Positive antiglobulin tests due to intravenous

immunoglobulin in patients who received bone

marrow transplant
V. M. ROBERTSON, E. H. ROMOND,AND R. c . ASH
L. G . DICKSON,

To investigate an increased frequency of positive direct (DAT) and indirect (IAT) anti-
globulin tests in bone marrow transplant (BMT) patients who received intravenous
immunoglobulin (IVIG). serologic testing was performed weekly on blood samples from
94 consecutive BMT patients. Group 1 (47 patients) did not receive IVIG. Group II (47
patients) received high-dose IVlG a s prophylaxis for cytomegalovirus infections. Before
transplantation no alloantibodies were found in the serums of 92 patients and anti-E was
found in the serums of two patients. DATs were negative in all patients before BMT. Four
percent of Group I had a positive IAT and 13 percent had a positive DAT. In contrast,
25.5 percent of Group II patients had a positive IAT and 49 percent had a positive DAT,
usually within 1 week after initiation of IVlG therapy (p < 0.001). Antibodies identified in
serums and eluates of patients in Group I were anti-A and anti-B. Antibodies identified in
serums and eluates of patients in Group II were anti-A, -B, -D, and -K. Twenty-one lots of
IVlG were tested and antibodies identified were anti-A, -6, -D. and -K. The data suggest
that the higher frequency of positive serologic tests in Group II was due to passively
acquired antibodies from high-dose IVIG. TRANSFUSION 1987;27:28-31.

INTRAVENOUSADMINISTRATION of pooled human
plasma and its derivatives containing high titers of
antibodies to cytomegalovirus (CMV) may provide
effective prophylaxis against CMV disease in patients
undergoing allogeneic bone marrow transplantation
(BMT).'.* Recently, intravenous immunoglobulin
(IVIG) has been introduced for similar use for CMV
pro phyla xi^.^ Human plasma and immunoglobulin
derivatives contain isohemagglutinins and alloanti-
bodies to human blood group antigens that may be
passively acquired by patients who receive these
product^.^" Serologic abnormalities that might be
expected in patients receiving high-dose IVIG include
positive direct and indirect antiglobulin tests. These
serologic findings would affect crossmatching, anti-
body screening, and antibody identification testing,
and would complicate decisions for blood bank per-
sonnel regarding blood component therapy. They
might also lead to uncertainty in the clinical interpre-
tation of hyperbilirubinemia and suspected hemoly-
sis in BMT patients.
This report describes the serologic findings in 47
BMT patients who received high-dose IVIG as pro-
phylaxis for CMV infections, as compared with the
serologic findings in 47 similarly managed BMT
From the Departments of Pathology and Medicine, University
of Kentucky Medical Center, and Lexington Veterans Adminis-
tration Medical Center, Lexington, Kentucky.
Supported by the University of Kentucky and the Veterans Ad-
ministration.
Received for publication July 31, 1985; revision received Janu-
ary 10, 1986, and accepted February 3, 1986.
patients who did not receive IVIG. Our analysis sug-
gests that high-dose IVIG is the probable cause of
many positive antiglobulin tests (ACT) in these
patients.
Materials and Methods
Serologic test results were examined in 94 consecutive
patients who underwent BMT at the University of Ken-
tucky Medical Center between September 1982 and May
1985. With informed consent, treatment was administered
according to the protocols approved by the University of
Kentucky Institutional Review Board on Human Investi-
gation. Group I consisted of 47 patients who underwent
BMT, without IVIG therapy, between September 1982 and
May 1984. Group I1 consisted of 47 patients who under-
went BMT between April 1984 and May 1985 and received
weekly infusions of IVIG in doses of 20 ml per kg or 1000
mg per kg as described by Winston et al.' In most cases
IVIG therapy began a few days before BMT and continued
for 4 months. Patients in Group 1 and Group I I received
similar component therapy with platelets and other com-
ponents containing plasma. Red cells were washed and
fresh-frozen plasma and cryoprecipitates were minor
ABO-group-compatible. The plasma volume of minor
ABO-group-incompatible plateletapheresis components or
pooled platelets was reduced to 50 ml. The number of
major incompatible bone marrow transplants was similar
in both groups.
Standard techniques following manufacturers' instruc-
tions were used for all serologic studies.' A C T were read
macroscopically and microscopically. Before BMT and at
weekly intervals after BMT, serum samples were tested for
ABO isohemagglutinins with commercial A1 and B cells
(Cooper Biomedicals, Malvern, PA). Tests for alloanti-
bodies were performed with low-ionic-strength saline solu-
tion (LISS, Cooper) indirect antiglobulin tests (IAT) with
routine screening cells, antibody identification panels, and
polyspecific antihuman globulin (Cooper). Ficin-1 AT stud-

28
TRANSFUSION POSITIVE ANTIGLOBULIN TESTS DUE TO IVlG 29
1987 Val. 27. No. I

Table 1. IAT and antibody specificity in the serums of patients after BMT’
~~ ~ ~

Specificity in serums
Positive IAT
Anti-A
Patients Total nt (%I n or -B Anti-D Anti-K
Group I $ 47 2 (4.3) 2 2 0 0
Group 115 47 12 (25.5) 12 10 2 1
p < 0.01
~~~ ~ ~ ~~

Testing before BMT was negative in all except two patients.

t n-Number of patients; some patients had more than one antibody.
$ Group I did not receive IVIG.
5 Group II received IVIG.

ies with anti-lgG (Ortho Diagnostics Systems, Raritan,
NJ) were performed on serums that showed positive anti-
body screens. Routine and ficin antibody screening cells
and identification panels were obtained commercially
(Gamma Biologicals, Houston, TX).
Red cell antigen typing was performed on all patients
before BMT with commercial typing serums (Cooper and
Gamma). Polyspecific antihuman globulin (AHG) and
anti-IgG direct antiglobulin tests (DAT) were performed
on samples of the patients’ cells that were anticoagulated
with EDTA before BMT, and at weekly intervals after
BMT. Eluates of patients’ cells that showed positive DATs
were obtained with an acid elution technique (Elu Kit 11,
Gamma) and tested through IAT phase with commercially
prepared A 1 , B, routine, and ficin antibody identification
panels.
Samples from 21 lots of WIG (Cutter Laboratories,
Berkley, CA) were tested for ABO isohemagglutinins and
alloantibodies. Polyspecific AHG and anti-IgG were used
with routine and ficin panels, respectively. IVlG samples
with positive results were titrated by standard twofold
serial saline dilution through IAT phase with selected cells.
The incidence of positive IATs and DATs in patients
who received IVIG was compared with that in patients
who did not receive this product. Statistical comparison of
the two groups was performed with Student’s t test.
Results
Screening of serums collected from 92 patients before
BMT was negative. The serums of two additional patients
contained anti-E before BMT; this antibody is not included
in our analysis. Serums from 92 of the 94 patients demon-
strated the expected isohemagglutinins with A l and B cells;
appropriately, the two patients with severe combined
immune deficiency syndrome (SCIDS) did not have anti-A
and -B. Serums from 12 of the Group I I patients tested
weekly during IVlG therapy showed unexpected reactivity
with A l , B, or 0 cells within 1 week after the beginning of
IVlG therapy. In contrast, serums from only two of the
Group I patients (who did not receive IVIG) demon-
strated similar reactivity (p<O.OI, Table I ) . Anti-D and/or
anti-K were identified in the serums of two of the 12
Group I1 patients who had positive lATs after lVlG ther-
apy. One patient with D-positive red cells received a
D-graft, and had anti-D within I week of IVIG adminis-
tration. The second patient (phenotype D-, K-), who had
no previous transfusions, received a D-,K -graft and had
anti-D and anti-K within 7 to 10 days of lVlG administra-
tion. Neither anti-D nor anti-K was in the serums of
Group I patients.
Twenty-three of the Group I1 patients who had a nega-
tive DAT before WIG therapy had a positive DAT after
IVIG therapy began (Table 2). Only six of Group 1
patients who did not receive IVlG had a positive DAT
(p<O.OOl). Anti-A and anti-B were identified in the eluates
of 16 patients who received W I G and six patients who did
not. Anti-D and/or anti-K were identified in the eluates of
13 patients who received WIG, but in none of the patients
who did not. Of the 13 patients with anti-D in eluates of
their red cells 12 were D+ patients who received D+ grafts
and one was a D- patient who received a D+ graft. Thus,
passively acquired antibodies were identified more fre-
quently in both the serums and eluates of red cells from
patients who received high-dose IVIG than in the serums
and eluates of red cells from patients who did not receive
this product.
We investigated the possibility of IVlG as a source of
the increased incidence of positive DATs and IATs in our
BMT patients. Twenty-four lots of IVlG were adminis-

Table 2. DAT and antibody specificity in eluates of patients after BMT’

Specificity in eluates
Positive DAT
Anti-A
Patients Total nt (%) n or -B Anti-D Anti-K
Group I$ 47 6 (12.8) 6 6 0 0
Group 115 47 23 (48.9) 23 16 13 1
p < 0.001

Testing before BMT was negative in all patients.

t n-Number of patients; some patients had more than one antibody.
4 Group I did not receive WIG.
5 Group II received IVIG.
30 ROBERTSON ET AL.
tered to BMT patients over 13 months; three lots were not
tested. The 21 lots of WIG tested contained anti-A and -B.
ACT titers of anti-A ranged from 2 to 32, and AGT titers
of anti-B ranged from 1 to 16. Two lots of WIG also con-
tained anti-D with titers of 2 and 8 in the AGT phase. One
additional lot of IVIG contained anti-D with a titer of 2
and anti-K with a titer of 1 in the ACT phase. The
observed patterns of anti-D and anti-K reactivity were
identical in patients’ serums and in red cell eluates and in
the IVIG lots studied. A few patients received single
administrations of IVIG from lots that contained anti-D
and anti-K, but a positive AGT did not develop due to
anti-D or anti-K.
Discussion
Apart from infrequent anaphylactoid reactions,
high-dose IVIG administered for CMV prophylaxis
has not been reported to have significant adverse
effect^.^ Positive DATs without associated hemolysis
have occurred in patients who received IVIG in
lower doses (8 ml/ kg or 400 mg/ kg for 5 days) for
chronic idiopathic thrombocytopenic purpura (ITP).8
Positive DATs with mild and transient hemolysis in
ITP patients receiving low-dose IVIG have been
r e p ~ r t e d . ~In
. ’ ~another ITP study,” however, posi-
tive DATs after similar low-dose WIG treatment
were not found. In our series, the BMT patients
who received high doses of IVIG for 4 months had a
higher frequency of positive AGTs than patients who
did not receive IVIG (p<O.OOl). Slight hemolysis
was observed in some of our patients who had posi-
tive DATs and IATs, but it was not proven to be
directly related to the passively acquired alloanti-
bodies detected in the IVIG given.
Positive ACTS in BMT patients may derive from
several sources. In Group I, the presence of positive
DATs and IATs was presumably related to either
passive acquisition of isohemagglutinins contained in
minor ABO-group-incompatible platelets or endoge-
nous isohemagglutinins reacting against newly en-
grafted red cells of donor ABO type in major ABO-
group-incompatible transplants. Some of the patients
in Group I1 may have had positive DATs and IATs
for the same reasons. However, since all 94 patients
received similar blood component therapy and the
number of major incompatible transplants was sim-
ilar in both groups, it appears that this increased
incidence of positive ACTS due to anti-A and -B was
attributable to frequent administrations of high-dose
WIG. Moreover, none of the patients in Group I
had anti-D or -K in their serums or red cell eluates,
whereas 15 patients in Group I1 did. These 15
patients had received lots of IVIG that contained
anti-D and/or anti-K. Since all platelet and plasma
donors had been screened and found negative for
alloantibodies, anti-D and -K reactivity could have
been passively acquired only from the administered
TRANSFUSION
No. I - 1987
Vol. 27.
lots of IVIG. In addition, in most cases those
patients who received IVIG had positive DATs
and/or positive IATs within 1 week of the initiation
of IVIG therapy.
The 21 lots of IVIG tested for blood group anti-
bodies showed anti-A and -B. Anti-D and -K also
were detected in three lots of IVIG. In general, the
patients who acquired anti-D and -K received more
administrations of the lots of IVIG that contained
those antibodies than did the few patients who did
not have unexpected positive DATs and IATs.
The IVIG used is manufactured from pooled hu-
man plasma Cohn fraction I1 subjected to dithio-
threitol reduction, iodoacetamide alkylation, and
stabilization with glycine and 10 percent maltose;
it contains approximately 50 mg IgG per ml. Al-
though Cohn fraction I1 contains IgG anti-A and -B
isoagglutinins, the final product has been assumed
not to have sufficient anti-A or -B to cause destruc-
tion of either A or B red cells.“
We did not observe clinically significant hemolysis
associated with the passive transfer of blood group
antibodies contained in IVIG. Hyperbilirubinemia
may be frequently observed in BMT patients due to
complications such as graft-versus-host disease, hepa-
titis, drug effects involving the liver, and engraftment
of major ABO-group-incompatible donor cells. Thus,
a specific cause of hyperbilirubinemia or hemolysis
in individual BMT patients may be difficult to
determine. We believe, however, that our observa-
tions of the passively acquired antibodies contained
in IVIG may be an important consideration in
patients receiving IVIG.
In the absence of evidence of clinically significant
hemolysis from WIG, we continue to manage BMT
patients’ red cell requirements based on the compati-
bility of red cell products with patients’ serums.I3
However, if evidence of significant red cell destruc-
tion appears in patients who have positive DATs
while receiving IVIG, red cells for transfusion should
be selected to avoid the antibodies identified in
patients’ red cell eluates.
Acknowledgments
The authors thank Ms. Ann Hamlin and the staff of the Uni-
versity of Kentucky Medical Center Pharmacy for assistance in
obtaining samples of IVIG lots and for their excellent record-
keeping practices and also Ms. Linda Taylor for secretarial
assistance.
References
I . Winston DJ, Pollard RB, Ho WG, et al. Cytomegalovirus
immune plasma in bone marrow transplant recipients. Ann
Intern Med I982;97: I 1-8.
2. Condie RM, O’Reilly RJ. Prevention of cytomegalovirus
infection by prophylaxis with an intravenous, hyperimmune,
TRANSFUSION POSITIVE ANTIGLOBULIN TESTS DUE TO lVlG
1987-Vol. 27, No. I
native, unmodified cytomegalovirus globulin. Am J Med
1984;76(Suppl 3A):13441.
3. Winston DJ, Ho WG, Rasmussen LE, et al. Use of intrave-
nous immune globulin in patients receiving bone marrow
transplants. J Clin lmmunol 1982;2:2(Suppl):42S-7S.
4. Oberman HA, Beck ML. Red blood cell sensitization due to
unexpected Rh antibodies in immune serum globulin. Trans-
fusion 1971;11:3824.
5. Niosi P, Lundberg J, McCullough J, Park BH, Biggar DW.
Blood group antibodies in human immune serum globulin. N
Engl J Med 1971;285:1435-6.
6. Lang GE, Veldhuis B. Immune serum globulin: A cause for
anti-Rh,(D) passive sensitization. Am J Clin Pathol 1973;
60~205-7.
7. Widmann FK, ed. Technical manual. 8th ed. Washington:
American Association of Blood Banks, 1981.
8. Warrier I, Lusher JM. Intravenous gamma globulin treat-
ment for chronic idiopathic thrombocytopenic purpura in
children. Am J Med 1984;76(Suppl 3A):193-8.
9. Salama A, Kiefel V, Mueller-Eckhardt C. Effect of intrave-
nous immunoglobulin in immune thrombocytopenia. Lancet
1983;2:193-5.
10. Salama A, Kiefel V, Amberg R. Mueller-Eckhart C. Treat-
ment of autoimmune thrombocytopenic purpura with rhesus
antibodies (anti-Rh,(D)). Blut 1984;49:29-35.
I I. Carroll RR, Noyes WD, Rosse WF, Kitchens CS. Intrave-
nous immunoglobulin administration in the treatment of
severe chronic immune thrombocytopenic purpura. Am J
Med 1984;76(Suppl 3A):181-6.
12. Romer J, Morgenthaler J-J, Scherz R, Skavil F. Characteri-
zation of various immunoglobulin preparations for intrave-
nous application. Vox Sang 1982;42:62-73.
13. Lasky LC, Warkentin PI, Kersey JH, Ramsey NKC, McGlave
PB, McCullough J. Hemotherapy in patients undergoing
31
blood group incompatible bone marrow transplantation.
Transfusion 1983;23:277-85.
Addendum
Since this manuscript was accepted for publication, an article
appeared in the Journal in which hemolysis due to passive transfer
of antibodies occurred after intravenous immune globulin therapy
in two patients.-The Editor.
Reference
I. Copeland EA, Strohm PL, Kennedy MS, Tutschka PJ.
Hemolysis following intravenous immune globulin therapy.
Transfusion 1986;26:410- 12.
~~~ ~
Vickie M. Robertson, MT(ASCP)SBB, Blood Bank BMT
Coordinator, University of Kentucky Medical Center Hospital
Blood Bank, Room HL423, 800 Rose Street, Lexington, KY
40536-0084. [Reprint requests]
Larry G. Dickson, MD, Medical Director, Clinical Laboratory,
Department of Pathology, University of Kentucky Medical
Center.
Edward H. Romond, MD, Assistant Professor of Medicine,
Department of Medicine, Hematology-Oncology Division, Uni-
versity of Kentucky Medical Center.
Robert C. Ash, MD, Director, Bone Marrow Transplant Pro-
gram, Assistant Professor of Medicine, Department of Medicine,
Hematology-Oncology Division, University of Kentucky Medical
Center.