Detection of Circulating Donor White Blood Cells in Patients Receiving Multiple
Transfusions
By Paul T. Adams, Robertson D. Davenport, David A. Reardon, and Mark S. Roth
Significant morbiditiesare associated with the routine administration of blood products. Although the exact etiology of
these complications may be unknown, many are thought to
arise from the incidental cotransfusion of ”donor” lymphocytes. We have developed an assay to detect small numbers
of male white blood cells (WBCs) circulating in female
patients who have received multiple blood transfusions
using the polymerase chain reaction (PCR). Twenty female
patients undergoing major surgical procedures were studied
and received an average of 9.3 U of packed red blood cells
(4.8 U from male donors) and 11.7 U of platelets (6.1 U from
male donors). DNA was extracted from whole blood or
peripheral blood buffy coats posttransfusion and PCR per-
formed using oligonucleotides designed to amplify a segment within the repetitive Y-chromosome DYZl locus. Posttransfusion, 15 of 20 women showed evidence of circulating
male WBCs for an average of 2.0 days (range, 1 to 6). We
conclude that (1) DYZl PCR analysis is a useful approach for
the detection of small numbers of circulating transfused male
WBCs in female patients; and (2) circulating donor WBCs
persist for a mean of 2.0 days in the majority of women
receiving multiple transfusions. Future application of this
technique may detect persisting or proliferating WBCs and
lead to an improved understanding of common transfusionrelated morbidities.
o 1992by TheAmerican Society of Hematology.
T
with a range of 1to 87 U. The number of male units transfused was
identified retrospectively through the records of the Southeastern
Michigan Red Cross. Patients averaged 4.8 U of transfused male
PRBCs (range, 1 to 10 U), and averaged 6.1 U of transfused male
platelets (range, 0 to 48 U). The average total number of male units
transfused was 10.8 U (range, 2 to 56 U). Median patient follow-up
from the day of surgery was 8.0 days, with a range of 5 to 53 days.
After institutional review board approval and informed consent,
peripheral blood (PB) samples were obtained on the day before
surgery and on subsequent hospital days. Samples were collected
concomitantly with clinical blood draws and, as a result, were not
obtained for each hospital day in all patients. A median of 9.0
blood samples were obtained for each patient (range, 5 to 28).
DNA preparation. DNA was prepared from PB bufFy coats or
whole blood as previously described.’l In brief, bufFy coats or whole
blood were resuspended in a sucrose lysis buffer (0.32 mol/L
sucrose, 10 mmol/L Tris HCI [pH 7.51, 5 mmol/L MgC12, 1%
Triton X-100) and pelleted three times to lyse RBCs and obtain a
nuclear pellet. The pellet was then resuspended in 500 pL of PCR
lysis buffer with nonionic detergents (50 mmol/L KCI, 10 mmol/L
Tris HCI [pH 8.31, 2.5 mmol/L MgC12, 0.1% gelatin, 0.45% NP40
[Sigma Chemical Company, St Louis, MO], 0.45% Tween 20
[Sigma], and proteinase K [0.6 mg/mL]), incubated for 1 hour at
56”C, and then placed in a boiling water bath for 10 minutes.
Twenty-five microliters of this mixture was then used directly for
amplificationby PCR.
Y-chromosome-specific gene amplijication and detection. Oligonucleotide primers synthesized on an Applied Biosystems synthesizer model 380B (Applied Biosystems, Foster City, CA) were
HE INCIDENTAL transfusion of white blood cells
(WBCs) has been implicated as the principal cause of
significant morbidities associated with the routine administration of blood products. These morbidities include fever,’
leukoagglutination,2 viral tran~mission,~
increased refractoriness to platelet transfusions: and transfusion-associated
graft-versus-host disease (TA-GVHD).S-7The latter complication was initially reported in immunocompromised patients receiving blood products, but more recently has been
described in immunocompetent patients as well: Fatal
postoperative erythroderma, caused by TA-GVHD, has
been reported to occur in approximately 0.2% of Japanese
patients undergoing open heart surgery. This observation
within the homogeneous Japanese population, in addition
to the development of TA-GVHD in patients receiving
HLA haplo-identical blood products: suggests that HLA
compatibility between blood donor and recipient may
influence the survival of cotransfused “donor” WBCs.
Prolonged survival may in turn predispose immunocompetent patients to the development of TA-GVHD. Previous
studies using radionucleotide labels have shown half-lives
of from 5 to 8 hours for autologous granulocyte transfus i o n ~ .Circulating
~,~
transfused donor cells have been detected in four children with severe combined immunodeficiency syndrome using restriction fragment length
polymorphism analysis.1° At present, little is known about
the incidentally transfused “donor” WBCs’ ability to survive within the circulation of an immunocompetent host. To
address this issue, we have developed an assay using
Y-chromosome-specific polymerase chain reaction (PCR)
to detect small numbers of circulating transfused male
WBCs after transfusion of male blood products into female
patients. Using this assay, circulating transfused donor
WBCs were detected in 15 of 20 patients for a mean of 2.0
days after transfusion.
MATERIALS AND METHODS
Patient material. A total of 20 female patients undergoing open
heart surgery (18 patients) or liver transplantation (2 patients) at
the University of Michigan Medical Center between September
1989 and March 1990 were studied (Table 1). The study patients
received an average of 9.3 U of packed red blood cells (PRBCs),
with a range of 2 to 25 U. Sixteen of 20 patients also received
platelet transfusions. These patients averaged 11.7 U of platelets,
Blood, Vol80, No 2 (July 15). 1992: pp 551-555
From the Departments of Medicine, Pathology, and Pediatrics,
University of Michigan Medical School, Ann Arbor.
Submitted November 27,1991; accepted March 27,1992.
Supported in part by grants from the American Cancer Society,
Children’s Leukemia Foundation of Michigan, and National Institutes
of Health Grants No. R29-DK43670-01, T32-HLO7622, and F2CA090224.
Presented in part at the American Society of Hematology annual
meeting Boston, December 1990.
Address reprint requests to Paul T.Adams, MD, F7828 Mott, 1500
E Medical Center Dr, Ann Arbor, MI 481 09-0247.
The publication costs of this article were defrayed in part by page
charge payment. This article must therefore be hereby marked
“advertisement” in accordance with 18 U.S.C. section 1734 solely to
indicate this fact.
0 I992 by The American Society of Hematology.
0006-4971/92/8002-0025$3.00/0
551
ADAMS ET AL
552
Table 1. Summary of Transfusions Received and DYZl Detection
Posttransfusion
Patient
No.
RBCs
RBCs
Male
Platelets
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
Mean
SD
Median
11
8
6
3
6
13
8
5
8
2
5
11
14
12
7
10
25
7
12
13
9.3
4.9
8
6
5
2
2
2
10
7
4
4
2
2
4
5
7
3
6
8
1
8
7
4.8
2.5
4.5
6
6
6
0
6
12
12
0
6
0
0
6
12
18
6
1
87
7
30
12
11.7
19.1
6
Male
Platelets
Total
Maleunits
Positive
Days'
F/U
Dayst
2
1
4
0
4
3
7
8
6
6
2
6
13
14
4
7
2
2
6
12
13
6
7
56
6
26
14
10.8
11.9
6.5
1
1
3
0
2
4
2
4
0
1
0
3
1
1
0
0
1
1
1
6
2.0
16
6
8
6
11
8
8
7
12
5
11
6
31
53
8
18
18
9
12
7
13
11.2
8.5
0
3
0
0
2
7
6
3
1
48
5
18
7
6.1
10.7
3
'The number of days a positive signal is seen after the transfusion of
male blood products and before the transfusion of additional male
blood products. Mean number of positive days refers only to the 15
DYZl(+) patients.
tThe last day after surgery that a blood sample was obtained for
analysis.
designed to amplify a 300-bp segment within the Y-chromosomespecific sequence DYZ112 (Table 2). PCR was performed in a 50
pL volume containing a 25-pL aliquot of the DNA preparation, 0.2
mmol/L dXTF's, 1 U Taq polymerase, 25 mmol/L Tris HCl (pH
8.3), 25 mmol/L KCl, 4 mmol/L MgC12, and 5 mmol/L dithiothreito1 (DTT) and 0.005 OD260 units of each specific primer (A and B)
(Table 2). Forty cycles of amplification were performed consisting
of denaturing at 94°C for 45 seconds, annealing at 55°C for 30
seconds, and extension at 72°C for 60 seconds. PCRs were
terminated with a 10-minute extension at 72°C. One-fifth of the
PCR product was size-fractionated by gel electrophoresis in
standard 2% agarose (Bethesda Labs, Gaithersburg, MD) gels and
visualized by ethidium bromide staining. The PCR products were
then transferred to a nylon filter (Hybond-N; Amersham, Inc,
Arlington, IL) by alkaline transfer and hybridized as previously
described." To obtain an internal DYZ1-specific DNA probe,
PCR was performed using nested internal PCR primers C and D
(Table 2). The internal PCR product was size-fractionatedon a 1%
low melting point agarose gel (Bethesda Research Labs), excised,
and radiolabeled using random hexamer priming as previously
des~ribed.'~.'~
Blots were exposed by autoradiography at -80°C
with a single intensifying screen for 1 to 4 days. The detection of a
300-bp fragment confirmed the presence of the Y-chromosomespecific segment. Parallel mock DNA preparations were amplified
as negative controls and known male blood samples were used as
positive controls. For each DNA sample, the presence of intact
DNA for amplification was verified by performing PCR on a
separate aliquot with primers specific for non-Y-chromosome
DNA sequence; anti-thrombin I11 or P-globin as previously described"; or using phospho-glycerate kinase-specific primers (5'
primer from positions +186 to +220 and 3' primer from +718 to
+752).15
Statistical analysis. The Mann-Whitney test was used to determine the association of amount and type of blood product
transfused with the presence or absence of detectable circulating
"donor" WBCs. This test was chosen over the t-test because of the
considerable variability between patients in quantities of blood
products received.
RESULTS
Sensitivity and specificity of PCR. Figure 1 shows the
results of a PCR mixing study performed using purified
DNA corresponding to the content of approximately 100,000
female cells mixed with decreasing amounts of male DNA.
The upper panel shows the ethidium bromide-stained
agarose gel, while the lower panel shows the same samples
after Southern blot analysis. In both panels, the detection
of Y-chromosome sequence by DYZl PCR is seen to be
specific for male cells (lane 2). From the reconstitution
experiment, DNA corresponding to as few as 10 male cells
can be detected in the presence of approximately 100,000
female cells by direct visualization on an agarose gel (lane
6, upper panel). After Southern transfer and hybridization
with a radiolabeled probe, a quantity of DNA equivalent to
the content of single male cell could be detected (lane 7,
lower panel). PCR from serial mixtures of male and female
cells, rather than purified DNA, resulted in similar sensitivity (data not shown). The absolute specificity of this method
was further confirmed by the analysis of DNA prepared
from 25 separate female patient blood samples and 25
separate male patient blood samples. Each male sample
was positive for the presence of the DYZl segment of the
Y-chromosome, whereas all female samples were negative.
In addition, preoperative blood samples from all 20 female
patients were studied and found to be DYZl negative.
Detection of circulating transfused W C s . Fifteen of 20
female patients showed evidence of circulating male cells
after transfusion (Figs 2 through 4 and Table 1). Thirteen of
18 female patients undergoing open heart surgery had
Table 2. DYZ1-Specific PCR Primers
Primer
A
B
C
D
Sequence (5'-3')
Location
TTCCAATCCATTCCllTCCmCGCTTGCA
TGGATAGTAATCGACTGGAGTG~TGGAC
TKCA'ITCTATTCCCTTCTACTGCATACAA
TGGAAAGGAATGGACTCAG~GGGC
11-40
281-310
41-70
251-280
Strand
+
-
+
-
Primers C and Dare nested primers located internal to primers A and B, respectively.Location refers to base pair sequence as reported by Nakahori
et a1.12
Abbreviations: +, sense strand; -, antisense strand.
553
DETECTION OF CIRCULATING MALE WBCs
1 2 3 4 5 6 7
from malc donors, and 6 U of platclcts, 1 U from a malc
donor), with no DYZI-containing malc cclls dctcctcd on
days 2 through 6 (days 5 and 6 not shown).
Fiftccn of thc 20 patients showcd cvidcncc of circulating
malc cclls for a mcan of 2.0 days, with a rangc of 1 to h days
aftcr thcir last transfusion of blood product (Tablc 1). Thc
15 patients with dctcctablc malc cclls rcccivcd a mcan of
10.1 U of PRBCs (5.1 from malc donors). and 14.7 U of
platclcts (7.6 from malc donors). Thosc patients without
cvidcncc for circulating malc WBCs rcccivcd an avcragc of
6.6 U of PRBCs (3.4 from malc donors), and 2.6 U of
platclcts (1.4 from malc donors) (Tablc 3).
Thc Mann-Whitncy tcst was uscd to analyzc for diffcrcnccs bctwccn patients with or without circulating transfused malc cclls (Tablc 3). Thosc patients with cvidcncc for
circulating WBCs werc significantly morc likcly to rcccivc
larger quantities of platclcts (P = .On) and malc platclcts
(P = .OS).
DISCUSSION
Fig 1. Sensitivity and specificity of DYZl PCR. The upper panel
shows PCR product?.
size-fractionated on an ethidium bromidestained agarose gel and the lower panel shows Southern blot analysis
of the same gel using a DYZ1-specific hybridization probe (see text).
pBR plasmid digested with Mspl was run as a size marker (lane 1).
Lane 2, DYZ1 PCR product with template DNA extracted from male
cells; lane 3, DYZl PCR product with template DNA extracted from
female cells. Lanes 4 through 7 show DYZl PCR products obtained
from serial dilutions of male DNA mixed with female DNA with the
total corresponding t o the DNA content of lo5 cells. Lane 4, DNA
equivalent t o l o 3 male cells; lane 5, 10' male cells; lane 6, 10 male
cells; and lane 7. one male cell.
dctcctablc malc cclls, as did both livcr transplant patients
(patients 14 and 17, Tablc 1). In fivc patients. pcrsistcnt
circulating male cclls wcrc dctcctcd for 3 or morc days aftcr
thc last transfusion of blood products. Two patients showcd
cvidcncc of circulating malc cclls for 2 days posttransfusion,
and cight patients had circulating cells for only 1 day after
transfusion. All fivc of thc patients without dctcctablc
circulating malc WBCs had samplcs analyzcd from thc first
3 postopcrativc days.
Figurcs 2 through 4 show typical patterns of transfuscd
WBC dctcction aftcr transfusions. Figurc 2 shows thc
analysis of multiplc blood samplcs from patient 17 (Table
1). N o DYZl signal was sccn in thc prcopcrativc blood
samplc (day O), hut a positivc signal was ohscwcd during
thc operation (lanes A through E), and on subscqucnt
postsurgical days 1 through h. In addition to rccciving
PRBCs and platclcts on thc day of surgery, this patient
continued to rcccivc daily platclct transfusions through
postopcrativc day 6. Blood samplcs from 10 subscqucnt
days showed no cvidcncc of circulating malc WBCs ( d a y 9
through 18 not shown).
DNA analysis in patient 6 (Fig 3) showcd circulating
malc cclls up to 4 days aftcr thc last transfusion (day h
postsurgcry). For patient 2 (Fig 4). a singlc positivc samplc
was obtained thc day aftcr tramfusion ( 8 U of PRBCs. 5 U
This rcport dcscribcs thc application of PCR to dctcct
thc prcscncc of circulating malc donor WBCs in fcmalc
patients aftcr multiple blood transfusions. Circulating malc
cclls wcrc shown in 15 of 20 fcmalc transfusion rccipicnts.
A statistically significant corrclation was ohscrvcd bctwccn
thc amount of platclcts transfuscd and thc dctcction of
circulating malc WBCs.
post-surgery
'
OABCDE12345678'
I
a a a a a
-8 3
3333
!i!i!i!i!I
7U RBC,
22U PLT
3 a O O h O Y )
c
Fig 2. Detection of circulating male WBCs in patient 17. Blood
samples were obtained preoperatively (lane 0) and during the operative procedure (lanes A through E). In addition, blood samples were
obtained on postoperative days 1 through 18 (days 1 through 8 are
depicted in lanes 1 through 8). Times and amounts of male blood
products transfused are indicated at the bottom. Analysis of additional samples from days 9 through 18 detected no evidence of
circulating male WBCs (not shown). Abbreviations: RBC, red blood
cells; PLT, platelets; U, units.
554
ADAMS ET AL
Dav
post-surgery
post-surgery
'0 1 2 3 4 6 7 8 '
'0 1 2 3 4 '
J
m
5+ 5
5
4
Fig 3. Penlrtent detection of circulating male WBCs in patient 6.
The blood sample labeled day 0 was obtained on the day of surgery
(before transfusion) with subsequent samples obtained postoperatively, as indicatedat the top. Male blood products were administered
as noted at the bonom.
Fig 4. Transient detection of circulating male WBCs in patient 2
Blood samples were obtained preoperatively on day 0 and postoperatively on days 1 through 6 (days 1 through 4 shown). Male blood
products were administered as noted at the bottom. Samples obtained on days 5 and 6 were also negative (not shown).
The amplification of Y-chromosome-specific sequences
by the PCR has been described previously for fetal sex
determination and to detect residual male host cells after
allogeneic bone marrow transplant.161xThese studies have
also found that Y-chromosome-specific PCR is both scnsitive and specific. The Y-chromosome-spccific sequence
DYZl used in the present study is of particular utility as
each Ythromosomc contains between 3,000 and 5,000
copies of this repetitive sequence." This high copy number
markedly enhances the sensitivity of the PCR assay, facilitating easy detection of even a single male ccll (Fig 1).
Dcspite this high sensitivity, a very low level of persistent
donor WBCs could have been missed. Each blood sample
obtained in this study represents only approximately 0.001
of the total patient blood volume, with only 1/20 of the
DNA from each sample used for the PCR reaction. Thus, a
positive signal indicates the presence of 2 20,000 total male
cells in the patient's circulation. In addition, only the
nonmarginated, nonsequestercd, circulating pool of WBCs
is being evaluated.
A negative DYZl PCR assay docs, however, suggest a
significant clearance of transfused WBCs. Each unit of
PRBCs contains on the order of 1 x 10"WBCs,'" whereas a
unit of platelets contains 1oX W B C S . Therefore,
~
a negative
DYZl PCR assay indicates the clearance of 99.9% (after
platelet transfusion) to 99.99% (after PRBC transfusion) of
transfused WBCs.
Table 3. Association of Amount and Type of Blood Productwrth
Detection of "Donor" WBCs
-
1 or More
0 Positive
Days (N-
RBCs
Male RBCs
Platelets
Male
platelets
Total male
units
Median
Mean
Median
Mean
Median
Mean
Median
Mean
Median
Mean
7.0
6.6 f 2.7
3
3.4 f 1.7
1
2.6 f 3.1
1.o
1.4 f 1.5
6.0
4.8 f 2.6
5)
Positive
Days (N+ = 15)
11.0
10.1 f 5.2
5
5.1 f 2.6
7
14.7 f 21.3
4
7.6 f 12.0
8.0
12.7 f 13.1
Mean f SD.
'Statistical analysis using Mann-Whitneytest.
P
Value'
.13
.19
.03
.05
.12
DETECTION OF CIRCULATING MALE WBCS
555
Patients who received larger quantities of blood products
in our study were more likely to have samples positive for
circulating male WBCs (Table 3). This finding may result
from delayed clearance of larger numbers of cotransfused
WBCs, or may be due in part to an increased likelihood of
receiving blood products from donors with shared HLA
phenotypes. An estimated 30% of white Americans share
one of nine common extended HLA haplotypes.2l The
degree of reactivity in mixed-lymphocyte culture (MLC)
between unrelated individuals matched for these common
haplotypes is as low as that found between HLA-identical
siblings.*l The greatest risk for overt TA-GVHD is believed
to be in patients who are heterozygous for a common
extended haplotype and who receive blood products from a
donor homozygous for that same common extended haplotype. However, this situation is estimated to occur in only 2
of every 1,OOO transfusions between white Americans,2l
making it unlikely that such a phenomena is occurring in
our patient population, as we evaluated only 216 transfusions of male blood products. Clinically apparent TAGVHD does not occur with this frequency, suggesting that
all transfusion recipients who may be at risk do not develop
the syndrome. This may be due to decreased survival of
progenitor cells with increasing blood storage time, or may
be due in part to rapid clearance of donor WBCs by the host
through recognition of minor histocompatibility antigens.
Although HLA data was not available from our patient
population, the rapid clearance of “donor” WBCs shown
may help explain this low frequency of TA-GVHD in the
immunocompetent patient.
PCR Y-chromosome detection is thus a highly sensitive
and specific technique for detecting circulating transfused
male WBCs in female recipients. This assay is capable of
detecting circulating donor cells at much lower numbers
than HLA typing, cytogenetics, or Southern blotting. The
physiologic significance and characteristics of such cells
remain to be determined.Application of the DYZ1, Y-chromosome-specific PCR assay to detect persisting WBCs
after transfusion or “subclinical” proliferation of transfused progenitor cells in different clinical settings may
provide important insights into the relationship of donor
WBC survival and common transfusion-associated morbidities.
ACKNOWLEDGMENT
The authors thank David Ginsburg for helpful discussions,
Judith Bromberg for statistical analysis, and Judith Castagna for
help in preparation of this manuscript.
REFERENCES
1. Brubaker DB: Clinical significanceof white cell antibodies in
febrile nonhemolytic transfusion reactions. Transfusion 30:733,
1990
2. Hammerschmidt DE, Jacob HS: Adverse pulmonary reactions to transfusion.Adv Intern Med 27511,1982
3. Schrier RD, Nelson JA, Oldstone MBA Detection of human
cytomegalovirus in peripheral blood lymphocytes in a natural
infection. Science 230:1048,1985
4. Saarinen UM, Kekomaki R, Siimes MA, Myllyla G Effective
prophylaxis against platelet refractoriness in multitransfused patients by use of leukocyte-free blood components. Blood 75512,
1990
5. Juji T, Takahashi K, Shibata Y, Ide H, Sakakibara T, Ino T,
Mori S: Post-transfusiongraft-versus-host disease in immunocompetent patients after cardiac surgery in Japan (letter). N Engl J
Med 32156,1989
6. Thaler M, Shamiss A, Orgad S, Huszar M, Nussinovitch N,
Meisel S , Gazit E, Lavee J, Smolinsky A. The role of blood from
HLA-homozygous donors in fatal transfusion-associated graft
versus host disease after open-heart surgery. N Engl J Med 321:25,
1989
7. Anderson KC, Weinstein JH: Transfusion-associated graft
versus host disease. N Engl J Med 323:315,1990
8. Weiblen BJ, Forstrom L, McCullough J: Studies of the
kinetics of indium-111-labeled granulocytes. J Lab Clin Med
94246,1979
9. McCullough J, Weiblen BJ, Clay ME, Forstrom L Effect of
leukocyte antibodies on the fate in vivo of indium-111-labeled
granulocytes. Blood 58:164,1981
10. Blazar BR, Filipovich AH: Identificationof transfused blood
cells in children with severe combined immunodeficiencysyndrome
by analysis of multiple cell lineages using restriction fragment
length polymorphisms. Bone Marrow Transplant 5327,1990
11. Roth MS, Antin JH, Bingham EL, Ginsburg D: Use of
polymerase chain reaction detected sequence polymorphisms to
document engraftment following allogeneic bone marrow transplant. Transplantation 49:714,1990
12. Nakahori Y, Mitani K, Yamada M, Nakagome Y: A human
Y-chromosomespecificrepeated DNA family (DYZ1) consists of a
tandem array of pentanucleotides. Nucleic Acids Res 14:7569,1986
13. Feinberg AP,Vogelstein B: A technique for radiolabeling
DNA restriction endonuclease fragments to high specific activity.
Anal Biochem 132:6,1983
14. Feinberg AP, Vogelstein B: Addendum: A technique for
radiolabeling DNA restriction endonuclease fragments to high
specific activity. Anal Biochem 137266,1984
15. Michelson AM, Blake CC, Evans ST, Orkin SH: Structure of
the human phosphoglycerate kinase gene and the intron-mediated
evolution and dispersal of the nucleotide-binding domain. Proc
Natl Acad Sci USA 82:6965,1985
16. Pinckert TL, Lebo RV, Golbus MS: Rapid determination of
fetal sex by deoxyribonucleicacid amplification of Y chromosomespecific sequences. Am J Obstet Gynecol161:693,1989
17. Witt M, Erickson R P A rapid method for detection of
Y-chromosomal DNA from dried blood specimens by the polymerase chain reaction. Hum Genet 82:271,1989
18. Lawler M, McCann SR, Conneally E, Humphries P Chimaerism following allogeneic bone marrow transplantation: Detection
of residual host cells using the polymerase chain reaction. Br J
Haematol73:205,1989
19. Mijovic V, Brozovic B, Hughes ASB, Davies TD: Leukocytedepleted blood, a comparison of filtration techniques. Transfusion
23:30,1983
20. van Marwijk Kooy M, van Prooijen HC, Borghuis L, Moes
M, Akkerman JWN Filtration, a method to prepare white cellpoor platelet concentrates with optimal preservation of platelet
viability.Transfusion 3034,1990
21. Kruskall MS, Alper CA, Yunis EJ: HLA homozygosity and
transfusion-associatedgraft versus host disease (letter). N Engl J
Med 332:1005,1990