colleagues primarily African American donors from the general vol-
suggested a model to reduce alloimmunization in SCD unteer donor pool. This program works well; nearly 13% of
patients by using RBC units only from ethnically similar blood donors in the associated blood center self-identify as
(African American) donors, based on the fact that 93% of African American. In other parts of the country in which the
African Americans shared the typical SCD phenotype of E, general donor pool does not have such a high percentage of
C, Fya, Jkb, and K antigen-negativity, whereas only 7% of the African American donors, special recruitment efforts within
Caucasion population had the same phenotype.93 Ambruso the African American community would need to be made
and associates92 reported that such a limited donor antigen- to procure the required number of phenotypically similar
matching program can diminish by tenfold the incidence RBC units for SCD patients.
of alloimmunization in transfused SCD patients. However, The standard-setting association for transfusion medi-
administrative problems reported in Ambruso’s study cine, the AABB (formerly the American Association of Blood
included difficulties in recruiting eligible donors from the Banks; www.aabb.org) is developing recommended guide-
African American community, in inventory management, lines for limited phenotypic matching RBC transfusion pro-
and in distribution and transfusion of the matched blood, tocols for SCD patients. These guidelines are sensitive to the
as well as increased expense of the program. Tahhan and different approaches to limited antigen-matching protocols
coworkers90 reported that the cost of operating an antigen- for SCD patients used by different centers. The recommen-
matching (not a limited donor) program was 1.5 to 1.8 times dations will review the current literature regarding limited
that of a standard transfusion protocol. RBC antigen matching and its merits, suggest possible strat-
Hillyer and colleagues94 conducted a retrospective review egies for antigen-matching programs, and assist in engaging
of the records of 85 patients who were enrolled in a limited the community of physicians, nurses, transfusion medicine
donor pool antigen-matching program (Partners for Life, experts, blood banks, and blood centers in developing pro-
PFL) between January 1993 and August 2000 and reported grams to suit the needs of their individual hospitals and
III that the average number of donors per PFL patient was 9.5, patients.
average number of RBC units transfused before PFL was
AUTOIMMUNIZATION TO RBC ANTIGENS
532 17.8, and the average number of RBC units transfused dur-
ing participation in PFL was 39.4. Development of autoantibodies to RBC antigens in asso-
The overall alloimmunization rate for PFL patients was ciation with transfusions in certain SCD patients has been
7% (6 of 85 developed new RBC antibodies while enrolled described.96 In 1999, Castellino and associates97 reported on
in PFL). Of the 6 patients who developed new antibodies, the frequency, characteristics, and significance of erythro-
3 had previously identified antibodies at enrollment, and 3 cyte autoantibodies in a large group of multiply transfused
had no previously identified antibodies. One patient with children with SCD.97 The rate of warm (immunoglobulin G)
previously identified antibodies developed anti-V, another RBC autoantibody formation in this group was 7.6%; 29% of
developed anti-Goa, and the third developed anti-Cw and patients with erythrocyte autoantibodies had clinically sig-
anti-Kpa (all antigens not routinely matched for in PFL). nificant hemolysis thought to be caused by the autoantibody.
Two of the three patients without previously formed anti- All patients with clinically significant hemolysis had both
bodies developed anti-Fya (an antigen not matched for in immunoglobulin G and complement detected on the surface
PFL patients without previously formed alloantibodies). The of their RBCs. There was a strong association between auto-
other patient without previously formed antibodies devel- antibody formation and the presence of RBC alloantibodies:
oped an anti-E after being for in PFL 1.5 years with no anti- 86% of patients with autoantibodies also had one or more
body formation; however, she was known to have received alloantibodies.
nonantigen-matched RBCs at a nonparticipating institution The phenomenon of RBC autoantibody formation in
immediately prior to her anti-E formation. association with blood transfusion is not well understood,
Although this limited donor pool program was successful but several theories exist. Alloantibodies may bind to trans-
in mitigating RBC alloimmunization (7% overall rate) for fused cells and cause conformational changes in the RBC
this group of patients, it was not successful in limiting the antigenic epitopes, leading to stimulation of autoantibody
exposure of SCD patients to only this small, dedicated blood formation.97 Alternatively, some SCD patients may simply
donor pool of 9 to 10 donors, due to technical difficulties in have a predisposition to develop RBC autoantibodies, per-
having these units available at the time they were needed by haps because of an overall dysfunction of their immune
the patients. Only 6% of PFL patients received all of their systems.97,98 For example, the loss of a functional spleen
RBC units from their dedicated donors. The majority (57%) in patients with SCD could lead to immune dysregula-
received a combination of RBC units from their dedicated tion, because some experimental evidence suggests that the
donors and antigen-matched RBC units from the general spleen is important in the regulation of RBC autoantibody
volunteer donor pool. The 7% alloimmunization rate for formation.97,99
TRANSFUSION OF THE HEMOGLOBINOPATHY PATIENT
Whatever the cause, physicians should be aware that a groups reported that erythrocytapheresis does not obviate the
syndrome of clinically significant post-transfusion hemoly- need for chelation therapy in those patients with previously
sis may occur in SCD patients in which both autologous and accumulated iron.20–22 In general, ferritin levels decreased in
transfused RBCs are destroyed (bystander hemolysis) and patients undergoing chronic erythrocytapheresis who received
that hemolysis may be exacerbated by further transfusions.100 concurrent chelation therapy, and they either mildly decreased
Serologic findings may be negative or simply not helpful in or stabilized in patients who were not receiving chelation ther-
identification of the autoantibody.97 In most cases, cortico- apy. However, at-risk patients who began erythrocytapheresis
steroids with or without intravenous immunoglobulin are without a long history of previous chronic simple transfu-
beneficial in slowing hemolysis and allowing for successful sions maintained very low serum ferritin levels not requiring
continuation of necessary transfusions.100 chelation therapy.20–22
Therefore, it appears that chronic erythrocytapheresis
ALLOIMMUNIZATION TO HLA-SPECIFIC
may be most beneficial when it is initiated early in the course
OR PLATELET-SPECIFIC ANTIGENS
of chronic transfusion therapy, before significant iron accu-
As bone marrow/stem cell transplantation (BMT) becomes mulation occurs. Nevertheless, chronic erythrocytapheresis
a viable option for selected patients with SCD,101–107 alloim- does appear to stabilize or decrease serum ferritin levels in
munization to platelets presents more serious problems for patients who have already developed significant iron over-
this group. Friedman and coworkers108 reported that 85% load and continue on chelation therapy.20–22
of SCD patients receiving 50 or more transfusions, 48% of The primary potential problems with the chronic eryth-
SCD patients receiving 1 to 49 transfusions, and no non- rocytapheresis transfusion protocol (compared with chronic
transfused SCD patients demonstrated alloimmunization to simple transfusion protocols) are increased blood product
human leukocyte antigen (HLA)-specific or platelet-specific exposure, with concomitant increased risks of alloimmuni-
antigens. Because platelet refractoriness is a serious compli- zation to RBCs and platelets and of transfusion-transmitted
cation during BMT, prevention of platelet alloimmunization infection, and the increased cost (i.e., increased numbers of
appears prudent in this group of patients. blood products used, increased cost of phenotypically simi-
Most transfusion experts support the use of leukoreduced lar units, if chosen for transfusion, and the added cost of the
cellular blood products to prevent or reduce platelet allo- automated procedures themselves).
immunization and refractoriness in SCD patients, a prac- The four published reports19–22 indicated that SCD
tice employed for a variety of multiply transfused patient patients’ blood product exposures increase in chronic eryth-
groups.109–112 rocytapheresis protocols, with reported increases in blood
utilization rates ranging from 52% to almost 100% (i.e., one
Nonimmune-Related Adverse Effects to two times more RBC units transfused than with previ-
ous simple transfusions of the same patients). However, of
IRON OVERLOAD
the combined 43 patients studied, only 1 patient developed
Iron overload resulting in hemosiderosis is a serious long- an alloantibody20 during the period of erythrocytapheresis 38
term complication of chronic transfusion in SCD patients. treatment. Three of these centers used antigen-matched RBC
The most informative reports regarding transfusion-associ- units for the erythrocytapheresis procedures: Singer and 533
ated iron overload were described in patients with thalassemia; associates22 matched for C, E, and K; Hilliard and colleagues21
see “Thalassemias” for a discussion of the pathophysiology for C, E, K, Fya, and Jkb; and Adams and associates20 for C, E,
of iron overload. Patients who develop iron overload may K, and Jkb. When evaluating the very low alloimmunization
be treated with long-term chelation therapy in the form of rates that have been reported for chronic erythrocytapheresis
deferoxamine; however, this therapy is expensive, and due to protocols, it is important to realize that the majority of SCD
multiple side effects and a history of difficulty of administra- patients studied received RBC units matched for at least the
tion of the drug, the compliance rate has been notoriously C, E, and K antigens.
poor among patients.113 However, with the November 2005 The high cost of erythrocytapheresis is an important issue.
FDA approval of the first oral iron chelator available in the Hilliard and colleagues21 compared the total cost of 1 year
United States,114 compliance to long-term chelation therapy of erythrocytapheresis ($36,085) with the total annual cost
may improve as a result of increased ease of use. for simple transfusion ($26,058) and found an economically
One potential transfusion methodology for the preven- significant difference. They suggested that the added cost of
tion of iron overload currently being investigated in SCD chelation therapy ($29,480) with simple transfusion (for a
patients is chronic erythrocytapheresis. Chronic erythro- total of $62,143) makes erythrocytapheresis without chela-
cytapheresis procedures may be performed at 3- to 4-week tion a much less expensive alternative. However, for patients
intervals. In contrast to simple transfusions, the patient’s who have significant iron accumulation at the time erythro-
own sickled RBCs are removed while an equal volume of cytapheresis therapy is initiated, chelation therapy must be
normal donor RBCs is infused. The obvious potential benefit continued to achieve serum ferritin level reduction or stabi-
of chronic erythrocytapheresis compared with simple trans- lization.19–21 This cost comparison provides further evidence
fusion is the prevention of long-term iron accumulation and that, if it is technically feasible, early initiation of chronic
hemosiderosis. erythrocytapheresis in SCD patients before significant iron
Although it has not been universally implemented, chronic accumulation occurs may be preferable to long-term chronic
erythrocytapheresis appears to be clinically effective in reduc- simple transfusion and the resulting complications of iron
ing iron overload in chronically transfused SCD patients. Four overload and the need for chelation therapy.
investigative teams19–22 have described their individual experi-
TRANSFUSION-TRANSMITTED INFECTIONS
ences with chronic erythrocytapheresis transfusion protocols
for SCD patients. All suggested that erythrocytapheresis does The risk of transmission via transfusion of most known infec-
limit iron accumulation in SCD patients, but three of the four tious agents, particularly viruses, has become substantially
TRANSFUSION MEDICINE
reduced in the developed world since the 1990s,115,116 largely A primary consequence of the ineffective erythropoiesis is
owing to improved screening tests for donated blood prod- increased iron absorption and progressive accumulation of
ucts. Because the risk of contracting the most harmful trans- iron in tissues. Anemia, increased erythropoiesis, and hyper-
fusion-transmitted diseases in the United States and many splenism also cause marked expansion of the plasma volume
other developed countries is quite low (e.g., an estimated and blood volume. Extramedullary hematopoiesis can cause
1 in 2 to 4 million chance of contracting human immuno- pressure symptoms from perivertebral masses. Expanded
deficiency virus [HIV] per RBC unit transfused116), it is a marrow activity causes skeletal changes, including osteopenia
generally accepted practice that blood products should not and characteristic deformities in the skull and face.123
be withheld (if an appropriate clinical indication for trans-
fusion exists) for the sole purpose of preventing a low-risk Goals of Transfusion Therapy
transfusion-transmitted disease.
In developing countries, however, the same cannot be said; The goals of therapy in β-thalassemia major are to increase
in some countries in Africa, such as Zambia and Botswana, oxygen-carrying capacity by correcting the anemia, pre-
HIV may affect upward of one third of the country’s popula- venting progressive hypersplenism, suppressing erythro-
tion.117 However, the risks of all adverse effects of transfusion poiesis, and reducing increased gastrointestinal absorption
(see Table 38–5), including transfusion-transmitted disease, of iron.120 Transfusion therapy is begun early in life to ame-
should be balanced against the clinical need for transfusion liorate the symptoms and signs of anemia and to support
on a case-by-case basis. normal growth and development. Adequate transfusion
reduces progression of hypersplenism, delaying the need
for splenectomy.124 Suppression of erythropoiesis prevents
THALASSEMIAS skeletal changes, prevents complications of extramedullary
hematopoiesis, and reduces pathologic fractures and other
complications from osteopenia.125 Suppression of ineffective
Pathogenesis and Clinical Pathology
erythropoiesis decreases transfusion requirements by reduc-
Thalassemias are among the most prevalent genetic disor- ing blood volume126 and suppresses the increased intestinal
ders caused by a single gene, occurring in high frequency in absorption of iron.127
Southeast Asia, southern China, Indonesia, India, the Middle
East, Africa, and the Mediterranean basin.118 α-thalassemias Indications for Transfusion
are caused by mutations that reduce the synthesis of the α-
globin chain of hemoglobin, and β-thalassemias from muta- Transfusion therapy is initiated in childhood when the symp-
tions that reduce β-globin synthesis. The complex molecular toms and signs of anemia are present, including growth
genetics of these disorders is beyond the scope of this dis- retardation and failure to thrive. Transfusions are occasion-
cussion and has been reviewed elsewhere.119–121 The only ally initiated in β-thalassemia intermedia and hemoglobin H
III definitive cure for β-thalassemia major is BMT, which has disease to prevent facial and skull deformity from expansion
cured more than 1000 individuals with β-thalassemia major of the medullary bone space. Progressive hypersplenism may
534 worldwide. BMT is considered in all β-thalassemia major require transfusion to postpone splenectomy in β-thalasse-
children with a suitable donor. mia intermedia. Splenectomy is indicated to prevent excessive
By definition, chronic transfusion is required to maintain transfusion requirements.125
wellness in homozygous individuals and compound hetero-
zygotes with β-thalassemia major. Individuals with β-thalas- Methods of Transfusion
semia intermedia have complex genetics and may have severe
anemia that requires episodic transfusion. Hemoglobin H Simple transfusion of leukoreduced RBCs to maintain a
disease is caused by structural or functional loss of globin Hb level greater than 9.5 g/dL is the standard approach to
synthesis from three of the normal four α-globin genes. transfusion in thalassemia.121 The older practice of main-
Individuals usually are not transfusion dependent, but taining higher pretransfusion Hb levels124–126 has been
they may require transfusion support for complications. associated with excessive iron loading and generally is not
Compound heterozygotes with β-thalassemia and hemo- advocated.120,128,129 Splenectomy is recommended when
globin E disease, which is very common in individuals from hypersplenism increases the transfusion requirement beyond
East Asia, can have clinical manifestations ranging from 200 to 250 mL/kg/year.124,129
mild microcytic anemia to severe transfusion-dependent β- Another approach to reducing iron loading has been the
thalassemia major.122 use of neocytes prepared by cell separators, cell processors, or
Clinical manifestations of severe β-thalassemias include other density means.130–137 The use of “neocyte” transfusions
transfusion-dependent anemia, expansion of medullary bone was shown to allow a 15% extension in transfusion inter-
and extramedullary hematopoiesis, severe iron overload, val while maintaining the same pretransfusion Hb level.137
increased infections, retarded growth and development, and The costs of this approach are increased blood use, increased
osteopenia.120 Anemia results from severe ineffective erythro- donor unit exposure, and an estimated fivefold increase in
poiesis and hemolysis. The excess α-globin chain from unbal- the cost of transfusion.137 This approach is not presently
anced globin chain synthesis leads to arrested development advocated by most thalassemia experts.
and accelerated intramedullary apoptosis. Erythropoietic Erythrocytapheresis has also been applied to thalassemia
activity may be increased tenfold, but more than 95% of the transfusion therapy. The use of automated RBC exchange,
erythropoietin produced may be ineffective. Precipitated returning patient and donor “neocytes” and removing the
α-globin in mature erythrocytes alters membrane proteins “gerocytes,” resulted in a 30% reduction in RBC transfusion
and causes oxidant damage, shortening RBC survival time. requirement and a 43% increase in transfusion interval.138,139
Progressive hypersplenism further reduces RBC survival. Further clinical trials of this approach are warranted.
TRANSFUSION OF THE HEMOGLOBINOPATHY PATIENT
Adverse Effects of Transfusion patients. Ethnic differences have been proposed, but there is
little data to support this claim.144–148 Leukoreduction has
Iron Overload been suggested as a possible means of decreasing alloim-
A primary consequence of the ineffective erythropoiesis is munization in thalassemics, but no data exists to support
increased iron absorption and progressive accumulation of this hypothesis. Splenectomy has also been suggested as a
iron in tissues. Anemia, increased erythropoiesis, and hyper- trigger for increased alloimmunization to RBCs, but Singer
splenism also cause marked expansion of the plasma volume and colleagues found that although such a trend existed in
and blood volume. Extramedullary hematopoiesis can cause their group of patients studied, it did not reach statistical
pressure symptoms from perivertebral masses. Expanded significance.149
marrow activity causes skeletal changes, including osteope- Risks of alloimmunization in thalassemic patients include
nia and characteristic changes in the skull and face.123 delay of transfusion due to difficulties in finding compat-
Individuals with β-thalassemia major are transfusion- ible RBC units, if patients make multiple alloantibodies; a
dependent from early life; indeed, chronic transfusion sup- potential for hemolytic disease of the newborn in thalasse-
port has markedly improved the prognosis of patients with mic patients who become pregnant; and a possible relation-
thalassemias. Transfused children develop iron overload from ship between RBC alloantibody development and secondary
increased absorption caused by ineffective erythropoiesis RBC autoantibody development and/or autoimmune hemo-
and iron administered during transfusion. Cardiac, hepatic, lytic anemia.149–151
and endocrine failure results in death in the second or third Some studies have shown that limited phenotypic RBC
decade without effective therapy to remove excess iron. antigen matching reduces the rate of alloimmunization in
Although no ideal approach to treatment of transfusion- thalassemia patients, particularly in those patients whose
related iron overload exists, until recently, the treatment first transfusion occurred after age 12 months.145,146,149 More
of choice consisted of subcutaneous infusion of deferox- data is required, however, before evidence-based recommen-
amine over 8 to 12 hours, 5 to 7 days a week. This approach dations can be made regarding the feasibility of RBC pheno-
appeared to control iron accumulation and prevent cardiac typic matching for thalassemic patients.
and liver damage, thereby improving life expectancy in
patients with thalassemia. However, the difficulty of admin-
istration of deferoxamine and the many side effects caused a REFERENCES
poor compliance rate in most patients.
On November 9, 2005, the FDA announced the approval of 1. Heller P, Best WR, Nelson RB, et al. Clinical implications of sickle cell
trait and glucose-6-phosphate dehydrogenase deficiency in hospital-
deferasirox, the first orally absorbed iron chelator developed ized black male patients. NEJM 1979;300:1001–1005.
to treat chronic iron overload due to multiple blood trans- 2. Beutler E. The sickle cell diseases and related disorders. In Beutler E,
fusions available for use in the United States.114 Deferasirox Lichtman MA, Coller BS, et al (eds). Williams Hematology, 6th ed.
was approved through the FDA’s accelerated approval pro- New York, McGraw-Hill, 2001, p 585.
3. Bunn HF. Mechanisms of disease: pathogenesis and treatment of sickle
gram; further clinical studies from the company who makes 38
cell disease. NEJM 1997;337:762–769.
the drug are required to maintain the drug’s approval and 4. Steinberg MH. Management of sickle cell disease. NEJM 1999;340:
confirm clinical benefits. Clinical trials of less than 1 year’s 1021–1030. 535
duration showed reduction in hepatic iron concentration fol- 5. Platt OS, Brambilla DJ, Rosse WF, et al. Mortality in sickle cell dis-
lowing once-daily (oral) deferasirox therapy in patients who ease: life expectancy and risk factors for early death. NEJM 1994;330:
1639–1644.
had been receiving multiple transfusions. Side effects such as 6. Reed W, Vichinsky EP. New considerations in the treatment of sickle
nausea and abdominal pain and other less common difficul- cell disease. Annu Rev Med 1998;49:461–474.
ties were noted.140 However, the greater ease of administra- 7. Schmalzer EA, Lee JO, Brown AK, et al. Viscosity of mixtures of
tion of this drug and its apparent efficacy with typically mild sickle and normal red cells at varying hematocrit levels. Transfusion
1987;27:228–233.
side effects may markedly improve the care and treatment 8. Eckman JR. Techniques for blood administration in sickle cell patients.
of patients with transfusion-dependent iron overload.141 In Semin Hematol 2001;38:23–29.
other countries, such as India, different oral iron chelators 9. Wayne AS, Kevy SV, Nathan DG. Transfusion management of sickle
have been widely used, safely and with good success, for the cell disease. Blood 1993;81:1109–1123.
treatment of transfusion-dependent iron overload in patients 10. Sharon BI. Transfusion therapy in congenital hemolytic anemias.
Hematol Oncol Clin North Am 1994;8:1053–1086.
with thalassemia.142 11. Vengelen-Tyler V (ed). AABB Technical Manual, 13th ed. Bethesda,
Md., American Association of Blood Banks, 1999, p 486.
Alloimmunization to RBC Antigens 12. Embury SH, Vichinsky EP. Sickle Cell Disease. In Hoffman R, Benz EJ
Reported rates of alloimmunization to RBC antigens in Jr, Shattil SJ, et al (eds). Hematology: Basic Principles and Practice, 3rd
ed. Philadelphia, Churchill Livingstone, 2000, p 539.
chronically transfused thalassemia patients range from 5% to 13. Ohene-Frempong K. Indications for red cell transfusion in sickle cell
30%.143–147 The largest of these studies followed alloimmuni- disease. Semin Hematol 2001;38:5–13.
zation rates in 973 Greek patients, primarily children, and 14. Piomelli S, Seaman C, Ackerman K, et al. Planning an exchange trans-
found an alloimmunization rate of 23% in this group.146 fusion in patients with sickle cell syndromes. Am J Pediatr Hematol
Some evidence exists that thalassemia major patients who Oncol 1990;12:268–276.
15. Hillyer KL, Hillyer CD. Packed red blood cells and related products. In
begin chronic transfusion therapy at younger ages develop Hillyer CD, Hillyer KL, Strobl FJ, et al (eds). Handbook of Transfusion
fewer alloantibodies.145,146 Michail-Merianou and colleagues Medicine. Philadelphia, Academic Press, 2001, p 29.
reported that children who began transfusion therapy before 16. Strauss R. Data-driven blood banking practices for neonatal RBC
age 12 months had a 9% alloimmunization rate; children transfusions. Transfusion 2000;40:1528–1540.
17. Hall T, Barnes A, Miller J, et al. Neonatal mortality following trans-
who started chronic transfusions after age 12 months had a fusion of red cells with high plasma potassium levels. Transfusion
39% alloimmunization rate.145 1993;33:606–609.
There are no agreed-upon hypotheses as to what specifi- 18. Cohen A, Kron E, Brittenham G. Toxicity of transfusional iron over-
cally affects increased alloimmunization rates in thalassemia load in sickle cell anemia. Blood 1984;64(Suppl 1):47a.
TRANSFUSION MEDICINE 19. Kim H, Dugan N, Silber J. Erythrocytapheresis therapy to reduce iron 49. Hassell KL, Eckman JR, Lane PA. Acute multiorgan failure syndrome:
overload in chronically transfused patients with sickle cell disease. a potentially catastrophic complication of severe sickle cell pain epi-
Blood 1994;83:1136–1142. sodes. Am J Med 1994;96:155–162.
20. Adams D, Schultz W, Ware R, et al. Erythrocytapheresis can reduce 50. Powars D, Wilson B, Imbus C, et al. The natural history of stroke in
iron overload and prevent the need for chelation in chronically sickle cell disease. Am J Med 1978;65:461–471.
transfused pediatric patients. J Pediatr Hematol Oncol 1996;18: 51. Russell MO, Goldberg HI, Hodson A, et al. Effect of transfusion ther-
46–50. apy on arteriographic abnormalities and on recurrence of stroke in
21. Hilliard LM, Williams BF, Lounsbury AE, et al. Erythrocytapheresis sickle cell disease. Blood 1984;63:162–169.
limits iron accumulation in chronically transfused sickle cell patients. 52. Peglow CH, Adams RJ, McKie V, et al. Risk of recurrent stroke in
Am J Hematol 1998;59:28–35. patients with sickle cell disease treated with erythrocyte transfusion. J
22. Singer ST, Quirolo K, Nishi K, et al. Erythrocytapheresis for chroni- Pediatr 1995;126:896–899.
cally transfused children with sickle cell disease: an effective method 53. Williams J, Goff JR, Anderson HR Jr, et al. Efficacy of transfusion
for maintaining a low hemoglobin S level and reducing iron overload. therapy for one to two years in patients with sickle cell disease and
J Clin Apheresis 1999;14:122–125. cerebrovascular accidents. J Pediatr 1980;96:205–209.
23. Lawson SE, Oakley S, Smith NA, et al. Red cell exchange in sickle cell 54. Cohen AR, Martin MB, Silber JF, et al. A modified transfusion program
disease. Clin Lab Haematol 1999;21:99–102. for prevention of stroke in sickle cell disease. Blood 1992;79:1657–1661.
24. Vichinsky EP. Current issues in blood transfusion in sickle cell disease. 55. U.S. Department of Health and Human Services. National Institutes
Semin Hematol 2001;38:14–22. of Health. National Heart, Lung, and Blood Institute. Clinical Alert,
25. Embury SH. The clinical pathophysiology of sickle cell disease. Ann December 5, 2004. Available at http://www.nhlbi.nih.gov/health/prof/
Rev Med 1986;37:361–376. blood/sickle/clinical-alert-scd.htm.
26. Goldstein AR, Anderson MJ, Serjeant GR. Parvovirus-associated 56. Wang WC, Kovnar EH, Tonkin IL, et al. High risk of recurrent stroke
aplastic crisis in homozygous sickle cell disease. Arch Dis Child 1987; after discontinuance of five to twelve years of transfusion therapy in
62:585–588. patients with sickle cell disease. J Pediatr 1991;118:377–382.
27. Serjeant GR, de Ceulaer K, Maude GH. Stilboestrol and stuttering 57. Adams R, McKie V, Nichols F, et al. The use of transcranial ultrasonog-
priapism in homozygous sickle-cell disease. Lancet 1985;2:1274. raphy to predict stroke in sickle cell disease. NEJM 1992;326:605–610.
28. Vichinsky E. Transfusion. In Embury SH, Hebbel RP, Mohandas N, et 58. Adams RJ, McKie VC, Carl EM, et al. Long-term stroke risk in children
al (eds). Sickle Cell Disease: Basic Principles and Clinical Practice. New with sickle cell disease screened with transcranial Doppler. Ann Neurol
York, Raven Press, 1994, pp 261–283. 1997;42:699–704.
29. Castro W, Brambilla DJ, Thorington B, et al. The acute chest syn- 59. Koshy M, Burd L, Wallace D, et al. Prophylactic red cell transfusions in
drome in sickle cell disease: incidence and risk factors. Blood 1994;84: pregnant patients with sickle cell disease. NEJM 1988;319:1447–1452.
643–649. 60. El-Shafei AM, Dhaliwal JK, Sandhu AK, et al. Indications for blood
30. Quinn C, Rogers Z, Buchanan G. Survival of children with sickle cell transfusion in pregnancy with sickle cell disease. Aust NZ J Obstet
disease. Blood 2004;103:4023–4027. Gynaecol 1995;35:405–408.
31. Gray J, Anionwu EN, Davies SC, et al. Patterns of mortality in sickle 61. Tuck SM, James CE, Brewster EM, et al. Prophylactic blood transfusion
cell disease in the United Kingdom. J Clin Pathol 1991;44:459. in maternal sickle cell syndromes. BJOG 1987;94:121–125.
32. Dreyer ZE. Chest infections and syndromes in sickle cell disease of 62. Koshy M. Sickle cell disease and pregnancy. Blood Rev 1995;9:
childhood. Semin Respir Infect 1996;11:163–172. 157–164.
33. Davies SD, Luce PJ, Win AA, et al. Acute chest syndrome in sickle-cell 63. Morgan AG, Gruber CA, Serjeant GR. Erythropoietin and renal func-
disease. Lancet 1984;1:36–38. tion in sickle cell disease. BMJ 1982;285:1686–1688.
34. Mallouh AA, Asha M. Beneficial effect of blood transfusion 64. Weil JV, Castro O, Malik AR, et al. Pathogenesis of lung disease in sickle
on children with sickle cell chest syndrome. Am J Dis Child hemoglobinopathies. Am Rev Respir Dis 1993;148:249–256.
III 1988;142:178–182. 65. Simmons B, Santhanam V, Castaner A, et al. Sickle cell heart disease:
35. Emre U, Miller ST, Gutierez M, et al. Effect of transfusion in acute Two-dimensional echo and Doppler ultrasonographic findings in
536 chest syndrome of sickle cell disease. J Pediatr 1995;127:901–904. the hearts of adult patients with sickle cell anemia. Arch Intern Med
36. Emond AM, Collis R, Darvill D, et al. Acute splenic sequestration in 1988;148:1526–1528.
homozygous sickle cell disease: natural history and management. J Pe- 66. Sutton L, Castro O, Cross D, et al. Pulmonary hypertension in sickle
diatr 1985;107:201–206. cell disease. Am J Cardiol 1994;74:626–628.
37. Seeler RA, Shwiaki MZ: Acute splenic sequestration crises (ASSC) in 67. Castro O. Systemic fat embolism and pulmonary hypertension in sickle
young children with sickle cell anemia. Clin Pediatr 1972;11:701–704. cell disease. Hematol Oncol Clin North Am 1996;10:1289–1303.
38. Sheehy TW, Law DE, Wade BH. Exchange transfusion for sickle cell 68. Vichinsky E. Pulmonary hypertension in sickle cell disease. NEJM
intrahepatic cholestasis. Arch Intern Med 1980;140:1364–1366. 2004;350:857–859.
39. Ohene-Frempong K, Weiner SJ, Sleeper LA, et al. Cerebrovascular 69. Hankins J, Jeng M, Harris S, et al. Chronic transfusion therapy for
accidents in sickle cell disease: Rates and risk factors. Blood 1998;91: children with sickle cell disease and recurrent acute chest syndrome. J
288–294. Pediatr Hematol Oncol 2005;27:158–161.
40. Russell MO, Goldberg HI, Reis L, et al. Transfusion therapy for 70. Gladwin M, Sachdev V, Jison M, et al. Pulmonary hypertension
cerebrovascular abnormalities in sickle cell disease. JAMA 1979;242: as a risk factor for death in patients with sickle cell disease. NEJM
2317–2318. 2004;350:886–895.
41. Adams R, McKie V, Brambilla D, et al. Stroke prevention trial in sickle 71. Rifkind S, Waisman J, Thompson R, et al. RBC exchange pheresis for
cell anemia. Control Clin Trials 1998;19:100–129. priapism in sickle cell disease. JAMA 1979;242:2317–2318.
42. Adams RJ, McKie VC, Hsu L, et al. Prevention of a first stroke by trans- 72. Walker EM, Mitchum EN, Rous SN, et al. Automated erythrocytapher-
fusions in children with sickle cell anemia and abnormal results on esis for relief of priapism in sickle cell hemoglobinopathies. J Urol
transcranial Doppler ultrasonography. NEJM 1998;339:5–11. 1983;130:912–916.
43. Vichinsky EP, Haberkern CM, Neumayr L, et al. A comparison of 73. Seeler RA. Intensive transfusion therapy for priapism in boys with
conservative and aggressive transfusion regimens in the perioperative sickle cell anemia. J Urol 1973;110:360–361.
management of sickle cell disease. NEJM 1995;333:206–213. 74. Seeler RA. Priapism in children with sickle cell anemia. Clin Pediatr
44. Koshy M, Weiner SJ, Miller ST, et al. Surgery and anesthesia in sickle 1971;10:418–419.
cell disease. Blood 1995;86:3676–3684. 75. Miller ST, Rao SP, Dunn KE, et al. Priapism in children with sickle cell
45. Haberkern CM, Neumayr LD, Orringer EP, et al. Cholecystectomy disease. J Urol 1995;154:844–847.
in sickle cell anemia patients: perioperative outcome of 364 cases 76. Baruchel S, Rees J, Bernstein ML, et al. Relief of sickle cell pria-
from the National Preoperative Transfusion Study. Blood 1997;89: pism by hydralazine: report of a case. Am J Pediatr Hematol Oncol
1533–1542. 1993;15:115.
46. U.S. Department of Health and Human Services. National Institutes of 77. Siegel JF, Rich MA, Brock WA. Association of sickle cell disease, pria-
Health. The Management of Sickle Cell Disease (ed 4). NIH Publica- pism, exchange transfusion and neurological events: ASPEN syn-
tion 02–2117. Bethesda, Md., NIH, 2002. drome. J Urol 1993;150:1480–1482.
47. Riddington C, Williamson L. Preoperative blood transfusions for sickle 78. Platt OS, Thorington BD, Brambilla DJ, et al. Pain in sickle cell disease.
cell disease. Cochrane Database Syst Rev 2001; 3:CD003149. NEJM 1991;325:11–16.
48. Serjeant GR. Sickle Cell Disease, 2nd ed. Oxford, Oxford Medical 79. Koshy M, Entsuah R, Koranda A, et al. Leg ulcers in patients with sickle
Publishers, 1992, p 332. cell disease. Blood 1989;74:1403–1408.
TRANSFUSION OF THE HEMOGLOBINOPATHY PATIENT
80. Miller JP, AuBuchon JP. Leukocyte-reduced and cytomegalovirus- 106. Ferster A, Corazza F, Vertongen F, et al. Transplanted sickle-cell dis-
reduced-risk blood components. In Mintz PD (ed). Transfusion ease patients with autologous bone marrow recovery after graft failure
Therapy: Clinical Principles and Practice. Bethesda, Md., AABB Press, develop increased levels of fetal haemoglobin which corrects disease
1999, p 313. severity. Br J Haematol 1995;90:804–808.
81. Stack G, Judge JV, Snyder EL. Febrile and nonimmune transfusion 107. Souillet G. Indications and results of progenitor cell transplant in con-
reactions. In Rossi EC, Simon TL, Moss GS, et al (eds). Principles of genital haemopathies (except Fanconi anaemia). Bone Marrow Trans-
Transfusion Medicine, 2nd ed. Baltimore, Williams & Wilkins, 1996, plant 1998;21:S28–S33.
pp 773–784. 108. Friedman D, Lukas M, Jawad A, et al. Alloimmunization to platelets
82. Orlina AR, Unger PJ, Koshy M. Post-transfusion alloimmunization in in heavily transfused patients with sickle cell disease. Blood 1996;88:
patients with sickle cell disease. Am J Hematol 1978;5:101–106. 3216–3222.
83. Davies SC, McWilliam AC, Hewitt PE, et al. Red cell alloimmunization 109. Slichter SJ. Platelet transfusion therapy. Hematol Oncol Clin North
in sickle cell disease. Br J Hematol 1986;63:241–245. Am 1990;4:291–311.
84. Reisner EG, Kostyo DD, Phillips G, et al. Alloantibody responses 110. Sniecinski I, O’Donnell MR, Nowicki B, et al. Prevention of refractori-
in multiply transfused sickle cell patients. Tissue Antigens 1987;30: ness and HLA-alloimmunization using filtered blood products. Blood
161–166. 1988;71:1402–1407.
85. Cox JV, Steane E, Cunningham G, et al. Risk of alloimmunization and 111. Saarinen UM, Kekomaki R, Simes MA, et al. Effective prophylaxis
delayed hemolytic transfusion reactions in patients with sickle cell against platelet refractoriness in multitransfused patients by use of
disease. Arch Intern Med 1988;148:2485–2489. leukocyte-free blood components. Blood 1990;75:512–517.
86. Rosse WF, Gallagher D, Kinney T. Transfusion and alloimmunization 112. Brand A, Claas FHJ, Voogt PJ, et al. Alloimmunization after leuko-
in sickle cell disease. Blood 1990;76:1431–1437. cyte-depleted multiple random donor platelet transfusions. Vox Sang
87. Vichinsky EP, Earles A, Johnson RA, et al. Alloimmunization in sickle 1988;54:160–166.
cell anemia and transfusion of racially unmatched blood. NEJM 113. Cohen AR, Martin MB: Iron chelation therapy in sickle cell disease.
1990;322:1617–1621. Semin Hematol 2001;38(Suppl 1):69–72.
88. Coles SM, Klein HG, Holland PV: Alloimmunization in two multi- 114. U.S. Food and Drug Administration. FDA Approves First Oral Drug for
transfused patient populations. Transfusion 1981;21:462–466. Chronic Iron Overload. FDA News Release P05–86, November 9, 2005.
89. Luban NL. Variability in rates of alloimmunization in different groups Available at http://www.fda.gov/bbs/topics/news/2005/NEW01258.html.
of children with sickle cell disease: effect of ethnic background. Am J 115. Schrieber G, Busch M, Kleinman S, et al. The risk of transfusion-trans-
Pediatr Hematol Oncol 1989;11:314–319. mitted viral infections. NEJM 1996;334:1685–1690.
90. Tahhan HR, Holbrook CT, Braddy LR, et al. Antigen-matched donor 116. Stramer SL, Glynn SA, Kleinman SH, et al. Detection of HIV-1 and
blood in the transfusion management of patients with sickle cell HCV infections among antibody-negative blood donors by nucleic
disease. Transfusion 1994;34:562–569. acid amplification testing. NEJM 2004;351:760–768.
91. Osby M, Shulman I. Phenotype matching of donor red blood cell 117. World Health Organization. Regional Office for Africa, Harare,
units for non-alloimmunized sickle cell disease patients: a survey Zimbabwe. HIV/AIDS Epidemiological Surveillance Update for the
of 1182 North American Laboratories. Arch Path Lab Med 2005; WHO African Region 2002, September 2003. Available at http://www.
129:190–193. afro.who.int/aids/surveillance/resources/hiv_surveillance_report_
92. Ambruso DR, Githens JH, Alcorn R, et al. Experience with donors 2002.pdf.
matched for minor blood group antigens in patients with sickle cell 118. Weatherall DJ, Clegg JB. Thalassemia: a global public health problem.
anemia who are receiving chronic transfusion therapy. Transfusion Nat Med 1996;2:847–849.
1987;27:94–98. 119. Higgs DR. Molecular mechanisms in α thalassemia. In Steinberg MH,
93. Sosler S, Jilly B, Saporito C, et al. A simple, practical model for reduc- Forget BG, Higgs DR, et al (eds). Disorders of Hemoglobin: Genetics,
ing alloimmunization in patients with sickle cell disease. Am J Hema- Pathophysiology, and Clinical Management. Cambridge, Cambridge
tol 1993;43:103–106. University Press, 2001, p 405.
38
94. Hillyer KL, Hare VW, Eckman JR, et al. Decreased alloimmunization 120. Oliveri NF. The β-thalassemias. NEJM 1999;341:99–100.
rates in chronically-transfused sickle cell disease patients in a directed- 121. Higgs DR. α-Thalassemia. Baillieres Clin Haematol 1993;6:117–150. 537
donor, red blood cell antigen-matching program entitled “Partners for 122. Weatherall DJ. Hemoglobin E beta-thalassemia: an increasingly com-
Life” (PFL). Blood 2001;98:2274. mon disease with some diagnostic pitfalls. J Pediatr 1998;132:765–767.
95. Vichinsky EP, Luban NL, Wright E, et al. Prospective red blood cell 123. Rioja L, Girot R, Garabedian M, et al. Bone disease in children with
phenotype matching in a stroke-prevention trial in sickle cell anemia: homozygous beta-thalassemia. Bone Miner 1990;8:69–86.
a multicenter transfusion trial. Transfusion 2001;41:1086–1092. 124. Piomelli S, Graziano J. Reduction in iron overload in thalassemia.
96. Wenz B, Gurtlinger A, Wheaton D, et al. A mimicking red blood cell Birth Defects 1982;18:339–346.
autoantibody accompanying transfusion and alloimmunization. 125. Fosburg MT, Nathan DG. Treatment of Colley’s anemia. Blood 1990;
Transfusion 1982;22:147–150. 76:435–444.
97. Castellino SM, Combs MR, Zimmerman SA, et al. Erythrocyte auto- 126. Propper R, Button L, Nathan D. New approach to the transfusion
antibodies in paediatric patients with sickle cell disease receiving management of thalassemia. Blood 1980;55:55–60.
transfusion therapy: frequency, characteristics, and significance. Br J 127. Pootrakul P, Kitcharoen K, Yansukon P, et al. The effect of erythroid
Haematol 1999;104:189–194. hyperplasia on iron balance. Blood 1988;71:1124–1129.
98. Test ST, Woolworth VS: Defective regulation of complement by the 128. Cazzola M, Borgna-Pignatti C, Locatelli F, et al. A moderate transfu-
sickle erythrocyte: evidence for a defect in control of membrane attack sion regimen may reduce iron loading in β-thalassemia major with-
complex formation. Blood 1994;83:842–852. out producing excessive expansion of erythropoiesis. Transfusion
99. Cox KO, Finlay-Jones JJ. Impaired regulation of erythrocyte auto- 1997;37:135–140.
antibody production after splenectomy. Br J Exp Pathol 1979;60: 129. Olivieri NF, Brittenham GM. Iron-chelation therapy and treatment of
466–470. thalassemia. Blood 1997;89:739–761.
100. Cullis JO, Win N, Dudley TM, et al. Post-transplant hyperhaemolysis 130. Piomeli S, Seaman C, Reibman J, et al. Separation of younger red cells
in a patient with sickle cell disease: use of steroids and intravenous with improved survival in vivo: An approach to chronic transfusion
immunoglobulin to prevent further red cell destruction. Vox Sang therapy. Proc Natl Acad Sci USA 1978;74:3473–3478.
1995;69:355–357. 131. Corash L, Klein H, Deisseroth A, et al. Selective isolation of young
101. Vermylen C, Cornu G. Bone marrow transplantation in sickle cell erythrocytes for transfusion support of thalassemia major patients.
anaemia. Blood Rev 1993;7:1–3. Blood 1981;57:599–606.
102. Vermylen C, Cornu G, Ferster A, et al. Haematopoietic stem cell trans- 132. Bracey AW, Klein HG, Chambers S, et al. Ex vivo selective isolation
plantation for sickle cell anaemia: the first 50 patients transplanted in of young red blood cells using the IBM-2991 cell washer. Blood
Belgium. Bone Marrow Transplant 1998;22:1–6. 1983;61:1068–1071.
103. Walters MC, Patience M, Leisenring W, et al. Bone marrow transplan- 133. Cohen AR, Schmidt JM, Martin MB, et al. Clinical trial of young red
tation for sickle cell disease. NEJM 1996;335:369–376. cell transfusions. J Pediatr 1984;104:865–868.
104. Platt OS, Guinan EC: Bone marrow transplantation in sickle cell ane- 134. Marcus RE, Wonke B, Bantock HM, et al. A prospective trial of young
mia: the dilemma of choice. NEJM 1996;335:426–428. red cells in 48 patients with transfusion-dependent thalassemia. Br J
105. Brichard B, Vermylen C, Ninane J, et al. Persistence of fetal hemoglobin Haematol 1985;60:153–159.
production after successful transplantation of cord blood stem cells in 135. Kevy SV, Jacobson MS, Fosburg M, et al. A new approach to neocyte
a patient with sickle cell anemia. J Pediatr 1996;128:241–243. transfusion: preliminary report. J Clin Apheresis 1988;4:194–197.
TRANSFUSION MEDICINE 136. Simon TL, Sohmer P, Nelson EF. Extended survival of neocytes pro- 144. Ameen R, Al-Shemmari S, Al-Humood R, et al. RBC alloimmuniza-
duced by a new system. Transfusion 1989;29:221–225. tion and autoimmunization among transfusion-dependent Arab thal-
137. Collins AF, Dias GC, Haddad S, et al. Evaluation of a new neocyte assemia patients. Transfusion 2003;43:1604–1610.
transfusion preparation vs. washed cell transfusion in patients with 145. Michail-Merianou V, Pamphili-Panousopoulou L, Piperi-Lowes L, et al.
homozygous beta thalassemia. Transfusion 1994;34:517–520. Alloimmunization to red cell antigens in thalassemia: comparative
138. Berdoukas VA, Moorew RC. A study of the value of red cell exchange study of usual versus better match transfusion programmes. Vox Sang
transfusions in transfusion dependent anemias. Clin Lab Haematol 1987;52:95–98.
1986;8:209–220. 146. Spanos T, Karageorga M, et al. Red cell alloantibodies in patients with
139. Cohen AR, Porter JB. Transfusion and iron chelation therapy in thal- thalassemia. Vox Sang 1990;58:50–55.
assemia and sickle cell anemia. In Steinberg MH, Forget BG, Higgs 147. Hmida S, Mojaat N, Maamar M, et al. Red cell alloantibodies in patients
DR, et al (eds). Disorders of Hemoglobin: Genetics, Pathophysiology, with haemoglobinopathies. Nouv Rev Fr Hematol 1994;36:363–366.
and Clinical Management. Cambridge, Cambridge University Press, 148. Sirchia G, Zanella A, Parravicini A, et al. Red cell alloantibodies in thal-
2001, p 982. assemia major. Transfusion 1985;25:110–112.
140. Capelleni MD. Iron-chelating therapy with the new oral agent ICL670 149. Singer ST, Wu V, Mignacca R, et al. Alloimmunization and erythrocyte
(Exjade). Best Pract Res Clin Hematol 2005;18:289–298. autoimmunization in transfusion-dependent patients of primarily
141. Cunningham MJ, Nathan DG. New developments in iron chelators. Asian descent. Blood 2000;96:3369–3373.
Curr Opin Hematol 2005;12:129–134. 150. Mizon P, Cossement C, Mannessier L, et al. Severe hemolysis relation
142. Naithani R, Chandra J, Sharma S: Safety of oral iron chelator deferi- to an association of erythrocye allo- and autoantibodies in thalassemia
prone in young thalassemics. Eur J Haematol 2005;74:217–220. patients. Transfusion Clin Biol 1996:3;257–261.
143. Bhatti FA, Salamat N, Nadeem A, et al. Red cell immunization 151. Kruatrachue M, Sirisinha S, Pacharee P, et al. An association between
in beta-thalassemia major. J Coll Physicians Surg Pak 2004;14: thalassemia and autoimmune hemolytic anemia (AIHA). Scand J Hae-
657–660. matol 1980;25:259–263.
III
538
Chapter 39
Transfusion to Bone Marrow or
Solid Organ Transplant Recipients
Richard M. Kaufman ● Steven R. Sloan
Recipient Donor RBC PLT 1st Choice PLT Next Choice* FFP
A B O AB A, B, O AB
O O A, AB B, O A, AB
AB A, O AB A, B, O AB
B A O AB B, A, O AB
O O B, AB A, O B, AB
AB B, O AB B, A, O AB
O A O A, AB B, O A, AB
B O B, AB A, O B, AB
AB O AB A, B, O AB
AB A A, O AB A, B, O AB
B B, O AB B, A, O AB
O O AB A, B, O AB
*
Volume reduction recommended for “next choice” platelets.
All cellular products must be irradiated.
FFP; fresh frozen plasma; PLT, platelets; RBC, red blood cells.
TRANSFUSION MEDICINE
factors (e.g., infection, GVHD).71,72 Often, these clinical fac- or BU (partially homozygous) matched platelets provide the
tors are more important determinants of bleeding risk than highest probability of a successful response.87
the platelet count alone. Approximately two thirds of bleed-
ing events after HSCT occur in patients with platelet counts
above 20,000/μL.17,73 SOLID-ORGAN TRANSPLANTATION
Alloimmunization and Platelet The total number of organ transplants performed in the
United States has been gradually increasing over the past sev-
Refractoriness
eral years. Indeed, between 2000 and 2004, the total number
Platelet refractoriness may be defined as an inappropriately of transplants performed in the United States increased by
low platelet increment following repeated platelet transfu- 16%, to 27,036 per year. During this time period, the num-
sions. Several nonimmune factors can cause platelet refrac- bers of kidney and liver transplants increased while the num-
toriness, including fever, sepsis, disseminated intravascular ber of heart transplants remained stable (United Network
coagulation, medications, bleeding, splenomegaly, hepatic of Organ Sharing, www.unos.org). The blood transfusion
veno-occlusive disease, and GVHD. In a subset of cases, plate- requirements for a transplant depend on several factors,
let refractoriness is caused by alloimmunization. Platelets including the type of organ being transplanted, the clinical
express HLA class I antigens, ABO antigens,74 and several condition of the patient, the experience of the transplant
platelet-specific antigens. Any of these molecules could poten- center, and use of antifibrinolytic agents such as aprotinin.
tially serve as an immune stimulus in a transfusion recipient.
Antibodies directed against HLA molecules are responsible
for most cases of immune-mediated platelet refractoriness. Blood Product Support
Notably, less than half of all platelet-refractory patients have
Liver Transplants
demonstrable antiplatelet or anti-HLA antibodies.75,76 Platelet
counts obtained at 10 minutes and 1 hour post-transfusion Although liver transplants often require substantial transfu-
that repeatedly fail to demonstrate an adequate corrected sion of blood components, the transfusion needs of these
count increment usually indicate an immune mechanism patients has declined as centers have gained experience.93 The
of platelet destruction.77 If the 10 minute and 1 hour post- transfusion requirements can vary significantly and depend
transfusion platelet count shows a reasonable increment but on the medical center and on the complexity of the case and
the platelet count falls back to baseline by 18 to 24 hours, a the patient’s condition.94,95 Red cells are transfused in 80% to
nonimmune mechanism of refractoriness may be presumed. 90% of cases, and the median number of RBC units required
HLA antibody screening (percent reactive antibody, or PRA) ranges from 4 to 12, depending on the center.93,96–98
provides valuable supporting evidence that allosensitization Although not all patients undergoing liver transplanta-
has occurred.78,79 A patient with a PRA greater than 70% tion require platelet transfusions, some patients with liver
III may be considered to be severely immunized 80 and a good failure develop thrombocytopenia that can be compounded
candidate for HLA-matched platelets. by blood loss during surgery. Platelets are transfused in about
542 Although they express HLA class I antigens, platelets 55% of cases,93 and the median number of platelets trans-
themselves are fairly weak immunogens. It has been shown fused is about 10 whole blood–derived platelet units.93,96–98
that contaminating leukocytes in platelet products are pri- Plasma transfusions are also used to support about 75%
marily responsible for stimulating HLA antibody formation of liver transplants.93 Patients with more severe coagulopa-
in platelet transfusion recipients.81–84 Thus, removal of white thies and greater blood loss during surgery are more likely to
cells from blood products helps prevent alloimmunization require plasma transfusions.94,95 The use of plasma transfu-
and platelet refractoriness. The definitive study showing sions varies greatly between institutions,93 with the median
this was the Trial to Reduce Alloimmunization to Platelets number of units of plasma used ranging from 0 to 19.93,96–98
(TRAP study),84 which compared alloimmunization rates in Additionally, some institutions use cryoprecipitate for some
530 newly diagnosed acute myelogenous leukemia patients of their liver transplants.97,98
randomized to receive either unmodified, pooled platelet
concentrates (control); filtered, pooled platelet concentrates Heart Transplants
(F-PC); filtered single-donor apheresis platelets (F-AP); or Although heart transplant patients do not usually have a
UVB-irradiated pooled platelet concentrates (UVB-PC). coagulopathy or thrombocytopenia prior to surgery, they
Anti-HLA antibodies were detected in 45% of controls, invariably require blood component support because of
compared with 17% to 21% of patients receiving modified the blood volume diverted to the cardiopulmonary bypass
platelets. Of the control group patients, 13% became plate- circuit and because platelets are damaged when flowing
let-refractory, versus only 3% in the F-PC group, 4% in the through bypass circuits. There is significant variability in
F-AP group, and 5% in the UVB-PC group. product usage between institutions, with these patients using
Once platelet refractoriness has been demonstrated, sev- a median of 3 to 9 units of red cells, 3 to 10 units of plasma,
eral strategies may help in obtaining therapeutic platelet and 0 to 13 units of platelets.96,99,100
increments in vivo. A trial of ABO-matched, fresh (1 to 2 days
old) platelets may be helpful.85 In cases of immune-mediated Lung Transplants and Heart-Lung Transplants
refractoriness, a trial of HLA-matched platelets,78,86–88 anti- Transfusion support for single or double lung transplants
gen-negative platelets,89 or crossmatched platelets90–92 should differs significantly, because double lung transplants require
be considered. Due to the high degree of polymorphism of cardiopulmonary bypass. Single lung transplants have been
the HLA loci, it is often not possible to find perfect HLA- reported to require a mean of 1.7 units of red cells, 1.5 units of
A and HLA-B locus matches, leading to the use of platelets platelets, 1.3 units of plasma, and 0.8 units of cryoprecipitate101;
mismatched at one or more loci. Grade A (HLA identical) double lung transplants or heart-lung transplants require 6.4
TRANSFUSION TO BONE MARROW OR SOLID ORGAN TRANSPLANT RECIPIENTS
to 13.3 units of red cells, 3 to 13.6 units of platelets, 4 to 8 units the Kupffer cells may be able to remove antigen–antibody
of plasma, and 0 to 15.8 units of cryoprecipitate.100–102 One complexes.109 Indeed, HLA matching is not usually per-
report found that the use of aprotinin was able to reduce the formed for liver allografts.110,111 Early studies suggested that
use of red cell and whole blood units from a mean of 13.3 to 7, ABO barriers could be ignored because patients rarely devel-
the use of plasma units from a mean of 6.1 to 2, and to eliminate oped hyperacute rejection.112 However, additional experience
the use of platelets or cryoprecipitate.102 revealed that ABO-incompatible liver allografts can undergo
ABO-mediated hyperacute rejection113,114 and the allografts
Other Solid Organ Transplants have a significantly diminished survival. For instance, one
Small bowel transplants are relatively uncommon. They can group reported a 30% 2-year graft survival rate for ABO-
require a median of 7 units of red cells, 3 units of plasma, and incompatible livers compared with a 76% to 80% 2-year graft
8 units of platelets.96 The far more common kidney transplants survival rate for ABO-compatible liver allografts.115 Other
usually only require 0 to 2 units of red cells.96 centers have also found a significant decrease in 1-year or
2-year survival of ABO-incompatible liver allografts.105,116
ABO-incompatible liver transplants appear to be safer in
young children. In patients less than 1 year old, the 5-year
ABO
survival following an ABO-incompatible liver transplant
was 76% at one Japanese center.117 Similar results have been
Overview seen in an American hospital with 81% of ABO-incompat-
Endothelial cells express ABO antigens.103,104 Hence, the ible allografts surviving long-term following transplants into
potential exists for antibody-mediated rejection when a trans- children who were less than 3 years old.118
planted vascular organ expresses ABO antigens that are absent One report suggests that group A2 livers can be safely
in the patient (i.e., there is a major ABO mismatch). However, transplanted into blood group O patients.119 However, fur-
because the supply of organs is severely limited and organs ther work is needed in this area because this report only
are sometimes needed emergently or are provided by living included 6 patients who suffered nine episodes of rejection
related donors, a variety of centers have attempted to perform that responded to standard treatment.
ABO mismatched organ transplants using a variety of strate- Plasmapheresis has been used by some groups to remove
gies. Overall, success has been mixed. Currently, although 70% incompatible isohemagglutinins either before or after trans-
to 80% of ABO-compatible hearts, livers, or cadaveric kidneys plant.105,120,121 However, there is no evidence that plasma-
survive 1 year, only 50% to 60% of ABO-incompatible trans- pheresis improves survival of the graft or the patient since
plants survive at least 1 year, even though those patients often the patients generally received intensive immunosuppressive
undergo intensive immunosuppression.105 treatments and graft survival rates were usually no better
Results from many reports of transplanting ABO- than 60%. Additionally, plasmapheresis may be associated
mismatched organs need to be viewed with caution because with increased septic complications in these patients.122
of the paucity of controlled studies. Even without special 39
ABO Compatibility in Heart
protocols, the results of ABO-incompatible organ trans-
and Lung Transplants
plants are variable; whereas some transplanted organs are 543
rapidly rejected, others survive for long periods of time. Also, ABO compatibility is a requirement for heart transplants
in addition to ABO issues, transplant results are affected by in adults. Five of eight cases of unintentional transplants of
the medical status of patients prior to transplant and by ABO-incompatible hearts resulted in hyperacute rejection.123
the immunosuppressive treatments used to prevent or treat An unintentional transplant of an ABO-incompatible heart
rejection. and lung also resulted in hyperacute rejection,124 whereas an
One strategy to breach the ABO barrier is to limit the unintentional ABO-incompatible lung transplant was able
transplants to those in whom only low levels of the ABO to be sustained for at least 3 years through the addition of
antigen are expressed. Specifically, some transplanters have antigen-specific immunoadsorption, anti-CD20 monoclo-
limited ABO-incompatible transplants to those in which the nal antibody, and recombinant soluble complement receptor
donor types as an A2 and the recipient lacks the A antigen type 1 to an immunosuppressive regimen of cyclophospha-
entirely.106,107 mide, antithymocyte globulin, cyclosporine, mycophenolate
Another strategy for performing ABO-incompat- mofetil, and prednisone.125
ible transplants is to limit the transplants to patients whose Increasing evidence demonstrates that infants can be
immune system is less likely to reject the ABO-incompat- transplanted successfully with ABO-incompatible hearts.
ible organ. One way to identify such patients may be to limit This finding was ariginally reported in 2001 by West and col-
ABO-incompatible transplants to patients with low titers of leagues126; experience in almost 50 patients at several medi-
the relevant isohemagglutinin. Another set of patients that cal centers has shown that many young infants can be safely
may not reject ABO-incompatible organs is very young chil- transplanted with ABO-incompatible hearts.127
dren with immature immune systems. Isohemagluttinins are Although most ABO-incompatible heart transplants have
usually absent in the first few months of life and the titers been performed in infants too young to produce isohemag-
usually remain low in the first year of life.108 glutinins, the blood banks supporting such transplants have
A third strategy for performing ABO-incompatible solid implemented special procedures. Starting prior to a possible
organ transplants is to suppress the immune system with some ABO-incompatible transplant, patients are usually provided
combination of drugs, plasmapheresis, and splenectomy. with blood components lacking isohemagglutinins (i.e., AB
plasma and washed cellular blood components). Even with
ABO Compatibility in Liver Transplants these precautions, a patient’s plasma may contain incom-
In comparison to other vascular organs, the liver is thought patible isohemagglutinins that were passively transfused,
to be less susceptible to humoral rejection, possibly because of maternal origin, or were produced by the infant’s own
TRANSFUSION MEDICINE
immune system. Such isohemagglutinins are usually removed Transplant Study.105 However, data from these selected
immediately prior to the transplant either by plasmapheresis reports must be considered cautiously; long-term analysis of
or by whole blood exchange as the patient is placed on car- the original patients reveals that 23 of 31 ABO-incompatible
diopulmonary bypass for surgery. Incompatible anti-A and/ renal transplants from living related donors have survived at
or anti-B antibodies that are detected after the transplant least 15 years, which is a long-term survival rate equivalent
may be removed by plasmapheresis. to that seen with ABO-compatible renal transplants at that
Although the mean age of patients that have been success- center.141 The specific immunosuppressive treatment is espe-
fully transplanted with incompatible hearts is 117 days,127 cially important in ABO-incompatible transplants, with one
patients as old as 14 months have received transplants.128 report finding that the addition of mycophenolate mofetil
However, experience with patients older than age 8 months is allowed ABO-incompatible kidneys to survive as long as
limited; one patient who received a transplant at age 9 months ABO-compatible transplants.144
developed rejection mediated by an isohemagglutinin that Several centers have continued to perform transplants using
was successfully treated with rituximab.128 this or similar approaches. In some of these protocols, patients
only receive transplants if plasmapheresis can reduce the anti-
ABO Compatibility in Renal Transplants A/B antibody titer to no more that 4:1 or 8:1 before the trans-
Early experiences demonstrated that transplantation of an plant. Recent studies demonstrate that pretransplant removal
ABO-incompatible kidney could lead to hyperacute or acute of anti-A/B antibodies is necessary for successful transplants
rejection; the 1-year survival of such transplanted kidneys of ABO-incompatible kidneys (other than A2 kidneys).145,146
was only 4%.129–133 However, many attempts have been made Some studies also suggest that splenectomy is unnecessary for
to transplant ABO-incompatible kidneys because the wait ABO-incompatible renal transplants, especially if patients are
list for cadaveric kidneys is very long and many patients are treated with newer potent immunosuppressive treatments that
offered ABO-incompatible kidneys from family members. can modify B-cell responses, such as rituximab, mycophenolate
This has been of particular importance in Japan where tradi- mofetil, or intravenous immunoglobulin (IVIG).141,145,147–151
tionally very few cadaveric kidneys have been made available
for transplant. Non-ABO Antibodies
Many reports suggest that the safest ABO-incompatible
renal transplants are ones in which the kidney is from a Although transplanted organs can induce alloantibodies to
donor whose cells express the A2 antigen and the recipient erythrocyte antigens,152 such antibodies have little impact on
is of blood group O or B with low titers of anti-A antibod- graft survival. Early studies suggested that renal allografts sur-
ies. An early report of A2-incompatible transplants using vived longer if the Lewis type of the kidney donor matched
the standard immunosuppression of the time found limited the type of the patient.153 However, this has not been seen
success with 12 of 20 such grafts surviving long-term and 8 in more recent studies. It is now accepted that kidneys can
of the grafts being lost within 1 month.134 Subsequent work be safely transplanted into patients with preexisting anti–red
III suggested that patients with an immunoglobulin M titer of cell antibodies such as anti-RhD.154–156
<64 against A2 red blood cells were less likely to reject an
544 A2 kidney than patients with higher anti-A2 titers.135 Some Pretransplant Transfusion for Kidneys
American transplant programs currently transplant kidneys
in which the A2 antigen confers major incompatibility. These to Increase Survival
programs have restricted the transplants to patients whose Starting in 1973, studies have shown that blood transfu-
initial anti-A antibody titer is ≤4 or ≤8, which in some cases sions prior to renal transplantation prolong survival and
had been reduced by pretransplant plasmapheresis.135–139 reduce rejection of renal allografts.96,157 The beneficial effect
Even with these protocols occasional cases of acute rejection of pretransplant transfusion had only been seen for kidney
have occurred. However, these centers have reported at least transplants. Then, new immunosuppressive drugs such as
90% 1-year and 85% 5-year survival of the grafts, which is cyclosporine were developed and it was unclear whether
similar to graft survival rates that these centers have achieved transfusions conferred any additional benefit to patients
with ABO-compatible transplants.138,139 who were receiving these drugs.96,158 Because it had also
Starting in the 1980s, some centers tried transplanting become clear that transfusions can transmit infectious dis-
kidneys with major ABO incompatibility with the recipients, eases and induce alloimmunization, the practice of pretrans-
based on expression of A1 and/or B antigens on the renal plant transfusion to prolong the survival of renal allografts
allografts. Several of these centers modeled their immuno- was largely abandoned.
suppressive treatment protocols on the approach used by a Several recent studies have found that even with the use
Belgian group that included pretransplant antibody removal of immunosuppressive drugs such as cyclosporine or tacro-
by plasmapheresis or immunoadsorption, steroids, azathio- limus, pretransplant transfusion may still decrease the risk
prine, cyclosporine or tacrolimus, antilymphocyte globulin, of rejection of the renal allograft.159–162 In the pediatric set-
and splenectomy.140 In some of these series patients also ting, this benefit was observed in patients receiving one to
received pretransplant infusion of the relevant A and/or B five transfusions, but was not seen in patients receiving more
substance.141 In most cases in which the transplanted kid- than five transfusions.161 Despite these studies, most centers
neys survived, incompatible isohemagglutinins returned to are not currently using transfusions for this purpose.
the patient’s plasma and the kidney continued to express the
ABO antigen.141–143
At least 309 ABO-incompatible renal transplants have Passenger Lymphocyte Syndrome
been reported with a combined 1-year graft survival rate Donor B lymphocytes contained in solid organ allografts can
of 83%,130 which is equivalent to the survival rate seen with sometimes produce anti-erythrocyte antibodies, resulting in
ABO compatible grafts in the multicenter Collaborative a condition known as the passenger lymphocyte syndrome.
TRANSFUSION TO BONE MARROW OR SOLID ORGAN TRANSPLANT RECIPIENTS
Typically, the antibodies are detected 1 to 2 weeks after the Several centers have also reported success with treating
transplant as a positive DAT.96 Elution reveals anti-A and/ acute humoral rejection of cardiac allografts with an immu-
or anti-B. These patients may develop hemolysis, which can nosuppressive regimen that includes plasmapheresis.173–179
be severe in rare cases. Although at least one death has been In some cases in which the rejection failed to respond to
attributed to this hemolysis, most patients recover and the multiple pharmacologic interventions, there was significant
antibody usually disappears in about 1 month.96,163 clinical improvement following plasmapheresis.175
Most cases of passenger lymphocyte syndrome involve Plasmapheresis has also been used to remove anti-HLA
antibodies to ABO antigens. As reported by Ramsey’s exten- antibodies prior to renal transplants. This approach has been
sive analysis of the literature,163 the observed rate of ABO used with apparent success in renal transplant patients receiv-
antibodies and hemolysis from passenger lymphocyte syn- ing either living donor allografts180,181 or cadaveric allografts.182
drome is lowest in kidney transplant patients (17% and 9%, Plasmapheresis combined with other immunosuppressive
respectively), intermediate in liver transplant patients (40% therapy such as IVIG has also been used to successfully treat
and 29%, respectively), and highest in heart-lung transplant acute humoral rejection of renal allografts.178,179,183,184
patients (both 70%). Notably, most of these reports were Plasmapheresis for primary allograft nonfunction after liver
prior to the advent of cyclosporine or tacrolimus. In kidney transplantation has been attempted with mixed results.185,186
transplant patients, cyclosporine increases the incidence of
passenger lymphocyte syndrome so that 30% of patients
develop antibodies and 17% develop hemolysis. TRANSFUSION-RELATED COMPLICATIONS
No risk factors for passenger lymphocyte syndrome IN IMMUNOCOMPROMISED PATIENTS
have been proven. It has been hypothesized that passenger
lymphocyte syndrome is more likely when the donor has a
Cytomegalovirus Infection
high-titer antibody or when the recipient is a nonsecretor or
is of the A2 subtype.163,164 Although it is possible that these Cytomegalovirus (CMV) is a member of the herpesvirus
donor and recipient characteristics are risk factors for pas- group, which includes the herpes simplex viruses 1 and
senger lymphocyte syndrome, no large studies have been 2, Epstein-Barr virus, and varicella-zoster virus. CMV is
performed, and individual case reports demonstrate that capable of infecting a variety of cell types, including mature
secretors and non-A2 patients can develop passenger lym- white blood cells and their progenitors. In the immuno-
phocyte syndrome. The only prophylactic treatment that has compromised host, CMV infection can lead to significant
been shown to reduce the incidence of passenger lympho- morbidity and mortality. Allogeneic HSCT patients in par-
cyte syndrome in renal transplant patients is post-transplant ticular are at risk to develop serious manifestations of CMV
irradiation of the graft.163,165 disease, which include pneumonitis, gastroenteritis, hepa-
Passenger lymphocyte syndrome resulting in hemolysis titis, encephalitis, and retinitis. CMV disease is much less
has rarely been caused by non-ABO antibodies including commonly seen in the autologous HSCT setting.187 Because
anti-RhD, anti-Rhe, anti-Rhc, anti-RhE and anti-Jka antibod- of the risk of transfusion-transmitted CMV disease in 39
ies163,166–168 Case reports suggest that passenger lymphocyte transplant recipients, “CMV safe” blood products should be
syndrome involving these antibodies only occurs when the provided to CMV-negative transplant recipients. 545
donor is alloimmunized to the antigen prior to donating the Serologic screening of blood products is currently the
organ. most effective means of reducing the risk of transfusion-
transmitted CMV. The rare cases of transfusion-transmitted
Plasmapheresis to Prevent or Treat CMV disease still observed in large part reflect donations
Anti-HLA Antibody-Mediated Rejection made in the preseroconversion window period.188 Because
the seroprevalence of CMV ranges in different communities
of Transplanted Organs
from 40% to 80%, it is logistically difficult for blood banks to
No randomized, controlled prospective trials have ana- maintain an inventory of CMV-seronegative components.189
lyzed the efficacy of plasmapheresis in preventing or treat- Several studies have demonstrated that leukoreduction can
ing anti-HLA antibody-mediated rejection of transplanted be used as an alternative to serologic screening to provide
solid organs. However, several transplant centers incorporate blood that is “CMV safe.”187,190–195 Although unquestionably
plasmapheresis into their treatment protocols based on suc- efficacious, leukoreduction appears to be slightly inferior to
cess from case reports, case series, and retrospective analyses providing seronegative products for the prevention of CMV
of local data. Plasmapheresis in this setting has been used for transmission.190,196,197 Overall, although transfusion-trans-
cardiac and renal transplants, and very rarely with hepatic mitted CMV remains a problem in the HSCT and solid-
transplants. organ transplant population, the transmission rates seen
Patients who have significant levels of anti-HLA anti- using either seronegative or leukoreduced products are far
bodies at the time of cardiac transplant have reduced lower than in the past, due in large part to improvements in
long-term survival,169 and patients with antibodies that both surveillance and preemptive therapy.198
react against the heart donor’s HLA antigens appear to
have a very high mortality rate.170 If patients with anti- Transfusion-Associated
HLA antibodies are treated by plasmapheresis followed by
Graft-versus-Host Disease
IVIG immediately before transplant, patient and cardiac
allograft survival rates improve to levels comparable to Transfusion-associated GVHD (TA-GVHD) is a devas-
those of patients without pre-existing anti-HLA antibod- tating complication of transfusion that primarily affects
ies.171,172 Alternatively, some institutions have had success immunocompromised patients, including hematopoietic
performing multiple plasma exchanges on alloimmunized stem cell and solid organ transplant recipients. TA-GVHD
patients awaiting heart transplants. occurs when immunocompetent lymphocytes in a blood
TRANSFUSION MEDICINE mobilized peripheral blood as measured by the quantitative in vivo
product proliferate in the transfusion recipient. The donor
lymphocytes attack host organs, including bone marrow, SCID-repopulating cell assay. Blood 1997;89:3919–3924.
13. Rubinstein P, Carrier C, Scaradavou A, et al. Outcomes among 562
skin, liver, and intestines. TA-GVHD usually becomes clin- recipients of placental-blood transplants from unrelated donors.
ically apparent 7 to 10 days after transfusion. Patients usu- NEJM 1998;339:1565–1577.
ally present with fever, skin rash, elevated results on liver 14. Benito AI, Diaz MA, Gonzalez-Vicent M, et al. Hematopoietic stem
function tests, and gastrointestinal symptoms. Severe pan- cell transplantation using umbilical cord blood progenitors: review of
current clinical results. Bone Marrow Transplant 2004;33:675–690.
cytopenia is often observed. In classical GVHD, the bone 15. Rocha V, Sanz G, Gluckman E. Umbilical cord blood transplantation.
marrow is typically spared, because donor and recipient Curr Opin Hematol 2004;11:375–385.
bone marrow are selected to be HLA-compatible. In con- 16. Barker JN, Weisdorf DJ, DeFor TE, et al. Transplantation of 2 partially
trast, TA-GVHD involves immune rejection of host mar- HLA-matched umbilical cord blood units to enhance engraftment in
row cells, and as a result generally carries a worse prognosis adults with hematologic malignancy. Blood 2005;105:1343–1347.
17. Bernstein SH, Nademanee AP, Vose JM, et al. A multicenter study
than classical GVHD. TA-GVHD is almost always fatal of platelet recovery and utilization in patients after myeloablative
within 4 weeks of transfusion, with most deaths attribut- therapy and hematopoietic stem cell transplantation. Blood 1998;91:
able to bleeding or infection resulting from marrow fail- 3509–3517.
ure. 18. Klaesson S, Ringden O, Ljungman P, et al. Reduced blood transfusions
requirements after allogeneic bone marrow transplantation: results
No effective therapy for TA-GVHD currently exists, but of a randomised, double-blind study with high-dose erythropoietin.
it is prevented by gamma-irradiation of blood products Bone Marrow Transplant 1994;13:397–402.
before transfusion. TA-GVHD was previously reported to 19. Link H, Boogaerts MA, Fauser AA, et al. A controlled trial of recombi-
occur following blood product irradiation to 1500 rads,199 nant human erythropoietin after bone marrow transplantation. Blood
and other studies indicated that irradiation to 2000 rads 1994;84:3327–3335.
20. Biggs JC, Atkinson KA, Booker V, et al. Prospective randomised dou-
is needed to reduce mitogen-responsive lymphocytes by ble-blind trial of the in vivo use of recombinant human erythropoietin
5 to 6 logs.200 Irradiated blood products are now required in bone marrow transplantation from HLA-identical sibling donors.
to receive 2500 rads (25 Gy) to the center of the container, The Australian Bone Marrow Transplant Study Group. Bone Marrow
with a minimum dose of 1500 rads (15Gy) to any other Transplant 1995;15:129–134.
21. Miller CB, Lazarus HM. Erythropoietin in stem cell transplantation.
point.201 Cellular blood products are required to be irradi- Bone Marrow Transplant 2001;27:1011–1016.
ated in the following situations: recipients identified to be 22. Weissinger F, Sandmaier BM, Maloney DG, et al. Decreased transfu-
at risk for TA-GVHD, donations made by blood relatives of sion requirements for patients receiving nonmyeloablative compared
the recipient, and donors selected to be HLA-matched with with conventional peripheral blood stem cell transplants from HLA-
the recipient.202 identical siblings. Blood 2001;98:3584–3588.
23. Ruiz-Arguelles GJ, Morales-Toquero A, Lopez-Martinez B, et al.
Bloodless (transfusion-free) hematopoietic stem cell transplants: the
Mexican experience. Bone Marrow Transplant 2005;36:715–720.
REFERENCES 24. Rowley SD. Hematopoietic stem cell transplantation between red cell
incompatible donor-recipient pairs. Bone Marrow Transplant 2001;28:
1. Schmitz N, Barrett J. Optimizing engraftment—source and dose of 315–321.
III stem cells. Semin Hematol 2002;39:3–14. 25. Buckner CD, Clift RA, Sanders JE, et al. ABO-incompatible marrow
2. Singhal S, Powles R, Kulkarni S, et al. Comparison of marrow and transplants. Transplantation 1978;26:233–238.
546 blood cell yields from the same donors in a double-blind, randomized 26. Mielcarek M, Leisenring W, Torok-Storb B, et al. Graft-versus-host dis-
study of allogeneic marrow vs blood stem cell transplantation. Bone ease and donor-directed hemagglutinin titers after ABO-mismatched
Marrow Transplant 2000;25:501–505. related and unrelated marrow allografts: evidence for a graft-versus-
3. Bensinger WI, Martin PJ, Storer B, et al. Transplantation of bone mar- plasma cell effect. Blood 2000;96:1150–1156.
row as compared with peripheral-blood cells from HLA-identical rela- 27. Mielcarek M, Torok-Storb B, Storb R. ABO incompatibility and relapse
tives in patients with hematologic cancers. NEJM 2001;344:175–181. risk in patients undergoing allogeneic marrow transplantation for
4. Blaise D, Kuentz M, Fortanier C, et al. Randomized trial of bone mar- acute myeloid leukemia. Bone Marrow Transplant 2002;30:547–548.
row versus lenograstim-primed blood cell allogeneic transplantation 28. Kim JG, Sohn SK, Kim DH, et al. Impact of ABO incompatibility on
in patients with early-stage leukemia: a report from the Societe Fran- outcome after allogeneic peripheral blood stem cell transplantation.
caise de Greffe de Moelle. J Clin Oncol 2000;18:537–546. Bone Marrow Transplant 2005;35:489–495.
5. Heldal D, Tjonnfjord G, Brinch L, et al. A randomised study of alloge- 29. Goldman J, Liesveld J, Nichols D, et al. ABO incompatibility between
neic transplantation with stem cells from blood or bone marrow. Bone donor and recipient and clinical outcomes in allogeneic stem cell
Marrow Transplant 2000;25:1129–1136. transplantation. Leuk Res 2003;27:489–491.
6. Powles R, Mehta J, Kulkarni S, et al. Allogeneic blood and bone mar- 30. Klumpp TR. Immunohematologic complications of bone marrow
row stem cell transplantation in haematological malignant diseases: a transplantation. Bone Marrow Transplant 1991;8:159–170.
randomised trial. Lancet 2000;355:1231–1237. 31. Worel N, Kalhs P, Keil F, et al. ABO mismatch increases transplant-
7. Schmitz N, Linch DC, Dreger P, et al. Randomised trial of filgrastim- related morbidity and mortality in patients given nonmyeloablative
mobilised peripheral blood progenitor cell transplantation versus allogeneic HPC transplantation. Transfusion 2003;43:1153–1161.
autologous bone marrow transplantation in lymphoma patients. Lan- 32. Canals C, Muniz-Diaz E, Martinez C, et al. Impact of ABO incompat-
cet 1996;347:353–357. ibility on allogeneic peripheral blood progenitor cell transplantation
8. Vigorito AC, Azevedo WM, Marques JF, et al. A randomised, prospec- after reduced intensity conditioning. Transfusion 2004;44:1603–1611.
tive comparison of allogeneic bone marrow and peripheral blood 33. Griffith LM, McCoy JP Jr, Bolan CD, et al. Persistence of recipient
progenitor cell transplantation in the treatment of haematological plasma cells and anti-donor isohaemagglutinins in patients with
malignancies. Bone Marrow Transplant 1998;22:1145–1151. delayed donor erythropoiesis after major ABO incompatible non-
9. Kiss JE, Rybka WB, Winkelstein A, et al. Relationship of CD34+ cell myeloablative haematopoietic cell transplantation. Br J Haematol
dose to early and late hematopoiesis following autologous periph- 2005; 128:668–675.
eral blood stem cell transplantation. Bone Marrow Transplant 34. Bolan CD, Leitman SF, Griffith LM, et al. Delayed donor red cell chi-
1997;19:303–310. merism and pure red cell aplasia following major ABO-incompatible
10. Haas R, Mohle R, Fruhauf S, et al. Patient characteristics associated nonmyeloablative hematopoietic stem cell transplantation. Blood
with successful mobilizing and autografting of peripheral blood pro- 2001;98:1687–1694.
genitor cells in malignant lymphoma. Blood 1994;83:3787–3794. 35. Gmur JP, Burger J, Schaffner A, et al. Pure red cell aplasia of long dura-
11. Lane TA. Umbilical cord blood grafts for hematopoietic transplantation tion complicating major ABO-incompatible bone marrow transplan-
in adults: a cup half empty or half full? Transfusion 2005;45:1027–1034. tation. Blood 1990;75:290–295.
12. Wang JC, Doedens M, Dick JE. Primitive human hematopoietic cells 36. Bensinger WI, Buckner CD, Thomas ED, et al. ABO-incompatible
are enriched in cord blood compared with adult bone marrow or marrow transplants. Transplantation 1982;33:427–429.
TRANSFUSION TO BONE MARROW OR SOLID ORGAN TRANSPLANT RECIPIENTS
37. Lee JH, Choi SJ, Kim S, et al. Changes of isoagglutinin titres after 60. Young PP, Goodnough LT, Westervelt P, et al. Immune hemolysis
ABO-incompatible allogeneic stem cell transplantation. Br J Haematol involving non-ABO/RhD alloantibodies following hematopoietic stem
2003;120:702–710. cell transplantation. Bone Marrow Transplant 2001;27:1305–1310.
38. Maschan AA, Skorobogatova EV, Balashov DN, et al. Successful treat- 61. Zupanska B, Zaucha JM, Michalewska B, et al. Multiple red cell alloan-
ment of pure red cell aplasia with a single dose of rituximab in a child tibodies, including anti-Dib, after allogeneic ABO-matched peripheral
after major ABO incompatible peripheral blood allogeneic stem cell blood progenitor cell transplantation. Transfusion 2005;45:16–20.
transplantation for acquired aplastic anemia. Bone Marrow Transplant 62. Ting A, Pun A, Dodds AJ, et al. Red cell alloantibodies produced after
2002;30:405–407. bone marrow transplantation. Transfusion 1987;27:145–147.
39. Lee JH, Lee KH, Kim S, et al. Anti-A isoagglutinin as a risk factor for 63. de La Rubia J, Arriaga F, Andreu R, et al. Development of non-ABO
the development of pure red cell aplasia after major ABO-incompati- RBC alloantibodies in patients undergoing allogeneic HPC trans-
ble allogeneic bone marrow transplantation. Bone Marrow Transplant plantation. Is ABO incompatibility a predisposing factor? Transfusion
2000;25:179–184. 2001;41:106–110.
40. Damodar S, George B, Mammen J, et al. Pretransplant reduction of iso- 64. Leo A, Mytilineos J, Voso MT, et al. Passenger lymphocyte syndrome
haemagglutinin titres by donor group plasma infusion does not reduce with severe hemolytic anemia due to an anti-Jka after allogeneic PBPC
the incidence of pure red cell aplasia in major ABO-mismatched trans- transplantation. Transfusion 2000;40:632–636.
plants. Bone Marrow Transplant 2005;36:233–235. 65. Platelet transfusion therapy. National Institutes of Health Consensus
41. Bar BM, Van Dijk BA, Schattenberg A, et al. Erythrocyte repopula- Conference. Transfus Med Rev 1987;1:195–200.
tion after major ABO incompatible transplantation with lymphocyte- 66. Beutler E. Platelet transfusions: The 20,000/μL trigger. Blood 1993;81:
depleted bone marrow. Bone Marrow Transplant 1995;16:793–799. 1411–1413.
42. Benjamin RJ, Connors JM, McGurk S, et al. Prolonged erythroid 67. Gmur J, Burger J, Schanz U, et al. Safety of stringent prophylactic plate-
aplasia after major ABO-mismatched transplantation for chronic let transfusion policy for patients with acute leukaemia. Lancet 1991;
myelogenous leukemia. Biol Blood Marrow Transplant 1998;4: 338:1223–1226.
151–156. 68. Heckman KD, Weiner GJ, Davis CS, et al. Randomized study of pro-
43. Maciej Zaucha J, Mielcarek M, Takatu A, et al. Engraftment of early ery- phylactic platelet transfusion threshold during induction therapy
throid progenitors is not delayed after nonmyeloablative major ABO- for adult acute leukemia: 10,000/μL versus 20,000/μL. J Clin Oncol
incompatible haematopoietic stem cell transplantation. Br J Haematol 1997;15:1143–1149.
2002;119:740–750. 69. Rebulla P, Finazzi G, Marangoni F, et al. The threshold for prophy-
44. Veelken H, Wasch R, Behringer D, et al. Pure red cell aplasia after lactic platelet transfusions in adults with acute myeloid leukemia.
allogeneic stem cell transplantation with reduced conditioning. Bone Gruppo Italiano Malattie Ematologiche Maligne dell’Adulto. NEJM
Marrow Transplant 2000;26:911–915. 1997;337:1870–1875.
45. Peggs KS, Morris EC, Kottaridis PD, et al. Outcome of major ABO- 70. Wandt H, Frank M, Ehninger G, et al. Safety and cost effectiveness of a
incompatible nonmyeloablative hematopoietic stem cell transplan- 10 × 109/L trigger for prophylactic platelet transfusions compared with
tation may be influenced by conditioning regimen. Blood 2002;99: the traditional 20 × 109/L trigger: a prospective comparative trial in
4642–4643. 105 patients with acute myeloid leukemia. Blood 1998;91:3601–3606.
46. Santamaria A, Sureda A, Martino R, et al. Successful treatment of pure 71. Gil-Fernandez JJ, Alegre A, Fernandez-Villalta MJ, et al. Clinical results
red cell aplasia after major ABO-incompatible T cell-depleted bone of a stringent policy on prophylactic platelet transfusion: nonrandom-
marrow transplantation with erythropoietin. Bone Marrow Trans- ized comparative analysis in 190 bone marrow transplant patients
plant 1997;20:1105–1107. from a single institution. Bone Marrow Transplant 1996;18:931–935.
47. Paltiel O, Cournoyer D, Rybka W. Pure red cell aplasia following ABO- 72. Schiffer CA, Anderson KC, Bennett CL, et al. Platelet transfusion for
incompatible bone marrow transplantation: response to erythropoi- patients with cancer: clinical practice guidelines of the American Soci-
etin. Transfusion 1993;33:418–421. ety of Clinical Oncology. J Clin Oncol 2001;19:1519–1538.
48. Heyll A, Aul C, Runde V, et al. Treatment of pure red cell aplasia after 73. Nevo S, Enger C, Hartley E, et al. Acute bleeding and thrombocytope-
39
major ABO-incompatible bone marrow transplantation with recom- nia after bone marrow transplantation. Bone Marrow Transplant 2001;
binant erythropoietin. Blood 1991;77:906. 27:65–72. 547
49. Rabitsch W, Knobl P, Prinz E, et al. Prolonged red cell aplasia after 74. Cooling LL, Kelly K, Barton J, et al. Determinants of ABH expression
major ABO-incompatible allogeneic hematopoietic stem cell trans- on human blood platelets. Blood 2005;105:3356–3364.
plantation: Removal of persisting isohemagglutinins with Ig-Therasorb 75. Doughty HA, Murphy MF, Metcalfe P, et al. Relative importance of
immunoadsorption. Bone Marrow Transplant 2003;32:1015–1019. immune and nonimmune causes of platelet refractoriness. Vox Sang
50. Verholen F, Stalder M, Helg C, et al. Resistant pure red cell aplasia after 1994;66:200–205.
allogeneic stem cell transplantation with major ABO mismatch treated 76. Yankee RA, Graffs KS, Dowling R, et al. Selection of unrelated
by escalating dose donor leukocyte infusion. Eur J Haematol 2004;73: compatible platelet donors by lymphocyte HLA matching. NEJM
441–446. 1973;288:760.
51. Bierman PJ, Warkentin P, Hutchins MR, et al. Pure red cell aplasia fol- 77. Daly PA, Schiffer CA, Aisner J, et al. Platelet transfusion therapy. One-
lowing ABO mismatched marrow transplantation for chronic lympho- hour posttransfusion increments are valuable in predicting the need
cytic leukemia: response to antithymocyte globulin. Leuk Lymphoma for HLA-matched preparations. JAMA 1980;243:435–438.
1993;9:169–171. 78. Duquesnoy RJ, Filip DJ, Rodey GE, et al. Successful transfusion of
52. Labar B, Bogdanic V, Nemet D, et al. Antilymphocyte globulin for platelets “mismatched” for HLA antigens to alloimmunized thrombo-
treatment of pure red cell aplasia after major ABO incompatible mar- cytopenic patients. Am J Hematol 1977;2:219–226.
row transplant. Bone Marrow Transplant 1992;10:471–472. 79. Lee EJ, Schiffer CA. Serial measurement of lymphocytotoxic antibody
53. Petz LD, Garratty G. Immune Hemolytic Anemias, 2nd ed. Philadelphia, and response to nonmatched platelet transfusions in alloimmunized
Churchill Livingstone, 2004. patients. Blood 1987;70:1727–1729.
54. Lopez J, Steegmann JL, Perez G, et al. Erythropoietin in the treatment 80. Nance ST, Hsu S, Vassallo RR, et al. Review: Platelet matching for
of delayed immune hemolysis of a major ABO-incompatible bone alloimmunized patients—room for improvement. Immunohematol
marrow transplant. Am J Hematol 1994;45:237–239. 2004;20:80–88.
55. Buckner CD, Clift RA, Sanders JE, et al. ABO-incompatible marrow 81. Oksanen K, Kekomaki R, Ruutu T, et al. Prevention of alloimmuniza-
transplants. Transplantation 1978;26:233. tion in patients with acute leukemia by use of white cell-reduced blood
56. Lasky LC, Warkentin PI, Kersey JH, et al. Hemotherapy in patients components—a randomized trial. Transfusion 1991;31:588–594.
undergoing blood group incompatible bone marrow transplantation. 82. Saarinen UM, Kekomaki R, Siimes MA, et al. Effective prophylaxsis
Transfusion 1983;23:277–285. against platelet refractoriness in multitransfused patients by use of
57. Bolan CD, Childs RW, Procter JL, et al. Massive immune haemolysis leukocyte-free components. Blood 1990;75:512.
after allogeneic peripheral blood stem cell transplantation with minor 83. van Marwijk Kooy M, van Prooijen HC, Moes M, et al. Use of leuko-
ABO incompatibility. Br J Haematol 2001;112:787–795. cyte depleted platelet concentrates for the prevention of refractoriness
58. Hows J, Beddow K, Gordon-Smith E, et al. Donor-derived red blood and primary HLA alloimmunization: a prosective, randomized trial.
cell antibodies and immune hemolysis after allogeneic bone marrow Blood 1991;77:201.
transplantation. Blood 1986;67:177–181. 84. The Trial to Reduce Alloimmunization to Platelets Study Group. Leu-
59. Franchini M, Gandini G, Aprili G. Non-ABO red blood cell alloanti- kocyte reduction and ultraviolet B irradiation of platelets to prevent
bodies following allogeneic hematopoietic stem cell transplantation. alloimmunization and refractoriness to platelet transfusions. NEJM
Bone Marrow Transplant 2004;33:1169–1172. 1997;337:1861–1869.
TRANSFUSION MEDICINE 85. Friedberg RC, Donnelly SF, Boyd JC, et al. Clinical and blood bank 114. Rego J, Prevost F, Rumeau JL, et al. Hyperacute rejection after ABO-
factors in the management of platelet refractoriness and alloimmuni- incompatible orthotopic liver transplantation. Transplant Proc 1987;
zation. Blood 1993;81:3428–3434. 19:4589–4590.
86. Yankee RA, Grumet FC, Rogentine GN. Platelet transfusion: the selec- 115. Gugenheim J, Samuel D, Reynes M, et al. Liver transplantation across
tion of compatible platelet donors for refractory patients by lympho- ABO blood group barriers. Lancet 1990;336:519–523.
cyte HL-A typing. NEJM 1969;281:1208–1212. 116. Reding R, Veyckemans F, de Ville de Goyet J, et al. ABO-incompat-
87. Moroff G, Garratty G, Heal JM, et al. Selection of platelets for refrac- ible orthotopic liver allografting in urgent indications. Surg Gynecol
tory patients by HLA matching and prospective crossmatching. Trans- Obstet 1992;174:59–64.
fusion 1992;32:633–640. 117. Egawa H, Oike F, Buhler L, et al. Impact of recipient age on outcome
88. Murphy MF, Brozovic B, Murphy W, et al. Guidelines for platelet transfu- of ABO-incompatible living-donor liver transplantation. Transplanta-
sions. British Committee for Standards in Haematology, Working Party tion 2004;77:403–411.
of the Blood Transfusion Task Force. Transfus Med 1992;2:311–318. 118. Varela-Fascinetto G, Treacy SJ, Lillehei CW, et al. Long-term results
89. Petz LD, Garratty G, Calhoun L, et al. Selecting donors of platelets for in pediatric ABO-incompatible liver transplantation. Transplant Proc
refractory patients on the basis of HLA antibody specificity. Transfu- 1999;31:467–468.
sion 2000;40:1446–1456. 119. Fishbein TM, Emre S, Guy SR, et al. Safe transplantation of blood type A2
90. Rachel JM, Summers TC, Sinor LT, et al. Use of a solid phase red blood livers to blood type O recipients. Transplantation 1999;67:1071–1073.
cell adherence method for pretransfusion platelet compatibility test- 120. Mor E, Skerrett D, Manzarbeitia C, et al. Successful use of an enhanced
ing. Am J Clin Pathol 1988;90:63–68. immunosuppressive protocol with plasmapheresis for ABO-incom-
91. von dem Borne AE, Ouwehand WH, Kuijpers RW. Theoretic and practical patible mismatched grafts in liver transplant recipients. Transplanta-
aspects of platelet crossmatching. Transfus Med Rev 1990;4: 265–278. tion 1995;59:986–990.
92. Gelb AB, Leavitt AD. Crossmatch-compatible platelets improve cor- 121. Tanaka A, Tanaka K, Kitai T, et al. Living related liver transplantation
rected count increments in patients who are refractory to randomly across ABO blood groups. Transplantation 1994;58:548–553.
selected platelets. Transfusion 1997;37:624–630. 122. Farges O, Kalil AN, Samuel D, et al. The use of ABO-incompatible
93. Ozier Y, Pessione F, Samain E, et al. Institutional variability in trans- grafts in liver transplantation: a life-saving procedure in highly selected
fusion practice for liver transplantation. Anesthesia and Analgesia patients. Transplantation 1995;59:1124–1133.
2003;97:671–679. 123. Cooper DK. Clinical survey of heart transplantation between ABO
94. Findlay JY, Rettke SR. Poor prediction of blood transfusion require- blood group-incompatible recipients and donors. J Heart Transplant
ments in adult liver transplantations from preoperative variables. J Clin 1990;9:376–381.
Anesth 2000;12:319–323. 124. Campion EW. A death at Duke. NEJM 2003;348:1083–1084.
95. Steib A, Freys G, Lehmann C, et al. Intraoperative blood losses and 125. Pierson RN 3rd, Loyd JE, Goodwin A, et al. Successful management of
transfusion requirements during adult liver transplantation remain an ABO-mismatched lung allograft using antigen-specific immuno-
difficult to predict. Can J Anaesth 2001;48:1075–1079. adsorption, complement inhibition, and immunomodulatory therapy.
96. Triulzi DJ. Specialized transfusion support for solid organ transplanta- Transplantation 2002;74:79–84.
tion. Curr Opin Hematol 2002;9:527–532. 126. West LJ, Pollock-BarZiv SM, Dipchand AI, et al. ABO-incompatible
97. Laine E, Steadman R, Calhoun L, et al. Comparison of RBCs and heart transplantation in infants. NEJM 2001;344:793–800.
FFP with whole blood during liver transplant surgery. Transfusion 127. West LJ, Pollock-BarZiv SM, Ang A, et al. ABO-incompatible infant
2003;43:322–327. heart transplantation: The World Experience. J Heart Lung Transplant
98. McNicol PL, Liu G, Harley ID, et al. Blood loss and transfusion require- 2005;24:S63–S64.
ments in liver transplantation: experience with the first 75 cases. Anaes 128. Fan XH, Ang A, BarZiv SMP, et al. Donor-specific B-cell tolerance
Intensive Care 1994;22:666–671. after ABO-incompatible infant heart transplantation. Nature Med
99. Danielson CF, Filo RS, O’Donnell JA, et al. Institutional variation 2004;10:1227–1233.
III in hemotherapy for solid organ transplantation. Transfusion 1996; 129. Cook DJ, Graver B, Terasaki PI. ABO incompatibility in cadaver donor
36:263–267. kidney allografts. Transplant Proc 1987;19:4549–4552.
548 100. Hunt BJ, Sack D, Amin S, et al. The perioperative use of blood components 130. Rydberg L. ABO-incompatibility in solid organ transplantation. Trans-
during heart and heart-lung transplantation. Transfusion 1992; 32:57–62. fus Med 2001;11:325–342.
101. Triulzi DJ, Griffith BP. Blood usage in lung transplantation. Transfu- 131. Starzl T, Marchioro T, Holmes J, et al. Renal homografts in patients
sion 1998;38:12–15. with major donor–recipient blood group incompatibilities. Surgery
102. Peterson KL, DeCampli WM, Feeley TW, et al. Blood loss and transfu- 1964;55:195–200.
sion requirements in cystic fibrosis patients undergoing heart-lung or 132. Murray J, Merill J, Damin G, et al. Study on transplantation immu-
lung transplantation. J Cardiothorac Vasc Anesth 1995;9:59–62. nity after total body iradiation: clinical and experimental investigation.
103. Westhoff CM, Reid ME. ABO and related antigens. In Roush KS (ed). Surgery 1960;48:272–284.
Blood Banking and Transfusion Medicine: Basic Principles and Prac- 133. Hume D, Merill J, Miler B, et al. Experiences with renal allotrans-
tice. Philadelphia, Churchill Livingstone, 2003, pp 21–30. plantation in the human: report of nine cases. J Clin Invest 1955;34:
104. Oriol R. Tissular expression of ABH and Lewis antigens in humans and 372–382.
animals: expected value of different animal models in the study of ABO- 134. Rydberg L, Breimer ME, Samuelsson BE, et al. Blood group ABO-
incompatible organ transplants. Transplant Proc 1987;19:4416–4420. incompatible (A2 to O) kidney transplantation in human subjects:
105. Wu A, Buhler LH, Cooper DKC. ABO-incompatible organ and bone a clinical, serologic, and biochemical approach. Transplant Proc
marrow transplantation: current status. Transpl Int 2003;16:291–299. 1987;19:4528–4537.
106. Rydberg L. ABO incompatibility in solid organ transplantation. Trans- 135. Welsh KI, van Dam M, Koffman CG, et al. Transplantation of blood
fus Med 2001;11:325–342. group A2 kidneys into O or B recipients: the effect of pretransplant
107. Fishbein TM, Emre S, Guy SR, et al. Safe transplantation of blood type A2 anti-A titers on graft survival. Transplant Proc 1987;19:4565–4567.
livers to blood type O recipients. Transplantation 1999;67: 1071–1073. 136. Alkhunaizi AM, de Mattos AM, Barry JM, et al. Renal transplanta-
108. Fong SW, Qaqundah BY, Taylor WF. Developmental patterns of ABO tion across the ABO barrier using A2 kidneys. Transplantation 1999;
isoagglutinins in normal children correlated with the effects of age, 67:1319–1324.
sex, and maternal isoagglutinins. Transfusion 1974;14:551–559. 137. Nelson PW, Landreneau MD, Luger AM, et al. Ten-year experience in
109. Wardle EN. Kupffer cells and their function. Liver 1987;7:63–75. transplantation of A2 kidneys into B and O recipients. Transplantation
110. Gordon RD, Fung JJ, Iwatsuki S, et al. Immunological factors influenc- 1998;65:256–260.
ing liver graft survival. Gastroenterol Clin North Am 1988;17:53–59. 138. Nelson PW, Shield CF 3rd, Muruve NA, et al. Increased access to
111. Poli F, Scalamogna M, Aniasi A, et al. A retrospective evaluation of transplantation for blood group B cadaveric waiting list candidates by
HLA-A, B and -DRB1 matching in liver transplantation. Transpl Int using A2 kidneys: time for a new national system? Am J Transplant
1998;11(Suppl 1):S347–S349. 2002;2:94–99.
112. Starzl TE, Ishikawa M, Putnam CW, et al. Progress in and deterrents 139. Norman DJ, Prather JC, Alkhunaizi AM, et al. Use of A2 kidneys for
to orthotopic liver transplantation, with special reference to survival, B and O kidney transplant recipients: report of a series of patients
resistance to hyperacute rejection, and biliary duct reconstruction. transplanted at a single center spanning a decade. Transplant Proc
Transplant Proc 1974;6:129–139. 2001;33:3327–3330.
113. Demetris AJ, Jaffe R, Tzakis A, et al. Antibody-mediated rejection of 140. Alexandre GP, Squifflet JP, De Bruyere M, et al. Present experiences in
human orthotopic liver allografts. A study of liver transplantation a series of 26 ABO-incompatible living donor renal allografts. Trans-
across ABO blood group barriers. Am J Pathol 1988;132:489–502. plant Proc 1987;19:4538–4542.
TRANSFUSION TO BONE MARROW OR SOLID ORGAN TRANSPLANT RECIPIENTS
141. Squifflet J-P, De Meyer M, Malaise J, et al. Lessons learned from ABO- 165. Ishida H, Tanabe K, Tokumoto T, et al. The evaluation of graft irradia-
incompatible living donor kidney transplantation: 20 years later. Exper tion as a method of preventing hemolysis after ABO-mismatched renal
Clin Transplant 2004;2:208–213. transplantation. Transplant Int 2002;15:421–424.
142. Bach FH, Ferran C, Hechenleitner P, et al. Accommodation of vascular- 166. Hareuveni M, Merchav H, Austerlitz N, et al. Donor anti-Jka causing
ized xenografts: expression of “protective genes” by donor endothelial hemolysis in a liver transplant recipient. Transfusion 2002;42:363–367.
cells in a host Th2 cytokine environment. Nat Med 1997;3:196–204. 167. Fung MK, Sheikh H, Eghtesad B, et al. Severe hemolysis resulting from
143. Reding R, Squifflet JP, Latinne D, et al. Early postoperative monitoring D incompatibility in a case of ABO-identical liver transplant. Transfu-
of natural anti-A and anti-B isoantibodies in ABO-incompatible living sion 2004;44:1635–1639.
donor renal allografts. Transplant Proc 1987;19:1989–1990. 168. Larrea L, delaRubia J, Arriaga F, et al. Severe hemolytic anemia due to
144. Tanabe K, Tokumoto T, Ishida H, et al. ABO-incompatible renal anti-E after renal transplantation. Transplantation 1997;64:550–551.
transplantation at Tokyo Women’s Medical University. In Checka JM, 169. Loh E, Bergin JD, Couper GS, et al. Role of panel-reactive antibody
Terasaki P (eds). Clinical Transplants 2003. Los Angeles, UCLA Immu- cross-reactivity in predicting survival after orthotopic heart transplan-
nogenetics Center, 2004, pp 175–181. tation. J Heart Lung Transplant 1994;13:194–201.
145. Ishida H, Koyama I, Sawada T, et al. Anti-AB titer changes in patients 170. Singh G, Thompson M, Griffith B, et al. Histocompatibility in car-
with ABO incompatibility after living related kidney transplantations: diac transplantation with particular reference to immunopathology
survey of 101 cases to determine whether splenectomies are necessary of positive serologic crossmatch. Clin Immunol Immunopathol 1983;
for successful transplantation. Transplantation 2000;70:681–685. 28:56–66.
146. Shimmura H, Tanabe K, Ishikawa N, et al. Role of anti-A/B antibody 171. Pisani BA, Mullen GM, Malinowska K, et al. Plasmapheresis with
titers in results of ABO-incompatible kidney transplantation. Trans- intravenous immunoglobulin G is effective in patients with elevated
plantation 2000;70:1331–1335. panel reactive antibody prior to cardiac transplantation. J Heart Lung
147. Warren DS, Simpkins CE, Cooper M, et al. Modulating alloim- Transplant 1999;18:701–706.
mune responses with plasmapheresis and IVIG. Current Drug Target 172. Robinson JA, Radvany RM, Mullen MG, et al. Plasmapheresis followed
2005;5:215–222. by intravenous immunoglobulin in presensitized patients awaiting
148. Tanabe K, Tokumoto T, Ishida H, et al. Excellent outcome of ABO- thoracic organ transplantation. Therap Apheresis 1997;1:147–151.
incompatible living kidney transplantation under pretransplantation 173. McOmber D, Ibrahim J, Lublin DM, et al. Non-ischemic left ventricu-
immunosuppression with tacrolimus, mycophenolate mofetil, and lar dysfunction after pediatric cardiac transplantation: treatment with
steroid. Transplant Proc 2004;36:2175–2177. plasmapheresis and OKT3. J Heart Lung Transplant 2004;23:552–557.
149. Mannami M, Mitsuhata N. Improved outcomes after ABO-incompat- 174. Grauhan O, Knosalla C, Ewert R, et al. Plasmapheresis and cyclophos-
ible living-donor kidney transplantation after 4 weeks of treatment phamide in the treatment of humoral rejection after heart transplanta-
with mycophenolate mofetil. Transplantation 2005;79:1756–1758. tion. J Heart Lung Transplant 2001;20:316–321.
150. Sonnenday CJ, Warren DS, Cooper M, et al. Plasmapheresis, CMV 175. Berglin E, Kjellstrãm C, Mantovani V, et al. Plasmapheresis as a rescue
hyperimmune globulin, and anti-CD20 allow ABO-incompatible therapy to resolve cardiac rejection with vasculitis and severe heart
renal transplantation without splenectomy. Am J Transplant 2004;4: failure. A report of five cases. Transplant Int 1995;8:382–387.
1315–1322. 176. Malafa M, Mancini MC, Myles JL, et al. Successful treatment of acute
151. Tyden G, Kumlien G, Fehrman I. Successful ABO-incompatible kidney humoral rejection in a heart transplant patient. J Heart Lung Trans-
transplantations without splenectomy using antigen-specific immu- plant 1992;11:486–491.
noadsorption and rituximab. Transplantation 2003;76:730–731. 177. Partanen J, Nieminen MS, Krogerus L, et al. Heart transplant rejection
152. Cummins D, Contreras M, Amin S, et al. Red cell alloantibody devel- treated with plasmapheresis. J Heart Lung Transplant 1992;11:301–305.
opment associated with heart and lung transplantation. Transplanta- 178. Takemoto SK, Zeevi A, Feng S, et al. National conference to assess anti-
tion 1995:59:1432–1435. body-mediated rejection in solid organ transplantation. Am J Trans-
153. Lenhard V, Hansen B, Roelcke D, et al. Influence of Lewis and other plant 2004;4:1033–1041.
blood group systems in kidney transplantation. Proc Eur Dialysis 179. Jordan SC, Vo AA, Tyan D, et al. Current approaches to treatment of
39
Transplant Assoc 1983;19:432–437. antibody-mediated rejection. Pediatr Transplant 2005;9:408–415.
154. White AG, Kumar MS, Abouna GM. HLA, MLR, P and Lewis antigens 180. Thielke J, DeChristopher PJ, Sankary H, et al. Highly successful living 549
and living donor renal transplantation in a single centre in the Middle donor kidney transplantation after conversion to negative of a previ-
East. Tissue Antigens 1986;27:279–284. ously positive flow-cytometry cross-match by pretransplant plasma-
155. Etheredge EE, Bettonville P, Sicard GA, et al. Anti-erythrocyte antibod- pheresis. Transplant Proc 2005;37:643–644.
ies, leukocytotoxins and human renal allograft survival. Tissue Anti- 181. Sonnenday CJ, Ratner LE, Zachary AA, et al. Preemptive therapy with
gens 1982;19:205–212. plasmapheresis/intravenous immunoglobulin allows successful live
156. Bryan CF, Mitchell SI, Lin HM, et al. Influence of the RhD blood group donor renal transplantation in patients with a positive cross-match.
system on graft survival in renal transplantation. Transplantation Transplant Proc 2002;34:1614–1616.
1998;65:588–592. 182. Alarabi A, Backman U, Wikstrom B, et al. Plasmapheresis in HLA-
157. Opelz G, Sengar DP, Mickey MR, et al. Effect of blood transfusions on immunosensitized patients prior to kidney transplantation. Int J Arti-
subsequent kidney transplants. Transplant Proc 1973;5:253–259. ficial Organs 1997;20:51–56.
158. Egidi MF, Scott DH, Corry RJ. The effect of transfusions on renal 183. White NB, Greenstein SM, Cantafio AW, et al. Successful rescue ther-
allograft survival in the cyclosporine era: a single center report. Clini- apy with plasmapheresis and intravenous immunoglobulin for acute
cal Transplant 1993;7:240–244. humoral renal transplant rejection. Transplantation 2004;78:772–774.
159. Reinsmoen NL, Matas AJ, Donaldson L, et al. Impact of transfusions 184. Montgomery RA, Zachary AA, Racusen LC, et al. Plasmapheresis and
and acute rejection on posttransplantation donor antigen-specific intravenous immune globulin provides effective rescue therapy for
responses in two study populations. Cooperative Clinical Trial in refractory humoral rejection and allows kidneys to be successfully
Transplantation Research Group. Transplantation 1999;67:697–702. transplanted into cross-match-positive recipients. Transplantation
160. Higgins RM, Raymond NT, Krishnan NS, et al. Acute rejection after 2000;70:887–895.
renal transplantation is reduced by approximately 50% by prior thera- 185. Mandal AK, King KE, Humphreys SL, et al. Plasmapheresis: an effec-
peutic blood transfusions, even in tacrolimus-treated patients. Trans- tive therapy for primary allograft nonfunction after liver transplanta-
plantation 2004;77:469–471. tion. Transplantation 2000;70:216–220.
161. Chavers BM, Sullivan EK, Tejani A, et al. Pretransplant blood trans- 186. Skerrett D, Mor E, Curtiss S, et al. Plasmapheresis in primary dysfunc-
fusion and renal allograft outcome: a report of the North American tion of hepatic transplants. J Clin Apher 1996;11:10–13.
Pediatric Renal Transplant Cooperative Study. Pediatr Transplant 187. Wingard JR, Chen DYH, Burns WH, et al. Cytomegalovirus infec-
1997;1:22–28. tion after autologous bone marrow transplantation with compari-
162. Galvao MM, Peixinho ZF, Mendes NF, et al. Stored blood—an effec- son to infection after allogeneic bone marrow transplantation. Blood
tive immunosuppressive method for transplantation of kidneys 1988;71:1432.
from unrelated donors. An 11-year follow-up. Braz J Medical Biol R 188. American Association of Blood Banks. Leukocyte reduction for the
1997;30:727–734. prevention of transfusion-transmitted cytomegalovirus (TT-CMV).
163. Ramsey G. Red cell antibodies arising from solid organ transplants. AABB Bulletin 1997;2:1.
Transfusion 1991;31:76–86. 189. Roback JD. CMV and blood transfusions. Rev Med Virol 2002;12:
164. Sternberg AJ, Lee G, Croxton T, et al. Severe haemolysis after an ABO 211–219.
unmatched kidney transplant—a nonsecretor transplanted from a 190. Bowden RA, Slichter SJ, Sayers M, et al. A comparison of filtered
donor with high anti-A titre. Transfusion Med 2000;10:87–89. leukocyte-reduced and cytomegalovirus (CMV)-seronegative blood
TRANSFUSION MEDICINE products for the prevention of transfusion-associated CMV infection 197. Vamvakas EC. Is white blood cell reduction equivalent to antibody
after marrow transplant. Blood 1995;86:3598. screening in preventing transmission of cytomegalovirus by transfu-
191. Bowden RA, Slichter SJ, Sayers MH, et al. Use of leukocyte-depleted sion? A review of the literature and meta-analysis. Transfus Med Rev
platelets and cytomegalovirus-seronegative red blood cells for preven- 2005;19:181–199.
tion of primary cytomegalovirus infection after marrow transplant. 198. Zaia JA. Prevention and management of CMV-related problems after
Blood 1991;78:246–250. hematopoietic stem cell transplantation. Bone Marrow Transplant
192. De Witte T, Schattenberg A, Van Djik BA, et al. Prevention of primary 2002;29:633–638.
cytomegalovirus infection after allogeneic bone marrow transplanta- 199. Lowenthal RN, Challis DR, Griffiths AE, et al. Transfusion-associated
tion by using leukocyte-poor blood products from cytomegalovirus- graft-versus-host disease: report of an occurence following the admin-
unscreened blood bank donors. Transplantation 1990;50:964. istration of irradiated blood. Transfusion 1993;33:524.
193. Gilbert GL, Hayes K, Hudson IL, et al. Prevention of transfusion- 200. Pelszynski MM, Moroff G, Luban N, et al. Effect of gamma irradiation
aquired cytomegalovirus infection in infants by blood filtration to of red blood cell units on T-cell inactivation as assessed by limiting
remove leukocytes. Lancet 1989;1:1228. dilution analysis: Implications for preventing transfusion-associated
194. Miller WJ, McCullough J, Balfour HH Jr, et al. Prevention of cyto- graft-versus-host disease. Blood 1994;83:1683.
megalovirus infection following bone marrow transplantation: a ran- 201. U.S. Food and Drug Administration. FDA memorandum to registered
domized trial of blood product screening. Bone Marrow Transplant blood establishments: recommendations reguarding license amend-
1991;7:227–234. ments and proceedures for gamma irradiation of blood products.
195. Lang DJ, Ebert PA, Rodgers BM, et al. Reduction of postperfusion Washington, D.C., July 22, 1993.
cytomegalovirus infections following the use of leukocyte depleted 202. Menitove JE (ed). Standards for Blood Banks and Transfusion Services,
blood. Transfusion 1977;17:391. 21st ed. Bethesda, Md., American Association of Blood Banks, 2002.
196. Nichols WG, Price TH, Gooley T, et al. Transfusion-transmitted cyto-
megalovirus infection after receipt of leukoreduced blood products.
Blood 2003;101:4195–4200.
III
550
Chapter 40
Transfusion of the Platelet-Refractory
Patient
Thomas S. Kickler
Platelet transfusion therapy has improved over the past 2 transfusion strategy, using a platelet count of 20,000/μL as the
decades with new methods for collection, storage, and process- transfusion trigger, has been commonly employed.8
ing of platelets. With the development of aggressive forms of More recently, in a randomized study of prophylactic platelet
chemotherapy, organ transplantation, and new strategies to transfusion, Rebulla and coworkers5 showed that giving trans-
treat aplastic anemia and related bone marrow failure disor- fusions only when the platelet count dips below 10,000/μL can
ders, the clinician relies heavily on the availability of platelet decrease platelet use with only a small adverse effect on bleed-
transfusions. Consequently, practitioners of platelet transfusion ing and no effect on mortality. It therefore appears that, with
therapy must know how to use platelet transfusions effectively amegakaryocytic thrombocytopenia, prophylactic transfusions
so that resources are not wasted. This chapter describes the should be given if the count falls below 5000/μL. At values
use of platelet transfusions in a variety of medical conditions, between 5000 and 10,000/μL, transfusion may be withheld if
especially in patients requiring multiple platelet transfusions. the patient is stable and if no other conditions make spontane-
Because refractoriness to platelet transfusions is a serious prob- ous bleeding likely.5 Such conditions include blast crisis, rapidly
lem, methods to circumvent or prevent it are also described. falling platelet count, anticoagulation with heparin for dissemi-
nated intravascular coagulation (DIC), drugs that affect platelet
function, uremia, and recent invasive procedures, including spi-
INDICATIONS FOR PLATELET nal taps or placement of central venous catheters.9
TRANSFUSIONS
40
In patients with thrombocytopenia, the risk of hemorrhage PLATELET TRANSFUSION
increases progressively once the platelet count drops below REFRACTORINESS 551
100,000/μL.1–3 Many studies have attempted to define the
bleeding time or platelet count necessary to achieve hemosta- Failure to achieve an expected increment with a platelet trans-
sis in surgical patients, with conflicting results.3–7 Some gener- fusion is called platelet transfusion refractoriness. Refractoriness
alities do exist, however. With normally functioning platelets, may be caused by an immune or a nonimmune condition.
most major surgery can be performed safely if the count is Clinically, one can assess the response to a platelet transfusion
maintained in the range of 50,000 to 75,000/mL. A higher by measuring the increment in platelet count 1 to 18 hours
range may be necessary for longer and more technically dif- after the transfusion. The post-transfusion platelet response
ficult procedures involving extensive incisions or exposure of should be calculated on the basis of the patient’s body surface
large surface areas. Performance of surgery on the central ner- area in square meters and corrected for the number of plate-
vous system should be done only with platelet counts greater lets transfused. The corrected platelet count increment (CCI)
than 100,000/μL.3 Table 40–1 shows general guidelines for the is calculated by the following formula:
transfusion of platelets in different clinical situations.
In amegakaryocytic thrombocytopenia, platelet transfusions
are given prophylactically or therapeutically and for the perfor-
mance of invasive procedures. There is considerable interest in
trying to define the lowest safe platelet concentration, so that CCI =
[
body
surface
2
area (m ) ][
platelet
× count
increment ]
× 1011
Platelet products: The benefits of pooled platelets or single-donor platelets are similar; the two products can be used
interchangeably. Single-donor platelets from selected donors are used when histocompatible platelet transfusion
(i.e., HLA-A and HLA-B antigen matched) are needed.
Prophylactic Platelet Transfusion Thresholds:
Acute leukemia: For adult patients a threshold of 10,000 / μL is recommended. Transfusion at higher levels may be necessary in
the newborn or in patients with hemorrhage, high fever, hyperleukocytosis, rapid fall in platelet count, or coagulation
abnormalities.
Hematopoietic stem cell transplantation: Same as for acute leukemia, with similar exceptions.
Chronic stable severe thrombocytopenia: Many patients can be observed without prophylactic transfusion, reserving transfusion
for episodes of hemorrhage or during times of active treatment.
Solid tumors: Evidence supports the benefit of prophylactic transfusion at a threshold of 10,000 /μL or less. A threshold of
20,000 /μL should be considered for patients receiving aggressive therapy for bladder cancer, as well as for those with
demonstrated necrotic tumors.
Surgical or invasive procedures: A platelet count of 40 to 50,000 /μL is sufficient to perform major invasive procedures safely,
in the absence of associated coagulation abnormalities. Certain procedures, such as bone marrow aspiration/biopsy, can be
performed safely with counts <20,000 /μL; lumbar puncture in children is safe at platelet counts >10,000 /μL.
testing, similar to those used to select compatible red blood nonimmune causes of shortened platelet survival. None of
cells. These procedures are well known to transfusion services the placebo group (five patients) achieved a satisfactory 1-
and involve phenotyping of donors and recipients and iden- hour CCI. By t-test, the post-treatment mean 1-hour CCI
tification of the specificity of antibodies present in a patient’s values were significantly greater in those who received IVIG
serum. Specifically for platelets, HLA phenotyping of donors than in the control group.32
and recipients, and identification of the HLA antibody speci- If all conventional methods fail to increase the platelet
ficities are required. Petz and coworkers29 extensively evalu- count to hemostatic levels, the only remaining alternative
ated this approach, and it appears to be highly reliable and that has been tried is continuous transfusion of platelets
successful in identifying donors who would ordinarily be (massive transfusion). It has been argued that, although the
excluded if only exact HLA matches were considered. These platelet count is not increased, transfused platelets still exert
investigators coined the phrase antibody specificity prediction some effect, permitting platelet plug formation or mainte-
(ASP) method to describe these procedures. nance of endothelial integrity. These arguments are based on
Petz and coworkers29 reported data on the utility of ASP clinical observations. In one well-established animal model
for 1621 platelet transfusions in 114 persons with platelet of alloimmune thrombocytopenia, if the platelet count did
transfusion refractoriness. They compared the effective- not increase above 60,000/μL, capillary leakage or bleeding
ness of platelets selected by the ASP method with that of still persisted.2,3
platelets selected on the basis of HLA matching or cross-
matching or, if selected components were not available, Other Approaches
on a random basis. They concluded that the ASP method
was as effective as HLA matching or crossmatching, and In uncontrolled studies and in small controlled studies, the
that all three methods were superior to the random selec- antifibrinolytic agents aminocaproic acid and tranexamic
tion of platelets. Further, in a file of HLA-matched donors, acid have been used.31 The results have been mixed in terms
III many more potential donors were identifiable by the ASP of reduction of microvascular hemorrhage and reduction
method than by HLA matching, which makes the acquisi- of platelet transfusion requirements. The differing results
554 tion of compatible platelets for alloimmunized refractory probably are related to the small number of patients stud-
patients much more feasible. This approach appears to be ied. Immunosuppressive agents, splenectomy, and plasma
logistically simple on a regional basis when the testing is exchange have not been proven effective.2,3
offered by blood centers, and it promises to harmonize the
selection of platelets with the selection process used for red
blood cell transfusions. PREVENTION OF ALLOIMMUNIZATION
III
556
Chapter 41
Autoimmune Hemolytic Anemias
Leslie E. Silberstein ● Melody J. Cunningham
SPECTRUM OF AUTOIMMUNE the cold and that a heat-labile serum factor lysed the eryth-
HEMOLYTIC SYNDROMES rocytes at 37°C.
The first report of cold agglutinin disease appeared in
Autoimmune hemolytic anemias refer to a spectrum of dis- 1918, but the fact that cold agglutinating serum antibod-
orders in which autoantibodies against antigens on the ies were found in healthy individuals initially obscured the
erythrocyte membrane cause shortened survival of native significance of cold autolysins. In 1937, these cold-reactive
as well as transfused red blood cells (RBCs). Three catego- antibodies were discovered to occur in much higher,
ries of antierythrocyte autoantibodies exhibit distinctive thus pathologic, levels in affected patients. Dameshek
serologic properties and result in characteristic clinical and Schwartz established the first experimental model of
disorders (Table 41–1). Immunoglobulin G (IgG) warm immune hemolytic anemia, inducing hemolysis by injection
autoantibodies attach to erythrocytes at 37°C, IgM cold of heterologous antierythrocyte antibodies into guinea pigs.
autoantibodies clump RBCs at cold temperatures, and IgG Yet the idea of an “autoimmune” form of hemolytic anemia
Donath-Landsteiner antibodies bind to RBC membranes in was resisted for several reasons, punctuated by the difficulty
the cold and cause hemolysis at 37°C; in this chapter, the in making the diagnosis.
associated clinical entities are referred to as warm autoim- The antiglobulin test, designed to detect nonagglutinat-
mune hemolytic anemia (AIHA), cold agglutinin disease, and ing anti-erythrocyte antibodies, was introduced into clinical
paroxysmal cold hemoglobinuria (PCH), respectively. All of medicine by Coombs and associates2 in 1945. Within 1 year,
these antibodies are capable of simply attaching to the RBC this test was used to diagnose autoimmune hemolytic ane-
membrane without having any pathologic effect or induc- mia.3 In 1954, autoimmune hemolytic anemia in dogs was
ing fulminant hemolysis.1 The antibodies may be idiopathic reported,4 and in 1958, the first easily bred animal model of 41
or may develop secondary to another disease process or in the disease, the NZB mouse, was described.5 This last dis-
response to exposure to a drug. covery was a turning point in the development of a scientific 557
basis for the study of autoimmunization.
Historical Background
WARM AUTOIMMUNE
The first recognized form of hemolytic anemia was PCH, HEMOLYTIC ANEMIA
probably because its clinical manifestation, the passage of
black urine after exposure to cold, is so striking. Reports of
apparent cold-induced hematuria began to appear in medical
Epidemiology in Children and Adults
literature in the mid-1800s. By 1884, the association of PCH The incidence of AIHA is estimated to be approximately 1 in
with syphilis was noted. In 1904, Donath and Landsteiner 100,000 adults and less than 0.2 in 100,000 children. The
determined that an autolysin fixed to the patient’s RBCs in disorder is less common than immune thrombocytopenia.6 In
Type of AIHA
Paroxysmal Cold
Characteristic Warm-Reactive Cold Agglutinin Disease Hemoglobinuria
Ig, immunoglobulin.
TRANSFUSION MEDICINE
teenagers and adults, AIHA is more common in women than of the antibody attached and the extent to which complement
in men. In children, boys are somewhat more affected than is activated. IgA, IgM, IgG1, and IgG3 can all fix complement,
girls. AIHA occurs at all ages but more commonly in midlife. although IgA is a rare cause of AIHA.14 If complement is acti-
In pediatric cases, the mortality is less than 10% and primar- vated through the C5 to C9 membrane attack complex, intra-
ily occurs in adolescents with a chronic refractory course.7 vascular hemolysis occurs.15 The RBCs coated with simply
Approximately half of cases of AIHA are idiopathic. In antibody or complement component C3b are phagocytized
children, AIHA in patients younger than age 2 and older by macrophages and destroyed extravascularly in the spleen or
than age 12 is more likely to have a chronic unremitting liver, respectively. Cells coated with IgG are destroyed primarily
course.8 Some cases are induced by drugs, and others occur in the spleen; those with IgM, in the liver.16
concomitantly with another autoimmune disease or malig- The interaction of macrophages with RBCs coated with
nancy. A substantial proportion of cases develop in patients IgG or C3b (or both) occurs through receptors specific for the
with systemic lupus erythematosus (SLE), B-cell lympho- Fc portion of IgG (especially IgG1 and IgG3) and for C3b.17,18
mas, or chronic lymphocytic leukemia (CLL). A number of The presence on the erythrocyte membrane of both IgG and
other diseases have also been complicated by AIHA, but only C3b accelerates immune clearance,19–21 suggesting that the Fc
as unusual exceptions.9 Treatment of the underlying process and C3b macrophage receptors act synergistically.
can often resolve the AIHA. In other instances, presumably
by affecting T cells more than B cells, drugs induce a distur- Red Blood Cell Injury
bance in immunoregulatory T cells and trigger the onset of
hemolytic anemia.10,11 The opsonized RBC may be phagocytosed and destroyed
As more diseases, malignant and nonmalignant, are treated entirely by macrophages. Alternatively, proteolytic enzymes
with bone marrow transplantation (BMT) and peripheral may digest part of the membrane surface, producing sphero-
blood stem cell transplantation (PBST), the numbers of cases cytes. The spherocytes are less deformable and consequently
of post-transplant autoimmune hemolysis are expected to are hemolyzed in the spleen. The predominant mechanism
increase. An overall 2.6% incidence was reported in a recent of destruction of erythrocytes coated with IgG with or
study in post-BMT patients. A 5% incidence was found in without C3b occurs extravascularly in AIHA. The amount
those patients surviving 6 months.12 The timing of develop- of immune clearance is also mediated by the entire reticu-
ment of post-BMT AIHA appears to depend on the antibody loendothelial system; thus, viral or bacterial infections may
mediating the hemolysis, possibly because IgM reconstitutes exacerbate hemolysis.22 In vitro assays of the ability of blood
earlier than IgG. In this clinical setting, three distinct mecha- monocytes from patients with viral infections to phagocy-
nisms can mediate the hemolysis. Investigations to determine tose immunoglobulin-coated RBCs have shown marked
the etiology of hemolysis will appropriately direct therapy, deviations from normal.23,24
expected duration, and prevention.13 The three possibilities
are true autoimmune hemolysis, immune hemolysis medi- Clinical Findings
III ated by passenger lymphocytes, and immune hemolysis in
the setting of chimerism and major blood group mismatch13 The clinical findings in AIHA are variable. They are deter-
558 (Table 41-2). mined by the rate of hemolysis and by the abilities of the body
to process breakdown products and of the bone marrow to
mount a reticulocytosis. Some of the signs are associated with
Pathophysiology
hyperdynamic circulation secondary to anemia and a con-
Attachment of autoantibody to the surface of the RBC may comitant decrease in oxygen-carrying capacity. They include
lead to intravascular or extravascular hemolysis. The pathologic hepatomegaly and, in more severe cases, pulmonary edema,
effect is determined by the class or subclass as well as the avidity lethargy, and obtundation. Splenomegaly can also occur
AIHA, autoimmune hemolytic anemia; GVHD, graft-versus-host disease; RBC, red blood cells.
AUTOIMMUNE HEMOLYTIC ANEMIAS
from an increase in white pulp. Other signs and symptoms, Indirect Antiglobulin Test (IAT)
such as jaundice, fever, and renal insufficiency, are caused by
the breakdown products and subsequent vasoconstriction. It Washed RBCs and
has now been demonstrated that decreased renal perfusion, patient serum
not injury due to free hemoglobin, is the mechanism leading
to renal insufficiency.
Autoimmune hemolytic anemia may have a fulminant RBC coated
presentation, with rapid onset of profound anemia, or may with -IgG
develop gradually, with concomitant physiologic compensa-
tion. Occasionally, unsuspected AIHA is diagnosed through
⫹ Antihuman
a positive direct antiglobulin test (DAT) result in an anemic globulin
patient who has been referred for transfusion. The presence
of lymphadenopathy, fever, hypertension, renal failure, rash,
petechiae, or ecchymoses necessitates careful investigation Agglutination
for an underlying malignancy or collagen vascular disease. -Positive IAT
Laboratory Evaluation
Direct and Indirect Antiglobulin Tests
Figure 41–2 Indirect antiglobulin test (IAT). IgG, immunoglobulin G;
It can be difficult to distinguish AIHA from other forms of RBC, red blood cell.
hemolytic anemia on the basis of laboratory data. The posi-
tive result of a DAT, also known as the Coombs’ test, is con-
sidered pathognomonic of immune-mediated hemolysis. In approximately 80% of patients with AIHA, the auto-
The test detects the presence of IgG or complement bound antibodies are present in serum as well as on RBC mem-
to the RBC membrane (Fig. 41–1). Severe disease usually branes.27 IAT detects the presence of these serum antibodies
produces a strong DAT response, but this finding does not in the patient’s serum. These may be autoantibodies in a
always correlate with the degree of hemolysis.1 The indirect patient with AIHA or they may be alloantibodies induced
antiglobulin test (IAT; indirect Coombs’ test), which detects by blood transfusion or maternal-fetal incompatibility.
the presence of antibodies in the patient’s serum, usually has Alloantibodies, present only in the serum, have specificity for
a positive result in AIHA (Fig. 41–2). RBC antigens not present on the patient’s erythrocytes. The
A positive DAT result is occasionally noted in a healthy per- DAT result is therefore negative in alloimmunization as long
son without anemia or evidence of hemolysis.25 Conversely, as the patient has not recently been transfused with RBCs
patients with known AIHA may have a negative DAT result; that have the target antigen. In the setting of a recent trans- 41
this latter finding can be caused infrequently by laboratory fusion, the alloantibodies may bind to recently transfused
error or by the presence of IgA, IgM, or low-affinity IgG RBCs, yielding a positive DAT result. 559
autoantibodies.26
Other Supportive Laboratory Investigations
More commonly, the test is not sensitive enough to
detect small numbers of erythrocyte-bound IgG molecules; It is difficult to distinguish intravascular from extravascu-
this occurs most often in AIHA associated with lymphoma lar hemolysis. The hemoglobin, hematocrit, lactate dehy-
or CLL. If the DAT result is positive, specific reagents are drogenase (LDH), bilirubin, and haptoglobin values are
required to identify the erythrocyte-bound protein. similarly affected in both types of hemolysis. A significant
urine hemosiderin level can indicate intravascular hemoly-
sis yet often appears too late to be a helpful clinical tool. An
unchanged hemoglobin or hematocrit value obtained within
Direct Antiglobulin Test (DAT) 4 hours after a transfusion of RBCs is an indicator of intra-
vascular hemolysis. The intravascular destruction occurs so
RBC coated with rapidly that no evidence of transfusion is reflected in the lab-
-IgG oratory values. Intravascular hemolysis is important to rec-
-C3 ognize, because the patient may require greater supportive
care to treat the anemia and consequent organ injury from
intravascular hemolysis.
⫹ Antihuman Like clinical symptoms, laboratory findings reflect the
globulin intensity of the hemolytic process as well as the ability of
the body to process the RBC breakdown products and of the
Agglutination
-Positive DAT
bone marrow to respond to the anemia. In fulminant cases,
with an RBC life span of less than 5 days, the anemia is severe
and erythropoiesis increases eight- to tenfold. As a result,
the reticulocyte count rises, sometimes to levels greater than
40% of RBCs. If the regenerative capacity of the bone mar-
Polyspecific antihuman row lags only slightly behind the rate of RBC destruction,
Monospecific antihuman
globulin
globulin a mild anemia with an elevated reticulocyte count results.
Figure 41–1 Direct antiglobulin test (DAT). C3, complement compo- Between these extremes are many variations. Inspection of
nent 3; IgG, immunoglobulin G; RBC, red blood cell. the blood smear in a typical case reveals polychromatophilia,
TRANSFUSION MEDICINE
spherocytes, a few fragmented RBCs, nucleated RBCs, and, BOX 41–1Transfusion Therapy and
occasionally, erythrophagocytosis. Examination of the bone
Autoimmune Hemolysis
marrow, which is rarely indicated, shows erythroid hyper-
plasia, often with megaloblastoid features. Occasionally, RBC There are times when a patient requires transfusion of incom-
autoantibodies or parvovirus B1928 cause reticulocytopenia patible RBCs or when transfusion must occur before completion
and dyserythropoiesis, thereby contributing to the severity of the blood bank evaluation. Autoimmune hemolytic anemia
of anemia.29 always complicates and prolongs the blood bank evaluation,
Patients with severe hemolytic anemia and markedly which may take up to 24 to 48 hours to complete. It is impera-
increased erythropoiesis occasionally experience folate defi- tive, however, that the patient receive RBCs expeditiously, even
though they are crossmatch-incompatible, if the hematocrit is not
ciency and frank megaloblastosis. The growth of hemato-
stabilized or cardiac or cerebral function is compromised. In
poietic tissue in the bone marrow also leads to moderate these situations, the patient should be very closely monitored
increases in the white blood cell and platelet counts. The and should receive the smallest volume of blood necessary to
absence of reticulocytosis does not exclude the diagnosis alleviate the life-threatening symptoms.
of AIHA but portends a serious prognosis.30–33 In addi-
tion, reticulocytopenia may represent excessive apoptosis of
erythroblasts.30,31,33,34 Presumably as a result of destruction
of young erythrocytes by the autoantibody, the reticulocyto-
Transfusion
penia aggravates the severity of the anemia and increases the
need for RBC transfusions. Some cases of AIHA may be life threatening and may neces-
sitate transfusion with RBCs (Box 41–1). It is important to
recognize that, in the majority of cases, the patient receives
Therapy crossmatch-incompatible blood.43–46 The presence of autoan-
tibodies complicates and prolongs the evaluation performed
General Principles by the blood bank. In situations of fulminant hemolysis,
The severity of AIHA may range from indolent to life threat- transfusion of incompatible blood or transfusion performed
ening. The impetus to initiate treatment, as well as the deter- on an emergency basis before completion of the evaluation
mination about whether the treatment required is immediate may be imperative and lifesaving. Severe anemia may cause
transfusion or an attempt to modulate the immune system’s high-output cardiac failure and subsequent pulmonary
production of autoantibody, must be based on a thor- edema, somnolence, and even obtundation, which require
ough appraisal of symptoms and the extent of the clinical immediate intervention with RBC transfusion. The hemo-
compromise. globin level at which these symptoms occur varies according
Rapidly developing anemia with a hematocrit less than to the rate of fall of the hemoglobin level, the capacity for
20% requires urgent management. In less aggressive forms of cardiac compensation, and other underlying clinical features.
III the disease, however, it may be prudent to allow physiologi- Occasionally (1% to 2% of cases), relative specificity of the
cally compensated anemia rather than to institute treatment. autoantibodies can be demonstrated. This specificity usually
560 The management of AIHA depends in part on whether the occurs within the Rh system, and RBCs lacking the corre-
disease is primary or is secondary to such disorders as a B-cell sponding Rh antigen survive better in vivo than those that
malignancy or SLE.35,36 This, too, demands a careful assessment express the antigen.47–50 Specificities of IgG autoantibodies
before any treatment begins. In some cases of AIHA secondary for multiple other blood groups have been described.51
to lymphoma or CLL, the pathogenic autoantibody (usually In addition, the blood bank must look for alloantibodies
monoclonal) is secreted by the neoplastic B cells. Combination that may be masked by the autoantibodies. Alloantibodies,
chemotherapy or irradiation of the underlying malignancy usually with specificity for the Rh or Kell blood group sys-
often brings the hemolytic anemia under control.37–39 tems, occur in approximately 30% of patients with AIHA
In other cases, however, the autoantibodies (usually who have a history of blood group immunization by mater-
polyclonal) do not originate from the B-cell neoplasm but nal-fetal incompatibility or previous transfusions.52–54 The
probably result from abnormal immune regulation insti- nonspecific serum autoantibodies that react with nearly
gated by the neoplastic B cells. Treatment of the latter type all normal RBCs must be removed to ensure that no con-
of secondary AIHA with immunosuppressive agents may comitant alloantibodies are present. An adsorption test is
improve the anemia but may also trigger an exacerbation.40 performed, using either the patient’s cells or cells of known
Multiple chemotherapeutic agents and immunosuppressive phenotype to absorb the autoantibodies from the serum—
medications interfere with T-cell function and can thus trig- a process known as autologous adsorption or heterologous
ger autoimmune processes in general and AIHA specifically. adsorption, respectively.
Fludarabine and cladribine given as therapy for CLL have These tests are not performed by all laboratories and
been demonstrated to precipitate autoimmune processes by should not be required before transfusion of a patient in need.
interfering with the balance of T- and B-cell functions.11 However, standard antibody detection and identification tests
The ultimate goal of therapy is control of the B-cell popu- with both the patient’s serum and an eluate prepared from
lations that secrete pathogenic autoantibodies. However, so the patient’s cells should be performed whenever possible.
little is known about such cells41,42 that the currently avail- Titration of the eluate and the serum against RBCs of various
able therapy is, by default, nonspecific and often aimed at Rh phenotypes can indicate an autoantibody specificity (or
reducing RBC clearance by macrophages. The desired thera- preference) within the Rh system. Any such specificity should
peutic effect is eradication of the abnormal hemolytic pro- be respected in the selection of donor units.55,56
cess, not reversal of the serologic abnormalities. Indeed, DAT Red blood cell substitutes have been transfused in a few
results often remain positive in the presence of a hematologic situations of severe hemolytic anemia and have demon-
response. strated benefit to the patients.57,58 Further investigation of
AUTOIMMUNE HEMOLYTIC ANEMIAS
these substitutes is necessary to determine their efficacy and accessory spleen in AIHA is meager. Faced with such a rare
safety. They could potentially be of benefit in the short-term finding in a patient with relapse, many hematologists would
emergent situation when the presence of underlying alloanti- recommend its removal. The role of splenectomy in patients
bodies cannot be ruled out or when the hemolysis is so brisk with mixed IgG, IgM, or mixed cold- and warm-reactive IgG
that transfusion of least incompatible cells does not result antibodies is unclear.
in any increase in hemoglobin and thus in oxygen-carrying
Rituximab Therapy
capacity.59
Rituximab, a chimeric anti-CD20 monoclonal antibody
Corticosteroids with a well-established, favorable safety profile, has possi-
Corticosteroids are the first line of treatment for most patients bilities for widespread clinical application in autoimmune
with symptomatic, unstable AIHA, either idiopathic or sec- disease in general and AIHA specifically.68,69 Rituximab
ondary. The clinical response to prednisone results primarily induces cell death through complement-dependent lysis,
from its ability to disable macrophages from clearing IgG- or antibody-dependent cellular toxicity, and cellular apopto-
C3b-coated erythrocytes. Corticosteroids interfere with both sis.68 Although plasma cells, fully differentiated B cells, are
the expression and function of macrophage Fc receptors. CD20-negative, the response rates of patients with AIHA and
This interference is probably the earliest, and perhaps even thrombocytopenia suggest that earlier CD20-positive B cells
the primary, mechanism in the ability of steroids to dimin- are producing antibody or that other as-yet not clearly delin-
ish the immune clearance of blood cells.60–63 Prednisone can eated effects of rituximab are effecting the disease remission.
also reduce autoantibody production, but only after several Reports that rituximab induces remission in patients with
weeks of therapy. idiopathic thrombocytopenic purpura within 1 week of first
The side effects of corticosteroids often preclude the long- infusion suggest that the mechanism of action may involve
term use of high-dose therapy. The cushingoid features that more than elimination of B cells.69
develop can lead to noncompliance, especially in adolescents. The efficacy and safety of rituximab have been dem-
The associated osteoporosis and immunosuppression as well onstrated in multiple large trials in adult patients with
as the risk of gastric bleeding may warrant discontinuation non-Hodgkin’s lymphomas.69 Infusion-related side effects,
of therapy. including fever, respiratory distress, and hypotension, are
reported to occur in a small population of lymphoma
Splenectomy patients.69–71 No side effects have precluded completion of
Splenectomy has been used as therapy for AIHA for many planned therapy in patients with autoimmune disease.72 The
years.64 Indications for splenectomy include failure to occurrence and severity of side effects seem to be related to
respond to prednisone, need for prednisone dosages higher B-cell level and, thus, tumor burden.70 It is anticipated and
than 10 to 20 mg/day, and intractable corticosteroid side has been demonstrated that the side effect profile will be
effects. The procedure can be highly effective, presumably even more favorable in patients with autoimmune disease
through removal of the major reticuloendothelial site of who have a lower level of B cells.73 41
RBC destruction; an animal model demonstrated that IgG- No large, prospective trials have studied the efficacy and
coated RBCs are removed almost exclusively by the spleen.60 safety of rituximab in patients with warm AIHA. However, 561
In addition, the procedure eliminates many phagocytosing the results of multiple case reports, case series, pilot stud-
macrophages and autoantibody-producing B cells. ies, and small therapeutic trials indicate that rituximab war-
In most young adults with chronic AIHA, the question of rants further study in patients with autoimmune disease and
splenectomy arises almost inevitably. However, in an elderly have prompted some experts in the field to attempt a trial of
patient with a stable but incomplete remission, mainte- rituximab prior to splenectomy in warm AIHA. The litera-
nance therapy with prednisone at a dose of 10 mg/day for ture reveals successful use of anti-CD20 treatment in mul-
an indefinite period may be the better alternative. There is a tiple autoimmune diseases.6,7,74–79
slight risk that overwhelming sepsis by encapsulated organ-
Rituximab Therapy: Mechanism(s) of Action
isms may develop immediately and up to 25 years after sple-
nectomy.65 The risk is higher in children, especially those Rituximab’s mechanism of action appears to be multifac-
younger than age 6, so a conservative approach to splenec- eted, complex, and incompletely understood. It is known
tomy is prudent in this age group. The risk of overwhelming that it induces cell death through complement-depen-
sepsis is lessened by immunization with pneumococcal and dent lysis; antibody-dependent cellular toxicity; antibody-
meningococcal vaccines, which are optimally administered dependent phagocytosis mediated by Fc, complement, and
at least 2 weeks preoperatively, and by the prompt use of phosphatidylserine receptors; direct antibody effects of
antibiotics for febrile illness.66 The Haemophilus influenzae CD20 ligation leading to inhibition of proliferation, apop-
series of vaccines should also be completed in children tosis, and sensitization to chemotherapy; and induction of
before splenectomy. active immunity.68 Although plasma cells are CD20-nega-
The response to splenectomy does not correlate with the tive, the response rate of patients with AIHA and throm-
age of the patient, the presence or absence of an underlying bocytopenia suggest that earlier, CD20-positive B cells are
B-cell disorder, the strength of the antiglobulin test result, producing antibody or that other as-yet not clearly delin-
prior response to prednisone, or the pattern of sequestration eated mechanisms are effecting disease remission. Reports
of chromium 51 (51Cr)-labeled RBCs. Between 50% and 60% that rituximab induces remission in patients with AIHA
of patients with classic AIHA have a good to excellent initial within 1 to 3 weeks of first infusion suggest that the mech-
response to splenectomy. They will need less than 15 mg/day anism of action may involve more than elimination of B
of prednisone to maintain an adequate level of hemoglo- cells. Table 41-3 summarizes the case reports and small case
bin.67 Information regarding the clinical implications of an series of patients with AIHA treated with rituximab.
III
TRANSFUSION MEDICINE
562
Table 41–3 Summary of Studies and Case Report Data for Rituximab for AIHA
15 pediatric/idiopathic79 375 mg/m2 Weekly × 2–4 87 (13/15) Two or more S, C, I, CyA, Az 2/15 7–28 months
1 pediatric/SLE,78 idiopathic158 375 mg/m2 Weekly × 2 100 (2/2) S, C, I, CyA 0/2 5–7 months
6 pediatric/idiopathic77 375 mg/m2 Weekly × 4 100 (6/6) S, I, Ph, CyA, Az 2/6 15–22 months
Weekly × 12 (2pt)
4 pediatric/idiopathic159 375 mg/m2 Weekly × 4–6 100 (4/4) S, V, CyA, Az, I, C, T 2/4 3–14 months
1 pediatric/idiopathic160 375 mg/m2 Weekly × 4 then repeat 100 (1/1) S, I, C, CyA, ATG, Ph, Az Yes 19 months
3–1 CAD, 2 WAIHA161 375 mg/m2 Weekly × 4 33 (1/3) S, C, Az 0/4 96 months
6 adult/5 lymphoma; 375 mg/m2 Weekly × 4 17 (1/6) PR in 4/6 3 untreated S, C, Cl, Chl 1/6 6–14 months
1 idiopathic123
1 pediatric/Hurler’s post-BMT162 375 mg/m2 Weekly × 3 100 (1/1) I, S, CyA 0/1 21 months
1 pediatric post-BMT/ 375 mg/m2 Weekly × 2 100 (2/2) S, I 0/2 3–12 months
β-thalassemia,163 WAS164
Weekly × 4
1 adult/idiopathic CAD+ 375 mg/m2 Weekly × 4, then 100 (1/1) S × 2 long tapers No 7 months
WAIHA3 1 month later
Weekly × 4
9 adult/CAD +/− 375 mg/m2 Weekly × 4 100 (9/9) S, CyA, Chl, Az, C, MM, 1/9 9–36 months
lymphoma/CLL165–172 CHOP for lymphoma
6 adults/CAD173 375 mg/m2 Weekly × 4 1 CR, 4 PR, 1 NR S, Chl, C, Cl 1/6 6–14 months
8 adult/CLL174 375 mg/m2 Q 4 weeks 100 (8/8) S, F, C, Chl 0/8 7–23 months
Given w/ C and D
27 adults/CAD {123} 375 mg/m2 Weekly × 4 +/− IFN ICR, 19 PR S, C, MP Not reported 2–42 months
*
Regimens: ATG, antithymocyte globulin; Az, azathioprine; C, cyclophosphamide; Ch l, chlorambucil; Cl, cladribine; CyA, cyclosporine; I, intravenous immune globulin; IFN, interferon; MM,
mycophenolate mofetil; MP, mercaptopurine; Ph, plasmapheresis; S, corticosteroids; T, tacrolimus; V, vincristine.
AUTOIMMUNE HEMOLYTIC ANEMIAS
Intravenous Immune Globulin there are some reports that this modality is efficacious in
IgG-mediated disease.92 Occasional dramatic responses have
Intravenous immune globulin (IVIG) has been found
been reported in patients being prepared for surgery or when
effective in the management of selected cases of AIHA.
plasma exchange was used as a temporizing measure after
The soluble IgG in the material may lengthen the life span
initiation of immunosuppressive therapy.92,93 This therapy
of IgG-coated RBCs by saturating Fc receptors on macro-
is reserved for patients in critical condition whose AIHA is
phages. In a study of patients who had AIHA associated
unresponsive to transfusion because of rapid destruction
with lymphoproliferative disorders, a long-term benefit
and clearance of the RBCs.
was observed with a maintenance dosage schedule of
intravenous IgG every 21 days. A decrease in antiglobu-
lin titer was found in these patients, suggesting a mecha- COLD AGGLUTININ DISEASE
nism other than blockade of Fc receptors by intravenous
IgG.80 The mechanism of action of IVIG has been further
Cold agglutinin disease refers to a group of disorders caused
elucidated by Samuelsson and colleagues.81 They inves-
by anti-erythrocyte autoantibodies (e.g., cold agglutinins)
tigated the mechanism of protection in a murine model
that preferentially bind RBCs at cold temperatures (4° C to
of immune thrombocytopenia that may have relevance
18°C) and may or may not induce hemolysis. Virtually all
to mechanism of action in other autoimmune diseases.
sera from healthy individuals contain low-titer cold agglu-
Their model demonstrated that the inhibitory Fc recep-
tinins, which are regarded as benign or harmless RBC auto-
tor FcγRIIB was necessary for IVIG to confer protection
antibodies and are considered polyclonal. Similarly, cold
against platelet destruction.81 This finding suggests that
agglutinins that arise after certain infections are also poly-
modulation of inhibitory signals in macrophages could
clonal and usually benign; in rare cases, a transient form
possibly be involved in autoimmunity and could be investi-
of cold agglutinin disease ensues. By contrast, monoclonal
gated as therapeutic targets in other autoimmune diseases,
cold agglutinins are generally pathogenic and are derived
such as AIHA.
from clonal B-cell expansions (as in idiopathic or chronic
Immunosuppressive Therapy cold agglutinin disease), which may be a prelude to frank
lymphoma.
Most experience with immunosuppressive drugs in the treat-
ment of AIHA has been with alkylating agents (cyclophos-
phamide and chlorambucil) and thiopurines (azathioprine Chronic Cold Agglutinin Disease
and mercaptopurine).82 The basis for the clinical use of these The most common type of cold agglutinin disease, a chronic
drugs is their inhibitory effect on the immune system, pos- form characterized principally by a stable anemia of moder-
sibly affecting both B cells and T cells.83,84 ate severity and attacks of acrocyanosis precipitated by ex-
Cyclophosphamide and azathioprine, like prednisone, can posure to cold, constitutes about one third of all cases of
induce numerous side effects. The early side effects include immune hemolytic anemia. Cold agglutinins cause the car- 41
bone marrow suppression and impairment of the immune dinal abnormalities of the disease. The acrocyanosis stems
response (particularly T cell–mediated immunity) that occur from intra-arteriolar agglutination of erythrocytes in the 563
concomitantly with therapy. After sustained administration, relatively cool tips of the fingers, feet, ear lobes, and nose.
cyclophosphamide may damage ovarian function, inhibit The occurrence of this hemolytic anemia depends on the
spermatogenesis,85–88 and cause bladder fibrosis.89 Acute capacity of the cold agglutinins to initiate activation of
myeloid leukemia can develop years after administration of the complement cascade on the surface of the RBC. Most
this drug.83 By contrast, the prolonged use of azathioprine patients with chronic cold agglutinin disease are in the fifth
has not been associated with a statistically significant increase to eighth decade of life and have a B-cell neoplasm or lym-
in malignant diseases. All of these considerations mandate phoma, Waldenström macroglobulinemia, or CLL. The cold
careful monitoring of any patient treated with either cyclo- agglutinin in cases secondary to such diseases is monoclo-
phosphamide or azathioprine. nal, almost always IgMκ, and may show up as a monoclonal
Cyclosporine, a powerful T-cell modulator, has been used band in the γ region of the serum protein electrophoretic
alone and in combination to elicit successful and sometimes pattern. In the absence of a B-cell neoplasm, the spleen and
durable remission in patients with AIHA and Evans’s syn- lymph nodes are rarely enlarged; therefore, the finding of
drome.90 Cyclosporine and other immunosuppressive agents splenic and lymph node enlargement warrants a search for
are discussed here and have been reported in the literature. the neoplasm.
However, their long-term use is not recommended, because
the benefits do not usually offset the side effects inherent in
the prolonged use of these agents. Transient Cold Agglutinin Disease
A second type of cold agglutinin disease, usually acute and
Plasma Exchange always self-limited, occurs as a rare complication of several
Because a single-volume plasma exchange replaces only infectious diseases, most notably Mycoplasma pneumoniae
about 60% of the patient’s plasma volume, its therapeutic infection and infectious mononucleosis. Patients with this
advantage lies in the removal of plasma antibodies to IgG, form of cold agglutinin disease are therefore much younger
IgM, or both, which mediate the hemolysis. Unfortunately, than those with chronic cold agglutinin disease. The onset is
continuous antibody production and the large extravascu- abrupt, occurring as the infection wanes, and the anemia can
lar distribution of IgG limit the long-term efficacy of plasma be severe. Cold agglutinin titers are moderately elevated, and
exchange in IgG-mediated AIHA. On cessation of therapy, the cold agglutinins are polyclonal. Often, these polyclonal
the rate of return to pretreatment levels of autoantibody cold agglutinins coincide with high-titer warm-reactive IgG
depends on the rate of autoantibody production.91 However, RBC autoantibodies.
TRANSFUSION MEDICINE ANTIGENIC TARGETS OF COLD produce a picture of immune complex–mediated vasculitis,
AGGLUTININ DISEASE with vascular purpura, arthritis, and nephritis as the domi-
nant complications. In occasional patients, the cryoglobulin
The antigenic specificity of cold agglutinins is usually identi- can also be a cold agglutinin.100–103
fied from their degree of reactivity with RBCs from adults
(blood group I) and from cord blood (blood group i). The
Pathophysiology
cold-reactive autoantibody produced after some cases of
M. pneumoniae infection has anti-I specificity,94 whereas the The pathogenic IgM autoantibody in cold agglutinin disease
antibody in infectious mononucleosis frequently, but not is highly efficient in activating the classic complement path-
always, has anti-i specificity.95,96 Additional specificities have way on the erythrocyte membrane.36,104 However, the thermal
been identified by tests with rare adult RBCs that lack the dependency of the antibody constrains its pathogenic effects.
I-antigen or with enzyme-treated erythrocytes. Rarely, cold The autoantibody rapidly elutes off RBCs at 37°C, the tem-
agglutinins are specific for the A blood group antigen.46 perature of the visceral circulation, but in the cool peripheral
circulation of the hands and feet, the cold agglutinin remains
Laboratory Evaluation on the erythrocyte membrane for at least a few seconds. That
amount of time is sufficient to activate the complement cas-
The usual laboratory findings in hemolytic anemia (i.e., ane- cade to the stage of C3b, which adheres to the RBC after it
mia, reticulocytosis, polychromatophilia, spherocytosis, ery- reenters the central circulation. In the hepatic circulation,
throid hyperplasia in the bone marrow, and elevations in serum C3b-positive RBCs encounter macrophages with receptors
bilirubin and LDH levels) are generally not striking in chronic specific for C3b19,105,106; however, C3b sensitization is only a
cold agglutinin disease. Hemagglutination may be visible to the weak signal for the activation of phagocytosis—the hepatic
unaided eye in blood drawn from a patient with cold aggluti- clearance of C3b-coated RBCs requires 500 to 800 C3b mol-
nin disease and can interfere with automated blood counts. The ecules per RBC. As a result, many C3b-positive RBCs escape
anemia is often mild and stable, because the C3b inactivator in unharmed into the systemic circulation, where they come
serum limits the extent of cold agglutinin–induced complement under the influence of the regulatory proteins of the com-
activation on the erythrocyte membrane. Exposure to cold plement system. The C3b inactivator system degrades C3b
may greatly augment the binding of cold agglutinins, however, into C3dg, C3d, or both. The result is a cohort of erythro-
exceeding the restraints of the inactivator system. That occur- cytes coated with C3d but not with the IgM autoantibody.107
rence can lead to a sudden drop in hematocrit value, with com- Macrophages bind to C3d with even lower avidity than to
plement-mediated intravascular hemolysis and renal failure. C3b, thus the C3d-positive erythrocytes tend to have near-
In a distinctive subset of patients with aggressive cold agglu- normal survival in vivo despite a heavy coating with that
tinin disease, the cold agglutinin titer is relatively low but the degradation product of C3.108,109 It is important to recognize
autoantibody has a high thermal amplitude. Recognition that if transfusion is necessary, the transfused cells will not
III that a patient has this variant of cold agglutinin disease is have the protection conferred by C3d and therefore may be
important because it may respond to prednisone,62 whereas rapidly lysed.109
564 high-titer cold agglutinin disease usually does not. These limits on the pathogenicity of cold agglutinins
In typical cases of chronic cold agglutinin disease, the cold account for the subdued hematologic picture in most patients
agglutinin titer is very high (>1:105 and occasionally >1:106). with cold agglutinin disease. If, however, the regulatory C3b
The antibodies are most reactive in the cold, and hemagglu- inactivator proteins are impaired, limiting cleavage of RBC-
tination disappears as the temperature rises toward 37° C. In bound C3b, or if the production of IgM autoantibodies with
some cases, however, the antibody is reactive at relatively high a high thermal amplitude is impaired, permitting comple-
temperatures and occasionally even at 37°C. The reactivity tion of the complement cascade in the visceral circulation,
of the cold agglutinin at high temperatures (i.e., its thermal severe extravascular hemolysis can occur. Several patients
amplitudes), not the titer of the antibody, most accurately with high titers of IgA cold agglutinins have been reported.
predicts the severity of the disease. The DAT result is positive Such cases are not associated with cold agglutinin disease,
because of erythrocyte-bound C3d, but results of tests with which may relate to the lack of complement activation by
anti-IgG reagents are negative. The result of the IAT, which IgA antibodies.109–118
is conducted at 37° C, is negative. In addition to monoclonal
IgM cold agglutinins, IgG-IgM mixed cold agglutinins have
been reported.97–99 Besides the usual high titers of IgM cold Therapy
agglutinins, some patients with cold agglutinin disease have
low titers of IgG and IgA cold agglutinins. Chronic Cold Agglutinin Disease
Cold agglutinins are not cryoglobulins. The latter are Therapy for the cold agglutinin syndromes depends on the
most often monoclonal IgM immunoglobulins that, in the gravity of the symptoms, the serologic characteristics of the
cold, either self-associate and precipitate from solution (type I autoantibody, and any underlying disease. In the idiopathic,
cryoglobulinemia) or precipitate as complexes with poly- or primary, form of chronic cold agglutinin disease, pro-
clonal IgG molecules (type II cryoglobulinemia, often due longed survival and spontaneous remissions and exacerba-
to a monoclonal IgM rheumatoid factor). Type III cryo- tions are not unusual. The anemia is generally mild, and the
globulins consist of a mixture of polyclonal IgM and poly- simple measure of avoiding exposure to cold temperatures can
clonal IgG immunoglobulins. The clinical manifestations of avoid exacerbations, especially if the cold agglutinin respon-
the cryoglobulinemic syndromes are highly variable: type sible has a low thermal amplitude. Prednisone has been ben-
I and type II cryoglobulinemias occur in B-cell neoplasms eficial in rare cases in which there are relatively low titers of
(Waldenström macroglobulinemia, multiple myeloma, lym- cold agglutinins of a high thermal amplitude or an IgG cold-
phoma, and CLL); type II and type III cryoglobulinemias can reactive antibody is produced. However, prednisone is not
AUTOIMMUNE HEMOLYTIC ANEMIAS
useful therapy in most patients with primary IgM-induced characterized by constitutional symptoms with fulminant
cold agglutinin disease, and its administration should not be intravascular hemolysis and its associated signs of hemoglo-
undertaken lightly, given the chronicity of the disease.119,120 binemia, hemoglobinuria, jaundice, severe anemia, and some-
Plasma exchange may help as a temporary measure in acute times renal failure. The disease is self-limited, usually lasting
situations.93 Splenectomy is usually ineffective because the 2 to 3 weeks, although it can be life threatening because of the
liver is the dominant site of sequestration of RBCs heavily severity of the hemolysis and consequent anemia.
sensitized with C3b. However, rare cases in patients with
enlarged spleens have responded to splenectomy; in some
Laboratory Evaluation
of these patients, a localized splenic lymphoma was found,
whereas in others, only lymphoid hyperplasia was evident. The IgG antibody responsible for PCH is found in the patient’s
It is essential to seek evidence of a B-cell neoplasm before serum through incubation of normal erythrocytes, fresh
therapy for chronic cold agglutinin disease is initiated. Oral normal serum as a source of complement, and the patient’s
alkylating agents (chlorambucil or cyclophosphamide) help serum, first at 4°C and then at 37°C, with appropriate controls.
many patients with the secondary form of cold agglutinin The Donath-Landsteiner antibody fixes the first two compo-
disease because of their effect on the B-cell neoplasm, but nents of complement in the cold and completes the cascade
only occasionally do they benefit patients with the primary on warming to 37°C.128 The DAT result is almost always nega-
form of the disease.121,122 When cold agglutinin disease is part tive, but occasionally weak reactions for erythrocyte-bound
of an established B-cell malignancy, the severity of hemoly- complement are manifested. The IAT result is negative. Most
sis often waxes and wanes in parallel with the activity of the Donath-Landsteiner antibodies have specificity for the P
neoplasm. blood group system,129,130 but other specificities have been
Patients with IgM-mediated hemolysis can be dramati- described.131–133 The diagnosis depends on recognition of
cally helped with plasma exchange. Because of its large size, the clinical picture, because tests for the Donath-Landsteiner
the antibody is located primarily in the intravascular space antibody are not routinely performed.
and is efficiently removed by plasma exchange. Some patients
with Waldenström macroglobulinemia are maintained over Therapy
the long term with this therapy.
A recent publication of a prospective, phase II study of No specific treatment for PCH has been found. Unlike the
patients with primary cold agglutinin disease reported a effectiveness of steroids in most IgG-mediated autoim-
54% response rate in 27 patients treated with 37 courses of mune diseases, prednisone is not useful for PCH. The best
rituximab. Some of the patients who did not respond to the approach consists of supportive care, transfusions to alleviate
first course of four weekly doses of rituximab were treated symptoms, and avoidance of cold temperatures. The patient
with a combination of rituximab and interferon-α. Of the should be kept in a warm room, and transfusions should be
responses, one patient achieved complete remission and 19 given through a blood warmer.
partial remission. The complete remission was sustained at 41
42 months, although all but one of the patients with partial
remission relapsed at a mean of 11 months.123 DRUG-ASSOCIATED IMMUNE 565
HEMOLYTIC ANEMIA
Transient Cold Agglutinin Disease
Transient cold agglutinin disease is a rare form that is always Drug-associated immune hemolytic anemia can be either
self-limited. Supportive measures, including transfusions induced by or dependent on a drug. Four distinct mecha-
and avoidance of cold, may suffice to tide the patient over the nisms are associated with the disorder. The first involves any
episode of hemolysis. Corticosteroids are usually not helpful, drug that can bind to the RBC membrane. The patient then
and splenectomy is almost never indicated. makes antibodies against the drug (e.g., penicillin), which
combine with the erythrocyte-bound drug, opsonizing and
preparing the RBC for destruction. Discontinuation of the
PAROXYSMAL COLD HEMOGLOBINURIA drug brings the hemolytic anemia to a rapid halt, because
the antibodies have no specificity for antigens on the RBC
membrane. Clues to the diagnosis are the appropriate clini-
Clinical Features
cal setting, a positive DAT result, a negative IAT result, and
Paroxysmal cold hemoglobinuria was historically associ- failure of antibodies eluted from the patient’s RBCs to bind
ated with tertiary syphilis, which is rarely seen today. PCH to normal erythrocytes. The diagnosis is established when
is now more commonly seen primarily in children after a both the eluate and the patient’s serum contain antibodies
viral or, much less commonly, bacterial illness.124 Most com- directed against the drug-coated cells. In the case of penicil-
monly, the viral etiology is not known but is associated with lin,134 hemolytic anemia occurs only when large amounts are
an upper respiratory tract infection. However, case reports administered; in patients treated with lower doses, a posi-
in both adults and children have reported an association of tive DAT result without hemolytic anemia is not unusual,
PCH with varicella. because the production of low-avidity IgG antipenicillin
Although the Donath-Landsteiner antibody often occurs antibodies is a common event.
in tertiary or congenital syphilis, it generally does not The second mechanism involves immune complexes. The
cause hemolytic disease in this situation. On exposure to offending drug, or drug metabolite, binds to a plasma protein,
cold, an occasional patient experiences paroxysms of hemo- forming an immunogenic conjugate. If the patient develops
globinuria and constitutional symptoms (fever, back pain, an antibody to the conjugate, it is usually IgM. This antibody
leg pain, abdominal cramps, and rigors) followed by hemo- then binds to the immunogenic conjugate, forming an
globinuria. In contrast, the postviral form of PCH125–127 is immune complex that adheres to RBCs. The resulting clinical
TRANSFUSION MEDICINE
picture consists of intravascular hemolysis and concomitant develop—reticulocytosis, spherocytosis, a shortened RBC
hemoglobinemia, hemoglobinuria, and even renal failure survival time, and splenomegaly.141
through efficient activation of complement on the eryth- Okamoto and colleagues142 developed a transgenic murine
rocyte membrane. This chain of events accounts for most model of AIHA. The symptoms in this model range from
reported examples of drug-induced immune hemolytic ane- unaffected to severe anemia. The B1 subpopulation has been
mia. Reports concerning the nonsteroidal drug diclofenac demonstrated to mediate the AIHA.143 They are activated
have shown that autoimmune hemolytic anemia is induced in both T cell–dependent and T cell–independent ways.144
by sensitization to the glucuronide conjugate of the drug.135 These cells are unique from B2, the more prevalent B cells,
Serologic findings in erythrocyte-bound immune complexes in several ways. B1 cells preferentially locate in the perito-
are similar to those of the first mechanism, except that the neal and pleural cavities, produce 50% of the natural serum
DAT reveals only complement bound to the RBC; the IgM IgM, and escape clonal deletion in these immunoprivileged
antibody is presumed to be no longer present after comple- sites. Interleukin-10 influences T cell–dependent prolifera-
ment activation. The patient’s serum reacts with RBCs (lack- tion of B1 cells and continuous administration of anti-inter-
ing antidrug antibody) in the presence of the offending drug, leukin-10 monoclonal antibody depletes B1 but not B2 cells
and the eluate from the patient’s RBCs generally does not in murine models. The influence of this Th2 cytokine on the
react with normal erythrocytes. behavior of the B1 cells in vivo145 suggests possible avenues
The third mechanism involves in vivo sensitization to for treatments.
drugs through the formation of immunogenic drug–RBC
Origins of Anti-erythrocyte Autoantibodies
complexes. In these cases, the specificity of the drug-induced
antibodies is contributed to not only by the drug (or its The vast improvement in our understanding of what pre-
metabolites), but also by defined RBC antigens, particularly vents autoimmunization has not yet informed us of the
those of the Rh and I/i systems.136 mechanism that causes autoimmunization. Very little is
The fourth mechanism of drug-associated immune known of the origins of warm-reactive IgG anti-erythro-
hemolytic anemia involves the induction of authentic auto- cyte autoantibodies, despite the availability of a thoroughly
antibodies against RBCs by a drug; methyldopa is the clas- investigated, spontaneous animal model of the disease (the
sic example.137 In as many as 20% of patients treated with NZB mouse) and stocks of pathogenic autoantibodies,
methyldopa, the DAT result turns positive, but few demon- which are readily obtained from patients with the disease.
strate hemolytic anemia. The DAT result may take several A major impediment to advances in our understanding of
months to a year or more after the start of drug therapy to how autoimmune hemolytic anemia originates is that the
become positive. In patients with hemolytic anemia, dis- autoantigens are for the most part unknown. Even in those
continuation of the drug results in only gradual cessation cases in which blood group specificity of the autoantibodies
of the hemolytic anemia and disappearance of the autoan- has been identified, the relevant structures have not been
tibody, because the drug itself is not required for the hemo- elucidated. Leddy and associates146 have succeeded in iden-
III lytic process, only for the initiation of antibody production. tifying four proteins on the RBC membrane that bind to
Curiously, the autoantibody is usually specific for antigens anti-erythrocyte autoantibodies; they are the band 3 anion
566 of the Rh system. The serologic findings are indistinguish- transporter, glycophorin A, and two polypeptides, probably
able from those in primary AIHA; they consist of a positive related to the Rh family of antigens. Various combinations
DAT result, usually a positive IAT result, and an eluate that of those four autoantibody specificities were found in a
reacts with normal erythrocytes. Patients taking methyldopa group of 20 patients with AIHA.
often have other antibodies in addition to the RBC auto- The association of AIHA with SLE and with immune
antibodies. The mechanism by which methyldopa induces thrombocytopenia (Evans’s syndrome), the induction of the
autoantibodies is unknown, but it may involve effects on disease by drugs that seem to perturb immune regulation,
immunoregulatory T cells. and the graft-versus-host model of AIHA all suggest that, at
Drug-induced immune hemolytic anemia was commonly least in some cases, there is antigen-independent activation
seen when penicillin was administered in large doses (i.e., of clones of B cells with the capacity to produce IgG anti-
>20 million U/day) and when methyldopa was widely used RBC autoantibodies. Such polyclonal B-cell activation may
in the treatment of hypertension.138,139 However, the disease account for the production of anti-erythrocyte autoantibod-
is unusual in present-day clinical practice.140 Numerous ies in patients with acquired immunodeficiency syndrome
drugs can induce hemolytic anemia. (AIDS).94,147 Hypergammaglobulinemia and other signs of
nonspecific activation of B cells are prominent in human
immunodeficiency virus infection.148
ANIMAL MODELS OF AUTOIMMUNE The immunologic basis of AIHA in patients with CLL or
HEMOLYTIC ANEMIA a B-cell lymphoma is equally obscure.149 In CLL, the autoan-
tibodies are IgG and often polyclonal,150 whereas the malig-
Insights into Pathogenesis nant CD5+ B cells of that disease generally produce only IgM
and Therapeutic Targets antibodies that are monoclonal. It is therefore likely that B
cells other than those constituting the leukemia produce the
NZB Mice autoantibodies. The large mass of CD5+ B cells in CLL might
The inbred NZB mouse is genetically programmed to induce nonneoplastic CD5− B cells to produce IgG autoanti-
develop AIHA at around age 6 to 8 months (the life span bodies, perhaps through a disturbance of immunoregulatory
of a normal mouse is approximately 2 years). Anti-eryth- idiotypic networks. The demonstration of the simultaneous
rocyte autoantibodies begin to appear at around age 3 presence of autoantibodies and anti-idiotypic antibodies on
months; by age 9 months, the DAT result is positive in RBCs in AIHA151 suggests that such networks may indeed
60% to 80% of animals. Typical signs of hemolytic anemia have a role in the disease.
AUTOIMMUNE HEMOLYTIC ANEMIAS
In contrast to the antigens that incite warm-reactive auto- serum are not restricted to the VH4-34 gene segment. They
antibodies, the structures of the autoantigens of cold aggluti- are associated with different genes of the VH3 family as well
nin disease, the I/i system, are known152; this knowledge has as the VH4-34 gene.156 It therefore appears that B-cell neopla-
clarified our thinking about the immunology of this group sia is an important but not exclusive element in the asso-
of disorders. There is little reason to doubt that the very high ciation between VH4-34 and cold agglutinins. The correlation
levels of monoclonal cold agglutinins found in some patients with lymphomas has additional interest, because VH4-34 has
with B-cell neoplasms are produced by the malignant cells. been independently linked to B-cell lymphomas that do not
The demonstration that an idiotypic marker on monoclonal secrete cold agglutinins.157
cold agglutinins could be detected not only on the patients’
neoplastic B cells, but also on 3% to 10% of normal B cells153
supports the view that these autoantibodies are part of the PERSPECTIVE
normal immune repertoire; malignant transformation of
a cold agglutinin–producing B cell results in a lymphoma In previous years, there have been many debates concerning
complicated by chronic cold agglutinin disease. the factors that contribute to the severity of AIHA. Previous
The basis of the association of PCH with syphilis may be studies have focused to a large extent on the humoral aspect
antigenic mimicry, in which structural similarities between of the autoimmune response. The antibodies were easily
a microbial antigen and a self-antigen trigger an autoanti- available from peripheral blood, allowing for many serologic
body response. In the case of PCH, the infecting organism, investigations. However, none of the serologic parameters
Treponema pallidum, should possess two antigenic deter- by itself—the quantity of serum RBC autoantibodies, titer,
minants (epitopes), one recognized by T cells (the foreign thermal amplitude, or allotype—has proved useful in pre-
epitope) and the other by self-reactive B cells (the mim- dicting severity of the disease in patients. This fact is per-
icking epitope). Donath-Landsteiner antibodies would be haps not surprising, because RBC clearance in both cold
produced only by syphilitic patients whose class II major and warm AIHA occurs predominantly extravascularly
histocompatibility complex glycoproteins could present through the actions of macrophages rather than intravas-
the foreign epitope in an immunogenic form to T cells. cularly through antibody-mediated complement lysis. Also,
A similar mechanism could apply to postinfectious acute membrane receptors on macrophages, such as FcγΙIRB and
cold agglutinin disease, in which a crossreaction involving SIRPα (signal regulation protein-α), have been identified
antigenic determinants of M. pneumoniae and the I blood that have the ability to modulate RBC clearance and, thus,
group substance has been incriminated.154 severity of hemolysis. Therefore, it is possible that the activ-
Structural analyses of monoclonal anti-I and anti-i auto- ity of these receptors may vary among patients as well, con-
antibodies from patients with B-cell neoplasms are begin- tributing to the severity of disease expression. Moreover,
ning to yield important clues about the origins of chronic these macrophage receptors may prove to be viable targets
cold agglutinin disease. A striking observation is the repeti- for the development of more specific immunotherapy than
tive use of the same immunoglobulin VH gene, VH4-34, in the methods currently used (i.e., corticosteroid therapy and 41
monoclonal IgM cold agglutinins, regardless of the anti-I IVIG). Another potentially exciting therapeutic approach is
or anti-i specificity of the autoantibody.155,156 In each case, targeting of the humoral immune response, either through 567
the VH4-34 heavy chain gene had a different CDR3 (comple- direct targeting of the B cells, such as with anti-CD20 anti-
mentarity-determining region 3); the light chains of cold body, or through interference with the interactions between
agglutinins with anti-I or anti-i specificity differed as well. B and T lymphocytes, such as with costimulatory molecule
The VH4-34 genes of these cold agglutinins contained few or blockade (e.g., anti-CD40 antibody).
no somatic mutations of the type that would lead to amino
acid substitutions (replacement mutations). This finding,
together with the variations in their CDR3 and in the light REFERENCES
chains, implies that the VH4-34 germline gene segment itself 1. Sloan S, Silberstein LE. Transfusion: in the face of autoantibodies.
encodes a binding site for the I and i antigens. In contrast to In Reid M, Nance SJ (ed). Red Cell Transfusion: A Practical Guide.
the heavy chain gene, the light chain genes of the cold agglu- Totowa, N.J., Humana Press, 1998.
tinins do contain replacement mutations, especially in their 2. Coombs R, Mourant AE, Race RR. A new test for the detection of weak
and incomplete Rh agglutinins. Br J Exp Pathol 1945;26:255.
hypervariable regions.156 3. Boorman K, Dodd BE, Loutit JF. Hemolytic icterus (acholuric jaun-
It appears from these results that the germline VH4-34 heavy dice): congenital and acquired. Lancet 1946;1:812.
chain encodes the dominant specificity of monoclonal cold 4. Lewis R, Henry WB, Thornton GW, et al. A syndrome of autoimmune
agglutinins, the somatic mutations of the light chain genes hemolytic anemia and thrombocytopenia in dogs. Proc J Am Vet Med
of the cold agglutinins are the result of an immune response, Assoc 1963;1:140.
5. Bielschowsky M, Helyer BJ, Howie JB. Spontaneous hemolytic ane-
and the VH CDR3 and the light chain confer fine specific- mia in mice of the NZB/Bl strain. Proc Univ Otago Med Sch 1959; 37:9.
ity (e.g., for I or i) on the cold agglutinin and influence its 6. Petschner F, Walter UA, Schmitt-Graff A, et al. “Catastrophic systemic
affinity. These data make a convincing case that monoclonal lupus erythematosus” with Rosai-Dorfman sinus histiocytosis. Suc-
cold agglutinins arise as the result of an immune response, cessful treatment with anti-CD20/rutuximab. Dtsch Med Wochenschr
2001;126:998–1001.
perhaps an autoimmune response to an autoantigen on 7. Stasi R, Pagano A, Stipa E, et al. Rituximab chimeric anti-CD20 mono-
erythrocytes. The results of these molecular studies of cold clonal antibody treatment for adults with chronic idiopathic thrombo-
agglutinins complement other evidence favoring a role for cytopenic purpura. Blood 2001;98:952–957.
antigen-mediated clonal selection in some types of B-cell 8. Sackey K. Hemolytic anemia: part 1. Pediatr Rev 1999;20:152–159.
neoplasms. 9. Pirofsky B. Autoimmunization and the Autoimmune Hemolytic Ane-
mias. Baltimore, Williams & Wilkins, 1969.
In contrast to monoclonal cold agglutinins associated 10. Myint H, Copplestone JA, Orchard J, et al. Fludarabine-related auto-
with chronic cold agglutinin disease, the naturally occurring immune haemolytic anaemia in patients with chronic lymphocytic
IgM cold agglutinins that are present in low titers in normal leukaemia. Br J Haematol 1995; 91:341–344.
TRANSFUSION MEDICINE 11. Robak T, Blasinska-Morawiec M, Krykowski E, et al. Autoimmune body is derived from a chromosomally aberrant B cell clone. Blood
haemolytic anaemia in patients with chronic lymphocytic leukaemia 1986;67:1705–1709.
treated with 2-chlorodeoxyadenosine (cladribine). Eur J Haematol 40. Rosenthal MC, Pisciotta AV, Komninos ZD, et al. The auto-immune
1997;58:109–113. hemolytic anemia of malignant lymphocytic disease. Blood 1955;10:
12. Drobyski WR, Potluri J, Sauer D, et al. Autoimmune hemolytic anemia 197–227
following T cell-depleted allogeneic bone marrow transplantation. 41. Hamblin TJ, Oscier DG, Young BJ. Autoimmunity in chronic lympho-
Bone Marrow Transplant 1996;17:1093–1099 cytic leukaemia. J Clin Pathol 1986;39:713–716.
13. Sokol RJ, Stamps R, Booker DJ, et al. Posttransplant immune-medi- 42. Sikora K, Krikorian J, Levy R. Monoclonal immunoglobulin rescue
ated hemolysis. Transfusion 2002;42:198–204. from a patient with chronic lymphocytic leukemia and autoimmune
14. Beckers EA, van Guvdener C, Overbeeke MA, et al. Intravascular hemo- hemolytic anemia. Blood 1979;54:513–518.
lysis by IgA red cell autoantibodies. Neth J Med 2001;58:204–207. 43. Petz LD. Transfusing the patient with autoimmune hemolytic anemia.
15. Janeway C, Travers P, Walport M, Capra JD (eds). In Immunobiology: Clin Lab Med 1982;2:193–210.
the Immune System in Health and Disease, 4th ed. New York, Garland 44. Plapp FV, Beck ML. Transfusion support in the management of immune
Publishing, 1999. haemolytic disorders. Clin Haematol 1984;13:167–183.
16. Domen RE. An overview of immune hemolytic anemias. Cleve Clin J 45. Rosenfield RE, Jagathambal. Transfusion therapy for autoimmune
Med 1998;65:89–99. hemolytic anemia. Semin Hematol 1976;13:311–321.
17. Huber H, Polley MJ, Linscott WD, et al. Human monocytes: distinct 46. Sokol RJ, Hewitt S, Booker DJ, et al. Patients with red cell autoantibodies:
receptor sites for the third component of complement and for immu- selection of blood for transfusion. Clin Lab Haematol 1988;10:257–264.
noglobulin G. Science 1968;162:1281–1283. 47. Dacie JV, Cutbush M. Specificity of auto-antibodies in acquired hae-
18. LoBuglio AF, Cotran RS, Jandl JH. Red cells coated with immunoglob- molytic anaemia. J Clin Pathol 1954;7:18–21.
ulin G: binding and sphering by mononuclear cells in man. Science 48. Hogman C, Killander J, Sjolin S. A case of idiopathic auto-immune
1967;158:1582–1585. haemolytic anaemia due to anti-e. Acta Paediatr 1960;49:270–280.
19. Fleer A, van der Meulen FW, Linthout E, et al. Destruction of IgG- 49. Sachs V. Anti-C as a sole autoantibody in autoimmune hemolytic ane-
sensitized erythrocytes by human blood monocytes: modulation of mia. Transfusion 1985;25:587–588.
inhibition by IgG. Br J Haematol 1978;39:425–436. 50. Weiner W, Battey DA, Cleghorn TE, et al. Serological findings in a case
20. Schreiber AD, Frank MM. Role of antibody and complement in the of haemolytic anaemia, with some general observations on the patho-
immune clearance and destruction of erythrocytes. I. In vivo effects of genesis of this syndrome. BMJ 1953;2:125–128.
IgG and IgM complement-fixing sites. J Clin Invest 1972;51:575–582. 51. Issitt P, Anstee DJ. In Montgomery (ed). Applied Blood Group Serol-
21. Schreiber AD, Frank MM. Role of antibody and complement in the ogy, 4th ed. Durham, N.C., Scientific Publications, 1998.
immune clearance and destruction of erythrocytes. II. Molecular 52. James P, Rowe GP, Tozzo GG. Elucidation of alloantibodies in autoim-
nature of IgG and IgM complement-fixing sites and effects of their mune haemolytic anaemia. Vox Sang 1988;54:167–171.
interaction with serum. J Clin Invest 1972;51:583–589. 53. Laine ML, Beattie KM. Frequency of alloantibodies accompanying
22. Meite M, Leonard S, Idrissi ME, et al. Exacerbation of autoanti- autoantibodies. Transfusion 1985;25:545–546.
body-mediated hemolytic anemia by viral infection. J Virol 2000;74: 54. Wallhermfechtel MA, Pohl BA, Chaplin H. Alloimmunization in patients
6045–6049. with warm autoantibodies. A retrospective study employing three donor
23. Brown DL, Lachmann PJ, Dacie JV. The in vivo behaviour of comple- alloabsorptions to aid in antibody detection. Transfusion 1984;24:482–485
ment-coated red cells: studies in C6-deficient, C3-depleted and nor- 55. Mollison PL. Measurement of survival and destruction of red cells in
mal rabbits. Clin Exp Immunol 1970;7:401–421. haemolytic syndromes. Br Med Bull 1959;15:59–67.
24. Munn LR, Chaplin H Jr. Rosette formation by sensitized human red 56. Petz L, Garratty G (ed). Acquired Immune Hemolytic Anemias. New
cells—effects of source of peripheral leukocyte monolayers. Vox Sang York, Churchill Livingstone, 1980.
1977;33:129–142. 57. Gould SA, Rosen AL, Sehgal LR, et al. Fluosol-DA as a red-cell substi-
III 25. Gorst DW, Rawlinson VI, Merry AH, et al. Positive direct antiglobulin tute in acute anemia. NEJM 1986;314:1653–1656.
test in normal individuals. Vox Sang 1980;38:99–105. 58. Mullon J, Giacoppe G, Clagett C, et al. Transfusions of polymerized
568 26. Unger LJ. A method for detecting Rho antibodies in extremely low bovine hemoglobin in a patient with severe autoimmune hemolytic
titer. J Lab Clin Med 1951;37:825–837. anemia. NEJM 2000;342: 1638–1643.
27. Issitt PD, et al. Anti-Wrb and other autoantibodies responsible for positive 59. Klein HG. The prospects for red-cell substitutes. NEJM 2000;342:
direct antiglobulin tests in 150 individuals. Br J Haematol 1976;34:5–18. 1666–1668.
28. Smith MA, Shah NS, Lobel JS. Parvovirus B19 infection associated 60. Atkinson JP, Schreiber AD, Frank MM. Effects of corticosteroids and
with reticulocytopenia and chronic autoimmune hemolytic anemia. splenectomy on the immune clearance and destruction of erythro-
Am J Pediatr Hematol Oncol 1989;11:167–169. cytes. J Clin Invest 1973;52:1509–1517
29. Roush GR, Rosenthal NS, Gerson SL, et al. An unusual case of autoim- 61. Fries LF, Brickman CM, Frank MM. Monocyte receptors for the Fc
mune hemolytic anemia with reticulocytopenia, erythroid dysplasia, portion of IgG increase in number in autoimmune hemolytic anemia
and an IgG2 autoanti-U. Transfusion 1996;36:575–580. and other hemolytic states and are decreased by glucocorticoid ther-
30. Conley CL, Lippman SM, Ness P. Autoimmune hemolytic anemia with apy. J Immunol 1983;131:1240–1245.
reticulocytopenia. A medical emergency. JAMA 1980;244:1688–1690 62. Schreiber AD. Clinical immunology of the corticosteroids. Prog Clin
31. Conley CL, Lippman SM, Ness PM, et al. Autoimmune hemolytic anemia Immunol 1977;3:103–114.
with reticulocytopenia and erythroid marrow. NEJM 1982;306:281–286 63. Schreiber AD, Parsons J, McDermott P, et al. Effect of corticosteroids
32. Crosby WH, Rappaport H. Reticulocytopenia in autoimmune hemo- on the human monocyte IgG and complement receptors. J Clin Invest
lytic anemia. Blood 1956;11:929–236. 1975;56:1189–1197.
33. Mangan KF, Besa EC, Shadduck RK, et al. Demonstration of two dis- 64. Collins PW, Newland AC. Treatment modalities of autoimmune blood
tinct antibodies in autoimmune hemolytic anemia with reticulocyto- disorders. Semin Hematol 1992;29:64–74.
penia and red cell aplasia. Exp Hematol 1984;12:788–793. 65. Schwartz SI, Bernard RP, Adams JT, et al. Splenectomy for hematologic
34. Van De Loosdrecht AA, Hendriks DW, Blom NR, et al. Excessive apopto- disorders. Arch Surg 1970;101:338–247.
sis of bone marrow erythroblasts in a patient with autoimmune haemo- 66. Committee on Infectious Disease. American Academy of Pediatrics:
lytic anaemia with reticulocytopenia. Br J Haematol 2000;108:313–315. Red Book: report of the Committee on Infectious Diseases, 24th ed.
35. Frank M. NIH conference. Pathophysiology of immune hemolytic Elk Grove Village, Ill., American Academy of Pediatrics, 1997.
anemia. Ann Intern Med 1977;87:210–222. 67. Parker A, MacPherson AI, Richmond J. Value of radiochromium inves-
36. Pruzanski W, Shumak KH. Biologic activity of cold-reacting autoanti- tigation in autoimmune haemolytic anemia. BMJ 1977;6055:208.
bodies (first of two parts). NEJM 1977;297:538–542. 68. Gopal AK, Press OW. Clinical applications of anti-CD20 antibodies.
37. Crisp D, Pruzanski W. B-cell neoplasms with homogeneous cold- J Lab Clin Med 1999;134:445–450.
reacting antibodies (cold agglutinins). Am J Med 1982;72:915–922. 69. Grillo-Lopez AJ, White CA, Varns C, et al. Overview of the clinical
38. Silberstein LE, Goldman J, Kant JA, et al. Comparative biochemical development of rituximab: first monoclonal antibody approved for
and genetic characterization of clonally related human B-cell lines the treatment of lymphoma. Semin Oncol 1999;26:66–73.
secreting pathogenic anti-Pr2 cold agglutinins. Arch Biochem Bio- 70. Dillman RO. Infusion reactions associated with the therapeutic use of
phys 1988;264:244–252. monoclonal antibodies in the treatment of malignancy. Cancer Metas-
39. Silberstein LE, Robertson GA, Harris AC, et al. Etiologic aspects of tasis Rev 1999;18:465–471.
cold agglutinin disease: evidence for cytogenetically defined clones of 71. Dillman RO. Monoclonal antibodies in the treatment of malignancy:
lymphoid cells and the demonstration that an anti-Pr cold autoanti- Basic concepts and recent developments. Cancer Invest 2001;19:833–841.
AUTOIMMUNE HEMOLYTIC ANEMIAS
72. Waldmann T, Levy R, Coller BS. Emerging therapies: spectrum of 99. Tschirhart DL, Kunkel L, Shulman IA. Immune hemolytic anemia asso-
applications of monoclonal antibody therapy. In Hematology (Am Soc ciated with biclonal cold autoagglutinins. Vox Sang 1990;59:222–226.
Hematol Educ Program). 2000;394. 100. Deutsch HF. Properties and modifications of a cryomacroglobulin
73. Hagberg H, Holmbom E. Risk factors for side effects during first infu- possessing cold agglutinin activity. Biopolymers 1969;7:21–37.
sion of rituximab-definition of a low risk group. Med Oncol 2000; 101. Kuenn JW, Weber R, Teague PO, et al. Cryopathic gangrene with
17:218–221. an IgM lambda cryoprecipitating cold agglutinin. Cancer 1978;42:
74. Berentsen S, Tjonnfjord GE, Brudevold R, et al. Favourable response 1826–1833.
to therapy with the anti-CD20 monoclonal antibody rituximab in pri- 102. Tsai CM, Zopf DA, Yu RK, et al. A Waldenström macroglobulin that is
mary chronic cold agglutinin disease. Br J Haematol 2001;115:79–83. both a cold agglutinin and a cryoglobulin because it binds N-acetyl-
75. Borradori L, Lombardi T, Samson J, et al. Anti-CD20 monoclonal neuraminosyl residues. Proc Natl Acad Sci USA 1977;74:4591–4594.
antibody (rituximab) for refractory erosive stomatitis secondary to 103. Umlas J, Kaufman M, MacQueston C, et al. A cryoglobulin with cold
CD20+ follicular lymphoma-associated paraneoplastic pemphigus. agglutinin and erythroid stem cell suppressant properties. Transfusion
Arch Dermatol 2001;137:269–272. 1991;31:361–364.
76. Heizmann M, Itin P, Wernli M, et al. Successful treatment of paraneo- 104. Ruddy S, Gigli I, Austen KF. The complement system of man. I. NEJM
plastic pemphigus in follicular NHL with rituximab: report of a case 1972;287:489–495.
and review of treatment for paraneoplastic pemphigus in NHL and 105. Borsos T, Rapp HJ. Complement fixation on cell surfaces by 19S and
CLL. Am J Hematol 2001;66:142–144. 7S antibodies. Science 1965;150:505–506.
77. Quartier P, Brethon B, Philippet P, et al. Treatment of childhood 106. Konig AL, Kather H, Roelcke D. Autoimmune hemolytic anemia by
autoimmune haemolytic anaemia with rituximab. Lancet 2001;358: coexisting anti-I and anti-Fl cold agglutinins. Blut 1984;49:363–368.
1511–1513. 107. Sokol RJ, Booker DJ, Stamps R. The pathology of autoimmune haemo-
78. Zecca M, De Stefano P, Nobili B, Locatelli F. Anti-CD20 monoclonal lytic anaemia. J Clin Pathol 1992;45:1047–1052.
antibody for the treatment of severe, immune-mediated, pure red cell 108. Abramson N, Gelfand EW, Jandl JH, et al. The interaction between
aplasia and hemolytic anemia. Blood 2001;97:3995–3997. human monocytes and red cells. Specificity for IgG subclasses and IgG
79. Zecca M, Nobili B, Ramenghi U, et al. Rituximab for the treatment fragments. J Exp Med 1970;132:1207–1215.
of refractory autoimmune hemolytic anemia in children. Blood 109. Atkinson JP, Frank MM. Studies on the in vivo effects of antibody. Inter-
2003;101:3857–3861. action of IgM antibody and complement in the immune clearance and
80. Besa EC. Rapid transient reversal of anemia and long-term effects of destruction of erythrocytes in man. J Clin Invest 1974;54:339–348.
maintenance intravenous immunoglobulin for autoimmune hemo- 110. Ambrus M, Bajtain G. A case of an IgG-type cold agglutinin disease.
lytic anemia in patients with lymphoproliferative disorders. Am J Med Haematologia 1969;3:225.
1988;84:691–698. 111. Angevine CD, Andersen BR, Barnett EV. A cold agglutinin of the IgA
81. Samuelsson A, Towers TL, Ravetch JV. Anti-inflammatory activity of IVIG class. J Immunol 1966;96:578–586.
mediated through the inhibitory Fc receptor. Science 2001; 291:484–486. 112. Dellagi K, Brouet JC, Schenmetzler C, et al. Chronic hemolytic ane-
82. Murphy S, LoBuglio AF. Drug therapy of autoimmune hemolytic ane- mia due to a monoclonal IgG cold agglutinin with anti-Pr specificity.
mia. Semin Hematol 1976;13:323–334. Blood 1981;57:189–191.
83. Fauci AS, Dale DC, Wolff SM. Cyclophosphamide and lymphocyte 113. Garratty G, Petz LD, Brodsky I, et al. An IgA high-titer cold agglutinin
subpopulations in Wegener’s granulomatosis. Arthritis Rheum 1974; with an unusual blood group specificity within the Pr complex. Vox
17:355–361. Sang 1973;25:32–38.
84. Steinberg A. Cytotoxic drugs in treatment of nonmalignant diseases. 114. Moore J, Chaplin H Jr. Autoimmune hemolytic anemia associated with
Ann Intern Med 1972;76:619–642. an IgG cold incomplete antibody. Vox Sang 1973;24:236.
85. Fahey JL. Cancer in the immunosuppressed patient. Ann Intern Med 115. Pruzanski W, Cowan DH, Parr DM. Clinical and immunochemical
1971;75:310–312. studies of IgM cold agglutinins with λ type light chains. Clin Immunol
86. Floersheim GL. A comparative study of the effects of anti-tumour and Immunopathol 1974;2:234–245.
41
immunosuppressive drugs on antibody-forming and erythropoietic 116. Shulman IA, et al. Autoimmune hemolytic anemia with both cold and
cells. Clin Exp Immunol 1970;6:861–870. warm autoantibodies. JAMA 1985;253:1746–1748. 569
87. Miller JJ 3rd, Williams GF, Leissring JC. Multiple late complications of 117. Silberstein LE, Berkman EM, Schreiber AD. Cold hemagglutinin disease asso-
therapy with cyclophosphamide, including ovarian destruction. Am J ciated with IgG cold-reactive antibody. Ann Intern Med 1987;106:238–242.
Med 1971;50:530–535. 118. Tonthat H, et al. A new case of monoclonal IgAκ cold agglutinin with
88. Qureshi M, Pennington JH, Goldsmith HJ, Cox PE. Cyclophophamide. anti-Pr1d specificity in a patient with persistent HB antigen cirrhosis.
Lancet 1972;7790:1290. Vox Sang 1976;30:464–468.
89. Johnson WW, Meadows DC. Urinary-bladder fibrosis and telangi- 119. Andrzejewski C Jr, Gault E, Briggs M, et al. Benefit of a 37°C extracor-
ectasia associated with long-term cyclophosphamide therapy. NEJM poreal circuit in plasma exchange therapy for selected cases with cold
1971;284:290–294. agglutinin disease. J Clin Apher 1988;4:13–17.
90. Emilia G, et al. Long-term salvage treatment by cyclosporin in refrac- 120. Park JV, Weiss CI. Cardiopulmonary bypass and myocardial protec-
tory autoimmune haematological disorders. Br J Haematol 1996; tion: management problems in cardiac surgical patients with cold
93:341–344. autoimmune disease. Anesth Analg 1988;67:75–78.
91. Orlin JB, Berkman EM. Partial plasma exchange using albumin 121. Hippe E, Jensen KB, Olesen H, et al. Chlorambucil treatment of
replacement: removal and recovery of normal plasma constituents. patients with cold agglutinin syndrome. Blood 1970;35:68–72.
Blood 1980;56:1055–1059. 122. Schubothe H. The cold hemagglutinin disease. Semin Hematol 1966;3:
92. Silberstein LE, Berkman EM. Plasma exchange in autoimmune hemo- 27–47.
lytic anemia (AIHA). J Clin Apher 1983;1:238–242. 123. Berentsen S, Ulvestad E, Gjersten BT, et al. Rituximab for primary
93. Kutti J, Wadenvik H, Safai-Kutti S, et al. Successful treatment of refrac- chronic cold agglutinin disease: a prospective study of 37 courses of
tory autoimmune haemolytic anaemia by plasmapheresis. Scand J therapy in 27 patients. Blood 2004;103:2925–2928.
Haematol 1984;32:149–152. 124. Godder K, Pati AR, Abhyankar SH, et al. De novo chronic graft-ver-
94. Rapoport AP, Rowe JM, McMican A. Life-threatening autoimmune sus-host disease presenting as hemolytic anemia following partially
hemolytic anemia in a patient with the acquired immune deficiency mismatched related donor bone marrow transplant. Bone Marrow
syndrome. Transfusion 1988;28:190–191. Transplant 1997;19:813–817.
95. Capra JD, Dowling P, Cook S, et al. An incomplete cold-reactive 125. Gottsche B, Salama A, Mueller-Eckhardt C. Donath-Landsteiner auto-
gamma G antibody with i specificity in infectious mononucleosis. Vox immune hemolytic anemia in children. A study of 22 cases. Vox Sang
Sang 1969;16:10–17. 1990;58:281–286.
96. Rosenfield RE, Schmidt PJ, Calvo RC, et al. Anti-i, a frequent cold 126. Heddle NM. Acute paroxysmal cold hemoglobinuria. Transfus Med
agglutinin in infectious mononucleosis. Vox Sang 1965;10:631–634. Rev 1989;3:219–229.
97. Silberstein, LE, Shoenfeld Y, Schwartz RS, et al. A combination of IgG 127. Nordhagen R, Stensvold K, Winsnes A, et al. Paroxysmal cold haemo-
and IgM autoantibodies in chronic cold agglutinin disease: Immu- globinuria. The most frequent acute autoimmune haemolytic anaemia
nologic studies and response to splenectomy. Vox Sang 1985;48: in children? Acta Paediatr Scand 1984;73:258–262.
105–109. 128. Hinz D, Picken MD, Lepow IH. Studies on immune human hemolysis
98. Szymanski IO, Teno R, Rybak ME. Hemolytic anemia due to a mixture II: the Donath-Landsteiner reaction as a model system for studying
of low-titer IgG lambda and IgM lambda agglutinins reacting opti- the mechanism of action of complement and the role of C1 and C1
mally at 22°C. Vox Sang 1986;51:112–116. esterase. J Exp Med 1961;113:193.
TRANSFUSION MEDICINE 129. Levine P, Celano MJ, Falkowski, F. The specificity of the antibody 154. Costea N, Yakulis VJ, Heller P. Inhibition of cold agglutinins (anti-I) by
in paroxysmal cold hemoglobinuria (PCH). Transfusion 1963;3: M. pneumoniae antigens. Proc Soc Exp Biol Med 1972;139:476–479.
278–280. 155. Pascual V, Victor K, Lelsz D, et al. Nucleotide sequence analysis of
130. Worlledge SM, Rousso C. Studies on the serology of paroxysmal cold the V regions of two IgM cold agglutinins. Evidence that the VH4-21
haemoglobinuria (PGH), with special reference to its relationship with gene segment is responsible for the major cross-reactive idiotype.
the P blood group system. Vox Sang 1965;10:293–298. J Immunol 1991;146:4385–4391.
131. Engelfriet CP, Borne AV, Moes M, et al. Serological studies in autoim- 156. Silberstein LE, Jefferies LC, Goldman J, et al. Variable region gene
mune haemolytic anaemia. Bibl Haematol 1968;29:473–478. analysis of pathologic human autoantibodies to the related i and I red
132. Judd WJ, Wilkinson SL, Issitt PD, et al. Donath-Landsteiner hemo- blood cell antigens. Blood 1991;78:2372–2386.
lytic anemia due to an anti-Pr-like biphasic hemolysin. Transfusion 157. Stevenson FK, Spellerberg MB, Treasure J, et al. Differential usage of
1986;26:423–425. an Ig heavy chain variable region gene by human B-cell tumors. Blood
133. Weiner W, Gordon EG, Rowe D. A Donath-Landsteiner antibody 1993;82:224–230.
(non-syphilitic type). Vox Sang 1964;9:684–697. 158. Perrotta S, Locatelli F, La Manna A, et al. Anti-CD20 monoclonal anti
134. Petz LD, Fudenberg HH. Coombs-positive hemolytic anemia caused body (rituximab) for life-threatening autoimmune haemolytic anae-
by penicillin administration. NEJM 1966;274:171–178. mia in a patient with systemic lupus erythematosus. Br J Haematol
135. Bougie D, Johnson ST, Weitekamp LA, et al. Sensitivity to a metabolite 2002;116:465–467.
of diclofenac as a cause of acute immune hemolytic anemia. Blood 159. Motto DG, Williams JA, Boxer LA. Rituximab for refractory childhood
1997;90:407–413. autoimmune hemolytic anemia. Isr Med Assoc J 2002;4:1006–1008.
136. Salama A, Mueller-Eckhardt C. On the mechanisms of sensitization 160. McMahon C, Babu L, Hodgson A, et al. Childhood refractory auto-
and attachment of antibodies to RBC in drug-induced immune hemo- immune haemolytic anaemia: is there a role for anti-CD20 therapy
lytic anemia. Blood 1987;69:1006–1010. (rituximab)? Br J Haematol 2002;117:480–483.
137. Carstairs K, Breckenridge A, Dollery CT, Worlledge SM. Incidence of 161. Zaja F, Iacona I, Masolini P, et al. B-cell depletion with rituximab as
a positive direct Coombs test in patients on alpha-methyldopa. Lancet treatment for immune hemolytic anemia and chronic thrombocyto-
1966;7455:133. penia. Haematologica 2002;87:189–195.
138. Petz LD. Drug-induced immune haemolytic anaemia. Clin Haematol 162. Corti P, Bonanomi S, Vallinoto C, et al. Rituximab for immune hemo-
1980;9:455–482. lytic anemia following T- and B-cell-depleted hematopoietic stem cell
139. Worlledge SM. Immune drug-induced hemolytic anemias. Semin transplantation. Acta Haematol 2003;109:43–45.
Hematol 1973;10:327–344. 163. Hongeng S, Tardtong P, Worapongpaiboon S, et al. Successful treat-
140. Danielson DA, Douglas SW III, Herzog P, et al. Drug-induced blood ment of refractory autoimmune haemolytic anaemia in a post-
disorders. JAMA 1984;252: 3257–3260. unrelated bone marrow transplant paediatric patient with rituximab.
141. Theofilopoulos AN, Dixon FJ. Murine models of systemic lupus ery- Bone Marrow Transplant 2002;29:871–872.
thematosus. Adv Immunol 1985;37:269–390. 164. Ship A, May W, Lucas K. Anti-CD20 monoclonal antibody therapy
142. Okamoto M, Murakami M, Shimizu A, et al. A transgenic model of for autoimmune hemolytic anemia following T cell-depleted, haplo-
autoimmune hemolytic anemia. J Exp Med 1992;175:71–79. identical stem cell transplantation. Bone Marrow Transplant 2002;29:
143. Sakiyama T, Ikuta K, Nisitani S, et al. Requirement of IL-5 for induc- 365–366.
tion of autoimmune hemolytic anemia in anti-red blood cell autoanti- 165. Lee EJ, Kueck B. Rituxan in the treatment of cold agglutinin disease.
body transgenic mice. Int Immunol 1999;11:995–1000. Blood 1998;92:3490–3491.
144. Nisitani S, Honjo T. Breakage of B cell tolerance and autoantibody 166. Layios N, Van Den NE, Jost E, et al. Remission of severe cold agglutinin
production in anti-erythrocyte transgenic mice. Int Rev Immunol disease after rituximab therapy. Leukemia 2001;15:187–188.
1999;18:259–270. 167. Zaja F, Russo D, Fuga G, et al. Rituximab in a case of cold agglutinin
145. Nisitani S, Sakiyama T, Honjo T. Involvement of IL-10 in induction of disease. Br J Haematol 2001;115:232–233.
III autoimmune hemolytic anemia in anti-erythrocyte Ig transgenic mice. 168. Bauduer F. Rituximab: A very efficient therapy in cold agglutinins
Int Immunol 1998;10:1039–1047. and refractory autoimmune haemolytic anaemia associated with
570 146. Leddy JP, Falany JL, Kissel GE, et al. Erythrocyte membrane proteins CD20-positive, low-grade non-Hodgkin’s lymphoma. Br J Haematol
reactive with human (warm-reacting) anti-red cell autoantibodies. J 2001;112:1085–1086.
Clin Invest 1993;91: 1672–1680. 169. Chemnitz J, Draube A, Diehl V, Wolf J. Successful treatment of steroid
147. Bloy C, Blanchard D, Lambin P, et al. Human monoclonal antibody and cyclophosphamide-resistant hemolysis in chronic lymphocytic
against Rh(D) antigen: partial characterization of the Rh(D) polypep- leukemia with rituximab. Am J Hematol 2002;69:232–233.
tide from human erythrocytes. Blood 1987;69:1491–1497. 170. Pulik M, Genet P, Lionnet F, Touahri T. Treatment of primary chronic
148. Lane HC, Masur H, Edgar LC, et al. Abnormalities of B-cell activation cold agglutinin disease with rituximab: maintenance therapy may
and immunoregulation in patients with the acquired immunodefi- improve the results. Br J Haematol 2002;117:998–999.
ciency syndrome. NEJM 1983;309:453–458. 171. Engelhardt M, Jakob A, Ruter B, et al. Severe cold hemagglutinin disease
149. Kipps TJ, Carson DA. Autoantibodies in chronic lymphocytic leukemia (CHD) successfully treated with rituximab. Blood 2002;100:1922–1923.
and related systemic autoimmune diseases. Blood 1993;81:2475–2487. 172. Gharib M, Poynton C. Complete, long-term remission of refractory
150. Leddy JP, Bakemeier RF. Structural aspects of human erythrocyte idiopathic cold haemagglutinin disease after Mabthera. Br J Haematol
autoantibodies. I. L chain types and electrophoretic dispersion. J Exp 2002;117:248–249.
Med 1965;121:1–17. 173. Berentsen S, Tjonnfjord GE, Brudevold R, et al. Favourable response
151. Masouredis SP, Branks MJ, Victoria EJ. Antiidiotypic IgG crossreactive to therapy with the anti-CD20 monoclonal antibody rituximab
with Rh alloantibodies in red cell autoimmunity. Blood 1987;70: 710–715. in primary chronic cold agglutinin disease. Br J Haematol 2001;
152. Hakomori S. Blood group ABH and Ii antigens of human erythrocytes: 115:79–83.
Chemistry, polymorphism, and their developmental change. Semin 174. Gupta N, Kavuru S, Patel D, et al. Rituximab-based chemotherapy for
Hematol 1981;18:39–62. steroid-refractory autoimmune hemolytic anemia of chronic lympho-
153. Stevenson FK, Smith GJ, North J, et al. Identification of normal B-cell cytic leukemia. Leukemia 2002;16:2092–2095.
counterparts of neoplastic cells which secrete cold agglutinins of anti-I
and anti-i specificity. Br J Haematol 1989;72:9–15.
Chapter 42
Transfusion in Economically Restricted
and Developing Countries
Audrey N. Schuetz ● Kenneth A. Clark
Until the human immunodeficiency virus (HIV) pandemic In the third place, proper laboratory testing of donated
in the 1980s, transfusion services in many economically blood and adequate record keeping present challenges to
restricted countries remained poorly developed, primarily blood centers with financial and logistical restrictions.
due to economic constraints. With the institution of HIV Testing and storage of blood is limited by the lack of reliable
testing in 1985 in the United States and in other developed sources of electricity and refrigeration, as well as by the cost
countries, economically restricted countries were encour- of machinery, reagents, and testing kits.
aged to place a stronger focus on transfusion safety practices. Finally, transfusion safety relies on appropriate blood
Although transfusion safety has improved somewhat over administration to recipients. Given the difficulties in donor
the past few decades, many economically restricted countries screening and laboratory testing of blood, consideration
continue to struggle with inadequate resources and infra- must be given to balancing the risk of exposure to a trans-
structure that hinder establishment of a safer blood supply. fusion-transmitted infection versus the risk of no transfu-
According to the World Health Organization (WHO), only sion. Furthermore, because children comprise the largest
20% of the worldwide supply of safe and screened blood is proportion of transfusion recipients in certain regions and
available to people living in economically restricted coun- pregnancy- or malaria-associated anemia also rank among
tries, where approximately 80% of the world’s population the highest indications for transfusion in many economically
resides.1,2 restricted countries, the morbidity and mortality associated
Enforcement of a successful transfusion service depends with a transfusion-transmitted infection may be potentially
on several levels of commitment, from the government high among certain populations. 42
administration to the staff delivering the blood. Optimally, Although two decades have passed since relatively sim-
blood transfusion services consist of education, recruitment ple HIV tests became available to screen blood donors and 571
and adequate selection of safe blood donors; collection, pro- products, the lack of economic and political resources has
cessing, and storage of blood products; performance of sero- compromised the blood supply in many areas of the world.
logic and other tests on the products before transfusion at The overwhelming challenges are evident when faced with
the collection facility; and, finally, transportation and subse- the fact that approximately 10% of all acquired immuno-
quent release of the blood to patients with appropriate need deficiency syndrome (AIDS) cases in certain areas of the
for transfusion. world are transfusion associated.5,6 However, improvements
Commitment of governments to a well-developed and have already begun; Zimbabwe was the third country in
organized blood transfusion service is the first key to sus- the world to routinely test donated blood for HIV.7 Various
tainable and successful blood transfusion systems. Many organizations are currently gathering data on methods to
economically restricted countries are faced with a high dis- improve governmental infrastructure, to understand disease
ease burden but lack the infrastructure to address the prob- epidemiology, and to adjust current transfusion practices
lem. Various organizational structures of blood services are and attitudes toward blood safety in economically restricted
described in this chapter, as is the commitment of the WHO countries.
to worldwide blood safety.
A second key to optimizing blood transfusion services
includes selection of safe blood donors by recruitment of GOVERNMENTAL ORGANIZATION
nonremunerated (nonpaid or voluntary) individuals. In the OF BLOOD SERVICES
developed world, donor screening focuses on excluding indi-
viduals who carry a higher “risk” of infectious disease than Development and organization of blood transfusion services
the general population; such individuals include intravenous in economically restricted countries has not only lagged
drug users, men who have sex with men, and commercial sex behind blood services in developed countries but has also
workers. In economically restricted countries, such as Africa, lagged behind other parts of the health care system as well.8
where heterosexual transmission has accounted for as much The sustainability of high-quality transfusion systems in eco-
as 80% of all HIV infections, such focused donor screening is nomically restricted countries is best achieved with national
less effective.3,4 A strategy employed by economically restricted oversight and a national commitment to safe blood transfu-
countries to select safe blood donors involves predonation sion as part of the health care system. The WHO claims that
laboratory testing of blood with rapid diagnostic tests. a well-organized blood transfusion service is a prerequisite
TRANSFUSION MEDICINE collect and process the blood in economically restricted
Table 42–1 Key Organizational Elements
countries, typically including the ministries of health, social
of a National Blood Transfusion Service
security systems, the armed forces, private organizations, and
Formalization of government commitment and support nongovernmental organizations. Blood transfusion services,
Establishment of a national blood policy or plan which oversee the delivery of blood products to patients, are
Establishment of legal authority through legal or regulatory usually operated by health care facilities, such as hospitals.19
framework The WHO recommends that governments of economi-
Establishment of a responsible organization to oversee
and implement the national blood transfusion service
cally restricted countries take responsibility to ensure a safe
Appointment of a national blood service medical director, and adequate blood supply. Indicating the need for national
a quality manager, and appropriate advisory groups support of blood transfusion programs, the 58th World
Development of financial and budgetary systems to ensure Health Assembly (WHA) passed resolution WHA58.13 in
sustainability of the national transfusion service May 2005, urging member states to support full implementa-
Establishment of a national quality system9
tion of well-organized, nationally coordinated, and sustain-
able blood programs with appropriate regulatory systems.24
In particular, the resolution advocated government commit-
for the safe and effective use of blood products while empha- ment and support for national blood programs with quality
sizing government commitment as a strategic priority.9,10 management systems, by means of a legal framework and a
The WHO also recommends that governments establish national blood safety policy and plan. Many economically
blood transfusion services as a separate entity with an ade- restricted countries have made significant organizational
quate budget, effective management, and trained staff.9,11 advances in the past decade, but improvements are still
The key organizational elements of a national blood trans- required.
fusion service in an economically restricted country, as sug-
gested by the WHO, are listed in Table 42–1. Organization of
blood services varies considerably throughout the world with BLOOD DONOR RECRUITMENT,
regard to geographic regions and availability of resources. SELECTION, AND SCREENING
The organizational structure of blood services is outlined in
Table 42–2. In 1975, the WHA adopted Resolution WHA28.72, which
Several types of organizational blood banking structures required its members to promote the development of
operate worldwide. Hospital-based systems predominate in national transfusion services based on the use of voluntary,
economically restricted countries; centralized or regional- nonpaid blood donors.2,25 Replacement donors are recruited
ized blood transfusion systems operate largely in developed from among the family or relatives of hospitalized patients
countries.12,14,15 In sub-Saharan Africa, where blood services who require blood transfusions. A number of studies have
are predominantly hospital-based, blood transfusion sys- shown that voluntary donors often have a lower prevalence
III tems are moving toward a more centralized or regionalized and incidence of infectious transfusion-transmitted diseases
approach in countries such as South Africa, Uganda, Kenya, than do family/replacement donors and paid donors.17,26–33
572 Zimbabwe, and Togo.16 In addition, some northern African In Kenya, a study comparing volunteer donors from mobile
countries bordering the Mediterranean Sea have made con- blood units with donors recruited by family members
siderable progress toward centralization. Centralized blood showed that HIV seroprevalence among voluntary donors
collection systems are more advantageous than small hospital- was significantly lower than among family-recruited donors
based systems, because they have a greater number of highly (1.7% and 9.1%, respectively; odds ratio, 5.9).34 Information
trained specialists with the ability to handle a larger number collected in the Global Database on Blood Safety has repeat-
of blood units.17 In contrast, blood programs in economi- edly shown the importance of recruiting voluntary blood
cally restricted areas of South and Central America are largely donors.35
decentralized, some to a high degree. For example, Argentina Blood centers in sub-Saharan Africa routinely rely on fam-
has more than 1000 blood banks.18–20 A similar situation exists ily/replacement or paid donors for a significant percentage of
in economically restricted areas of Southeast Asia, where there their blood collection needs, often due to shortages in supply
is extensive decentralization of the blood system.18 of donor blood in the blood banks. A lack of resources to
Blood services in economically restricted countries, train personnel and to provide transport for mobile blood
just as in more developed countries, should be defined by campaigns contributes to blood shortages as well. From
a legal framework that states the national blood policy and 50% to 70% of donations are collected from family/replace-
describes the processes governing the collection, process- ment and paid donors worldwide, often in countries with
ing, and transfusion of blood.21–23 A variety of organizations relatively high prevalence of HIV, hepatitis B virus (HBV),
Centralized One national blood center operates the blood services for the country, with or without regional
blood centers.
Regionalized The country is divided into regions of varying degrees of autonomy, with different mechanisms
to achieve national coordination.
Hospital-based Each hospital operates its own blood collection and processing system, with or without national
control.
Combination of above There may be a national blood transfusion system, but hospitals may find the national coverage
unsatisfactory and choose to operate their own systems.12,13
and hepatitis C virus (HCV) infection.25,36,37 A blood bank
42
Table 42–3 Percentage Prevalence of Infectious Diseases among Blood Donors in Selected Countries
Location HIV HCV HBV Syphilis HTLV T. cruzi Reference No. 575
†
Voluntary donors.
* Majority replacement donors.
‡
100% paid or replacement donors.
§ JB Gorlin, personal communication, November 17, 2005.
TRANSFUSION MEDICINE In contrast to developed countries, economically restricted Finally, although hundreds of cases of human infection
countries continue to rely on syphilis testing as a means of with Babesia spp. have been reported in the United States,
disease screening. Both the reagin tests, such as the Venereal only a few cases have been described in China, Egypt, Mexico,
Disease Research Laboratory (VDRL) test, and the trepone- South Africa, and Taiwan, currently indicating a low need for
mal tests, such as the Treponema pallidum hemagglutination Babesia screening in economically restricted countries.113
assay (TPHA-FTA ABS), detect syphilis antibodies. Some Other potential laboratory surrogate markers for infectious
countries with a low prevalence of syphilis use TPHA-FTA disease risk have been examined in economically restricted
ABS, because its sensitivity and specificity are higher than countries, such as serum alanine aminotransferase (ALT) lev-
those of the reagin tests.103,104 Yet, the TPHA-FTA ABS costs els, yet they have met with limited success. Studies of serum
more and is more complex than the reagin tests. In countries ALT in Indian blood donors have shown no correlation with
with high syphilis prevalence, the VDRL test may be more viral seropositivity of HBV, HCV, HIV, or CMV.114
appropriate in diagnosing the infectious phase of syphilis.
In Thailand, at a clinic for laborers, HIV-1 seropositivity Transfusion-Associated Infectious Risks
correlated positively with positive TPHA-FTA ABS results,
with an odds ratio of 1.8 (p = 0.015).105 Such findings sup- Transfusion-transmitted infections are among the largest risks
port the continued use of syphilis screening as both a marker of receiving a blood transfusion in economically restricted
for other diseases and as a means of disease diagnosis in countries. Among the infectious diseases, HIV has one of the
economically restricted countries. most serious adverse consequences. In contrast to data from
Antibody testing for HTLV-I, the etiologic agent of adult the United States, where significantly less than 1% of HIV
T-cell leukemia, is generally not performed in economically infections are estimated to be transfusion transmitted, data
restricted countries. HTLV-I is endemic in southwest Japan, from economically restricted countries show that 5% to 10%
the Caribbean, Central and South America, sub-Saharan of new and existing HIV infections are due to blood transfu-
Africa, Papua New Guinea, Australia, the Solomon Islands, sion.31,115–117
and western Asia but is not a required test in many of these Although estimates on transfusion-transmitted HIV
regions.106,107 proved difficult to gather during the early stages of the
Screening for Chagas disease in endemic areas of Latin AIDS epidemic, as many as 25% of HIV-infected children
America is performed using EIA methods to detect T. cruzi and women in Africa are believed to have acquired HIV
antibodies. The main serologic methods for identification of from blood transfusions.118–120 McFarland and colleagues
T. cruzi antibodies are the indirect hemagglutination assay, the estimate that in sub-Saharan Africa alone, 25% of pediatric
indirect immunofluorescence assay, and the enzyme-linked AIDS cases, 20% of adult female AIDS cases, and 10% of
immunosorbent assay (ELISA). The indirect hemagglutina- all other AIDS cases may be associated with transfusion.121
tion assay, although easy to perform and widely used, is the Blood transfusions in sub-Saharan Africa rank third among
least sensitive of the three tests.108,109 Blood centers in Brazil modes of HIV transmission, most notably affecting children
III most often use the ELISA, according to a series of external under age 5.122,123 In the Caribbean, 1.1% of AIDS cases are
quality control assessments of blood bank laboratories per- due to blood transfusion.124 In Brazilian blood donors,
576 formed in 1999 and 2000.108 The use of two different screen- a history of blood transfusion is thought to be the etiology
ing methods, most often the combination of an ELISA test in 15% of HCV-positive cases.125 Transfusion before 1992 is
and an indirect hemagglutination test, minimizes the risk of considered to be the major route of transmission for HCV
false negatives.108,110 The indirect immunofluorescence assay in Japan.126
is less commonly used. The risk of receiving HIV-positive blood in Africa ranges
Although recommended in malaria-endemic areas, malaria from less than 1% to more than 20% per unit of blood,
screening in very high prevalence areas, such as sub-Saharan depending on the country studied.6 In a 2001 survey of
Africa, is neither practical nor effective, because the majority 1482 blood units in six government hospitals in Kenya, 31
of donors and recipients would test positive. In areas of low (~2%) units tested HIV-positive on follow-up at the refer-
malaria prevalence, such as Southeast Asia, malaria screen- ence laboratory, of which 14 had tested negative by the hos-
ing with Geimsa-stained blood films is common but not pital and were not removed from the blood supply.34 HIV
universal. test results for 1290 donor–recipient pairs showed that 26 of
Cytomegalovirus (CMV) antibody testing also is not rou- the 31 HIV-positive donations had been given to individuals
tinely performed in the economically restricted world, and who were HIV-negative before transfusion. As a result, the
very few studies are available on its use. In a recent study estimated HIV transmission rate was 2% in these hospitals.
in India, none of the 200 voluntary blood donors in Delhi In 1985 in Kinshasa, Democratic Republic of Congo, 561
tested positive for CMV immunoglobulin M (IgM) antibody, HIV-positive blood units were transfused to children in one
but 95% were positive for CMV IgG antibody.111 Testing of pediatric hospital ward.127
donors would be superfluous, because very few seronegative Studies continue to show unacceptably high rates of HIV
blood units would be available for transfusion. transmission from blood transfusion. In a pediatric tertiary
Very few studies on human herpes virus-8 (HHV-8) care center in India, 33 of the 285 HIV-positive children were
have been performed in economically restricted countries. infected through blood transfusion from 1994 to 2001.128
HHV-8 antibodies among 306 blood donors in South Africa A study in rural China in a region with a large percentage
were present in 2% to 10% of individuals, with increas- of paid donors showed that 1% of the 115 HIV-positive
ing prevalence with age.112 Because the seropositivity for residents obtained infection through blood transfusion.129
HHV-8 was relatively high in these blood donors, screen- Finally, in a consecutive survey of 222 women at prenatal
ing for this herpesvirus may be useful, given the association and pediatric clinics who had received transfusions between
between HHV-8 and Kaposi sarcoma. 1980 and 1985 in Kigali, Rwanda, HIV seroprevalence was
TRANSFUSION IN ECONOMICALLY RESTRICTED AND DEVELOPING COUNTRIES
45%, compared to 28% in nontransfused women presenting [a]lthough Tanzania’s health policy is that all blood should be
to the clinics.130 screened anonymously for infections which can be transmit-
The probability of seroconversion from transfusion with a ted on transfusion, health experts say that a lack of systematic
single HIV-positive unit of blood is 90%.131 The risk of HIV and comprehensive blood-screening support programme has
undermined this goal. There is no screening for other infec-
from a screened unit of blood in Ivory Coast in 1992 was two
tions, because the health ministry does not routinely supply
to three times greater than the risk of acquiring HIV from test kits to hospitals due to financial constraints.45
a seropositive needlestick exposure.132,133 Approximately 142
to 276 units of potentially infected blood units were trans- Small blood banks may exist as unregulated commercial
fused annually in the early 1990s in the Ivory Coast, despite operations without quality control.141
routine testing. Innovative answers to logistical problems in economically
Paid plasma donors are becoming an increasingly recog- restricted countries are needed. One such example is the
nized reservoir of infection in the blood bank community. development of the slide Polybrene method for pretransfu-
Paid donors typically become infected at plasmapheresis sion compatibility testing, in response to the lack of centri-
centers through unsafe practices, such as the use of non- fuges in some areas.142 Simple to use, the slide Polybrene can
sterile needles, sharing of intravenous lines, and injection of be performed by personnel with minimal training.
donors with blood to hyperimmunize them for anti-Rho and During an attempt to simulate suboptimal storage con-
serum typing production.134,135 Unlicensed plasmapheresis ditions in economically restricted countries, researchers
centers began appearing in the 1990s in China, where poor tested blood units for HIV with inappropriately stored or
peasants in communities with few risk factors for HIV began expired rapid antibody assays in Zambia.143 The sensitivity
selling their plasma at unsanitary centers and became subse- and specificity of the tests were 11% to 18% lower than the
quently infected.90,136 Some of these centers directly contrib- manufacturer’s claim, with a risk of contracting HIV through
uted to spread of disease in donors through the pooling of transfusion at least six times higher than expected.
blood from donors of the same ABO type before returning Blood screening for infectious diseases proves cost effec-
the red cells to each donor. High rates of HIV seropositivity tive, especially in areas with large numbers of HIV-positive
have been discovered among repeat blood donors, especially individuals.27,61,144 Supplies and equipment comprise the
paid donors.137,138 At one illegal plasma collection site dis- majority of cost in blood transfusion.145 To prevent one
covered in central China between 1998 and 1999, 71 of the case of HIV in Africa, measures to improve blood transfu-
96 paid donors were positive for HIV.139 At another site, 50% sion safety may cost from $US 20 to $US 1000.146 One study
and 100% of all donors tested positive for syphilis and HCV, reported that savings exceeded costs by a factor of 2.7 to 3.5
respectively, and the entire group of 142 tested samples were in a region of Zambia with an HIV seroprevalence among
positive for HIV.90 donors of 16%.87 A Tanzanian study showed that safe trans-
Multifaceted approaches to curb the use of unscreened fusion practices can be assured at an annual cost of $US 0.07
or unsafe blood are needed. In response to the high use of per capita.147 One method of recovering the cost of screening
unscreened blood for transfusion in Nigeria, the government blood involves charging all inpatients a fee that covers the 42
instituted a law to regulate blood transfusions by requiring screening costs on admission.13 Charging all patients rather
blood banks to register their blood donors, with fines and than only those patients who received blood by transfusion 577
imprisonment for offenders.140 is easier to institute and does not lay the cost burden on the
Hemovigilance data systems throughout the world are women and children who require the majority of transfusions.
incomplete, particularly in economically restricted coun- Poor funding and lack of government support also lead
tries. More data on infectious disease prevalence are needed to suboptimal practices. In a survey of 62 hospitals in the
to assess effectiveness of screening procedures and guide fur- Democratic Republic of Congo in 1991, only 29% regu-
ther laboratory testing. larly received HIV testing kits due to lack of funding.148 As
a result, 28% of the blood units transfused during the study
period were not screened for HIV. Logistic and economic
Logistics and Finances of Laboratory issues should be better understood to assess the feasibility of
Screening laboratory testing in economically restricted countries.
Although basic compatibility testing is performed in eco-
nomically restricted countries, many blood banks run only
ABO grouping and Rh typing in lieu of crossmatching. More BLOOD USAGE
extensive testing, such as additional alloantibody or auto-
antibody identification, extends beyond the current capa- Clinicians must weigh several factors when administering
bilities of blood banks in economically restricted countries blood products, including the recipient’s underlying need for
due to limited technical expertise and a lack of automation, transfusion, infectious contamination risks, and other nonin-
reagents, and adequate training. fectious risks of transfusion. Despite the considerations that
Logistical problems often challenge laboratories in eco- should be made in deciding to transfuse, blood transfusions
nomically restricted countries. Laboratories may lack a stan- continue to be performed unnecessarily in some instances.
dardized reporting system for results or have poor record
keeping. Educational systems for blood bank personnel are
Transfusion Recipient Populations
also lacking. The cold chain for laboratory support may be
compromised, with lack of a standard refrigerator, often Transfusion recipient populations in economically restricted
with inadequate storage space for blood components. One countries differ from those in more industrialized countries.
researcher emphasizes the logistical problems by pointing Blood recipients in developed countries generally include
out that: cancer or transplant patients, adults undergoing surgery, and
TRANSFUSION MEDICINE hemophiliacs. In economically restricted countries, recipi- Finally, thalassemia patients comprise another major
ents typically include children, pregnant women, sickle cell population group that receives a large proportion of
disease patients, victims of trauma, and persons with hemo- blood transfusions in economically restricted countries.
globinopathies, severe parasitic infections, and nutritional Thalassemia is an important public health problem in
anemia. Southeast Asia, Africa, India, and the Middle East. Adequate
In many areas of the world, children comprise a large data on the transfusion needs of people with thalassemia are
percentage of those receiving transfusions. Between 19% not generally available from economically restricted coun-
and 67% of hospitalized children in Africa receive transfu- tries. A small study of 330 multiply transfused patients with
sions.67,118,119,127,149,150 In Uganda, 65% of all transfusions are thalassemia in rural Bengal, India, diagnosed 3 (approxi-
performed on children under age 5, the majority of whom mately 1%) HIV-positive patients who had received over 10
suffer from malaria and other infections, malnutrition, and transfusions each.159
sickle cell disease.151 Often, the youngest children require the
majority of transfusions, due in part to the length of time Appropriate Uses of Blood Transfusion
required for children to develop immunity to certain patho-
gens, specifically malaria. A study conducted among six gov- In deciding whether to transfuse, the clinician weighs the risks
ernment hospitals in Kenya found that 58% of transfusions of obtaining potentially infected blood against the risk of the
were requested for children age 10 or younger, and 37% for patient’s not receiving blood. Often, blood is requested emer-
children under age 2.34 Pediatric transfusions were primarily gently and may not be properly screened. As shown in Table
administered for malaria (49%), but were also administered 42–4, a high number of blood transfusions are deemed inap-
for surgery (14%) and chronic anemia (11%). Finally, a retro- propriate in many countries. The authors of one Tanzanian
spective study of blood usage in Mozambique showed that the study found no difference in mortality between transfused
pediatric department ordered nearly half of the blood used and nontransfused anemic children, unless the child showed
for transfusion, primarily for management of anemia.152 signs of cardiac failure, acute hemorrhage, or pneumonia.163
A second population group that receives a large propor- However, the subjects were only followed for 8 weeks in this
tion of transfusions in economically restricted countries is study, which did not take into account morbidity associated
pregnant women. In sub-Saharan Africa, pregnant women with transfusion-transmitted infections. In another deci-
often suffer from anemia due to malaria, iron and folate sion analysis, African children with severe malarial anemia
deficiency, or sickle cell trait or disease, which may be com- showed no transfusion benefit when the chance of survival
pounded by malnutrition or other health problems such as without transfusion was greater than 95%.164
hookworm infestation.153 Tanzanian studies show that preg- Researchers have attempted to quantify the risks and ben-
nant women receive 8% to 11% of all transfusions.154,155 efits of transfusions to certain patient populations in order
Patients with sickle cell disease are another large popu- to provide blood banks with precise guidelines of whom to
lation sector requiring multiple transfusions. Most of the transfuse.165–171 Implementation of such guidelines has proven
III world’s population with sickle cell disease resides in Africa, practical as well as effective. For instance, an observational
with the remainder in the Middle East or India.156,157 Other study showed that Kenyan children benefited from transfu-
578 areas of the world have reported significant numbers of sion under the following circumstances: when the hemoglo-
sickle cell disease as well. Since implementation of neonatal bin level is less than 5.0 g/dL and when they are transfused
sickle cell screening in 2000, one case of sickle cell disease within 48 hours of admission.118,172
per 1196 births in Brazil has been diagnosed.158 Insufficient The effect of implementing the above consensus guide-
transfusion facilities unfortunately hinder the application of lines on appropriate use of blood transfusions in Tanzania
long-term transfusion protocols for these patients. was examined, and it was found that the proportion of avoidable
42
583
B. Complications of Transfusion
i. Infectious Complications
Chapter 43
Hepatitis A, B, and Non-A, Non-B,
Non-C Viruses
Roger Y. Dodd
0 4 8 12 16 20 24 28 32 36 52 100
Weeks after exposure
Figure 43–1 Typical serologic course of acute hepatitis B virus infec-
tion with recovery. Serologic markers of hepatitis B virus (HBV) infection
vary, depending on whether the infection is acute or chronic. The first IgM anti-HBc
serologic marker to appear after acute infection is hepatitis B surface
antigen (HBsAg), which can be detected as early as 1 or 2 weeks and
as late as 11 or 12 weeks (mode, 30–60 days) after exposure to HBV.
In persons who recover, HBsAg is no longer detectable in serum after
0 4 8 12 16 20 24 28 32 36 52 Years
an average of about 3 months. Hepatitis B e antigen (HBeAg) is gener-
ally detectable in patients with acute infection; the presence of HBeAg Weeks after exposure
in serum correlates with higher titers of HBV and greater infectivity. Figure 43–2 Typical serologic course of progression to chronic hepa-
A diagnosis of acute HBV infection can be made on the basis of the titis B virus infection. In patients with chronic hepatitis B virus (HBV)
detection of immunoglobulin M (IgM) class antibody to hepatitis B infection, both hepatitis B surface antigen (HBsAg) and immunoglobu-
core antigen (IgM anti-HBc) in serum; IgM anti-HBc is generally detect- lin G (IgG) antibody to hepatitis B core antigen (anti-HBc) remain per-
able at the time of clinical onset and declines to subdetectable levels sistently detectable, generally for life. Hepatitis B e antigen (HBeAg) is
with 6 months. IgG anti-HBc persists indefinitely as a marker of past variably present in these patients. The presence of HBsAg for 6 months
infection. Anti-HBs becomes detectable during convalescence after the or longer generally indicates chronic infection. In addition, a negative
disappearance of HBsAg in patients who do not experience chronic test result for IgM anti-HBc together with a positive test for HBsAg in a
infection. The presence of anti-HBs after acute infection generally indi- single serum specimen usually indicates that an individual has chronic
cates recovery and immunity from re-infection. (Courtesy of Centers for HBV infection. (Courtesy of Centers for Disease Control and Prevention;
Disease Control and Prevention; from a publicly available slide set.) from a publicly available slide set.)
TRANSFUSION MEDICINE world. As of late 2005, one test had been licensed for use in
Table 43–1 Interpretation of Results of
the United States, and clinical trials have been completed for
Hepatitis B Test Panel
another, in a triplex format that includes HIV and HCV test-
Tests Results Interpretation ing.34 However, it is clear that the current minipool approach
will not have any significant benefits, because NAT has essen-
HBsAg Negative Susceptible tially the same sensitivity as the newest of the serologic tests
Anti-HBc Negative for HBsAg.35 Nevertheless, clinical trials of one of the U.S.
Anti-HBs Negative
tests did generate a measurable yield of about 1:250,000 HBV
HBsAg Negative Immune because of natural DNA-positive, seronegative donations, even in a pooled test-
Anti-HBc Positive infection
Anti-HBs Positive ing format. Such NAT for HBV DNA has been implemented
HBsAg Negative Immune because of hepatitis B
in some European locations (where anti-HBc testing is not
Anti-HBc Negative vaccination performed) and by some manufacturers of pooled plasma
Anti-HBs Positive products.36 The potential added value of single-unit (i.e.,
HBsAg Positive Acutely infected nonpooled) NAT for HBV DNA has not been well-defined,
Anti-HBc Positive although it will certainly prevent cases of post-transfusion
lgM anti-HBc Positive HBV infection. Perhaps the greatest value of HBV NAT will
Anti-HBs Negative
be in regions where testing for anti-HBc is not used, because
HBsAg Positive Chronically infected it could detect those samples that are positive for anti-HBc
Anti-HBc Positive
lgM anti-HBc Negative and viremic for HBV. However, such samples generally have
Anti-HBs Negative a low level of DNA, and a sensitive test will be needed.
HBsAg Negative Four interpretations possible*
Anti-HBc Positive
Anti-HBs Negative
Mutants and Vaccines
*
A number of mutants of HBV have been recognized, and
The patient (1) may be recovering from acute HBV infection;
(2) may be distantly immune (the test is not sensitive enough to
there is evidence that the representation of such mutants is
detect very low levels of anti-HBs in serum); (3) may be susceptible increasing. Perhaps of greatest concern for transfusion medi-
with a false-positive anti-HBc test result; or (4) may have an cine are those mutants that do not express HBsAg.37–39 This
undetectable level of HBsAg in serum, so is actually a carrier. group includes those described as escape mutants—a term
Anti-HBc, antibody to hepatitis B core antigen; anti-HBs, signifying an ability to evade the protective effects of HBV
antibody to hepatitis B surface antigen; HBsAg, hepatitis B surface
antigen; Ig, immunoglobulin. Courtesy of Centers for Disease vaccines (which are currently based on HBsAg). However,
Control and Prevention. infection with these mutants still provokes the formation of
detectable levels of anti-HBc, so that most such infections
would be detected, at least in the United States and other
III countries where anti-HBc testing of donations is routine.
Estimates of the residual risk of HBV transmission are Currently available hepatitis B vaccines prepared from
588 surprisingly high; in 1996, Schreiber and colleagues31 sug- recombinant antigens are safe and effective and are rec-
gested a figure of 1:63,000 donations. Some researchers ommended for universal use in infants. It is to be expected
express concern that this figure may be inappropriately high, that consistent use of these vaccines will eventually lead to
on the basis of questions about the underlying assumptions significant reduction in HBV infection rates and thus to a
and the absence of observed post-transfusion hepatitis B. decline in the prevalence of chronic infection. Current vac-
In fact, in 2001, even with the same assumptions made by cines provoke only anti-HBs and so do not interfere with
Schreiber,31 the risk of HBV transmission was estimated to routine blood donation testing. However, it should be noted
have at least halved as a result of the decreased incidence of that HBsAg itself may be detectable in the circulation for a
HBV among U.S. donors. The most recent estimate of the few days after inoculation. Although there is no need to defer
residual risk of post-transfusion HBV infection is 1:205,000 a recent vaccine recipient for safety reasons, it is wise to avoid
among repeat donations or 1:144,000 among all donations, collection of blood from such a person for about 7 days after
in the absence of NAT.32,33 However, such estimates inherently vaccination. Otherwise, there is a risk that the donor will
depend on a number of assumptions and extrapolations, and be confirmed positive for HBsAg and will thus be deferred.
the frequency of observed cases of post-transfusion hepati- Finally, although the risk of transmission of HBV by blood
tis B is insufficient to support this estimate. Additionally, components has almost been eliminated and although there
it seems likely that many of the cases of post-transfusion should be no cases from manufactured plasma products as a
hepatitis that are reported are not, in fact due to transfusion result of careful donor management and viral inactivation of
transmission, as recently shown by Alter and her associates the final product, chronic users of blood and blood products
(presented at the Advisory Committee on Blood Safety and should certainly receive the HBV vaccine.
Availability, August 26 to 27, 2004; see http://www.hhs.gov/
bloodsafety/presentations/MiriamAlter.pdf). She found that
only one of 49 cases of supposed post-transfusion hepatitis HEPATITIS D
could be unequivocally linked to transfusion of an infectious
unit. On the other hand, there is little in the way of organized Hepatitis D is caused by a small satellite virus, originally
lookback for hepatitis B, at least in the United States, but data termed the delta virus, that can replicate only in the presence of
from Japan clearly show that window-period HBV infection HBV infection. The virus is an RNA virus and, almost unique
may occur. among animal viruses, has a circular genome. It encodes a
Nucleic acid amplification testing for HBV is available and single peptide, originally observed in infected hepatocytes
has been implemented in a number of countries around the and termed delta antigen. The infectious form of HDV is
HEPATITIS A, B, AND NON-A, NON-B, NON-C VIRUSES
coated with HBsAg. Co-infection with HBV and HDV results and is found at high prevalence among individuals who
in a more serious disease than does HBV infection alone. In have undergone multiple transfusions. However, it has not
truth, HDV has little current relevance to transfusion safety, proved possible to demonstrate that infection with HGV
because measures designed to detect infectivity for HBV also is associated with hepatitis or even with signs of mild liver
detect all individuals co-infected with HDV.40,41 disease, such as elevated ALT levels. Indeed, HGV appears
to be a virus that is currently in search of a disease. The
term hepatitis in its name may be a misnomer, attributable
HEPATITIS E only to the fact that the virus was found in association with
hepatitis in the first place. It is also important to recognize
Hepatitis E virus (HEV) causes an epidemic form of hepa- that the worldwide distribution of HGV clearly shows that
titis that is self-limited.42,43 The disease is somewhat similar it is not a new virus but rather one that has coexisted with
to hepatitis A, although it is much more severe in pregnancy. humans for many centuries.
Transmission is by the fecal-oral route and is most often There have, however, been some intriguing observations
water-borne. The virus is related to the calicivirus group and, that clearly suggest that HGV/GBV-C may have an impact on
as such, is a nonenveloped virus that has an RNA genome. the course of HIV disease. For example, studies have shown
Although there is some evidence (based largely on sero- lower mortality among co-infected individuals relative to
prevalence studies) that HEV is present in the United States, those with HIV only.51,52 The mechanism for this effect is
it is found predominantly in tropical countries. Indeed, most unclear, but it may be due to the effects of infection on the
cases identified in the United States appear to have resulted levels of a number of chemokines.53
from infections that occurred in countries where HEV is
endemic. HEV infection is self-limited and acute, and it has
generally been thought that there is little risk of transmission TTV AND SEN-V
by transfusion. Until recently, no cases of transfusion-associ-
ated HEV infection have been reported.44 However, a num- Curiously, another pair of viruses were separately identified
ber of recent cases of such transmission have been reported among individuals with hepatitis and were also shown to
from nonendemic areas, emphasizing that acute infections be poorly, if at all, associated with hepatitis. These viruses
are also transmissible by transfusion, as long as there is an were also found to cause prolonged viremia and, in some
asymptomatic viremic phase.45,46 This has, of course, been cases, turned out to be present in up to 90% of the popu-
abundantly illustrated by the example of West Nile virus in lation. Like HGV, they were readily transmitted by transfu-
the United States.47 sion. Both viruses were thought to be representatives of the
circovirus group: small, nonenveloped viruses with a circular
DNA genome. This group of viruses had not previously been
HEPATITIS G VIRUS/GBV-C described among humans.
Workers in Japan used representational difference anal- 43
The vast majority of cases of post-transfusion hepatitis have ysis to isolate DNA sequences from three patients with
been shown to be caused by HBV or HCV. Nevertheless, unexplained post-transfusion hepatitis. The sequence was 589
there continue to be some residual cases of hepatitis asso- established as viral, and the virus was named TTV, reflecting
ciated with transfusion. For example, in Alter’s continuing the initials of the patient from whom it was isolated.54 A con-
studies at the National Institutes of Health, approximately siderable amount of research has revealed some of the key
12% of cases of non-A, non-B hepatitis could not be attrib- features of this agent.55 It is a small virus with a covalently
uted to either HBV or HCV. These residual cases appear to be closed, circular DNA genome of approximately 3800 bases.
mild and self-limited, and some may not even have an infec- The virus is thought to be nonenveloped and is most closely
tious etiology. Such cases have, however, led to a continuing related to the circoviruses, which are responsible for a num-
search for additional hepatitis viruses. The first of such puta- ber of diseases in plants and a handful of mammalian and
tive hepatitis viruses was identified by two separate groups in avian disease states. The classification of TTV is currently
the late 1990s. In one case, scientists at Abbott Laboratories incomplete, and proposals have been made to place it in at
looked for genomic sequences related to those of an existing least two genera. What has become apparent is that TTV is a
isolate known as the GB virus (GBV), which had previously member of an extremely diverse group of viruses, as demon-
been associated with hepatitis in a physician. Three viruses strated by considerable variation in genomic sequence.
were identified, one of which (termed GBV-C) was found Epidemiologic studies have confirmed that TTV is
among a number of human sources.48 Working in paral- a widely distributed virus and have clearly established that it
lel, but using a different approach, scientists at Gene-Labs is transmissible by transfusion. Interestingly, it also appears
isolated viral RNA sequences and characterized a virus they to be transmitted by the vertical, fecal-oral, and perhaps
termed hepatitis G virus (HGV).49 It is generally accepted that other routes.
these two isolates were, in fact, representatives of essentially A key issue is the clinical significance of this group of
the same virus group, which is now known as HGV. viruses. Although the original source of the virus and its
HGV, like HCV, appears to be closely related to the apparent association with ALT elevations implied that TTV
Flavivirus group. It is found among a relatively high pro- was indeed a hepatitis virus, this identification no longer
portion of the normal population, as exemplified by blood seems tenable. Indeed, there are far more infections with-
donors. Its presence has been demonstrated both by sero- out ALT elevations than with such evidence of liver disease.
prevalence studies, in which the frequency of antibodies is Even in clear, transfusion-associated transmission of TTV,
3% to 15%, and more interestingly by detection of viral RNA the recipients did not manifest ALT elevations in any pat-
in the plasma of 1% to 3% of normal subjects.50 Perhaps not tern that could be associated with the infection. Thus, at
surprisingly, the virus is readily transmissible by transfusion this stage, there is little evidence that this virus expresses
TRANSFUSION MEDICINE any pathogenic potential. However, it is certainly too early 18. Ling CM, Overby LR. Prevalence of hepatitis B virus antigen as revealed
to conclude that this entire group of viruses is without any by direct radioimmune assay with 125–1-antibody. J Immunol 1972;109:
834–841.
clinical significance. 19. Dodson SF, Issa S, Araya V, et al. Infectivity of hepatic allografts with
After the recognition of TTV, the search for hepatitis antibodies to hepatitis B virus. Transplantation 1997;64:1582–1584.
viruses continued, and Primi and colleagues56,57 used degener- 20. Lin OS, Keeffe EB. Current treatment strategies for chronic hepatitis B
ate primers derived from TTV to probe samples from selected and C. Annu Rev Med 2001;52:29–49.
21. Lok ASF, McMahon BJ. Chronic hepatitis B: update of recommenda-
patients. An isolate was identified and named SEN-V (after the tions. Hepatology 2004;39:1–5.
initials of the source patient, an HIV-infected injection drug 22. Lau GKK, Piravisuth T, Luo KX, et al. Peg-interferon alfa-2a, lamivu-
user). Eventually, at least eight different strains were isolated, dine, and the combination for HBeAg-positive chronic hepatitis B.
termed A through H. The SEN group has been shown to be NEJM 2005;352:2682–2695.
one branch of the TTV group, seeming to share its epidemio- 23. Hadziyannis SJ, Tassopoulos NC, Heathcote EJ, et al. Long-term therapy
with adefovir dipivoxil for HBeAg-negative chronic hepatitis B. NEJM
logic characteristics. Although two of the strains (D and H) 2005;352:2673–2681.
have been associated with evidence of transfusion-associated 24. McQuillan GM, Coleman PJ, Kruszon-Moran D, et al. Prevalence of
hepatitis, a causal relationship has not been established.56,57 hepatitis B virus infection in the United States: The National Health and
Thus, attempts to define the etiologic agent(s) of non-A Nutrition Examination Surveys, 1976 through 1994. Am J Public Health
1999;89:14–18.
through E hepatitis do not appear to have been successful 25. Dodd RY. Germs, gels, and genomes: A personal recollection of 30 years
to date. The availability of powerful genomic techniques has in blood safety testing. In Stramer SL (ed): Blood Safety in the New
certainly led to the recognition of previously undescribed Millenium. Bethesda, Md., American Association of Blood Banks, 2001,
viruses, but it has not been possible to associate any partic- pp 97–122.
ular disease with these remarkably widespread viruses. No 26. CDC. Hepatitis Surveillance Report No. 60. Atlanta: U.S. Department of
Health and Human Services, Centers for Disease Control and Preven-
doubt other such orphan viruses will be identified. However, tion, 2005.
it will be important to avoid an automatic assumption of 27. Almeida JD, Rubenstein D, Stott EJ. New antigen–antibody system in
causality when such viruses are isolated from patients with Australia-antigen positive hepatitis. Lancet 1971;2:1225–1227.
any given disease state. It may also be important to question 28. Mahoney FJ. Update on diagnosis, management, and prevention of
hepatitis B virus infection. Clin Microbiol Rev 1999;12:351–366.
whether all residual transfusion-associated hepatitides do 29. Busch MP. HIV, HBV and HCV: new developments related to transfu-
indeed have an infectious etiology. sion safety. Vox Sang 2000;78:253–256.
30. Kleinman SH, Kuhns MC, Todd DS, et al. Frequency of HBV DNA
detection in U.S. blood donors testing positive for the presence of anti-
REFERENCES HBc: implications for transfusion transmission and donor screening.
Transfusion 2003;43:696–704.
1. Cuthbert JA. Hepatitis A: Old and new. Clin Microbiol Rev 2001;14: 31. Schreiber GB, Busch MP, Kleinman SH, Korelitz JJ. The risk of transfu-
38–58. sion-transmitted viral infections. NEJM 1996;334:1685–1690.
2. Hollinger FB, Khan NC, Oefinger PE, et al. Posttransfusion hepatitis 32. Dodd RY, Notari EP, Stramer SL. Current prevalence and incidence
type A. JAMA 1983;250:2313–2317. of infectious disease markers and estimated window-period risk in
3. Azimi PH, Roberto RR, Guralnick J, et al. Transfusion-acquired hepa- the American Red Cross blood donor population. Transfusion 2002;42:
III titis A in a premature infant with secondary nosocomial spread in an 975–979.
intensive care nursery. Am J Dis Child 1986;140:23–27. 33. Dodd RY. Current safety of the blood supply in the United States. Int J
590 4. Giacoia GP, Kasprisin DO. Transfusion-acquired hepatitis A. South Med Hematol 2004;80:301–305.
J 1989;82:1357–1360. 34. Stramer SL. Pooled hepatitis B virus DNA testing by nucleic acid ampli-
5. Lee KK, Vargo LR, Le CT, Fernando L. Transfusion-acquired hepatitis fication: Implementation or not. Transfusion 2005;44:1242–1246.
A outbreak from fresh frozen plasma in a neonatal intensive care unit. 35. Biswas R, Tabor E, Hsia CC, et al. Comparative sensitivity of HBV NATs
Pediatr Infect Dis 3 1992;11:122–123. and HBsAg assays for detection of acute HBV infection. Transfusion
6. Mosley JW, Nowicki MI, Kasper CK, et al. Hepatitis A virus transmission 2003;43:788–798.
by blood products in the United States. Vox Sang 1994;67(Suppl)1:24–28. 36. Roth WK, Weber M, Seifried E. Feasibility and efficacy of routine PCR
7. Bower WA, Nainan OV, Han XH, Margolis HS. Duration of viremia in screening of blood donations for hepatitis C virus, hepatitis B virus, and
hepatitis A virus infection. J Infect Dis 2000;182:12–17. HIV-1 in a bloodbank setting. Lancet 1999;353:359–363.
8. Gowland P, Fontana S, Niederhauser C, Taleghani BM. Molecular and 37. Blum HE. Hepatitis B virus: Significance of naturally occurring mutants.
serologic tracing of a transfusion-transmitted hepatitis A virus. Trans- Intervirology 1993;35:40–50.
fusion 2004;44:1555–1561. 38. Hsu HY, Chang MH, Liaw SH, et al. Changes of hepatitis B surface anti-
9. CDC. Hepatitis A outbreak associated with green onions at a restau- gen variants in carrier children before and after universal vaccination in
rant—Monaca, Pennsylvania, 2003. MMWR Morb Mortal Wkly Rep Taiwan. Hepatology 1999;30:1312–1317.
2003;52:1155–1157. 39. Jongerius IM, Wester M, Cuypers HTM, et al. New hepatitis B virus
10. Soucie JM, Robertson BH, Bell BP, et al. Hepatitis A virus infections mutant form in a blood donor that is undetectable in several hepatitis B
associated with clotting factor concentrate in the United States. Transfu- surface antigen screening assays. Transfusion 1998;38:56–59.
sion 1998;38:573–579. 40. Lai MMC. The molecular biology of hepatitis delta virus. Annu Rev
11. Mannucci PM, Gdovin S, Gringeri A, et al. Transmission of hepatitis Biochem 1995;64:259–286.
A to patients with hemophilia by factor VIII concentrates treated with 41. Liaw YF, Tsai SL, Sheen IS, et al. Clinical and virological course of
organic solvent and detergent to inactivate viruses. Ann Intern Med chronic hepatitis B virus infection with hepatitis C and D virus mark-
1994;120:1–7. ers. Am J Gastroenterol 1998;93:354–359.
12. Johnson Z, Thornton L, Tobin A, et al. An outbreak of hepatitis A among 42. Purcell RH. The discovery of the hepatitis viruses. Gastroenterology
Irish haemophiliacs. Int J Epidemiol 1995;24:821–828. 1993;104:955–963.
13. Lawlor E, Graham S, Davidson F, et al. Hepatitis A transmission by fac- 43. Thomas DL, Yarbough PO, Vlahov D, et al. Seroreactivity to hepati-
tor IX concentrates. Vox Sang 1996;71:126–128. tis E virus in areas where the disease is not endemic. J Clin Microbiol
14. Lee WM. Hepatitis B virus infection. NEJM 1997;337:1733–1745. 1997;35:1244–1247.
15. Blumberg BS, Alter HJ, Visnich S. A “new” antigen in leukemia sera. 44. Mateos ML, Camarero C, Lasa E, et al. Hepatitis E virus: Relevance in
JAMA 1965;191:541–546. blood donors and other risk groups. Vox Sang 1998;75:267–269.
16. Blumberg BS, Gerstley BJS, Hungerford DA, et al. A serum antigen 45. Matsubayashi K, Nagaoka Y, Sakata H, et al. Transfusion-transmitted
(Australia antigen) in Down’s syndrome, leukemia and hepatitis. Ann hepatitis E caused by apparently indigenous hepatitis E virus strain in
Intern Med 1967;66:924. Hokkaido, Japan. Transfusion 2004;44:934–940.
17. Gocke DJ, Greenberg HB, Kavey NB. Correlation of Australia antigen 46. Boxall EH, Herborn A, Kochethu G, et al. Transfusion transmitted
with posttransfusion hepatitis. JAMA 1970;212:877–879. hepatitis E in a “non endemic” country. Transfus Med 2006;16:79–83.
HEPATITIS A, B, AND NON-A, NON-B, NON-C VIRUSES
47. Pealer LN, Marfin AA, Petersen LR, et al. Transmission of West Nile 53. Xiang J, George SL, Wunschmann S, et al. Inhibition of HIV-1 replica-
virus through blood transfusion in the United States in 2002. NEJM tion by GB virus C infection through increases in RANTES, MIP-1α,
2003;349:1236–1245. MIP-1β, and SDF-1. Lancet 2004;363:2040–2046.
48. Leary TP, Muerhoff AS, Simons JN, et al. Sequence and genomic orga- 54. Nishizawa T, Okamoto H, Konishi K, et al. A novel DNA virus (TTV)
nization of GBV-C. A novel member of the Flaviviridae associated with associated with elevated transaminase levels in posttransfusion hepatitis
human non-A-E hepatitis. J Med Virol 1996;48:60–67. of unknown etiology. Biochem Biophys Res Commun 1997;241:92–97.
49. Linnen J, Wages J Jr, Zhang-Keck ZY, et al. Molecular cloning and dis- 55. Bendinelli M, Pistello M, Maggi F, et al. Molecular properties, biology,
ease association of hepatitis G virus: a transfusion transmissible agent. and clinical implications of Tf virus, a recently identified widespread
Science 1996;271:505–508. infectious agent of humans. Clin Microbiol Rev 2001;14:98–113.
50. Allain JP. Emerging viral infections relevant to transfusion medicine. 56. Umemura T, Yeo AE, Sottini A, et al. SEN virus infection and its
Blood Rev 2000;14:173–181. relationship to transfusion-associated hepatitis. Hepatology 2001;33:
51. Tillmann HL, Heiken H, Knapik-Botor A, et al. Infection with GB 1303–1311.
virus C and reduced mortality among HIV-infected patients. NEJM 57. Tanaka Y, Primi D, Wang RY, et al. Genomic and molecular evolution-
2001;345:715–724. ary analysis of a newly identified infectious agent (SEN virus) and its
52. Xiang J, Wunschmann S, Diekema DJ, et al. Effect of coinfection with relationship to the TT virus family. J Infect Dis 2001;183:359–367.
GB virus C on survival among patients with HIV infection. NEJM
2001;345:707–714.
43
591
Chapter 44
Hepatitis C
Roger Y. Dodd
Table 44–1 Enzyme Immunoassay Tests for Anti-HCV: Components and Performance Characteristics
Version of Test (manufacturer) Peptides Sensitivity* Specificity*
*
Based on manufacturers’ claims in product inserts.
†
Based on a diagnosis of chronic non-A, non-B hepatitis—ALT elevated >6 months, HBsAg negative.
‡
Includes c33 and c100 sequences.
HEPATITIS C
automated platforms are available elsewhere in the world and ●
Negative: No band with a greater intensity than the weak
are soon expected to be licensed in the United States. Most positive control.
diagnostic reagents are defined as devices, but those that are ●
Indeterminate: Only one band reactive or any pattern in
used in the preparation of blood and blood products are association with a reactive SOD band.
required to meet the more stringent biologics requirements,
Table 44–2 summarizes the results of testing a large num-
because blood itself is defined as a biologic.
ber of volunteer blood donations with HCV EIA and SIAs.
The sensitivity of these tests was defined in trials using
Unlike the situation with HIV, there does not seem to be a
patients diagnosed with NANBH. Because this diagnosis is not
common pattern for the sequence of appearance of reactive
specific, the significance of the sensitivity figures is unclear. Of
bands in the blot during seroconversion with HCV. However,
more importance now, at least in the context of blood donor
a number of studies have defined the relationship between
screening, is the ability of the test method to detect infection
particular bands, or band patterns, and the likelihood that a
at the earliest possible time during the seroconversion period.
sample will also contain detectable HCV RNA. For example,
Studies of seroconversion panels suggest that currently avail-
Dow and colleagues35 reviewed data from 177 blood donor
able tests detect antibodies, on average, 70 to 80 days after
specimens that were reactive on EIA and tested positive on the
exposure to the source of infection and 40 to 60 days after the
version 3.0 SIA. Among 82 samples with four positive bands,
initial detection of HCV RNA in the plasma. In addition, it
69 (84.1%) were RNA-positive. Of the 54 samples with three
should be noted that earlier versions of the EIA tests clearly
positive bands, 40 (74.1%) were RNA-positive, whereas only 14
failed to detect some infected individuals.38 This is not sur-
(34.1%) of the 41 samples with two positive bands were RNA-
prising, because only a very limited number of viral epitopes
positive. Among the samples with indeterminate SIA results,
were included in the capture reagent. Current tests for anti-
the frequencies of RNA-positive results were 3 of 154 for c22, 1
HCV are whole-antibody assays. No tests are yet available for
of 220 for c33, 1 of 191 for c100, and 0 of 380 for NS5.35 Thus,
specific detection of immunoglobulin M anti-HCV.
a few of the indeterminate patterns may be associated with
Although these EIA tests have high sensitivity and speci-
the presence of RNA and, therefore, of active HCV infection.
ficity, there is a possibility that some reactive test results
Similar data were published by Dodd and Stramer.36
are nonspecific. When the EIA is used to screen blood
donors, its actual positive predictive value is 70% to 80%.36
Consequently, it is recommended that all asymptomatic TESTS FOR HCV RNA
individuals who test as reactive on EIA repeatedly should be
further tested with an additional, more specific procedure. Tests for HCV RNA serve an important role in diagnosis
Currently, only one such immunologic method is licensed and patient management.9,39 A variety of procedures is avail-
and available in the United States. This is a strip immuno- able, all of which depend on nucleic acid amplification, with
blot assay (RIBA 3.0) that is constructed by application of one exception. The reverse transcriptase polymerase chain
recombinant or synthetic viral peptides representing c22, reaction (RT-PCR) is perhaps the most familiar. Viral RNA
c33, c100, 5-1-1, and NS5 regions to nitrocellulose strips in a 44
is reverse-transcribed to DNA, and two primers are used to
fixed pattern. The c-100 and 5-1-1 peptides are present in the define a sequence for repetitive amplification using a temper-
same band on the strip. The expression carrier protein for 595
ature-resistant DNA polymerase and a temperature-cycling
the recombinant viral antigens is also applied (superoxide
protocol. A variety of methods may be used to detect the
dismutase, SOD), as are strong and weak positive controls.
resulting amplified sequence, including visualization in gels,
Patient or donor samples are added to the strips and, after
hybridization with labeled probes, and detection of ampli-
washing, adherent antibodies are detected by an appropriate
cons in real time. Both qualitative and quantitative proce-
enzyme-conjugated antiglobulin and visualization reaction. dures are available.5,40,41 In most cases, a conserved segment
The number and intensity of bands are scored, and the result
of the 5′ untranslated region of the genome is selected for
is interpreted as follows: amplification. PCR-based assays are available commercially
●
Positive: Two or more bands with an intensity equal to, (e.g., Roche Molecular Systems) as well as from independent
or greater than, that of the weak positive control band, reference laboratories; alternatively, they may be developed
plus nonreactive SOD band. in-house with the use of standard technologies.
Table 44–2 Results of Testing 19.2 Million Blood Donations with HCV 3.0 EIA and Version 3.0 SIA
and Percentage of RNA-Positive Samples among Subgroups of SIA-Positive Subjects
*
Sample of 200 tested.
†
2347 tested.
EIA, enzyme immunoassay; HCV, hepatitis C virus; NA, not applicable; RR, repeatedly reactive; SIA, Strip immunoblot assay.
Data courtesy of Susan L. Stramer, Ph.D.(personal communication).
TRANSFUSION MEDICINE Another technique, known as nucleic acid sequence– DIAGNOSTIC ALGORITHM
based amplification (NASBA; a proprietary technology from
Organon-Technika),42 and the very similar transcription- The Centers for Disease Control and Prevention (CDC)
mediated amplification (TMA; a proprietary technology has published an algorithm for diagnostic testing for
from Gen-Probe Inc.) can be performed without tempera- HCV among asymptomatic individuals (Fig. 44–1).14 This
ture cycling. Two enzymes are used to produce an RNA algorithm is somewhat different from that recommended
amplicon, which can be detected by a variety of methods, for blood donor screening, in that it explicitly permits
including the hybridization protection assay. It should be the use of NAT to confirm a repeatedly reactive EIA
noted that Chiron owns patent rights to the HCV genomic result. However, NAT-negative samples must be further
sequence used for amplification. evaluated by an SIA. This algorithm also provides useful
Sample collection, stability, and preparation for ampli- guidance in the context of treatment, but a more compre-
fication are all important and must be properly controlled. hensive guide to treatment was published in 2004.9 Given
HCV RNA is quite labile, and it is preferable, if not essen- that blood donors are now routinely evaluated by NAT for
tial, to collect specimens in EDTA. Samples should be main- HCV RNA, it is hoped that these results may be incorpo-
tained at refrigerated temperatures and tested with minimal rated into the notification and management of seropositive
delay. A number of different methods may be used to pre- donors.37,45
pare the RNA for testing, including conventional extraction
from ultracentrifugal pellets, extraction on silica, and probe-
capture techniques. IMPACT OF BLOOD DONOR SCREENING
Finally, in a method known as the branched-chain DNA AND TESTING FOR HCV
assay (B-DNA), a probe is labeled with a large branched
DNA molecule that carries many copies of the detection Results of Testing
label. This method does not amplify the target nucleic
acid; rather, it provides a system in which numerous label In 2001, the frequency of positive results among first-
molecules may be associated with a single target sequence. time blood donations as defined by RIBA was 0.3%,
As might be expected, this technique is not as sensitive which is about one fifth of the national prevalence rate
as amplification. However, the lack of sensitivity can be of 1800 per 100,000. The incidence of new infections in
exploited to differentiate those patients with high levels of the donor population is 1.9 new infections per 100,000
circulating RNA.43,44 person-years; the corresponding national incidence fig-
Within the United States, almost all blood donations have ure is 13.4 per 100,000 person-years.47 Thus, the donor
been tested for HCV RNA since 1999.34 Two methods are rate is about one fifth of the national rate. These differ-
in use, RT-PCR(Roche) and TMA (Chiron/Gen-Probe). To ences are attributable, at least in part, to the procedures
date, all such testing has been performed on small pools of used to recruit safer populations for donation and to the
III plasma samples, with current pool sizes of 24 for the Roche measures used to question presenting donors about their
procedure and 16 for the Chiron/Gen-Probe method. Both medical and behavioral histories. However, it should be
596 tests have been licensed for routine use by the FDA and both
have an analytical sensitivity of 12 or fewer copies of RNA/
mL by probit analysis and a 95% detection level of 30 to 60
copies/mL. The overall sensitivity of the testing is, of course, Negative
(nonreactive)
proportionately reduced in pooled testing. EIA for anti-HCV Stop
A number of methods are available for the determi-
Positive
nation of genotype and subtype. Not all methods are (repeatedly reactive)
able to discriminate every genotype or subtype, however. OR
Available methods may be broadly separated into those
that depend on immunologic differentiation and those RIBA™ Negative RT-PCR
for anti-HCV for HCV RNA
that directly detect variation in nucleic acid sequences. In
the former case, specific peptides derived from the NS4
Negative Indeterminate Positive Positive
region are used to probe for antibodies in the specimen.
Nucleic acid–based techniques include sequencing ampli- Stop Additional Medical
cons from selected genomic areas, PCR using genome- laboratory evaluation
specific primers, DNA-enzyme immunoassay (DEIA), evaluation
(e.g., RT-PCR, ALT)
restriction fragment length polymorphism (RFLP)
analysis of amplicons, and differential hybridization of
amplicons with specific probes. Two of these approaches Positive RT-PCR
Negative RT-PCR
are commercially available (GEN-ETI-K DEIA kit and and normal ALT or abnormal ALT
INNO-LiPA HCVI and HCV II). The reader is referred ALT Alanine aminotransferase
Anti-HCV Antibody to HCV
elsewhere for further details.23,45 Stop
EIA Enzyme immunoassay
Relatively little information is available about the impact RIBA™ Recombinant immunoblot assay
of HCV genotype on the sensitivity of diagnostic tests, and RT-PCR Reverse transcriptase polymerase
chain reaction
some data suggest little variation in the analytic sensitivity
of NAT.46 However, a recent publication reviewing the per- Figure 44–1 Algorithm for hepatitis C virus (HCV) testing among
asymptomatic individuals. (From Centers for Disease Control and
formance characteristics of numerous blood screening assays Prevention: Recommendations for prevention and control of hepatitis
for HCV antibodies suggests that most tests are not seriously C virus (HCV) infection and HCV-related chronic disease. MMWR Morb
impacted by genotype variations (El-Nageh, in press). Mortal Wkly Rep 1998;487[RR-19]:1–39.)
rates for HCV antibody53 suggest that this risk is not chang-
HEPATITIS C
noted that only a very few donors are actually deferred
on the basis of their response to risk questions, in part ing significantly, at least through the end of 2003.
because donors do not always provide complete answers48 It is interesting to speculate about the higher incidence
but also perhaps because potential donors make a con- of infections among first-time donors, but there are no
scious decision to avoid giving blood so as not to have to definitive data. One possibility is that there are more test-
answer the questions. seekers among people presenting to donate for the first time,
although this is not necessarily supported by published data
on interviews of RNA NAT-positive donors, because both
Risk of Post-transfusion HCV Infection
first-time and repeat donors were found to be positive and
It is clear from a number of published studies that donor to acknowledge risk factors for infection.17 It is also possible
screening and testing measures have had a profound that the process of giving blood has an educational impact
impact on the incidence of post-transfusion hepatitis and that individuals who learn about risk factors for blood-
C. Indeed, prospective studies have not demonstrated borne infections defer themselves from future donation.
any infections since the implementation of the so-called Finally, it should be noted that there are cases in which
multi-antigen tests for antibodies to HCV. But there is HCV has been transmitted by blood units that have been
also substantial evidence for the efficacy of a variety of fully tested for HCV, even after the implementation of
screening approaches used even before this time. The NAT.54 Such cases have usually been detected as a result of
essentially complete elimination of commercial donation lookback and are too few to offer any quantitative estimate
of whole blood had a major effect on reducing the frequency of residual risk. Window-period theory would strongly
of post-transfusion hepatitis. It is also believed that a suggest that such cases would occur, and this is supported
further reduction was seen as a result of more stringent by the detection of HCV RNA in window-period sam-
donor questioning to reduce the risk of transmission of ples, using ultrasensitive test methods. It is of interest to
HIV and acquired immunodeficiency syndrome (AIDS), speculate on the minimal infectious dose of HCV. Busch
although subsequent information implies that the major has developed models based on the conservative assump-
effect would have come from a reduction in the number tion that it is as low as one genome equivalent per 20 mL
of injection drug users. of plasma. Using this assumption and back-calculating, the
The first study to clearly demonstrate the impact of theoretical window period from the observed doubling rate
testing measures on hepatitis C infection was published of HCV plasma RNA during early infection has led to esti-
by Nelson and colleagues.49 They evaluated samples from mates that are largely compatible with currently accepted
a large population of patients undergoing cardiac surgery, window period estimates.55
using the first-generation test for anti-HCV. These research-
ers found that the risk of infection was 0.45% per unit prior
to the implementation of any testing. After implementa- LOOKBACK
tion of testing for ALT and anti-HBc, the rate of infection 44
dropped to 0.19%. Finally, once the version 1.0 EIA test was A positive test result for HCV antibodies does not provide any
implemented, the rate dropped to 1 per 3300 units (0.03%), information about the duration of infection, even if accom- 597
a reduction of 84%.49 A subsequent reevaluation using the panied by a positive finding for HCV RNA. Consequently,
more sensitive version 2.0 test on the blood recipients sug- if a blood donor is found to be HCV antibody positive, it
gested that the actual risk was closer to 1:1700.50 Once the is possible that prior donations from that individual were
version 2.0 test was introduced for blood donor screening, infectious for HCV. This could occur by two broad mecha-
however, the frequency of residual infection declined pro- nisms. First, a previous donation could have been made in
foundly. As pointed out previously, cases of hepatitis C were the infectious but seronegative window period. Second, the
no longer observed in prospective studies, and risk estimates prior donations could have been collected at a time when a
then had to be developed on the basis of the length of the less sensitive test was in use or even before testing was ini-
window period (as determined from post-transfusion infec- tiated. Accordingly, a focused lookback program has been
tions) and the incidence of new infections within the donor initiated to locate, notify, test, and, if appropriate, treat recip-
population. In a landmark publication in 1996, Schreiber ients of such potentially infectious prior donations. Studies
and colleagues estimated the residual risk of HCV infection in Canada and elsewhere suggest that up to 70% of such
at 1 per 103,000 donations, based on a window period of recipients may indeed have been infected.56,57 However, in
82 days and an incidence rate of 4.84 per 100,000 person- the United States, the effort appears to have been less pro-
years.51 Subsequent evaluations account for a 12- to 13-day ductive.
reduction in the window period attendant on the imple- A team from the CDC reported that an estimated 98,484
mentation of the version 3.0 EIA and an overall decline in blood components were identified as potentially infectious.58
the incidence of HCV infection to 2.09 per 100,000 per- Of these, 85% had been transfused. This interim study found
son-years. The latter figure translated to a residual risk of that lookback had been completed for 80% of the transfused
1:276,000 repeat donations. The addition of HCV NAT was products; 69% of the recipients had died. Of those living,
estimated to markedly reduce this risk to 1:1.935 million 78% were successfully notified that they had received a poten-
repeat donations. In the same paper, it was observed, on tially infectious blood component. It was estimated that, of
the basis of the results of NAT, that the incidence of HCV recipients notified, 49.5% were tested for anti-HCV; of those,
infections was 2.4 times higher among first-time donors.52 18.9% were seropositive, but 32% of these individuals were
Allowing for a 23% frequency of first-time donors in the already aware that they were seropositive. Thus, at the time of
Red Cross system, this translated to an overall risk of 1:1.39 publication of the study, it was estimated just over 1000 indi-
million donations.3 The relative stability of the detection viduals were newly notified of unexpected HCV infection;
rates for HCV RNA and the lack of change in prevalence this estimate translated to a national figure of 1520, on the
TRANSFUSION MEDICINE assumption that the lookback process was to be completed. REFERENCES
The figure represents fewer than 1% of all individuals who
1. Aach RD, Szmuness W, Mosley JW, et al. Serum alanine aminotransferase
may have been infected as a result of transfusion.58 It should of donors in relation to the risk of non-A, non-B hepatitis in recipients.
be noted that this component of the lookback program was The Transfusion-Transmitted Viruses Study. NEJM 1981;304:989–994.
restricted only to donations that were identified as a result of 2. Alter HJ, Purcell RH, Holland PV, et al. Donor transaminase and recipient
testing with multi-antigen tests (i.e., versions 2.0 or 3.0). It is hepatitis: Impact on blood transfusion services. JAMA 1981;246: 630–634.
3. Dodd RY. Current safety of the blood supply in the United States. Int
likely that, as lookback is extended to cover donors initially J Hematol 2004;80:301–305.
identified by the version 1.0 test, the proportional yield will 4. Alter HJ, Seeff LB. Recovery, persistence, and sequelae in HCV infection:
be greater, but the efficiency of the process will certainly be a perspective on long-term outcome. Semin Liver Dis 2001;20:17–35.
affected by availability of required records. 5. Management of Hepatitis C. NIH Consensus Statement 1997;15:1–41.
On the other hand, the essential completion of achiev- 6. Liang TJ, Rehermann B, Seeff LB, Hoofnagle JH. Pathogenesis, natural
history, treatment, and prevention of hepatitis C. Ann Intern Med 2000;
able lookback affecting recipients of blood donated prior to 132:296–305.
the implementation of testing means that almost all avail- 7. Carrat F, Bani-Sadr F, Pol S, Rosenthal E, et al. Pegylated interferon
able cases now represent lookback from incident cases of alfa-2b vs standard interferon alfa-2b, plus ribavirin, for chronic hepa-
HCV infection. The yield of this aspect of lookback is now titis C in HIV-infected patients: A randomized controlled trial. JAMA
2004;292:2839–2848.
extremely low, and such yield will be further reduced as the 8. Jacobson IM, Gonzalez SA, Ahmed F, et al. A randomized trial of
impact of NAT is manifested. The ethical imperatives for pegylated interferon alpha-2b plus ribavirin in the retreatment of
lookback continue, but the public health benefits of this chronic hepatitis C. Am J Gastroenterol 2005;100:2453–2462.
process are becoming vanishingly small. 9. Strader DB, Wright T, Thomas DL, Seeff LB. Diagnosis, management
and treatment of hepatitis C. Hepatology 2004;39:1147–1171.
10. Shepard CW, Finelli L, Alter MJ. Global epidemiology of hepatitis C
virus infection. Lancet Infect Dis 2005;5:558–567.
COMMENT AND SUMMARY 11. Frank C, Mohamed MK, Strickland GT, et al. The role of parenteral
antischistosomal therapy in the spread of hepatitis C virus in Egypt.
In its way, the history of hepatitis C has been as remarkable Lancet 2000;355:887–891.
as that of HIV and AIDS. In retrospect, hepatitis C was the 12. Alter MJ, Kruszon-Moran D, Nainan OV, et al. The prevalence of hepa-
titis C virus infection in the United States, 1988 through 1994. NEJM
blood-borne agent most commonly transmitted by transfu- 1999;341:556–562.
sion in the United States, right up until the early 1990s. The 13. Alter MJ. Hepatitis C virus infection in the United States. J Hepatol
legacy of this problem will be seen for many years, as the 1999;31:88–91.
long-term health consequences of HCV infection become 14. CDC. Recommendations for prevention and control of hepatitis C virus
(HCV) infection and HCV-related chronic disease. MMWR Morb Mor-
manifest. The almost complete elimination of the problem tal Wkly Rep 1998;47 (RR-19):1–39.
of transfusion-transmitted HCV is the result of many years 15. Alter MJ. Epidemiology of hepatitis C and lookback. Hematology
of dedicated study of an intractable problem. There were 1999;1999:418–421.
no simple solutions; the virus was refractory to laboratory 16. Conry-Cantilena C, VanRaden M, Gibble J, et al. Routes of infection,
study, and there was a truly frustrating inability to identify viremia, and liver disease in blood donors found to have hepatitis C
III virus infection. NEJM 1996;334:1691–1696.
any serologic markers of infection. Indeed, only at the very 17. Orton SL, Stramer SL, Dodd RY, Alter MJ. Risk factors for HCV infec-
598 end of the 20th century was a naturally occurring, circulat- tion among blood donors confirmed to be positive for the presence of
ing viral antigen recognized and then only in the few weeks HCV RNA and not reactive for the presence of anti-HCV. Transfusion
preceding seroconversion. 2004;44:275–281.
18. Choo Q-L, Kuo G, Weiner AJ, et al. Isolation of a cDNA clone derived
As with HIV, the first laboratory approach to abrogating from a blood-borne non-A, non-B viral hepatitis genome. Science
transfusion-transmitted HCV infection was the develop- 1989;244:359–362.
ment of serologic tests. In the case of HCV, however, this 19. Kuo G, Choo Q-L, Alter HJ, et al. An assay for circulating antibodies
development was achieved through the combination of years to a major etiologic virus of human non-A, non-B hepatitis. Science
of study of the disease, development of animal models, and 1989;244:362–364.
20. Simmonds P. Genetic diversity and evolution of hepatitis C virus—15
the painstaking application of new recombinant technology. years on. J Gen Virol 2004;85:3173–3188.
It is fitting that Harvey Alter and Michael Houghton received 21. Wakita T, Pietschmann T, Kato T, et al. Production of infectious hep-
the 2000 Lasker award for this work, but hundreds of others atitis C virus in tissue culture from a cloned viral genome. Nat Med
contributed over many years. 2005;11:791–796.
22. Zhong J, Gastaminza P, Cheng G, et al. Robust hepatitis C virus infec-
Serologic and nucleic acid testing have essentially elimi- tion in vitro. Proc Natl Acad Sci USA 2005;102:9739–9740.
nated the risk of transfusion-transmitted HCV. Plasma deriv- 23. Lau JYN, Mizokami M, Kolberg JA, et al. Application of six hepatitis C
atives are prepared from highly tested starting material and virus genotyping systems to sera from chronic hepatitis C patients in
are further treated with advanced inactivation procedures. It the United States. J Infect Dis 1995;171:281–289.
is of interest to note, however, that whereas the early tests 24. Simmonds P, Holmes EC, Cha T-A, et al. Classification of hepatitis C
virus into six major genotypes and a series of subtypes by phylogenetic
removed a substantial fraction of antibody-positive units, analysis of the NS-5 region. J Gen Virol 1993;74:2391–2399.
they failed to identify all infectious units, probably leading 25. McOmish F, Yap PL, Dow BC, et al. Geographical distribution of hepa-
to the unexpected occurrence of HCV in recipients of some titis C virus genotypes in blood donors: an international collaborative
immunoglobulin products. survey. J Clin Microbiol 1994;32:884–892.
26. Simmonds P, Alberti A, Alter HJ, et al. A proposed system for the
Despite the success of testing for HCV, there continue nomenclature of hepatitis C viral genotypes. Hepatology 1994;19:
to be barriers to other aspects of management of this virus. 1321–1324.
As with many viral diseases, treatment options for HCV 27. Zein NN, Persing DH. Hepatitis C genotypes: Current trends and future
are limited and incomplete. More baffling, however, are implications. Mayo Clin Proc 1996;71:458–462.
the difficulties inherent in understanding and manipulat- 28. Martin P. Hepatitis C genotypes: the key to pathogenicity. Ann Intern
Med 1995;122:227–228.
ing the interactions between the immune system and HCV. 29. Cooreman MP, Schoondermark-Van de Ven EME. Hepatitis C virus:
Development of an effective vaccine seems to be a particu- Biological and clinical consequences of genetic heterogeneity. Scand
larly elusive goal. J Gastroenterol 1996;31(Suppl 218):106–115.
HEPATITIS C
30. Busch MP, Korelitz JJ, Kleinman SH, et al. Declining value of alanine 44. Halfon P, Khiri H, Gerolami V, et al. Impact of various handling and
aminotransferase in screening of blood donors to prevent posttransfu- storage conditions on quantitative detection of hepatitis C virus RNA.
sion hepatitis B and C virus infection. Transfusion 1995;35:903–910. J Hepatol 1996;25:307–311.
31. Busch MP. HIV, HBV and HCV: New developments related to transfu- 45. Lau JYN, Davis GL, Prescott LE, et al. Distribution of hepatits C virus
sion safety. Vox Sang 2000;78:253–256. genotypes determined by line probe assay in patients with chronic hep-
32. Couroucé AM, Le Marrec N, Bouchardeau F, et al. Efficacy of HCV core atitis C seen at tertiary referral centers in the United States. Ann Intern
antigen detection during the preseroconversion period. Transfusion Med 1996;124:868–876.
2000;40:1198–1202. 46. Forman MS, Valsamakis A: Increased sensitivity of the Roche COBAS
33. Tanaka E, Ohue C, Aoyagi K, et al. Evaluation of a new enzyme immu- AMPLICOR HCV test, version 2.0, using modified extraction tech-
noassay for hepatitis C virus (HCV) core antigen with clinical sensitivity niques. J Mol Diagn 2004;6:225–230.
approximating that of genomic amplification of HCV RNA. Hepatol- 47. Dodd RY. Current estimates of transfusion safety worldwide. In Vyas
ogy 2000;32:388–393. GN, Williams AE (eds). Advances in Transfusion Safety. Basel, Karger,
34. Stramer SL, Glynn SA, Kleinman SH, et al. Detection of HIV-1 and 2005, pp 3–10.
HCV infections among antibody-negative blood donors by nucleic 48. Williams AE, Thomson RA, Schreiber GB, et al. Estimates of infectious
acid-amplification testing. NEJM 2004;351:760–768. disease risk factors in US blood donors. JAMA 1997;277:967–972.
35. Dow BC, Buchanan I, Munro H, et al. Relevance of RIBA-3 supplemen- 49. Donahue JG, Murioz A, Ness PM, et al. The declining risk of post-trans-
tary test to HCV PCR positivity and genotypes for HCV confirmation fusion hepatitis C virus infection. NEJM 1992;327:369–373.
of blood donors. J Med Virol 1996;49:132–136. 50. Nelson KE, Ahmed F, Ness P, Donahue JG. The incidence of post-trans-
36. Dodd RY, Stramer SL. Indeterminate results in blood donor testing: What fusion hepatitis: Reply. NEJM 1993;328:1280–1281.
you don’t know can hurt you. Transfus Med Rev 2000;14:151–160. 51. Schreiber GB, Busch MP, Kleinman SH, et al. The risk of transfusion-
37. Farci P, Shimoda A, Coiana A, et al. The outcome of acute hepatitis C pre- transmitted viral infections. NEJM 1996;336:1685–1690.
dicted by the evolution of the viral quasispecies. Science 2000;288: 339–344. 52. Dodd RY, Notari EP, Stramer SL. Current prevalence and incidence of infec-
38. Alter MJ, Margolis HS, Krawczynski K, et al. The natural history of community- tious disease markers and estimated window-period risk in the American
acquired hepatitis C in the United States. NEJM 1992;327: 1899–1905. Red Cross blood donor population. Transfusion 2002;42:975–979.
39. Gretch DR, Dela Rosa C, Carithers RL Jr, et al. Assessment of hepatitis 53. Zou S, Notari EP, Stramer SL, et al. Patterns of age- and sex-specific
C viremia using molecular amplification technologies: correlations and prevalence of major blood-borne infections in United States blood
clinical implications. Ann Intern Med 1995;123:321–329. donors, 1995 to 2002: American Red Cross blood donor study. Transfu-
40. Lunel F, Mariotti M, Cresta P, et al. Comparative study of conventional sion 2004;44:1640–1647.
and novel strategies for the detection of hepatitis C virus RNA in serum: 54. Schüttler CG, Caspari G, Jursch CA, Willems WR, Schaefer S. Hepati-
Amplicor, branched-DNA, NASBA and in-house PCR. J Virol Methods tis C virus transmission by a blood donation negative in nucleic acid
1995;54:159–171. amplification tests for viral RNA. Lancet 2000;355:41–42.
41. Hawkins A, Davidson F, Simmonds P. Comparison of plasma virus loads 55. Busch MP, Glynn SA, Stramer SL, et al. A new strategy for estimating
among individuals infected with hepatitis C virus (HCV) genotypes 1, risks of transfusion-transmitted viral infections based on rates of detec-
2, and 3 by Quantiplex HCV RNA assay versions 1 and 2, Roche moni- tion of recently infected donors. Transfusion 2005;45:254–264.
tor assay, and an in-house limiting dilution method. J Clin Microbiol 56. Long A, Spurll G, Demers H, Goldman M. Targeted hepatitis C look-
1997;35:187–192. back: Quebec, Canada. Transfusion 1999;39:194–200.
42. Damen M, Sillekens P, Sjerps M, et al. Stability of hepatitis C virus RNA 57. Christensen PB, Groenboek K, Krarup HB, Danish HVL. Transfusion-
during specimen handling and storage prior to NASBA amplification. acquired hepatitis C: the Danish lookback experience. Transfusion
J Virol Methods 1998;72:175–184. 1999;39:188–193.
43. Sangiovanni A, Morales R, Spinzi GC, et al. Interferon alfa treatment of 58. Culver DH, Alter MJ, Mullan RJ, Margolis HS. Evaluation of the effec-
HCV RNA carriers with persistently normal transaminase levels: a pilot tiveness of targeted lookback for HCV infection in the United States—
randomized controlled study. Hepatology 1998;27:853–856. interim results. Transfusion 2000;40:1176–1181.
44
599
Chapter 45
HIV, HTLV, and Other Retroviruses
Eberhard W. Fiebig ● Edward L. Murphy ● Michael P. Busch
This chapter provides a general overview of retroviruses, with endogenous retroviral elements (as contrasted with vertically
emphasis on those aspects of human retrovirus epidemiol- transmitted exogenous retroviruses) are present in many
ogy, pathophysiology, and detection of greatest relevance to species, including humans, and in some species account for
specialists in transfusion medicine. The four major known up to 10% of total genomic DNA.2
pathogenic human retroviruses—human immunodeficiency Disease manifestations of retroviruses are highly vari-
virus types 1 and 2 (HIV-1 and HIV-2) and human T-cell able. Many animal species harbor exogenous or endogenous
lymphotropic virus types 1 and 2 (HTLV-1 and HTLV-2) are retroviruses that appear to be benign and may in fact be
considered in detail. Strategies to further reduce the risk of beneficial in restricting infection by related pathogenic ret-
retrovirus transfusion transmission are addressed. roviruses. On the other hand, retroviruses became the focus
of intense research in the 1960s and 1970s because of their
capacity to induce malignancies in a wide range of species.
DEFINITION, LIFE CYCLE, AND The demonstration that certain retroviruses (termed acute
DISTRIBUTION OF RETROVIRUSES RNA tumor viruses) rapidly transform target cells in vitro
and induce tumors within days to weeks of inoculation into
Retroviruses were among the first viruses described in the animals greatly facilitated experimental investigation of viral
scientific literature. They constitute a major class of mem- carcinogenesis.4 Studies of the molecular differences between
brane-coated, diploid, single-stranded RNA viruses with these acute retroviruses and genetically related slow viruses
wide distribution in nature; examples exist in genera rang- (which failed to transform cells in vitro and only occasion-
ing from insects to reptiles to virtually all mammals.1 The ally caused tumors many months after inoculation) led to
III human retroviruses HIV and HTLV belong to the lentivirus the discovery of viral oncogenes, which were responsible for
and oncornavirus groups of the retrovirus family, respec- tumorigenesis. This achievement was followed by the revo-
600 tively.2 Characteristic features of retroviruses are a distinct lutionary discovery that these viral oncogenes had in fact
genomic organization, the presence of viral particle (virion)- arisen by recombination events between slow viruses and key
associated reverse transcriptase, and a unique replication cellular genes termed proto-oncogenes.4 Thus, investigation
cycle (Fig. 45–1). of retrovirus-induced cancers in animals led to unparalleled
The first step of infection is attachment of virus particles insights into normal cell biology and disease pathogenesis
to the cell membrane. In the case of HIV, which has tropism in humans. Retroviruses have also attracted interest from
for T cells and macrophages, virus glycoprotein 120 (gp120) researchers who explore nonpathogenic strains as vehicles
attaches to CD4 molecules expressed on the surface of these for therapeutic gene transfer.
cells. Efficient infection also requires engagement of viral
proteins with chemokine coreceptors, identified as CCR5 on
macrophages and CXCR4 on T cells.3 After entry into a host DISCOVERY OF HUMAN RETROVIRUSES
cell, typically by fusion of the virion and host cell membranes,
the reverse transcriptase enzyme copies viral RNA into com- The first report of successful isolation and characteriza-
plementary double-stranded DNA (cDNA). Virion-associ- tion of a bona fide human retrovirus (later termed HTLV-1)
ated integrase then mediates integration of this cDNA into occurred in 1980 and involved a patient with a rare type
random sites in the host cell’s genome, forming integrated of leukemia (now called adult T-cell leukemia/lymphoma
viral cDNAs termed proviruses. Subsequent transcription, [ATL]).5 Subsequently, HTLV-1 was also identified as the
processing, and translation of viral genes are mediated prin- cause of a rare neurologic condition known as HTLV-asso-
cipally by host cell enzymes, although both viral and host cell ciated myelopathy/tropical spastic paraparesis (HAM/TSP).6
regulatory gene products influence the level and pattern of A second, closely related human retrovirus (HTLV-2) was
viral gene expression and replication. The classic retrovirus isolated in 1982 from a patient with a somewhat more com-
life cycle is completed when nascent particles associate and mon type of leukemia (hairy cell leukemia)7; further surveys
bud from the plasma membrane to form progeny virions, failed to show a relationship between HTLV-2 and hairy cell
which can then infect other cells and other organisms. leukemia, but established the virus as a causative agent of
Retroviruses can also spread horizontally, by fusion of HAM/TSP.6,8 Both HTLV-1 and HTLV-2 are thought to be
infected and uninfected cells, or vertically, by replication of derived from simian T-lymphotropic retroviruses transmit-
integrated viral DNA along with cellular DNA during mitosis ted to humans over the past hundreds to thousands of years.
or meiosis. Indeed, integrated proviruses have evolved that The full pathogenic potential of human retroviruses
are passed congenitally through the germline; these so-called was not realized until HIV was established as the cause of
HIV, HTLV, AND OTHER RETROVIRUSES
Figure 45–1 Replication cycle of
human immunodeficiency virus type 1.
(a) After attachment of virus particles to
the CD4 receptor molecule, virus enters
the cell by a pH-dependent mecha- Mature HIV
nism and/or endocytosis. Not shown (h) Budding
(a) Attachment Membrane
is the required interaction of viral pro-
teins and chemokine co-receptors. (b) Cytoplasm (g) Virus assembly
The outer lipid envelope of the virus is
removed when the particle undergoes Entry and
fusion with cytoplasmic vacuoles. (c) The Genomic RNA
(b) partial
core particle that remains is the site for
uncoating
reverse transcription of the virion RNA
into DNA. (d) After translocation into Regulatory proteins Structural proteins Maturation proteins
the nucleus, integration into the DNA of (tat, rev, nef) (gag, pol, env, (proteinase,
the cell occurs. (e) The integrated pro- vpx?, vit?) vpu?, vpr?)
virus genome is transcribed by cellular RNA
RNA polymerase II. (f) Translation of viral gag proteins
messenger mRNA produces regulatory (c) Reverse
transcription Viral mRNA-
proteins, which stimulate synthesis of Ribosome
maturation proteins and the structural Viral DNA Viral protein
proteins of the virion. (g) Accumulation (f) Translation
of structural proteins in the cell mem-
brane permits the assembly of virus (d) Integration
particles. (h) Maturation and release Provirus genome Host DNA
from the cell by budding. (From Mayer
A, Busch MP. Human immune defi- LTR LTR
ciency viruses. In Anderson KC, Ness P (e) Transcription
(eds). The Scientific Basis of Transfusion
Medicine: Implications for Clinical Practice. Nucleus
Philadelphia, WB Saunders, 1994.) RNA export
AIDS in 1984. It is now thought that the virus originated in during reverse transcription, which has been attributed to
nonhuman primate species and was introduced into native the negligible proofreading exonuclease activity of the reverse
people of the central African rain forest.9 There may have transcriptase.12 Nucleotide misincorporations may occur at
been multiple introductions of the virus into humans dating an astonishingly high rate of 5 to 10 per HIV genome per
back several centuries (based on phylogenetic molecular clock replication cycle, a phenomenon known as hypermutation. 45
analyses); so far, however, the earliest documented evidence Insertions, deletions, and intergenic recombination add to
of HIV infection in humans comes from a blood sample col- the trend toward genetic diversification.13 Within an HIV-1– 601
lected in 1959.10 Sub-Saharan Africa remains the area worst infected individual, a swarm of “quasispecies,” or HIV vari-
affected by the epidemic today, with an estimated 60% of the ants, develops over time,14 and dual and multiple infections
global caseload.11 According to World Health Organization with recombination between the infecting HIV-1 strains
figures, 4.1 million people were newly infected in 2005, rais- have been reported.15,16
ing the projected number of infected people to a staggering On the basis of relatedness of genomic sequences, the
38.6 million worldwide.11 HIV-1 family is divided into main (M), outlier (O), and non-M,
non-O (N) groups, with 11 distinct subtypes or clades (A
through K) recognized within group M.17 Of note is the cur-
HUMAN IMMUNODEFICIENCY VIRUS rently almost exclusive prevalence of clade B strains in the
United States and, to a lesser degree, Europe; other clades are
more predominant in South America and Asia. The greatest
Genomic Organization and Virion Structure
genetic diversity of HIV-1 strains is found in central Africa,
The organization of the HIV RNA genome is shown in Figure in keeping with this area’s role as the presumed point of ori-
45–2. The products of the gag, pol, and env genes that are gin of the pandemic. The worldwide distribution of HIV-1
shared among all retroviruses give rise to the structural ele- and -2 subgroups and clades is depicted in Figure 45–4. A
ments of the virion, which is shown in Figure 45–3. Similar recent survey of 292 U.S. blood donors detected by anti-HIV-
to other lentiviruses, HIV contains at least six additional 1/HIV-2 screening in the late 1990s identified only 7 (2.3%)
genes, which encode for regulatory proteins and virulence non-B subtypes, which nonetheless reflects a trend toward
factors that appear to play an important role in pathogenic- increasing diversity relative to earlier periods (e.g., 3 [0.8%]
ity. Detection of specific antibodies against viral proteins or of 383 donors from 1993 through 1996).18 This is a concern
identification of viral nucleic acid sequences in infected hosts because the seroconversion window period from beginning
provides the basis for screening and supplemental assays used of infectiousness to detectability of anti-HIV is prolonged
in blood donor eligibility testing. with non-B strains when testing is performed with assays in
routine use for blood donor screening in the United States,
which are optimized for subtype B detection.18
HIV Diversity: HIV-1 Subtypes and HIV-2
Group O viral strains are most common in Cameroon and
Extensive genetic diversity is a hallmark of HIV and other surrounding West African countries, where an estimated 1%
lentiviruses. A contributing factor is the high error rate to 2% of HIV infections are caused by these viral strains.19,20
TRANSFUSION MEDICINE Enhances viral
infectivity at level
Viral envelope of reverse
Promotes viral Causes CD4 proteins mediating transcription;
infectivity at level degradation in ER; CD4 and chemokine downregulates
of reverse required for efficient receptor binding and surface CD4
transcription virion budding membrane fusion and MHC I
Nucleocapsid
core proteins nef
U3 R U5 gag vif vpu env
pol
5' LTR 3' LTR
tat
rev
vpr
gp120env(SU) Reverse
surveillance testing for HIV-1 divergent strains did not
transcriptase detect any group O viral isolates among 1072 serum samples
III gp41env(TM) (RT) from high- and low-risk population groups in the United
States and Puerto Rico.21 Nonetheless, concern arose in the
p6gag(NC) mid-1990s when studies demonstrated that some group O
602 Integrase (IN)
Vpr isolates were not reliably detected by a number of HIV-1 and
HIV-1/HIV-2 combination assays,22 including some that are
Lipid bilayer Protease (PR) used for blood donor screening.23 Antibody assays employ-
ing synthetic peptides or recombinant antigens on the solid
Single- phase, and those using the so-called third-generation anti-
stranded gen-sandwich format, were especially prone to false-negative
Host proteins HIV-1 RNA results. Since then, test manufacturers have moved quickly to
p24gag(CA)
enhance their assays’ sensitivity to unusual variants such as
p7gag(NC) p17gag(MA)
subtype O. As an added precaution, the U.S. Food and Drug
Figure 45–3 Schematic of the HIV-1 virion. Each of the virion pro- Administration (FDA) continues to recommend permanent
teins making up the envelope (gp120env and gp41env) and inner deferral of blood and plasma donors who were born, resided,
core (p24gag, p17gag, p7gag, and p6gag) is identified. In addition,
the diploid RNA genome is shown associated with reverse transcrip- or traveled in West Africa since 1977 or had sexual contact
tase (RT), an RNA- and DNA-dependent DNA polymerase. Integrase with someone identified by these criteria.24
(IN) and protease (PR) are also found in the mature HIV-1 virion. The HIV-2, discovered in 1985 in several countries in West
auxiliary protein Vpr is incorporated into the HIV-1 virion through an Africa,25 was initially called HTLV-4 and lymphadenopathy-
interaction with the p6gag protein, which comprises the carboxyl ter-
minus of the p55gag precursor protein. CA, capsid protein; MA, matrix
associated virus type 2. Although still primarily concentrated
protein; NC, nucleocapsid protein; SU, surface protein; TM, transmem- in West Africa, HIV-2 has now spread throughout Western
brane protein. (Reproduced with permission from Geleziunas R, Greene Europe, where a substantial number of infected blood donors
WC. Molecular insights into HIV-1 infection and pathogenesis. In Sande have been detected, and cases of transfusion transmission
MA, Volberding PA (eds). The Medical Management of AIDS, 6th ed. have been documented.26 HIV-2 is transmitted in the same
Philadelphia, WB Saunders, 1999, p 25.)
manner as HIV-1 (i.e., by sexual contact, intravenous drug
use, and, at a lower rate than for HIV-l, from mother to child)
and causes progressive immunodeficiency, with susceptibility
Outside this geographic area, group O isolates have rarely to an array of opportunistic infections similar to those seen
been seen. The current risk of HIV-1 group O infection in with HIV-1. Rates of disease progression and secondary viral
the United States is very low. Only two such infections have transmission appear to be lower, however, in persons infected
been reported; both involved immigrants from West Africa with HIV-2 than in those infected with HIV-1, possibly owing
who had never donated blood or plasma. Furthermore, to lower viral burden in HIV-2 infection.27,28
HIV, HTLV, AND OTHER RETROVIRUSES
Figure 45–4 Distribution of human
immunodeficiency virus (HIV) through-
out the world according to the preva-
lence of its types and subtypes (clades).
The distribution of HIV-1 clades is indi-
cated by letters A to I (more prevalent B,a,c,d,g,h
in capital letters). Clades of HIV-2 are B,a,c,d,f,o
not shown separated. (Reproduced B,a,d,e,o HIV-2
with permission from Diaz RS, Busch HIV-2 B,i B,e
MP. Human immunodeficiency viruses. B,e,o
In Anderson KC, Ness PM (eds).
C,a,b,e
Scientific Basis of Transfusion Medicine: B
A,O,d,g,h HIV-2 C,E,b
Implications for Clinical Practice, 2nd
ed. Philadelphia, WB Saunders, 2000, HIV-2 C
b
p 508.) A,D,e E,b
B
B,f,c
C,a,d
B
B,e
HIV-l and HIV-2 are highly (>50%) homologous at the proliferating virus. In most cases signs and symptoms of acute
nucleic acid sequence level, and they crossreact immunologi- HIV disease subside within weeks to a few months, giving way
cally to a great extent (particularly the core and polymerase to a clinically silent period that can last several years. The lack
antigens). For this reason, up to 90% of sera from HIV-2– of clinical symptoms during this stage of the disease is decep-
infected persons have been found to test positive with FDA- tive, however, because viremia persists in untreated infection
licensed anti-HIV-1 assays, with variable reactivity on HIV-1 and the number of CD4+ lymphocytes, the primary target of
Western blot analysis.29,30 This crossreactivity undoubtedly HIV, is gradually declining.40 Paralleling the loss of CD4+ lym-
prevented transfusion transmission of HIV-2 by blood phocytes are clinical disease manifestations, such as opportu-
screened for anti-HIV-1 in areas where the type 2 virus nistic infections, numerous conditions related to suppression
was present before implementation of combination HIV-1/ and dysfunction of the immune system, direct viral effects on
HIV-2 screening tests. This high-level crossreactivity has also multiple organ systems, and ultimately death after a median
facilitated surveillance for HIV-2 in regions where it was rare survival of 8 to 10 years from exposure. Although it appeared 45
or absent.31,32 From 1987, when the first case of HIV-2 infec- initially that disease manifestations and clinical course of
tion was diagnosed in the United States, to 1998, a total of 79 transfusion-transmitted HIV infection follows a more rapid 603
persons with HIV-2 infections were documented, of which course than HIV infection due to other routes of transmission,
approximately two thirds had been born in West Africa.33 subsequent analysis did not confirm this impression.40
Combination HIV-l/HIV-2 assays were developed in the Also not confirmed were concerns that allogeneic trans-
late 1980s,34–36 and mandatory implementation of either a fusions—via immune stimulation by donor leukocytes—
combination test or a separate anti-HIV-2 test in the donor may result in accelerated disease progression and shortened
screening setting was required by the FDA effective June 1, survival in HIV-infected transfusion recipients. The Viral
1992.37 From the implementation of HIV-2 screening in 1992 Activation by Transfusion Study (VATS), a large multicenter
through 1996, three prospective U.S. blood donors were found U.S. clinical trial, specifically addressed this issue and found
to be HIV-2 positive at the time of attempted donation.38 One no evidence of significant activation of HIV replication or
was a U.S.-born woman without identifiable risks for HIV accelerated disease progression in HIV-1–infected recipients
infection; the other two were men, born in France and Liberia, of either leukoreduced or nonleukoreduced transfusions.41
West Africa respectively, who had resided for years in West Widespread availability of potent antiretroviral therapy in
Africa. These three cases of HIV-2 infection were detected western countries since the late 1990s has changed the course
from screening of more than 50 million whole blood dona- of HIV disease to that of a manageable chronic condition
tions, indicating that the prevalence of HIV-2 is very low (less with reduction in opportunistic infections and other HIV-
than 1 in 15 million screened donations). A more recent study related complications and has markedly prolonged survival.
identified a single HIV-2–infected donor among 7.2 million Unfortunately, current treatment regimens are not capable
donations at 18 U.S. blood centers from 1997 through mid- of eradicating HIV from cell- and tissue-based sanctuaries
2000.18 As of 2001, no cases of HIV-2–infected transfusion throughout the body, and therapy-related side effects and
recipients had been reported in the United States. development of resistant viral strains add to the problems of
managing HIV-infected patients.42
Clinical HIV-1 Disease
Efficiency of HIV Transmission by Blood
In approximately 60% of acute HIV infections newly infected and Blood Products and Transmission Risk
persons experience a nonspecific flulike illness, usually within
Prior to Blood Donor Screening
2 to 4 weeks of exposure.39 This retroviral syndrome coincides
with appearance of antibody to HIV antigens and is thought Human immunodeficiency virus is sensitive to drying43 but
to represent a reaction of the immune system to the rapidly survives refrigeration of blood and freezing of plasma. Virus
TRANSFUSION MEDICINE present in the bloodstream of an undetected infected donor 2
is therefore readily passed on by transfusion. Data from the
Transfusion Safety Study (TSS), which traced and enrolled
First TA-AIDS case reported:
recipients of retrospectively identified HIV-1–seropositive
Ab
RNA
LS-Ab
LS-Ab OD cutoff
p24 Ag
0 I II III IV V VI
ecl. Recent infection Early chronic infection
45
Investigators have frequently used calculations based The result suggests a rather modest decrease in risk, from 607
on the incidence-window period model to project the risk approximately 1 per 2 million component units transfused
of transmission of HIV and other viruses by transfusion with MP-NAT, to 1 in 3 to 4 million units with ID-NAT,82 at
in different geographic areas and at various periods of the considerably higher cost and potentially delayed availability
epidemic. Experience over time has confirmed the valid- of transfusable cellular blood components. Not addressed by
ity of the approach, and the risk projections derived from the novel strategy of estimating residual transmission risk of
the model are generally accepted as meaningful and reliable HIV based on viral replication dynamics is the finding of
estimates of the true risk.94,95 Decreasing projections of the low-level (estimated 1 to 10 copies/mL) viral “blips” in the
magnitude of risk in the same population over time reflect 2- to 3-week long period between infection and ramp-up
removal of high-risk donors and improvements in blood phase of viremia, sometimes referred to as the eclipse phase
donor screening. Recently, widespread use of NAT assays in of HIV infection (see Fig. 45–6). Such “blips” have been
blood donor screening led to application of a new strategy observed at intervals of approximately 1 to 3 weeks before
for estimating risk of HIV transmission based on rates of HIV ramp-up viremia in 6 of 15 plasma donors with newly
detection of recently infected donors.72 The method takes acquired HIV infection.96 The added risk from blood dona-
into account the expected increase of viral genomic equiva- tions during the eclipse phase is unknown, but might con-
lents (viral load) in blood donor samples during the initial tribute slightly to the overall risk estimate associated with
phase of viral replication before appearance of HIV anti- current NAT blood donor screening.96
bodies.72,82 By back-extrapolation of the viral load from a
reference point such as appearance of p24 antigen or detec-
tion of HIV infection by minipool NAT to a minimal viral HUMAN T-CELL LYMPHOTROPIC VIRUS
concentration that is thought to be infective (e.g., 1 viral
copy per 20 mL plasma), the window period between the Although eclipsed by concerns about HIV, HTLV is none-
infectious threshold and the reference point can be pro- theless relevant to the safety of blood transfusion. Both
jected. By substituting the reference point with a theoretical HTLV-1 and HTLV-2 are transmitted by blood transfusion,
threshold such as assay sensitivity of a new NAT, the window cause chronic retroviral infection of humans, and are asso-
period for that assay can be estimated. A practical applica- ciated with serious disease outcomes. Serologic testing for
tion of the new method is the projection of risk reduction HTLV-1 in U.S. blood donors has been in place since 1988.
for transfusion transmission of HIV by testing individual This policy has led to the unexpected discovery that at least
donor samples rather than minipools of 16 to 24 samples. half of blood donors testing positive for HTLV-1 are in fact
TRANSFUSION MEDICINE infected with HTLV-2, for which disease outcomes are less Epidemiology and Modes of Transmission
well described.97–99
HTLV-1 is endemic in sub-Saharan Africa; the Caribbean
region and parts of South America, including Colombia, Peru,
Virology and Brazil as a result of the slave trade; southwestern Japan;
Both HTLV-1 and HTLV-2 are human retroviruses of the and parts of Melanesia and Australia.106 HTLV-1 has also been
oncornavirus class. As previously mentioned, the first reported from Iran, India, and Taiwan as well as in countries
report of HTLV-1 discovery was published in 1980 and containing significant immigrant populations from the main
that of HTLV-2 in 1982.5,7 With RNA genomes of approxi- endemic areas. Given the high HTLV-1 prevalence in south-
mately 8000 nucleotides in length, the genetic organiza- western Japan, the absence of HTLV-1 infection in mainland
tions of HTLV-1 and HTLV-2 are similar to those of other China and in Korea as well as other parts of Southeast Asia
retroviruses (Fig. 45–7).100,101 The HTLV genome contains poses an epidemiologic puzzle. Molecular epidemiologic stud-
gag regions coding for viral core proteins, pol regions ies have indicated the presence of at least two viral subtypes
coding for viral reverse transcriptase, protease and inte- in Africa; a cosmopolitan subtype found in Africa, Japan, the
grase env regions coding for the viral envelope proteins, Caribbean, and other more recent foci of HTLV-1 infection;
and finally the tax or px region (analogous to the HIV and a distantly related Melanesian subtype.107
tat gene), which is responsible for transcriptional regula- HTLV-2 has also been found to be endemic in a num-
tion of HTLV-1 and HTLV-2. HTLV-1 and HTLV-2 have ber of Amerindian tribes throughout South, Central, and
approximately 60% nucleotide sequence homology. Many North America.108–110 Not all tribes are infected, and tribes
of their viral proteins crossreact serologically, although with the least intermingling with Western settlers, such
peptides eliciting specific immune responses and allowing as the Kayapo of Brazil, appear to have the highest preva-
differential serologic diagnosis of HTLV-1 versus HTLV-2 lence rate.108 The other endemic focus of HTLV-2 infection
have been discovered. appears to be among Pygmies in sub-Saharan Africa, again
HTLV-1 predominantly infects CD4+ T lymphocytes, in tribes having relatively little contact with European set-
whereas HTLV-2 has a broader tropism with preference for tlers.111 HTLV-2 infection is also prevalent among injection
CD8+ lymphocytes and, to a lesser degree, CD4+ lympho- drug users and their sexual partners in the United States,
cytes, B lymphocytes, and macrophages.102,103 Neither virus Brazil, and Europe.112 Given the relative recency of injection
has high levels of cell-free viremia, perhaps accounting for drug use behavior, there appears to have been an epidemic
lower transmissibility than either hepatitis B virus or HIV. spread of HTLV-2 over the past 40 or 50 years as a result of
Also in contrast to HIV, there is relatively little active replica- the introduction of endemic HTLV-2 into the injection drug
tion of HTLV-1 and HTLV-2 in infected humans. Instead, use population and subsequent spread through sharing of
most expansion of the pool of infected lymphocytes appears contaminated needles.113
to occur by lymphocytic division and the proliferation In endemic populations, both HTLV-1 and HTLV-2 show
III of clones of infected lymphocytes.104 Corollaries of these age-specific seroprevalence rates that rise steadily with age,
observations are that descriptions of HTLV viral load refer from 1% or less among infected children to 5% to 10% of older
to measurements of proviral DNA in the cellular compart- individuals.113–116 Female seroprevalence is generally greater
608 than male seroprevalence, presumably owing to more efficient
ment. There has been only one report of measurable HTLV
RNA in cell-free plasma105; however, further research will be sexual transmission from males to females than vice versa.
necessary to exclude the possibility of a significant viremia In one large study of blood donors in the United States in
in cell-free plasma. the 1990s, HTLV-1 seroprevalence was approximately 10 per
0 1000 2000 3000 4000 5000 6000 7000 8000 9032 Figure 45–7 Structure and organization of
the HTLV genome. The HTLV provirus genome
Non-translated is shown at the top of the figure. Positions of
U3 R U5 region U3 R U5 known genes are indicated. Sizes and posi-
gag env tions of the proteins encoded by the provirus
are shown beneath the provirus. The struc-
pro ture of the three messenger RNA species pro-
gp46(SU) p21(TM)
p2 A)
5( )
)
pol
M
N
4(
env An
tax, rex An
HTLV-I/II CRS
readily transmitted than HTLV-2 may not be true.120 The
45
617
Chapter 46
Human Herpesvirus Infections
John D. Roback
EGFR were resistant to CMV infection.19 Because hemato- from infected cells by exocytosis,31 eventually resulting in
poietic target cells do not express EGFR, other unidentified host cell destruction.
docking receptors for gB are probably involved. gB also acti- Cytomegalovirus may also assume a latent state when it
vates Toll-like receptors, conserved cellular membrane pro- infects target cells that are not permissive for viral replica-
teins involved in initiating antipathogen responses.20 tion. Latency, the presence of viral DNA in an infected cell in
Although attachment complexes are initially dissociable, the absence of active viral replication, may persist indefinitely
they can be stabilized by membrane fusion. After attach- because the host cell is not destroyed by the virus. The latent
ment, interactions between gB and its receptors may ini- CMV genome retains the capacity to reactivate viral gene
tiate fusion. In addition, the viral protein complex gp86 expression, produce infectious virions, and enter lytic growth
(gcIII, UL75:UL115, or gH:gL) appears to be involved in at a later time. Studies with human and murine CMV have
these processes.21 Evidence indicates that gp86 is composed demonstrated that latency can be established in hematopoietic
of the viral gH and gL proteins, as well as a third compo- cells, primarily those of the granulocyte-monocyte lineage, as
nent not yet identified.22 The cell membrane receptor for well as in endothelial cells.32–38 The possibility of CMV latency
gp86 has been identified as a 92.5-kD constitutively phos- in other cell types has not been excluded. The molecular mech-
phorylated glycoprotein.23,24 Binding of gB and gp86 to the anisms regulating CMV latency and reactivation from latency
target cell initiates signaling processes that may be impor- have not been completely elucidated. In the herpesvirus EBV,
tant to CMV infection. Interactions between gB and its cel- a clearly defined set of viral proteins have been identified that
lular receptor(s) activates signal transduction through the control latency and reactivation.39 Similarly, CMV latency-
interferon-response pathway, leading to induction of the associated transcripts (LATs) have been detected in 0.01% to
46
interferon-responsive genes OAS and ISG54,25 and binding 0.001% of sorted CD33+ lineage-committed hematopoietic
of gp86 to the cellular 92.5-kD protein can alter intracellular progenitors from the peripheral blood of naturally infected
calcium concentration.26 individuals.36,40 CMV LATs are transcribed from the viral IE 619
After fusion, the viral capsid penetrates into the host cell gene locus and encode immunogenic viral proteins that are
and releases viral DNA. The molecules mediating these pro- targets of naturally arising antibodies in CMV-seropositive
cesses have not been identified. Additional proteins have also individuals.32,40,41 The function of LATs is currently unknown.
been hypothesized to mediate viral infection. For example, Circularization of the CMV genome is associated with latency
cellular human leukocyte antigen (HLA) class I proteins in CD14+ peripheral blood monocytes in CMV-seropositive
may promote viral attachment through interactions with β2- individuals,6 a phenomenon also seen during latent infections
microglobulin attached to the virion membrane.27 with other herpesviruses.42–44 The possibility of persistent viral
infections, an intermediate state between active and latent
VIRAL LIFE CYCLE infection in which low levels of virus are produced, remains a
topic of considerable debate.45,46
Active and Latent Infections After the steps of viral
attachment, fusion, and penetration, active infection occurs Viral Genes Important to Pathogenesis In addition to
if the target cell is permissive for the complete sequence of viral proteins that control expression of viral genes and virion
viral gene expression, viral genome replication, and produc- assembly and provide structural support for the viral par-
tion of progeny virions. During active infection, viral genes ticle, CMV also encodes proteins that favor viral replication
are expressed in coordinated waves. Three distinct kinetic at the expense of host cell metabolism and disrupt the host’s
classes of viral genes have been identified: immediate early ability to combat viral infection. For example, CMV infec-
(α), followed by delayed early (β) and then late (γ).4 α-Class tion alters the expression, accumulation, and activity of the
gene transcription is controlled by a combination of consti- cellular tumor suppressor proteins, cyclins, and cyclin-asso-
tutively expressed host cell proteins and viral proteins pres- ciated kinases. These alterations in the cellcycle machinery
ent in the infecting virion. Thus, α genes can be transcribed act to simultaneously promote progression toward the G1/S
in the presence of pharmacologic inhibitors of protein transition but prevent cellular DNA synthesis and cell divi-
synthesis. Viral α proteins, in turn, are required for expres- sion, resulting in cell-cycle arrest and cellular aneuploidy. It
sion of viral genes of the β and γ classes.28–30 The β pro- has been hypothesized that in the arrested state cellular DNA
tein products perform viral DNA replication and metabolic synthesis would be blocked but the cellular milieu would
functions; the γ genes encode structural proteins required contain abundant nucleotides and other metabolic pre-
for assembly of progeny virions. Finally, mature virions are cursors that could support viral replication.47–50 One viral
transported through the Golgi apparatus and are released protein involved in this process is the immediate-early IE1
TRANSFUSION MEDICINE 72-kDa protein, which complexes with the cellular Rb-related including saliva, tears, breast milk, urine, stool, and semen.
protein p107 and blocks its ability to repress E2F-responsive Community-acquired CMV infection is usually the result of
promoters.51 The IE1-mediated derepression at the level of close contact with a person shedding CMV. The incidence of
E2F, in turn, allows expression of cellular genes that promote community-acquired CMV infection varies with the study
cell-cycle progression.51 CMV infection also activates cyclin- population. For example, the yearly CMV seroconversion
dependent kinase 2 (CDK2), a cellular protein that controls rate in health care workers has been estimated at 0.6% to
progression through the G1 and S phases of the cell cycle. 3.3%,65 similar to rates of 2.0% to 6.3% reported in middle-
The importance of CDK2 to viral replication was illustrated class women during and between pregnancies.66 In contrast,
by blocking CDK2 activity with either a dominant-negative rates as high as 13% per year have been observed in ado-
mutant or the pharmacologic inhibitor roscovitine. In both lescents.67 In blood donors, the CMV seroconversion rate is
cases, inhibition of CDK2 activity prevented CMV replication estimated at approximately 1% per year.68 Most studies have
and production of progeny virus.50 shown that 50% to 80% of the population is CMV seroposi-
Cytomegalovirus has also developed mechanisms to inter- tive,4 although the incidence can be higher in some urban
fere with antiviral immune function (reviewed in Hengel populations and lower in some groups of blood donors.69
and colleagues52 and Reddehase53). The UL37 ORF of CMV Most individuals contracting community-acquired CMV
encodes vMIA, an anti-apoptotic protein that localizes to the infection are immunocompetent, and the infection is often
mitochondria and protects infected cells from immune-medi- asymptomatic. However, a mild self-limited infectious mono-
ated apoptosis by blocking the effects of Fas, tumor necrosis nucleosis syndrome can occur, with symptoms that include
factor receptor-1 (TNFR-1), and granzyme B.54 Monocytes fever, malaise, hepatosplenomegaly, and a rash.70 CMV can
are a prominent site of CMV latency, and monocyte-derived be isolated from bodily secretions during the symptomatic
macrophages can support active CMV replication. During phase. The infected individual mounts both a humoral and
differentiation to macrophages, CMV in monocytes displays cell-mediated immune response and viral symptoms rapidly
delayed replication kinetics and viral particles are retained in resolve, leading to a complete recovery. However, despite
the Golgi apparatus, which may facilitate immune evasion effective control of CMV infection by the competent host
until sufficient progeny virions have been produced.55 In immune system, the virus is not completely eliminated but
contrast, in patients with compromised antiviral immunity, instead becomes latent.
CMV can replicate with rapid kinetics, displaying viral doubling Transplacental transmission of CMV to a developing
times approaching 24 hours.56 fetus is an important viral cause of birth defects.71,72 Fetal
Despite sophisticated viral mechanisms of immune eva- infection occurs in 40% to 50% of cases in which a seronega-
sion, clinical and experimental evidence demonstrates that tive mother contracts a primary CMV infection during preg-
the competent immune system can effectively suppress viral nancy.73,74 CMV disease occurs in 5% to 15% of the infected
replication. For example, the murine CMV gp40 and gp48 gly- infants, presenting most often with intrauterine growth
coproteins (ORFs m152 and m06, respectively) can decrease retardation, deafness, mental retardation, blindness, and
III expression of cellular class I major histocompatibility com- thrombocytopenic bleeding.72,73 However, when mothers are
plex (MHC) proteins during infection of fibroblasts and thus seropositive before pregnancy, maternal antiviral immunity
620 decrease CMV antigen presentation to CD8+ T cells. However, can limit congenital CMV infection and disease. For exam-
the significance of these mechanisms during natural infection ple, in one study of seropositive mothers the rate of vertical
are unclear because CMV infection does not disrupt class I transmission was approximately 1%.73 Furthermore, no cases
MHC expression and antigen presentation in macrophages, the of symptomatic CMV infections were seen in 27 congenitally
professional antigen-presenting cell most important in initiat- infected infants born to seropositive mothers.73
ing the anti-CMV immune response.57 Furthermore, although Cytomegalovirus can also be transmitted by blood trans-
human CMV has also evolved mechanisms to interfere with fusion or transplantation of hematopoietic stem cells and
antigen presentation by infected cells,58,59 the immune system solid organs from infected donors. When the recipients are
circumvents these obstacles by utilizing structural proteins immunocompromised, CMV transmission through these
in the infecting viral particle as immunodominant epitopes mechanisms can produce serious clinical consequences.
for an immune response.60,61 Thus, the immune response Prevention of CMV infection is an important concern in
can be initiated before expression of antiviral proteins that transfusion medicine.
halt antigen presentation by the infected cell. CMV has also
developed strategies to interfere with interferon-γ (IFN- γ) sig- CMV Infection of Peripheral Blood
nals that normally upregulate MHC expression during viral and Marrow Cells
infection.62 Interestingly, downregulation of class I MHC cell
CELL TROPISM
surface expression by CMV should lead to destruction of the
infected cells by host natural killer (NK) cells.63 However, From the perspective of transfusion medicine, the most
expression of the viral class I MHC homologue m144 by important target cells of CMV infection are peripheral blood
murine CMV decreases the susceptibility of the infected cell leukocytes and their progenitors. Under appropriate condi-
to NK-mediated lysis.64 tions, these cell types can either harbor latent CMV or allow
active viral replication, and thus are well-suited to mediate
transfusion-transmitted CMV infection.
CMV Infection, Immune Response, CMV infection of bone marrow hematopoietic progeni-
and Diagnosis tor cells likely occurs during primary infection75 (Fig. 46–1).
Most evidence suggests that these cells restrict viral repli-
Transmission, Prevalence, and Epidemiology cation but support viral latency,32,33 although some studies
During CMV infection, active viral replication results in have shown low levels of CMV replication in bone marrow–
shedding of infectious virions into plasma and bodily fluids, derived cells in culture.75,76 CMV DNA has been identified
HUMAN HERPESVIRUS INFECTIONS
Figure 46–1 Hypothetical model Allogeneic Cytokines
integrating long-term latency of cellular
cytomegalovirus (CMV) in the hema- interactions
topoietic compartment with trans-
mission of CMV by blood transfusion
and hematopoietic stem cell trans-
plantation. CMV infects CD34+ mul-
tipotent progenitors during primary
infection, and latently infected cells
retain the viral genome during self- CD34⫹
self-renewing CD33⫹ lineage-
renewal. Committed progenitors committed myeloid Tissue
in the marrow may also be directly stem cells
progenitor Monocyte macrophage
infected with CMV. Either cell type
may transmit CMV to a seronegative
transplant recipient. CMV remains
latent as CD33+ progenitors differ-
entiate into circulating peripheral
Latent CMV
blood monocytes. Latently infected
monocytes subsequently differenti-
ate into tissue macrophages, either
in the original host or after transfu- CMV CMV
sion into a recipient. The allogeneic CMV
or cytokine-mediated signals mono-
cytes encounter during differentia-
VASCULATURE CMV reactivation
tion render them permissive to CMV
reactivation and viral replication. See BONE MARROW PERIPHERAL TISSUES
text for detailed discussion.
CMV
by polymerase chain reaction (PCR) in sorted multipotent infected leukocytes in transfused blood components may
CD34+ progenitor cells from the bone marrow of seroposi- contribute to the variable incidence of transfusion-transmit-
tive, and in some cases seronegative, donors.33,36,77,78 Because ted CMV observed clinically.
of their capacity for self-renewal, latently infected hemato- Cytomegalovirus is also found associated with other cell
poietic progenitor cells represent a potential long-term reser- types in the peripheral blood and marrow. In immunocom-
voir of latent virus. In fact, when CMV-infected CD34+ cells promised patients with CMV infections, polymorphonuclear
are grown in suspension cultures, transfer of CMV DNA to leukocytes (PMNs) phagocytose and contain large amounts 46
progeny cells during mitosis has been demonstrated.34 CMV of virus.80 Although PMNs do not appear to support the com-
DNA has also been identified in myeloid-lineage-committed plete viral replication cycle, they can retain CMV in a viable 621
CD33+ progenitor cells in the marrow or mobilized into the and infectious form under experimental circumstances.81
peripheral blood by granulocyte-macrophage colony-stimu- CMV can also productively infect megakaryocytic precur-
lating factor. Lineage-committed progenitors appear to be sors and mature megakaryocytes.82 Plasma free virus appears
latently infected, as indicated by the presence of CMV LATs in to be less stable than intracellular virus, and the presence of
0.01% to 0.001% of sorted CD33+CD14+ or CD33+CD15+ free virus in plasma is usually transient.83 For example, in
bone marrow cells from seropositive donors.36 These find- one study of recently infected adolescents, only a minority
ings support a model for latency in which early hemato- (25% to 40%) had plasma viremia, which was rarely identi-
poietic progenitor cells are latently infected during primary fied more than 4 months after seroconversion.67 Based on
CMV infection and thereafter serve as viral reservoirs. the available evidence, these other peripheral blood sources
Furthermore, latently infected marrow progenitor cells are a of CMV are unlikely to be as important as latently infected
likely vector for transmission of CMV infection by hemato- monocytes to the pathogenesis of transfusion-transmitted
poietic stem cell transplantation. CMV.
As CD33+ progenitors continue to differentiate they
VIRAL LOADS
enter the peripheral blood. Monocytes appear to retain
latent virus, but as they differentiate into macrophages Quantitation of peripheral blood CMV load is clinically
CMV replication with production of progeny virus has been useful in immunocompromised patients, where viral loads
observed35,79 (see Fig. 46–1). Cells of the monocytic lineage can reach 108 copies per milliliter of plasma or greater,84 and
have been hypothesized to mediate transfusion-transmitted correlate with the severity of viral disease.85–90 For example,
CMV, but the prevalence of latently infected monocytes in in liver transplant recipients a 50% probability of CMV dis-
the peripheral blood appears to be low. It has been estimated ease was associated with a viral load of 105.1 genomes/mL
that 0.004% to 0.01% of peripheral blood mononuclear cells of blood, and a 90% probability with 105.5 genomes/mL.87
(PBMCs) mobilized by granulocyte colony-stimulating fac- Infectious virus can also be cultured readily from the blood
tor38 and 0.01% to 0.12% of PBMCs from healthy seroposi- of immunocompromised patients with active infections.91
tive blood donors35 contain CMV DNA, with a range of 2 to In contrast, the peripheral blood viral load in CMV-
13 viral genomes per infected cell.38 Because approximately infected but otherwise healthy individuals is much lower and
5% of PBMCs are monocytes, latently infected monocytes rarely quantitated. For example, in a series of published stud-
may comprise only 1 to 25 of every million peripheral blood ies CMV could be cultured from only 2 of over 1500 buffy
white blood cells (WBCs). The low numbers of latently coat samples from healthy blood donors.70,92,93 Nonetheless,
TRANSFUSION MEDICINE a consideration of CMV loads in healthy individuals who limited control over CMV infection and disease. Antibody
are potential blood donors is useful in understanding the expression is typical of other humoral responses, with tran-
biology of transfusion-transmitted CMV (Fig. 46–2). WBC- sient anti-CMV IgM synthesis followed by persistent expres-
associated viremia is often detectable for 6 months after sion of antiviral immunoglobulin G (IgG) (see Fig. 46–2). In
infection,67,94 in contrast to plasma free virus, which often a study of recently infected adolescents, anti-CMV antibod-
disappears by 4 months. In recently seroconverted adoles- ies were usually detectable by 6 to 8 weeks postinfection, a
cents, 75% to 80% of WBC samples were positive for CMV time of high peripheral blood viral loads.67 However, despite
DNA by PCR within the first 4 months of infection, com- the fact that anti-CMV, anti-gB, as well as viral neutralizing
pared with 25% to 40% of plasma samples within this same antibodies could be detected during the humoral response,
period.67 A study of 98 seroconverting blood donors likewise they were insufficient to produce a precipitous decline in
identified a low frequency of transient plasma viremia.83 In CMV DNA in either the plasma (free virus) or WBCs. Plasma
recently infected pregnant women, WBC-associated CMV and WBC-associated viral DNA were present in 25% to 40%
DNA was detected in 100% of samples during the first month and 75% to 80% of individuals, respectively, during the first
of infection using PCR and in 90% of samples during the 16 weeks of infection, and could still be detected in some
second month of infection. None of the samples were posi- individuals at 48 weeks of infection. Development of an
tive after 6 months of infection.94 WBC-associated viral load antibody response likewise failed to immediately suppress
decreased substantially during early infection. During the viral shedding because CMV could be isolated from 59%
first month 60% of positive samples had viral loads of more of urine specimens during the first 80 weeks of infection.67
than 10 CMV genome-equivalents (GE)/105 WBCs (range Consistent with these observations, infectious virus has also
of 10 to 398), whereas only 3.3% of positive samples dur- been identified in saliva and cervical secretions of remotely
ing the second month of infection had more than 10 GE/105 infected seropositive individuals.95,96 These findings indicate
WBCs.94 Thus, shortly after infection of the immunocom- that anti-CMV antibodies, including those with neutralizing
petent host, the patient’s viral load peaks. The subsequent activities in vitro, may not completely prevent viral infectiv-
decline in viral load correlates with the development of host ity in vivo.
anti-CMV immunity. Nonetheless, anti-CMV antibodies can protect against
sequelae of CMV infections in some circumstances. In a
Host Immune Response
study of neonatal CMV infection, 10 of 10 (50%) infants
HUMORAL IMMUNITY born to seronegative mothers who subsequently contracted
CMV infection initiates both a humoral and cellular immune transfusion-transmitted CMV developed serious or fatal
response in the host, although anti-CMV antibodies exert CMV disease. In contrast, 32 infants born to seropositive
mothers contracted CMV infections, but none developed
CMV disease, suggesting that passively acquired maternal
CMV antibodies abrogated disease severity.97 Therapeutic
III administration of antiviral antibodies, such as those present
WBC in intravenous immunoglobulin (IVIg) preparations, can
CMV DNA also be efficacious in altering CMV disease course in some
622
Anti-CMV IgG circumstances.98
Anti-CMV CTL Although anti-CMV antibodies generated during natural
infection display a spectrum of specificities, among the more
important targets is viral gB. Radioimmunoprecipitation
assays using recombinant gB demonstrated the presence
of anti-gB antibodies in the serum of all 48 seropositive
Plasma donors tested.99 Furthermore, the anti-gB antibodies were a
CMV DNA
significant component of the CMV-neutralizing activity in
the serum samples. When anti-gB antibodies were absorbed
Anti-CMV IgM with recombinant gB protein, viral neutralizing titer in the
serum was reduced an average of 48%.99 Similarly, other
studies have demonstrated that 40% to 88% of serum CMV-
Window phase
neutralizing activity in naturally infected donors is directed
Infection 2 months 4 months 6 months against gB.100,101 These results lend support for the use of
Figure 46–2 Temporal relationships between detection of plasma recombinant gB as a subunit CMV vaccine.
and white blood cell (WBC)-associated cytomegalovirus (CMV) DNA
and the development of an immune response after primary infec- CELLULAR IMMUNE RESPONSE
tion. CMV DNA can be detected by polymerase chain reaction in both
the plasma and WBC peripheral blood compartments during the first Cellular immune responses play an important role in the
month of infection, and subsequently declines to undetectable levels control of CMV infection. In bone marrow transplant (BMT)
over 4 to 6 months. The curves are not meant to be quantitative, but recipients, a patient population at high risk for CMV infec-
rather to illustrate that WBC-associated virus is more frequently detect-
able than plasma CMV during primary infection and also persists for
tion, development of a class I HLA-restricted CD8+ cytotoxic
a longer time. A humoral response is usually detectable by 8 weeks T-lymphocyte (CTL) response to CMV was significantly
postinfection, and persists indefinitely, along with a cytotoxic T-lympho- associated with the effective control of CMV infections.102–106
cyte response. The window phase represents the period between the In a study of 58 allogeneic BMT recipients with low or
initial presence of CMV in the peripheral blood and the first serologic absent anti-CMV CTL activity at enrollment, 43 developed
evidence of CMV infection. Seronegative blood components obtained
from donors in the window phase may explain some episodes of break- CMV infections, which were lethal in 12 of the patients.102
through transfusion-transmitted CMV in patients transfused exclusively Detectable anti-CMV CTL activity developed in all sur-
with seronegative blood. See text for detailed discussion. vivors of CMV infection, but in only 2 of the 12 patients
HUMAN HERPESVIRUS INFECTIONS
who succumbed to CMV disease. NK cell activity was also assays have been developed in multiple configurations,
depressed in the patients with fatal CMV infection but not including indirect hemagglutination, complement fixation,
in those who controlled their infections.102,103 In another solid phase fluorescence immunoassay, enzyme immunoas-
study, 10 of 20 recipients of allogeneic BMT developed CMV- say (EIA), latex or particle agglutination, and solid phase red
specific CTL activity by 3 months post-transplantation. Six of cell adherence, although the first three of these techniques
the 10 patients who failed to develop anti-CMV CTL activity are no longer frequently used.116–120 These assays detect anti-
died of CMV pneumonitis, and all 10 patients with a detect- CMV antibodies of the IgG, and in some cases IgM and IgA,
able CMV CTL response were protected.104 Similar conclu- classes. Direct comparisons of the sensitivities and speci-
sions regarding the protective effects of CMV CTLs were ficities of the latter four methodologies are difficult to per-
reached with CMV-seropositive patients who underwent form owing to the lack of good standards. Some EIAs have
autologous peripheral blood stem cell or bone marrow trans- stated sensitivities and specificities of approximately 99%
plantation. In these patients, whose preexisting CMV-specific and because of their objective readouts may have advantages
CTLs were suppressed by the preparative regimen, the reap- over techniques such as latex agglutination. However, anti-
pearance of anti-CMV CTL activity was positively correlated CMV antibodies may not be detected by serology until 6 to
with control of CMV infections (p = 0.002).105 The investi- 8 weeks after primary infection,67 and serology cannot accu-
gators also noted that a CD4 T-helper cell response to CMV rately identify or quantitate the extent of active CMV infec-
always preceded the reappearance of anti-CMV CTLs, and tion. Although viral culture can be used for these purposes,
appeared to be obligatory for the CTL response.105,106 conventional tube cultures can require 2 weeks or more to
Using a murine model, investigators specifically depleted yield results, and the more recently implemented shell vial
CD4+ and CD8+ T cells from the animals before experi- methodology may still require 24 to 48 hours to detect the
mental CMV infection to determine the contribution of presence of infectious CMV.121,122 Furthermore, these assays
each subset to antiviral immunity. These studies showed are only quantitative in a limiting-dilution format, which is
that CMV-primed CD8+ CTLs were capable of controlling labor intensive and not suited to routine clinical use.
CMV infections in the absence of CD4+ cells, except in sali- The CMV antigenemia assay (which uses immunostain-
vary glands.107,108 Furthermore, the mice deficient in CD4+ ing to identify and quantitate peripheral blood leukocytes
cells did not make high levels of anti-CMV antibodies, that contain CMV proteins) and CMV PCR have solved some
indicating that when the antiviral CTL response is intact of these problems.123–127 The antigenemia assay can be used
a humoral response is unnecessary for effective control of for both viral diagnosis and surveillance. Significantly, this
CMV infection.107 In the murine model, NK cells are also methodology is sensitive enough for early quantitative detec-
important for control of acute CMV infection.64,109 IFN-γ tion of CMV infections, allowing the institution of preemp-
production appears to be one of the principal mechanisms tive (presymptomatic) antiviral therapies.128,129 Qualitative
through which CD8+ and NK cells exert this effect.109–114 PCR allows even earlier detection of CMV infection than
The observation that cellular immunity plays a critical role does antigenemia.86,130–132 However, due to its sensitivity,
in controlling CMV infection has led to successful early- in some earlier studies PCR displayed poor positive predic- 46
stage clinical trials in which CMV-immune CTLs were tive value for identifying patients at risk for CMV disease
adoptively transferred into immunocompromised BMT because some patients with low but detectable viral loads 623
recipients.115 did not develop disease.86,129,131 The recent introduction of
quantitative PCR assays for CMV may provide a more rapid,
Laboratory Diagnosis of CMV Infection sensitive, and specific predictor of patients at risk for CMV
Accurate detection of CMV infection enables the identifi- disease.133 For example, the results obtained with a moder-
cation of transfusion recipients at risk for CMV infection, ately sensitive (400 copies of CMV DNA/mL) quantitative
as well as blood donors whose components are potentially CMV PCR assay strongly correlated with results from the
infectious. Furthermore, quantitation of the degree of viral antigenemia method and with development of CMV dis-
replication is important for guiding appropriate use of anti- ease.134 Advantages of the PCR method included reduced
viral therapies, such as ganciclovir, cidofovir, and foscarnet, turnaround time, smaller sample requirements (200 μL
in immunocompromised patients. The standard approach plasma versus 3 to 5 mL blood), simplified specimen pro-
for identifying a previously infected individual is through cessing, improved stability of specimens before processing,
detection of anti-CMV antibodies (Table 46–2). Serologic and ability to test samples from patients with leukopenia.
Category A: Clear morbidity and mortality from transfusion-transmitted CMV; CMV-safe components proven efficacious in
decreasing incidence of transfusion-transmitted CMV, and should be used for all transfusions.
• Low-birth-weight infants (<1500 grams) of seronegative mothers
• Seronegative recipients of autologous or seronegative allogeneic bone marrow transplantation
• Seronegative recipients of seronegative solid organ transplants, excluding renal and cardiac
Category B: Identified risk of morbidity and mortality from transfusion-transmitted CMV; benefit of CMV-safe products is possible
or likely, but not proven; consider using CMV-safe components when availability allows.
• Seronegative pregnant women requiring antepartum blood transfusion or intrauterine transfusion
• Seropositive women requiring intrauterine transfusion
• Low-birth-weight infants or seronegative immunosuppressed patients requiring granulocyte transfusions
• Seronegative HIV-infected patients
• Children born to HIV-infected mothers
• Low-birth-weight infants of seropositive mothers
• Seronegative patients who may be candidates for bone marrow transplantation
Category C: Morbidity and mortality of transfusion-transmitted CMV low or poorly documented, but likely greater than
Category D; consider CMV-safe products on a case-by-case basis.
• Infants with birthweight > 1500 grams, born to seronegative mothers
• Neonates receiving ECMO or extensive transfusion support (e.g., exchange transfusions)
• Seronegative recipients of seronegative renal or cardiac transplantation
• Seronegative patients receiving chemotherapy
• Seronegative patients experiencing major trauma or splenectomy
Category D: Low morbidity and mortality of transfusion-transmitted CMV; CMV-safe components not indicated.
• Infants with birthweight > 1500 grams, born to seropositive mothers
• All other transfusion recipients not listed above
Modified with permission from Preiksaitis JK. The cytomegalovirus-”safe” blood product: Is leukoreduction equivalent to antibody screening?
Transfus Med Rev 2000;14(2):112–136.
in liver function tests. The illness can progress to dissemi- seronegative recipients of seronegative marrow or autolo-
nated tissue-invasive CMV disease, including CMV hepatitis, gous transplants, transfusion is the primary mechanism for
retinitis, interstitial pneumonitis, encephalitis, and gastroen- CMV infection.158–160 Solid-organ transplant recipients are
teritis, including esophagitis.163 Progression to disease is more also susceptible to CMV infection and disease. In contrast 46
likely in patients with elevated viral loads. CMV infections to marrow transplantation, the most important source of
are also associated with, and may predispose to, other com- CMV infection is the donor organ, with transfusion-trans- 625
plications in immunocompromised patients, including graft- mitted CMV being less significant.144,176–179 In seronegative
versus-host disease (GVHD) in allogeneic marrow transplant recipients of seronegative organs, transfusion of unscreened
recipients,164–166 accelerated solid organ graft rejection,167,168 blood products has been associated with an incidence of
and other opportunistic infections, including invasive fungal CMV ranging from 0 to 33%, with a cumulative incidence of
disease.163,169 approximately 9% (reviewed in Roback147 and Preiksaitis170).
Seronegative infants transfused with unscreened blood Among organ transplant recipients, those receiving heart,
products had a 13.5% incidence of transfusion-transmitted heart-lung, liver, and pancreas transplants usually require
CMV.97 When greater than 50 mL of packed red cells were numerous transfusions and thus have an increased risk of
transfused, the incidence of transfusion-transmitted CMV transfusion-transmitted CMV. Early studies showed that
increased to 24%. Of the infants who acquired transfusion- even in heavily transfused organ transplant recipients, the
transmitted CMV, five (50%) developed serious symptoms or use of seronegative blood products could effectively prevent
fatal disease, all of them weighing less than 1200 grams.97 In transfusion-transmitted CMV.180–182
other studies, seronegative neonates weighing less than 1250 In addition to these patient populations with well-defined
to 1500 grams also experienced a high incidence of trans- susceptibility to transfusion-transmitted CMV, there are
fusion-transmitted CMV (reviewed in Preiksaitis170), likely other groups that may be susceptible and may also benefit
due to their immature immune systems. It should be noted, from transfusion of CMV-safe blood components (see Table
however, that low-birth-weight infants born to seropositive 46–3). For example, it is well-documented that primary
mothers can also be at risk for lethal CMV infection, despite CMV infection during pregnancy carries high risks of con-
the transfer of humoral immunity.171 genital fetal infection. Although there is no direct evidence
Marrow transplant recipients are at significant risk of that primary maternal infection resulting from transfusion-
morbidity and mortality from CMV infections. Up to one transmitted CMV can in turn lead to fetal infection,68,156
third of those patients who contract CMV infection can it is prudent to provide CMV-safe blood components to
develop CMV pneumonitis, a frequently fatal complica- pregnant women who are seronegative. Because of the high
tion.172 In seropositive marrow recipients, CMV infection is incidence of CMV reactivation and infection in seropositive
usually due to viral reactivation, making CMV seropositivity marrow recipients, 69% in one study,172 transfusion-trans-
the most important risk factor for CMV infection and dis- mitted CMV is not a significant concern in these patients.
ease.172–175 CMV infection also occurs frequently in seroneg- However, special components may be considered for sero-
ative recipients of seropositive marrow.158–160 However, in negative patients who are candidates for BMT, including
TRANSFUSION MEDICINE immunosuppressed oncology patients, to prevent infection Overall, the infection rate with transfusion of fresh blood
before transplant. has been documented at 10% to 59% (reviewed in Lee and
coworkers161). A decline in infectivity with storage has also
Blood Components Implicated been shown experimentally. When naturally infected blood
in Transfusion-Transmitted CMV obtained from AIDS patients with CMV viremia was refrig-
Most evidence suggests that the primary vector for trans- erated under standard conditions, CMV infectivity was
fusion-transmitted CMV is the CMV-infected leukocyte rapidly lost during the first 5 days of storage.187 However,
(Table 46–4). Transfusion-transmitted CMV has not been not all studies have demonstrated an effect of product stor-
observed in patients receiving blood components that are age interval on the incidence of transfusion-transmitted
free of WBCs, arguing that plasma free virus is not signifi- CMV.70,149,153,188
cantly involved in the pathogenesis of transfusion-trans- It is also useful to consider the percentage of donated
mitted CMV.183 For example, there was no evidence of blood components that can transmit CMV infection. In
transfusion-transmitted CMV in a group of 21 immunosup- a review of 10 studies published between 1968 and 1988,
pressed seronegative recipients of seronegative BMT under- including data from 2806 patients, Ho calculated the chang-
going total plasma exchange, although they were exposed ing risk of contracting transfusion-transmitted CMV over
to an average of 47.6 +/− 19.5 units of unscreened fresh this period.186 The risk per unit of unscreened and unfiltered
frozen plasma (FFP).184 The absence of CMV after transfu- blood was calculated at 11% to 12% in 1968 to 1970, and
sion of FFP may be due to the scarcity of plasma free virus remained at greater than 1% per unit until the early 1980s.
in healthy seropositive donors, as well as neutralization of However, the risk subsequently fell to 0.4% or less by 1988.186
virus by anti-CMV antibodies. Possible reasons for changes in the epidemiology of transfu-
In contrast, there is an abundance of evidence that trans- sion-transmitted CMV include decreasing use of fresh blood,
fusion-transmitted CMV can be mediated by WBCs in blood improved tools for CMV serologic screening of blood donors
components and that the incidence of transfusion-transmit- and recipients, implementation of improved protocols to
ted CMV correlates with the WBC load. For example, multi- screen for other viral infections, and exclusion of blood from
ple studies have demonstrated transfusion-transmitted CMV homosexual men starting in the mid-1980s.186 The risk of
after granulocyte transfusions, most often by seropositive transfusion-transmitted CMV also varies with different
granulocyte preparations transfused to seronegative recipi- groups of transfusion recipients. In studies where recipi-
ents.154,172,185 Red blood cell and platelet transfusions are also ents were immunosuppressed, 2.5% to 12% of unscreened
known to transmit CMV infections97 (see Table 46–4). blood components were estimated to be infectious.70,97,144
Comparison of data over the past 4 decades reveals a cor- Adler estimated that although 5% of all units were capable
relation between decreased usage of fresh blood since the of producing CMV infection, 15% of seropositive units were
1960s and a decline in the incidence of transfusion-trans- potentially infectious.189 In nonimmunosuppressed patients,
mitted CMV produced by unscreened units over the same in contrast, the risk may be as low as 0.14% of randomly
III period.186 Most early evidence suggested that fresh blood selected units, or 0.38% of seropositive units.
(donated within 24 hours) from seropositive donors was
Pathobiology of Transfusion-
626 more infectious than stored blood.145,155 For example, in the
Transmitted CMV Infections
initial descriptions, transfusion-transmitted CMV typically
occurred following open-heart surgery in which the patient Clinical evidence for the involvement of leukocytes in trans-
was exposed to fresh whole blood.138–140 In nonimmunosup- fusion-transmitted CMV include the observations that CMV
pressed seronegative transfusion recipients, 6 of 7 patients is transmitted with high frequency by granulocyte transfu-
who seroconverted were transfused with fresh whole blood. sions from seropositive donors,154 whereas the incidence of
In contrast, among 585 patients who did not seroconvert, transfusion-transmitted CMV can be attenuated by remov-
only 53 had received fresh blood (p < 0.001).155 Similar ing leukocytes from blood components.146,190 Furthermore,
findings were made in a pediatric study where transfusion- these WBCs are in almost all cases latently infected. When
transmitted CMV occurred in 13 of 15 children (87%) who more than 1500 buffy coat samples from healthy blood
received fresh blood (<24 hours old) as compared to 1 of 6 donors were subject to viral culture in multiple studies, only
children (17%; p = 0.01) receiving blood more than 24 hours 2 samples (both seropositive) grew infectious CMV.70,92,93
old.161 In the same study, none of the children receiving only These results demonstrate that when WBC-associated CMV
CMV-seronegative blood developed CMV, compared to the is present, it is nearly always in the latent state. Experimental
36% rate of CMV in children receiving unscreened blood.161 studies have indicated that WBCs of the monocyte lineage are
*
Not specifically tested, but other plasma-derived components have not been shown to be infectious for CMV.
n/a, not applicable.
the most likely to carry latent CMV in seropositive donors,36 Prevention of Transfusion-Transmitted CMV