Professional Documents
Culture Documents
Hígado Páncreas y Transplante Ahpba Capitulo Pancreas
Hígado Páncreas y Transplante Ahpba Capitulo Pancreas
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. Requests can be sent to Beaux Books
Publishing, c/o Trina Chu, P.O. Box 53364, Shreveport, Louisiana 71135. This book and the individual
contributions contained in it are protected under copyright by the Publisher (other than as may be noted
herein).
NOTICES
Knowledge and best practice in the field are constantly changing. As new research and
experience broaden our understanding, changes in research methods, professional practices, or
medical treatment may become necessary.
This work is provided “as is” and contains information obtained from authentic and highly
reliable sources. While all reasonable efforts have been made to publish reliable and up to date data
and information, neither the Publisher nor the editors, authors, contributors, the Americas-Hepato-
Pancreato- Biliary Association (AHPBA) can accept any legal responsibility or liability for any
errors or omissions that may be made. The Publisher wishes to make clear that any views or
opinions expressed in this book by individual editors, authors, contributors, or AHPBA are personal
to them and do not necessarily reflect the views/opinions of the Publisher. The editors, authors,
contributors, AHPBA, and the Publisher disclaim any and all warranties, express or implied,
including any warranties as to accuracy, comprehensiveness, or currency of the content of this
work.
Practitioners and researchers must always rely on their own experience and knowledge in
evaluating and using any information, methods, compounds, or experiments described herein. In
using such information or methods, they should be mindful of their own safety and the safety of
others, including parties for whom they have a professional responsibility. This work is no substitute
for individual patient assessment based upon healthcare professionals’ examination of each patient
and consideration of, among other things, age, weight, gender, current or prior medical conditions,
medication history, laboratory data and other factors unique to the patient. The Publisher does not
provide medical advice or guidance and this work is merely a reference tool.
To the fullest extent of the law, neither the Publisher nor the editors, authors, contributors, or
AHPBA assume any liability for any injury and/or damage to persons or property as a matter of
products liability, negligence or otherwise, or from any use or operation of any methods, products,
instructions, or ideas contained in the material herein. Healthcare professionals, and not the
Publisher, are solely responsible for the use of this work including all medical judgments and for any
resulting diagnosis and treatments.
Given continuous, rapid advances in medical science and health information, independent
profession verification of medical diagnoses, indications, appropriate pharmaceutical selections and
dosages, and treatment options should be made and healthcare professionals should consult a
variety of sources. When prescribing medication, healthcare professionals are advised to consult the
product information sheet (the manufacturer’s package insert) accompanying each drug to verify,
among other things, conditions of use, warnings and side effects and identify any changes in dosage
schedule or contradictions, particularly if the medication to be administered is new, infrequently used
or has a narrow therapeutic range. The reader is strongly urged to consult the drug companies’
printed instructions, and their websites before administering any of the drugs or selecting any
procedures recommended in this book. This book does not indicate whether a particular evaluation
and/or treatment is appropriate or suitable for a particular individual. Ultimately, it is the sole
responsibility of the medical professional to make his/her own professional judgements, so as to
advise and treat patients appropriately.
The Publisher, editors, authors, contributors, and AHPBA have also attempted to trace the
copyright holders of all material reproduced in this publication and apologize to copyright holders if
permission to publish in this form has not been obtained. If any copyright material has not been
acknowledged, please write and let us know so that we may rectify in any future reprint.
Except as permitted under U.S. Copyright Law, no part of this book may be reprinted,
reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now
known or hereafter invented, including photocopying, microfilming, and recording, or in any
information storage or retrieval system, without written permission from the Publisher.
Trademark Notice: Product or corporate names may be trademarks or registered trademarks,
and are used only for identification and explanation without intent to infringe.
Administrative Assistants: Marjorie Malia, Beverly Wright, Ashley Abrams, and Yvette Sanchez
Cover Design: Jeremy Ricci
Format Designers: Jeremy Ricci, Thuy-Tien Chu
Proofreaders: Barbara Pratt, Dr. Phillip Estes, Nina Tindall
Legal Assistant: Trina Chu, Esq
Artists: Yen Chu, Iwan Yasin Agustin, Kaylee Messina
Printed in The United States of America
Quyen D. Chu, MD, MBA
To my parents, Trinh V. Chu and Nhan T. Chu, who risked everything,
crossing a vast and perilous ocean so that we children could have a future in
America. To my wife, Trina Chu, and my two children, Thuy-Tien and Yen,
who endured countless days and nights without having their husband and
father at home. To my mentors and patients who have taught me not only the
art and science of surgery, but also the humility of knowing that there is much
more that I do not know than what I do know. To my alma mater, Dartmouth
College, who took a chance on me and gave me this wonderful life. Finally,
to this great nation of ours that has given me the opportunity to rise from a
world of poverty to one of limitless possibilities.
Charles M. Vollmer, MD
To my wife, Beth, and three children, Andrew, Julia, and Eric, for their
steadfast patience, understanding, and support. Also, to my parents, Charles
and Judith Vollmer, for providing me all the opportunities to become the
surgeon I always wanted to be. It is my sincere hope that this endeavor will
improve decision-making and outcomes for the next generation of HPB
surgeons.
Gazi B. Zibari, MD
To my parents, the late Baderkhan S. Zibari and Khanzad K. Zibari, who are
up in heaven. To my caring and passionate wife, Maysoon M. Zibari, who
single-handedly raised our children while bravely beating breast cancer. To
our four children, Susan, Renas, Rona, and Lehat. To my siblings, especially
my oldest brother, General Zibari, who was the father figure for all of us,
helping guide us through difficult life-and-death decisions. To my mentors,
colleagues, students, and organ donors and their families. To the AHPBA
Foundation Board and the leadership and management team for entrusting us
with this worthy project. To all authors for their outstanding contributions. To
my co-editors for their steadfast support, especially to Dr. Chu for
spearheading this noble work. To the victims of the Kurdistan of Iraq who
had to endure the brutal Anfal Campaign and the Halabja chemical attack. To
the Kurdish Peshmerga who, on the behalf of the world, fought bravely
against ISIS. Finally, to the United States of America, the best country in the
world, for allowing me to take refuge and empowering me to achieve the
American Dream.
Susan L. Orloff, MD
To all of my patients, who have given me the privilege and honor of caring
for them and who have provided me with true meaning in my life as a
physician and surgeon. To my parents, Ann and Marshall, both physicians,
who guided my five siblings and me on a path of embracing education and
humanitarianism, and gave all of us opportunities towards this endeavor. To
my husband, Bob, my son, Jackson, my daughter, Annie, and our various and
sundry animals that create our menagerie. To Marjie Malia, Director of
Association Management for the AHPBA, and Katherine Rose Franklin, my
Administrative Coordinator, who were incredibly helpful in navigating the
difficult task of ”herding cats” towards the goal of making this textbook
happen. To Jeremy Ricci, for his amazing and beautiful formatting, and to
Iwan Agustin, for his creative artistic talents. Finally, to all of the awesome
authors who participated in the content of this textbook that has led to the
creation of a novel and, hopefully, very impactful textbook. To the
readership, please read, learn, enjoy, and become invigorated by our field.
René Adam, MD
Head Department of Hepatobiliary Surgery, Cancer and Transplantation
Director Research Team
Hôpital Universitaire Paul Brousse
Centre Hépato-Biliaire
Villejuif, France
Marc-Antoine Allard, MD
Department of HPB Surgery & Transplantation
Centre Hépatobiliaire
Université Paris Sud
Paul Brousse Hospital
Villejuif, France
Mario Almada, MD
Assistant Surgeon
University of Montevideo
Montevideo, Uruguay
Yumi Ando, MD
Fellow, Gastroenterology
Oregon Health & Sciences University
Portland, Oregon
Gustavo Andreoli, MD
Associate Professor of Surgery
Department of Surgery
University of the Republic (UDELAR)
Uruguay
Oscar C. Andriani, MD
Assistant Professor
HPB Surgery
Swiss Medical Group
Universidad Austral
Buenos Aires, Argentina
Jenna Aziz, MS
Georgetown University
Washington, DC
Salim Aziz, MD
Clinical Professor of Surgery
George Washington University &
Howard University School of Medicine
Washington, DC
Michelle Babicky, MD
Senior Clinical Fellow in Surgical Oncology
Memorial Sloan Kettering Cancer Center
New York, New York
Wesley Barry, MD
Research Fellow
Children’s Hospital Los Angeles
University of Southern California
Los Angeles, California
Joal D. Beane, MD
Fellow
Complex Surgical Oncology
University of Pittsburgh Medical Center
Pittsburgh, Pennsylvania
Rachel E. Beard, MD
Assistant Professor of Surgery
Division of Surgical Oncology
Hepatobiliary and Pancreatic Surgery
Alpert Medical School of Brown University
Rhode Island Hospital
Providence, Rhode Island
Hillary J. Braun, MD
Resident Physician, General Surgery
University of California, San Francisco
San Francisco, California
L.D. Britt, MD, MPH, D.Sc. (Hon), FACS, FCCM, FRCSEng (Hon),
FRCSEd (Hon), FWACS (Hon), FRCSI (Hon), FCS(SA) (Hon)
Henry Ford Professor and Edward J. Brickhouse Chairman
Eastern Virginia Medical School
Surgery Department
Norfolk, Virginia
Zachary J. Brown, DO
Surgical Oncology Research Fellow
Thoracic and GI Oncology Branch
Center for Cancer Research
National Cancer Institute | National Institutes of Health
Bethesda, Maryland
Lee Cartagena, MD
Senior Trauma/Critical Care Fellow
Department of Trauma and Burns
John H. Stroger Hospital of Cook County
Chicago, Illinois
François Cauchy, MD
HPB Unit of Liver Surgery and Transplantation
Hopital Beaujon
Paris, France
Guillermo Cervio, MD
Chief of the Clinical Area Hospital Garrahan
Buenos Aires, Argentina
Turkmen T. Ciftci, MD
Associate Professor of Radiology
Hacettepe University
Department of Radiology
Nonvascular Interventional Radiology
Ankara, Turkey
Jordan M. Cloyd, MD
Assistant Professor of Surgery
Division of Surgical Oncology
The Ohio State University Wexner Medical Center
Columbus, Ohio
Jashodeep Datta, MD
Fellow, Complex General Surgical Oncology
Memorial Sloan Kettering Cancer Center
New York, New York
Allison David
Sidney Kimmel Medical College
Thomas Jefferson University
Philadelphia, Pennsylvania
Siddharth Desai, MD
Chief Resident in Surgery
University of Mississippi Medical Center
Jackson, Mississippi
Mashaal Dhir, MD
Assistant Professor of Surgery
Upstate Medical University
Syracuse, New York
Geraldine C. Diaz, DO
Associate Professor of Anesthesia & Critical Care
University of Illinois at Chicago
Chicago, Illinois
Maren Donato, MD
Clínica de Imágenes
Neuquén
Neuquén, Argentina
Austin Dosch, MD
Resident in Surgery
University of California, Irvine
Irvine, California
Fernando Duek, MD
Staff Surgeon
Liver & Transplant Unit
Hospital Dr. Cosme Argerich
Buenos Aires, Argentina
Elizabeth W. Duggan, MD
Assistant Professor of Anesthesiology
Chief of Service for Division of General
& Transplant Anesthesiology
Emory University School of Medicine
Atlanta, Georgia
Brett L. Ecker, MD
Resident, Department of Surgery
University of Pennsylvania
Philadelphia, Pennsylvania
Alberto Espino, MD
Assistant Professor
Department of Gastroenterology
Director, Endoscopy Unit
Pontificia Universidad Católica de Chile
Santiago, Chile
Caetano Finger, MD
Percutaneous Surgeon,
Staff DAICIM Foundation,
Buenos Aires, Argentina
Kathryn J. Fowler, MD
Assistant Professor
Mallinckrodt Institute of Radiology
Washington University in St Louis
St Louis, Missouri
Wesley P. Francis, MD, FACS
Associate Lecturer
University of the West Indies
School of Clinical Medicine and Research
Nassau, Bahamas
Catherine H. Frenkel, MD
Fellow
University of Pennsylvania
Philadelphia, Pennsylvania
Danielle M. Fritze, MD
Assistant Professor of Surgery
University Transplant Center
University of Texas Health San Antonio
San Antonio, Texas
Armando Ganoza, MD
Assistant Professor of Surgery
Thomas E. Starzl Transplantation Institute
Children’s Hospital of Pittsburgh of UPMC
University of Pittsburgh Medical Center
Pittsburgh, Pennsylvania
Georgios Gemenetzis, MD
Postdoctoral Fellow
Department of Surgery
Johns Hopkins University School of Medicine
Baltimore, Maryland
Nicolas Goldaracena, MD
Clinical Associate in Transplantation
Department of Surgery
University of Toronto
Toronto, Canada
Jamie Golden, MD
Pediatric Surgery Research Fellow
Children’s Hospital Los Angeles
Los Angeles, California
Claire Goumard, MD
Research Fellow
University of Texas MD Anderson Cancer Center
Houston, Texas
Elliot I. Grodstein, MD
Assistant Professor of Surgery
Donald and Barbara Zucker School of Medicine at Hofstra/Northwell
Division of Transplantation
Department of Surgery
Manhasset, New York
Esteban Halac, MD
Staff Surgeon
Liver Transplant Division
Hospital Juan P. Garrahan
Buenos Aires, Argentina
Andrew D. Hardie, MD
Assistant Professor of Radiology
Medical University of South Carolina
Charleston, South Carolina
Jonathan M. Hernandez, MD
Surgical Oncology and HepatoPancreatoBiliary Surgery
Thoracic and GI Oncology Branch | Center for Cancer Research
National Cancer Institute | National Institutes of Health
Bethesda, Maryland
Eduardo J. Houghton, MD
Staff DAICIM Foundation.
Associate Professor
University of Buenos Aires
Staff, Mini Invasive Surgery
Bernardino Rivadavia Hospital
Buenos Aires, Argentina
Abhinav Humar, MD
Chief, Division of Abdominal Transplantation Surgery
Clinical Director, Thomas E. Starzl Transplantation Institute
Thomas E. Starzl Professor in Transplantation Surgery
University of Pittsburgh Medical Center
Pittsburgh, Pennsylvania
Oscar Imventarza, MD
Chief of Liver Transplantation Hospital Garrahan & Hospital Argerich
President International Hepato Pancreatic Biliary Association
President Elect STALYC (Latin American Transplant Society)
Buenos Aires, Argentina
Kyle R. Jackson, MD
General Surgery Resident
Johns Hopkins Hospital
Baltimore, Maryland
Sandeep Jhajj, MD
Resident
Department of Surgery
University of Arizona
Tucson, Arizona
Zeljka Jutric, MD
Assistant Professor, Department of Surgery
Division of Hepatobiliary & Pancreas Surgery
University of California, Irvine Medical Center
Orange, California
Piyush Kalakoti, MD
Postdoctoral Research Fellow
LSU-Health Sciences Center- Shreveport
Shreveport, Louisiana
Na Eun Kim, MD
Boston Medical Center
General Surgery Resident
Boston, Massachusetts
Nisa Kubiliun, MD
Assistant Professor of Internal Medicine
Director of Endoscopic Services
Clinical Chief, Division of Digestive and Liver Diseases
UT Southwestern Medical Center
Dallas, Texas
Aleksandra Kukla, MD
Associate Professor of Internal Medicine and Nephrology
Mayo Clinic
Rochester, Minnesota
Rupak D. Kulkarni, MD
Assistant Professor of Surgery
Division of Transplantation
Department of Surgery
University of Texas Medical Branch
Galveston, Texas
Michael E. Lidsky, MD
Fellow, Hepatopancreatobiliary Surgery & Complex Surgical Oncology
Memorial Sloan Kettering Cancer Center
New York, New York
Michael Loudin, MD
Gastroenterology Fellow
Division of Gastroenterology and Hepatology
Oregon Health & Science University
Portland, Oregon
Wuttiporn Manatsathit, MD
Department of Gastroenterology
University of Nebraska Medical Center
Omaha, Nebraska
Ronald S. Mowad, MD
General Surgery Resident
Department of Surgery
LSU Health Sciences Center- Shreveport
Shreveport, Louisiana
Arvind R. Murali, MD
Assistant Professor of Medicine
Gastroenterology & Hepatology
University of Iowa
Iowa City, Iowa
David Nasralla, MD
Clinical Research Fellow in Transplant Surgery
Nuffield Department of Surgical Sciences
University of Oxford
Oxford Transplant Centre
Churchill Hospital
Oxford, United Kingdom
Willscott Naugler, MD
Associate Professor of Medicine
Medical Director of Liver Transplantation
Oregon Health & Science University
Portland, Oregon
Sergio Pacheco MD
Digestive Surgeon
Pontificia Universidad Catolica de Chile
Santiago, Chile
Alessandro Paniccia, MD
General Surgery Resident
Department of Surgery
University of Colorado Anschutz Medical Campus
Aurora, Colorado
Rebecca Plevin, MD
Clinical Instructor of Surgery
Trauma Fellow
Zuckerberg San Francisco General Hospital
UCSF, School of Medicine
San Francisco, California
Alessandra Pulvirenti, MD
Research Fellow
Hepatopancreatobiliary Service
Department of Surgery
Memorial Sloan-Kettering Cancer Center
New York, New York
Majed A. Refaai, MD
Associate Professor, Pathology and Laboratory Medicine
Director, Hemostasis and Thrombosis Laboratory
Associate Director, Clinical Laboratory Medicine
University of Rochester Medical Center
Rochester, New York
Jeffrey Reha, MD, FACS
Division of Surgical Oncology
Department of Surgery
Roger Williams Medical Center
Providence, Rhode Island
Sean Rogers, MD
University of Missouri-Kansas City School of Medicine
Kansas City, Missouri
Vinayak S. Rohan, MD
Assistant Professor of Surgery
Division of Transplant Surgery
Medical University of South Carolina
Charleston, South Carolina
Robert E. Roses, MD
Assistant Professor
Department of Surgery
Hospital of the University of Pennsylvania
University of Pennsylvania Perelman School of Medicine
Philadelphia, Pennsylvania
Thomas M. Runge, MD
Division of Gastroenterology and Hepatology
University of North Carolina, Chapel Hill
Chapel Hill, North Carolina
Hrishikesh Samant MD
Assistant Professor of Clinical Medicine
Division of Gastroenterology/ Hepatology
LSU Health Sciences Center- Shreveport
Transplant Hepatology
JCM Transplant Center, WK System
Shreveport, Louisiana
Guillermo P. Sangster, MD
Professor of Clinical Radiology and Anesthesiology
LSU Health Sciences Center - Shreveport
Director, Body Imaging Division
Department of Radiology
Shreveport, Louisiana
Amit Sastry, MD
HPB Surgery Fellow
Carolinas HealthCare System
Charlotte, North Carolina
David L. Scott, MD
Associate Professor of Surgery
Director of Kidney and Pancreas Transplantation
Oregon Health and Science University
Portland, Oregon
Aarti Sekhar, MD
Associate Professor of Radiology
Department of Radiology and Imaging Sciences
Emory University School of Medicine
Atlanta, Georgia
Edgardo Serra, MD
Director of Centro CIEN
Bariatric Surgery, Argentina
Staff of the Percutaneous Image Guided Surgery TEAM of
DIACIM Fundation/IHU-IRCAD
Strasbourg, France
Adil Shah, MD
Surgery Resident
Howard University
Washington, DC
Neeraj Singh, MD
Associate Professor of Internal Medicine and Nephrology
LSU Health Sciences Center- Shreveport
Medical Director, Kidney and Pancreas Transplant
John C. McDonald Regional Transplant Center,
Shreveport, Louisiana
Melanie K. Sion, MD
Advanced GI Surgery Fellow
Department of Surgery
Thomas Jefferson University
Philadelphia, Pennsylvania
Thomas Soeprono, MD
Director of Transplant Psychiatry
Associate Residency Director
Attending Physician
Psychiatric Consultation and Telepsychiatry Program
Department of Psychiatry
University of Washington
Seattle, Washington
Divya Sood, MD
General Surgery Resident
Cancer Therapeutics Research Fellow
Moores Cancer Center
University of California, San Diego
La Jolla, California
Zachary E. Stiles, DO
Research Fellow and Resident in Surgery
Department of Surgery
University of Tennessee Health Science Center
Memphis, Tennessee
Lynne Strasfeld, MD
Associate Professor of Medicine
Division of Infectious Diseases
Oregon Health & Science University
Portland, Oregon
Ashley N. Suah, MD
General Surgery Resident
University of Chicago Medicine
Chicago, Illinois
Vijay Subramanian, MD
Washington University School of Medicine
St. Louis, Missouri
Miguel Tobon, MD
General Surgery Resident
Wayne State University/Detroit Medical Center
Detroit, Michigan
Anthony M. Villano, MD
Resident Physician
Medstar Georgetown University Hospital
Washington, DC
Angela S. Volk, MD
General Surgery Resident
Tulane University School of Medicine
New Orleans, Louisiana
Mariano M. Volpacchio, MD
Head of Computed Tomography
Hospital de Clínicas, Universidad de Buenos Aires
Head of Body Imaging
Buenos Aires, Argentina
Scott Westphal, MD
Assistant Professor of Medicine
University of Nebraska Medical Center
Omaha, Nebraska
Steven A. Wisel, MD
UCSF Department of Surgery
San Francisco, California
Brent T. Xia, MD
General Surgery Resident
University of Cincinnati Medical Center
Cincinnati, Ohio
7 Is There a Role for Liver Resection in Patients with Non- Colorectal,
Non-Neuroendocrine Metastases?
Anthony M. Villano and Lynt B. Johnson
10 NASH, CASH, and Steatohepatitis: What Are the Concerns and How
to Diagnosis Them?
Hrishikesh Samant and Wuttiporn Manatsathit
11 How Should the HPB Surgeon Manage the Patient with Polycystic
Liver Disease?
Zaheer S. Kanji and Adnan A. Alseidi
P /C D
18 A Patient with Colorectal Cancer and Synchronous Liver Metastases:
A) Simultaneous Resection of Primary and Metastases:
Roberto Hernandez-Alejandro and Kerollos Nashat Wanis
B) Liver Resection First (Reverse Strategy):
Marc-Antoine Allard and René Adam
C) Resect the Primary First:
Somala Mohammed, Steven A. Curley, and Avo Artinyan
I D
23 How to Manage Tumors Involving the Retrohepatic Inferior Vena
Cava
Alan W. Hemming, and Jennifer Berumen
P D
26 What Are the Post-Hepatectomy Complications and How Should I
Manage Them?
Miguel-Angel Mercado and Hiram Raul Muñoz Gonzalez
P D
27 How to Avoid Injuring the Biliary Tree During Laparoscopic
Cholecystectomy and How to Manage it When it Does Happen?
Dominic Sanford and Steven M. Strasberg
P D
49 How Should I Manage Postoperative Bile Leak?
Blayne A. Sayed and David A. Kooby
P D
50 A Patient Needs a Pancreatectomy But Has A Cardiac History
Demanding Anticoagulation Therapy. How Do You Manage This?
Zachary Brown and Jonathan Hernandez
53 The Patient Presents with a Pancreatic Head Mass with Jaundice and
a Bilirubin of 20. Should You Operate Soon or Stent the Bile Duct?
Tyler S. Wahl and John D. Christein
P /C D
59 A 1.5 cm Neuroendocrine Tumor is Identified in the Pancreatic Head.
Should I Operate or Not?
A) The Case for Resection:
Catherine H. Frenkel and John D. Allendorf
B) The Case for Observation:
Alessandra Pulvirenti and Peter J. Allen
I D
61 A Pancreatic Cancer at the Portal Confluence Requires a Vein
Resection. How Do You Reconstruct the Vein – Primary Repair,
Venous Conduit, Synthetic Graft?
Jordan M. Cloyd and Jeffrey E. Lee
P /C D
69 Should I Place an Intraoperative Drain for Whipple’s Resection?
A) The Case for Drainage:
George van Buren 2nd and William Fisher
B) The Case Against Drain Use:
John W. Kunstman and Ronald R. Salem
P D
70 A Septic Patient with a Completely Dehisced Pancreatico-
jejunostomy (ISGPF Grade C) Requires a Reoperation on POD#7.
How Do You Manage This?
Rachel E. Beard and Mark P. Callery
P /C D
79 Should I Use a Somatostatin Analogue After a Whipple Resection?
A) Pro:
Shailesh V. Shrikhande and Esha Pai
B) Con:
Divya Sood and Andrew Lowy
LIVER TRANSPLANTATION
P D
80 How Should I Optimize a Patient With Acute Cholecystitis and End-
Stage Liver Disease?
Jared A. White and Devin Eckhoff
98 How Do You Decide Whether to Use the Right Lobe or Left Lobe in a
Live Donor Liver?
Dong-Hwan Jung and Sung-Gyu Lee
100 How Should We Match Donor Organ Quality with Recipient Severity
of Illness?
Kenneth Washburn and Sylvester M. Black
I D
105 Management of Hepatorenal Syndrome, Renal Insufficiency
Intraoperatively-Intraoperative CRRT?
Dragos M. Galusca and Marwan S. Abouljoud
111 When to Use Venoveno Bypass? (Rarely Needed- But There Are
Some Indications)
Shannon Orr and Christopher D. Anderson
116 Is tPA Protocol with Infusion into the Recipient Hepatic Artery in
Donation After Circulatory Death (DCD) Livers to Prevent Ischemic
Cholangiopathy Useful?
Nicolas Goldaracena and Paul D. Greig
P D
117 How Should I Manage Hepatic Artery Thrombosis After Liver
Transplantation?
Mary Decoteau, Jamie Case, and Christopher Lee Marsh
120 How Should I Manage Biliary Strictures and Leaks After Liver
Transplantation?
Juan F. Guerra and Alberto Espino
121 What Are The Indications and Timing for Liver Retransplantation?
Erin C. Maynard
122 What Should I Look for in a Liver Transplant Recipient Who Has a
Fever?
Sui Kwong Li and Lynne Strasfeld
123 What Are the Common Perioperative Cardiac Events During and
After Liver Transplantation?
Elliot I. Grodstein and David P. Foley
127 What Are the Indications for Early Reoperation Following Liver
Transplantation?
Ashley Suah, Adam S. Bodzin, and J. Michael Millis
PANCREAS TRANSPLANTATION
128 Is There a Role for Islet Cell Transplantation in Patients with Type I
Insulin-Dependent Diabetes Mellitus? Should Islet Cell Transplant Be
Considered for Type II Diabetes Mellitus Patients?
Steven A. Wisel and Peter G. Stock
131 Who Should Get a Pancreas Transplant, Which Technique and How
to Manage Its Rare Complications?
Hosein Shokouh-Amiri and Gazi B. Zibari
133 A Patient is Two Years Post Pancreas Transplant with Bleeding from
the Donor Arterial Conduit Aneurysm. What is the Role of Surgery
and Interventional Radiology?”
Armando Ganoza and Abhinav Humar
HEPATIC TRAUMA
139 What is the Role of Vascular Shunting for Major HPB Trauma?
Chad G. Ball and Amar Gupta
140 What is the Role of Hepatic Transplantation for Hepatobiliary
Trauma?
Rupak D. Kulkarni and Jeffrey H. Fair
PANCREATICODUODENAL TRAUMA
SPECIAL TOPICS
P D
160 Will Technology, Along with Proper Training and Teaching, Improve
Patient Outcome?
Jorge Cardoso Cuneo and Andrés Garelli
162 How Many Classifications Are There for Acute Pancreatitis &
Which One Should We Be Using?
Michael G. Sarr
P /C D
170 Preoperative HPB Biopsy:
A) Rarely Needed:
Perparim Limani, Claire Goumard, and Claudius Conrad
b) Always:
Claudemiro Quireze, Jr
I /T D
172 Image Fusion, Virtual Reality and Hybrid Imaging in HPB Surgery -
Landmark Innovations or New-Fangled Technology?
Michele Diana and Jacques Marescaux
P /C D
176 Palliative Treatment of Malignant Jaundice - Percutaneous vs.
Endoscopic & Echoendoscopic Approaches?
A) Percutaneous:
Eduardo Houghton
B) Endoscopic:
Thomas M. Runge and Todd H. Baron
P D
179 Bioethics - Implications of Use of Technology - Where Are We
Going?
Bruce Schirmer
CASE SCENARIO
A 68-year-old man is referred by his medical oncologist after a recent
pathologic diagnosis of pancreatic adenocarcinoma obtained during ERCP
and stenting for progressive jaundice. The patient has a history of
hypertension and rheumatic heart disease that required a mitral valve
replacement 8 years prior and is currently being treated with warfarin 5mg
daily. His most recent INR was 3.0. On review of the CT scan, the tumor
appears to be localized to the head of the pancreas with clear fat planes
surrounding the major vascular structures.
Management
The patient described will need to be seen in consultation with his
cardiologist, as well as by a staff hematologist who will continue to follow
throughout the perioperative period. The patient and consulting physicians
will need to be made aware of the potential complications and, specifically,
any complication that may require procedural intervention that could
necessitate delay or full reversal of anticoagulation. Given this patient’s
mechanical mitral valve, he is considered high-risk for cessation of
anticoagulation (see Table 1 ) and should therefore receive bridging therapy
with subcutaneous low molecular weight heparin. Warfarin should be
stopped 5 days prior to the operation. INR should be checked during the
ensuing days and corrected accordingly the day prior to resection with
vitamin K to a goal around 1.5.
TABLE 1: Overview of Risk Evaluation and Need for
Bridging Therapy. TIA: Transient Ischemic Attack, VTE:
Venous Thromboembolism, a-fib: atrial fibrillation, HTN:
hypertension, CHF: Congestive Heart Failure, CHADS2:
CHF, HTN, Age >75, Diabetes Mellitus, Stroke
DISCUSSION
Warfarin, a vitamin K antagonist, inhibits the synthesis of clotting factors II,
VII, IX, and X, as well as protein C and S. It was originally utilized as a
rodenticide in the 1950s and remains one of the most common
anticoagulation medications encountered in clinical practice today. The name
Warfarin was derived from its founding organization, the Wisconsin Alumni
Research Foundation, and “arin” from a similar molecule, coumarin.
Coumarin is present in high quantities in rotten sweet clover and was
responsible for the hemorrhagic deaths of large numbers of grazing cattle in
the 1920s. For patients requiring a Whipple procedure, or any major
operation, warfarin should be stopped approximately 5 days before the
operation, with a goal INR of about 1.5 at the start of the procedure. Vitamin
K should be administered as appropriate, with fresh-frozen plasma reserved
for immediate perioperative use.
The need for bridging anticoagulation after warfarin cessation is
dependent upon the risk of developing adverse cardiovascular and/or
thrombotic events. High, moderate, and low risk refers to the rate of
occurrence of acute thromboembolism per year in absence of anticoagulation
and is approximated at >10%/year (high), 5-10%/year (moderate, and
<5%/year (low). High- risk patients should undergo bridging therapy after
stopping antithrombotic medication, whereas low-risk patients typically do
not require anything beyond cessation of the medication. In patients
determined to be of moderate risk for perioperative acute thromboembolism,
there are no clear guidelines, and the decision should be made on an
individual basis. This information is summarized in Table 1 .
The incidence of prosthetic valve thrombosis in patients not appropriately
anticoagulated is approximately 2% per patient-year, whereas the incidence
of any thromboembolic event resulting in death, stroke, or peripheral
ischemia is 4% per patient-year. Of note, thromboembolic events following
aortic valve replacement are significantly lower than that of mitral valve
replacement. Patients with mitral valve prostheses require bridging
anticoagulation, which typically consists of subcutaneous low molecular
weight heparin at a dose of 1 mg/kg twice a day of actual (not ideal) body
weight. The last dose of low molecular weight heparin should be given 24
hours before the operation, or if the patient is admitted to the hospital and
receiving intravenous heparin as a bridge, the continuous infusion should be
stopped 4-6 hours before surgery.
CASE SCENARIO 2
A 62-year-old woman presents with a pathologic diagnosis of
adenocarcinoma of the head of the pancreas obtained during endoscopic
ultrasound. The patient’s CT scan reveals clear fat planes surrounding all
major vascular structures and a small hypodense mass in the uncinate
process. She has a history of atrial fibrillation and is being treated with
rivaroxaban (Xeralto). In addition, the patient has hypertension and type II
diabetes mellitus but no congestive heart failure, and has never suffered
symptoms of a transient ischemic attack or stroke.
Management
The patient described will need to be seen in consultation with her
cardiologist. As she has normal renal function, her anticoagulation
medication should be held ≥1 day prior to surgery. The half-life of
rivaroxaban is about 7 hours and there are no active metabolites. Her
CHADS2 score is two, with points assigned for hypertension and diabetes
mellitus. This CHADS2 score places her at low risk for a perioperative
cardiovascular event. No bridging therapy is required (Table 1 ).
Following completion of the procedure, prophylactic subcutaneous
heparin (we prefer heparin over other low molecular weight formulations)
should begin at 8 hours and continue throughout the hospitalization. This
patient is at low risk for perioperative events, and we believe patients with a
similar risk profile should generally not be restarted on their antithrombotic
therapy until the time of the first postoperative clinic visit. Moreover, this
patient may be at higher risk for a postoperative pancreatic fistula given the
size and location of her tumor, meaning she may require additional
postoperative interventions. In general, patients with low-risk profiles, such
as the patient presented in this scenario, will not require prophylactic
heparin therapy at home. This decision should of course be made in concert
with the consulting physicians.
DISCUSSION
Until recently the mainstay of anticoagulation therapy has been warfarin, but
over the past several years, new oral anticoagulant medications have
emerged. Both the intrinsic and extrinsic clotting cascades converge to
generate activated factor Xa, making it a potent target. Factor Xa, along with
factor Va, form the active prothrombinase complex. The prothrombinase
complex generates thrombin (from prothrombin), and ultimately the platelet-
fibrin thrombus. The new oral anticoagulation medications are summarized in
Table 2 .
CASE SCENARIO 3
A 58-year-old man is referred after presenting with weight loss and
symptoms of pancreatic exocrine insufficiency. He has a resectable
hypodense mass in the head of his pancreas and a serum Ca 19-9 level of
150. The patient has coronary artery disease and suffered a myocardial
infarction 10 months ago, for which he underwent percutaneous coronary
intervention with placement of two drug-eluting stents. The patient is
currently on anti-hypertensive medication, as well as aspirin and
clopidogrel.
Management
The patient described will need to be seen in consultation with his
cardiologist, as well as by a staff internist/cardiologist who will follow
throughout the perioperative period. Following completion of all necessary
preoperative cardiac assessment, the patient should be instructed to stop
clopidogrel 5 days before the operation. His drug-eluting stents were placed
greater than 6 months ago, meaning the likelihood of stent thrombosis is
significantly decreased when compared to the risk at earlier time points. The
patient should continue to take aspirin, as the available data indicates a
benefit from continuation of therapy for patients with coronary artery or other
cardiovascular disease and moderate- to high-risk for perioperative
cardiovascular events.
Following completion of the procedure, prophylactic heparin (we prefer
heparin over other low molecular weight formulations) should begin at 8
hours and continue throughout the hospitalization. Aspirin should be resumed
approximately 24 hours after surgery, and may be given per rectum if
necessary. Given that our patient’s coronary vessels were stented more than 6
months ago, we suggest delaying resumption of clopidogrel until the time of
discharge or at the time of the first follow-up visit. Clopidogrel is an
irreversible platelet ADP receptor antagonist and will necessitate platelet
transfusion if hemorrhage ensues or an additional procedure is required. This
decision regarding restarting of clopidogrel should of course be made in
concert with the cardiologist.
DISCUSSION
The incidence of cardiac stent thrombosis is 2-5% based on several
retrospective studies of patients undergoing operations within two years of
stent placement, which may seem small but the consequences can be
catastrophic. Patients with coronary artery stents usually require antiplatelet
therapy such as aspirin and clopidogrel to prevent stent thrombosis. Aspirin
is an irreversible cyclooxygenase inhibitor, while clopidogrel is an ADP
receptor antagonist. In general, no monitoring is required for antiplatelet
medications. See Table 3 for more information regarding antiplatelet agents.
If bleeding occurs and urgent reversal is required, the treatment of choice is
platelet transfusion.
TABLE 3: Comparison of postoperative outcomes by distal
pancreatectomy technique
Management
Despite the insistence, this patient has a borderline resectable cancer and
should not undergo an operation but rather referral to a medical oncologist
for systemic chemotherapy. Moreover, he is at high risk for any intervention
given his advanced age and multiple co-morbidities. This patient is unlikely
to complete the prescribed 6 months of chemotherapy without incident. He
should, however, undergo repeat imaging following completion of systemic
chemotherapy to evaluate the extent of his disease, assuming his candidacy
for resection remains viable. If resection is entertained, realistic expectations
of life expectancy with and without resection should be thoroughly explained,
preferably using a comorbidity risk estimation tool, along with the potential
complications associated with pancreatic resections. He should be seen in
consultation with his cardiologist for the usual battery of work up, such that
perioperative risk can be accurately assessed.
DISCUSSION
In our opinion, patients with potentially resectable disease should not be
denied consideration for operative intervention. However, in some patients
the risks simply outweigh the potential benefits. We would be remiss without
a reminder that the median survival after resection is about 2 years, and that
80% of patients undergoing resection will die of disease recurrence within 5
years. For the patient in question, using a life expectancy risk tool based on
comorbidities (and not accounting for the cancer diagnosis), his life
expectancy is approximately 2 years. In other words, accepting all the risks
associated with a pancreaticoduodenectomy in this patient, including
management of his antithrombotic medication, makes little sense if he is
unlikely to survive beyond 2 years because of his other medical
comorbidities.
SALIENT POINTS
Pancreaticoduodenectomy is a high-risk procedure necessitating
cessation of anticoagulation and antiplatelet therapy prior to resection.
As there is rarely, if ever, an urgent need for a
pancreaticoduodenectomy, appropriate planning in concert with
consulting physicians is strongly advised.
The use of bridging therapy should be stratified depending on the
individual patient’s risk for perioperative adverse events.
Patients on antiplatelet medication for coronary artery or other
cardiovascular disease who are considered moderate- to high-risk for
perioperative cardiovascular events likely benefit from continuation of
aspirin therapy.
Given the increased morbidity and mortality associated with
postpancreatectomy hemorrhage, and recognizing that this complication
can occur after discharge, we prefer to delay reinstitution of
antithrombotic therapy until the time of the first postoperative visit for
most patients, except those at the highest risk for thrombotic
complications.
SELECTED REFERENCES
1. Baron TH, Kamath PS, McBane RD. Management of antithrombotic
therapy in patients undergoing invasive procedures. N Engl J Med.
2013. 368(22): 2113-24.
2. Bauer KA. Pros and cons of new oral anticoagulants. Hematology Am
Soc Hematol Educ Program, 2013. 2013: 464-70.
3. Douketis JD, Spyropoulos AC, Spencer FA, Mayr M, Jaffer AK,
Eckman MH, et al. Perioperative management of antithrombotic therapy:
Antithrombotic Therapy and Prevention of Thrombosis, 9th ed:
American College of Chest Physicians Evidence-Based Clinical
Practice Guidelines. Chest, 2012. 141(2 Suppl): e326S-50S.
4. Panduranga P, Al-Mukhaini M, Al-Muslahi M, Hague MA, Shehab A.
Management dilemmas in patients with mechanical heart valves and
warfarin-induced major bleeding. World J Cardiol, 2012. 4(3): 54-9.
5. Wente MN, Veit JA, Bassi C, Dervenis C, Fingerhut A, Gouma DJ, et al.
Postpancreatectomy hemorrhage (PPH): an International Study Group of
Pancreatic Surgery (ISGPS) definition. Surgery, 2007. 142(1): 20-5.
█ BACKGROUND
Operative management with an R0 resection margin offers patients with
pancreatic adenocarcinoma the best odds at extended survival. Numerous
studies have demonstrated better short and long-term outcomes for patients
with pancreatic neoplasms operated on at high volume centers, and when
performed by high-volume surgeons. In addition, studies from MD Anderson
Cancer Center and Technische Universität München – two highly specialized
centers – have shown successful outcomes in patients undergoing resection
after being deemed inoperable at other centers.
These patients need to be treated by a team of specialists focused on the
diagnosis and management of pancreatic disease. The patient must be treated
by consensus, not ego or the tunnel vision of a single discipline. The
surgeon’s role is first to determine resectability. If initially unresectable, the
surgeon determines specific endpoints for which achievement may allow for
resectability. Collaboration amongst the multidisciplinary group then moves
toward achieving the endpoints while improving or maintaining quality of
life. The patient is regularly presented at a multidisciplinary conference
where relevant disciplines review and provide expert input, assess
achievement of objectives, and document next steps.
Approximately 50-70% of pancreatic tumors are considered unresectable
at diagnosis because of metastases, or arterial involvement. R0 resections
are possible in a proportion of the patients without metastases. The
remainder will progress during neoadjuvant therapy, or occult metastasis
will be identified at exploration. We therefore argue that locally advanced
pancreatic cancer (LAPC) must be treated by a multidisciplinary group with
high volume commitment to treating LAPC, not just experience treating
resectable disease. This will limit unnecessary explorations, and provide
safer outcomes with neoadjuvant and operative therapy.
DEFINITIONS
There is little variation in the definitions of potentially resectable, borderline
resectable, and LAPC among the commonly referenced criteria: MD
Anderson Cancer Center, National Comprehensive Cancer Network, and the
International Study Group of Pancreatic Surgery. No one set of criteria is
necessarily better than the other. Confusion mostly arises around borderline
resectable and LAPC, where there is great interpretative variability by
different institutions and the consequent treatments offered.
Potentially resectable disease can be characterized as tumor without
superior mesenteric artery, superior mesenteric vein, portal vein, celiac axis,
or hepatic artery involvement. These patients should go straight to surgery, or
enter a clinical trial evaluating neoadjuvant therapy. Although we have seen
patients advised by non-surgeons they were unresectable because of splenic
artery or vein involvement, this alone does not stand in the way of surgery. At
the opposite end of the continuum, metastatic disease is defined by the
presence of biopsy-confirmed adenocarcinoma in any distant organ, the
peritoneum, or lymph nodes that do not directly drain the pancreas. Under
such circumstances, surgery is contraindicated even if there is a response to
chemotherapy.
Borderline resectable cases are categorized based on venous or arterial
involvement. When the vein involvement can be resected and repaired by
simple suture or patch, we bring the patient directly to surgery. When there is
venous involvement requiring circumferential transection with or without
interposition grafting, we advocate neoadjuvant chemotherapy prior to
surgery.
When there is common hepatic artery or ≤180° superior mesenteric artery
involvement, we advocate neoadjuvant chemotherapy and radiotherapy prior
to surgery. This is because of the effectiveness of modern-era regimens in
triggering tumor regression away from essential vasculature and improving
R0 rates. Operative morbidity can interfere with the delivery of adjuvant
therapy, so upfront chemo- and radio- therapy in an optimally vascularized
and oxygenated field is practical. Further, the neoadjuvant course allows
insight into tumor biology, and better patient selection.
LAPC describes non-metastatic disease with >180° involvement of the
superior mesenteric artery, celiac axis, long segments of the hepatic artery, or
superior mesenteric/portal vein involvement without targets for
reconstruction. No tumor categorized as LAPC on a radiology report should
be considered unresectable until the appropriate imaging is reviewed by an
experienced surgeon, radiologist and multidisciplinary panel.
Misclassification based on non-specific imaging reports is common in our
experience and can be the difference between palliative and curative-intent
therapies. Even when reporting is done according to a standard template,
there is variability in interpretation, confusion over contact (solid versus
stranding), and templates typically will not comment on the length of
involvement. This is the first reason why a specialized surgeon should be
involved in the care of LAPC.
NEOADJUVANT CHEMOTHERAPY
In LAPC, we strongly encourage patients to receive their neoadjuvant therapy
at our center. This allows frequent presentation at multidisciplinary meetings,
a rigorous concentrated approach to the patient, and avoidance of
preventable derailing from the treatment course. If the patient is seen at
another comprehensive cancer center, the surgeon must follow the imaging
closely. Reports of local progression may be meaningless to the final
operative plan, but may be interpreted as progression to unresectability by a
provider not intimately involved with the process, or experienced with
LAPC.
Consistent with NCCN guidelines, we recommend use of either
FOLFIRINOX or gemcitabine-based therapy. No patient being considered
for surgery should receive mono-therapy. If this is all they can
physiologically tolerate, it is highly unlikely they can handle the stress of
operative management. Therapy is continued for approximately two months
(4 cycles of FOLFIRINOX or 2 cycles of gemcitabine/abraxane) before
patients are reimaged, markers are drawn and presentation at
multidisciplinary conference occurs. This is where center experience is
important. Local progression does not warrant regimen change unless there is
a major issue with tolerance. Rarely will radiotherapy be initiated after this
short course, but if there has been rapid regression, radiotherapy and sooner
to surgery may be recommended. The majority of patients go on to complete
another two months of neoadjuvant chemotherapy (4 cycles of FOLFRINOX
or 2 cycles of gemcitabine/abraxane). If they remain operable, then radiation
is pursued. If inoperable, chemotherapy is continued with palliative intent.
NEOADJUVANT RADIATION
In all patients with locally advanced disease, radiation therapy is delivered
after post-chemotherapy imaging rules out metastatic disease. Stereotactic
body radiation therapy or intensity-modulated radiation therapy can be used
at the discretion of the multidisciplinary team. After the delivery of radiation,
we maintain the patient off systemic chemotherapy and perform pancreas
protocolled imaging 4-6 weeks later with immediate plans for surgery in the
absence of metastatic disease.
When treating LAPC, a working assumption is that even the best imaging
will over stage some patients after neoadjuvant therapy. According to the
International Study Group of Pancreatic Surgery, imaging specificity ranges
from 67-91% due to the inadequate differentiation between soft tissue scar
and true cancerous invasion. We achieved R0 resections in 80.4% of 56
patients with locally advanced NCCN-defined unresectable pancreatic
cancer based on cross-sectional imaging after FOLFIRINOX or gemcitabine-
based neoadjuvant therapy. Although tumor shrinkage is desired, this seldom
occurs because of the desmoplastic nature and relative chemo resistance of
pancreatic adenocarcinoma. A combination of failure of progression, decline
in tumor markers, decrease in surrounding inflammation, and changes in
attenuation on imaging are seen as treatment success. Then, the only way to
confirm the presence of malignant tissue around the vasculature is by
exploration, palpation and biopsy as the dissection proceeds.
SURGERY
LAPC by definition involves the neck of the pancreas. Whether the burden of
tumor extends to the right or left dictates whether a pancreaticoduodenectomy
or distal pancreatectomy is indicated, respectively. It is expected that a
pancreaticoduodenectomy will extend just to the left of the aorta, whereas a
distal pancreatectomy will extend to the gastroduodenal artery. The areas of
high risk for extirpative failure, venous inflow and outflow, vascular control,
and potential conduits, need to be planned before entering the operating room
based on detailed study of the imaging. There needs to be clear limits in the
mind of the surgeon before the case begins as to what circumstances the
operation will be terminated before full commitment. This will prevent going
too far and harming the patient if the operation cannot be completed or result
in an R2 resection. Operations are started early in the day when all hospital
resources are available. Laparoscopy is performed to prevent futile
laparotomy. Any suspicious lesions are biopsied for frozen section. If
metastatic lesions are confirmed, additional tissue is obtained for complete
pathological evaluation.
The operation should move immediately to dissection beyond the
extremes of arterial disease, as this will determine whether the resection
proceeds. For example, if a modified Appleby is anticipated, dissection
begins along the cephalad portion of the pancreas widely around the celiac
axis down to the aorta to determine involvement (Figure 2 & 3). The
combination of neoadjuvant therapies and tumor-desmoplasia make tissue
differentiation difficult, so suspicious tissues should be sent for frozen
section to guide operative progression. The objective is to dissect in normal
soft tissue planes, while minimizing violation of tumor planes. Mobilization
of viscera may be necessary during these steps, but transection should be
avoided unless there is a need for bypass. Attention should next be turned to
identifying sites of vascular control. At this point we will perform test
clamps to make sure there is sufficient arterial flow to terminal organs if
arterial resection without reconstruction is expected.
FIGURE 2: Modified Appleby Operation. The original operation
was described for gastric cancer that involves the celiac axis. The
modified Appleby operation is for resection of locally advanced
tumors of the pancreatic body and tail that involve the celiac axis.
In this procedure, the celiac axis is ligated and flow to the hepatic
artery is established via retrograde flow through the intact
pancreaticoduodenal arcade and the gastroduodenal artery (GDA)
to maintain liver perfusion. CA: Celiac Axis; LGA: Left Gastric
Artery; SA: Splenic Artery; CHA: Common Hepatic Artery; GDA:
Gastroduodenal Artery; SMA: Superior Mesenteric Artery; IPDA:
Inferior Pancreaticoduodenal Artery; AIPDA: Anterior Inferior
Pancreaticoduodenal Artery, PSPDA: Posterior Superior
Pancreaticoduodenal Artery; SMV: Superior Mesenteric Vein
FIGURE 3: Completed Modified Appleby Operation
VENOUS RECONSTRUCTION
CONSIDERATIONS
Venous involvement is common and surgeons treating pancreatic cancer must
always be prepared for resections and reconstructions. The distal portal vein
target is almost always available; the portal vein is long and can be extended
by separating it from lymphatics and other structures in the hepatoduodenal
ligament, or by mobilizing the right liver. Although not essential, preserving
the cardiac vein can prevent a problematic gastropathy if the splenic vein
needs to be sacrificed.
The superior mesenteric vein is typically the critical target, because
proximally it is the confluence of multiple tributaries that may need to be
individually anastomosed to a graft. Under these circumstances, an internal
jugular graft with short distances of the middle and superior thyroid veins
can be used for multiple anastomoses. Alternatively, mesenteric veins
leading to the superior mesenteric vein can be joined side-to-side and then
anastomosed around the perimeter to the larger jugular graft. It is important to
examine the proximal jejunum prior to anastomosis. Performing the pancreas,
choledochal and gastric anastomoses to congested or poorly vascularized
intestine is sure to result in leaks and complications, so a further length of
jejunum should be resected. We do not routinely heparinize these
anastomoses during the recovery period, however, if we are concerned with
flow along these tributaries, we will start a heparin gtt in the operating room
with a goal of about 75% therapeutic ptt and discontinue this in the first 48-
72 hours if there is no clinical evidence of thrombosis.
POST-OPERATIVE CARE
After these complex procedures, most patients are recovered in the SICU for
24 hours to prevent hypo- or hypertension, both which risk vascular
anastomoses. Routine post-operative vascular imaging is not performed;
ultrasonography is employed in response to clinical symptoms, abruptly
rising AST/ALT, or excessive ascites drainage. Otherwise, no special steps
are taken with regard to feeding, mobilization, drain management, or
anticoagulation outside of our standard post-operative protocols.
ADJUVANT THERAPY
The decision for adjuvant chemotherapy is based on the success of the
neoadjuvant regimen, and final pathologic interpretation. Thus, this decision
should be made on an individual basis considering nodal status, failure to
achieve an R0 resection, and evidence of lymphovascular or perineural
involvement. Radiation therapy is not repeated in the adjuvant setting.
Adjuvant therapy is initiated within 6-12 weeks. Beyond this point, a plan
based on surveillance imaging and tumor markers should be considered.
CONCLUSION
Only an experienced surgeon who understands the technical details of the
operations for LAPC and inherent risks can identify the potential resectable
patients. As such, they should be the advocate for the patient, monitoring
progress through neoadjuvant therapy, reviewing all imaging, and being
attentive at multidisciplinary conference. The surgeon cannot be a passive
observer, waiting for the patient to be prepared and delivered.
When properly planned, these procedures can be performed safely at high
volume centers. The work of Dr. Joseph G. Fortner should not be used as a
deterrent to more radical pancreatic resections. Since Fortner’s pioneering
work, operative and peri-operative safety have improved. More importantly,
we are in a new era in chemotherapy and targeted radiotherapy, allowing for
better treatment and selection of patients that may benefit from these
operations. Neoadjuvant therapy allows the multidisciplinary committee
approximately 6 months to assess tumor biology, identify distant metastasis,
and properly select patients. Further study is unquestionably necessary.
Randomized operative trials are not feasible, but the comparative group of
inoperable patients is readily available.
SALIENT POINTS
Locally advanced pancreatic cancer (LAPC) must be treated by a
multidisciplinary group with high volume commitment to treating
LAPC, not just experience treating resectable disease.
A radiologist committed to pancreas imaging, definitive in judgement,
and versed in pancreatic surgery possibilities is an essential team
member. Imaging should be standardized and follow a structured
reporting template.
Decisions about surgery for pancreatic cancer cannot be age-based.
Careful consideration of comorbidities and performance status are
necessary.
FOLFIRINOX or gemcitabine-based neoadjuvant therapy followed by
radiation therapy is necessary before considering curative-intent
surgery for LAPC.
Patients must be presented at multidisciplinary conference every time
imaging is performed.
The specificity of contrast-enhanced cross sectional imaging to
differentiate between soft tissue scar and cancerous vascular invasion
is limited.
Final operative decisions are based on multidisciplinary review and an
identified plan to achieve an R0 margin
There needs to be clear limits in the mind of the surgeon before the
case begins as to what circumstances the operation will be terminated
before full commitment. This will prevent going too far and harming
the patient if the operation cannot be completed or result in an R2
resection.
Irreversible electroporation is an increasingly popular therapy for
LAPC. The technology is still being studied with regard to safety and
mid- and long-term efficacy.
SELECTED REFERENCES
1. Michalski CW, Kleeff J, Bachmann J, Alkhatib J, Erkan M, Esposito I,
et al. Second-look operation for unresectable pancreatic ductal
adenocarcinoma at a high-volume center. Ann Surg Oncol.
2008;15(1):186-92.
2. Al-Hawary MM, Francis IR, Chari ST, Fishman EK, Hough DM, Lu
DS, et al. Pancreatic Ductal Adenocarcinoma Radiology Reporting
Template: Consensus Statement of the Society of Abdominal Radiology
and the American Pancreatic Association. Radiology 2014; 270(1);
248-60.
3. Bockhorn M, Uzunoglu FG, Adham M, Imrie C, Milicevic M, Sandberg
AA, et al. Borderline resectable pancreatic cancer: a consensus
statement by the International Study Group of Pancreatic Surgery
(ISGPS). Surgery. 2014;155(6):977-88.
4. Kluger MD, Rashid MF, Rosario VL, Schrope BA, Steinman J, Hecht
EM, Chabot JA. Resection of locally advanced pancreatic cancer
without regression of arterial encasement after modern-era neoadjuvant
therapy. Journal of Gastrointestinal Surgery. Accepted 8/13/17. PMID:
28895032
█ BACKGROUND
Pancreatic cancer remains the fourth leading cause of cancer deaths in both
men and women in the United States. In 2017, there will be an estimated
53,670 new cases of pancreatic cancer, with 43,090 estimated deaths. As life
expectancy in the United States increases and the population ages, an
increasing numbers of older patients are diagnosed with pancreatic cancer.
Currently, the median age of diagnosis is 70 years old, with 80% of patients
over the age of 60 and approximately 13% over the age of 84.
The treatment for pancreatic cancer primarily depends on the resectability
at presentation (Table 1 , Figure 2 ), determined by the presence or absence
of vascular invasion or distant metastatic disease. Approximately one third
of patients present with locoregional disease (resectable, borderline
resectable, locally unresectable) and even fewer have truly resectable
disease without significant involvement of the superior mesenteric vein
(SMV), portal vein (PV), superior mesenteric artery (SMA), or celiac axis at
the time of presentation. In resectable patients, multimodality therapy with
surgery and adjuvant or neoadjuvant chemotherapy offers the best survival.
While the overall 5-year survival is only 7%, multimodality therapy in
patients with resectable disease has a 2-year survival of 41%.
FIGURE 2: Algorithm showing the management options for
octogenarian patients with pancreatic cancer by resectability status
and functional status. Note that age is not a factor in the algorithm.
MULTIMODALITY THERAPY
Although surgery is essential for curative treatment, adjuvant chemotherapy
has been shown to improve survival. Given the high incidence of recurrence
and poor survival in pancreatic cancer, it is thought to be a systemic disease
at the time of presentation. Therefore, in resectable patients, multimodality
therapy with surgery and adjuvant or neoadjuvant chemotherapy offers the
best survival and is the current standard of care. There is significant debate
over whether chemotherapy should be delivered in the adjuvant or
neoadjuvant setting, with pros and cons to each approach. The approaches
are difficult to compare, as most studies are not analyzed on an intent-to-treat
basis and randomized trials have been difficult especially in resectable
disease. However, recent population-based data suggests that survival is
similar in patients who receive both modalities, regardless of timing.
Multimodality therapy for locoregional disease in all patients has increased
over the last decade, from 31.3% in 2004 to 37.9% in 2011 (P<0.0001).
However, rates are lower in older in patients. Only 11% of Medicare
beneficiaries with locoregional pancreatic cancer patients received
multimodality therapy, with chemotherapy delivered in an adjuvant setting in
93.1% and neoadjuvant setting in 6.9%.
ADJUVANT THERAPY
Adjuvant treatment has been shown to improve survival in multiple
controlled trials including ESPAC-1, CONKO-001, ESPAC-3, RTOG 9704,
and GITSG. However, older patients are less likely to receive adjuvant
therapy following surgery. In Medicare patients undergoing surgery first,
more than half (51.2%) did not receive adjuvant chemotherapy. The underuse
of adjuvant therapy in older patients is likely multifactorial. The higher
complication rates following surgery may delay or preclude adjuvant
therapy. Additionally, it has been our experience that many older patients
decline adjuvant therapy when offered, as in our case scenario.
NEOADJUVANT THERAPY
While there is significant debate over the timing of chemotherapy, of those
receiving multimodality therapy, >90% of patients nationally receive it in the
adjuvant setting. While adjuvant therapy improves survival, postoperative
complications can delay or prevent receipt of chemotherapy, especially in the
older population. Over the last decade, the use of neoadjuvant therapy in
truly resectable disease has increased. Moreover, neoadjuvant therapy is
favored in the setting of borderline resectable disease; neoadjuvant therapy
can convert patients to resectable disease and has been associated with
higher R0 resection rates. This approach may be beneficial in older patients,
especially in those who present a high surgical risk.
In a retrospective study evaluating the receipt of multimodality therapy
over a nine-year period (2000-2008), 179 patient >70 years of age were
treated for resectable or borderline resectable pancreatic cancer; 153 (85%)
were treated with neoadjuvant chemoradiation with intent to resect, and 26
(15%) underwent surgery first. In the neoadjuvant group, 48% completed
multimodality therapy, while 52% progressed during chemotherapy or had
poor performance status and would not tolerate surgery. In the initial surgical
resection group, 11 of 26 (42%) received adjuvant therapy. Survival was
similar between the two groups. Lack of progression during chemotherapy
helps select the patients who will most likely benefit from surgery. If disease
progresses, surgical resection can be avoided. A neoadjuvant approach
allows the disease biology to become evident. Finally, the older patient’s
ability to tolerate chemotherapy may be a marker of their physiologic reserve
and predict surgical outcomes.
Newer neoadjuvant regimens such as FOLFIRINOX are more effective
but not well tolerated, especially in older patients. FOLFIRINOX is
associated with a higher incidence of adverse effects including neutropenia,
febrile neutropenia, thrombocytopenia, diarrhea, and sensory neuropathy.
FOLFIRINOX is only recommended in young patients with good cardiac
function and normal bilirubin levels. Additionally, older patients are
underrepresented in studies of neoadjuvant and adjuvant therapy.
CHEMORADIATION IN NONSURGICAL
CANDIDATES
In older patients who are not surgical candidates, either for locally
unresectable disease or poor functional/performance status, chemoradiation
is also possible. Chemoradiation without surgery is not curative, but may
slow tumor growth or cause tumor regression. It may also palliate symptoms
associated with unresected pancreatic cancer including gastrointestinal
bleeding due to tumor invasion of the duodenum. Chemoradiation prolongs
survival for months. Patients that undergo chemoradiation for unresectable
disease experience multiple side-effects of treatment, such as with
FOLFIRINOX. Overall, chemoradiation alone prolongs survival less than
surgery alone.
PALLIATION
In some cases, despite resectable disease, patients may not be candidates for
surgical resection. In older patients with a poor performance status, frailty,
poor functional status, and severe medical comorbidities, surgery may not be
an option. In addi tion, one must consider the patient’s preferences and goals
of care. Given the poor prognosis of pancreatic cancer and the toxicity
associated with treatment, the decision often balances the quality and quantity
of life. Many older patients may choose to forego aggressive treatment given
these factors.
Once the decision is made to not pursue aggressive therapy in the case of
resectable (or unresectable disease), palliation becomes tantamount. The
goal of palliative care is to improve quality of life. Most common
complications associated with pancreatic cancer are related to involvement
of adjacent structures and include obstructive jaundice, gastric outlet
obstruction, and intractable pain from celiac plexus tumor infiltration.
One half to three quarters of patients with pancreatic cancer develop
obstructive jaundice. With advances in endoscopy, most patients with
obstructive jaundice who are not candidates for surgical resection can be
managed endoscopically. Endoscopic stent placement has a success rate of
>90%. The median patency for plastic stents is 2-4 months; they are prone to
obstruction and bacterial overgrowth and should only be placed in the setting
of planned resection or short expected survival. In cases of palliation without
planned surgical resection in patients with a projected survival of 4 to 6
months, self-expanding metal stents are recommended due to longer patency
(7-10 months).
In the rare setting of failed endoscopic management, percutaneous
transhepatic biliary drainage is an alternative, nonoperative procedure. With
percutaneous transhepatic drainage, the bile ducts are accessed through the
liver parenchyma and a wire is passed through the obstruction. Once the
obstruction is crossed, metal stents can be place and the internal/external
drains removed. If the site of obstruction is not passed, obstruction can only
be relieved by proximal drainage and will be associated with significant
daily fluid and bile salt loss.
Gastric outlet obstruction is another common complication of unresected
pancreatic cancer. Historically, this was managed surgically with a
gastrojejunostomy; however, in older patients who are poor surgical
candidates, this is ideally avoided. Endoscopic stent placement for outlet
obstruction is now an option and is often successful in the short-term with
approximately 90% of patients resuming diet within 24 hours. Stent migration
can occur in approximately 27% of patients, therefore it is recommended to
stent only symptomatic patients. Stent occlusion is also common and most
patients can tolerate only clear or full liquid diets, but they can at least
manage their own secretions. Newer endoscopic technology allows for
creation of an endoscopic gastrojejunostomy, but this technology is in its
infancy, without good outcomes data. In our preliminary experience,
complication rates are high, but when successful, patients experience better,
more durable return to diet with endoscopic gastrojejunostomy compared to
stenting. Furthermore, surgery is avoided.
Finally, invasion of celiac plexus will produce severe abdominal and
back pain. In addition to oral analgesics, pain can be managed with
neurolysis of celiac access. Overall, patients have decreased pain at four and
eight weeks post procedure and consume decreased amount of opioids
overall.
Hospice involves a team-oriented approach to medical care, pain
management, and emotional and spiritual support expressly tailored to the
individual patient during the final stage of serious illness. Palliative care and
hospice teams help clarify patient and family goals of care, and the benefits
of hospice care have been well established. In a population-based study of
older patients who died of pancreatic cancer, hospice was underused. Only
59.6% of patients who died of pancreatic cancer were enrolled in hospice
care, and only 35.9% were enrolled for four weeks or more. In addition,
aggressive care at the end of life increased over time, and admission to the
ICU and receipt of chemotherapy during the last month increased from 15.5
to 19.6 % (P<.0001) and from 8.1% to 16.4% (P<0.001), respectively.
Hospice care should be considered early in the course once patients decide
against aggressive therapy.
CONCLUSION
We present two cases to highlight the challenge. In both, the patient was over
80 years of age, but in good health and living independently. In the latter
case, the patient has a better prognosis tumor. In our first clinical case, the
patient was initially told she was not a surgical candidate, and she could
have missed an opportunity to live an additional six years. She understood
the risks of surgery and chose to proceed. However, she declined adjuvant
therapy despite healing from surgery. She ultimately succumbed to pancreatic
cancer but enjoyed good quality of life in the interim. In the second case, the
patient failed to thrive and died shortly after surgery, despite readmission
and no obvious postoperative complications. He was never able to return
home and back to his prediagnosis quality of life. In fact, surgery shortened
his life significantly. These cases highlight the difficulty in making these
decisions. Despite all we know, we cannot reliably predict who will benefit
and who will not.
The management of resectable pancreatic cancer in older patients
requires careful consideration of the staging of disease; the patient’s frailty,
functional status, and comorbidities; the disease prognosis; as well as the
patient’s treatment goals and preferences. All patients with locoregional
pancreatic cancer should undergo surgical evaluation by a pancreatic
surgeon, as patients cannot make an informed decision if they are not fully
evaluated, not offered all the options, and do not understand the risks and
benefits of each option. In discussing surgery, age alone should not dictate the
patient’s management, as surgical benefit does not diminish with age. Older
patients with good functional status (living at home, no dementia, tending to
their own activities of daily living) should be considered for surgical
resection. It is critical to communicate the risks and disclose the potentially
higher morbidity, mortality, and failure-to-rescue rates. In addition, we
should discuss the long-term survival/prognosis as well as the potential
impact on quality of life, including inability to return to home or independent
living after surgery. The impact of complications is difficult to communicate
unless a person has experienced them. Older patients should also understand
the benefits of surgical resection and multimodality therapy.
Multidisciplinary care teams should be involved in the care of these complex
and vulnerable patients and the surgical and multimodality therapy options,
risks, and benefits should be discussed with the multidisciplinary team, the
patient, and the family. Based on the data presented, surgical resection should
be directed to high-volume centers and/or surgeons to optimize outcomes in
this vulnerable group when aggressive therapy is reasonable and chosen by
the patient and treatment team. Only after a complete discussion of the
benefits and risks of procedure as well as alternatives, can a patient can
make a decision that is truly right for them. This decision will be different for
each patient.
SALIENT POINTS
In discussing surgery, age alone should not dictate a patient’s
management.
The morbidity, mortality, and failure-to-rescue rates are higher in older
patients undergoing surgical resection for pancreatic cancer. Despite
these facts, older patients can still derive a survival benefit from
surgical resection.
The decision to resect an older patient with pancreatic cancer must
consider the complete picture, including tumor characteristics and
prognosis, patient comorbidity/ability to tolerate surgery, the opinion
of the multidisciplinary care team, and patient preferences and goals of
treatment.
Multimodality therapy is the standard of care, but many factors
challenge our ability to fully deliver this in older patients.
Consider neoadjuvant in older patients; this will allow the tumor
biology to become clear and avoid surgery in patients who will likely
not benefit.
While the current data suggest underuse of surgical resection this, in
part, reflects good patient selection.
When surgery is performed, it should be done at high-volume centers.
SELECTED REFERENCES
1. Parmar AD, Vargas GM, Tamirisa NP, Sheffield KM, Riall TS.
Trajectory of care and use of multimodality therapy in older patients
with pancreatic adenocarcinoma. Surgery. 2014; 156(2):280-9.
2. Bilimoria KY, Bentrem DJ, Ko CY, Stewart AK, Winchester DP,
Talamonti MS. National failure to operate on early stage pancreatic
cancer. Ann Surg. 2007; 246(2):173-80.
3. Riall TS, Sheffield KM, Kuo YF, Townsend CM, Goodwin JS.
Resection Benefits Older Adults with Locoregional Pancreatic Cancer
Despite Greater Short‐Term Morbidity and Mortality. J Am Geriatr
Soci. 2011;59(4):647-54.
4. Finlayson E, Fan Z, Birkmeyer JD. Outcomes in octogenarians
undergoing high-risk cancer operation: a national study. J Am Coll of
Surg. 2007;205(6):729-34.
5. Riall TS, Reddy DM, Nealon WH, Goodwin JS. The Effect of Age on
Short-Term Outcomes After Pancreatic Resection: A Population-based
Study. Ann Surg. 2008; 248(3):459-67.
6. Scurtu R, Bachellier P, Oussoultzoglou E, Rosso E, Maroni R, Jaeck D.
Outcome after pancreaticoduodenectomy for cancer in elderly patients.
J of Gastrointest Surg. 2006;10(6):813-22.
7. Tamirisa NP, Parmar AD, Vargas GM, Mehta HB, Kilbane EM, Hall
BL, et al. Relative contributions of complications and failure to rescue
on mortality in older patients undergoing pancreatectomy. Ann of Surg.
2016; 263(2):385-91.
8. Dimou F, Sineshaw H, Parmar AD, Tamirisa NP, Jemal A, Riall TS.
Trends in Receipt and Timing of Multimodality Therapy in Early-Stage
Pancreatic Cancer. J Gastrointest Surg. 2016;20(1):93-103.
9. Sheffield KM, Boyd CA, Benarroch‐Gampel J, Kuo YF, Cooksley CD,
Riall TS. End‐of‐life care in Medicare beneficiaries dying with
pancreatic cancer. Cancer. 2011;117(21):5003-12.
█ BACKGROUND
PREOPERATIVE CONSIDERATIONS
Approximately 65-75% of patients suffering from pancreatic cancer develop
symptomatic biliary obstruction. Debate exists about whether patients
experience higher mortality following pancreatoduodenectomy in the setting
of hyperbilirubinemia. Further, controversy remains over whether to
manipulate the biliary system preoperatively with stent placement given
associations with morbidity and mortality postoperative, including higher
surgical site infection rates and wound-related complications.
At our institution, we follow current National Comprehensive Cancer
Network (NCCN) guidelines for preoperative biliary drainage and treat
symptomatic, obstructive hyperbilirubinemia preoperatively to improve
symptoms and liver function. We recommend preoperative biliary
decompression whenever feasible for symptomatic obstructive jaundice. If a
biliary stent cannot be successfully placed via endoscopic technique, we
recommend percutaneous biliary drainage. The NCCN also recommends
preoperative biliary decompression among septic, coagulopathic patients, or
those with impaired renal function prior to surgery. Additionally,
symptomatic patients suspected to have delays in surgery >1 week are also
stented for decompression. We are a high-volume center offering neoadjuvant
therapies for pancreatic adenocarcinoma and routinely place a stent for
preoperative biliary drainage in this sort of patient prior to undergoing
neoadjuvant therapy.
Optimization
If your patient is optimized for surgery, biliary stenting for decompression is
indicated for symptomatic obstructive jaundice in this patient with a bilirubin
of 20 or greater. Should other comorbidities or risk assessment require
further testing or intervention prior to surgery, we recommend biliary decom
pression with stenting while further tests or interventions are performed
preoperatively.
SELECTED REFERENCES
1. House MG, Choti MA. Palliative therapy for pancreatic/biliary cancer.
Surg. Clin. North Am. 2005;85(2):359-71.
2. National Comprehensive Cancer Network (NCCN). Pancreatic Cancer
(Version1.2017).2017;https://www.nccn.org/professionals/physician_gl
s/pdf/pancreatic.pdf. Accessed March 21, 2017.
█ BACKGROUND
Surgical resection provides the only chance for prolonged survival and cure
in patients with pancreatic adenocarcinoma. Pancreatectomy has become
safer over the last three decades due to advancements in surgical technique
and critical care. As experience has grown, some surgeons have taken a more
aggressive approach to locally advanced tumors of the pancreas. While
significant strides have been made, post-operative morbidity remains high
and performing multi-visceral resection during pancreatectomy remains
controversial. Proponents argue that the dismal outcomes provided by other
local therapies necessitates an aggressive surgical approach in order to
afford the best chance for prolonged survival for the patient. Others would
suggest that tumor invasion into major vascular structures or adjacent organs
is a sign of poor tumor biology and portends a dismal outcome irrespective
of an aggressive surgical approach. Herein, we discuss our experience in
managing this complex clinical scenario and how this compares to the
limited data in the literature.
An analysis of patients at Indiana University between November 2006
and May 2011 who underwent pancreatectomy (PD, distal pancreatectomy,
and total pancreatectomy) and simultaneous colectomy (P+C) were analyzed
and outcomes compared to patients who underwent pancreatectomy alone
(P). Of 840 patients who underwent pancreatectomy, twenty underwent
pancreatectomy with concomitant colectomy (2.4%).
Pancreatoduodenectomy was performed in 9 patients (9/519, 1.7%) and
distal pancreatectomy in 11 (11/301, 3.7%). Outcomes are shown in Table 1 .
Mortality appeared to be increased following P+C compared to P alone
(10% vs 3% NS), and there was a statistically significant increase in
morbidity (70% vs 33%, p < 0.01). The increased morbidity was largely due
to increases in organ space infection (Figure 2 ). Given the substantial
increases in both morbidity and mortality, we concluded that careful patient
selection and strategies to prevent and/or control anastomotic leaks are
necessary to improve outcomes in patients undergoing simultaneous
pancreatectomy and colectomy.
SALIENT POINTS
Pancreatoduodenectomy with colectomy is considered an aggressive
surgical therapy given the potential for increased morbidity and paucity
of data showing a clear survival benefit
While select patients may benefit, our practice is to refer patients for
consideration of neoadjuvant therapy prior to consideration of surgery
Patients should be treated at a high-volume pancreatic center and
managed by a multidisciplinary team
SELECTED REFERENCES
1. House MG, Kilbane M, Ceppa EP, Nakeeb A, Howard TJ, Schmidt
CM, et al. Simultaneous Pancreatectomy and Colectomy: A Safe
Combination? Paper presented at: Pancreas Club Annual Meeting 2012;
San Diego, CA.
2. Temple SJ, Kim PT, Serrano PE, Kagedan D, Cleary SP, Moulton CA,
M, et al. Combined pancreaticoduodenectomy and colon resection for
locally advanced peri-ampullary tumours: analysis of peri-operative
morbidity and mortality. HPB (Oxford). 2014;16(9):797-800.
3. Harris JW, Martin JT, Maynard EC, McGrath PC, Tzeng CW. Increased
morbidity and mortality of a concomitant colectomy during a
pancreaticoduodenectomy: an NSQIP propensity-score matched
analysis. HPB (Oxford). 2015;17(9):846-54.
4. Kimchi ET, Nikfarjam M, Gusani NJ, Avella DM, Staveley-O’Carroll
KF. Combined pancreaticoduodenectomy and extended right
hemicolectomy: outcomes and indications. HPB (Oxford).
2009;11(7):559-64.
█ BACKGROUND
Chronic pancreatitis is marked by severe, intractable, and debilitating
abdominal pain with associated devastating impairments in quality of life.
Patients may develop attendant diabetes and malnutrition due to endocrine
and exocrine pancreatic failure, respectively. Frontline therapies include
addressing inciting factors for inflammation and fibrosis, pain control, and
medical support. Intervention is undertaken for pain relief. Endoscopic
procedures may be employed in an attempt to optimize organ drainage.
Surgical therapies are considered once medical and endoscopic interventions
fail. Operative algorithms are based primarily on ductal anatomy and organ
morphology. Patients with a dilated main pancreatic duct (>6 mm) can
benefit from operative drainage (lateral pancreaticojejunostomy or localized
pancreatic head resection combined with lateral pancreaticojejunostomy).
Patients with small pancreatic ducts may experience pain relief with focused
resection of a dominantly affected pancreatic head or tail. Some patients,
however, have diffuse, small duct pancreatitis. These patients may benefit
from total pancreatectomy with islet autotransplantation (TPIAT) (Figure 2,
3 ). TPIAT can also be effective for pain relief and improvements in quality
of life in patients with genetic pancreatitis and as salvage therapy in patients
who have failed lesser surgeries.
SURGICAL EVALUATION
In particular, patient selection is recognized as being paramount to a
successful outcome after TPIAT, but is still being elucidated. Patients
presenting for evaluation for TPIAT are a heterogeneous and complex group
of patients. They are marked by debilitating abdominal pain with associated
poor quality of life and attendant social dysfunction; they have encountered
(and failed) multiple medical and endoscopic interventions; and they often
have complicated medical histories. Current best practice includes following
strict criteria for patient selection to better sort through these complex
patients and to optimally apply this invasive treatment option.
Genetic pancreatitis
Genetic causes for pancreatitis are increasingly being recognized. In 1996,
Whitcomb and colleagues described the mutated cationic trypsinogen gene
(PRSS1) as the cause for hereditary pancreatitis, an autosomal dominant
disorder with 80% penetrance conveying the inappropriate activation of
trypsin. Patients affected develop recurrent acute pancreatitis and chronic
pancreatitis often beginning in early childhood and carry a significantly (up
to 70 times) increased risk for pancreatic cancer, particularly with smoking.
Since then, several other genes have been identified that predispose patients
to pancreatitis, predominantly through the inappropriate activation or
regulation of trypsin. Mutations of the pancreas secretory trypsin inhibitor
gene (serine protease inhibitor Kazal type 1, SPINK-1) and of the
chymotrypsin C gene (CTRC) cause impaired regulatory inhibition of trypsin
activity. Mutations in the calcium-sensing receptor gene (CASR) result in
altered intracellular calcium regulation and mutations of the gamma glutamyl
transferase 1 gene (GGT1) cause dysregulation of cellular glutathione levels,
both of which lead to activated trypsin and a predisposition towards
pancreatitis. Some mutations in the cystic fibrosis transmembrane
conductance regulator gene (CFTR) can compromise pancreatic ductal
flushing leading to pancreatitis. Finally, the claudin gene locus (CLDN2) on
the X-chromosome has been associated with alcoholic pancreatitis, which
suggests that even environmentally related etiologies for pancreatitis may
have an underlying genetic component. Currently PRSS1, SPINK1, CFTR,
and CASR mutations have readily commercially available tests and should
be a routine part of the evaluation for TPIAT. The presence of a pancreatitis-
associated gene mutation may support a diagnosis of chronic pancreatitis in a
patient with equivocal imaging, and may influence the decision for a TPIAT
as an initial surgical intervention rather than a partial resection or drainage
procedure.
PAIN ASSESSMENT
The clinical hallmark of pancreatitis is severe intractable abdominal pain.
This pain is the primary factor for determining intervention. Patients
considered for TPIAT should have pain that is significantly disruptive to
their work, school, family, and societal roles, as well as impairments in
quality of life. While patients with chronic pancreatitis, and in particular
hereditary pancreatitis, have an increased risk for pancreatic cancer, current
evidence does not support TPIAT in patients without pain for cancer
prevention.
During the preoperative optimization period, patients should have their
pain management optimized, both with pharmacological and behavioral
medicine support. Identifying a pain management provider during the
perioperative and postoperative period is ideal.
CONSIDERATION OF LESSER
INTERVENTIONS
Surgery is not the frontline therapy for pancreatitis, and thus medical
optimization and endoscopic options should be thoroughly explored as an
initial step. In patients who fail the nonoperative approach, surgery is
considered based on organ morphology and ductal anatomy. In patients with
dilated ducts, a drainage procedure is appropriate. In patients with
nondilated ducts, a targeted resection is indicated. Patients offered total
pancreatectomy are those that have small ducts and diffuse disease. The
potential exception to this principle is the patient with hereditary pancreatitis
(PRSS1 mutation). Once these patients demonstrate the pancreatitis
phenotype, they typically have a progressive disease course, and
consideration of TPIAT as the initial intervention appears warranted. Less is
currently known about the natural history of the other genetic etiologies of
pancreatitis, and thus the classic algorithm should not be altered for these
patients.
DETERMINATION OF PHYSIOLOGICAL
FITNESS
As in all elective major abdominal surgeries, patients need to be risk
stratified and medically optimized prior to considering surgical intervention,
as comorbidities are often the ultimate determinant of outcome.
Cardiopulmonary comorbidities should be evaluated. It is notable that
comorbid hepatic disease is a relative contraindication to islet transplant, as
the islet infusion is an embolic event to the terminal branches of the portal
vein. This normally small physiologic event can be significant in the setting
of underlying hepatic compromise.
Chronic pancreatitis patients are often malnourished on presentation due
to exocrine pancreatic insufficiency and pain, causing poor oral intake. A
formal nutritional evaluation and support is important prior to TPIAT.
Pancreatic enzyme replacement therapy should be instituted preoperatively to
assess patient tolerance of medications as well as to prehabilitate them for
surgery.
DIABETIC PATIENTS
Patients presenting for TPIAT may have comorbid diabetes, either from
pancreatitis-related endocrine failure, or due to other etiologies. Oral
glucose or mixed-meal tolerance testing can assess preoperative beta cell
function. In patients requiring insulin on initial evaluation, a stimulated C-
peptide study can assess the patient’s ability to synthesize endogenous
insulin. Patients with a stimulated C-peptide less than 0.7 nmol/L likely do
not have sufficient islet function to warrant the potential morbidity and cost
of an islet autotransplant.
DETERMINATION OF PSYCHOLOGICAL
FITNESS
Patients with chronic pancreatitis and chronic pain being evaluated for
TPIAT have significant attendant psychological comorbidities. There is a
high prevalence of depression (68%), opioid misuse (48%), and impaired
psychological quality of life. Patients often have poor coping mechanisms.
Behavior health psychology evaluation and support are essential prior to
surgery and postoperatively to optimize patient selection and outcomes. In
addition, behavior health specialists can teach patients coping skills and
adjunctive pain relief measures including biofeedback and guided imagery
that can improve opportunities for a successful perioperative period.
SALIENT POINTS
TPIAT is an effective means of pain relief and improves quality of life
in selected patients with chronic pancreatitis.
Patient selection is fundamental, but can be challenging in this complex
and heterogeneous patient group.
Following specific guidelines can facilitate optimal outcomes and is
essential as this emerging clinical modality evolves and is elucidated.
SELECTED REFERENCES
1. Sutherland DE, Radosevich DM, Bellin MD, Hering BJ, Beilman GJ,
Dunn TB, et al. Total Pancreatectomy and Islet Autotransplantation for
Chronic Pancreatitis. J Am Coll Surg. 2012; 214(4): 409-24.
2. Morgan K, Owczarski SM, Borckardt J, Madan A, Nishimura M,
Adams DB. Pain Control and Quality of Life After Pancreatectomy with
Islet Autotransplantation for Chronic Pancreatitis. J Gastrointest Surg.
2012; 16(1): 129-34.
3. Sutherland DE, Matas AJ, Najarian JS. Pancreatic islet cell
transplantation. Surg Clin North Am.1978; 58(2): 365-82.
4. Walsh TN, Rode J, Theis BA, Russell RCG. Minimal change chronic
pancreatitis. Gut. 1992; 33(11): 1566-71.
5. Wilson GC, Sutton JM, Smith MT, Schmulewitz N, Salehi M, Choe KA,
et al. Total pancreatectomy with islet cell autotransplantation as the
initial treatment for minimal-change chronic pancreatitis. HPB. 2015;
17(3): 232-8.
6. Whitcomb DC, Gorry MC, Preston RA, Furey W, Sossenheimer MJ,
Ulrich CD, et al. Hereditary pancreatitis is caused by a mutation in the
cationic trypsinogen gene. Nat Genet. 1996; 14(2):141-5.
7. Morgan KA, Borckardt J, Balliet W, Owczarski SM, Adams DB. How
are chronic pancreatitis patients selected for total pancreatectomy with
islet autotransplantation? Are there psychometric predictors? J Am Coll
Surg. 2015; 220(4): 693-8.
█ BACKGROUND
Chronic pancreatitis is a progressive inflammatory disorder characterized by
the hallmark symptom of persistent, disabling abdominal pain. In addition,
due to fibrosis and irreversible destruction of the pancreatic parenchyma,
patients can exhibit metabolic sequelae of endocrine and exocrine
insufficiency. Regardless of the disease etiology, environmental versus
genetic, more than half of all patients will require surgical intervention.
The most common indication for surgical intervention is intractable pain
in spite of optimal medical management. Multiple hospitalizations for acute
or chronic pancreatitis episodes interfere with a patient’s quality of life, and
secondary complications of pancreatic fibrosis (obstruction of the pancreatic
or common bile duct, and intestinal obstruction) and duct rupture (formation
of pseudocyst or pancreatic ascites) may also warrant surgical exploration.
In addition, the concomitant suspicion of malignancy reinforces the decision
to proceed with surgery.
Given the multifactorial etiology and pathophysiology of chronic
pancreatitis, various surgical options exist. Although randomized controlled
trials (RCT) have established the superiority of surgery over endoscopic
treatment, much debate exists with regard to the optimal operative approach.
Surgical techniques for chronic pancreatitis can be divided into two general
categories: drainage/decompressive procedures and resection procedures. A
cross-sectional study by van der Gaag et al. of 223 patients who underwent
surgical intervention for chronic pancreatitis demonstrated long-term pain
relief in two-thirds of patients, when the procedure was tailored to the
anatomical abnormalities. Thus, a surgeon dedicated to treating patients with
chronic pancreatitis should be facile with tailoring the operative approach to
the patient’s pancreatic parenchymal and ductal anatomy and should be
skilled in the available operative techniques. This chapter will highlight the
steps critical to working up a patient with head-dominant chronic pancreatitis
for surgery (Figure 2 ) and define the current operative procedures as well
as their indications, advantages, and disadvantages.
PREOPERATIVE WORKUP
The majority of patients will already have had numerous imaging studies
performed for acute exacerbations of pancreatitis. Prior to any surgical
intervention, the surgeon must be familiar with the patient’s pancreatic
parenchymal and ductal anatomy. Various imaging modalities are utilized to
gain this information. Computed tomography (CT) plays a key role in the
global assessment of chronic pancreatitis. It allows for delineation of
pancreatic parenchymal changes through the use of intravenous iodinated
contrast. The pancreas is enhanced homogenously after administration of
contrast, with optimal enhancement occurring 35-40 seconds after contrast
injection.
In addition to detecting parenchymal atrophy and calcifications, CT is
also useful in detecting pseudocysts. Multiplanar reconstruction and thin
slices allow for optimal visualization of the pancreatic duct and any
associated changes in caliber. The portal venous phase of enhancement
allows for detection of vascular complications, such as pseudoaneurysms or
venous thrombi.
Similar to CT, magnetic resonance imaging (MRI) allows for a global
assessment of the pancreas parenchyma and ductal anatomy. Normal pancreas
appears bright on T1-weighted noncontrast images, with loss of signal
alluding to fibrosis. Magnetic resonance cholangiopancreatography (MRCP)
uses T2-weighted images to evaluate ductal anatomy. Its sensitivity is
improved with secretin administration, which promotes duct distension (at
least 1 mm in the main pancreatic duct) by increasing pancreatic secretions
and sphincter of Oddi contraction.
Although MRCP and secretin administration improves assessment of the
ductal anatomy, it may be insufficient for small branch pancreatic ducts.
Ultimately, endoscopic retrograde cholangiopancreatography remains the
gold standard for assessing the pancreatic duct and may also be therapeutic
when complications are detected, such as obstruction of the right-sided main
pancreatic duct.
Finally, endoscopic ultrasound (EUS) is another useful modality in
assessing pancreatic parenchyma and ductal abnormalities. Due to the
increased risk of developing pancreatic adenocarcinoma in patients with
chronic pancreatitis compared to the general population, EUS with biopsy
may be useful in obtaining tissue diagnosis when there is a change in
appearance of the pancreas on serial imaging that is concerning for
development of cancer.
Resection Procedures
PANCREATICODUODENECTOMY
Traditionally, pancreaticoduodenectomy (PD) and pylorus-preserving PD
have been the main surgical approaches for chronic pancreatitis with head
enlargement and no upstream ductal dilatation, as the head and uncinate
process are viewed as the driver of chronic pancreatitis. At the University of
Cincinnati, we perform pancreatic head resection if the main pancreatic duct
is < 7 mm in caliber. The main argument in support of resection is that an
inadequate drainage procedure does not address the primary goal of pain
relief when leaving involved parenchyma behind.
Although the step-to-step details of PD vary, the procedure consists of
three anastomoses: pancreaticojejunostomy, choledochojejunostomy, and
duodeno- or gastrojejunostomy. When performed at high-volume centers, the
procedure is associated with mortality rates < 2%, though morbidity rates
remain up to 50%. In comparison to patients with periampullary
malignancies, those with chronic pancreatitis develop firm, fibrotic
parenchyma which facilitates anastomosis with the small bowel, resulting in
reduced rates of pancreatic fistulas and potentially less morbidity.
Pancreaticoduodenectomy Technique
A bilateral subcostal incision or midline incision is made to enter the
abdomen. After an initial abdominal exploration, the lesser sac is entered,
and the hepatic flexure is mobilized off the duodenum. Adhesions from the
pancreas head to the stomach, if present, are taken down and the
infrapancreatic SMV is identified. Next, an extended Kocher maneuver is
performed to mobilize the duodenum and pancreatic head to the level of the
superior mesenteric vein anteriorly and left renal vein posteriorly.
The portal dissection begins with a cholecystectomy. The common bile
duct is isolated, and the common hepatic duct is divided just above the
junction with the cystic duct. Next, the distal common bile duct is dissected
to the pancreatic head, and the gastroduodenal artery is identified and
ligated. The portal vein is identified, and the avascular plane between the
anterior portal vein and posterior pancreas is developed. Once this is
completed, the stomach is divided (or duodenum if pylorus-preservation is
desired), as well as the jejunum approximately 8-10 cm distal to the ligament
of Treitz. This is followed by division of the pancreas along the created
dissection plane from the superior mesenteric vein to its confluence with the
portal vein.
The reconstruction is performed by first performing a retrocolic end-to-
side pancreaticojejunostomy followed by a retrocolic end-to-side
hepaticojejunostomy. Finally an antecolic gastrojejunostomy (Figure 3 ) or
duodenojejunostomy (Figure 4 ) is performed. We selectively place feeding
jejunostomy or gastric tubes, taking into account the patient’s preoperative
nutrition status and prior use of parental nutrition, and weigh this with the
risk of developing delayed gastric emptying (DGE) or likelihood of needing
to supplement poor oral intake.
FIGURE 3: In the “classic” Whipple, the pylorus and distal part of
the stomach are removed along with the pancreatic head, and
reconstruction consists of a retrocolic end-to-side
pancreaticojejunostomy, followed by a retrocolic end-to-side
hepaticojejunostomy, and an antecolic gastrojejunostomy.
FIGURE 4: In the pylorus-sparing Whipple, the stomach and
pylorus remain intact, and reconstruction consists of a
duodenojejunostomy instead of a gastrojejunostomy.
BEGER PROCEDURE
Duodenum-preserving pancreatic head resection (DPPHR), in which the
duodenum is left intact and resection of the pancreatic parenchyma is limited,
has gained traction as an alternative to PD (Figure 5 ). This procedure was
first introduced by Dr. Hans G. Beger in 1972, under the notion that
conservative resection of inflamed pancreatic tissue would provide pain
relief and spare the patient from the high morbidity associated with an
extensive PD. The contemporary modification of the Beger procedure
consists of two pancreaticojejunostomies and a jejunojejunostomy. Similar to
PD, the indications are head-predominant, small duct chronic pancreatitis.
However, the Beger procedure is not appropriate when malignancy is
suspected.
Beger Technique
A bilateral subcostal incision or midline incision is made. The lesser sac is
opened, and adhesions from the pancreas are taken down. A Kocher
maneuver is performed, and the infrapancreatic superior mesenteric vein is
identified. After ligation of the right gastroepiploic arteries, the pancreas
neck is transected. Next, sutures are placed in the pancreatic head along the
C-loop of the duodenum, to aid in both hemostasis and traction (our
preference is to place two rows of 3-0 polypropylene).
Subtotal resection of the head and neck of the pancreas above the
portomesenteric axis is performed with electrocautery or the ultrasonic
dissector. Next, a cholecystectomy is completed, if a gallbladder is present.
Using a roux limb brought through the transverse mesocolon, two
pancreaticojejunostomies (side-to-side anastomosis to the resection cavity in
the pancreatic head, and end-to-end anastomosis to the distal pancreas) are
created, followed by a downstream jejunojejunostomy to reestablish
intestinal continuity.
FREY PROCEDURE
The Beger procedure was revised in 1987 by Dr. Charles F. Frey to consist
of duodenum-sparing head resection combined with a lateral
pancreaticojejunostomy, a modification which now bears his name (Figure 6
). Similar to PD and Beger procedures, the Frey procedure is indicated for
head-predominant, small duct chronic pancreatitis. In addition, it is also well
suited for patients with the presence of “chain-of-lakes,” or multiple
strictures in the proximal pancreatic duct.
FIGURE 6: Approach is similar to Beger procedure. The
pancreatic duct is opened from the pancreatic head to 1-2 cm of the
pancreatic tail. After a limited resection of the pancreatic head, a
roux jejunal limb is used to construct a lateral
pancreaticojejunostomy to the parenchyma and exposed duct.
Given that the Frey procedure avoids dissection of the pancreas above the
portal and mesenteric vasculature, the operative time and morbidity is
significantly shorter than the Beger procedure and PD. A single institution
RCT by Bachmann et al. found the Beger and Frey procedures to be
comparable in postoperative mortality, quality of life, and pain control.
There were also no differences in survival, endocrine insufficiency, or
exocrine insufficiency at 16-year follow-up.
Frey Technique
The initial operative approach is similar to the Beger technique–enter the
lesser sac, conduct a Kocher maneuver, and ligate the right gastroepiploic
arteries. Next, the pancreatic duct is assessed by palpation or with the use of
intraoperative ultrasound and is opened using electrocautery. The duct is
opened from the pancreatic head to 1-2 cm of the pancreatic tail. Prior to
coring out the pancreatic head, we prefer to place stay sutures to gain
hemostasis and traction. After a limited resection of the pancreatic head, a
roux jejunum limb is used to construct a lateral pancreaticojejunostomy to the
parenchyma and exposed duct. Finally, intestinal continuity is re-established
with a jejunojejunostomy.
BERNE PROCEDURE
The Berne procedure is a combination of the Beger and Frey procedures–a
subtotal pancreas head resection is performed but avoids division of the
pancreas neck and drainage of the pancreatic duct via creation of a lateral
pancreaticojejunostomy (Figure 7 ). Proponents of the Berne procedure
claim that the modification reduces morbidity while adopting the highlights
of both Beger and Frey procedures. A single institution RCT from the
University of Heidelberg demonstrated superior perioperative outcomes and
equivalent long-term patient quality of life, endocrine and exocrine
pancreatic function, and need for additional interventions. Larger, multi-
institutional trials are needed to evaluate the Berne procedure with Beger
and Frey procedures.
FIGURE 7: This is a combination of the Beger and Frey
procedures-a subtotal pancreas head resection is performed by
avoids division of the pancreas neck and drainage of the
pancreatic duct via creation of a lateral pancreaticojejunostomy.
CONCLUSION
Operative management of patients with chronic pancreatitis who have failed
conservative measures is dictated by anatomical involvement of the
parenchyma and ductal system. The primary goal of intervention is durable
pain relief, followed by preservation of functional pancreatic tissue.
Surgeons dedicated to caring for this unique patient population should be
familiar with both resection and drainage procedures.
For the majority of treated patients, a tailored surgical approach is
successful with regard to pain relief and narcotic independence. Up to 25%
of patients, however, may fail to achieve symptomatic relief following index
operation. Our practice is to reevaluate these patients and tailor the
revisional operation based on the anatomical constraints. Among patients
with diffuse involvement of the (remnant) pancreas gland, we utilize
completion pancreatectomy with islet cell transplant as salvage therapy. Our
experience with revisional surgery has been effective in relieving pain, with
a quarter of these patients demonstrating narcotic independence and more
than half maintaining decreased narcotic requirements. We have previously
demonstrated total pancreatectomy with islet cell transplant (TP/ICT) as an
effective option in patients with genetically linked pancreatitis and minimal-
change chronic pancreatitis.
Future work is needed on determining preoperatively which patients are
at risk for failure after the index operation and may benefit from TP/ICT as
the initial surgical intervention. Figure 9 summarizes our surgical algorithm
for patients with head-predominant chronic pancreatitis.
FIGURE 9: Surgical Algorithm for Patients with Head-Dominant
Chronic Pancreatitis
SALIENT POINTS
Operative approach for chronic pancreatitis is dictated by anatomical
abnormalities.
Pancreaticoduodenectomy is mandated for clinical suspicion and
diagnostic workup suggestive of malignancy.
Duodenal-preserving pancreatic head resection and
pancreaticoduodenectomy are equally effective with regard to
achieving durable postoperative pain relief.
Frey procedure is warranted for small duct chronic pancreatitis with
multiple duct strictures.
The modified Puestow drainage procedure is warranted for large duct
(> 7 mm) chronic pancreatitis ± multiple duct strictures.
SELECTED REFERENCES
1. Ahmad SA, Wray C, Rilo HL, Choe KA, Gelrud A, Howington JA.
Chronic pancreatitis: recent advances and ongoing challenges. Curr
Probl Surg. 2006; 43(3): 127-238.
2. van der Gaag NA, van Gulik TM, Busch OR, Sprangers MA, Bruno MJ,
Zevenbergen C. Functional and medical outcomes after tailored surgery
for pain due to chronic pancreatitis. Ann Surg. 2012; 255(4): 763–70.
3. Bachmann K, Tomkoetter L, Erbes J, Hofmann B, Reeh M, Perez D.
Beger and Frey procedures for treatment of chronic pancreatitis:
comparison of outcomes at 16-year follow-up. J Am Coll Surg. 2014;
219(2): 208–16.
4. Klaiber U, Alldinger I, Probst P, Bruckner T, Contin P, Köninger J.
Duodenum-preserving pancreatic head resection: 10-year follow-up of
a randomized controlled trial comparing the Beger procedure with the
Berne modification. Surgery. 2016; 160(1): 127-35.
5. McClaine RJ, Lowy AM, Matthews JB, Schmulewitz N, Sussman JJ,
Ingraham AM. A comparison of pancreaticoduodenectomy and
duodenum-preserving head resection for the treatment of chronic
pancreatitis. HPB (Oxford). 2009; 11(8): 677-83.
6. Diener MK, Rahbari NN, Fischer L, Antes G, Büchler MW, Seiler CM.
Duodenum-preserving pancreatic head resection versus
pancreatoduodenectomy for surgical treatment of chronic pancreatitis: a
systematic review and meta-analysis. Ann Surg. 2008; 247(6): 950–61.
7. Nealon WH, Matin S. Analysis of surgical success in preventing
recurrent acute exacerbations in chronic pancreatitis. Ann Surg. 2001;
233(6): 793-800.
Chu Q, Vollmer C, Zibari G, Orloff S, Williams M, Gimenez M (eds).
Hepato-Pancreato-Biliary and Transplant Surgery: Practical Management of Dilemmas
CASE SCENARIO
A 45-year-old male undergoes a CT for assessment of abdominal pain after a
vehicle accident. Imaging is unremarkable for trauma-related findings, and
the patient is set for discharge. However, an incidental finding of a 2 cm
mass in the head of the pancreas is noted (Figure 1 ). The lesion has cystic
characteristics and is consistent with a branch-duct Intraductal Papillary
Mucinous Neoplasm (BD-IPMN). The patient is a smoker with no significant
past medical history. His father died of metastatic pancreatic
adenocarcinoma at the age of 61. Should I operate or follow?
FIGURE 1: Computed tomography in a 45-year-old patient with a
2 cm branch-duct IPMN in the pancreatic head (arrow).
█ BACKGROUND
The complexity of IPMN management proves to be one of the biggest puzzles
of modern surgical oncology; the million-dollar question is how to determine
who should undergo surgical resection, who needs surveillance, and who
needs no further attention. The clinical decision is difficult, since the risk of
malignancy in not operating must be balanced against perioperative
morbidity and mortality, and potential long-term complications, such as
endocrine insufficiency, in patients undergoing surgical resection.
IPMNs are mucin-producing cystic neoplasms arising within the
pancreatic ducts. These lesions are precursors to invasive pancreatic ductal
adenocarcinoma (PDAC). The progression of IPMN to invasive cancer
occurs through an adenoma-to-carcinoma sequence analogous to colorectal
polyps. Despite the precursor nature of IPMN, evidence suggests that the
overwhelming majority of these lesions will not progress to cancer. Thus, the
management of IPMN consists of thoughtful selection of patients for resection
versus observation. Patients with features associated with an increased risk
of either high-grade dysplasia (carcinoma in situ ) or invasive cancer will
benefit from resection. The evidence we use to stratify that risk is limited,
and we have little data on the long-term biological behavior of most
individuals with IPMN. Our current understanding of features associated
with high-grade dysplasia (HGD) or invasive cancer is based on
retrospective series of resected patients. These studies lack a common
denominator of all patients with IPMN lesions, since they are already
selected for surgery. Due to the lack of definitive evidence on numerous
aspects of IPMN biology, expert consensus statements have been developed.
These guidelines are based on judgment of experts in the field using extended
clinical experience and limited evidence.
The examined scenario is typical of what is frequently seen in the
Multidisciplinary Pancreatic Cyst Clinic of Johns Hopkins Hospital. Our
approach to a patient with a pancreatic cystic neoplasm is to start with the
premise that the clear majority of these lesions are benign at the time of
diagnosis and only a very small percentage will progress to malignancy.
Therefore, the decision to offer surgical resection should only be made to the
patients who have a high probability of benefitting–a small minority of the
total population. For each patient, we need to assess a variety of different
clinical, radiological, and pathological parameters, assign a risk of
malignancy, and conduct an individualized treatment plan. We generally
follow either the International Association of Pancreatology or the European
Consensus guidelines (ICG/ECG) in the management of our patients. The
general workup includes a careful assessment of age, gender, and medical
history. In addition, high-quality pancreas-protocol cross-sectional imaging
is obtained, if not already available; this will include either a CT scan or an
MRI. Based on this information alone, surgical resection can sometimes be
recommended and no further work-up is needed. If the initial analysis
identifies no indications for surgical resection (per ICG/ECG), we often
perform an endoscopic ultrasound (EUS) with fine needle aspiration (FNA)
and cyst fluid analysis for a cyst larger than 1 cm. The main indication for
this procedure is to evaluate for severe atypia and the possible presence of a
solid component. If these elements are absent, the patient can safely undergo
surveillance. With these things in mind, the following is how we would think
of the management of the individual presented in this scenario.
PRESENTATION
In a 45-year-old male with a pancreatic cyst of this appearance the most
likely diagnosis is a BD-IPMN and less likely an oligocystic serous
cystadenoma or a cystic pancreatic neuroendocrine tumor (PNET). Based on
appearance and demographics, this lesion is highly unlikely to be a solid
pseudopapillary neoplasm or a mucinous cystic neoplasm. Thus, we have
given this gentleman the presumed diagnosis of BD-IPMN. The workup will
be to confirm the diagnosis of BD-IPMN as best as possible with current
methodology. Once we make the presumptive diagnosis of BD-IPMN, we
then focus on ruling out any “high-risk stigmata” that would warrant
resection. As mentioned above, the vast majority of BD-IPMN can be
observed. In the assessment of this patient we will address the following
questions:
DOES THE PATIENT HAVE ANY “HIGH-
RISK STIGMATA ” THAT WARRANT
RESECTION OR “ WORRISOME FEATURES”
THAT WARRANT FURTHER ASSESSMENT
WITH EUS-FNA?
Imaging
The radiological assessment of BD-IPMNs is the big weapon in our arsenal
to stratify, as correctly as possible, patients who are at high-risk of invasive
cancer and may need surgery. In our practice, we have adopted a
combination of the ICG/ECG criteria to assess the imaging characteristics in
BD-IPMN patients (Table 1 ). This patient already has a high-quality CT
scan that does not show a mass lesion, ductal dilatation, or other concern for
malignancy. With this in mind, he will undergo a Magnetic Resonance
Imaging/Cholangiopancreatography (MRI-MRCP, Figure 2A ) to assess
further and more accurately the incidental BD-IPMN. Utilization of MRI-
MRCP as the imaging modality of choice increases the chances of a
definitive radiologic diagnosis and provides substantial information on a
possible communication of the BD-IPMN with the main pancreatic duct
(MPD). If MRI is not available, a multidetector computed tomography (CT)
with <2 mm sections over the area of interest is sufficient.
FIGURE 2: MRI (A) and Endoscopic Ultrasound (B) of the Same
Patient.
Recommendation
The patient in this scenario has a presumed BD-IPMN with no features
associated with increased risk of high-grade dysplasia or invasive
malignancy. He will not benefit from resection at this time and should
undergo close observation.
CONCLUSIONS
Most 2 cm BD-IPMN cases fall into the category where surgical resection is
not clearly indicated. The optimal therapeutic approach is to assess the
patient’s performance status, clinical signs, and radiology findings, and
consult with them over the advantages and disadvantages of surveillance and
surgery. An assessment of the case under a multidisciplinary setting can
provide an individual comprehensive approach. The main parameters that
dictate the management of a young patient with a 2-cm BD-IPMN are
available in Table 2 . The surgeon must keep in mind that each patient is
different, and all available clinical and imaging information are valuable in
determining the therapeutic approach. The patient must be actively involved
in the treatment plan, and if available, a multidisciplinary committee should
be consulted.
SALIENT POINTS
Utilize MRI-MRCP or a multidetector CT for accurate assessment of
the lesion.
Enhancing solid component in the cystic wall or a dilated MPD is an
indication for surgical resection, if clinically appropriate.
Perform an EUS-guided FNA if the lesion has “worrisome features” in
order to assess for the presence of severe atypia or solid component
not seen on cross-sectional imaging.
Absence of clinical symptoms is favorable towards surveillance.
Patients with jaundice, severe pain, and unintentional weight loss have
an indication for surgical resection, if clinically appropriate.
Pancreatitis, pancreatic insufficiency, and recent-onset diabetes
mellitus are risk factors of malignant progression, and resection may be
considered when accounting for other factors.
Measure blood CA19-9 and consider high values as suspicious in
lesions with “worrisome features.”
Proceed to cyst fluid analysis to evaluate atypical cells and fluid
biomarkers in a patient with difficult diagnosis.
Smoking is not considered a risk factor for BD-IPMN progression;
however, always advise cessation, since there is a proven correlation
between smoking and pancreatic cancer
A positive family history of pancreatic cancer in first-degree relatives
increases the risk for pancreatic cancer; that risk is substantially higher
with 2 first-degree relatives.
SELECTED REFERENCES
1. Tanaka M, Fernandez-Del Castillo C, Kamisawa T, et al. Revisions of
International Consensus Fukuoka Guidelines for the management of
IPMN of the Pancreas. Pancreatology. 2017 Jul 13. [Epub ahead of
print]
2. Del Chiaro M, Verbeke C, Salvia R, Kloppel G, Werner J, McKay C, et
al. European experts’ consensus statement on cystic tumours of the
pancreas. Dig Liver Dis. 2013; 45(9):703-11.
3. Gemenetzis G, Bagante F, Griffin JF, et al. Neutrophil-to-lymphocyte
Ratio is a Predictive Marker for Invasive Malignancy in Intraductal
Papillary Mucinous Neoplasms of the Pancreas. Ann Surg. 2017
Aug;266(2):339-345.
4. Rezaee N, Khalifian S, Cameron JL, Pawlik TM, Hruban RH, Fishman
EK, et al. Smoking is not associated with severe dysplasia or invasive
carcinoma in resected intraductal papillary mucinous neoplasms. J
Gastrointest Surg. 2015;19(4):656-65.
5. He J, Cameron JL, Ahuja N, Makary MA, Hirose K, Choti MA, et al. Is
it necessary to follow patients after resection of a benign pancreatic
intraductal papillary mucinous neoplasm? J Am Coll Surg. 2013;
216(4):657-65; discussion 665-7.
█ BACKGROUND
Groove pancreatitis is an uncommon presentation of chronic pancreatitis and
is frequently unrecognized by radiologists who are unpracticed with cross-
body imaging in the diagnosis of chronic pancreatitis. Groove pancreatitis is
a form of focal chronic pancreatitis within the pancreas tissue that lies
between the duodenal wall and the intrapancreatic portion of the common
bile duct. It has both a cystic and solid presentation and may be described as
cystic dystrophy of the duodenal wall, duodenal dystrophy, and paraduodenal
pancreatitis. The variations of nomenclature relate to the pathogenesis of this
disorder. An embryologic variation in the differentiation and fusion of the
pancreas and duodenum leads to ectopic or heterotopic location of
pancreatic tissue. This variation in pancreatic anatomy does not usually
present until the adult period. It is likely that the development of chronic
pancreatitis in ectopic pancreas tissue requires a genetic predisposition as
well as environmental factors such as tobacco and alcohol abuse as seen in
the index case.
Groove pancreatitis involves progressive cystic degeneration of ectopic
pancreas rests that lie within the duodenal wall. Pancreatic rests can occur in
other locations, but when located within the pancreaticoduodenal groove, the
development of fibrotic and inflammatory tissue leads to a typical radiologic
and clinical presentation. Groove pancreatitis may not involve the rest of the
pancreas with the inflammatory process, though the pancreatic duct and
biliary system are frequently secondarily involved in the process, as was the
case with this patient. Identification of this entity and its varied appearances
as a distinct pathology is important in order to direct the patient to a surgeon
who is willing and able to perform pancreaticododenectomy for chronic
pancreatitis. It is not uncommon for the diagnosis to be overlooked on initial
cross-sectional imaging, though the experienced radiologist readily identifies
the characteristic cystic and inflammatory periduodenal changes.
The imaging appearances and clinical features of groove pancreatitis may
overlap with pancreatic adenocarcinoma and EUS may have a role in
diagnosis when the question of a pancreatic mass is present. CA 19-9 may be
helpful in ruling out a cancer diagnosis. The typical patient is similar to the
patient in this case, a 50-year-old male with a history of chronic relapsing
abdominal pain and weight loss associated with alcohol and tobacco abuse.
Serum laboratory values usually show only mildly elevated amylase and
lipase that help to differentiate from typical acute interstitial or acute
necrotic pancreatitis. Serum liver transaminases may be mildly elevated.
Jaundice and hyperbilirubinemia are less common but can occur mimicking
ductal carcinoma. Patients may also present with symptoms of gastric outlet
obstruction when the inflammatory process leads to duodenal stenosis in the
descending duodenum.
Groove pancreatitis is evident histologically as chronic inflammation and
scarring in the anatomical space between the medial wall of the duodenum
and the head of the pancreas, which is termed the “pancreaticoduodenal
groove.” Initial presentations can be subtle both clinically and on imaging.
The earliest imaging finding is focal inflammation and fibrosis in the
pancreaticoduodenal groove without involvement of the pancreatic or biliary
ductal systems. Ductal dilatation may present at a later stage of the disease
process and can be related to increased fibrosis. Ductal obstruction may be
precipitated by the development of the cystic changes that exert a mass effect
on both the pancreatic and bile ducts near the ampullary region. The
accessory duct of Santorini at the minor papilla is the duct most frequently
involved in the obstructive process. When an associated obstructive
pancreatopathy develops due to focal inflammation and fibrosis, ductal
dilation, proteinaceous plugs, and intraductal calcific stones may develop. At
this stage the appearance of groove pancreatitis may overlap with typical
appearances of advanced chronic calcific pancreatitis. The features of
chronic calcific pancreatitis correspond to the severity of ductal
abnormalities, with chronic stenosis leading to upstream stasis and eventual
intraductal stone formation as demonstrated in this patient.
There are two established typical gross morphologic forms of this
disease: the pure form and the segmental form. In the “pure” form, the
abnormalities are confined to the pancreaticoduodenal groove, and the head
of the pancreas is spared. In the “segmental” form, the inflammation and
scarring extend to directly involve the head of the pancreas, which frequently
leads to intrapancreatic abnormalities including cysts and fibrotic masses,
which was the case with this patient.
Prior to undertaking pancreatic resection, a trial of smoking and alcohol
cessation is warranted. Endoscopic management of terminal biliary stenosis
and pseudocyst drainage may have a role while the benefits of smoking and
alcohol cessation are given some time to work. When chronic inflammatory
changes of an obstructive pancreatopathy develop in the body and tail of the
pancreas, resection of the head of the pancreas has a lower complication rate
because the anastomosis is constructed to a fibrotic pancreas with a dilated
duct. When that is the case, the postoperative course is similar to that of this
patient in which no pancreatic fistula developed, and the patient had an
expeditious discharge to home. Although operative complications related to
anastomotic leak and fistula formation are higher in the patient without
secondary inflammatory changes in the remnant pancreas, long-term patient
outcomes are better because of the absence of chronic inflammatory changes
in the remnant pancreas.
CASE SCENARIO
A 33-year-old woman was referred for surgical management of recurrent
acute pancreatitis associated with pancreas divisum. She had a 5-year history
of recurrent abdominal pain and nausea associated with multiple emergency
room visits and several hospitalizations with associated serum lipase
elevation and mild changes of acute pancreatitis on CT scan. Past surgical
history was significant for morbid obesity managed with a laparoscopic
gastric bypass. One year previously the patient had undergone on open minor
duct sphincteroplasty that decreased emergency department visits and
hospitalizations but failed after 6 months and led to a subsequent
pancreaticoduodenctomy, which also failed after 12 months. A completion
pancreatectomy with islet autotransplantation was performed. The resected
pancreas was atrophic and fibrotic with consequent poor islet cell yield.
█ BACKGROUND
The strategies utilized in the surgical management pancreas divisum are those
used in the surgical management of chronic pancreatitis: improve drainage,
resect damaged tissue, or combine the first two with a simultaneous resection
and drainage procedure. This patient demonstrates how all three can be
utilized in an ineffective way that delays optimal therapy. The prudent
surgeon always remembers that pancreas divisum is as common as left-
handedness: ten percent of the population has pancreas divisum. The concept
was promulgated in the 1970s that in pancreas divisum there is an anatomic
impediment to the normal drainage of pancreatic exocrine secretions that
results in an obstructive pancreatopathy. When divisum anatomy is present,
the dorsal duct, formerly known as the duct of Santorini, is called the
dominant duct. Variations of the dominant dorsal duct are many and
infrequent enough so that radiologic and endoscopic delineation may be
confusing. The classic radiographic image has the dominant pancreatic duct
crossing the terminal bile duct and entering the descending duodenum
(Figure 2 ). Usually a small ductal system drains the ventral head of
pancreas and enters the duodenum through the major papilla with the terminal
bile duct (Figure 3 ). Beware of the entity called acquired pancreas divisum
that is associated with chronic pancreatitis and malignancy. Either process
may result in total occlusion of the ventral duct, causing the duct of Santorini
to assume responsibility for pancreatic exocrine outflow via the minor
papilla.
FIGURE 2: MRCP demonstrating dorsal duct dominant pancreas
divisum with duct of Santorini coursing horizontally across distal
common bile duct to its downstream union with the duodenum
proximal to the ampulla of Vater.
FIGURE 3: The normal adult pancreas is formed from the fusion
of the dorsal duct and ventral duct. The ventral duct drains 70% of
the pancreas via the major papilla. Failure of the ventral duct to
fuse with the dorsal duct results in pancreas divisum, a name given
because the pancreas is now drained by two ducts. As a result 70%
of the pancreatic drainage is via the minor papilla. Although not
entirely cleared, it is thought that pancreatitis stems from the the
minor papilla being too narrow to adequately drain the pancreas.
SALIENT POINTS
Groove pancreatitis is a relatively rare disease, which is best
described as chronic inflammation of ectopic pancreatic tissue in the
duodenal wall.
Groove pancreatitis is also known as paraduodenal pancreatitis and
duodenal cystic dystrophy.
Groove pancreatitis is associated with alcohol and tobacco abuse.
Groove pancreatitis may be complicated by duodenal stenosis and
biliary and pancreatic ductal obstruction.
Groove pancreatitis may mimic pancreatic cancer and vice versa.
Pancreas divisum occurs in about ten percent of the population.
In combination with environmental and genetic factors, patients with
pancreas divisum may develop recurrent acute pancreatitis and
subsequent chronic pancreatitis.
Endoscopic minor duct sphincterotomy has a role in the management of
pancreas divisum.
Operative sphincteroplasty may also have a role in patient
management, as may other resection and drainage procedures.
The exact timing and selection of resection and drainage procedures
including total pancreatectomy with islet autotransplantation has not
been determined.
SELECTED REFERENCES
1. Hungerford JP, Neill Magarik MA, Hardie AD. The breadth of imaging
findings of groove pancreatitis. Clin Imaging. 201;39(3):363-6.
2. Egorov VI, Vankovich AN, Petrov RV, Starostina NS, Butkevich AT,
Sazhin AV. Pancreas-preserving approach to “paraduodenal
pancreatitis” treatment: why, when, and how? Experience of treatment
of 62 patients with duodenal dystrophy. Biomed Res Int. 2014; Epub
2014 Jun 5.
3. Morgan KA, Romagnuolo J, Adams DB. Transduodenal
sphincteroplasty in the management of sphincter of Oddi dysfunction and
pancreas divisum in the modern era. J Am Coll Surg. 2008; 206(5):
908-14.
4. Schnelldorfer T, Adams DB. Outcome after lateral
pancreaticojejunostomy in patients with chronic pancreatitis associated
with pancreas divisum. Am Surg. 2003; 69 (12):1041-4.
5. Adams DB, Cote G. Pancreas Divisum and Other Variants of Dominant
Dorsal Duct Anatomy. Current Surgical Therapy. 12th Ed. Elsevier,
New York, 2017, Chapter 95.
█ BACKGROUND
Pancreatic neuroendocrine tumors (PNETs) represent a minority, less than
4%, of pancreatic neoplasms. Compared to other gastrointestinal endocrine
tumors, PNETs are of greater concern because they have significantly poorer
survival. These neoplasms are not homogeneous. Rather, different PNET
diagnoses portend variable long-term outcomes dependent upon presentation
and histology. Management strategies are not uniform, can often be
controversial, and are typically individualized to the patient. Given the
increasing frequency with which high-resolution, cross-sectional imaging is
utilized, there is an increasing abundance of incidental, small, asymptomatic
PNETs. Management of such tumors with surgery versus observation is
controversial. However, a surgical approach is favored in the context of
several clinical scenarios, which are described by this chapter in detail.
TUMOR CHARACTERISTICS
There are many tumor-dependent characteristics of a PNET which favor
surgical resection. These characteristics include functionality, classification,
anatomy, and diagnostic ambiguity.
FUNCTIONAL TUMORS
Sporadic hormone-producing PNETs warrant resection in the absence of
metastatic disease. Resection offers patients symptomatic relief from
hormone overproduction and may improve overall and disease-free survival.
Functional PNETs associated with genetic syndromes, such as Multiple
Endocrine Neoplasia Type I (MEN-1), are discussed later in the chapter.
Hormone-associated tumor syndromes include gastrinomas in Zollinger-
Ellison syndrome, insulinomas leading to the Whipple triad, VIPoma in
Vemer-Morrison Syndrome, carcinoid syndrome, adrenocorticotropic
hormone-omas producing Cushing’s syndrome, glucagonomas,
somatostatinomas, growth hormone releasing factor tumors, parathyroid
hormone-related peptide-omas, and cholecystokininoma. Given the wide
variety of potential functional syndromes, a thorough history and physical, in
addition to metabolic workup is essential after PNET diagnosis. Failure to
identify a functional lesion coupled with the decision for observation over
surgical resection could lead to significant adverse health effects.
NONFUNCTIONAL TUMORS
Multiple classification systems exist for PNETs, and consideration of more
than one system is advisable when embarking upon a surgical pathway for
nonfunctional sporadic tumors. The American Joint Committee on Cancer
(AJCC) staging system utilizes the features of the primary tumor (T) and the
presence of regional lymph node involvement (N) or distant metastases (M)
to inform management and prognosis. T-status is substratified based on size
greater than 2 cm (T2) and whether the tumor extends beyond the pancreas
(T3) and involves the celiac axis or superior mesenteric artery (T4).
It is accurate, albeit controversial, to state that surgical resection is a
viable option for all T1-T3 nonmetastatic, non functioning, nonsyndromic
PNETs. Some experts, however, would dismiss this strategy as an
oversimplification and argue that observation is preferable to surgery for
select T1 lesions. This selection process, unfortunately, is not well defined
and is often surgeon dependent. Importantly, mere classification as an
asymptomatic T1 lesion is not, on its own, sufficient enough information to
safely defer surgical intervention.
AJCC staging characteristics are too limited to dictate management
decisions alone. For example, they do not account for tumor histology, and
they are not specific to PNETs; they also apply to pancreatic
adenocarcinomas, which herald an overall worse prognosis. Additionally,
they fail to consider tumor growth rate. For example, a 0.75 cm tumor that
doubles in size over a 3- to 6-month interval is concerning and warrants
resection. Furthermore, there is a well defined risk of lymph node metastasis
for tumors less than 2 cm (T1). Failure to identify regional lymph node status
withholds valuable prognostic information from both the patient and the
treating physician. Therefore, several reasons to resect a T1 lesion exist and
are further described.
Intermediate and high-grade T1 PNETs, based on the World Health
Organization (WHO) classification of neuroendocrine tumors, should be
strongly considered for resection. The WHO determines histologic grade by
the number of mitoses per high-powered field (HPF), and Ki-67 proliferation
index, which is a marker of tumor cell multiplication and is expressed during
chromosomal replication. Well-differentiated/low-grade (G1) tumors have
less than 2 mitoses/10 HPF, and a low (less than 3%) Ki-67 index. Grade 3,
or poorly differentiated tumors, should be resected irrespective of size
criteria, although there is some debate regarding the exact Ki-67 cutoff, in
favor of a higher threshold. Nevertheless, the literature has clearly shown
that tumors with Ki-67 less than 2% have a distinctly less aggressive initial
disease course when compared to lesions with Ki-67 between 2 and 20%,
and greater than 20%.3 Intermediate grade tumors may also have strong
consideration in favor of surgery, with greater emphasis given to Ki-67
proliferation index than to number of mitoses.
Grading may be difficult to assess in the preoperative setting. Ki-67
expression on FNA is frequently concordant with postoperative histologic
analysis, but agreement is less than 90%. Tumor heterogeneity is also a well-
recognized phenomenon that may lead to misclassification and
undertreatment. Other tumor characteristics, such as lack of somatostatin
receptor positivity in nonfunctioning tumors, may correlate with more
aggressive tumors and warrant consideration as well. When there is
diagnostic ambiguity about the lesion in question, surgical resection is often
the preferred alternative.
Perhaps the most convincing data favoring surgical resection for T1
lesions is that nodal metastases occur in PNETs despite their small size, and
nodal status directly correlates with patient outcome. Failure to address
nodal metastases contradicts the primary tenant of PNET management, which
is to resect all disease, thereby preventing distant metastasis and extending
patient longevity. When Hashim et al. applied an aggressive surgical
strategy–resection and lymphadenectomy were uniformly applied to 98% of
studied patients irrespective of tumor size– it was demonstrated that 12% of
patients with subcentimeter lesions and 13% of patients with tumors less than
1.5 cm had N1 disease. Patients with tumor diameter greater than 1.5 cm
were 4.7 times more likely to have regional lymph node metastases than
smaller tumors (95% CI, 1.81 to 12.35, P = 0.002). When one considers this
statistic, it is clear that size criteria cannot be considered alone in deciding
to resect or observe a lesion. A T1 lesion with imaging suggestive of
positive lymph node status is not appropriate for observation.
Numerous published reports have demonstrated that nodal metastases,
disease recurrence, and death from PNET disease, occur infrequently albeit
undeniably in T1 lesions, even those less than 1 cm. Although PNETs tend to
be indolent tumors, many experts agree that the data is not good enough to
reliably distinguish those small, aggressive tumors in order to support more
conservative stances that favor observation over surgery. This is particularly
true of well-conditioned patients with a reasonable life expectancy and the
absence of clear medical contraindications to surgery.
Particular attention should be given to tumor location when considering
surgery for a PNET. Pancreatic head lesions have been shown to be 2.8 times
more likely to have regional lymph node metastases (95% CI, 1.3 to 5.8, P =
0.007).3 In extrapolated data, the probability of N1 disease for tumors of the
pancreatic head was at least 1 in 5 patients, even with small tumors. Tumor
location may also impact the likelihood of duct obstruction, jaundice,
pancreatic insufficiency, or pain, all of which will influence the decision to
resect the lesion.
When considering a strategy of observation one must also critically
appraise published data on nonoperative management of small tumors. These
studies are not randomized and therefore have built in biases, particularly
selection bias. There may be a significant conversion rate to surgery after
imaging surveillance. For example, Sadot et al. converted 26 of 104 patients
selected for observation to the surgical arm after a median of 30 months.
Cited reasons for these conversions appear arbitrary rather than being based
on objective data–38% patient and 27% physician preference, making it
difficult to create a cohesive guideline that challenges surgery as the standard
of care. Most importantly, advocates of observation report on patients with
N0 disease, which can be presumed but not guaranteed prior to surgery.
Also, long-term follow-up on patients undergoing observation instead of
surgery is limited, especially in the context of the slow-growing but not
always benign evolution that typifies a PNET.
PATIENT CHARACTERISTICS
The majority of patients with sporadic functioning and nonfunctioning PNETs
should undergo surgical resection, unless patient characteristics result in
operative contraindications. Examples of contraindications include patients
with a high-risk profile, or patients for whom resection would unreasonably
interfere with their quality of life. Lifestyle factors include pre-diabetics or
patients with good oral hyperglycemic control, for whom blood glucose
control could become challenging. Advanced age has been shown to be an
independent factor associated with more aggressive disease course, thus it
should not be used in isolation to select out patients otherwise well-suited to
surgery.3 Patients who are unreliable and unlikely or unable to present for
sequential imaging may be better served with initial surgery. Some patients,
like the woman in the clinical scenario above, experience significant anxiety
after being given their diagnosis and prefer surgery over the fear of the
unknown.
SURGICAL MANAGEMENT
Once the choice in favor of surgery is established, one must decide upon the
extent of surgery. Options to address pancreatic parenchyma include a
limited resection (enucleation or parenchymal-sparing procedures, e.g.
central or middle pancreatectomy) versus formal resection
(pancreaticoduodenectomy or distal pancreatectomy). Lymph node resection
may be included, and if it is performed, it may be completed in a limited or
extended fashion. Insulinomas and gastrinomas may be enucleated if their
anatomy is favorable. Enucleation instead of formal resection for other
functioning PNETs is rarely recommended. This is, in part, due to the fact
that these other functional subtypes tend to present with more advanced
disease and behave in a more aggressive fashion, warranting regional
lymphadenectomy. In patients with functional tumors who are anticipated to
require long-term octreotide, cholecystectomy should be performed at time of
initial surgery.
Nonfunctional resectable lesions greater than 2 cm in size (T2 or T3)
typically warrant formal surgical resection. For lesions less than 2 cm and
limited to the pancreas (T1), limited resection can be an appropriate, less
invasive, alternative to traditional resection techniques. Tumors considered
appropriate for enucleation are typically less than 2 cm, superficial, away
from the pancreatic duct, and encapsulated. Review of the Surveillance
Epidemiology and End Results (SEER) database from 1998-2012 highlights
that, while surgeons are significantly more likely to consider enucleation for
smaller tumors, a minority of patients (57 patients in 981) undergo
enucleation.
Lymph node sampling with frozen section and possible node dissection may
be performed at the time of limited resection, but the utility of this is widely
debated. NCCN guidelines favor lymphadenectomy based on staging
criteria–lymphadenectomy is recommended for T2 lesions and suggested for
T1 lesions greater than or equal to 1 cm. However, Kaplan-Meier survival
statistics using SEER data show no difference in overall or disease-free
survival for T1 or T2 N0 versus NX patients, and extended
lymphadenectomy (greater than 10 lymph nodes) for all T groups and any N
status has not been associated with improved survival. Therefore, Conrad et
al. recommend that lymphadenectomy is best utilized to confirm
prognostication in patients with any size tumor and radiographic suspicion of
nodal involvement and argue that the risks associated with routine
aggressive, extended lymphadenectomy may not be justified.
INHERITED SYNDROMES
The most essential patient factor impacting surgical management is the
presence of an inherited syndrome. Genetic predisposition to PNETs occurs
in MEN-1, von Hippel-Lindau disease (VHL), von Recklinghausen
disease/neurofibromatosis type 1 (NF-1), and tuberous sclerosis (TS).
PNETs are relatively uncommon manifestations of NF-1 and tuberous
sclerosis, and resection of these lesions is recommended.
MEN-1 is characterized by hormone production, larger and multifocal
lesions, and metastatic disease. Malignant islet tumors are the major cause of
premature death in these patients. Surgery is rarely curative. Therefore, a
more conservative approach to surgery is preferred. Enucleation is often
reasonable when feasible because this strategy favors a curative outcome and
spares the majority of pancreas parenchyma. Formal pancreatic resection for
patients with MEN-1 is intended to minimize metastatic risk and maximize
survival. A Whipple may be indicated in a young patient with a large mass
(greater than 1 to 2 cm), a functional tumor refractory to medical
management, or rapid tumor growth. However, given the genetic basis for
tumor development in these patients, additional lesions are more likely to
develop in the unresected pancreatic tissue than in patients with sporadic
disease, and completion or total pancreatectomy is to be avoided in
syndromic patients.
VHL PNETs, as in MEN-1, are typically multiple. However, unlike in MEN-
1, they are usually asymptomatic and nonfunctioning. Metastatic potential is
unclear due to the limited number of patients with this disease, and primary
causes of morbidity tend to be renal or cerebral tumors, rather than those
arising in the pancreas. Surgical management for PNETs in VHL is heavily
dependent on size criteria. Tumors less than 1 cm are typically monitored
with serial imaging, tumors greater than 3 cm are resected, and intermediate
tumors are managed individually. Enucleation is generally favored, if
feasible.
CONCLUSION
In summary, most patients with resectable PNETs are best suited to surgical
resection, with consideration given to limited resection based on anatomic
and size criteria. Observation could be considered for nonfamilial T1 lesions
with grade 1 his tology, no evidence of lymph node involvement, no evidence
of hormone production, without interval growth, if surgery is of prohibitive
risk, or the patient prefers to forgo resection. However, data supporting
observation is limited. For the majority of patients with this diagnosis,
surgical management remains the standard of care.
The patient presented in this scenario will almost certainly undergo
resection. The notion of observing a solid pancreatic neoplasm is an
anathema to most patients even with detailed counseling, especially when
they have first-hand experience watching someone close to them die of
pancreatic cancer. Furthermore, at the age of 56, this patient will need
surveillance at least annually for the next 40 years. Given the nature of cross-
sectional imaging, there will undoubtedly be some change on imaging within
this time period (either real or created by technique and interpretation) that
elevates concern enough to proceed with resection.
From our perspective, the question is not whether this lesion should be
resected, but rather, when it should be resected and how extensive the
resection should be. Understanding the limitations of the available data, in an
otherwise healthy, nonsyndromic patient with a 1.5 cm PNET, our practice is
to offer a resection that achieves a microscopically negative margin and
provides a sampling of the regional lymph nodes, while preserving as much
pancreatic parenchyma as possible, using the most minimally invasive
technique feasible based on the tumor location and its proximity to the main
pancreatic ductal system.
Finally, we would like to reinforce that the NCCN guidelines recommend
resection for this patient. Supporting this recommendation are decades of
experience by multiple experts in the field of PNET management at many of
the leading institutions across the country. Deviation from these guidelines
should require compelling evidence, which in our estimation does not exist.
SALIENT POINTS
Functioning and/or symptomatic PNETs should be resected and
consideration should be given to whether cholecystectomy is also
appropriate.
Formal surgical resection is appropriate for T2 and T3 PNETs.
Sporadic T1 lesions should be strongly considered for surgical
resection, especially for young patients without contraindications and
poorly compliant patients. Greater weight should be given to resect
lesions in the pancreatic head or adjacent to ductal structures, those
with grade 2 or 3 histology, size greater than 1 cm, or exhibiting
diagnostic ambiguity or interval growth.
Regional nodal status should not be presumed to be negative based on
small tumor size.
PNETs associated with inherited genetic syndromes should be
considered separately from sporadic lesions, with management based
on existing guidelines.
SELECTED REFERENCES
1. Chiruvella A, Kooby DA. Surgical Management of Pancreatic
Neuroendocrine Tumors. Surg Oncol Clin N Am. 2016;25(2):401-21.
2. NCCN Clinical Practice Guidelines in Oncology (NCCN Guidelines®).
Neuroendocrine Tumors. (Version 2.2016).
https://www.nccn.org/professionals/physician_gls/pdf/neuroendocrine.
pdf. Accessed February 14, 2017.
3. Hashim YM, Trinkaus KM, Linehan DC, Strasberg SS, Fields RC, Cao
D, et al. Regional lymphadenectomy is indicated in the surgical
treatment of pancreatic neuroendocrine tumors (PNETs). Ann Surg.
2014; 259(2):197-203.
4. Sadot E, Reidy-Lagunes DL, Tang LH, Do RK, Gonen M, D’Angelica
MI, et al. Observation versus Resection for Small Asymptomatic
Pancreatic Neuroendocrine Tumors: A Matched Case-Control Study.
Ann Surg Oncol. 2016; 23(4):1361-70.
5. Conrad C, Kutlu OC, Dasari A, Chan JA, Vauthey JN, Adams DB, et al.
Prognostic Value of Lymph Node Status and Extent of Lymphadenectomy
in Pancreatic Neuroendocrine Tumors Confined To and Extending
Beyond the Pancreas. J Gastrointest Surg. 2016;20(12):1966-1974.
6. Alexakis N, Connor S, Ghaneh P, Lombard M, Smart HL, Evans J, et al.
Hereditary pancreatic endocrine tumours. Pancreatology. 2004;
4(5):417-33; discussion 434-5.
█ BACKGROUND
Surgical Management of Small Pancreatic
Neuroendocrine Tumors
The finding of a solid, well-circumscribed hypervascular lesion in the
pancreas on CT imaging is strongly suggestive of a pancreatic
neuroendocrine tumor (PanNet). PanNets are rare neoplasms that originate
from the endocrine cells of the pancreas and exhibit a wide spectrum of
behavior ranging from an indolent low-grade neoplasm with slow growth, to
highly aggressive neoplasms with a propensity for early metastases and poor
outcomes.
Over the last decade very small (≤2 cm) PanNets have been increasingly
diagnosed in asymptomatic patients due to the widespread use of high-quality
imaging techniques. Resection is the only curative treatment for localized
PanNets, and represents the traditional first-line therapy. However, the
increasing identification of small nonfunctional resectable neoplasms, often
in elderly patients, with an uncertain malignant potential has questioned the
role of resection, as many patients will not experience progression for many
years. In order to avoid overtreatment, operative recommendations must take
into account, and balance, the risk of malignant progression and the risks
associated with resection--including the development of perioperative
complications and the development of pancreatic exocrine and endocrine
insufficiency.
Several surgical series have demonstrated that incidentally discovered
nonfunctional-PanNets (NF-PanNets) that are <3cm in diameter are in the
majority represented by a low-grade pathology (84-95%) and no nodal
involvement (86-99%). A recent study by our group investigated the natural
history of small NF-PanNet and reported a 99% 5-year overall survival, and
none of the >100 patients in the nonoperative group experience metastatic
progression. In this case-matched evaluation, there was no difference in
survival between patients who were selected for radiographic surveillance
and size-matched controls who had undergone resection. The median follow-
up was almost four years and suggests that selected patients may be safely
monitored.
The patient presented above presents with a small lesion in the head of
the pancreas that is amenable to pancreaticoduodenectomy (PD) or possibly
enucleation (EN). Both of these procedures have a high rate of postoperative
morbidity (46-50%) and a measurable mortality rate (≤2%), even in
specialized centers. Based on the data presented above, treatment
recommendations must consider the competing risks of surveillance and
resection. If pancreaticoduodenectomy is recommended for 100 patients with
a similar presentation as above, 2 will die from complications of the
operation. If radiographic surveillance is recommended for 100 similar
patients, then we would expect all to be alive in four years. In addition, the
related long-term complications from pancreatic resection should be
considered in patients treated for indolent diseases. The rate of exocrine and
endocrine insufficiency for patients undergoing PD has been reported to be
approximately 55% and 5%, respectively.
Although some would argue that PD is not necessary, it should be
emphasized that parenchyma- sparing resections are not always feasible.
Enucleation is reserved for small neoplasms located superficially that are not
immediately adjacent to the main pancreatic duct. Enucleation has also been
reported to have significant rates of morbidity. In one study by Jilesen et al.,
the leak rate was 40%, and the rate of exocrine and endocrine insufficiency
for patients undergoing EN has been reported to be approximately 19% and
7%, respectively.
Before a surveillance strategy should be implemented, an accurate
diagnosis and thorough staging evaluation should be obtained. A proper
clinical evaluation should investigate for the presence of symptoms referable
to a hormone-related syndrome such as diarrhea, flushing, hypo- or
hyperglycemic events, and rarely a rash. When these symptoms exist, the
specific syndrome should be supported by an increase in the appropriate
serum hormonal marker and, when present, represents an indication for
resection. Staging evaluation should include anatomic (CT/MRI) assessments
and, in selected patients, functional techniques (111 In-pentreotide
scintigraphy/68 Gallium DATATOC-PET scan). Imaging and functional
studies may be complementary and should be used to determine the location
and size of the primary tumor and to assess for the presence of regional and
distant metastases. Usually, small NF-PanNets are incidentally discovered
when abdominal imaging is performed for unrelated indications. PanNets
appear typically on CT and MRI imaging as well-circumscribed
hypervascular lesions. Rarely, PanNets may present as a cystic lesion, have
heterogeneous enhancement, and very rarely may present as a hypovascular
mass with pancreatic duct dilation. These latter findings may resemble other
pancreatic cystic lesions, or pancreatic adenocarcinoma, and can make the
differential diagnosis challenging.
Functional studies exploit the expression of somatostatin receptors
(SSTR) on PanNets and include 111 In-pentreotide scintigraphy (111 In-
SPECT) and 68 Gallium DATATOC-PET scan (68 Ga-PET). These tests are
able to diagnose well-differentiated tumors that highly express SSTR and
may be more sensitive in the identification of metastatic disease. In
particular, 68 Ga-PET has been found to detect distant metastases with higher
sensitivity and specificity than conventional techniques, and when functional
studies are combined with anatomical studies, a change in management has
been reported in 33% of the cases. Given its high specificity, 68 Ga-PET
may be used in selected patients to confirm the diagnosis of well-
differentiated NF-PanNet. For both functional techniques, false positives are
rare and include intra-pancreatic spleen and pancreatic renal clear cell
metastases.
When the anatomic and functional imaging evaluation suggest a stage I
(American Joint Committee Cancer 7th ed.) NF-PanNet ≤2 cm with a typical
radiographic pattern (well-circumscribed hypervascular lesion), no further
evaluations are necessary and patients may be considered candidates for
radiographic surveillance. In the setting of atypical radiographic findings, or
when there is a lack of radiotracer uptake, an endoscopic ultrasound with
fine needle aspiration biopsy (EUS/FNAB) should be performed to confirm
the diagnosis of PanNet and to evaluate for an intermediate or higher grade
lesion. The concordance between the cytological and histological Ki67 range
between 83-89%, and it is almost completely consistent for tumors smaller
than 18 mm.
Small PanNet with a high-grade component (Ki-67>20%) are rare, and
surgical resection is recommended regardless of the tumor size due to the
more aggressive tumor behavior. PanNet with intermediate grade component
(Ki-67 2-20%) are characterized by uncertain behavior and have a higher
chance to develop distant metastases. These patients should generally
undergo surgical resection or, in selected cases (patient not fit for surgery),
for close radiographic surveillance (3-6 months).
SALIENT POINTS
Small NF-PanNets are neoplasms characterized by low-grade
pathology with an indolent behavior.
Pancreatic surgery is curative but has a high morbidity rate, a
measurable mortality rate, and long-term sequela such as exocrine and
endocrine pancreatic insufficiency.
An initial observation approach is reasonable and safe in patients with
presumed small (≤2cm), low-grade, incidentally discovered PanNet.
A thorough diagnostic and staging workup must be performed to
confirm diagnosis before a surveillance strategy can be recommended.
SELECTED REFERENCES
1. Sadot E, Reidy-Lagunes DL, Tang LH, Do RK, Gonen M, D’Angelica
MI, et al. Observation versus Resection for Small Asymptomatic
Pancreatic Neuroendocrine Tumors: A Matched Case-Control Study.
Ann Surg Oncol. 2016;23(4):1361–70.
2. Bettini R, Partelli S, Boninsegna L, Capelli P, Crippa S, Pederzoli P, et
al. Tumor size correlates with malignancy in nonfunctioning pancreatic
endocrine tumor. Surgery. 2011;150(1):75–82.
3. Jilesen APJ, Van Eijck CH, Busch OR, Van Gulik TM, Gouma DJ, Van
Dijkum EJ. Postoperative Outcomes of Enucleation and Standard
Resections in Patients with a Pancreatic Neuroendocrine Tumor. World
J Surg. 2016;40(3):715–28.
4. Sadowski SM, Neychev V, Millo C, Shih J, Nilubol N, Herscovitch P, et
al. Prospective study of 68Ga-DOTATATE Positron Emission
Tomography/Computed Tomography for Detecting Gastro-Entero-
Pancreatic Neuroendocrine Tumors and Unknown Primary Sites. J Clin
Oncol. 2016;34(6):588–96.
5. Díaz del Arco C, Díaz Pérez JÁ, Ortega Medina L, Sastre Valera J,
Fernández Aceñero MJ. Reliability of Ki-67 Determination in FNA
Samples for Grading Pancreatic Neuroendocrine Tumors. Endocr
Pathol. 2016;27(4):276–83.
█ BACKGROUND
Pancreatic ductal adenocarcinoma (PDA) is notorious for its challenging
biology and potentially perilous location among the vascular structures of the
retroperitoneum. Despite substantial ongoing research, only modest gains
have been made in the treatment and life expectancy following the diagnosis
of this disease. According to National Cancer Institute SEER database
cancer statistics from the year 2012, the 5 year survival after diagnosis of
pancreas cancer was 7.7%. New data from the Pancreatic Cancer Action
Network suggests the 5 year survival may now be up to 9%. Although PDA
is twelfth on the list of new cancer diagnoses annually (53,070 new cases in
2016), it is projected to become the second leading cause of cancer-related
deaths in the United States (41,780 deaths due to pancreas cancer in 2016).
Sadly, only one-half of patients with Stage I disease come to surgery in
academic centers, and fewer in community hospital settings (typically 30%).
Thus, it is clear that the identification, rapid referral, and employment of
aggressive treatment strategies need improvement. With the goal of
increasing the number of long term survivors following diagnosis with PDA,
we in the pancreas cancer treatment community have yet to universally agree
upon the best treatment algorithm for patients such as the one in the above
clinical scenario.
HIGH-VOLUME CENTERS
It has been incontrovertibly shown that pancreas cancer is best treated at
high-volume centers. We strongly support this and receive many referrals,
such as the one described in the above clinical scenario, to our institution.
Regardless of tumor extent or clinical stage, the challenges of treating PDA
patients are many, and all cases should be reviewed by expert,
multidisciplinary teams.
OPTIMAL IMAGING
Our approach to these patients is expeditious and begins (if the patient has
had a less than optimal CT or MR scan) with a requisite “pancreas protocol”
CT scan, which is specifically designed to visualize pancreas tumors, liver
parenchyma and relevant visceral vasculature. Using a multidetector CT
scanner, the optimal protocol obtains cross-sectional images at four different
time points which are then reconstructed in 3-dimensions as 3 mm slices.
First, the patient will orally take in 500 mL of water to distend the
duodenum. Oral contrast is not used. The initial images obtained are without
intravenous contrast. Next, images are captured at three different phases
following administration of a 100 mL intravenous iodinated contrast bolus.
When a bolus-tracker device is placed over the liver and contrast is
detected, the early arterial images are obtained. Approximately ten seconds
elapse and the late arterial images are obtained. Finally, approximately 35
seconds later, the portal-venous images are then obtained. As with many
images performed outside of high-volume centers, the timing of contrast in
the portal venous phase was sub-optimal in the case scenario above. We
proceeded to acquire CT images at our own facility following these
standard, recommended guidelines.
In this case scenario, we reviewed the patient’s new multidetector CT
images carefully in all planes and, in fact, found that the superior mesenteric
vein (SMV) and the portal vein (PV) appeared to be free and clear of tumor.
We routinely evaluate the venous vasculature in the axial, coronal, and
sagittal views, as the composite information can be quite reassuring
regarding the degree of difficulty in the dissection. Additionally, upon review
of our CT scan, we noted a previously unmentioned, replaced right hepatic
artery arising off of the SMA (Figure 2 ), which requires attention at the time
of surgery. This among other nuances, such as hepatic parenchyma
enhancement, are key features to treating pancreas cancer patients
safely,confirming our desire to see these patients in referral.
FIGURE 2: CT scan with RHA (L=Liver, GB=Gallbladder, r-
RHA=replaced right hepatic artery, SMA=Superior mesenteric
artery, IVC=Inferior vena cava, Ao= Aorta, B= Common bile duct
with endoprosthesis, D=Duodenum, P=Pancreas, S=Spleen).
Despite numerous efforts, national experts have yet to fully agree upon the
precise definitions of resectability. The definitions vary slightly between
academic centers and surgical societies with respect to tumor relationship
with the surrounding vasculature. This makes communicating clinical
decisions more difficult and makes standardizing research efforts more
challenging. It is our preference to follow the guidelines outlined by the
National Comprehensive Cancer Network (NCCN) in terms of definitions
and treatment protocols.
In brief review: two pancreas tumor locations within the gland are
broadly defined by the NCCN as head/uncinate process or body/tail (Table 1
). The arterial and venous systems are each independently evaluated in
relation to the tumor location. A resectable tumor located anywhere in the
pancreas is one that displays no arterial contact, with a clear fat plane around
the celiac axis (CA), the superior mesenteric artery (SMA), and the common
hepatic artery. Additionally, a resectable tumor does not contact greater than
180 degrees of the SMV or PV and causes no vein contour irregularity. An
unresectable tumor in the head/uncinate process contacts greater than 180
degrees of the SMA, CA, or contacts the first jejunal SMA branch. In the
body/tail, an unresectable tumor contacts greater than 180 degrees of the
SMA or CA or has evidence of contact with the CA and aortic involvement.
With regard to the venous system, a PDA anywhere in the gland is deemed
unresectable if the SMV/PV are believed to be unreconstructable either due
to tumor involvement or occlusion, or if there is tumor involvement with the
first draining venous branch of the jejunum. (Of note, in rare instances we
have resected PDAs that meet this last venous unresectability criterion, and
we have not done a reconstruction, leaving the patient’s own varicosities
intact, to allow venous drainage from the small and large bowel.)
Identifying and acting expeditiously once a resectable tumor has been
discovered, we believe, is crucial in the treatment of early stage PDA.
Surgery is the cornerstone of curative therapy, as no other treatment modality
has the capacity to cure this disease. Achieving a margin negative, termed an
R0, surgical resection is the best chance for a cure that we can provide to a
patient. Therefore, we believe the best initial approach is to remove these
tumors, provided we are able to do so safely. It has been shown that when an
R0 resection is achieved, overall survival is improved and cure, even
without adjuvant therapy, is possible, thus making this an important goal in
therapy. While attributing improved prognostic implications of an R0
resection entirely to the resection itself is controversial; nonetheless, there is
no evidence to the contrary.
It is our current practice to treat all patients with resectable PDA with
upfront or initial surgery and, once knowledge of the pathologic stage is
gained, treat with post-operative adjuvant therapies as indicated. We obtain
standard templated surgical consents (for pancreatectomy) and educate
patients well in advance of their actual operation. All patients are offered
participation in various prospective randomized clinical trials (RCTs) that
we may offer. Alternative treatment strategies (such as neoadjuvant
chemotherapy options) are also explained and offered to the patient in a
multidisciplinary fashion.
When a borderline resectable tumor is discovered, the treatment
algorithm is variable between clinicians. Individual surgeon interpretation of
the radiographic studies also affects how to approach these cases. As
previously mentioned, the terminology used in the definition of borderline
resectable disease differs between national experts and, furthermore, results
in variable approaches. The area of greatest variation involves tumors that
demonstrate venous involvement. The key factors to note regarding venous
involvement are whether there is:
Greater or less than 180 degrees of tumor contact or abutment with the
SMV-PV confluence;
Presence or absence of venous occlusion or thrombosis, or wall contour
irregularity amenable to resection or reconstruction.
Again, our stance is aligned with the NCCN guidelines (Table 1 ). We
believe many borderline cases, in particular those due to minor venous
abutment of the porto-splenic confluence, warrant and benefit from an
exploration. In our practice, these patients are taken to the operating room
without delay, as long as the venous structures do not appear deformed or
show evidence of contour irregularity. In doing so, direct assessment and
visualization of the anatomy can be performed at surgery. A consensus
statement was published on this point, authored by Bockhorn et al in 2014
from the International Study Group of Pancreatic Surgery (ISGPS). It was
recommended that if venous reconstruction were possible, a patient should
be offered an exploration and resection. Furthermore, whether or not venous
involvement should even be classified as borderline has been called into
question. Although resectability in the face of vascular involvement remains
controversial in terms of improvement in overall survival, there is a general
acceptance nationally of the efficacy and appropriateness of venous
resections.
In cases of locally advanced PDA and particularly in cases of arterial
involvement, we are in agreement that neoadjuvant therapy should precede
surgical treatment strategies. Unfortunately, as a surgical community, we have
not been able to complete a phase III randomized-controlled trial to evaluate
the efficacy of neoadjuvant therapy in the context of resectable or borderline
resectable PDA. One study, first authored by Katz et al in 2012, performed
with gemcitabine-based chemotherapy showed that neoadjuvant therapy
made little difference in tumor burden size, resectability status, or ability to
achieve an R0 resection. Another 2015 study from the group at the
Massachusetts General Hospital, first authored by Ferrone et al, reported on
the results of a FOLFIRINOX-based neoadjuvant regimen resulting in no
correlation between post-treatment imaging and resectability status.
OUR TECHNIQUE
In the case of the patient highlighted in this chapter with pancreas
protocol CT images showing only venous abutment, we went through the
steps of preparing the patient for possible venous resection (see V. Venous
resections below). Indeed, we were pleased when the dissection for this
patient went smoothly; no vein resection was necessary, and ultimately a
standard pylorus-preserving pancreaticoduodenectomy was performed. To
describe this operation, we begin through a vertical midline incision from the
xiphoid process to umbilicus. Upon entering the peritoneal cavity we
perform a thorough exploration, ‘run’ the small intestine and look for
metastatic disease to the liver, ligament of Treitz, root of the mesentery, and
peritoneum. We begin the extirpative phase of this operation by first
performing a “top down” cholecystectomy. We dissect out the extrahepatic
biliary tree from the hepatoduodenal ligament, elevate the common hepatic
duct from the portal vein, and divide it. From there, we identify and dissect
out the gastroduodenal artery (GDA). In this particular case, the replaced
right hepatic artery was identified and carefully preserved. We test clamp the
GDA to ensure the absence of celiac stenosis and the presence of good flow
in the proper hepatic artery, and then divide it. We then elevate the first part
of the duodenum from the pancreatic neck and divide it with a linear dividing
stapler, approximately 3 cm below the pylorus, preserving the pylorus
proximally. We then take down filmy adhesions between the posterior distal
stomach and the anterior aspect of the pancreas. We identify and divide the
downstream GDA caudally, inferior to the stomach, which allows further
mobilization of the stomach to the left side. Next, we “Kocherize” the
duodenum, elevate the head of the pancreas up out of the retroperitoneum and
off of the IVC, and continue to dissect the uncinate process from the
transverse mesocolon. We then dissect out the SMV inferior to the pancreatic
neck. When able, in the absence of vascular involvement, we elevate the
pancreatic neck ventrally from the SMV-PV confluence and place a Penrose
drain through the tunnel we have created. We use electrocautery to divide the
neck and carefully dissect the neck and head of the pancreas away from the
SMV, PV, and SMA. Again, in this case, care was taken to preserve the
replaced right hepatic artery (Figure 2 ).
We then turn our attention to the ligament of Treitz, divide the proximal
jejunum 20 cm distally with a linear dividing stapler, and separate it from its
mesentery. The duodenojejunal junction is mobilized behind the mesenteric
vessels, and we deliver the proximal jejunal stump under the root of the
mesentery to the patient’s right side. We then divide the remaining
attachments between the specimen and the retroperitoneum (the so called
mesopancreas) and remove the specimen.
The reconstruction phase begins with an end-to-side invagination
pancreaticojejunostomy (PJ) using a 3-0 silk outer layer and 3-0 vicryl
continuous inner layer, performed over a pediatric feeding tube which is later
removed during the hepaticojejunostomy (HJ). Downstream from the PJ we
perform an end-to-side HJ using a single layer of interrupted 5-0 PDS suture,
beginning with the posterior layer and followed by the anterior layer,
performed over a T-tube modified to be an I-tube, which is also subsequently
removed intraoperatively. (Hence, no transanastomotic stents are left
behind.) Finally, for our duodenojejunostomy (DJ), we deliver the jejunum
through a rent in the transverse mesocolon and perform the anastamosis in
two layers, approximately 30 cm downstream from the HJ, using 3-0 silk and
3-0 vicryl suture. The DJ is then pulled caudally, to rest below the
mesocolon. We conclude by placing two intraperitoneal drains and closing
the mesenteric defects.
VENOUS RESECTION
It has been shown that patients who undergo a partial vein resection with
pancreaticoduodenectomy, when compared side by side with patients who
undergo a pancreaticoduodenectomy without vein resection, have the same
oncologic outcome, provided an R0 resection is achieved. Additionally,
there was no difference in survival in patients undergoing resection who
were found to have vein involvement. We have retrospectively collected an
unpublished series of patients considered to have borderline or locally
advanced disease who underwent neoadjuvant therapy, a quarter of which
ultimately came to surgery for resection. We have found in this small series
that no survival benefit was found for patients who received neoadjuvant
chemotherapy when compared to published survival rates of borderline
cases that undergo surgery first without neoadjuvant chemotherapy. From this
and other studies we conclude that in cases of isolated vein involvement,
proceeding with surgery followed by adjuvant therapy as indicated is a
reasonable approach and prevents the opportunity for a surgical delay and
disease progression to unresectable/locally advanced status. It has also been
shown that the addition of a venous resection is safe in experience hands and
does not cause any additional morbidity or mortality.
Despite optimal pre-operative imaging, we have found the degree of
venous involvement and associated challenges in dissection difficult to
interpret preoperatively. While some studies show a correlation between
venous invasion on preoperative imaging and postoperative histologic
venous involvement, our experience has found the degree of tumor adherence
unpredictable. This is due to misinterpretation of peritumoral inflammation
and other factors. If there is any question of vein involvement, we make the
appropriate preoperative preparations for a more aggressive operation, such
as a vein resection. We believe a patient with a tumor with venous
involvement that is less than or equal to 180 degrees of the PV-SMV
confluence without vein contour irregularity is one that can be safely taken to
the operating room with a surgery first approach. Often the dissection of the
vasculature and removal of the tumor will be more facile than predicted.
If we feel a venous resection may be necessary, we begin by evaluating
the type of reconstruction that may be required and evaluate available
vascular conduits for harvest. On the day of surgery, we communicate with
the anesthesia team to leave one internal jugular vein exposed and available
to us for harvest. We will sterilely prep out the patient accordingly, with one
side of the neck for potential internal jugular vein harvest, as well as
bilateral groins for saphenous vein harvest. We ask and ensure that the
anesthesia team has the necessary vascular access and blood products
readily available should we encounter blood loss requiring transfusion.
Intraoperatively, when we discover and decide that we will be
performing a venous reconstruction, we adhere to a number of principals.
First, we will complete all other aspects of the pancreatectomy before
attempting separation of the tumor or any part of the pancreas from the PV
and/or SMV. It is wise to have as much of the resection accomplished before
working on the areas of potential venous attachment or encasement. When
appropriate for the size of an anticipated small defect, we will harvest a
segment of the left renal vein to use for a patch venoplasty. Using the left
renal vein prevents the morbidity of a second operative site, while venous
drainage of the left kidney flows through the adrenal and gonadal veins. If we
believe a larger segment of vein is required, our options are to use saphenous
vein or internal jugular vein. Saphenous vein is useful in the event that we
need a longer segment interposition graft for segments of the SMV; internal
jugular vein we have found to be an excellent size-match for portal vein
reconstructions. Again, before attempting separation of the PV or SMV, we
will complete all other aspects of the resection. We obtain proximal control
of the SMV and side branches, splenic vein, and distal control of the portal
vein.
Four types of venous resections listed below and outlined in Figure 3 are
agreed upon by the ISGPS and range from simple closure to interposed
conduits. These are:
Type 1: Partial venous excision with direct closure (venorraphy) by
suture closure
Type 2: Partial venous excision using a patch venoplasty closure
Type 3: Segmental venous resection with primary veno-venous
anastomosis
Type 4: Segmental venous resection with interposed venous conduit and
at least two anastomoses
Optimally, and with the help of preoperative imaging, the most commonly
encountered situation is the focal adherence of the tumor to the side wall of
the SMV or PV and a simply venorraphy can be performed.
POST-OPERATIVE CARE
Approximately 80% of patients who undergo a pancreas resection at our
institution adhere to our standard clinical pathway protocol for postoperative
care. This pathway both intensively monitors patients early in the
perioperative period and ensures a timely discharge, generally on post-
operative day six or seven following pancreaticoduodenectomy and day 5
following distal pancreatectomy. Initially, these patients spend one night in
the ICU with clinical monitoring devices inclusive of a central venous
catheter, an arterial line, a Foley catheter and a nasogastric tube (NGT). On
post-operative day one, if the patient is clinically well, we begin de-
escalation of monitoring devices with removal of the NGT and arterial line.
We will transfer the patient to the general surgery floor and ensure that the
patient is out of bed and ambulating. We remove the Foley catheter and begin
early diuresis on post-operative day two and monitor the patient’s intake and
output as well as daily weight. One drain is typically removed on post-
operative day three. By post-operative day four, the patient is started on a
regular diet, all medications are taken orally, and if the remaining Jackson-
Pratt drain output is non-sinister, it is removed.
For the patient in the case scenario, her post-operative course followed
the standard pathway without deviation. Her pathology showed a 2.5 cm
moderately differentiated adenocarcinoma of the pancreatic head with all
surgical margins negative for tumor and zero out of fifteen lymph nodes with
evidence of metastatic tumor. As for adjuvant therapy, our preference is for
our patients to establish a relationship with the medical oncologists while
inpatient and schedule their port placement for four to five weeks after
discharge. In this case, the patient was enrolled in the APACT trial, a phase 3
randomized, multicenter trial evaluating the use of adjuvant nab-paclitaxel
plus gemcitabine versus gemcitabine alone. Generally we advise adjuvant
therapy to commence six weeks from the date of the pancreatectomy.
CLINICAL TRIALS
We have a number of investigator-initiated trials, which we explain and
enroll patients in routinely. In anticipation of a pancreaticoduodenectomy, we
will enroll appropriate patients in a trial entitled the WARP trial (Whipple
Accelerated Recovery Pathway, NCT #02517268), which will aid in
defining an optimal, streamlined post-operative clinical care pathway.
Patients on the experimental WARP trial arm have increased daily physical
activity, slower advancement of their diet, and discharge on post-operative
day five, with close telephone follow up and intense outpatient follow-up
through telehealth appointments. We also will enroll patients in the WASH
trial (abdominal lavage of distilled WAter or Saline at High volumes, NCT#
02757859) which evaluates intraoperative tumor cell shedding and survival,
using serial lavage and dilutions. As a result of our HYSLAR trial (the use of
a restrictive fluid regimen with 3% Hypertonic Saline versus Lactated
Ringers, NCT# 01428050), the patient discussed herein was intraoperatively
and postoperatively maintained on 3% hypertonic saline and did quite well
without any perioperative complications.
SUMMARY
Venous abutment does not preclude patients from a surgery- first approach to
treating PDA. Optimal CT imaging and interpretation in conjunction with the
attending surgeon is imperative for proper resectability status categorization
of PDA. The term “Borderline Resectable” and the treatment implications
from this categorization are not nationally agreed upon. A randomized
controlled trial is needed to elucidate the best initial approach to PDA
patients with resectable and borderline resectable disease. Venous resections
for PDA are well tolerated and rarely necessary; if anticipated, appropriate
preparations should take place.
SALIENT POINTS
All patients with PDA should be referred to high-volume centers and
treated by experienced individuals in a multidisciplinary fashion.
Achieving an R0 resection has been shown to improve overall
survival.
Avoid over-classifying tumors as locally advanced, unresectable or
borderline that on suboptimal imaging appear to have isolated venous
involvement, as many of these patients could benefit from a resection.
Carefully review all available radiographic studies yourself; avoid
misinterpreting sub-standard studies and prepare for all pre, peri, and
post-operative aspects of patient care.
All patients with PDA should be enrolled in clinical trials as
appropriate.
SELECTED REFERENCES
1. Bockhorn M, Uzunoglu FG, Adham M, Imrie C, Milicevic M, Sandberg
AA, et al. Borderline resectable pancreatic cancer: a consensus
statement by the International Study Group of Pancreatic Surgery
(ISGPS). Surgery. 2014; 155(6): 977-88.
2. Kelly KJ, Winslow E, Kooby D, Lad NL, Parikh AA, Scroggins CR, et
al. Vein involvement during pancreaticoduodenectomy: Is there a need
for redefinition of a “Borderline Resectable Disease?” J Gastrointest
Surg. 2013;17(7):1209-17.
3. Katz MH, Fleming JB, Bhosale P, Varadachary G, Lee JE, Wolff R, et
al. Response of Borderline Resectable Pancreatic Cancer to
Neoadjuvant Therapy Is Not Reflected by Radiographic Indicators.
Cancer. 2012; 118(23):5749-56.
4. Murakami Y, Uemura K, Sudo T, Hashimoto Y, Nakashima A, Kondo N,
et al. Benefit of Portal or Superior Mesenteric Vein Resection with
Adjuvant Chemotherapy for Patients with Pancreatic Head Carcinoma.
Journal of Surgical Oncology. 2013;107(4):414-21.
5. Maley WR & Yeo CJ. (2016). Vascular Reconstruction During the
Whipple Procedure. Cameron JL & Cameron AM (Eds.), Current
Surgical Therapy, 12th ed. (pp. 549-553). Philadelphia: Elsevier.
6. National Comprehensive Cancer Network, NCCN Guidelines, version
2.2016, Pancreatic Adenocarcinoma.
█ BACKGROUND
Pancreatic ductal adenocarcinoma (PDAC) is one of the leading causes of
cancer-related mortality in the United States. Only about 7% of patients
newly diagnosed with PDAC live 5 years or longer. Even patients who
present with a radiographically localized cancer and who undergo de novo
resection of their primary tumor and regional lymph nodes are likely to
develop recurrent disease after this “potentially curative” treatment. Eighty
percent of them will do so within 2 years following surgery, and their
recurrent cancer will invariably lead to their death. Although overall
survival following pancreatectomy for PDAC is definitively prolonged by
the postoperative administration of systemic chemotherapy, that improvement
is incremental (on the order of 3 months!), and as many as 50% of patients
who undergo surgery never receive intended chemotherapy thereafter.
Major advances in the care of patients with localized PDAC have come
with time, but they have come slowly. Indeed, a series of patients who were
treated with pancreatectomy de novo over four decades at a high-volume
pancreatic treatment center justifiably considered to be one of the world’s
finest showed important but only relatively marginal improvements in
survival over recent years. Today, a highly functional patient with localized
PDAC who undergoes pancreatectomy performed by an expert surgeon and
who receives postoperative chemotherapy as prescribed can expect to live
only slightly longer than 2 years. Additionally, although about 20% of
patients who undergo resection de novo are effectively cured of their cancer,
an equivalent or even higher percentage will die within a year of surgery.
Clearly, the strategy of treating patients with PDAC with surgery upon
diagnosis—though it remains the standard of care—is unacceptable.
The delivery of nonoperative therapies prior to surgical resection, instead
of following it, represents an alternative treatment strategy. Although high-
level studies have not demonstrated survival benefit to this approach relative
to the standard of care, significant empirical data and a sound theoretical
basis exist to support it. Purported benefits of preoperative therapy include
facilitation of a margin-negative resection, the early treatment of
micrometastatic disease that is present in all patients, the selection of a
physiologically robust population of patients with “biologically favorable”
disease, and the assurance that all patients who undergo surgery receive
combined modality therapy. Long discounted by surgeons concerned about
losing the window for a “potentially curative” operation, this strategy is
becoming increasingly accepted nationwide, and it was recently recognized
in guidelines published by the American Society for Clinical Oncology as
entirely appropriate for at least some patients with localized PDAC.
Here we argue that the administration of preoperative therapy is rational
for many—if not most—patients with localized PDAC because, as we will
show, the primary use of surgery for PDAC is contrary to everything we have
learned about the biology of this cancer and the physiology of the patients
who are afflicted with it.
Preoperative Therapy Makes Good Surgery Better
The safe conduct of a margin-negative resection represents the primary
obligation of the pancreatic surgeon. Unfortunately, both the anatomy and
biologic behavior of PDAC tumors are such that positive surgical margins
and locoregional cancer recurrence represent significant problems for even
the most technically gifted surgeon.
Because most pancreatic cancers arise in the head or uncinate process
within millimeters of the superior mesenteric artery, and because cancer
cells infiltrate into the perineural and lymphatic tissues of the
retroperitoneum, proper conduct of pancreatoduodenectomy requires
skeletonization of the right lateral aspect of the superior mesenteric artery
from the level of the first jejunal branch of the superior mesenteric vein to the
takeoff of the artery from the aorta. Even when this dissection technique is
used at pancreatoduodenectomy, however, microscopic disease is often left
behind in the local tumor bed (Figure 2 ). Although the likelihood of positive
margins at surgery can be minimized by the application of meticulous
surgical technique in patients with well-selected tumor anatomy, it cannot be
entirely eliminated. Indeed, series from large, high-volume pancreatic
treatment centers report microscopically positive resection margin (R1) rates
following pancreatoduodenectomy as high as 50%. Moreover, recent studies
using rigorous histopathologic protocols have found cancer cells in at least
one surgical margin in as many as 90% of all pancreatoduodenectomy
specimens—an observation congruent with the reality that up to 80% of
patients who die from PDAC do so with locally recurrent cancer.
FIGURE 1: Superior mesenteric artery margin distance estimated
radiographically and measured histopathologically in two patients.
Red line depicts path of surgical resection. Small arrow highlights
tissue between SMA and tumor; large arrow demonstrates cancer
cells in proximity of inked margin. A,B: A patient who underwent
pancreatoduodenectomy with tangential vein resection and who had
tumor cells within 1 mm of the inked retroperitoneal margin. C,D: A
patient who underwent pancreatoduodenectomy with a positive
retroperitoneal margin. From Katz et al. Effect of neoadjuvant
chemoradiation and surgical technique on recurrence of localized
pancreatic cancer. J Gastrointest Surg. 2012;16(1):68-78-79
OUTCOMES OF PREOPERATIVE
THERAPY
We recently reported our experience with the administration of preoperative
therapy at the University of Texas MD Anderson Cancer Center: a total of
622 patients who received chemotherapy and/or chemoradiation prior to
pancreatectomy over a 25-year period. Despite an increase in the relative
percentage of patients who were operated for “borderline resectable” or
locally advanced tumors as well as an increase in the use of vascular
resection and reconstruction concurrent with pancreatectomy over time, we
observed a steady increase in survival. The median overall survival of
patients treated between 2010 and 2014, the most recent time period we
examined, was 43 months, and no patient operated on within this period died
within 90 days of surgery. Although it is easy to overgeneralize these single-
institution results, they are clearly remarkable when compared to the median
survival duration of 28 months among patients treated with pancreatectomy
and the most modern postoperative therapy regimen and a nationwide
perioperative mortality rate of 7.4%.
A PLANNED REBUTTAL
Critics of preoperative therapy for PDAC will object to the arguments for
preoperative we have presented. They will lament the possibility that the
window for surgery—and therefore the potential for cure—will close during
the administration of therapies that have never been shown to be curative on
their own. Or, they will voice concerns about the adverse events associated
with preoperative chemotherapy or radiotherapy that may prohibit surgery. In
reality, however, isolated locoregional progression that precludes subsequent
resection is exceedingly rare, and treatment-related complications sufficient
to prevent surgery are also distinctly uncommon in the absence of systemic
disease progression, when care is delivered at experienced centers.
Critics will also note the total absence of high-level data that
unequivocally demonstrates the superiority of this alternative approach over
standard care. They will reference two randomized controlled trials (RCT)
from Europe, both of which randomized patients with resectable PDAC to
either surgery or preoperative chemoradiation, and neither of which detected
a difference in median overall survival between the treatment arms.
However, neither of these studies met their primary endpoint because they
failed to accrue patients: of a combined total of 318 patients targeted for
accrual, only 111 patients were enrolled. A larger European study that
randomizes patients to surgery or preoperative chemotherapy may be more
successful, but irrespective of its findings, the tide has already turned. There
is not a major academic treatment center in the United States that does not
have an active preoperative therapy trial open for accrual. ASCO guidelines
support the use of preoperative therapy at least for some patients. And, the
only two trials for localized pancreatic cancer offered through the National
Clinical Trials Network have been designed to evaluate preoperative
strategies, and neither of these randomized studies offers a surgery-first
control arm.
CONCLUSIONS
Significant advancements in the care of patients with localized PDAC
continue to lag behind the strides made elsewhere in oncology. It is clear
from decades of experience that the current dictum of surgery followed by
adjuvant therapy leads to unacceptable long-term results even in the best of
patients. In contrast, a sound theoretical argument can be made to justify the
delivery of nonoperative therapies in the preoperative setting, and increasing
empirical data support it. As systemic regimens improve in the near future,
decisions regarding treatment sequencing will only become more important.
Ultimately, the care of these patients is optimized when a community of
medical, radiation, and surgical oncologists alongside dedicated
radiologists, pathologists, gastroenterologists, and other team members are
engaged in their multidisciplinary management. In this chapter, we have
shown that the administration of preoperative therapy is appropriate for at
least some patients with PDAC—irrespective of the technical resectability of
the primary tumor. Therefore, the debate over this issue, we believe, should
now evolve to more mature questions regarding which regimens to use, the
optimal duration of therapy, the best sequencing of components, identifying
measures of response, personalizing therapy, introducing newer agents, and
more.
SALIENT POINTS
Potential advantages of a therapeutic approach for patients with
localized PDAC in which chemotherapy, radiotherapy, or novel agents
are administered prior to pancreatectomy include facilitation of a
margin-negative resection and improved local control, early systemic
treatment for a presumably systemic disease, optimization of
nutritional, physiologic and comorbid impairments prior to major
surgery, and ensuring delivery of all nonoperative therapies.
Although data from randomized controlled trials does not exist,
retrospective data demonstrates favorable long-term outcomes which
have improved over time
Complications from treatment, declines in performance status, or
locoregional progression occurring during neoadjuvant therapy that
prevent subsequent resection are rare.
SELECTED REFERENCES
1. Khorana AA, Mangu PB, Berlin J, Engebretson A, Hong TS, Maitra A,
et al. Potentially Curable Pancreatic Cancer: American Society of
Clinical Oncology Clinical Practice Guideline. J Am Soc Clin Oncol.
2016;34(21):2541-56.
2. Cloyd JM, Crane CH, Koay EJ, Das P, Krishnan S, Prakash L, et al.
Impact of hypofractionated and standard fractionated chemoradiation
before pancreatoduodenectomy for pancreatic ductal adenocarcinoma.
Cancer. 2016;122(17):2671-9.
3. Denbo JW, Bruno ML, Cloyd JM, Prakash L, Lee JE, Kim M, et al.
Preoperative Chemoradiation for Pancreatic Adenocarcinoma Does Not
Increase 90-day Postoperative Morbidity or Mortality. J Gastrointest
Surg. 2016; 20(12):1975-1985.
4. Rhim AD, Mirek ET, Aiello NM, Maitra A, Bailey JM, McAllister F, et
al. EMT and dissemination precede pancreatic tumor formation. Cell.
2012;148(1-2):349-61.
5. Cloyd JM, Katz MH, Prakash L, Varadhachary GR, Wolff RA, et al.
Preoperative Therapy and Pancreatoduodenectomy for Pancreatic
Ductal Adenocarcinoma: A 25-Year Single-Institution Experience. J
Gastrointest Surg. 2017; 21(1):164-174.
█ BACKGROUND
PDAC is one of the leading causes of cancer-related mortality in the United
States, largely due to the fact that the majority of patients have metastatic
disease at the time of presentation. Survival rates are better in patients with
localized disease amenable to surgical resection, especially when combined
with multimodality therapy. Historically, vascular involvement, whether
diagnosed radiographically prior to surgery or at the time of laparotomy, was
considered a contraindication to surgical resection. Therefore, patients with
localized cancers and major vessel involvement were treated solely with
chemotherapy and radiation. Since surgical resection is considered
necessary, though not necessarily sufficient, for curative intent, attempts to
extend the criteria of resectability to patients with vascular involvement have
long been of interest.
The first large series of vascular resection at the time of pancreatectomy
was described by Fortner as he introduced the concept of “regional
pancreatectomy” in the 1970s. Although Fortner’s original description
demonstrated the technical feasibility of venous and arterial resections with
pancreatectomy, the procedure was not associated with improved
oncological outcomes and therefore was largely abandoned. Nevertheless,
over the past several decades, developments in cross sectional imaging,
preoperative staging, vascular techniques, and multimodality therapy have
led investigators to again consider the merits of vascular resection at the time
of pancreatectomy. In fact, pancreatectomy with venous resection comprises
a substantial proportion of operations for PDAC at high volume centers and
large series suggest no worse survival in patients that are able to undergo a
margin negative resection.
PATIENT SELECTION
Since microscopic and macroscopic resections are associated with both
locoregional recurrence and reduced overall survival, the likelihood of
obtaining a margin negative resection must be critically assessed when
evaluating whether to proceed with surgical resection. As the ability to
safely perform complex vascular resections at the time of pancreatectomy has
evolved, a new radiographic classification system was needed to more
accurately characterize the anatomic relationship of the tumor to its nearby
vascular structures. Broadly defined as potentially resectable, borderline
resectable (BR) and locally advanced (LA) (Table), these designations
provide important prognostic information regarding the ability to achieve a
margin negative resection, the likely need for vascular resection, and the
selection of preoperative therapies prior to surgery.
TABLE 1: Criteria Defining Resectability Status (Data
derived from the National Comprehensive Cancer Network.
NCCN Guidelines, version 2.2016, Pancreatic
Adenocarcinoma).
SURGICAL ANATOMY
Venous Anatomy
The pancreas is anatomically intimately associated with critical vascular
structures including the PV, SMV, celiac artery (CA), hepatic artery (HA) and
superior mesenteric artery (SMA). The SMV drains the midgut and courses
posterior to the neck of the pancreas, joining the splenic vein (SV) in order to
form the PV. The IMV typically drains into the SV but occasionally enters in
to the SMV or forms a common confluence with SMV and SV. The
gastrocolic trunk is an early anterior tributary of the SMV and often divides
into the right gastroepiploic and middle colic vein. The first jejunal vein
typically runs posterior to the SMA to join the ileal branch in order to form
the SMV. The first jejunal vein carries particular significance as it can be
directly invaded by tumors of the uncinate process, preventing separation of
the tumor from the SMV. In addition, injuries of the first jejunal vein can lead
to excessive hemorrhage since vascular control at the root of the mesentery
can be extremely challenging.
Knowledge of the venous anatomy is critical to understanding and
applying sound technical principles for venous reconstruction. In general, the
limits of venous resection extend proximally from the bifurcation of the left
and right PV to the first jejunal/ileal branches in the root of the mesentery.
The two most important factors to consider are the location of the tumor
relative to the portosplenic confluence and the extent of involvement.
Classification systems exist that describe the extent of tumor involvement as
a means of guiding the reconstruction (Figure 2 ).
Arterial Anatomy
The SMA originates from the aorta and courses posterior to the pancreas as
it enters the root of the mesentery. The SMA maintains a fairly constant
posteromedial relationship to the SMV and is close in proximity to the SMV
throughout its length. Therefore, it is uncommon for pancreatic tumors to
directly invade the SMA in the absence of venous involvement. On the other
hand, the SMA is surrounded by a rich network of nerve and lymphatic
tissues that frequently harbor microscopic tumor cells directly invading from
adenocarcinomas of the head or uncinate process. For this reason, a
meticulous dissection during PD along the periadventitial plane of the SMA
is encouraged to maximize the retroperitoneal margin. Pancreatectomy with
concomitant SMA resections have historically been associated with
prohibitively high morbidity and mortality and are therefore not routinely
performed nor discussed further in this chapter.
TECHNICAL ASPECTS
Lateral Venorrhaphy
While the need for venous resection should be anticipated preoperatively
based on a thorough examination of cross-sectional imaging, a formal
assessment of the tumor-vein interface may be conducted after transection of
the pancreatic neck and obtaining vascular isolation and control. Venous
resection is indicated when, in the estimation of the surgeon, tumor adherence
prevents separation of the PV-SMV from the pancreatic head and uncinate
process. That is, venous resection is not performed routinely, for example, in
attempts to achieve a wider margin or enhance the lymphadenectomy.
When tumor involves only the right lateral wall of the SMV, a segment of
vein may be excised en bloc with the tumor. If the subsequent defect
comprises <25% of the vein diameter, lateral venorraphy with primary suture
may occasionally be performed. Transverse closure, as is done during
pyloroplasty for example, is the least likely method to result in stenosis. If
the venous defect is larger, or is located at the confluence, then patch
venoplasty with either autologous vein graft or bovine pericardial patch is
preferred (Figure 3 ).
FIGURE 3: Intraoperative photograph of patch venoplasty using
internal jugular autograft for repair of a V1 venous resection. This
reconstruction method is preferred when segmental resection is not
required but the venous defect is >25% of the diameter of the SMV-
PV or located at the portosplenic confluence. PV, portal vein; SV,
splenic vein; SMV, superior mesenteric vein
SEGMENTAL RESECTION AND
RECONSTRUCTION
For more extensive venous involvement, segmental resection should be
performed. Primary anastomosis is preferred when possible. In fact, defects
as long as 4cm can be repaired primarily after removal of the specimen and
complete mesenteric mobilization. When a tension free primary anastomosis
is not possible, an interposition graft should be used. Autologous grafts are
preferred over synthetic grafts due to their reduced risk of infection and
thrombosis. Our preference is the internal jugular (IJ) vein because of its
favorable size match to the PV-SMV and ease of harvest. Other autologous
options include the left renal vein (LRV) or saphenous vein. Alternatively, a
customized bovine pericardial tube graft (xenograft) can be fashioned to the
appropriate size using an endovascular stapler.
The possible need for interposition vein graft should be anticipated prior to
the operation and the patient prepped and positioned appropriately. If IJ vein
is the preferred conduit, a shoulder bump is placed and the head positioned
to the right in order to expose the left neck, which is then prepped and draped
at the start of the case. Once the tumor is deemed resectable and the surgeon
commits to proceed with PD (i.e., generally following division of the
pancreas and dissection of the specimen from the SMA), a longitudinal
incision is made along the left sternocleiodomastoid muscle and the internal
jugular vein is isolated and harvested from the clavicle to the facial vein; it
is then kept in heparinized saline until needed.
Once the SMA dissection is completed and the retroperitoneal tissues
divided, the tumor remains only attached to the PV-SMV. The right colon is
returned to its normal anatomic position in order to de-rotate the mesentery.
Low-dose systemic heparinization may be administered but is not required.
Vascular clamps are placed 2-3cm proximal and distal to the area of venous
involvement. SMA clamping is not routinely performed but may be used as a
means of preventing intestinal wall congestion that can interfere with
subsequent anastomoses, particularly if the reconstruction is anticipated to be
especially complex, or take a longer-than-average amount of time to
complete (e.g., oblique segmental interposition; or placement of a graft prior
to a difficult SMA dissection due to the presence of a large uncinate process
tumor, a close SMA margin or the presence of a replaced right hepatic
artery). Either primary anastomosis or insertion of an interposition graft is
then performed using 6-0 non-absorbable monofilament suture (Figure 4 ). In
the case of interposition grafting, we recommend performing the more
challenging side first. This is generally the distal anastomosis, as inadvertent
displacement of the distal vascular clamp can lead to significant hemorrhage,
which is difficult to control at the root of the mesentery. Occasionally,
however, the proximal anastomosis is more challenging and should be
performed first. This could be the case, for example, when the tumor is
located cephalad in the pancreatic head and extends to the undersurface of
the liver, forcing a portal venous anastomosis high in the hilum.
FIGURE 4: Intraoperative photograft of SMV-PV reconstruction
using internal jugular (IJ) vein as an interposition autograft. The
graft is fashioned using interrupted 6-0 Prolene and care is taken
to ensure that is not rotated. PV, portal vein; SMV, superior
mesenteric vein; LHA, left hepatic artery; RHA, right hepatic artery;
SMA, superior mesenteric artery
OUTCOMES
Short-Term Outcomes
PD with venous resection and reconstruction is a technically challenging
endeavor and one might expect higher rates of adverse postoperative events.
In fact, a study utilizing the National Surgical Quality Improvement Project
found higher postoperative morbidity (39.9% vs 33.3%) and mortality (5.7%
vs 2.9%) rates in patients undergoing PD with vascular resection compared
to those who did not undergo vascular resection. However, this finding, and
those of similar studies, are limited by the inability to differentiate planned
vascular resection and reconstruction from repair of inadvertent vascular
injuries at the time of surgery. Other single-institutional series, multi-
institutional studies, and meta-analyses have suggested that elective, planned
venous resection can be performed safely without significantly increased
morbidity rates compared to standard PD.
Long-Term Outcomes
Compared to non-operative therapies, PD with venous resection and
reconstruction leads to improved long term outcomes. In fact, it is now
universally accepted that venous involvement should not be an absolute
contraindication to surgical resection. Lygidakis et al randomized patients
with PDAC and limited vascular involvement to either PD with vascular
resection or double bypass and found 2 year survival rates of 81.8% vs 0%,
respectively. Other comparative studies have suggested similar oncologic
outcomes between patients with PDAC who underwent standard PD and
those who underwent PD with venous resection, especially when
preoperative therapy was utilized.
PATENCY
In general PV-SMV resection and reconstruction is associated with high
patency rates. Venous thrombosis can occur at any time after surgery, but
acute thrombotic events are associated with the highest morbidity. Late
thrombotic events are often related to cancer recurrence. Little literature
exists on risk factors for venous thrombosis after reconstruction, though some
studies have suggested that size mismatch, interposition grafting, duration of
the operation, and postoperative complications are associated with patency
failures. Primary suture repair, either lateral venorraphy or end-to-end
anastomosis, is associated with the lowest incidence of thrombosis.
SALIENT POINTS
With meticulous attention to preoperative cross-sectional imaging,
careful patient selection, the use of multimodality therapy, and sound
technical principles, the benefits of surgical resection can be safely
extended to those patients with PDAC and venous involvement.
The need for vascular resection should be anticipated based on the
preoperative imaging but critically assessed intraoperatively after
division of the pancreas.
Attempts at splenic vein preservation should be undertaken but not at
the expense of an oncologically inadequate SMA dissection.
The venous resection and reconstruction is deferred as the last step of
the operation and attempts should be made to perform primary repair
when possible.
When the venous defect cannot be repair primarily, an interposition
graft using autologous vein, should be fashioned.
When performed at experienced high volume centers in appropriately
selected patients, PD with concomitant venous resection is associated
with excellent short and long term outcomes.
SELECTED REFERENCES
1. Chua TC. & Saxena A. Extended pancreaticoduodenectomy with
vascular resection for pancreatic cancer: a systematic review. J.
Gastrointest. Surg. 2010; 14(9), 1442–52.
2. Katz MH, Pisters PW, Evans DB, Sun CC, Lee JE, Fleming JB, et al.
Borderline resectable pancreatic cancer: the importance of this
emerging stage of disease. J. Am. Coll. Surg. 2008; 206(5), 833-46;
discussion 846-8.
3. Katz MH, Fleming JB, Pisters PW, Lee, JE, Evans DB. Anatomy of the
superior mesenteric vein with special reference to the surgical
management of first-order branch involvement at
pancreaticoduodenectomy. Ann. Surg. 2008; 248(6), 1098–102.
4. Tseng JF, Raut CP, Lee JE, Pisters PW, Vauthey JN, Abdalla EK, et al.
Pancreaticoduodenectomy with vascular resection: margin status and
survival duration. J. Gastrointest. Surg. 2004; 8(8), 935-49; discussion
949-50.
5. Cloyd JM, Katz MH, Prakash L, Varadhachary GR, Wolff RA, et al.
Preoperative Therapy and Pancreatoduodenectomy for Pancreatic
Ductal Adenocarcinoma: a 25-Year Single-Institution Experience. J.
Gastrointest. Surg. 2017; 21(1), 164–174.
█ BACKGROUND
Pancreaticojejunostomy
On the other hand, PJ reconstruction remains the overwhelmingly applied
anastomotic technique around the world (89% of surgeons). From a technical
standpoint, there are myriad ways a duct-to-mucosa PJ can be created,
making head-to-head testing of efficacy nearly impossible. The author’s
preferred technique is described in great detail in the book chapter reference
provided below. Certainly, other approaches to PJ reconstruction with
different nuances are available, but this described duct-to-mucosa technique,
refined over hundreds of cases now, has resulted in a more than satisfactory
outcomes for us. This includes a 2.7% CR-POPF rate in the last 150 PJs,
including no fistulas in all (0/11) high-FRS cases.
The utility of these variations has been studied within the Pancreas Fistula
Study Group database. In a series of 459 High FRS PJs, 88
dunking/invaginating anastomoses were employed - most commonly when the
pancreatic duct was ≤1 mm. In this subset, it was not associated with a
decreased occurrence of CR-POPF (22.9% vs. duct-to-mucosa: 17.5%,
p=0.454). In summary, given this data, we use the dunking technique only in
the very rare circumstance when there is a 1mm or imperceptible duct (<3%
of all cases), where a duct-to-mucosa approach is practically impossible.
Duct Occlusion
Occlusion of the pancreatic remnant duct without anastomosis, as originally
implemented by Whipple during his initial cases, remains a possibility in
certain, usually extreme, operative circumstances. When a duct cannot be
identified and the operation need be hastened for physiologic instability or
other reasons, the distal pancreatic remnant can be closed in a number of
ways. Through-and-through 2-0 silk sutures may be placed in an interlocking
U fashion at the cut pancreatic end. If a duct can be visualized, it can
separately be suture ligated or sealed with a polymer prior to closing the
pancreatic capsule. Alternatively, a stapled closure, using an appropriate
sized staple for the parenchymal thickness encountered, is a swift alternative
in cases of physiologic instability.
Regardless of the method, the adjacent area should be externally drained
as these occlusions tend to leak pancreatic juice over the short term, while
rarely becoming a significant, externalized fistula. While exocrine
insufficiency may develop, islet cells are often preserved and overt diabetes
is infrequent. Due to its infrequent application, and lack of studies in the
modern era, there are no significant data to support its regular use in all PDs
– particularly given obligate exocrine insufficiency. We offer the Whipple
procedure as a “bail-out” maneuver in a case of extreme risk or physiologic
decompensation. This approach is quicker and less impactful than the
alternative of a “completion” total pancreatectomy, or external drainage of an
uncontrolled, open pancreatic duct.
TECHNOLOGICAL AND PHARMACOLOGIC
ADJUNCTS
Following creation of the anastomosis, there are several strategies available
to potentially mitigate the risk of pancreatic fistula. These include the
application of intraperitoneal drains, anastomotic stents, prophylactic
somatostatin analogues, and tissue sealants or autologous patches. As stated
earlier, the temptation may be that employing more of these is better – in
other words, “Throw everything and the kitchen sink at it”. However, we
believe that is not necessarily the best course.
Intraperitoneal Drainage
Contemporary randomized studies suggest that intraperitoneal drains are
valuable in cases of high fistula risk, and may decrease both the rate and
severity of CR-POPF. Following PD, the author routinely places one fluted
19-Fr Blake (silicon, closed suction) drain anterior to the anastomosis in
cases of moderate and high anastomotic risk (FRS 3-10). We avoid placing
rigid drains directly behind the anastomosis, as they can cause unnecessary
trauma to the posteriorly located portal and superior mesenteric veins. We
then adhere to a protocol of early drain removal, which is driven by the drain
amylase concentration on POD1. If the POD1 drain amylase is < 5000, and
there is no sinister effluent, then the drain is removed early (on POD3). This
has been successful and safe even in cases of a High FRS. If it is >5000, the
drain is removed later, at the surgeon’s discretion. It is important to note that
early drain removal for cases of POD1 amylase greater than 5000 has never
been tested in a randomized, controlled fashion, so we do not yet have
evidence that early drain removal in this scenario is either safe or valuable.
Transanastomotic Stents
Anastomotic stents can facilitate the creation of the duct-to-mucosa
anastomosis and may prevent its dehiscence by shunting digestive pancreatic
enzymes. Externalized transanastomotic stents diminished the rate of
clinically relevant fistula in two ISGPF-era RCTs of “risky” anastomosis
(i.e., non-dilated ducts +/- soft pancreatic texture). In a more rigorously risk-
adjusted analysis of the Pancreas Fistula Study Group, externalized stents
were not associated with improved fistula outcomes for Negligible, Low and
Moderate risk PDs (FRS 0-6). However, externalized stents in patients with
High FRS risk demonstrated markedly lower rates of CR-POPF. Moreover,
when stratifying comparisons by individual FRS risk factors in isolation,
there were no significant differences in rates of CR-POPF in the setting of
either a soft gland, a small duct, high-risk pathology, or elevated blood loss –
suggesting that the benefit of external stents may solely be in the cohort of
composite risk identified by multiple FRS factors (rather than individual risk
factors in isolation).
In contrast, according to a FRS risk-adjusted analysis, internalized stents
led to increased complications and should generally be avoided since they
do not diminish CR-POPF rates. It is unclear why this difference between
externalized vs. internal stents exists; what is even more intriguing is that
external stents appear to only be beneficial in the highest risk (i.e., High
FRS) anastomosis. One hypothesis involves competing factors between the
benefits of a stent, as mentioned above, and the detrimental influence of a
foreign body at the anastomosis. In cases of low and moderate risk, these
potential benefits may be outweighed; however, an externalized stent appears
to have a uniquely beneficial role in High FRS cases. Internal stents, by
definition, do not drain digestive enzymes to the external environment, and
thus may be a less effective shunt.
Tissue Patches and Sealants
In certain instances, the pancreas will have a fibro-fatty layer surrounding its
anterior capsule. If so, preserve this layer when diligently dissecting the
pancreas, and keep it attached to the pancreatic body if possible. It can be
tacked with silk sutures to the pancreaticobiliary limb to create a “hood” for
the anterior aspect of the anastomosis. The omentum, or alternatively the
umbilical ligament with its vascular pedicle, may also be used in a similar
fashion to “wrap” the anastomosis. There are no reasonable data to support
any value to either strategy in any degree of risk, but such tissue patches are
readily available, devoid of cost, and possibly beneficial. These adjuncts are
used always or frequently by just 13% of surgeons who perform PD.
In contrast, biologic sealants add increased cost to the operation without
proven efficacy. Interestingly, while infrequently used with regularity, 18%
of surgeons profess they use sealants “occasionally” (1-25% of their cases)
—probably reflective of cases with perceived higher risk for fistula
development. There is only one ISGPF-era RCT on the use of fibrin glue,
which distinguished between biochemical leaks and CR-POPFs; it did not
find a clinical value to this strategy. No inclusion criteria related to fistula
risk were employed; hence, the utility of sealants specifically in high-risk
cases is so far understudied. Likewise, an older RCT which included only
PDs where there was a soft gland and non-dilated duct did not find that
topical administration of a fibrin glue sealant altered the rate of fistula or
other intra-abdominal complications, with the caveat that this trial preceded
the consensus definition of CR-POPF. Neither sealants, nor patches, have
demonstrated any protective value in FRS-based risk adjusted analyses using
the Pancreas Fistula Study Group data.
Somatostatin Analogues
Prophylactic use of somatostatin, and its analogues—octreotide and
pasireotide—are some of the most studied, and controversial, adjuncts to
possibly mitigate the risk of pancreatic fistula. Octreotide prophylaxis has
been prospectively evaluated in over 20 trials; while large meta-analyses
have reported a decreased rate of pancreatic fistula, only three RCTs
employed established definitions for CR-POPF. Octreotide did not reduce
fistula rates in any of these contemporary studies, and the rates were instead
higher in the octreotide cohort in two of them. Moreover, none specifically
studied the high-risk anastomosis, identified by either the FRS or other
criteria.
On the other hand, retrospective, risk-adjusted analysis of the Pancreas
Fistula Study Group data has shown intriguing results. In fact, worse CR-
POPF and overall outcomes were seen with the application of prophylactic
octreotide; furthermore, the negative effect appears to be potentiated with
higher risk. In other words, when applied to High FRS scenarios, CR-POPFs
developed at an alarming rate of 65%, while that rate was just 18% when it
was not used in that setting. Based on these findings, the author has omitted
prophylactic octreotide from his practice and observed markedly improved
overall fistula outcomes in his last 100 PDs.
In contrast, the preliminary data for pasireotide is promising, with
reduced rates of clinically relevant fistulas in a single-center trial. However,
validation in a multi-institutional setting, increased access to the medication,
and data specific for the high-risk anastomosis are still lacking.
SALIENT POINTS
Risk adjustment methodologies, such as the Fistula Risk Score, are
useful for identifying high risk anastomoses as well as for comparing
outcomes of mitigation strategies across studies.
This situation demands the best skill and expertise available. It is not a
time to practice.
Stay true to your tenets: if you regularly use a duct-to-mucosa
pancreaticojejunostomy for reconstruction, this may be your best
choice. Alternative approaches (Dunking, Invagination, modified
Blumgart, single-layer) are options when there is a diminutive duct that
cannot be either identified or sewn.
Intraperitoneal drains and externalized transanastomotic stents may
improve patient outcomes in high-risk scenarios. There is insufficient
evidence to recommend the use of octreotide prophylaxis, biologic
sealants, Roux limbs, or tissue patches.
An optimal approach to the high-risk scenario has been identified,
which includes a pancreaticojejunostomy with an externalized
transanastomotic stent, an intraperitoneal drain, and the omission of
prophylactic octreotide.
SELECTED REFERENCES
1. McMillan MT, Malleo G, Bassi C, Sprys MH, Ecker BL, Drebin JA, et
al. Pancreatic Fistula Risk for Pancreatoduodenectomy: An International
Survey of Surgeon Perception. HPB(Oxford) 2017; 19(6):515-524.
2. Bassi C, Marchegiani G, Dervenis C, Sarr M, Abu Hilal M, Adham M,
et al. The 2016 Update of the International Study Group (ISGPs)
Definition and Grading of Postoperative Pancreatic Fistula: 11 Years
After. Surgery. 2017; 161 (3):584-591.
3. McMillan MT, Sprys M, Malleo G, Bassi C, and Vollmer CM. Defining
the Practice of Pancreatoduodenectomy Around the World.
HPB(Oxford). 2015; 17(12);1145-54.
4. Callery MP, Pratt WB, Kent TS, Chaikof E, and Vollmer CM. A
Prospectively Validated Risk Score Model for Pancreatic Fistula after
Pancreaticoduodenectomy. J Am Coll Surg. 2013; 216(1):1-14.
5. Vollmer CM. Pancreaticoduodenectomy: Pancreaticojejunostomy. In
Mulholland M, Hawn M, Hughes S, Albo D, Sabel M, and Dalman R,
eds. Operative Techniques in Surgery. First Edition. Lippincott
Williams & Wilkins, 2015.
6. McMillan MT, Soi S, Asbun HJ, Ball CG, Bassi C, Beane JD, et al.
Risk-Adjusted Outcomes of Clinically Relevant Postoperative Fistula
Following Pancreatoduodenectomy: A Model for Performance
Evaluation. Ann Surg. 2016; 264(2):344-52.
7. Ecker BL, McMillan MT, Asbun HJ, Ball CG, Bassi C, Vollmer CM, et
al. Characterization and Optimal Management of High-Risk Pancreatic
Anastomoses During Pancreatoduodenectomy. Ann Surg. 2017. In press.
█ BACKGROUND
“Throughout the surgical literature, in the past as well as currently, the
issue of drainage has been the subject of heated debate and remains the
source of abundant confusion…. The presence of a drain should not be
relied on instead of proper surgical management… drains cannot com-
pensate for poor surgical judgement and technique.” Margaret Levy, The
American Journal of Surgery, 1984.
EVIDENCE
The literature regarding routine drainage following distal pancreatectomy is
sparse, as most studies have focused on pancreatoduodenectomy rather than
distal resections. Table 1 summarizes key studies to date.
TABLE 1: Studies examining the use of routine surgical
bed drainage following distal pancreatectomy
CONCLUSION
Current evidence does not support the routine utilization of intraperitoneal
drains following elective distal pancreatectomy. Trials, including Level 1
evidence, suggest similar outcomes whether patients receive a drain or do
not. Ultimately, drain utilization should be at the discretion of the operating
surgeon.
SALIENT POINTS
Routine use of an intraperitoneal draiin after distal pancreatic resection
should not be considered mandatory.
The current literature suggests there is no advantage to intraperitoneal
drainage following elective distal pancreatectomy.
Inability to identify or ligate the main pancreatic duct or a soft remnant
gland may prompt consideration for drain placement.
Postoperative drain management should mimic that following
pancreatoduodenectomy.
SELECTED REFERENCES
1. Levy M. Intraperitoneal drainage. Am J Surg. 1984; 147(3):309-14.
2. Paulus EM, Zarzaur BL, Behrman SW. Routine peritoneal drainage of
the surgical bed after elective distal pancreatectomy: Is it necessary?
Am J Surg. 2012; 204(4):422–7.
3. Behrman SW, Zarzaur BL, Parmar A, Riall TS, Hall BL, Pitt HA.
Routine Drainage of the Operative Bed Following Elective Distal
Pancreatectomy Does Not Reduce the Occurrence of Complications. J
Gastrointest Surg. 2014; 19(1):72–9.
4. Correa-Gallego C, Brennan MF, D’Angelica M, Fong Y, DeMatteo RP,
Kingham TP, et al. Operative Drainage Following Pancreatic Resection.
Ann Surg. 2013;258(6):1051–8.
5. Van Buren G 2nd, Bloomston M, Schmidt CR, Behrman SW, Zyromski
NJ, Ball CG et al. A prospective randomized multicenter trial of distal
pancreatectomy with and without routine intraperitoneal drainage. Ann
Surg. 2017 Sep;266(3):421-31.
CASE SCENARIO
A 77-year-old male with a past medical history of hypertension, stage III
chronic kidney disease, hyperlipidemia, and a history of coronary artery
bypass grafting in 2002, presented with 2 weeks of vague right upper
quadrant fullness and unintentional 10-lb weight loss. Laboratory and
physical examinations were unremarkable. Computed tomography of the
abdomen demonstrated an irregular mass in the head/neck of the pancreas
with abutment, but not encasement, of the superior mesenteric vein, and
sparing of the superior mesenteric artery (SMA). He underwent 6 cycles of
neoadjuvant gemcitabine and nab-paclitaxel with excellent response, as
demonstrated by post-treatment imaging. He underwent exploration and
planned Whipple resection. FS analysis from the intraoperative neck margin
(IONM) demonstrated a focus of residual invasive adenocarcinoma. An
additional margin of approximately one centimeter was resected, which
returned negative for carcinoma, and pancreaticojejunostomy (PJ)
reconstruction was performed to this margin. The patient had an uneventful
recovery and remains disease-free 9 months later.
█ BACKGROUND
Does A Positive Pancreatic Neck Margin Reflect
Disease Biology Or Extent Of Initial Surgical
Resection?
Of the pancreatic margins in a Whipple’s resection (i.e., neck, radial,
SMA/uncinate, retroperitoneal), the IONM is most amenable to surgical
remediation; consequently, investigators have questioned whether the
practice of resecting additional parenchymal neck margin after initially
positive IONM FS is worthwhile, particularly if it converts the resection
from FS:R1 to R0 on permanent section (FS:R1→PS:R0). Although earlier
studies on this subject espoused an aggressive approach to neck margin
clearance in PDAC either by re-resection of the neck margin or via
conversion to total pancreatectomy (TP) in IONM-positive cases, the
preponderance of contemporary data suggest that this practice does not
impact survival.
The largest and most recent analysis of this question is a retrospective
study by Kooby et al. of 1399 PDAC patients undergoing
pancreaticoduodenectomy at 8 US academic medical centers between
January 2000 and August 2012. On FS, 152 patients (10.9%) were R1, and
51 patients (3.6%) had false-negative (FS-R0, PS:R1) margins. PS:R0
margins were achieved in 1196 patients (85.5%), 131 patients (9.3%)
remained PS:R1, and 72 patients (5.1%) were rendered FS:R1→PS:R0 by
additional resection. Median overall survival (OS) for PS:R0 patients was
21.1 months; however, OS for PS:R1 and FS:R1→PS:R0 patients were
similarly dismal (13.7 vs. 11.9 months; P<0.001; Figure 1 ). FS:R1→PS:R0
conversion was associated with a 55% increase in risk-adjusted mortality.
FIGURE 1: Data from Kooby et al. reporting on 1399 patients
demonstrating that patients converted from frozen section R1 to
permanent section R0 (FS:R1?PS:R0) margin have similarly dismal
outcomes compared with patients with R1 on PS (Reprinted with
permission from Wolters Kluwer in Kooby DA, Lad N, Squires M,
Maithel S, et al. Value of intraoperative neck margin analysis
during Whipple for pancreatic adenocarcinoma. A multicenter
analysis of 1399 patients. Ann Surg 2014;260:494-503).
The forerunner to this study was an analysis of 382 patients from Emory
University, some of whom were included in the previously discussed study
by Kooby et al., with similar conclusions—patients rendered FS:R1→PS:R0
and PS:R1 patients had similarly poor OS outcomes (11.3 vs. 11.0 months,
respectively). On multivariate analysis, poor histologic grade, larger tumors,
and positive retroperitoneal margin —but not FS of the IONM—were
independently associated with decreased OS. In a separate report of 202
Whipple’s resections for PDAC from the University of South Florida,
FS:R→PS:R0 (median OS 11 months) and PS:R1 (13 months) patients had
significantly worse OS compared with PS/FS:R0 patients (21 months). In the
discussion, the authors refer to the practice of re-resecting pancreatic neck
margins based on IONM-positivity as “doomed to failure because of
aggressive tumor biology.”
A closer dissection of these analyses underscores the bio logic
underpinnings of this phenomenon. In the multi-institutional study by Kooby
et al. discussed earlier, FS:R1→PS:R0 and PS:R1 patients had similarly
higher rates of larger tumors, nodal positivity, and perineural invasion
compared with PS/FS:R0 patients; moreover, when controlling for
SMA/retroperitoneal margin-positivity (which has confounded interpretation
of results from earlier studies), as well as after excluding patients who had
received neoadjuvant therapy, outcomes in the FS:R1→PS:R0 and PS:R1
cohorts were similarly dismal. These data speak to biologic tumor
aggressiveness in these patient subsets that is unlikely to be mitigated by
extent of surgical resection.
It is worthwhile discussing the role of conversion to TP—if necessary—
to achieve R0 resection in IONM-positive cases, as advocated by Schmidt et
al.; in 61 eligible patients for analysis (PS:R1 n=28 and FS:R1→PS:R0 via
TP n=33), median OS was significantly higher in patients undergoing TP
compared with PS:R1 patients (18 vs. 10 months, P=0.04). This study has
been criticized for failing to disregard the bias inherent in selecting patients
who are more likely to tolerate the morbidity of TP, as well as for its lack of
generalizability to practice patterns beyond high-volume tertiary care
centers. As such, long-term survival data of TP for PDAC derived from
national cancer registries have not reproduced these outcomes, even for R0
resection—in a recent review from the National Cancer Data Base (NCDB),
5-year actuarial survival was 13% and median OS 15.5 months in nearly
2600 patients (1998—2006). Of course, given its retrospective study design
and inability to decipher if TPs were being performed as a priori decisions,
or as a deliberate reaction to serially positive IONMs, these data are
difficult to interpret in the context of the question posed herein. Although
conversion to TP in IONM-positive cases is clearly not standard practice—
or our practice in most cases, for that matter—consideration may be given to
this approach with a critical view on patient fitness for TP, tolerability of the
ensuing exocrine/endocrine insufficiency, and assessment/interpretation of
disease biology, including response to neoadjuvant therapy (if applicable).
Suggested Approach
Therefore, in PDAC patients, our practice leans towards recognizing that
disease biology often trumps surgery, acknowledging that IONM is a
surrogate for biologic aggressiveness that is unlikely to be mitigated by
extent of surgical resection. That said, it is still our practice to routinely send
FS of the IONM during Whipple for PDAC; if positive, we perform another
transection of 1-2 cm thickness, and perform our PJ to this margin regardless
of FS result. Other experts in the field, however, have questioned the routine
practice of sending IONM FSs since it has little bearing on long-term
outcome—particularly in an era where cost containment and resource
limitations are a growing concern. In general, the HPB surgical community
agrees that the focus of our efforts should shift towards thoughtful and
evidence-driven patient selection, either by vetting disease biology with
constantly improving neoadjuvant systemic therapies or utilizing
molecular/immune prognosticators of outcome, in order to minimize the
incidence of IONM positivity in the first place.
Suggested Approach
When performing Whipple’s resection for MD-IPMN, in general, we adhere
to Fukuoka guidelines—we: (a) send FS of the IONM routinely; (b) take
additional margin if FS is positive for HGD or invasive carcinoma, but
desist further resection in the setting of negative margin or LGD; (c) consider
TP very selectively in preoperatively informed and medically fit patients if
margins are serially positive for HGD/invasion. Moreover, when PS reveals
HGD or invasive carcinoma at the final neck margin, undetected on FS, we
offer interval completion TP selectively to surgically fit and motivated
patients.
SALIENT POINTS
Although the practice of requesting intraoperative frozen section
analysis of neck margins from Whipple’s specimens in PDAC is fairly
ubiquitous, the data do not clearly support its routine application—
intraoperative conversion of R1 to R0 resection based on neck margin
analysis does not appear to impact long-term outcome; these patients
do similarly poorly compared with patients with definitive R1 margins
on permanent section.
It may be reasonable to consider re-resection of the pancreatic neck to
negative margins in PDAC patients demonstrating response to
neoadjuvant therapy; data guiding decision-making in this scenario are
scant.
In the era of improving anesthetic and perioperative surgical/medical
care, completion total pancreatectomy in serially neck margin-positive
PDAC cases may be considered very selectively in surgically fit
patients, particularly in cases demonstrating response to neoadjuvant
therapy.
Based on the current evidence, frozen section analysis of the neck
margin should be considered standard of care in MD-IPMN
Serial resection (and even completion total pancreatectomy in
appropriate candidates) should be offered to if HGD or invasive
carcinoma is detected at the margin; LGD does not mandate further
resection.
SELECTED REFERENCES
1. Kooby DA, Lad NL, Squires MH 3rd, Maithel SK, Sarmiento JM,
Staley, et al. Value of intraoperative neck margin analysis during
Whipple for pancreatic adenocarcinoma: a multicenter analysis of 1399
patients. Ann Surg. 2014;260 (3):494-501.
2. Schmidt CM, Glant J, Winter JM, Kennard J, Dixon J, Zhao Q, et al.
Total pancreatectomy (R0 resection) improves survival over subtotal
pancreatectomy in isolated neck margin positive pancreatic
adenocarcinoma. Surgery. 2007; 142(4):572-8.
3. Esposito I, Kleeff J, Bergmann F, Reiser C, Herpel E, Friess, et al.
Most pancreatic cancer resections are R1 resections. Ann Surg Oncol.
2008; 15(6):1651-60.
4. Tanaka M, Fernández-del Castillo C, Adsay V, Chari S, Falconi M, et
al. International consensus guidelines 2012 for the management of IPMN
and MCN of the pancreas. Pancreatology. 2012; 12(3):183-97.
5. Marchegiani G, Mino-Kenudson M, Sahora K, Morales-Oyarvide V,
Thayer S, Ferrone C, et al. IPMN involving the main pancreatic duct:
biology, epidemiology, and long-term outcomes following resection.
Ann Surg. 2015; 261(5):976-83.
█ BACKGROUND
Distal pancreatectomy is the procedure of choice for benign, premalignant,
and malignant lesions of the body and tail of the pancreas. In recent years
distal pancreatectomy has become a safe procedure that is being performed
more frequently because imaging studies are identifying more benign and
premalignant lesions. For patients with malignant lesions of the pancreas,
splenectomy is recommended to ensure an adequate lymphadenectomy. In
patients with benign lesions, distal pancreatectomy with preservation of the
spleen may be considered.
Spleen preserving distal pancreatectomy (SPDP) was first described by
Mallet-Guy and Vachon in 1943. This classic technique preserves the splenic
artery and vein by identification and ligation of the multiple small, short
vascular connections to the body and tail of the pancreas. In 1988, Warshaw
published an alternative technique in which the splenic artery and vein(s) are
ligated proximally and distally, with retention of the short gastric and left
gastroepiploic vessels to preserve blood flow to and from the spleen). In
subsequent years, the relative risks and benefits of distal pancreatectomy
with and without splenectomy have been studied, as well as method of spleen
preservation (splenic vessel ligation versus splenic vessel preservation).
There are several variables to take into account when considering distal
pancreatectomy with splenic preservation. The objective of this chapter is to
summarize the advantages and disadvantages to preservation of the spleen in
the setting of distal pancreatectomy and to compare methods of splenic
preservation.
SURGICAL TECHNIQUE
Minimally invasive pancreatic surgery has become increasingly common
over the last two decades. With improved laparoscopic technology and
increasing surgeon experience, a minimally invasive approach has been
shown to be comparable to open surgery with respect to morbidity and
mortality, and it is generally agreed upon that minimally invasive surgery is
not inferior to open surgery in terms of operative objective. A discussion of
the merits of surgical technique with respect to laparoscopic, robotic, or
open surgery is beyond the scope of this chapter and will be discussed
elsewhere. Basic considerations include surgeon expertise with each
approach, as well as patient-specific factors including body habitus,
pathology, location of the lesion, etc.
The technique of vessel preserving spleen preserving distal
pancreatectomy (VP-SPDP) is described below (Figure 2 ). The body and
tail of the pancreas are exposed by dividing the gastrocolic omentum, taking
care to stay outside the gastroepiploic vessels. The short gastric vessels
usually do not need to be divided, and every effort should be made to
preserve them if a splenic preserving procedure is being attempted. The
peritoneum is then incised along the inferior pancreatic border, and the
pancreatic body is separated from the retroperitoneum by means of sharp and
blunt dissection along its inferior border. The splenic vein can then be
identified from underneath the pancreas, and careful circumferential
dissection around the splenic vein is performed and a vessel loop placed
around the vein. The splenic artery can be identified from the under surface
of the pancreas by retracting on the vessel loop around the splenic vein, or it
can be identified along the superior border of the pancreas anteriorly. Once it
is dissected circumferentially, it is also controlled with a vessel loop. These
precautionary measures allow quick control of bleeding should a vascular
tear occur later in the procedure. Once the pancreatic body has been
adequately mobilized from the splenic vessels, the pancreatic parenchyma is
divided proximal to the lesion. Once the proximal pancreatic tissue is
divided, the specimen is grasped and gently retracted anteriorly to allow
further dissection of the vessels. The dissection proceeds toward the splenic
hilum in a medial-to-lateral direction. The pancreatic branches of the splenic
vein are sequentially identified, dissected free, and divided with ties, clips,
or an energy device. The branches of the splenic artery, which runs just
superior to the vein, are treated similarly. Special care must be taken as the
dissection approaches the hilum of the spleen.
FIGURE 2: Vessel preserving spleen preserving distal
pancreatectomy
LESION PATHOLOGY
Distal pancreatectomy with splenectomy is an operation that adheres to
oncologic surgical principles in that it is an en bloc resection and provides
adequate regional lymphadenectomy. Preserving the spleen necessarily limits
the regional lymphadenectomy, particularly with respect to the lymph nodes
within the splenic hilum, and in cases where the splenic vessels are spared,
lymph nodes along the vessels are potentially left in situ. Thus in cases of
adenocarcinoma or high grade neuroendocrine tumors of the pancreatic body
or tail, distal pancreatectomy with splenectomy is the appropriate operation.
In the setting of benign or premalignant lesions, e.g. intraductal papillary
mucinous neoplasm, mucinous cystic neoplasm, etc., distal pancreatectomy
without splenectomy can be considered.
In cases of chronic pancreatitis, it is our practice to remove the spleen as
our experience shows that patients have inferior postoperative outcomes with
respect to chronic pain when the spleen is spared.
SALIENT POINTS
Spleen preservation should be considered for benign and premalignant
lesions of the body and tail of the pancreas.
Splenic vessel preserving technique is superior to the vessel ligating
technique and splenectomy in terms of postoperative morbidity.
The primary constraints to splenic vessel preservation are vessel
anatomy and lesion location.
All patients should be vaccinated prior to distal pancreatectomy in
anticipation of potential splenectomy.
SELECTED REFERENCES
1. Mallet-Guy P, Vachon, A. (1943). Pancreatites Chroniques Gauches.
Paris: Paris, Masson et cie, 1943.
2. Warshaw AL. Conservation of the spleen with distal pancreatectomy.
Arch Surg. 1988; 123(5): 550-3.
3. Shoup M, Brennan MF, McWhite K, Leung DH, Klimstra D, Conlon
KC. The value of splenic preservation with distal pancreatectomy. Arch
Surg. 2002; 137:164-8.
4. Davidson RN, Wall RA. Prevention and management of infections in
patients without a spleen. Clin Microbiol Infect 2001;7(12):657–60.
5. Beane JD, Pitt HA, Nakeeb A, Schmidt CM, House MG, Zyromski NJ,
et al.
6. Splenic preserving distal pancreatectomy: does vessel preservation
matter? J Am Coll Surg. 2011;212(4):651-7.
7. Jean-Philippe A, Alexandre J, Laurent C, Collet D, Masson B,
Fernández-Cruz L, et al. Laparoscopic spleen-preserving distal
pancreatectomy: splenic vessel preservation compared with the
Warshaw technique. JAMA Surg. 2013;148(3):246-52.
█ BACKGROUND
This is a familiar scenario to many pancreatic surgeons. In fact, despite the
current advancements in imaging modalities, distant metastatic disease is still
diagnosed at the time of surgical exploration in approximately 12% of cases.
If we consider the patient described in the case vignette, and assume no
evidence of metastatic disease at exploration, we are faced with a potentially
treatable disease – perhaps with curative intent. Indeed, the survival rate for
surgically treated PDAC, with negative microscopic resection margins (R0),
is estimated at about 25% at five years (observed median overall survival
[OS] of 22–24 months) with approximately 4% considered to be long-term
survivors at ten years.
The presence of metastatic disease (stage IV) dramatically reduces the
expected OS of patients with PDAC, which is estimated between 2 and 6
months, without treatment.
Furthermore, despite significant improvements in modern
chemotherapeutic regimens, such as FOLFIRINOX or gemcitabine plus nab-
paclitaxel, the expected OS offered by these regimens, in patients with stage
IV PDAC, remains an unsatisfactory median of 11 and 8.5 months,
respectively.
Once again, let us consider the patient in our case vignette. She is brought
to the operating room with the expectation of having a complete resection of
her PDAC and a prospect of a 25% chance of being alive at five years;
however, with the discovery of the metastatic lesion, she suddenly becomes a
member of a whole different group with a far shorter life expectancy.
It has been established that complete surgical removal of PDAC is the
only chance of a cure, but its role in the setting of stage IV PDAC has not
been well investigated, especially in the era of modern chemotherapy. The
surgeon’s struggle becomes even more apparent when dealing with a locally
resectable PDAC that presents concomitantly with a completely resectable
metastatic lesion, as is illustrated in our case vignette. Therefore,
considering the recent success of resecting metastatic disease from other
solid organ malignancies such as colorectal cancer, sarcoma, and gastric
cancer, it is only natural to query if there is a role for metastatectomy in the
setting of PDAC.
One of the proposed advantages of resecting metastatic disease – if both
the primary lesion and the metastases can be completely resected – relies on
the concept that reducing the tumor burden would allow for a more
efficacious action of systemic chemotherapy; de facto permitting the
chemotherapy and the immune system to concentrate solely on undetected
microscopic disease. However, it remains unknown if the above stated
concept applies to PDAC once the tumor has already metastasized; thus,
extrapolating data from other cancers might not be possible nor advisable.
A downside of performing a major pancreatic resection and a
concomitant liver resection is represented by the potential mortality and
morbidity that can arise by combining both procedures. Although pancreatic
resection mortality has decreased significantly during the last decades – now
estimated around 5% for most centers and < 1% in high-volume pancreatic
centers – the morbidity remains a high 40-60% for all comers. Morbidity
following pancreatic surgery for PDAC often leads to delays in initiation of
systemic chemotherapy, changes in regimen or dosages, or, at its worst,
omission of systemic treatment altogether. In the setting of stage IV PDAC,
the inability to tolerate or the omission of systemic chemotherapy can have
severe and far-reaching consequences commonly resulting in a significant
decrease of life expectancy.
The current national and international guidelines for the treatment of
PDAC do not recommend resection of the primary tumor and synchronous
liver metastases (Tempero et al. 2014). Nonetheless, in the literature there
are isolated reports or case series where such an approach has been
undertaken. However, the available results are inconsistent and often in
conflict with each other, thus leaving an open question as to whether
complete resection of the PDAC with combined resection of liver metastases
will lead to a survival benefit.
Hackert et al. published the largest series on the subject: 128 patients
undergoing PDAC and metastases resection, between 2001 and 2014, of
whom 85 patients had liver oligometastatic disease and the remaining 43
patients had distant aortocaval lymph node metastases. In their study, 86% of
patients with liver metastases underwent a synchronous atypical resection,
while formal hepatic resection was reserved for only 14% of the patients,
and in most cases, it was carried on in a metachronous fashion. Most of the
resected lesions (72%) had a diameter measuring less than 2 cm and the
overall number of metastases in the resected specimen was up-to-three in
96% of the cases. The authors reported a 30-day surgical morbidity and
mortality of 45% and 3%, respectively, after resection of synchronous M1
tumors. These figures changed to 21.7% and 4.3%, respectively, after
resection of metachronous lesions. The median survival for patients
undergoing liver resection (synchronous or metachronous metastatectomy)
was 12.3 months with an estimated 5-year survival of 8%.
Data regarding the use of postoperative chemotherapy were available in
only 74% of patients, of which the majority (79.5%) received gemcitabine-
based regimen. It is worth noting that the 12-month median survival,
described in this study, is certainly superior to the 6- to 7-month median
survival reported with single-agent gemcitabine, yet this result is comparable
to the reported 11-month survival with modern FOLFIRINOX regimen,
therefore adding little benefit to the palliative systemic chemotherapy
approach in terms of overall survival.
Tachezy et al. reported on the experience of six large European pancreas
centers with resection of the primary PDAC and synchronous liver
metastases, between June 1994 and July 2014, focusing on the operative and
oncologic outcomes of such an approach. The authors identified 69 patients
who underwent synchronous pancreatic and liver resection, for PDAC, and
compared their outcomes to a matched cohort of patients, with liver
metastatic PDAC, undergoing surgical exploration without tumor resection.
The results of this study are particularly interesting as the authors reported a
definitive survival advantage for patients undergoing synchronous surgical
resection–when PDAC was localized in the pancreatic head–compared to
patients undergoing surgical exploration without resection (median OS 13.6
vs 7 months, P < .001). Though, in their study, the survival advantage offered
by surgical resection was no longer present when the primary PDAC was in
the body/tail of the pancreas (median OS 14 vs 15 months, P = .312).
However, no definitive explanation for this behavior was evident. In contrast
to the study conducted by Hackert et al. (previously described in this
chapter), the patients described by Tachezy et al. underwent solely non-
anatomic liver resections (atypical wedge resections). Once again, the data
regarding post-operative adjuvant therapy was available only in 83% of the
patients, and of these, 80% received adjuvant therapy mainly in the form of
single-agent gemcitabine (70% and 65% for the resected and nonresected
group, respectively), with a minority being treated with FOLFIRINOX (7%
and 6% for the resected and nonresected group, respectively). Of note, the
authors reported a significant increase in perioperative morbidity in the
resected group compared to the nonresected group (68% vs 48%, P = .025);
however, the reoperation rate and the 30-day mortality were similar between
the groups. Furthermore, 5-year survival was 0% in the nonresected group,
while 4 of 69 evaluable patients (5.8%), after combined resection, were still
alive after 5 years. The results of this study, although encouraging, must be
considered with caution due to the retrospective nature, the limited number of
patients described, the long-time interval spanning over two decades, and the
lack of definitive selection criteria for the allocation of patients in the
resection vs. nonresection group, among others. Nevertheless, the study
shows that synchronous resection, in the setting of metastatic PDAC to the
liver, is safe and feasible and is associated with acceptable morbidity and
mortality.
On the other end of the spectrum, several studies have argued against
surgical resection in patients with stage IV PDAC, suggesting that such an
aggressive approach is associated with minimal to no survival advantage –
compared to palliative systemic treatment alone – at a cost of increased
morbidity and potential mortality.
Takada et al. presented the first study assessing the survival benefit of
surgical resection of stage IV PDAC (with isolated liver metastases)
compared to palliative by-pass surgery. The authors reported on 33 patients,
between the years 1981 and 1995, who were diagnosed with PDAC and
synchronous liver metastases at the time of laparotomy. A total of 11 patients
in this group underwent a pancreaticoduodenectomy combined with
partial/wedge hepatic resection and the remaining 22 patients underwent a
palliative by-pass procedure (n=22). The size of the resected liver metastatic
lesions ranged from 0.5 to 5 cm, and the number of resected lesions per
patient ranged from 1 to 7. In the resected group, the median survival was 6
months and the longest survival was 10 months; in the palliative by-pass
group, the median survival was 4 months and the longest survival was 6
months. Moreover, the author reported a perioperative mortality of 9% in
both groups. Takada et al. observed no improvement in OS between the two
groups, but noted higher surgical morbidity and mortality rates in the group of
patients undergoing pancreatoduodenectomy with synchronous liver
resection. Furthermore, all 11 patients in the resected group died from
multiple recurrent liver metastases within 10 months of the resection. The
high perioperative mortality rate reported in this study appears excessive
compared to modern series, but it is representative of the study period
(1985–1991). Furthermore, the authors do not report on the use of systemic
chemotherapy in either of the two groups, thus adding additional limitations
to the interpretation of this study’s results.
Another example arguing against the benefit of surgical resection in the
setting of stage IV PDAC is offered by Gleisner et al. The authors described
a cohort of 17 PDAC patients on whom synchronous pancreatic resection and
hepatic metastatectomy were performed. The majority of patients, in their
cohort, underwent a nonanatomic liver resection (wedge resection), and most
patients selected for resection had a solitary hepatic metastasis, with a
median size of 0.6 cm. Of the 17 patients with PDAC who underwent
synchronous resection, only 6 received adjuvant chemotherapy. Ultimately,
the authors concluded that the OS in patients undergoing hepatic resection of
synchronous metastasis was not different from the OS of matched patients
who underwent palliative bypass (5.9 vs. 5.6 months, P=0.46).
Understanding the complex biology of PDAC is of the utmost importance,
especially when dealing with systemic disease. There is a concrete risk in
PDAC that visible resectable disease represents only the tip of the iceberg,
and undetected microscopic disease might already be present, in the liver
parenchyma or elsewhere, at the time of surgical resection. Furthermore, the
aggressive biology of PDAC, characterized by short tumor-doubling time
interval and by its well-known metastatic propensity, might make resection of
detectable metastatic disease futile without an effective mechanism to control
the occult microscopic disease and prevent its rapid resurgence.
The adoption of surgical resection, in stage IV PDAC, will require the
establishment of effective patient-selection strategy and the development of
effective tumor markers able to pre dict the tumor response to existent
systemic chemotherapeutic strategy.
Indeed, some patients might benefit from complete resection in stage IV
PDAC; however, the characteristics of this group remain for the most part
unknown and recommendations for routine resection in stage IV PDAC are,
at this stage, unsupported by reliable data.
Nonetheless, a true opportunity exists to evaluate the effect of complete
surgical resection, especially in the setting of more efficacious systemic
treatment as represented by FOLFIRINOX and nab-paclitaxel, a topic that
has not been fully investigated. The results of a clinical trial could
potentially provide further insight to guide patient-selection strategies and
provide valuable data on the efficacy of modern chemotherapeutic agents in
the setting of resected stage IV PDAC.
To conclude our case-vignette: based on the available data in the
literature and in accordance with the current guidelines, surgical resection
should not be performed. Nevertheless, the surgeon should assess the utility
of a palliative by-pass procedure (gastrojejunostomy with or without an
hepaticojejunostomy) to prevent or resolve a gastro/duodenal or choledochal
obstruction caused by a potentially enlarging pancreatic mass. In this
situation, every effort should be made to minimize morbidity and recovery
time, inevitably caused by the laparotomy, to allow for prompt initiation of
systemic therapy within the shortest interval from surgical exploration.
SALIENT POINTS
Only 10% of newly diagnosed PDAC presents with resectable
localized disease, 40-50% presents with locally advanced or
borderline resectable disease, and the remaining 50-60% of cases
presents with distant metastatic disease (ranging from an isolated liver
lesion to widely diffused disease).
Despite the current advancements in imaging modalities, distant
metastatic disease is still diagnosed at the time of surgical exploration
in approximately 12% of cases.
The current national and international guidelines for the treatment of
PDAC do not recommend surgical resection for stage IV disease,
without differentiating single isolated metastasis vs. diffuse disease.
Retrospective studies, which evaluate the role of surgical resection in
the setting of liver oligometastatic PDAC, consist of isolated case
reports or small case series. Their results are often discordant, patient
selection criteria are inconsistent, and the use of systemic
chemotherapy varies and mostly entails single agent gemcitabine.
The role of surgical resection, in the setting of stage IV PDAC with
oligometastatic liver disease, should be further investigated, and a
reliable answer to this controversial topic, in the era of modern
systemic chemotherapy, can only be generated through a randomized
control trial.
SELECTED REFERENCES
1. Gleisner AL, Assumpcao L, Cameron JL, Wolfgang CL, Choti MA,
Herman JM, et al. Is resection of periampullary or pancreatic
adenocarcinoma with synchronous hepatic metastasis justified? Cancer.
2017; 110(11): 2484–92.
2. Hackert T, Niesen W, Hinz U, Tjaden C, Strobel O, Ulrich A, et al.
Radical surgery of oligometastatic pancreatic cancer. EJSO 2017;
43(2): 358-63.
3. Tachezy M, Gebauer F, Janot M, Uhl W, Zerbi A, Montorsi M, et al.
Synchronous resections of hepatic oligometastatic pancreatic cancer:
Disputing a principle in a time of safe pancreatic operations in a
retrospective multicenter analysis. Surgery. 2016; 160(1): 136–44.
4. Takada T. Yasuda H, Amano H, Yoshida M, Uchida T. Simultaneous
hepatic resection with pancreato-duodenectomy for metastatic
pancreatic head carcinoma: does it improve survival?
Hepatogastroenterology. 1997; 44(14): 567–73.
5. Tempero MA, Malafa MP, Berhman SW, Benson AB 3rd, Casper ES,
Chiorean EG, et al. Pancreatic Adenocarcinoma, Version 2.2014:
featured updates to the NCCN guidelines. JNCCN. 2014; 12(8): 1083–
93.
Chu Q, Vollmer C, Zibari G, Orloff S, Williams M, Gimenez M (eds).
Hepato-Pancreato-Biliary and Transplant Surgery: Practical Management of Dilemmas
CASE SCENARIO
A 65-year-old man presents with painless jaundice and is found to have a
locally advanced pancreatic ductal adenocarcinoma (PDAC) in the head of
the pancreas (Figure 1 ). After placement of a metallic biliary stent, the
patient is enrolled in a neoadjuvant protocol, which includes 8 cycles of
FOLFIRINOX (5-FU, oxaliplatin, and irinotecan) followed by 28 treatments
of chemoradiation for a total of 50.4Gy, boosted to 58.8Gy around involved
blood vessels. Following a good radiologic response, the patient is taken to
the OR for a pancreaticoduodenectomy. Intraoperatively, the patient is noted
to have residual tumor involving the superior mesenteric artery.
Intraoperative biopsy confirms adenocarcinoma deeming the tumor
unresectable.
FIGURE 1: Computed axial tomography scan of a patient with
locally advanced pancreatic ductal adenocarcinoma (PDAC) in the
head of the pancreas with near circumferential involvement of the
superior mesenteric artery.
█ BACKGROUND
INTRODUCTION
Pancreatic cancer is a challenging disease to treat and cure. Approximately
80% of patients present with distant, or locally advanced unresectable,
disease. Of those who undergo surgical resection with curative intent, 5-year
overall survival remains low, in the order of 20%. Without treatment,
expected survival is short, ranging between 4 and 6 months. With
combinations of modern systemic agents (such as Gemcitabine/Abraxane or
FOLFIRINOX), with or without radiation therapy, and surgical resection,
median survival rates of up to 44 months have now been reported.
Improvements in multidisciplinary care with delivery of systemic therapies,
better treatment of associated toxicities, and optimization of surgical
technique and postoperative care have accounted for the considerable
improvements in survival following treatment of pancreatic cancer, including
in the palliative setting.
In this chapter, we focus on palliative treatment of locally advanced
unresectable disease. It is important to note that pancreatic cancer may be
considered unresectable because of either distant disease (lungs, liver, or
peritoneum), or locally advanced unresectable disease. The distinction is
potentially important, as implications for palliative therapy may differ based
on local extent of the tumor and anticipated progression of local disease.
In general, practice has shifted away from performing palliative
interventions at the time of discovering unresectable disease during surgical
exploration. Improvements in systemic regimens and interventional
techniques are thought to account for this paradigm shift. However, this is a
generalized observation, since locally advanced disease differs significantly
in its potential to cause biliary and/or gastrointestinal obstruction, compared
with small tumors that metastasize early, where the need for any palliative
intervention may be diminished.
We focus our discussion on a case of locally advanced unresectable
disease of the pancreatic head, where impending biliary and gastric outlet
obstruction may be present or anticipated, as an example of a periampullary
malignancy commonly encountered in the authors’ practice. With continued
improvements in the longevity of patients treated with palliative intent (e.g.
use of FOLFIRINOX in the palliative ACCORD-11 trial), the potential for
obstruction may be further increased. While substantial shrinkage of tumors
may be achieved, radiographic changes are difficult to assess. Therefore,
operative exploration may still be commonly performed. Given the potential
obstructive nature of locally advanced periampullary malignancies, practical
knowledge of surgical palliative techniques remains an important tool in the
armamentarium of the versatile pancreatic surgeon, who should be involved
early in the work up and management of all patients with localized pancreatic
cancer. This chapter addresses biliary bypass, gastric bypass, treatment of
pain, locally ablative procedures, and palliative pancreatectomy in the
management of locally advanced periampullary malignancies.
BILIARY BYPASS
In patients with periampullary cancer presenting with obstructive jaundice,
endobiliary stent placement offers non-operative placement of durable stents
with rapid recovery and low complication rates. Between endoscopic
techniques and percutaneous transhepatic access, cases in which access to
the bile duct is completely precluded are rare. However, endobiliary stent
placement and biliary instrumentation is still associated with complications,
particularly infectious morbidity, and often associated with need for repeated
interventions and hospital readmissions. Metallic biliary stents are now used
routinely in preoperative patients, including for prolonged periods in patients
undergoing neoadjuvant therapy.
Jaundice is the most common presenting symptom of patients who present
with periampullary cancer. Jaundice can lead to debilitating pruritus,
malnutrition, liver dysfunction, and death if not addressed. While endoscopic
management of malignant biliary obstruction is considered the gold-standard
modality in relieving jaundice, a role for surgical biliary bypass still exists
in patients who present with biliary obstruction who undergo surgical
exploration. The exact indications are difficult to define. For example, when
endobiliary stents have been placed preoperatively for jaundice, either with
or without neoadjuvant therapy, the surgeon may still choose to retain the
indwelling endobiliary stent and forego surgical bypass, particularly if a
self-expanding metallic stent is in place. However, the superiority and
durability of surgical bypass compared with endobiliary stenting is
indisputable, as even metallic stents can obstruct. There are no consensus
data by pancreatic surgeons to favor one modality over the other. On the
other hand, when exploration in a jaundiced patient is performed without a
biliary stent or even with a plastic stent, most experienced pancreatic
surgeons would favor a biliary bypass unless precluded for technical
reasons.
In cases where endobiliary stents are technically impossible to place (e.g.
following prior gastric-bypass or gastrectomy, or with duodenal obstruction),
the decision to perform surgical palliative bypass is an easier one, as
percutaneous access, while still reliable, frequently involves an external
drainage component initially and is associated with a higher complication
rate than endoscopic placement. Finally, in some cases, final determination
of resectability cannot be accomplished without dividing the bile duct,
thereby requiring biliary reconstruction.
The authors’ recommendation, based on the extensive experience of the
senior author with this scenario, is that if a jaundiced patient with a stent in
place undergoes exploration via laparotomy and is found to have a locally
unresectable cancer, then surgical palliative bypass is indicated, with stent
removal. If, however, the patient is found to be unresectable from a smaller
tumor that has metastasized within the abdomen, or when disease is detected
during staging laparoscopy, reliance on postoperative endobiliary stenting
may be justified. If in either of these scenarios a plastic stent is in place,
conversion to a metal stent can be performed at a later date. Finally,
regardless of tumor size, performance of a biliary bypass in a patient without
existing jaundice would seem unwarranted.
Multiple options exist for bypass of the biliary tree, which can be
performed via open or minimally-invasive techniques. Although Roux-en-Y
hepaticojejunostomy or choledochojejunostomy are most commonly
performed, other options include gallbladder bypass (cholecystogastrostomy,
cholecystoduodenostomy, or cholecystojejunostomy), as well as alternative
common bile duct or hepatic duct bypass options, such as
choledochoduodenostomy or hepaticoduodenostomy. These options are
discussed below.
CHOLECYSTOENTERIC BYPASS
The gallbladder provides the simplest conduit for enteric bypass of the
biliary tree, particularly given its enlargement in cases of obstructive
jaundice (Courvoisier’s sign). With simple mobilization of the gallbladder
fundus from the cystic plate, routine anastomoses, either handsewn or
stapled, to the stomach or duodenum may be carried out. However, such
procedures are largely of historic interest, due to either proximal extension
of the tumor into the bile duct or involvement of the proximal gastroduodenal
tract with tumor. Cholecystojejunostomy, however, is an equally simple
procedure that avoids these complications, requires no gallbladder
mobilization, and can also be safely performed using hand-sewn or stapled
techniques (Figure 2A-B ). The jejunal loop is brought up either in an
antecolic or retrocolic fashion, and a Roux limb or a Braun
enteroenterostomy is created to divert enteric contents from refluxing into the
gallbladder (Figure 2C-D ).
FIGURE 2: Cholecystojejunostomy performed using hand-sewn or
stapled techniques (Reprinted with permission from Lippincott
Williams & Wilkins and Wolters Kluwer Health. In Lillemoe K,
Jarnagin W (eds). Master Techniques in Surgery-Hepatobiliary and
Pancreatic Surgery. Philadelphia 2013).
PAIN MANAGEMENT
Pain control for locally advanced unresectable pancreatic cancer is an
important cornerstone in the palliative treatment algorithm. Pain is
classically central epigastric with radiation to the back, and gnawing pain
frequently affects patients during sleep at night, when it may be initially most
noticeable. Quality of life may eventually be significantly hindered due to
unremitting severe pain that disrupts activities of daily living. The disruption
to life can be significant, and the efficacy and side effects of narcotic
medications is often problematic. While endoscopic and percutaneous
techniques may be utilized to provide celiac plexus blockade, these
procedures are highly operator-dependent and are limited by access and
visualization.
Although the pendulum has swung away from doing more intraoperatively
for patients found to have locally advanced unresectable disease, Level 1
evidence from a prospective randomized-controlled trial from the Hopkins
group showed that prophylactic intraoperative celiac axis blockade is
indicated, given the benefit attained in preventing the development of pain,
with minimal additional risk or morbidity. Using 20ml of 50% alcohol
injections, the plexus, which lies behind the hepatic and splenic arteries, is
chemically ablated (Figure 6 ). The vessels can be palpated with relative
ease during surgical exploration and without the need for extensive
dissection. With tactile feedback, the ability to perform repeated injections
effectively illustrates the superior efficacy of this simple, quick, and cheap
modality over any endoscopic or percutaneous approach.
FIGURE 6:
SALIENT POINTS
If a jaundiced patient has undergone exploration via laparotomy and
found to have a locally advanced unresectable cancer, then palliative
surgical biliary bypass in the form of a Roux-en-Y
hepaticojejunostomy should be considered.
Given the high incidence of gastric outlet obstruction and the reduced
reliability of endoscopically-placed duodenal stents for long-term
palliation of locally advanced disease, loop gastrojejunostomy (open
or laparoscopic) is indicated at the time of exploration for most
patients with unresectable pancreatic cancers.
Prophylactic surgical celiac axis blockade is beneficial in preventing
the development of pain associated with locally advanced disease with
minimal risk or morbidity.
IORT for locally advanced unresectable disease appears to be
associated with improved survival in centers that possess technical
expertise in use of this locally ablative modality.
SELECTED REFERENCES
1. Lillemoe KD, Cameron JL, Hardacre JM, Sohn TA, Sauter PK,
Coleman J, et al. Is prophylactic gastrojejunostomy indicated for
unresectable periampullary cancer? A prospective randomized trial.
Ann Surg. 1999; 230(3): 322-8.
2. Lillemoe KD, Cameron JL, Kaufman HS, Yeo CJ, Pitt HA, Sauter PK.
Chemical splanchnicectomy in patients with unresectable pancreatic
cancer. A prospective randomized trial. Ann Surg 1993; 217(5): 447-
55.
3. Reid-Lombardo KM, Barton J, Sarr MG. (2013). Operative palliation
of pancreatic cancer. Master Techniques in Surgery: Hepatobiliary and
Pancreatic Surgery. By Lillemoe KD and Jarnagin WR. 1st Edition.
Philadelphia, PA. Lippincott Williams & Wilkins.
4. Bhutani N, Cheadle GA, Bahr MH, Vitale GC. Comparative efficacy of
bilateral thoracoscopic splanchnicectomy for intractable pain secondary
to pancreatic cancer vs. chronic pancreatitis. J Am Coll Surg 2017;
224(4): 566-571.
5. Conrad C, Lillemoe KD(2013). Palliative therapy for pancreatic cancer.
Current Surgical Therapy. By Cameron JL and Cameron AM. 11th
Edition. Philadelphia, PA. Elsevier Saunders.
█ BACKGROUND
The ampullary region is the anatomical area that includes the papilla of Vater,
the confluence of the main pancreatic duct, distal common bile duct, and the
sphincter of Oddi. Although uncommon, the ampullary region can harbor
benign and malignant neoplasms. They represent roughly 5% of all digestive
tract tumors.
The diagnosis of ampullary tumors may be incidental or due to the
investigation of digestive symptoms, such as jaundice, abdominal pain,
bleeding and/or acute pancreatitis episodes. The decision of whether a
patient should be treated or observed must be based on the type of tumor and
the presence of symptoms. Therapeutic options are serial endoscopic
surveillance, local resection (LR), or pancreatoduodenectomy (PD).
Endoscopic surveillance may be indicated for patients with asymptomatic
benign tumors without risk of malignant transformation.
Local resection (LR) can be performed either by a surgical approach
(transduodenal ampullectomy, TDA) or an endoscopic approach (endoscopic
ampullectomy, EA). LR may be considered for patients with benign tumors,
for small tumors that do not harbor carcinoma, and for poor surgical
candidates. LR presents lower complication and mortality rates but can be
associated with a high risk of local recurrence and inadequate resection
margins. EA has the disadvantage of not performing lymphadenectomy, and
its application to tumors at risk of metastases should be carefully considered.
TDA allows lymphadenectomy and node sampling in the peri-pancreatic
region.
PD is the treatment of choice for larger lesions, invasive lesions, and
malignant tumors with risk of lymph node involvement. PD follows
oncological principles, makes possible a radical resection with adequate
lymphadenectomy, and allows for correct staging of the disease. Morbidity of
the PD is around 40% and mortality can reach 6%. Often, these scenarios are
at a significant risk for postoperative pancreatic fistula (POPF), given that
they are often performed in the setting of a soft pancreatic parenchyma and a
normal to small, unobstructed main pancreatic duct. All PD candidates
should be clinically evaluated and fit for major surgery.
DIAGNOSTIC PROCEDURES
Occasionally the diagnosis of ampullary lesions is incidental in
asymptomatic patients, but in most cases it is identified during investigation
of upper gastrointestinal symptoms.
TYPE OF TUMOR
Nonadenomatous, Benign Tumors
Rarely seen in the ampullary area, lipomas, hamartomas, fibromas,
leiomyomas, and adenomyomas are benign tumors whose malignant potential
has not yet been established in the literature. For individuals with clinical
suspicion or pathological diagnosis of benign tumors, indication for surgical
resection will depend on the presence of symptoms.
In asymptomatic patients, close endoscopic observation is recommended.
For patients with symptoms of biliary obstruction, mechanical duodenal
obstruction, or pancreatitis, limited surgery is recommended. Endoscopic
sphincterotomy, EA, or TDA can be indicated in this situation. Whenever
possible the less-invasive procedure should be chosen. Pathological
examination of the surgical specimen should confirm the preoperative
diagnosis. Figure 3 suggests an algorithm for such benign ampullary lesions.
FIGURE 3: Management of Benign Ampullary Lesions EA:
Endoscopic Ampullectomy; TDA: Transduodenal Ampullectomy; PD:
Pancreatoduodenectomy
Adenomas
Similar to colon cancer, ampullary adenocarcinoma originates from an
adenoma-to-carcinoma sequence in a rate of about 30%. Complete resection
of ampullary adenomas is the goal for curative intent.
EA has been indicated for small adenomas (<3.0 cm) confined to
ampullary region, those not invading the muscularis propria, and those with
intraductal invasion < 1.0 cm, but without malignancy or evidence of lymph
node involvement. Recurrence rates range from 0% to 30%. However, a
correct preoperative diagnosis is not accurate. Endoscopy biopsies of
ampullary adenoma show false negative results for malignancy in between
11% and 60%. In contrast, the presence of high-grade dysplasia (HGD) and
ductal dilatation are associated with malignancy in ampullary adenoma. EUS
and MRI are valuable tools to define T and N stages.
Surgical resection (TDA or PD) is indicated when ampullary adenomas
do not fit the criteria for EA. Even in early-stage ampullary adenocarcinoma,
metastatic lymph nodes are present in 40% of cases and disease-free
survival is significantly longer after PD when compared to TDA. PD must be
the first-line treatment of choice for every patient in good clinical condition.
Recently a systematic review comparing surgical resection (TDA or PD)
and EA for ampullary adenoma showed surgical resection to be superior to
EA for complete primary resection, primary success, and recurrence
outcomes. There was no difference regarding complication rates between the
two groups. Figure 4 provides a framework of thought for the management of
ampullary adenomas.
Neuroendocrine Tumors
Ampullary neuroendocrine tumors (NET) are rare. They can be aggressive
and present high rates (35-45%) of lymph node metastasis at diagnosis.
Unfortunately, no clinical or pathological factors can accurately predict the
absence of nodal involvement.
Local excision (EA or TDA) for ampullary NETs is an inadequate
oncological resection because it may understage the disease and is
associated with high recurrence rates. For localized ampullary NETs, PD
should be considered the treatment of choice for patients with good
performance status. Figure 5 depicts an algorithm for managing this entity.
SELECTED REFERENCES
1. Baptiste GG, Postlewait LM, Ethun CG, Le N, Russell MC, Kooby DA,
et al. Is There an Optimal Surgical Approach to Neuroendocrine
Tumors of the Ampulla? A Single Institution Experience Over 15 Years.
Am Surg.2016;82(7):637-43.
2. Dixon E, Vollmer CM Jr, Sahajpal A, Cattral MS, Grant DR, Taylor BR,
et al. Transduodenal resection of peri-ampullary lesions. World J Surg.
2005;29(5):649-52.
3. Mendonça EQ, Bernardo WM, Moura EG, Chaves DM, Kondo A, Pu
LZ, et al. Endoscopic versus surgical treatment of ampullary adenomas:
a systematic review and meta-analysis. Clinics (Sao Paulo).
2016;71(1):28-35.
4. Panzeri F, Crippa S, Castelli P, Aleotti F, Pucci A, Partelli S, et al.
Management of ampullary neoplasms: A tailored approach between
endoscopy and surgery. World J Gastroenterol. 2015; 21(26):7970-87.
5. Song J, Liu H, Li Z, Yang C, Sun Y, Wang C. Long-term prognosis of
surgicaltreatment for early ampullary cancers and implications for local
ampullectomy.BMC Surg. 201;15:32.
█ BACKGROUND
Drains have traditionally been placed at the time of pancreatic resection to
evacuate pancreatic juice, bile, or lymphatic fluid that may accumulate in the
operative bed. They may also serve to herald post-pancreatectomy
hemorrhage. However, drains have been shown to be unnecessary or even
detrimental in other operations, such as splenectomy, gastrectomy, and
colectomy. In a recent global survey of experienced pancreatic surgeons,
14% reported they never leave drains and 27% use drains selectively. This
is mainly due to concern that drains, particularly in patients without
clinically relevant post-operative pancreatic fistula (CR-POPF), may serve
as a portal of entry for bacteria and increase complications. Also, surgeons
reason that an image-guided drain can be placed postoperatively if needed.
Until recently, the literature on this topic was limited and contradictory.
The need for routine placement of drains at the time of pancreatic
resection was first questioned by Jeekel in 1992. Since then, multiple
retrospective cohort studies and three randomized clinical trials have
addressed this question (Table 1 ). All of the cohort studies have shown
either no difference or a decreased overall complication rate with
elimination of routine drainage. However, these studies are retrospective,
small, single-institution cohort studies, which are susceptible to multiple
sources of bias. Changing of institutional practices and evolution of surgical
technique over time are not controlled. Most importantly, these studies lack
randomization, introducing a great deal of bias and potential confounding
differences between the two treatment groups.
TABLE 1: Studies assessing outcome of pancreatectomy
with and without drains
SUGGESTED APPROACH
Returning to our case, the risk of a fistula in this patient is negligible, but the
patient is an octogenarian with multiple co-morbidities, making the risk of
mortality higher in the presence of a CR-POPF. The wise course, in our
opinion, is to leave a drain at the time of resection, monitor the patient
closely in the early postoperative period, and, if the patient clinically
appears well and the drain amylase concentration is low, remove the drain
early. Otherwise, leave the drain in place, as it will mitigate the bad
outcomes associated with a CR-POPF.
SALIENT POINTS
If a patient develops a CR-POPF, absence of a drain placed at the time
of pancreaticoduodenectomy is associated with increased morbidity
and a 4-fold increase in mortality.
Placing a drain at the time of pancreaticoduodenectomy is very unlikely
to ever cause an operative mortality.
In patients who have few risk factors, and have demonstrated in the
early postoperative period by their clinical course and drain amylase
concentration to be at lower risk for CR-POPF, early drain removal
will decrease potential harm caused by routine drainage.
SELECTED REFERENCES
1. McMillan MT, Malleo G, Bassi C, Sprys MH, Vollmer CM Jr. Defining
the practice of pancreatoduodenectomy around the world. HPB
(Oxford). 2015; 17(12): 1145–54.
2. Jeekel J. No abdominal drainage after Whipple’s procedure. Br J Surg.
1992 ;79(2):182.
3. Heslin MJ, Harrison LE, Brooks AD, Hochwald SN, Coit DG, Brennan
MF. Is intra-abdominal drainage necessary after
pancreaticoduodenectomy? J Gastrointest Surg. 1998, 2 (4): 373-8.
4. Conlon KC, Labow D, Leung D, Smith A, Jarnagin W, Coit DG, et al.
Prospective randomized clinical trial of the value of intraperitoneal
drainage after pancreatic resection. Ann Surg. 2001, 234 (4): 487-93.
discussion 493–4.
5. Fisher WE, Hodges SE, Silberfein EJ, Artinyan A, Ahern CH, Jo E, et
al. Pancreatic resection without routine intraperitoneal drainage. HPB
(Oxford). 2011, 13 (7): 503-10.
6. Adham M, Chopin-Laly X, Lepilliez V, Gincul R, Valette PJ, Ponchon T.
Pancreatic resection: drain or no drain? Surgery. 2013; 154 (5): 1069-
77.
7. Correa-Gallego C, Brennan MF, D’Angelica M, Fong Y, Dematteo RP,
Kingham TP, et al. Operative drainage following pancreatic resection:
analysis of 1122 patients resected over 5 years at a single institution.
Ann Surg. 2013, 258 (6): 1051-8.
8. Van Buren G, Bloomston M, Hughes SJ, Winter J, Behrman SW,
Zyromski NJ, et al. A Randomized Prospective Multicenter Trial of
Pancreaticoduodenectomy With and Without Routine Intraperitoneal
Drainage. Ann Surg. 2014, 259 (4): 605-12.
9. Witzigmann H, Diener MK, Kienkötter S, Rossion I, Werner B, Pridöhl
O, et. Al. No Need for Routine Drainage After Pancreatic Head
Resection: The Dual-Center, Randomized, Controlled PANDRA Trial.
Annals of Surgery; 2016 (264)3:528-37.
10. McMillan MT, Vollmer CM Jr, Asbun HJ, Ball CG, Bassi C, Beane
JD, et al. The Characterization and Prediction of ISGPF Grade C
Fistulas Following Pancreatoduodenectomy. J Gastrointest Surg.
2016;20(2):262-76.
11. Bassi C, Molinari E, Malleo G, Crippa S, Butturini G, Salvia R, et al.
Early versus late drain removal after standard pancreatic resections:
results of a prospective randomized trial. Ann Surg. 2010;252(2):207-
14.
█ BACKGROUND
The surgical objective of pancreaticoduodenectomy (PD) has remained
constant since its description by Whipple, Kausch, and others in the first half
of the twentieth century – extirpation of the pancreatic head, distal common
bile duct, and duodenum followed by reestablishment of gastrointestinal
continuity. However, few procedures in the practice of general surgery have
evolved to the same degree as PD – nearly every aspect of the procedure and
its perioperative adjuncts have been scrutinized. It is vital to recall that in the
1970’s and 1980’s PD was nearly abandoned at many centers worldwide.
Then, as now, PD was the only curative option for malignancy of the
pancreatic head. The dismal overall cure rate, even with successful resection
coupled with a persistently high surgical mortality in excess of 20% at that
time, greatly limited the utility and indications for PD. Moreover, major
morbidity and prolonged inpatient hospitalization were considered the norm.
It is one of the great success stories of general surgery over the past thirty
years that PD and pancreatic surgery at large are now performed with
mortality rates of less than 3% and for an ever-growing list of indications,
including removal of pre-cancerous lesions. The transformation has been
multi-factorial. The introduction and relative standardization of more
effective surgical techniques by luminaries such as Cameron, Warshaw,
Blumgart, and Brennan was crucial, as was the establishment of surgical
societies dedicated to the field and the proliferation of high-volume
pancreatic surgeons and centers. Perhaps equally important, advances in both
diagnostic and interventional radiology greatly enhanced the pancreatic
surgeon’s ability to manage diverse causes of postoperative morbidity.
Improvements in anesthesia, nursing, and critical care have been important
contributors as well.
Operative drainage has been a constant component of PD since its outset.
Historically, drainage modalities have been multiple and varied. Biliary T-
tubes, pancreatic duct drains, and feeding gastrostomy or jejunostomy have
all been described as routine components of PD in the past; all are now
rarely employed. Currently, operative drains following PD are generally
understood to be one or two intraperitoneal drains placed at the time of
surgery in the region of the pancreatic and/or biliary anastomoses. Such
drains remain in regular employ, and descriptions of their placement,
management, and criteria for removal are included in most modern surgical
texts and atlases. The utility of such drains and the necessity of their usage
following PD is a topic of great controversy amongst pancreatic surgeons.
Morbidity following PD can mirror standard complications observed
following any major abdominal surgery, such as surgical site infections or
ileus. In addition, complications that are specific to pancreatic surgery that
occur after PD include delayed gastric emptying, postoperative pancreatic
fistula (POPF), and post-pancreatectomy hemorrhage. Of these, delayed
gastric emptying is troublesome but rarely results in prolonged incapacity.
Hemorrhage, particularly late extra luminal hemorrhage, although quite rare,
can be deadly to the patient if not recognized and managed expeditiously.
Conversely, POPF is a frequent occurrence and is defined by international
consensus as amylase concentration from any measurable volume of
intraabdominal fluid greater than three times serum amylase concentration on
or after postoperative day three. POPF, as shown in Table 1 , are classified
by severity as grade A (no clinical impact, biochemical diagnosis only),
grade B (requiring readmission or specific treatment such as percutaneous
drainage or antibiotics), and grade C (requiring reoperation or associated
with sepsis, organ dysfunction, or death). The profound improvements in
operative mortality and the widening indications for pancreatic resection
have resulted in an explosion of studies exploring means to improve surgical
morbidity following PD. Accordingly, a vast array of interventions have been
examined, attempting to mitigate the occurrence and severity of POPF,
including changes in surgical technique, pharmaceuticals such as
somatostatin analogues, adhesives, stents, and others. Remarkably, very few
of these myriad efforts have resulted in widespread practice changes due to
cost concerns, ineffectiveness, or feasibility. Notwithstanding this lack of
consensus, compared to a few short decades ago when the merits of
pancreatic resection were openly questioned, it is encouraging to see modern
clinical research questions focus on such granular details as the utility of
operative drains and their impact on outcomes.
SUGGESTED STRATEGY
It is our practice to employ operative drains in a highly selected fashion
following PD. While we currently place drains in less than 5% of cases, we
do not advocate total abandonment of surgical drainage in appropriate
patients as determined by sound surgical judgement. In general,
contamination of the surgical field or concern for anastomotic fidelity are the
most common indications for drainage in our practice. Specific instances
could include:
1. Presence of or concern for persistent leakage from the biliary
reconstruction.
2. Gross contamination of the surgical field, either from pre-existing
abscess or uncontrolled spillage of a large volume of infected bile or
succus from hollow viscera (i.e. during concomitant bowel resection).
3. Abnormally difficult or tenuous pancreatic anastomosis.
The third indication above is rarely applied in our practice and is not
universally utilized in the case of a pancreas thought to be high risk of POPF
formation due to a soft gland texture or small pancreatic duct if the
anastomosis is otherwise sound. In the case scenario above, in our practice
we would consider placement of a drain in the region of biliary anastomosis
if doubt persisted regarding its integrity. Should the operative field no longer
be soiled from spillage of stent-associated sludge, this would not necessarily
mandate drain placement, nor would the soft pancreas with a small duct.
The operative technique for reconstruction during PD we employ is as
follows: the pancreatic duct is transected sharply and
pancreaticojejunostomy is carried out in a two-layer end-to-side fashion with
duct-to-mucosa reconstruction. Single-layer hepaticojejunostomy in the
Blumgart fashion is similarly performed in a running fashion unless the duct
is small or tissue quality is poor. Following the re-establishment of biliary
continuity the area is copiously irrigated. Once the field is clean, an unsoiled
laparotomy sponge is firmly applied circumferentially to the biliary
anastomosis. Any evidence of biliary leakage, easily visible as greenish
discharge on the white laparotomy sponge, results in reinforcement or
revision of the anastomosis until its integrity is ensured. If drains are
employed, they are placed in a directed fashion (i.e. – if being placed for
biliary concerns, a second drain need not be placed adjacent to the
pancreatic reconstruction). During recovery, diet is re-instituted as soon as
patient tolerance allows (usually, liquids on day two and solids on day three
or four). Once oral intake has been established, the drains are examined (and
amylase concentration assessed, if the drain is peri-pancreatic) and removed
if no leakage is evident. For patients without drains, a high index of
suspicion is mandatory for POPF development, especially if a high-risk
gland exists (soft texture/small duct).
Slow return of dietary tolerance, unexplained fevers (even if intermittent),
leukocytosis, and vague abdominal discomfort should be closely monitored
with a low threshold for further investigation.
It is our belief that the majority of patients can undergo PD without
placement of operative drains. We recently retrospectively examined 343
patients that have undergone PD at Yale-New Haven Hospital over a ten-year
period by a single surgeon (RRS). Operative drains were employed in 23
(6.7%) for the reasons described above. Overall mortality was 1.2%, major
morbidity occurred in 16%, and clinically-relevant class B or C POPF
occurred in 9.6%. The sum of the above data suggests that doing so is safe
and may benefit the patient via reduced length of stay and morbidity. Should a
clinically-relevant POPF develop, we have not seen degradation in the
outcomes or limitations in the management of those patients without
operative drains. For surgeons who currently routinely employ operative
drains and are considering omitting them to gain the above potential benefits,
we encourage them to proceed judiciously. Complete familiarity with the
postoperative course is needed and it is imperative to remember that
placement of a surgical drain is necessary using sound surgical judgement in
certain patients. Finally, institutional expertise in the form of excellent house
staff, nursing, and interventional radiology is mandatory to allow efficient
diagnosis and treatment when POPF or other surgical complication ensues.
As additional data accrues and management techniques are refined, the
appropriate usage of operative drains following PD will continue to evolve,
as will the long course towards optimizing surgical outcomes all patients
undergoing pancreatic surgery.
SALIENT POINTS
Operative drainage has historically been an invariable component of
performing pancreaticoduodenectomy but this practice has been
questioned as routine drainage after most elective abdominal surgery
has been abandoned.
The first prospective trial and numerous large retrospective reports
from high-volume institutions suggest that forgoing routine placement of
drains decreases length of stay and the incidence of infectious
complications, delayed gastric emptying, and postoperative pancreatic
fistula formation (although perhaps not clinically-relevant grade B and
C fistulae). These benefits seem to be maximized when compared to
groups with operative drains left in place for a prolonged period of
time.
A single, recent prospective trial suggested that
pancreaticoduodenectomy without routine drain placement increased
operative mortality. However, this trial may not be generalizable to the
actual practice patterns of surgeons who do not routinely employ drains
and efforts are ongoing to objectively stratify risks of placing or
omitting drain placement.
The decreased morbidity associated with forgoing routine drain usage
can be safely achieved if the surgeon assertively investigates any signs
or symptoms of early morbidity and the institutional expertise is
available to assist with management (especially interventional
radiology).
Sound surgical judgement should never be supplanted by inflexible
doctrine regarding drain placement following
pancreaticoduodenectomy.
SELECTED REFERENCES
1. Kunstman JW, Kuo E, Fonseca AL, Salem RR. Evaluation of a recently
described risk classification scheme for pancreatic fistulae
development after pancreaticoduodenectomy without routine post-
operative drainage. HPB (Oxford). 2014; 16(11):987-93.
2. Conlon KC, Labow D, Leung D,Smith, Jarnagin W, Coit DG, et al.
Prospective randomized clinical trial of the value of intraperitoneal
drainage after pancreatic resection. Ann Surg. 2001;234(4):487-93;
discussion 93-4.
3. Mehta VV, Fisher SB, Maithel SK, Sarmiento JM, Staley CA, Kooby
DA. Is it time to abandon routine operative drain use? A single
institution assessment of 709 consecutive pancreaticoduodenectomies. J
Am Coll Surg. 2013;216(4):635-42; discussion 42-4.
4. Correa-Gallego C, Brennan MF, D’Angelica M, Fong Y, Dematteo RP,
Kingham TP, et al. Operative drainage following pancreatic resection:
analysis of 1122 patients resected over 5 years at a single institution.
Ann Surg. 2013; 258(6):1051-8.
5. Van Buren G, 2nd, Bloomston M, Hughes SJ, Winter J, Behrman SW,
Zyomski, et al. A randomized prospective multicenter trial of
pancreaticoduodenectomy with and without routine intraperitoneal
drainage. Ann Surg. 2014;259(4):605-12.
█ BACKGROUND
Over the past several decades the mortality following
pancreaticoduodenectomy has decreased dramatically and is now reported to
be less than 5% at high volume centers. The overall morbidity is still
reported to be as high as 60%, however. A leak from the pancreatic
anastomosis and subsequent pancreatic fistula remains a major cause of
morbidity following pancreaticoduodenectomy, and can present late in the
postoperative course as sepsis or as hemorrhage, often from the much-feared
complication of gastroduodenal (GDA) artery stump blowout. The incidence
of pancreatic fistula is still reported to be between 9 and 29%, with an
associated mortality of around 10%.
With the overarching aim of improving management and decreasing
morbidity, there has been an effort over the past decade to standardize the
definition and classification of pancreatic fistulae. In 2005 an international
study group comprised of 37 pancreatic surgeons from five continents met
with the aim of standardizing the much-debated definition of pancreatic
fistula. They generally defined a post-pancreaticoduodenectomy pancreatic
fistula as either a failure of healing of a pancreatic-enteric anastomosis or as
a parenchymal leak not directly related to the anastomosis, such as one
originating from the raw pancreatic surface. They suggested a diagnosis
based on any measurable drain output on postoperative day three or onwards
with amylase content greater than three times the upper limit of normal. In
addition, they established a clinical grading system that classified pancreatic
fistula as grade A, B or C based on the clinical severity and impact on
patient outcomes (Table 1 ). In brief, grade A was considered a transient
fistula with no clinical impact and minimal deviation from the expected
postoperative course. Grade B fistula mandates a change in management,
such as disallowing oral intake or treatment with parenteral nutrition, and
often leads to a delayed discharge. Grade C fistulae are defined by clinical
decline or instability, a major delay in discharge, and major deviation from
the expected clinical course including necessity of additional invasive
procedures and reoperation.
TABLE 1: Diagnosis and classification of postoperative
pancreatic fistulas following pancreatic surgery by the
International Study Group on Pancreatic Fistulas (Adapted
with permission from Elsevier in Bassi C, et al. The 2016
update of the International Study Group (ISGPS) definition
and grading of postoperative pancreatic fistula: 11 years
after. Surgery 2017;161(3):584-591).
SALIENT POINTS
Pancreatic fistula remains a significant complication following
pancreaticoduodenectomy, and development of a grade C fistula is a
major source of morbidity and mortality.
The traditional operative management of grade C pancreatic fistula
was completion pancreatectomy, a difficult procedure that commits that
patient to lifelong total exocrine and endocrine insufficiency.
Alternative management strategies include wide drainage, with or
without pancreatic duct occlusion, and conversion of the
pancreaticojejunostomy to a pancreatogastrostomy for restoration of
enteric continuity.
Our preferred approach is the bridge stent technique that avoids
excessive dissection but successfully diverts pancreatic secretion
while allowing for the formation of a ‘neo-anastomosis’.
SELECTED REFERENCES
1. Bassi C, Marchegiani G, Dervenis C, Sarr M, Hilal M, Adham M, et al.
The 2016 update of the International Study Group (ISGPS) definition
and grading of postoperative pancreatic fistula: 11 years after. Surgery
2017;161(3):584-591.
2. Balzano G, Pecorelli N, Piemonti L, Ariotti R, Carvello M, Nano R, et
al. Relaparotomy for a pancreatic fistula after a
pancreaticoduodenectomy: a comparison of different surgical strategies.
HPB (Oxford). 2014;16(1):40-5.
3. Fuks D, Piessen G, Huet E, Tavernier M, Zerbib P, Michot F, et al. Life-
threatening postoperative pancreatic fistula (grade C) after
pancreaticoduodenectomy: incidence, prognosis, and risk factors. Am J
Surg. 2009;197(6):702-9.
4. Callery MP, Pratt WB, Kent TS, Chaikof EL, Vollmer CM, Jr. A
prospectively validated clinical risk score accurately predicts
pancreatic fistula after pancreatoduodenectomy. J Am Coll Surg.
2013;216(1):1-14.
5. Bachellier P, Oussoultzoglou E, Rosso E, Scurtu R, Lucescu I, Oshita A,
et al. Pancreatogastrostomy as a salvage procedure to treat severe
postoperative pancreatic fistula after pancreatoduodenectomy. Arch
Surg. 2008;143(10):966-70; discussion 71.
6. Kent TS, Callery MP, Vollmer CM, Jr. The bridge stent technique for
salvage of pancreaticojejunal anastomotic dehiscence. HPB (Oxford).
2010;12(8):577-82.
█ BACKGROUND
This scenario should be recognized by all experienced pancreatic surgeons
as an ominous finding. Though the patient is stable and the event is tempting
to ignore, it may represent a “sentinel bleed”—bleeding that is relatively
minor and stops without any intervention yet precedes massive hemorrhage.
That this has occurred in delayed fashion relative to the index operation and
in the context of a pancreatic fistula is particularly worrisome.
Multiple algorithms have been developed for the evaluation and treatment
of post-pancreatectomy bleeding. In this chapter, the diagnostic and
therapeutic options used in the evaluation and treatment of post-
pancreatectomy hemorrhage are reviewed. After summarizing the
management options, an algorithm tailored toward delayed post-
pancreatectomy hemorrhage is presented.
DEFINITIONS
For years, the frequency, morbidity, and mortality of post-pancreatectomy
bleeding were difficult to quantify. That is because the term “post-
pancreatectomy bleeding/hemorrhage” was applied heterogeneously, and
standardized definitions were lacking. As a result, reported experiences
were difficult to compare. In 2007, the International Study Group of
Pancreatic Surgery (ISGPS) defined post-pancreatectomy hemorrhage (PPH)
based on a collective review of the literature. Post-pancreatectomy
hemorrhage is now differentiated based on the time of onset, the location and
cause, and the severity (Table 1 ). The time of onset is either early (≤ 24
hours after the end of the index operation) or late (≥ 24 hours after the end of
the index operation). The location is either intraluminal or extraluminal.
Severity is mild (small or medium volume blood loss managed
conservatively) or severe (larger volume blood loss with clinical
deterioration and the need for significant transfusion and/or invasive
treatment). To establish a clinical classification, three different grades of
post-pancreatectomy hemorrhage (grades A, B, C) were defined according to
these parameters and based on the clinical consequence to the patient. Grade
A bleeding is early, mild bleeding without much clinical significance. Grade
B bleeding is early severe or late mild bleeding that may require procedural
interrogation. Grade C bleeding is late, severe hemorrhage of major clinical
impact with potentially life-threatening ramifications. Grade C bleeds
invariably require intervention.
TABLE 1: ISGPS classification of post-pancreatectomy
hemorrhage (Reprinted with permission from Elsevier in
Wente M, Veit J, et al. Postpancreatectomy hemorrhage
(PPH)-An International Study Group of Pancreatic Surgery
(ISGPS) definition. Surgery 2007;142:20-5).
INTERVENTIONAL RADIOLOGY
Surgical exploration has traditionally been the principle therapeutic venture
in post-pancreatectomy hemorrhage. However, a clear shift has occurred
toward the use of interventional radiology in the diagnosis and treatment of
post-pancreatectomy bleeds. Re-exploration in this group of patients is
fraught with exceedingly high rates of morbidity and mortality. These patients
are often deconditioned, and dense peri-pancreatic adhesions are generated;
visualizing the site of bleeding requires manipulation of tenuous
anastomoses. Catheter-based techniques facilitate the treatment of arterial
injuries without hazardous dissection.
Post-pancreatectomy bleeds can originate from a variety of arterial
sources, including the hepatic arteries, gastroduodenal artery,
superior/inferior pancreaticoduodenal artery, splenic artery, and superior
mesenteric artery. In any endovascular evaluation, both the celiac and
superior mesenteric artery distributions must be interrogated, as multiple
arterial erosions can occur. The most common affected artery across series is
the gastroduodenal. Therapeutic options for treating this and other
pseudoaneurysms include coil embolization or exclusion of the bleeding site
with an endovascular stent. In a systematic review of delayed post-
pancreatectomy hemorrhage, the success of diagnostic angiography in
localizing a bleeding source was 88%, and the success rates of endovascular
measures in achieving hemostasis are essentially equivalent to surgical re-
exploration in modern series.
The interventional approach can fail for multiple reasons. First and
foremost is institutional availability. In many cases, the role for
interventional radiology depends upon patient stability and the time from
referral until the patient is on the table. Assuming this is no obstacle, other
issues can still arise. One is failure to identify the site of bleeding. This
occurs frequently, as bleeding may stop by the time of interrogation, or
originate from as a vascular injury that is too small to identify. Secondly, a
site of origin may be identified but have no treatment option. Since the
dominant regional arteries, including the SMA and hepatic, are skeletonized
in the course of dissection, and branches including the gastroduodenal and
inferior pancreaticoduodenal are often ligated near their origin, there is
concern about the ability to treat vascular issues without compromising
perfusion of the liver and intestine. Multiple complications associated with
endovascular treatment have been described, including re-bleeding, stent
stenosis/thrombosis, hepatic abscess, and bowel/liver infarction. Regardless,
a formal arteriogram that identifies but fails to treat a vascular erosion is still
of value as it can help to direct surgical re-exploration.
A matter of controversy in catheter-based treatment of these bleeds is
empiric embolization. As noted, bleeding in this scenario can originate from
numerous sites. This can create apprehension and skepticism toward empiric
embolization if no one site is suggested. However, empiric embolization
clearly has a role. Intraoperative findings or postoperative studies can
frequently direct attention to a likely bleeding source without absolute
evidence (Figure 1 ). Yekebas et al. reported an 80% success rate for
“blind” embolization based on surgeons’ clinical suspicion. My strategy is to
empirically embolize vessels that can be treated safely in patients with a
clear suggestion of arterial hemorrhage. For instance, empiric embolization
should be strongly considered in the setting of a late severe bleed with
concomitant pancreatic fistula, as the likelihood of an arterial erosion is
quite high. The best candidate for safe empiric embolization is usually the
gastroduodenal artery. This can sometimes be ligated a fair distance (>5mm)
from the hepatic artery, which allows for safe coil embolization. If no such
stump is present, empiric treatment may only be achievable using covered
metal stents. I have opted against empiric treatment with stents unless there is
a very high index of suspicion for a particular site, since stent thrombosis can
have disastrous consequences. If the patient is stable and there is not a high
index of suspicion for arterial disruption, I do not favor empiric embolization
of even safely-treatable vessels, since the results of a complication can be
devastating.
FIGURE 1: Index of suspicion in endovascular assessment. This
patient developed a sentinel bleed several days after a Whipple
procedure in the setting of a concomitant pancreatic fistula.
Although his CT angiogram was negative for a blush or a
pseudoaneurysm, the drain was found to essentially abut the
gastroduodenal artery creating suspicion that the GDA was the
source of the bleed (Figure 1A). The patient ultimately underwent a
formal arteriogram that demonstrated a GDA pseudoaneurysm
(Figure 1B). The drain is seen coursing adjacent to the GDA on the
arteriogram as well.
The first decision for the surgeon is whether a major arterial disruption is
strongly suggested. Either a small volume of bright red blood, or any amount
of venous-appearing blood not accompanied by hemodynamic change, is
ominous but does not clearly suggest arterial erosion. This may, however,
represent a sentinel bleed. Since the patient is hemodynamically stable, and
the problem is not clearly arterial in nature, I would favor a CT angiogram
with delayed images (multiphase CT). This is a good precursor to a formal
arteriogram. If a blush or pseudoaneurysm is noted, the event was clearly a
sentinel bleed and the vessel should be treated interventionally or surgically
with IR strongly favored. If no target is noted (no blush and no
pseudoaneurysm), I would transfer the patient to an intensive care setting and
monitor closely. One of two scenarios is most likely. One is that the bleeding
was not a true sentinel bleed and no major bleeding will occur over the next
few days. The second is a more significant bleeding event some time within
the next 72 hours, which more clearly evidences an arterial etiology.
If bright red blood is noted that is significant in volume and/or
accompanied by hemodynamic changes, a major arterial erosion is very
likely. In this event, I would favor proceeding straight to a formal
arteriogram in interventional radiology if the patient’s hemodynamics allow
for this. While stable patients should certainly be sent to IR, I favor sending
even a somewhat unstable patient as long as the hemodynamics can be
supported with transfusion. Avoidance of the CT scan gives the surgeon the
best chance for identifying a target that can be treated without delay. Ideally,
a target vessel can be identified and treated safely with endovascular coiling
or stenting. If an arterial problem can be identified but not treated, surgery is
the only option. A negative arteriogram (no active extravasation or
pseudoaneurysm) in this setting is not at all reassuring because the likelihood
of an arterial origin remains high. The expectation would be for another
potentially life-threatening event within the next few days. If the arteriogram
is negative but the gastroduodenal artery can be embolized safely,
embolization should be strongly considered. An argument can also be made
for leaving the endovascular sheath in for the next 24 hours in order to
facilitate the next arteriogram. Lastly, I favor obtaining a 3 phase CT in this
setting to determine whether the axial image strongly suggests a target for
empiric embolization (Figure 1 ). A second advantage of the CT is the
ability to visualize a complete anastomotic failure, in which case completion
pancreatectomy becomes a stronger consideration.
It is worth noting that endoscopy is minimized in this algorithm. The
algorithm is tailored toward an intra-abdominal bleed, and sentinel bleeding
that occurs in delayed fashion alongside a concomitant pancreatic leak is far
more likely to be solved in surgical or endovascular fashion. This is true
whether the bleeding is noted intraluminally or extraluminally. Though
endoscopy does have a diagnostic role, this patient is expected to require an
endovascular or surgical approach. I favor proceeding down the
interventional pathway rather than dedicating hours to what is likely to be a
diagnostic procedure at best.
Surgical re-exploration continues to have a pivotal role in the
management of this scenario. However, my preference is to avoid surgery in
the setting of delayed bleeds whenever possible. If the patient’s
hemodynamics do not allow for an arteriogram or there is no treatable target
in IR, surgery is often the only viable strategy. If the site of bleeding cannot at
all be compartmentalized from an ongoing pancreatic leak, the site may
rebleed. In a stable patient that is likely to tolerate completion
pancreatectomy, this should be considered. If the patient remains unstable or
postoperative inflammatory changes are severe, completion pancreatectomy
is fraught with morbidity and wide drainage is favored. Any regional
omental/peritoneal tissue that can be used to protect the ligated vessel from
the leak should be mobilized to accomplish this.
SALIENT POINTS
Massive delayed hemorrhage is often preceded by a sentinel bleed.
The most critical aspect of the sentinel bleed is the recognition that any
bleeding that occurs remote from surgery in a stable patient that has not
previously manifested ongoing bleeding mandates an urgent evaluation
for arterial erosion.
Anastomotic leak and intra-abdominal abscess are risk factors for the
development of arterial erosions and delayed hemorrhage. Sentinel
bleed in the setting of an anastomotic leak is especially ominous and
associated with high rates of mortality.
Although utilized frequently, the therapeutic efficacy of endoscopy in
the setting of post-pancreatectomy hemorrhage and especially late PPH
is low. Endoscopy continues to have a prominent diagnostic role.
Management of delayed post-pancreatectomy hemorrhage continues to
shift from re-exploration to interventional radiology techniques which
are associated with high success rates in some series. Despite this,
reoperation remains a central part of the algorithm.
SELECTED REFERENCES
1. Wente MN, Veit JA, Bassi C, Dervenis C, Fingerhut A, Gouma DJ, et al.
Postpancreatectomy hemorrhage (PPH): an International Study Group of
Pancreatic Surgery (ISGPS) definition. Surgery. 2007;142(1):20-5.
2. Yekebas EF, Wolfram L, Cataldegirmen G, Habermann CR, Bogoevski
D, Koenig AM, et al. Postpancreatectomy hemorrhage: diagnosis and
treatment: an analysis in 1669 consecutive pancreatic resections. Annals
of surgery. 2007;246(2):269-80.
3. Roulin D, Cerantola Y, Demartines N, Schafer M. Systematic review of
delayed postoperative hemorrhage after pancreatic resection. J
Gastrointestinal Surg. 2011; 15(6):1055-62.
█ BACKGROUND
A pancreaticoduodenectomy (Whipple procedure) is typically performed to
remove neoplasms involving the head of the pancreas, ampulla, distal
common bile duct, or duodenum while restoring continuity of the biliary-
pancreatic system. Significant improvements in outcomes for this complex
procedure have been achieved since it was first performed in 1935. For
decades after its introduction, mortality hovered at ~25%, while today the
perioperative mortality rate in high-volume centers in the United States is
less than 3%. Despite the improvement in mortality, morbidity remains high,
with the most common postoperative complications being pancreatic
leak/fistula (40%), delayed gastric emptying (18%), wound infection (7%),
and intra-abdominal abscess (6%).
Postpancreatectomy hemorrhage occurs in 1-8% of all pancreatic
resections; however, it accounts for a disproportionate share (11-38%) of
overall mortality. This chapter will focus on the definition, clinical
presentation, and management of postpancreatectomy hemorrhage.
DEFINITION OF POSTPANCREATECTOMY
HEMORRHAGE (PPH)
In 2007, the International Study Group of Pancreatic Surgery (ISGPS)
developed an objective definition of PPH based on literature review and
consensus clinical experience. Since 2007, this definition has gained
widespread acceptance and has been utilized in numerous studies aimed at
guiding appropriate management of PPH. The ISGPS defines PPH based on
three objective parameters: onset, location, and severity.
Onset is defined as either early (<24 hours following completion of the
index operation) or late (>24 hours).
The location of postpancreatectomy hemorrhage is defined as either
intraluminal (gastrointestinal) or extraluminal (intra-abdominal).
The severity of bleeding may be either mild or severe based on volume
of blood loss, degree of clinical impairment, and need for intervention.
PPH is divided into three grades:
Grade A -- early mild hemorrhage
Grade B -- early serious hemorrhage or delayed mild hemorrhage
Grade C – delayed life-threatening hemorrhage
Early PPH is virtually always secondary to a technical error, with failure
to achieve appropriate hemostasis during the index operation. In contrast,
late PPH typically occurs from complications of the operation, with a usual
delay of several days to even weeks postoperatively. There are numerous
causes of late bleeding. Erosion of a peripancreatic vessel secondary to a
postoperative pancreatic fistula associated with intra-abdominal infection is
a common cause. Pancreatic fistula occurs in ~15% of patients after
resection of the pancreatic head and in ~25% of patients after distal
pancreatectomy. The greatest risk factor for fistula development is a normal
pancreas, with soft parenchyma and a small pancreatic duct. More than half
of the mortality after pancreatectomy can be attributed directly to pancreatic
fistula. Other causes include anastomotic ulceration or bleeding in
association with an arterial pseudoaneurysm that has developed as a
consequence of vascular trauma at the time of surgery or chemical injury
(pancreatic enzymes). Leakage of pancreatic enzymes at the
pancreaticojejunal anastomosis can lead to enzymatic digestion of the blood
vessel wall by trypsin, elastase, and other exocrine pancreatic enzymes.
As mentioned above, the location of PPH is defined as either intraluminal
(gastrointestinal) or extraluminal (intra-abdominal). Intraluminal bleeding
can occur from a variety of sources and includes the common sources of
gastrointestinal bleeding seen in the inpatient population, such as peptic ulcer
disease and esophagitis. However, in the post pancreatectomy population,
bleeding from suture lines of the anastomoses (eg, duodenojejunostomy after
pylorus-preserving pancreaticoduodenectomy, gastrojejunostomy after
classical pancreaticoduodenectomy, pancreaticojejunal or hepaticojejunal
anastomoses) must also be considered as possible sources of intraluminal
blood loss. Most commonly, late intraluminal hemorrhage occurs from the
pancreatic anastomosis or marginal ulcers. However, a pseudoaneurysm with
erosion into one of the anastomoses can also present as intraluminal
hemorrhage and must always be considered.
Extraluminal (intra-abdominal) hemorrhage occurs secondary to
involvement of larger vessels such as the hepatic artery, the stump of the
gastroduodenal artery, inferior pancreaticoduodenal artery, or the superior
mesenteric artery. Bleeding secondary to a postoperative pseudoaneurysm is
another important cause of late intra-abdominal hemorrhage, and a high index
of suspicion must be maintained for its presence in the setting of PPH, as the
risk of mortality is high. Patients with a pseudoaneurysm will often present
with a herald bleed, and it is critical to consider this diagnosis and
investigate on initial presentation. The clinical presentation and management
of PPH secondary to a pseudoaneurysm will be discussed below.
MANAGEMENT OF PPH
Sentinel Bleed
All postoperative bleeding must be taken seriously and evaluated completely
until the etiology has definitively been identified. This is of particular
importance in the setting of PPH.
Even bleeding which is initially felt to be insignificant and self-limited can
be a harbinger of impending, life-threatening hemorrhage. Specifically, a
sentinel bleed is a well described postoperative occurrence in which a
seemingly minor bleeding episode serves as a warning sign of an impending
major hemorrhagic event. Sentinel bleeding has an overall prevalence of
33% and has been associated with higher relaparotomy and mortality rates
compared to patients without a sentinel bleed episode. Sentinel bleeding can
occur days after the index operation, and therefore, bleeding at any time in
the postoperative period warrants a great degree of vigilance and attention.
This includes even a small amount of blood that appears suddenly in the
surgical drains.
In cases of sentinel bleeding, CT angiography is recommended to make an
early diagnosis of a pseudoaneurysm (Figure 1 ). CT angiography is
preferred as the initial diagnostic modality as hemorrhage from ruptured
pseudoaneurysms cannot be completely ruled out based on negative
angiographic findings, and intermittent bleeding of ruptured pseudoaneurysms
is difficult to detect by angiography alone and may be missed. If a bleeding
site is identified, the patient should undergo emergent angiography to provide
endovascular treatment of the pseudoaneurysm. The mainstay of angiographic
treatment is coil embolization; however, in the setting of a short GDA stump,
placement of a covered stent in the common hepatic artery is often required
to maintain hepatic arterial patency.
Intraluminal PPH
When an intraluminal source of bleeding is suspected, endoscopy should be
performed. Historically, there has been concern regarding the performance of
endoscopy in the immediate postoperative period secondary to mechanical
forces from torque and air insufflation leading to anastomotic dehiscence. It
is now generally accepted that endoscopic intervention is safe, even in the
immediate postoperative period. When possible, carbon dioxide (CO2),
which is rapidly absorbed by the GI mucosa, should be used as the
insufflation agent for endoscopic procedures. CO2 is absorbed from the GI
tract approximately 160 times faster than nitrogen, the major gaseous
ingredient of ambient air. It is passively absorbed through the mucosal lining
into the bloodstream and eventually exhaled through the lungs. If a bleeding
source is identified, standard endoluminal hemostatic therapies (ie, thermal
and mechanical techniques) are safe in the postoperative period.
Extraluminal PPH
Blood in peritoneal drains is frequently an ominous sign and always warrants
performing an urgent CT angiogram to evaluate the source of bleeding. The
extraluminal sites of bleeding postpancreaticoduodenectomy are numerous
and include the GDA stump, portal vein tributaries, hepatic artery branches,
superior mesenteric vein tributaries (including uncinate vessels), branches of
the superior mesenteric artery (including jejunal mesenteric arterial branches
to the left and inferior pancreaticoduodenal artery to the right), the cut
pancreatic surface, and the gallbladder fossa. It is optimal for the surgical
team to be present in the angiography suite when these procedures are
performed to ensure that there is close communication between the surgeon
and interventional radiologist performing the procedure regarding all of the
potential sources of bleeding and pseudoaneurysm. For example, while
gastroduodenal or common hepatic artery pseudoaneurysms are most
common, a pseudoaneurysm of the inferior pancreaticoduodenal artery can
also occur and careful pull-back angiogram to evaluate the origin of this
vessel from the proximal SMA is critical to avoid missing this potential
bleeding source.
As described above, the most common etiology of extraluminal
hemorrhage in the early (<24 hours) postpancreatectomy population is the
result of technical failure to properly secure the inferior pancreaticoduodenal
artery, or less commonly the gastroduodenal artery stump. Many patients with
early extraluminal bleeding present with acute hemodynamic instability with
associated blood in the operative drains. Historically, these patients have
required emergent reoperation for hemostasis, and this remains the most
widely accepted approach to the management of early PPH.
In contrast to early PPH, a strategy of endoscopic and angiographic
intervention has been utilized in the setting of delayed PPH, with urgent
angiography as the mainstay of treatment. As previously discussed, late PPH
typically results from complications of the operation such as intra-abdominal
abscess, erosion of a peripancreatic vessel secondary to a pancreatic fistula,
ulceration of the site of an anastomosis, or in association with an arterial
pseudoaneurysm. Late PPH frequently presents after hospital discharge,
leading to delays in recognition and treatment, and therefore significant
associated mortality.
SUGGESTED APPROACH
Returning to the case, this patient underwent emergent CT angiography,
which revealed a large, contrast-filled cavity around the
pancreaticojejunostomy. A ruptured pseudoaneurysm from the GDA stump
was suspected. Conventional angiogram of the celiac trunk revealed a
pseudoaneurysm at the GDA stump with active extravasation. Transarterial
coil embolization was successfully performed (Figure 2 ).
SALIENT POINTS
All PPH must be evaluated urgently and completely until a source is
identified.
Grade B early PPH (<24 hours) is usually best treated by return to the
operating room.
Endoscopy and endoscopic interventions for intraluminal
gastrointestinal sources of bleeding are safe in the early postoperative
period.
Providers must maintain a high level of vigilance for extraluminal
bleeding presenting as luminal blood loss.
The surgical team must remain actively involved in the resuscitation
and management process, including being present in the Interventional
Radiology suite as angiography is taking place.
SELECTED REFERENCES
1. Leichtle SW, Kaoutzanis C, Mouawad NJ, Welch KB, Lampman R,
Hoshal VL Jr, et al. Classic Whipple versus pylorus-preserving
pancreaticoduodenectomy in the ACS NSQIP. J Surg Res
2013;18(1)3:170-6.
2. Cameron JL, Riall TS, Coleman J, Belcher KA. One thousand
consecutive pancreaticoduodenectomies. Ann Surg 2006;244(1):10-5.
3. Wente MN, Veit JA, Bassi C, Dervenis C, Fingerhut A, Gouma DJ, et al.
Postpancreatectomy hemorrhage (PPH): an International Study Group of
Pancreatic Surgery (ISGPS) definition. Surgery 2007;142(1):20-5.
4. Fogel EL, Shahda S, Sandrasegaran K, DeWitt J, Easler JJ, Agarwal, et
al. A Multidisciplinary Approach to Pancreas Cancer in 2016: A
Review. Am J Gastroenterol 2017;112(4):537-54.
5. Khalsa BS, Imagawa DK, Chen JI, Dermirjian AN, Yim DB, Findeiss
LK. Evolution in the Treatment of Delayed Postpancreatectomy
Hemorrhage: Surgery to Interventional Radiology. Pancreas
2015;44(6):953-8.
6. Committee AT, Lo SK, Fujii-Lau LL, Enestvedt BK, Hwang JH, Konda
V, et al. The use of carbon dioxide in gastrointestinal endoscopy.
Gastrointest Endosc. 2016;83(5):857-65.
7. Evans DB, Amini A, Christians KK. Chapter 26. Postpancreatectomy
Hemorrhage: Early and Late. In: Pawlik TM, Maithel SK, Merchant NB
(Eds.) Gastrointestinal Surgery. 271-280. New York, New York,
Springer New York; 2015
Acknowledgments: The authors would like to thank Dr. Sanjeeva Kalva,
Chief and Professor, Interventional Radiology, UT Southwestern Medical
Center, for the CT angiography and angiographic images provided as
figures in this chapter.
█ BACKGROUND
Chylous ascites (CA) is a rare complication following abdominal surgery,
with a frequency of only one in 20,464 abdominal operations. The main
cause of postoperative CA is due to injury to the cisterna chyli or its
tributaries. CA has been suggested to be more common following extensive
retroperitoneal dissection for malignant tumors. During pancreatic surgery
the proximity of the pancreas to the cisterna chyli indicates that CA could be
more frequently seen in this type of surgery. However, CA after pancreatic
surgery has been reported in relatively few cases and its incidence and
natural history is poorly known.
Recently, several series of CA in pancreatic surgery have described more
accurately the incidence, diagnosis, and management of this condition. CL
occur after pancreatic surgery about 1.8-13% of the time, between different
series, and a recent increase in the frequency of this complication has been
observed due to more advanced stages of pancreatic cancer that are
intervened upon, possibly with more extensive lymph node resections.
Although their incidence may be low, chyle leaks can be associated with
significant morbidity including dehydration, wound complications,
debilitation, malnutrition, immunosuppression, and delayed delivery of
adjuvant chemotherapy—thus affecting the long-term survival of these
patients.
Distinct definitions have been used to classify CA. It has been described
as the secretion of milky white fluid with high triglyceride content with a
secretion range from above 110 to 130 mg/dL. Another controversy on
defining this complication is the quantity of fluid drained necessary to be
classified as CA; because of this controversy, it is sometimes described
sometimes as a chyle leak (CL). Alternatively, it can be considered chylous
ascites when the volume drained is above 100 to 600 ml and there is diffuse
chyloperitoneum. The International Study Group of Pancreatic Surgery
(ISGPS) defined chyle leak as the output of milky-colored fluid from a drain,
drain site, or wound, on or after postoperative day 3, with a triglyceride
content ≥ 110 mg/dL. This contemporary definition is recommended to
standardize the report of complications after pancreatic surgery between
different centers.
CA is usually recognized until 5th or 6th postoperative day and generally
increases when oral intake is initiated. The majority of patients present with
abdominal distention and discharge of white milky fluid through either the
drain, former drain site, or surgical wound.
Many studies have analyzed factors associated with the formation of CL
after surgery; those found to significantly increase the incidence of CA are
early restoration of enteral feeding, number of lymph nodes harvested,
vascular resection and reconstruction, manipulation of the para-aortic area,
and retroperitoneal tumor invasion.
The type of surgery performed has also been related with the frequency of
CL. During pancreatoduodenectomy, the proximity of the head of the
pancreas to the cisterna chyli promotes the development of a postoperative
CL, although recently some studies have reported a comparable rate of CL
following distal pancreatectomy. The indication of surgery has also been
associated with CL with the majority of the cases relating to malignancy.
This is due to invasion of adjacent structures of malignant tumors, and the
requirement of more aggressive intervention to perform complete resections.
SALIENT POINTS
Chyle leak, as defined by IGSPS, is the output of milky-colored fluid
from a drain, drain site, or wound, on or after postoperative day 3,
with a triglyceride content ≥110 mg/dL.
Chyle leak, including chylous ascites, is a clinically-relevant
complication associated with longer hospital stay and complications
such as malnutrition, inmunosuppression, and generalized debilitation.
The initial treatment approach in these patients should be dietary
restrictions, as these are successful in the majority of cases.
Total parenteral nutrition should be considered when dietary
restrictions fail to reduce the degree of the leak.
Percutaneous catheter drainage of collections of chylous ascites is
advised only in the case of clinical symptoms.
Lymphangiography alone or with embolic agents, peritoneovenous
shunt, and surgical ligation should only be performed in cases of chyle
leak refractory to conservative management.
SELECTED REFERENCES
1. Kuboki S, Shimizu H, Yoshdome H, Ohtsuka M, Kato A, Yoshitomi H,
et al. Chylous ascites after hepatopancreatobiliary surgery. Br J Surg.
2013; 100(4): 522-7.
2. Besselink MG, Van Rijssen LB, Bassi C, Dervenis C, Montorsi M,
Adham M, et al. Definition and classification of chyle leak after
pancreatic operation: A consensus statement by the International Study
Group on Pancreatic Surgery. Surgery 2017;161(2):365-372.
3. Tabchouri N, Frampas E, Marques F, Blanchard C, Jirka A, Regenet N.
Chylous ascites management after pancreatic surgery. World J Surg.
2017; 41(4):1054-1060.
4. Kawasaki R, Sugimoto K, Fujii M, Miyamoto N, Okada T, Yamaguchi
M, et al. Therapeutic effectiveness of diagnostic lymphangiography for
refractory postoperative chylothorax and chylous ascites: correlation
with radiologic findings and preceding medical treatment. AJR Am J
Roentgenol. 2013;201(3):659-66.
5. Yarmohammadi H, Brody LA, Erinjeri JP, Covey Am, Boas FE, Ziv E,
et al. Therapeutic Application of Percutaneous Peritoneovenous
(Denver) Shunt in Treating Chylous Ascites in Cancer Patients. J Vasc
Interv Radiol. 2016;27(5):665–673.
█ BACKGROUND
POSTOPERATIVE ANASTOMOTIC
FAILURE FOLLOWING
PANCREATODUODENECTOMY
Reconstruction of the foregut during pancreatoduodenectomy (PD) involves
creation of three intestinal anastomoses: pancreatico-jejunostomy or -
gastrostomy, hepaticojejunostomy, and gastro- or duodenojejunostomy. Each
of these anastomoses carries the potential for dehiscence or leak. The suture
line at the blind end of the bilopancreatic limb might also dehisce, increasing
diagnostic and therapeutic difficulty. Leak from any of these sites can be
innocuous and self-limited, or they can be difficult to control, resulting in
clinically relevant morbidity and increased risk of mortality. In patients with
malignancy, anastomotic leak can also result in potential delay to receipt of
adjuvant chemotherapy.
The pancreaticojejunostomy (PJ) is by far the most likely of the three
anastomoses to fail. The technical components of the PJ, the risk factors
associated with clinically-relevant post-operative failure, and the
management strategies for pancreaticoenterostomy failure have been subject
to the most intense study. By comparison, the incidence of a leak or fistula
that does not involve the PJ is much lower. For instance, isolated
hepaticojejunostomy (HJ) leaks are reportedly between 3-8%. The etiology
and best practices concerning the management of leak at the HJ and other
sites of bilious fistula have been less well studied. There is no substantive
body of empiric evidence to guide management of these clinical problems.
As a result, we are left to extrapolate from the management principles of
other types of gastrointestinal fistulas or to rely on personal experience with
bile leak to make treatment decisions. In this chapter, we outline our
approach to the management of patients with early post-PD bile leak as
manifest by bilious output from the surgical drains on postoperative day two.
Figure 1 summarizes our clinical decision-making pathway.
FIGURE 1: Flow diagram summarizing management of early
hepaticojejunostomy leak after pancreatoduodenectomy. (POD:
Postoperative Day; HJ: Hepaticojejunostomy; POPF: Postoperative
Pancreatic Fistula)
CLINICAL DECISION-MAKING IN THE
EARLY UNCONTROLLED BILE LEAK
For postoperative gastrointestinal leak in general, the clinical condition of
the patient dictates early decision-making. This is true for patients following
simple operations, such as enterolysis for adhesive small bowel disease, and
it is true for early biliary leak or fistula following PD.
In rare instances, the presence of enteric organisms in the peritoneal
cavity can result in hemodynamic instability that places the patient at
immediate risk for multisystem organ failure and death. In these cases, the
enteric leak is uncontrolled by definition. The most prudent approach is re-
exploration with the intent to more definitively control the site of the leak.
Control can be achieved upon re-exploration with either immediate repair of
the anastomosis or surgical placement of better drainage. If the patient is in
septic shock as a result of an anastomotic leak two days following PD, urgent
re-exploration is the most appropriate next step in management. In these
circumstances, septic shock will be manifest as progressive hypotension
without another identifiable cause (e.g., no evidence of hemorrhage, acute
heart failure, or pulmonary embolus). Re-exploration will effectively identify
the site of the leak and affords the safest and most efficient method for
obtaining control.
What is done during the re-exploration depends on the location of the
leak. If the leak is from the PJ, it is almost always technically impossible to
improve or revise the prior anastomosis. Instead, repositioning the surgical
drains, or providing additional drainage of the site, is the only viable
management strategy. If the leak is at the HJ, the method of management
depends on the physical appearance of the proximal common hepatic or
common bile duct. If there is a healthy-appearing margin of bile duct above
the site of the dehiscence, the bilioenteric anastomosis could be redone. In
these circumstances, our practice is to place a T-tube in the duct above the
revision and run a short limb of the tube across the revised bilioenteric
anastomosis. A closed suction drain is then placed behind the revised
anastomosis. The T-tube is left to gravity drainage. When the clinical sepsis
has resolved, a cholangiography is performed through the T-tube. If there is
no evidence of contrast extravasation, the T-tube is capped, and the closed
suction drain removed. The T-tube is then routinely left in place for 6 weeks
regardless of drain output to act as a stent for the revised anastomosis. If, at
the re-exploration, there is no healthy-appearing margin of bile duct above
the site of the dehiscence, then the best approach is to reposition the surgical
drains to better control the site of the leak. The intent here is to control the
leak and to pursue a more definitive revision of the anastomosis after the
sepsis has resolved. In these cases, percutaneous placement of transhepatic
catheters may be employed as a second step in effort to avoid interval
surgical revision of the HJ weeks later. Most other sites of early leak that
manifest as hemodynamic instability (e.g., leak from the gastrojejunostomy,
the blind end of the biliopancreatic limb, or a missed enterotomy) may be
effectively revised at the time of the re-exploration.
Overall, an operative intervention is more likely to save the patient with
hemodynamic compromise associated with an early uncontrolled anastomotic
leak after PD than an image-guided percutaneous drainage or an unduly
prolonged resuscitation in the ICU.
SALIENT POINTS
Isolated hepaticojejunostomy (HJ) leaks from pancreatoduodenectomy
(PD) are rarer than leaks from the pancreaticojejunostomy (PJ), with
published rates ranging from 3-8%.
HJ leaks are diagnosed by examining the character of the surgical drain
effluent, measuring the amylase and bilirubin levels from the surgical
drain effluent, and confirming with imaging studies.
Patients with early HJ leaks and clinical instability should be re-
explored in the operating room with the intent to either revise the HJ or
to ensure appropriate drainage of the leak with drains (i.e., obtaining
control).
Use of percutaneous transhepatic cholangiocatheters is our preferred
management strategy for clinically stable patients with HJ leaks.
Inability to place or failure to improve with percutaneous transhepatic
cholangiocatheters typically requires re-exploration using a T-tube.
Adequate nutritional support and appropriate use of antibiotics is
essential to promote healing of the dehiscence.
SELECTED REFERENCES
1. Jester AL, Chung CW, Becerra DC, Molly Kilbane E, House MG,
Zyromski NJ, et al. The Impact of Hepaticojejunostomy Leaks After
Pancreatoduodenectomy: a Devastating Source of Morbidity and
Mortality. J Gastrointest Surg. 2017; 21(6):1017-1024.
2. Malgras B, Duron S, Gaujoux S, Dokmak S, Aussilhou B, Rebours V, et
al. Early biliary complications following pancreaticoduodenectomy:
prevalence and risk factors. HPB (Oxford). 2016;18(4):367-74.
3. Andrianello S, Marchegiani G, Malleo G, Pollini T, Bonamini D,
Salvia R, et al. Biliary fistula after pancreaticoduodenectomy: data from
1618 consecutive pancreaticoduodenectomies. HPB (Oxford).
2017;19(3):264-269.
4. Antolovic D, Koch M, Galindo L, Wolff S, Music E, Kienle P, et al.
Hepaticojejunostomy--analysis of risk factors for postoperative bile
leaks and surgical complications. J Gastrointesl Surg. 2007;11(5):555-
61.
5. Burkhart RA, Relles D, Pineda DM, Gabale S, Sauter PK, Rosato EL, et
al. Defining treatment and outcomes of hepaticojejunostomy failure
following pancreaticoduodenectomy. J Gastrointest Surg.
2013;17(3):451-60.
█ BACKGROUND
Distal pancreatectomy is most commonly performed for benign and malignant
disease of the pancreatic tail, body and in some cases, neck. This operation
can be performed open or with a minimally invasive approach, and can be
done with or without removal of the spleen, but is most commonly combined
with a splenectomy. We prefer the minimally invasive approach, and this is
most often done with a concurrent splenectomy after division of the splenic
artery and splenic vein. However, removal of the spleen is not without risk.
Patients who have undergone splenectomy are at increased risk of infectious
and thromboembolic events, some of which can be devastating.
Overwhelming Post Splenectomy Infection (OPSI) results in septicemia,
meningitis, and high rates of mortality. Therefore vaccination of these
patients is extremely important. The decision of when to administer these
vaccines, however, is less certain.
SPLENIC ANATOMY AND PHYSIOLOGY
The spleen lies in the posterior lateral portion left upper quadrant, protected
by the 9th, 10th , and 11th ribs. It is in close proximity to the stomach,
diaphragm, left kidney, and the aptly named “splenic flexure” of the colon.
Most importantly, for the pancreas surgeon, the tail of the pancreas lays in the
hilum of the spleen and is intimately associated with the splenic artery and
vein.
The spleen acts, in many ways, like a filter, to remove old or senescent
red blood cells, clear active red blood cells of nuclear remnants, and the
destruction of platelets or cells that have been coated with antibody. Most
importantly, the spleen is the largest lymphoid organ in the body and the B-
cell populations in the spleen are the main site of production of IgM
antibodies, which are integral to the opsonization and removal of
encapsulated organisms. It is also the main site for synthesis of opsonizing
molecules: opsonin, tuftsin and properdin. Removal of the spleen, therefore,
is not without consequence.
COMPLICATIONS OF SPLENECTOMY
Asplenic patients are more susceptible to infections related to encapsulated
bacteria. Of these, the most devastating is OPSI (Overwhelming Post
Splenectomy Infection) syndrome, which is defined as fulminating sepsis,
meningitis, or pneumonia. The risk of development of OPSI in asplenic
patients is 3.6%, which is 50 times higher than the general population, and is
associated with an overall mortality of 1.4-1.8%. However, these risks are
misleading for the pancreatic surgeon, since the majority of high risk patients
are those with genetic or hereditary disorders, infants, or young children. The
true risk of morbidity and mortality from splenectomy in healthy adults is less
than one percent.
The majority of post-splenectomy complications occur within the first
two years of resection, although there are reports of fulminating infection
occurring more than 20 years after splenectomy. No clear evidence exists
examining the time interval to OPSI for patients who have undergone
splenectomy for malignancy. However, the relative immunocompromised
state associated with such an operation, as well as the use of adjuvant
therapy, can be assumed to increase the risk in these patients.
The organisms implicated in OPSI are S. pneumonia, N. meningitides, and
H. influenza type b. Of these S. pneumoniae is the most common causative
organism, accounting from 50-90% of cases, followed by H. influenza type b
and N. meningitides. Other less frequently seen organisms include
Escherichia coli, Pseudomonas aeruginosa, Capnocytophaga canimorsus,
Enterococcus sp, Bacteroides sp, and Bartonella sp. Symptoms start
insidiously with a mild, influenza like presentation (Table 1 ). Patients report
fever, malaise, myalgias, headache, vomiting, diarrhea and abdominal pain.
This prodrome is quickly followed by septic shock leading to multi-system
organ failure and death. The mortality rate of OPSI ranges from 38-70% even
with treatment. Early institution of empiric antibiotics, followed by tailored
treatment based on culture results, is recommended.
TABLE 1: Signs and symptoms of Overwhelming Post-
Splenectomy Infection
GUIDELINES
Guidelines for prevention and treatment of infection in asplenic and
hyposplenic patients were first prepared by the British Committee for
Standards in Hematology in 1996 and subsequently reviewed in 2002. These
guidelines include three parts: education, vaccination, and antibiotic
prophylaxis. Operative asplenia patients should be aware of the signs and
symptoms of infection and notify their physician of any febrile illness. This
education is especially important for the operating surgeon, as the risk of
post-splenectomy complications continues beyond the follow up period, and
so patients must be taught to be their own advocate. Patients will require
repeat vaccinations and boosters. Travel precautions are also important due
to the risk of bacterial and parasitic infections, and patients must understand
the need for preventative measures prior to travel. There is clear evidence
demonstrating that properly educated patients have the lowest incidence of
post-splenectomy infection. We ascribe to the Advisory Committee for
Immunization Practices (ACIP) recommendations for immunization in
asplenic patients.
VACCINATIONS
Pneumococcus
Currently there are two different pneumococcal vaccines for S. pneumoniae:
the polysaccharide vaccine with 23 different serotypes (PPSV-23) and a 13
valent conjugate vaccine (PSV-13). The PPSV-23 reduces the occurrence of
OPSI in splenectomized patients but only covers for 90% of pneumococcal
infections. PSV-13 protects against fewer organisms, but there is evidence
that it has enhanced immunogenicity in adult patients. In our practice,
otherwise healthy patients receive the PSV13 vaccine 14 days prior to
surgery, followed by PPSV23 at least 8 weeks after PSV13 administration.
The optimal time for pre-operative administration of PPSV-23 is 14 days,
and this is the timing we perform for all our patients undergoing elective
distal pancreatectomy with splenectomy. In patients undergoing urgent or
emergent distal pancreatectomy with splenectomy, we delay administration
of vaccines until at least 14 days after resection, or until once we are
comfortable that they will mount an appropriate immunogenic response to the
vaccines. Vaccinations administered earlier that 14 days post-operatively
have been shows to be less effective. The PPSV-23 vaccination should be
repeated every five years, depending on the age and medical condition of the
patient as the protective effect of the pneumococcal vaccination declines five
years after vaccination.
H. influenza Type b
There are three monovalent Hib conjugate vaccines and three combination
Hib vaccines available. All are equally effective. For our patients, we
follow the ACIP guidelines and give the Hib vaccine as a single
administration 14 days prior to surgery to any patient who has not been
previously vaccinated. While it is true that patients who have been
previously vaccinated and completed the booster schedule are considered
immune and do not need re-vaccination, it is often difficult to document this
pre-operatively, and so if this cannot be adequately documented we give the
vaccine.
Meningococcal Immunization
There are three types of quadrivalent meningococcal vaccines; one
quadrivalent polysaccharide vaccine (MPSV4), three quadrivalent conjugate
vaccines, and serogroup B meningococcal vaccines. The MPSV4 protects
against strains A, C, Y, and W135 capsular polysaccharides. This vaccine is
preferred for adults older than 55 years who require a single dose and who
have not received a previous dose of meningococcal conjugate vaccine. The
majority of patients in the US, however, have already received the MPSV4
vaccine. For these patients, use of a quadrivalent conjugate vaccine
(MenACWY) is recommended, and this is our practice. MenACWY-CRM
and MenACWY-D are the two quadrivalent conjugate vaccines available in
the US. The MenACWY-CRM can be administered concomitantly with
PCV13 and, therefore, this is the vaccine we prefer. MenACWY-D must be
administered at least 4 weeks after PCV13 due to interactions leading to
decreased immune response to PCV13. In addition, there are now serogroup
B meningococcal vaccines (MenB-FHbp or MenB-4C) and patients
undergoing splenectomy should receive this vaccine as well.
The meningococcal immunizations are administered at least 14 days prior
to surgery. Patients should be re-vaccinated every five years.
The adverse effects of any of these vaccinations are minimal and should not
be used as a reason not to vaccinate patients. Local reactions include redness
or swelling and systemic reactions include arthralgia, fatigue, fever,
headache, nausea, myalgia and rash. These local and systemic reactions are
self-limiting and do not have severe sequelae.
COMPLIANCE
However, compliance with vaccination has been a factor in the development
of OPSI. The compliance rate for pneumococcal vaccination has been
reported as low as 42%, and vaccination against meningococcal and H.
influenza type b was even lower. In elective patients undergoing distal
pancreatectomy with splenectomy, pre-operative administration of vaccines
is especially effective. Management of patients undergoing distal
pancreatectomy encompasses the pre-operative decision making for
splenectomy. Indications for splenectomy should be reviewed and individual
patients screened for risks associated with splenectomy. A proposed
‘guideline’ for vaccination in patients with distal pancreatectomy can be seen
in Figure 2 .
SALIENT POINTS
Overwhelming post splenectomy infection syndrome results in
significant morbidity and high mortality in a small but real percentage
of patients undergoing distal pancreatectomy with splenectomy.
In certain patients, spleen preserving distal pancreatectomy can be safe
and effective and avoid risks associated with splenectomy.
We advocate for distal pancreatectomy with splenectomy in cases of
malignancy or in those with benign disease with concurrent splenic
vein thrombosis.
Vaccinations for S. pneumonia, N. meningitides, and H. influenza type
b should be administered for patients undergoing distal pancreatectomy
with splenectomy during pre-operative planning at least 14 days prior
to the operation.
SELECTED REFERENCES
1. Parikh PY, Lillemoe KD. Surgical management of pancreatic cancer--
distal pancreatectomy. Semin Oncol 2015; 42(1):110-22.
2. Bisharat N, Omari H, Lavi I, Raz R. Risk of infection and death among
post-splenectomy patients. J Infect 2001; 43(3):182-6.
3. Theilacker C, Ludewig K, Serr A, Schimpf J, Held J, Bögelein M, et al.
Overwhelming Postsplenectomy Infection: A Prospective Multicenter
Cohort Study. Clin Infect Dis 2016; 62(7):871-8.
4. Vaccines & Immunizations. Centers for Disease Control and Prevention
website. https://www.cdc.gov/vaccines/index.html. Updated April 21,
2017. Accessed June 14, 2017.
5. Shatz DV, Schinsky MF, Pais LB, Romero-Steiner S, Kirton OC,
Carlone GM. Immune responses of splenectomized trauma patients to
the 23-valent pneumococcal polysaccharide vaccine at 1 versus 7
versus 14 days after splenectomy. J Trauma 1998; 44(5):760-5.
Chu Q, Vollmer C, Zibari G, Orloff S, Williams M, Gimenez M (eds).
Hepato-Pancreato-Biliary and Transplant Surgery: Practical Management of Dilemmas
CASE SCENARIO
A 68-year-old female with a history of mild congestive heart failure and
obesity, undergoes pancreatoduodenectomy for pancreatic adenocarcinoma.
Her postoperative course is complicated by a pancreatic fistula requiring
prolonged maintenance of a surgical drain. She receives routine prophylactic
subcutaneous heparin perioperatively and during her inpatient recovery
without bleeding event. Should anticoagulation be continued for a prolonged
time-frame following discharge?
█ BACKGROUND
INTRODUCTION
Increasing scrutiny of perioperative morbidity and mortality in recent years
has been paralleled by the recognition that the traditional 30-day
perioperative window is too narrow to capture all surgery-related adverse
events. Indeed, the perioperative period has been redefined as extending 90
days from surgery. For hepatopancreatobiliary (HPB) operations, this
redefinition provides a more meaningful estimate of perioperative events
such as pancreatic fistula, bile leak, and liver insufficiency that take time to
unfold and may lead to morbidity and mortality late in the postoperative
course. Delayed presentation of complications, however, is not limited to
HPB-specific adverse events. Venous thromboembolism (VTE), while
traditionally attributed to intraoperative venous stasis, is increasingly
diagnosed beyond discharge. Scrutiny of larger institutional and nationwide
datasets suggest that a substantial minority of such events are diagnosed after
discharge from the hospital. These data and identification of at-risk
populations have driven consensus recommendations for prolonged VTE
prophylaxis from oncologic and surgical societies. Given the devastating
impact of pulmonary embolism and the risk of VTE in pancreas cancer
patients, adoption of prolonged prophylaxis protocols is gaining acceptance,
but has been slow. This may reflect concern regarding prophylaxis-related
complications, particularly bleeding.
IS PROLONGED PROPHYLAXIS
ENOUGH?
In most cases, the benefit of perioperative and prolonged VTE prophylaxis
justify the acceptably low risk of associated complication. Important to note,
however, is that these measures do not provide complete protection against
thromboembolic events. Areas for further study include additional
preoperative scrutiny of especially high-risk patients with duplex
sonography. In our own practice, class III obesity (BMI 40+), immobility,
and prolonged preoperative hospitalization are indications for preoperative
screening. The presence of preoperative deep vein thrombosis may guide
postoperative anticoagulation or selective preoperative placement of an
inferior vena cava (IVC) filter. In the postoperative setting, there may be a
role for the use of screening duplex ultrasonography as well. The
development of a DVT while on prophylactic anticoagulation may be an
indication for more intensive follow-up or for therapeutic dose
anticoagulation. Balancing risk and benefit remains the overarching
challenge.
SALIENT POINTS
Pancreatic cancer is one of the most thrombogenic malignancies.
Most patients who undergo pancreatectomy for pancreatic cancer are at
high-risk for VTE according to validated risk assessment models.
Numerous professional society guidelines recommend at least 28 days
of post-operative VTE prophylaxis for high-risk surgical oncology
patients.
Evidence consistently shows that extended VTE prophylaxis following
surgery for cancer reduces the incidence of asymptomatic DVT. Meta-
analysis supports extended prophylaxis for reduction of all VTE
events.
Extended VTE prophylaxis is safe following pancreatoduodenectomy.
SELECTED REFERENCES
1. Chew HK, Wun T, Harvey D, Zhou H, White RH. Incidence of venous
thromboembolism and its effect on survival among patients with
common cancers. Arch Intern Med. 2006;166(4):458-64.
2. Heit JA, O’Fallon WM, Petterson TM, Lohse CM, Silverstein MD,
Mohr DN, et al. Relative impact of risk factors for deep vein
thrombosis and pulmonary embolism: a population-based study. Arch
Intern Med. 2002;162(11):1245-8.
3. Fagarasanu A, Alotaibi GS, Hrimiuc R, Lee AYY, Wu C. Role of
Extended Thromboprophylaxis After Abdominal and Pelvic Surgery in
Cancer Patients: A Systematic Review and Meta-Analysis. Ann Surg
Oncol. 2016; 23(5)1422-30.
4. Agnelli G, Bolis G, Capussotti L, Scarpa RM, Tonellu F, Bonizzoni E,
et al. A clinical outcome-based prospective study on venous
thromboembolism after cancer surgery: the @RISTOS project. Ann
Surg. 2006;243(1):89-95.
5. Gould MK, Garcia DA, Wren SM, Karanicolas PJ, Arcelus JI, Heit JA,
et al. Prevention of VTE in nonorthopedic surgical patients.
Antithrombotic Therapy and Prevention of Thrombosis, 9th ed:
American College of Chest Physicians evidence-based clinical practice
guidelines. Chest. 2012;141(2 Suppl):e2275-2775.
6. Caprini J. Thrombosis risk assessement as a guide to quality patient
care. Dis Mon 2005;51:70-8.
7. Lyman GH, Bohlke K, Khorana AA, et al. Venous thromboembolism
prophylaxis and treatment in patients with cancer: American Society of
Clinical Oncology clinical practice guideline update 2014. J Clin
Oncol. 2015;33(6):654-656. doi:10.1200/JCO.2014.59.7351.
8. Streiff B, Holmstrom B, Ashrani A, et al. Cancer-Associated Venous
Thromboembolic Disease, Version 1.2015. J Natl Compr Canc Netw.
2015;13(9):1079.
9. Mandalà M, Falanga A, Roila F. Management of venous
thromboembolism (VTE) in cancer patients: ESMO Clinical Practice
Guidelines. Ann Oncol. 2011;22 Suppl 6(Supplement 6):vi85-92.
doi:10.1093/annonc/mdr392.
█ BACKGROUND
In our current healthcare climate, where cost and accountability is
increasingly scrutinized, hospital readmission rates have emerged as an
increasingly popular quality metric. With initiatives such as the National
Surgical Quality Improvement Program (NSQIP), and more public non-profit
organizations such as ProPublika’s surgeon’s report card, these hospital- and
surgeon-specific metrics are increasingly accessible to the public. This in
turn dangerously creates a dilemma for the surgeon when having to make
clinical decisions in the best interest of the patient versus acting due to
pressure from insurance companies and reimbursement policies. In patients
undergoing pancreaticoduodenectomy, 1 in 6 will experience a readmission
within 30-days, posing a significant management issue. In this chapter, we
discuss the causes and predictors of readmission after
pancreaticoduodenectomy, as well as our proposed management strategy for
the patient at high-risk of a readmission. What Are The Causes Of
Readmission After Pancreaticoduodenectomy?
In 2008, the Medicare Payment Advisory Commission reported that the cost
of preventable readmissions was $12 billion a year. However, such
preventable readmissions were in relation to patients with medical
conditions such as congestive heart failure and pneumonia, and it is important
to point out that generalization to the surgical patient is not accurate. In
patients undergoing pancreaticoduodenectomy, the majority of readmissions
occur as a result of procedure-specific complications rather than a lack of
care coordination. In an analysis of 1,173 consecutive patients undergoing
pancreaticoduodenectomy at our institution, the three most common causes of
readmission were intraabdominal abscess, pancreatic fistula, and wound
infection, accounting for 46% of all readmissions. Only 3.5% of patients
were readmitted without an identifiable medical cause, consistent with the
low rates (3-14%) of unrevealing readmission workups reported by other
institutions. As such, surgeons should tread carefully in their efforts to reduce
their readmission rates given that majority of patients requiring readmission
are a vulnerable cohort experiencing clinically significant complications that
require rescue interventions.
WHAT ARE THE PREDICTORS OF
READMISSION AFTER
PANCREATICODUODENECTOMY?
In an attempt to reduce readmissions after pancreaticoduodenectomy, similar
efforts have gone into predicting which patients are at risk of being
readmitted after discharge. In our institutional analysis, predictors of
readmission include patients requiring multivisceral resection (OR 4.0,
p=0.02), intensive care unit admission (OR 2.9, p<0.001), and patients with
pancreatic fistula (OR 1.9, p=0.004). Interestingly, surgeon volume was not a
significant predictor of readmission. In a population-based analysis,
Yermilov et al. demonstrated that increasing age, patient comorbidities, and
tumor T stage were independent predictors of readmission post-PD.
Similarly, in a SEER-Medicare analysis, Hyder et al. demonstrated that the
largest share of variation in readmission is attributable to patient-related
factors such as age and comorbidities (95.4%), as opposed to physician-
(0.3%) and hospital- (4.3%) related factors. This provides further evidence
that readmission after pancreaticoduodenectomy is largely a result of the
complexity of the disease process and surgical procedure necessary, rather
than lack of care coordination. However, it is important to point out that this
does not preclude care coordination efforts and diligent follow-up. At our
institution, a designated nurse practitioner calls every post-PD patient the
day following hospital discharge, and most have visiting nurse services.
Many other high-volume institutions have similar follow-up care mechanisms
in place, likely explaining the studies’ findings demonstrating that
readmissions post-PD are largely due to complexity of the disease and
operation.
TIMELINE OF READMISSION AFTER
PANCREATICODUODENECTOMY
So when do readmissions after pancreaticoduodenectomy actually occur? In
our institutional analysis, 50% of readmissions occur within 7 days after
discharge, and the rest in the subsequent 6 weeks (Figure 1 ). In analyzing
trends across 3 decades, we found that as we decreased the length of hospital
stay from 22 days in 1982 to 7 days in 2012, readmission rates increased
from 3% to 15% (Figure 2 ). Taken together, most readmissions may be a
safety mechanism to address these complications as we continue to reduce
length of hospital stay via enhanced recovery pathways. One argument is to
retain patients for an additional 3-4 days to manage late complications,
which would effectively decrease readmission rates by almost 50%.
However, it should be noted that more than 80% of patients do not require a
readmission, and it would be counterintuitive to be retaining those patients to
prevent readmissions in the 15% to 20% that would need them. Rather, we
should use readmissions as a fallback mechanism to allow us to safely
discharge patients as early as critical pathways dictate.
FIGURE 1: The distribution of readmissions across time after
discharge post pancreaticoduodenectomy (Adapted with permission
from Springer in Fong ZV, Ferrone CR, Thayer SP, et al.
Understanding hospital readmissions after
pancreaticoduodenectomy: can we prevent them? a 10-year
contemporary experience with 1,173 patients at the Massachusetts
General Hospital. J Gastrointest Surg 2014; 18:137-44).
DESTINATION OF READMISSION – NO
PLACE BETTER THAN HOME?
Given the complexity of PD, many institutions performing the procedure are
tertiary referral centers, servicing a significant proportion of patients who
travel beyond their local counties for the operation. This invariably
translates to a higher likelihood that patients requiring readmissions be
readmitted to a hospital that is different from the one that performed their
procedure (non-index hospital). In an analysis of the California state cancer
registry, Yermilov et al. reported that 49% of all PD readmissions were to
non-index hospitals. This has important implications for two reasons: (1)
there is loss of information when patients are not readmitted to their index
hospitals; (2) most non-index hospitals are not familiar with, or equipped to
care for, the post-PD patient with a complication. This is evident by the
findings of a Medicare study performed by Brooke et al. which demonstrated
that patients readmitted to non-index hospitals after a range of common
operations were associated with a 26% higher risk of 90-day mortality when
compared to patients readmitted to index hospitals. It is also not surprising
that they demonstrated that this effect was most significant for patients
undergoing pancreatectomy; patients readmitted to non-index hospitals were
associated with a 44% increase in risk of mortality.
In our institutional analysis, the mortality rate of the readmitted PD patient
was 5.1%, almost 5-fold higher than the non-readmitted patient (1.1%,
p<0.001). This readmitted PD cohort represents a vulnerable cohort that
requires prompt recognition of complications and timely treatment. It is
imperative that we maintain an open communication with our patients, and
facilitate readmission to patients’ home institutions whenever appropriate.
FUTURE DIRECTIONS
The challenge that remains is identifying high-risk patients that will benefit
from early detection of complications and treatment and low-risk patients
who are safe for early discharge, while minimizing readmission rates. A
promising biomarker to risk-stratify patients that has recently been
investigated is C-reactive protein (CRP). CRP is a marker of inflammation,
including infectious complications, and has been shown to be more sensitive
than the traditional white blood cell count used. In a meta-analysis of 7
cohort studies, Adamina et al. assessed the predictive value of CRP levels
for infectious complications in abdominal surgery across 7 major abdominal
procedures. The authors demonstrated that CRP levels 4 days after surgery
had the highest predictive value, with a sensitivity, specificity, positive
(PPV), and negative predictive value (NPV) of 68.5%, 71.6%, 50.4%, and
84.3% respectively. Taking the immediate postoperative inflammatory state
into account, the cutoff for pancreatic resection on postoperative day 4
specifically is 184 mg/l and had a PPV and NPV of 46% and 79%
respectively. While it is poor in predicting complications, its NPV of 79%
makes it an ideal biomarker to predict low risk patients that has been safely
tracked to enhanced recovery pathways for early discharge. It should be
noted that all the data on CRP’s predictive performance in major abdominal
surgery were from studies done in Europe, and this needs to be validated in a
North American cohort and treatment setting before incorporated in one’s
practice.
SALIENT POINTS
After PD, 1 in 6 patients will experience a readmission within 30 days
after discharge from their index hospitalization.
Predictors of readmissions after PD are largely patient-related factors,
such as increasing age, patient comorbidities, multivisceral resection,
postoperative pancreatic fistulas and intensive care unit admission.
The most common causes for readmission after PD were
intraabdominal and wound infections, while only 3-14% of
readmissions were unrevealing with no identifiable cause.
Half of all readmissions after PD occurred within 7 days after
discharge, and is a result of efforts aimed at decreasing length of
hospital stay over the years.
High-risk patients who are clinically ready for discharge should be
discharged, and readmission reserved for the 15% to 20% of patients
that will need it.
Patients requiring a readmission after PD should be readmitted to their
index hospital whenever appropriate, as it is associated with a 44%
decrease in mortality rates as compared to patients readmitted to non-
index hospitals.
SELECTED REFERENCES
1. Fong ZV, Ferrone CR, Thayer SP, Wargo JA, Sahora K, Seefeld KJ, et
al. Understanding hospital readmissions after
pancreaticoduodenectomy: can we prevent them?: a 10-year
contemporary experience with 1,173 patients at the Massachusetts
General Hospital. J Gastrointest Surg. 2014; 18(1):137-44; discussion
144-5.
2. Yermilov I, Bentrem D, Sekeris E, Jain S, Maggard MA, Ko CY, et al.
Readmissions following pancreaticoduodenectomy for pancreas cancer:
a population-based appraisal. Ann Surg Oncol. 2009; 16(3):554-61.
3. Hyder O, Dodson RM, Nathan H, Schneider EB, Weiss MJ, Cameron
JL, et al. Influence of patient, physician, and hospital factors on 30-day
readmission following pancreatoduodenectomy in the United States.
JAMA Surg. 2013; 148(12):1095-102.
4. Brooke BS, Goodney PP, Kraiss LW, Gottlieb DJ, Samore MH,
Finlayson SR. Readmission destination and risk of mortality after major
surgery: an observational cohort study. Lancet. 2015; 386(9996) 884-
95.
5. Adamina M, Steffen T, Tarantino I, Beutner U, Schmied BM,
Warschkow R. Meta-analysis of the predictive value of C-reactive
protein for infectious complications in abdominal surgery. Br J
Surg.2015; 102(6):590-8.
Chu Q, Vollmer C, Zibari G, Orloff S, Williams M, Gimenez M (eds).
Hepato-Pancreato-Biliary and Transplant Surgery: Practical Management of Dilemmas
CASE SCENARIO
A 74-year-old female undergoes pylorus-preserving Whipple for
cholangiocarcinoma. Her surgery is uneventful as is her postoperative
recovery. Her drain is removed on POD#4. She is cleared by physical
therapy and is discharged to home with her husband on POD#7. Three days
later, on POD#10, she presents to the emergency department with vague
complaints of not doing well at home and overall failure to thrive. Evaluation
fails to identify any postoperative complication to account for her situation.
What do you do?
█ BACKGROUND
Readmission following pancreatic resections and pancreaticoduodenectomy
(PD) in particular remains a significant problem with multifactorial
causality. It is an especially important topic given the increasingly prevalent
use of 30-day readmission rates as a surrogate for quality care, and the
creation of mandates by the Center of Medicare and Medicaid services that
potentially affect hospital payments if patients are readmitted for “a
condition that could have reasonably been prevented through the application
of evidence-based guidelines.” Recent publications have reported 30-day
readmission rates ranging from 15 to 23%. At 90 days, the readmission rate
increases to between 19 and 29%. Most major pancreatic surgery centers
have large and often geographically disparate referral bases, and it is
therefore likely that published readmission rates underestimate actual
readmission rates by failing to capture patients that are readmitted at outside
hospitals. The majority of readmissions occur within the first 30
postoperative days, with two studies reporting median times to readmission
of 8 and 9 days from initial discharge, and another reporting that 50% of
readmissions occurred within 7 days. Two studies reported that multiple
readmissions were ultimately required for 21% of those initially readmitted.
The total estimated additional cost for readmitted patients averaged $16,000
in one study, owing to index admission costs that were almost $6,000 more in
readmitted patients and readmission costs that exceeded $10,000.
The most common reasons for readmission vary depending on the time
frame from initial discharge. In a review of 1,173 patients who underwent
PD over a 10-year study period, Fong et al. reported that the most common
reasons for early (within 7 days) readmission were ileus, delayed gastric
emptying (DGE), and pneumonia. They reported that after 7 days, wound
infection, failure to thrive, and intra-abdominal hemorrhage were the
predominant reasons (6). Sadot et al. reviewed 490 patients who underwent
PD as well as distal and central pancreatectomies, and found that the main
reasons for readmission within 30 days were procedure-related infections
(58%), failure to thrive (21%), and non-surgical infections (7%). Between
31 and 90 days, the most common reasons were failure to thrive and
chemotherapy-related symptoms (38%), leak, fistula, abscess or wound
infection (28%), non-surgical infections (21%), and drain malfunctions
(14%). Ahmad et al. compiled medical records from six high-volume
institutions for 1302 patients who underwent PD over a five-year period and
found that the most common reasons for readmission within 30 days were
infectious complications (19.9%) and DGE (14.3%). After 90 days, the most
common reasons were wound infection (14.2%), intraabdominal abscess
(12.5%), and failure to thrive (13.5%). Kent et al. reviewed 578 pancreatic
resections over an eight-year period. Their most common reasons for
readmission were procedure-specific complications such as DGE,
pancreatic fistula, bile leak, and portal vein thrombosis (48%), followed by
general postoperative complications including wound infection and J-tube
damage (18%) and also failure to thrive (18%). In our experience,
postoperative complications, including pancreatic fistula and DGE, account
for many readmissions and can be obvious, or more subtle, presenting with
vague complaints like our patient above. Therefore, patients presenting for
readmission must be thor oughly evaluated for such postoperative
complications prior to attributing their problem to failure to thrive. However,
failure to thrive does comprise an important number of readmissions.
One way to approach the problem of readmission after pancreatic
resection is by attempting to identify which patients are at high risk for
readmission. Fong et al. found that pancreatic fistula (OR 1.86, p=0.004),
multivisceral resection at the time of PD (OR 4.02, p=0.02), length of
hospital stay (LOS) >7 days (OR 1.57, p=0.01), and admission to the
intensive care unit (ICU) (OR 2.9, p=0.0005) were independently associated
with readmissions. When Ahmad et al. performed a multivariate analysis,
they identified preoperative diagnosis of chronic pancreatitis (OR 2.356,
p=0.007), high transfusion requirement (OR 1.875, p=0.007) and
postoperative complications including wound infection (OR 2.2, p=0.005),
intra-abdominal abscess (OR 1.96, p=0.02), and pancreatic fistula (OR 2.4,
p=0.02) to be independently associated with readmission. Kent et al. found
that small pancreatic duct size (<3mm) was the only preoperative or
intraoperative factor associated with readmission (p=0.03). On multivariate
analysis, the following factors independently contributed to readmission: any
postoperative complication (OR 2.24, p=0.006), major postoperative
complication (OR 2.2, p=0033), clinically relevant pancreatic fistula (OR
5.03, p=0.0001), and latent fistula (OR 4.29, p=0.0001).
Of note, most previous publications relating to readmission after
pancreatectomy focus on identifying risk factors for readmission that have to
do with patient or disease factors but rarely take into account other
potentially important and modifiable factors such as disposition environment
and the true impact/availability of the patient’s family or friends in helping
care for the patient. For example, many institutions use fairly routine physical
therapy evaluation such that patients are “tested” at walking and stair
climbing. However, evaluation of completion of most activities of daily
living, like bathing or cooking, is extremely rare. Additionally, many studies,
as indicated above, report risk factors for readmission but do not report on
efforts to incorporate this knowledge into formal disposition planning.
Glass et al. reviewed 658 patients who underwent pancreatectomy over a
10-year period at a single institution and performed a root cause analysis to
identify reasons for readmission: whether or not the readmission was
potentially avoidable, the adequacy of initial discharge, and where failures
occurred. Their readmission rate was 18% at a median of 9 days following
discharge and the median length-of-stay once readmitted was 7 days. Most
common reasons for readmission were the development of postoperative
complications (64%), failure to thrive (14%), and need for diagnostic
evaluation (13%). They examined the disposition of readmitted patients and
found that the majority were discharged to home with services (48%),
followed by discharge to home without services (36%) and then discharge to
rehabilitation (12%). Following readmission, 21% of patients required an
escalation of care on their subsequent discharge. On reviewing the records of
patients requiring readmission, they noted a paucity of information regarding
the ability of the patients and families to manage care at home. In reviewing
the patients for whom these data were available, they concluded that 26% of
patient readmissions were potentially avoidable and that 11% of readmitted
patients had problems that may have been able to be managed on an
outpatient basis with appropriate resources.
Recurring themes were inadequate nutrition or oral intake prior to
discharge, and the inability of patients and families to access adequate help
and resources in managing new adjuncts to their care at home, such as drains
and medications, and in resuming baseline activities of daily living at the
preoperative level. Interventions, such as improved communication regarding
postoperative expectations, better assessment of support systems at home,
increased emphasis on nutritional support, and developing ways to facilitate
patient evaluation aside from inpatient admission may reduce the rate of
avoidable readmissions. Other authors suggest that shorter interval follow-up
after discharge, ideally within 7 days and therefore prior to the time that
readmission rates peak, may better allow for detection and management of
problems that would otherwise lead to readmission.
On the other hand, it has also been suggested that readmission following
pancreatic resection is unavoidable owing to the magnitude of the operation
and not necessarily always predictable. Gawlas et al. reviewed 787
pancreatic resections performed at a single institution over a five-year
period. They reported a 30-day readmission rate of 11.6% and on
multivariate analysis they identified young age (age ≤65 years, OR 1.86,
p=0.007) and development of a postoperative complication, either major or
minor, (Grade I or II, OR 6.5, p<0.001; Grade III or IV, OR 11.99, p<0.001)
to be independent predictors of readmission. The most common reasons for
readmission were leak, fistula, abscess and/or wound infection (45.1%) and
DGE (12.1%). Median time to readmission was 8 days and median length-
of-stay once readmitted was 6 days. When they reviewed their 91 readmitted
patients, they found that the majority of readmissions were related to
development of complications directly attributable to the initial operation.
They did not find any that correlate to the length of the initial hospital stay or
to the discharge disposition, and they did not find evidence that readmissions
were due to poor coordination of care or discharge planning. Only 12
patients were admitted for concerns or symptoms that resolved without an
identifiable cause, and the median time from discharge to readmission for
this patient group was 8 days, a time frame that does not suggest premature or
inappropriate discharge was a likely cause. The authors concluded that
because, to a certain extent, the development of these postoperative
complications is expected, and because there are no evidence-based
practices to prevent or decrease their rate of occurrence that readmissions
required for management should not be used as an indicator of quality care. It
has also been suggested that pushing to decrease short-term readmission rates
may lead to increased length of initial hospital stays and that denying hospital
coverage for reimbursement may limit access to care for patients with
complications and ultimately lead to increased mortality.
Failure to thrive is a significant and particularly challenging problem
following PD and other pancreatic resections. It is reported as the main
reason for readmission in both the short and long-term postoperative period
(12-21% at 30 days, 13.5-38% at 90 days). In these patients, an acute
process justifying in-hospital level of care is not found, yet some studies
suggest that readmissions for this reason trend toward longer lengths-of-stay
than patients with an acute postoperative complication (median 5 versus 3
days respectively, p=0.051). Contributing factors may include patient age,
cancer aggressiveness, and the initiation of adjuvant chemotherapy and
nutrition. Several studies have identified increased age as a predictor of
readmission, raising the possibility that older patients may be prone to
develop failure to thrive and require readmission. The correlation of
readmission with cancer aggressiveness is suggested by the findings of Sadat
et al., which found that elevated preoperative CEA is independently
predictive of late (31 to 90 day) readmission, and that postoperative CA19-9
levels were two-fold higher in the late readmission group than in non-
readmitted patients. In their study, the majority of patients with
adenocarcinoma received postoperative chemotherapy, which was a factor in
late readmission.
SUGGESTED APPROACH
The patient initially presented in the clinical scenario will be readmitted to
an inpatient service where she will undergo a thorough evaluation, including
laboratory studies and imaging, to identify or exclude the development of any
new postoperative complications or infectious issues. In the absence of
identification of postoperative complications to which her complaints could
be attributed, she can be diagnosed as failure-to-thrive. Subsequently, her
admission should be geared towards reevaluating her needs and supports at
home, as well as whether an alternate disposition (facility or visiting nurse)
might be safer for her. A careful assessment should be made of her nutritional
status and oral intake; this assessment should include investigation into what
food is available at home, whether she can prepare it herself, and/or whether
she has someone who can help to prepare it. Additionally important is
follow-up education on nutritional recommendations, such as for small,
frequent meals that may be better tolerated in the short-term. It is also
important to reassess for evidence of pancreatic insufficiency and need for
pancreatic enzymes. Both the patient and her family (and/or relevant others)
should be involved in discussions regarding the resources, should have help
available for managing care at home, including an understand of, and
compliance with, all new medications and other recommended treatments.
Their expectations regarding her continued postoperative recovery should be
reviewed, and it is possible that her disposition would need to be changed to
either home with services or rehabilitation. Safeguards should be put in
place so that any difficulty or failure to improve can be identified prior to
reaching a severity where readmission would be again required. It is
possible that factors such as high preoperative CEA or small pancreatic duct
size that would predict readmission could be identified; however it is
equally possible that none exist. While continued efforts should be made to
optimize patient recovery and disposition following pancreatic resection in
order to lower readmission rates overall, it is unlikely and ultimately
unrealistic to expect that they will ever cease entirely.
SALIENT POINTS
The majority of readmissions following pancreatic resection occur
early, with reported rates of 15-23% in the first 30 days and 19-29% in
the first 90 days. The major reasons for readmission vary depending on
the time frame, with infectious complications and DGE generally
accounting for more of the early readmissions, and failure to thrive, as
well as infectious complications, contributing to those that occur later.
On discharge following pancreatic resection, special attention should
be paid to the adequacy of nutrition and to the ability of patients and
families to manage new medical therapies as well as carry out baseline
activities of daily living.
Clarifying and communicating postoperative expectations to patients,
families, visiting nurses and other providers is crucial and may lead to
a decrease in avoidable readmissions.
Even with appropriate and well-coordinated discharge, the
development of postoperative complications following initial
discharge and leading to readmission is to be expected to some degree.
Therefore, the use of readmission following pancreatic resection as an
indicator of quality care or as a tool for decreasing hospital
reimbursement should be discouraged.
SELECTED REFERENCES
1. Gawlas I, Sethi M, Winner M, Epelboym I, Lee JL, Schrope BA, et al.
Readmission after pancreatic resection is not an appropriate measure of
quality. Ann Surg Oncol. 2013;20(6):1781-7.
2. Ahmad SA, Edwards MJ, Sutton JM, Grewal SS, Hanseman DJ,
Maithel SK, et al. Factors influencing readmission after
pancreaticoduodenectomy: a multi-institutional study of 1302 patients.
Ann Surg. 2012;256(3):529-37.
3. Sadot E, Brennan MF, Lee SY, Allen PJ, Gonen M, Groeger JS, et al.
Readmission after pancreatic resection: causes and causality pattern.
Ann Surg Oncol. 2014;21(13):4342-50.
4. Glass CC, Gondek SP, Vollmer CM, Jr., Callery MP, Kent TS.
Readmission following pancreatectomy: what can be improved? HPB
(Oxford). 2013;15(9):703-8.
5. Kent TS, Sachs TE, Callery MP, Vollmer CM, Jr. Readmission after
major pancreatic resection: a necessary evil? J Am Coll Surg.
2011;213(4):515-23.
6. Fong ZV, Ferrone CR, Thayer SP, Wargo JA, Sahora K, Seefeld KJ, et
al. Understanding hospital readmissions after
pancreaticoduodenectomy: can we prevent them?: a 10-year
contemporary experience with 1,173 patients at the Massachusetts
General Hospital. J Gastrointest Surg. 2014;18(1):137-44; discussion
44-5.
Chu Q, Vollmer C, Zibari G, Orloff S, Williams M, Gimenez M (eds).
Hepato-Pancreato-Biliary and Transplant Surgery: Practical Management of Dilemmas
CASE SCENARIO
A 50-year-old male, with a peri-ampullary adenocarcinoma, underwent a
pancreatoduodenectomy (PD). Intra-operatively, there was a soft and friable
pancreas with a posteriorly placed 3 mm duct. Duct-to-mucosa pancreato-
jejunostomy was performed with 5-0 polydiaxonone sutures. During the
anastomosis, two sutures cut through the pancreas making the operating
surgeon not very confident of the integrity and the ultimate outcome of the
anastomosis. Apart from taking all possible care and measures to ensure a
good outcome (i.e. pancreatic stents / specific drainage / falciform ligament
wrap / tissue glue etc), is there anything else which can be done in this high
risk situation?
█ BACKGROUND
Apart from an experienced surgeon and his / her team taking a fresh guard to
perform a flawless anastomosis appropriately tailored to the situation, in the
particular situation mentioned above, we would prophylactically administer
100 mcg of intravenous octreotide, even before commencing the anastomosis.
This is in view of the anastomosis being at high risk for a post-operative
pancreatic fistula (POPF). Although, there is much conflicting evidence on
this practice, and its routine use for all PDs has now been strongly debated,
we propose its benefit in some select scenarios, like the one described
above.
INTRODUCTION
Morbidity after PD persists at 20-40% despite a reduction in peri-operative
mortality of < 5% across continents. POPF is one of the major causes of
morbidity, prolonged hospitalization, and even mortality, after PD for
pancreatic and peri-ampullary cancer. The inherent corrosive nature of
activated pancreatic enzymes, often compounded by secondary infection, in
the vicinity of multiple vulnerable anastomoses makes early resolution of
POPF highly desirable in this life threatening situation. Apart from the
clinician monitoring the sick patient with a hawk’s eye, appropriate drainage,
antibiotics and nutrition, synthetic somatostatin analogues (octreotide,
lanreotide, vapreotide and pasireotide) have been used in clinical practice
with the sole goal of preventing POPF or its attending complications to
hasten post-operative recovery.
DO SOMATOSTATIN ANALOGUES
REALLY MAKE A DIFFERENCE?
Apart from the role of octreotide in influencing overall morbidity, its ability
to specifically reduce the POPF output, rate to spontaneous closure of POPF,
and impact on hospital stay have all been debated over the last few decades.
POPF
They have been shown to reduce overall POPF rates, but data is pre-
dated to the currently practiced standardized definition of clinically-
relevant POPF (Table 2: studies 1 to 4). A Cochrane review has shown
overall number of participants with postoperative complications was
significantly lower in the somatostatin analogue group (RR 0.70; 95%
CI 0.61 to 0.80; n = 1903). Incidence of POPF was lower in the
somatostatin analogue group (RR 0.66; 95% CI 0.55 to 0.79; n = 2206),
but on inclusion of trials that clearly distinguished clinically relevant
fistulas (CR-POPF), there was no significant difference between the
two groups (RR 0.69; 95% CI 0.38 to 1.28; n = 292).
The newer pasireotide (first US randomized controlled trial, albeit
single center, to show benefit of somatostatin analogue in impacting
POPF) has been found to significantly reduce Grade III or higher POPF,
leak or abscess (9 vs. 21%, p=0.006).
TABLE 2: Role of Somatostatin Analogues
Fistula Output
Some studies show that the use of octreotide or soma tostatin analogues
reduce pancreatic exocrine secretions and hence POPF output to about
50-70% after 24 hours of administration. (Table 2: study 9).
Peri-operative Mortality
There is unequivocal evidence to suggest that somatostatin and its
analogues do not decrease peri-operative mortality (Table 2: studies 1
and 4).
Re-operation Rates
Somatostatin and its analogues have no effect on re-operation rates (RR
1.26; 95% CI 0.58 to 2.70; n = 687).
Hospital Stay
Shorter hospital stay was noted with use of somatostatin analogues in
malignancy subgroups in initial studies but this effect was negated by
subsequent literature by the same investigators (MD -1.29 days; 95% CI
-2.60 to 0.03; n = 1314) between the groups.
CONCLUSION
Though some randomized controlled trials convincingly prove its advantage
(Table 2, studies 1 to 4), other trials have presented equivocal results (Table
2, studies 6 and 7 ), and meta-analyses advise adequately powered trials
with low risk of bias. This leads one to believe that evidence, Level I or
otherwise, can be practice-changing only to some extent, and, in the
remaining situations, it is left to the discretion of the surgeon on whether
somatostatin analogues should be used, requiring individualized treatment
decisions more than protocols and evidence-based practices.
In our experience of over a thousand Whipple resections while
documenting a CR-POPF rate of 13.4% makes us believe that selective use
of octreotide retains its place in the management of a high risk pancreatic
anastomosis. Although a Cochrane review suggests that based on the current
available evidence, somatostatin and its analogues are recommended for
routine use in people undergoing PD, we advocate its selective use in high-
risk anastomoses as described above.
SALIENT POINTS
The use of octreotide to reduce pancreatic exocrine secretions and
post-operative pancreatic fistula output after Whipple resection (PD)
remains a matter of discussion and debate.
We have convincing level I evidence to show that somatostatin
analogues reduce post-operative complications, but this may have
included biochemical POPF as well.
We also have level I evidence from highly specialized centers, which
showed no-benefit no-harm, but this may not be applicable to the
relatively lesser-experienced relatively-lower volume centers who
still document POPF rates in the range of 15-20%.
We already have retrospective evidence to show that the use of
octreotide in the high risk gland decreased incidence and morbidity of
clinically relevant POPF and translated into lower resource-usage and
cost-saving.
Until it is proven by further adequately powered trials with low risk of
bias, selective use of octreotide in the high risk anastomosis is our
choice, as its potential benefits may outweigh any otherwise noticed
incidental harm.
SELECTED REFERENCES
1. Shrikhande SV, Sivasanker M, Vollmer CM, Friess H, Besselink MG,
Fingerhut A, et al. Pancreatic anastomosis after
pancreatoduodenectomy: A position statement by the International Study
Group of Pancreatic Surgery (ISGPS). Surgery 2017; 161(5):1221-
1234.
2. Klempa I, Schwedes U, Usadel KH. Prevention of postoperative
pancreatic complications following duodenopancreatectomy using
somatostatin. Chirurg. 1979;50(7):427–31.
3. Gurusamy KS, Koti R, Fusai G, Davidson BR. Somatostatin analogues
for pancreatic surgery. Cochrane Database Syst Rev. 2013 ;
(4):CD008370.
4. Vanounou T, Pratt WB, Callery MP, Vollmer CM. Selective
administration of prophylactic octreotide during
pancreaticoduodenectomy: a clinical and cost-benefit analysis in low-
and high-risk glands. J Am Coll Surg. 2007; 205(4):546–57.
5. McMillan MT, Christein JD, Callery MP, Behrman SW, Drebin JA,
Kent TS et al. Prophylactic octreotide for pancreatoduodenectomy:
more harm than good? HPB (Oxford). 2014; 16(10):954-62.
SALIENT POINTS
Post-operative pancreatic fistulas are a significant source of morbidity
associated with pancreatic resections.
There is no clear signal in the literature to suggest a benefit to the
routine prophylactic use of somatostatin analogues following
pancreaticoduoedenectomy to prevent fistula formation.
By decreasing the high output fistula volume, therapeutic somatostatin
analogues may make fistulas easier to manage and limit their
associated complications.
SELECTED REFERENCES
1. Kitahata Y, Kawai M, Yamaue H. Clinical trials to reduce pancreatic
fistula after pancreatic surgery— review of randomized controlled
trials. Transl Gastroenterol Hepatol 2016;1:4.
2. Bassi C, Dervenis C, Butturini G, Fingerhut A, Yeo C, Izbicki J, et al.
Postoperative pancreatic fistula: An international study group (ISGPF)
definition. Surgery 2005; 138(1):8-13.
3. Hackert T, Werner J, Buchler MW. Postoperative pancreatic fistula.
Surgeon 2011; 9(4):211-7.
4. Callery MP, Pratt WB, Kent TS, Chaikof EL, Vollmer CM Jr. A
prospectively validated clinical risk score accurately predicts
pancreatic fistula after pancreaticoduodenectomy. J Am Coll Surg.
2013;216(1):1-14.
5. McMillan MT, Soi S, Asbun HJ, Ball CG, Bassi C, Beane JD, et al.
Risk-adjusted Outcomes of Clinically Relevant Pancreatic Fistula
Following Pancreaticoduodenectomy: Model for Performance
Evaluation. Ann Surg. 2016; 264(2):344-52.
6. Gurusamy KS, Koti R, Fusai G, Davidson BR. Somatostatin analogues
for pancreatic surgery. Cochrane Database of Syst Rev. 2013;
(4):CD008370.
7. Allen PJ, Gonen M, Brennan MF, Bucknor AA, Robinson LM, et al.
Pasireotide for postoperative pancreatic fistula. N Engl J Med 2014;
370(21):2014-22.
8. Gans SL, van Westreenen HL, Kiewiet JJ, Rauws EA, Gouma DJ,
Boermeester MA. Systematic review and meta-analysis of somatostatin
analogues for the treatment of pancreatic fistula. Br J Surg.
2012;99(6):754-60.