Anda di halaman 1dari 7

Blood Transfusions and Immunophenotypic Alterations of Lymphocyte

Subsets in Sickle Cell Anemia

By Wing-Yen Wong, Darleen R. Powars, Eva A. Operskalski, Joseph Hassett, John W. Parker, Sharada Sarnaik,
Charles H. Pegelow, Margaret W. Hilgartner, Cage S. Johnson, Yi Zhou, Pei-yu Ku, James W. Mosley,
and The Transfusion Safety Study Group

Transfusions purportedly inducedysfunctionofcell-medi- (aged 218 years) with SCA had increased total leukocytes
ated immunity insickle cell anemia (SCA). We studied hema- and monocytes and lymphocytes counts that remained level
tologic and lymphocytic indicesin 173 human immunodefi- insteadofdecreasing,asdidcomparablyagedcontrols.
ciency virus (HIVI-negativesubjects with SCA and 131 black Lymphocyte subsets typically increased in count, but their
controls. Children aged1 to 7 years with SCA had leukocyte percentage remained similar to children. The exceptionwas
counts and percentages of granulocytes, monocytes, natural CD56' cell counts, whichwere incrsaaed in adolescents and
killercells,andT-cell markers (CDZ'CDIlb', CD4'CD26', adults. No lymphocytic subset change suggested impaired
CD4+CD29') that were significantly higher than those for cellular immunity, and none could berelated to transfusion.
control children. Percent total lymphocytes was decreased Prophylactically transfused patients had higher granulocyte
for this age group,but the total number of lymphocytes and counts, but these may arise fromthe complications of SCA
T and B cell counts were similar to controls. Platelets were itseff.
not increased. Adolescents (aged 8 to 17 years) and adults 6 1995 by The American Societyof Hematology.

CA. Recruitment, evaluation, and data collection for other categories

A CQUIRED FUNCTIONAL asplenia in sickle cell anemia
(SCA) is associated with inefficient phagocytosis of
opsonized bacteria outside of the spleen.I4 Whether other
of subjects are described elsewhere.'.'' For patients with congenital
anemias, attention was focused on enrolling persons who required
transfusion most frequently and those who almost never required
components of the immune system, including cellular immu-
blood components. As controls, we recruited the patients' household
nity, are also compromised has been debated. The balance members.
of the evidence is that cellular immunity is either not affected The total of 205 subjects had hemoglobin SS and characterized
at all or at least is not clinically impaired. SCA patients, for themselves as black. For the present study, observations were made
example, do not have the incidence of opportunistic infec- when the patients were clinically welland crisis-free. Thirty-two
tions or severe forms of childhood viral diseases seen in patients have been excluded from the hematologic and immunologic
immunocompromised person^.^ On the other hand, it has analyses for the following conditions: HIV- 1 infection (two patients),
been suggested that the multiple transfusions necessary for excessive numbers of nucleated etythrocytes (nRBCs; seven pa-
many patients with congenital anemias may alter immune tients), age less than 12 months, malignancy, diabetes mellitus, preg-
function by one or several mechanisms.68 nancy, and administration of immunosuppressants (including corti-
We report a prospective, observational, multicenter study costeroids) and immunoglobulin preparations. For this SCA cohort
of 173 subjects, we recruited 131 household members in generally
of 173 human immunodeficiency virus (HIV)-negative SCA
good health who also identified themselves as black; they were not
patients (104 transfused, 69 untransfused) and 131 black evaluated for sickle cell trait.
controls. The major purpose for our study in 1984 was to Blood transfusion history. Patients were categorized as untrans-
monitor for HIV-type 1 (HTV-1) among persons receiving fused if review of their medical records at the clinic that they regu-
repeated blood transfusions. When enrollment began in Au- larly attended showed no blood component therapy in the 2 years
gust 1985, only two persons with SCA in our cohort had before study entry. If medical records were incomplete or not avail-
HIV- 1 infection. By that time, routine blood donor screening able, we reviewed blood bank records at study hospitals. For possible
for anti-HIV-l had already reduced this modest risk of transfusions at other hospitals, we relied on the subject's history.
transfusion acquisition of HIV-1 to an almost imperceptible
level, and observations of SCA were discontinued in Febru-
ary 1989. From the Departments of Pediatrics, Medicine, and Pathology,
The broad scope of the Transfusion Safety Study, nonethe- University of Southern California School of Medicine, Los Angeles,
less, also included the definition of any changes in immuno- CA; Mount Sinai Medical Center, New York, NY; the Department
logic function attributable to blood transfusion. The data we of Pediatrics, Wayne State University, Detroit, MI; the Department
of Medicine, University of Miami, Miami, FL;the Department of
collected gave us the opportunity to examine the hematologic
Pediatrics, Cornell University-New York Hospital Medical Center,
indices and lymphocyte immunophenotypic distributions in New York, NY; and other participating institutions (see Appendix).
SCA as afirst step. We then gave particular attention to Submitted June 20, 1994; accepted November 22, 1994.
differences in immunophenotypic lymphocyte populations Supported by Contracts No. NOI-HB-4-7002, NO]-HB-4-7003,
attributable to the administration of blood. and NOI-HB-9-7074 of the National Heart, Lung, and Blood Insti-
Address reprint requests to James W. Mosley, MD, University of
The national multicenter Transfusion Safety Study had all proto- Southern California School of Medicine, 1840 N Soto St (EDM 108),
cols and questionnaires approved by institutional review boards at Los Angeles, CA 90032.
each participating site. Written consents for participation, including The publication costs of this article were defrayed in part by page
specific mention of anti-HIV-l testing, were obtained from all sub- charge payment. This article must therefore be hereby marked
jects or their legal guardians. "advertisement" in accordance with I8 U.S.C. section 1734 solely to
Patients. The centers that participated in recruiting and observ- indicate this fact.
ing patients with congenital anemias were located inNew York, 6 1995 by The American Society of Hematology.
NY; Miami, FL; Detroit, MI; San Francisco, CA; and Los Angeles, aKxj-4971/pS/85oS-00I7$3.00/0

Blood, Vol 85, No 8 (April 151, 1995: pp 2091-2097 2091


Table 1. Peripheral Blood Mononuclear Cell Designations of lymphocyte subsets less than 85% of the lymphocyte count. We
and Associated Functions interpreted this cytogram to mean that the subsets were underesti-
PhenOtVDe Associated Functionkl
mated because of nRBCs in the lymphocyte gate. These seven SCA
patients have been excluded from the analyses.
CD14' Monocytes Statistical analyses. Values were statistically similar for males
CD20' All B cells and females in this evaluation, and theyhave been analyzed together.
CD20TD21' Resting B cells Age-specific comparisons were made using three strata: age 1 to 7
CDZO'CD21- Includes activated B cells years (n = 30),age 8 to 17 years (n = 65), and age 2 1 8 years (n
CD2' Includes all T and NK cells = 78). Each patient was represented only once. The number of
CDZ'CD11 b' Includes a subset of NK and CD8' cells children was too small to permit further substratification.
associated with suppressor activity To compare the values by age stratum, weusedthe Student-
CD2TD26' Includes T cells with enhanced proliferative Newman-Keds pairwise comparison to adjust for the significance
response and memory level of multiple testing. There was homogeneity of age distributions
CD4' Helperhnducer T cells within the 1- to 7- and 8- to 17-years strata by Wilcoxon rank sum
CD4TD45RA' Unprimedhaive T cells; includes suppressors for testing. There was nonhomogeneity in age distribution among adults;
lg production adult SCA patients were 7.9 years younger than controls. Compari-
CD4+CD45RA- Includes primed T cells involved in recall son of mean values for SCA and control adults with and without age
response to antigens adjustment showed only minor differences in the levels of statistical
CD4TD29' Primed T cells; includes helpers for Ig significance. The values we report for this stratum, therefore, are
production unadjusted.
CD8' Includes a subset of NK and suppressor/ The distributions in cell populations were normalized by a loga-
cytotoxic T cells rithmic or a square-root transformation"; only the distribution of
CDI'HLA-DR' Activated sutmressor/cvtotoxic
.. T cells percent CD4' cells was normal without transformation. The age
adjustments for each of the hematologic and mononuclear indices
were made by fitting regression modelstothenormal values for
During follow-up observations every 6 months, the kindand each individual. For most variables, this procedure showed age as
amount of blood components were recorded for the interval since a linear function; in some instances, however, a quadratic equation
the prior visit. The coordinating center made specific inquiry as to was more appropriate. The age chosen for standardization was 18
whether transfusions were given for (1) prophylaxis to prevent medi- years.
cal complications (eg, cerebrovasculopathy and chronic lung disease) The two-tailed t test was applied to compare values in the same
or (2) on demand (ie, transfusion to manage a limited clinical circum- age group between the two populations. Exact testing was also used.
stance). All P values are two-tailed.
Virologic testing. Anti-HIV-l status was tested at the coordinat-
ing center laboratory by enzyme-linked immunoassay (EIA) of both RESULTS
the initial and all serially collected sera during follow-up (Abbott HIV-l observations. At the time of entry into the study,
Laboratories, North Chicago, IL). Confirmation was by Western only 2 (1.0%) of the 205 persons with SCA were infected
immunoblot assay. During follow-up, a transfused person was not
with HIV-1. One was anuntransfused, asymptomatic woman
evaluated for anti-HIV-l seroconversion unless there was at least
one specimen tested 26 months later.''
who died shortly after entry in an automobile accident. The
Hematologic and lymphocyte immunophenotypic values. Each second patient was a woman withdiabetes mellitus and other
clinical center performed complete blood counts, usually by auto- conditions given 12 units of bloodin 1984. She was first
mated hemocytometers. If nNRBCs are present in a blood specimen, observed in 1986 and died in 1989.
these counters overestimate the total leukocyte count [white blood During the course of follow-up observations, a total of
cell (WBC) count] and the percentage of lymphocytes. For subjects 4,222 units of blood components were given to 146 of the
with a known congenital anemia, we performed a manual count subjects. There were no seroconversions in this group (280
and accordingly corrected the WBC count and the percentage of person-years of observation) or in the group of SCA non-
leukocytes. transfused patients and control subjects (383 person-years
Mononuclear cell subsets were estimated using a whole blood
of observation).
staining technique" and two-color flow cytometry (Coulter EPICS-
Hematologic values. Age-stratified comparison of SCA
C, Hialeah, FL). All immunology laboratories used the same lots
of monoclonal antibodies (Coulter Diagnostics, Hialeah, FL) and patients with black controls showed significantly increased
standardized proto~ols.'~.'~ The monoclonal antibodies to CD anti- numbers of WBCs ( P < .Owl),granulocytes ( P 5 .0001),
gens were paired to monitor broadly the peripheral blood immune and monocytes ( P 2 .0001; Fig 1). These populations were
cells, including T cells, B cells, and natural killer (NK) cells, and increased by 1.8- to 2.7-fold above the controls. The granulo-
to provide a detailed assessment of the CD4'andCD8' T-cell cyte percentage of the WBCs in SCA patients and controls
subsets (Table 1). was the most elevated in the children (56.4%and 38.9%,
Red blood cells were lysed with a modified ammonium chloride respectively; P = .0001); for monocytes, the only significant
solution (Immunolyse, Coulter Diagnostics). However, nRBCs are difference was among the adults (5.6% and 4.5%, respec-
variably ruptured by this procedure, so that no standard correction tively; P = .02).
based on number of nRBCs per 100 WBCs could be used. Attempts
Platelet counts were similar among children with SCA
to base a correction on anti-glycophorin to identify nRBCs were not
(3.88 x IO5 cells per microliter) and controls (3.60 X lo5
In our population of SCA patients, 18 (10%) had ?five nRBCs cells per microliter; Fig 2). Among patients with SCA aged
per 1 0 0 WBCs; the highest number was 49 nRBCs per 100 WBCs. 2 8 years, platelet counts were higher, reaching mean ranges
Seven patients in our SCA population had ?five nRBCs and a sum of 4.64 x lo5and 4.20 X IO5 cells per microliter for adoles-


. I . . . . . ...
71 71 81

.. .. .. ...

.. . . . . .
p---+" ...

loa I Age Group (yean)

.. ...

Fig 1. Mean E9546 confidence interval (Cl11 hematologic values

for three age groups of persons with SCA and black controls. The
indices are total leukocyte count (0). granulocyte count (W), lympho- 1-7 0-17 218 1-7 2188-17
cyte count (01, and monocyte count (C!). The means are categorical
variables; the connecting lines are for convenience in visualizing age Age Group (years)
Fig 3. Age group percentages of the absolute lymphocyte count
in SCA patients and black controls attributed by flow cytometry to
NK cells (CD56'; solid black), B cells (CD20'; hatched) and all T and
cents and adults, respectively. This compares with a signifi- NK cells (CDZ'; open).
cant age-specific decrease in platelet counts among adoles-
cent and adult controls (3.05 X 10' and 2.78 X 10' cells per
microliter, respectively). The ratios of both differences were z I8 years were 567, 805, and 836 cells/pL, respectively,
significant ( P = . M O 1 ). compared with 329, 332, and 249 cells/pL among the con-
Monocytes defined by manual differential were uniformly trols ( P I.006).
increased in SCA over controls by 1.2- to 1.3-fold, but only For children, lymphocyte counts were not significantly
the adult increase was statistically significant between SCA different between SCA (4,4 12/pL)and controls (3,782/pL).
and controls (P= .02; Fig 1). Monocytes defined as CD14' Lymphocyte percentages among WBCs, however, were sig-
by flow cytometry demonstrated more of a monocytosis in nificantly ( P < .OOOl) lower in SCA (36.1%) than in controls
SCA than the manual differential count. The counts for SCA (55.9%). In adolescence (age 8 to 17 years) and adulthood
patients aged 1 to 7 years, aged 8 to 17 years, and agt:d (age z 18 years), lymphocyte counts in SCA remained in
essentially the same numeric range as in childhood (ie,
greater than 4,2OO/pL), but their percentages among WBCs
decreased to34.0%and32.7%. respectively. Controls in
these two older age groups had age-related decreases in lym-
phocytes in respective counts (2,6631pL and 2,276/pL), as
I well as percentage (41.1%and 36.9%) decreases. Conse-
quently, lymphocyte counts were significantly higher in ado-
lescents and adults with SCA compared with controls
(Fig l).
Irnrnunophenotypic results. NK cells (CD56') in SCA
were numerically increased over controls in all three age
strata ( P S .OOOl): mean values for SCA patients were 509/
l pL, 415/pL, and 389/pL, respectively; for controls, I87/pL,
143/pL, and 182/pL, respectively. As a consequence, N K
cells comprised a significantly higher percentage ( P S . 0 0 0 1 )

1 x105 L" Age Group (years)

of lymphocytes across all age strata (Fig 3).
B cells (CD20') among SCA children were not signifi-
cantly increased over controls, either numerically or as a
percentage. Adolescent and adult SCA patients retained es-
sentially the same mean counts as the children (494/pL, 557/
Fig 2. Mean (k95% Cl) platelet counts for three age groups of
persons with SCA and blackcontrols. The means are categorical
pL, and 544/pL, respectively), while the average numbers
variables; the connecting lines are for convenience in visualizing age among controls declined from the childhood level (544/pL,
trends. 328/pL, and 223/pL, respectively). Consequently, total B

Table 2. Absolute Values for Lymphoeyt. Subsets in Patients With Sickle Cell Anemia Compared With Black Controls by Age Group
Mean No. CelldpL (95% Cl) for Age Group* Pairwise
Cell Population Group 1 11-7 yr) Group 2 (8-17 yr) Group 3 ( = l 8 yrl ( P S. .05t

CD20+CD21+(resting B cells)
SCA 281 (189. 417) 245 (188, 320) 163 (127, 211)
Controls 326 (227, 470) 139 (99, 197) 82 (68, 100)
CDZO'CD21- (includes activated B cells)
SCA 157 (115, 216) 249 (202, 308) 319 (263, 387)
Controls 162 (122, 214) 142 (109. 184) 117(101, 136)
CD2+CD1lb' (T suppressor activity)
SCA 573 (440. 747) 563 (473, 670) 528 (450, 619) NS
Controls 279 (206, 377) 264 (199, 351) 245 (210, 287) NS
CD2TD26' (activation marker of CD4'
and CD8' cells)
SCA 1,123 (863, 1,417) 1,220 (1,035, 1.420) 1,067(908,1,240) NS
Controls 527 (365, 718) 456 (315. 624) 570 (480, 668) NS
CD4' (helper/inducer)
SCA 1,559 (1,298, 1,873) 1,619 (1.433, 1,830) 1,566 (1,400, 1,753)
Controls 1,505 (1,255, 1,806) 1,l48 (968, 1,362) 1,024 (933, 1,125)
SCA 868 (687, 1.070) 704 (593, 824) 468 (386, 558)
Controls 910 (725, 1,115) 431 (314, 566) 260 (208, 316)
CD4+CD45RA- (primed)
SCA 639 (521, 787) 856 (747, 982) 1,042 (920, 1,181)
Controls 569 (455, 711) 651 (528, 802) 696 (621, 781)
SCA 508 (387, 646) 707 (608,813) 955 (849, 1,068)
Controls 372 (286, 469) 422 (335, 519) 571 (514, 631)
CD8' (suppressor/cytotoxic T cells)
SCA 983 (811, 1,192) 900 (792, 1,025) 866 (769, 975) NS
Cantrols 828 (680.1,009) 609 (506, 733) 561 (507, 620) (1, 2H1, 3)
CD8'HLA-DR' (activated T cells)
SCA 127 (93, 175) 147 (119, 182) 178 (146, 216) NS
Controls 94 (64, 137) 69 (49, 97) 54 (44, 64) (1.3)
Abbreviation: NS, not statistically significant ( P > .05).
For SCA subjects, n = 30, n = 65, and n = 78 for groups 1,2, and 3, respectively; for controls, n = 22, n = 25, and n = 84, respectively.
t Student-Newman-Keuls multiple pairwise comparison test.

cell counts were significantly higher in adolescents and in SCA and control children, butwerehigher in anemic
adults with SCA compared with controls (Fig 3). The per- adolescents and adults.
centage of B cells became significantly higher in SCA ado- In SCA patients, T cells expressing CD1 l b and CD26
lescents and adults ( P < .OO01; Fig 3). were much more numerous than in controls at all ages. Total
T cells (as estimated by C D 2 7 had the same age-specific numbers of CD4+ cells as well as CD8+ cells and subsets
trend as B cells in SCA and controls. Children in both catego- had patterns that paralleled the lymphocyte counts. CD4+,
ries had similar counts in childhood (3,308/pL and 2,8681 CD8+, and CD8+HLA-DR+ cells showed no downward or
pL, respectively). SCA patients continued to maintain that upward trend with SCA across age groups, compared with
level (greater than 3,0oO/pL) in adolescence and adulthood, steadily lower counts for adolescent and adult controls.
while controls had decreasing numbers (2,0771pL and 1,873/ To delineate the many changes in percentages, counts,
pL,respectively). This difference between SCA patients and and their proportions, we have summarized the differences
controls was significant for the two older age groups ( P 5 between SCA patients and controls in each age group by
.O001 for both strata). As a percentage of lymphocytes (Fig indicating the value for the former to that of the latter as a
3). only the high value (80.7%) among adult controls was ratio (Table 3). Because changes were very numerous in the
statistically significant ( P < .OOOl). percentage and/or number of each major subpopulation, the
Table 2 compares mean counts for subsets of B and T respective subsets were expressed as proportions of the "par-
cells. Resting B cell counts (CD20+CD21+)were the same ent" ontologic lineage. For example, CD20+CD21+ and
in children and became lower in both SCA patients and CD20TD21- are expressed as ratios of the percent and
controls with adolescence and adulthood. The decline, how- number of CD20+ cells.
ever, wasmuch less for SCA patients than for controls. Among children with SCA, monocytes and NK cells were
Activated B cell counts (CD20+CD21-) were also similar increased both as percentages (1.5-and 2.2-fold, respec-
CELL 2095

Tabla 3. Means of Age-Specific Ratiosof Percentages and Countsin SCA Patients ComparedWith Black Controls
~ ~

Ratio of Mean Counts Comparing SCA Patients

Ratio of Mean Percentages Cornparing SCA Patients With Controls With Controls

Cell Population Denominator* Age 1-7yr Age 8-17yr Age 218 yr Age 1-7 yr Age 8-17 yr Age >l8 yr

Monocytes (CD149 PBMC 1.5t 1.7*1.5* 1.7* 2.4* 3.4*

Lymphocytes WBC 0.6$ 0.8t 0.9t 1.9* NS 1.6*
CD56' 1.5$
ALC 2.2* 2.7$ NS 2.1* 3.0*
CD20' ALC NS NS 1.3* 2.4* NS 1.7*
CDZO'CD21' CD20' NS NS 0.8t NS 1.8t 2.0*
CDZO'CD21- CD20' NS NS NS 1.M 2.7*
CD2' ALC NS 0.9t 0.9* 1.7* NS 1.5*
CD2'CDllb' CD2' 1.7* 2.1*1.3t 1.3$ 2.2* 2.1*
CD2TD26' 1.7*
CD2' 1.6* NS 1.9* 2.1* 2.7$
CD4' CD2+ NS NS NS 1.5* NS 1.4*
CD4CD45RA' CD4' NS NS NS NS 1.6t 1.8*
CD4TD45RA- CD4+ NS NS NS NS 1.3t 1.5*
CD4'CD29+ CD4+ 1.3t NS NS 1.7$1.4t 1.7$
CD8' CD2' NS NS NS 1.5* NS 1.5*
CD8'HLA-DR' CD8' NS NS 2.1* NS 2.1* 3.3*
Abbreviations: PBMC, peripheral blood mononuclear cells; WBC, leukocyte count; ALC, lymphocyte count.
*The control denominator cell population for the ratio of the subpopulation in the first column.
t ,005 <: two-sided P .05 by pairwise t test.
* Two-sided P I,005 by pairwise t test.

tively) and counts (1.7- and 2.7-fold, respectively) compared Hematologic indices (Table 4) varied among the three
with controls. Lymphocytes were decreased (0.6) as a per- transfusion categories. Overall, the distributions of leukocyte
centage of WBCs, but their counts were similar. Children and granulocyte counts were statistically very significantly
had no T-cell changes other than in the percentage and higher with regular transfusion ( P S .001). Platelet and lym-
number expressing the CD1 l b and CD26 markers, and phocyte counts were not significantly different; monocyte
in an increased proportion and number of primed cells counts overall were at borderline significance ( P = .03), but
(CD4+CD29+)among CD4' cells. pairwise comparisons were not significant.
Among adolescents and adults with SCA, the size of all For the prophylactic transfusion category, the leukocyte
subpopulations and their subsets were significantly in- and granulocyte counts were both significantly higher by
creased. For proportionate changes, among SCA adults the pairwise comparisons than for either of the other two catego-
percentages of B cells (CD207 and activated B cells ries. Comparisons of the peripheral blood mononuclear cell
(CD20+CD2I-) significantly increased (1.3- and 1. l-fold); subpopulations showed no statistically significant differ-
the proportion of resting B cells declined (0.8). The percent- ences among the three transfusion categories.
age of T cells in adolescents and adults was slightly but
significantly lower (0.9). The percentages of some activation DISCUSSION
markers were variably increased. There are several plausible explanations for increased
Effect of transfusion on lymphocyte subsets. During the counts in the peripheral blood of patients with SCA. Altered
course of observation, 69 (40%) patients were given prophy- trafficking in asplenic subjects is probably a factor in throm-
lactic blood transfusions, 42 (24%) patients received occa- bocytosis, monocytosis, and lymphocytosis.16 The broad-
sional transfusions as needed, and 62 (36%) patients had based increases in T, B, and NK cells may also reflect lym-
been transfused in the 2 years before study entry. The age phocyte activation caused by infections and tissue damage
distributions in the three transfusion groups were nonhomo- and/or increased bone marrow production. Transient lym-
geneous; the medians were 14,22, and 15 years, respectively phocytosis has been reported during sickle cell c r i ~ i s , but
(Kruskal-Wallis test, P = .002). The percentages of children, the CD4+:CD8+ ratio is variably affected.l8.I9
adolescents, and adults receiving prophylactic transfusions In SCA, the number and proportion of CD2+CDllb+,
were 30%, 54%, and 30%, respectively. Accordingly, values CD2+CD26+, and CD4+CD29+ cells is increased in child-
have been adjusted to a standard age of 18 years for the hood (Table 3). This may reflect activation of primed T
comparisons of the transfusion effect. The patients on pro- cells as the frequency of infection increases with progressive
phylactic transfusions had a mean intensity of 12.6 units per hyposplenism. In adolescence and adulthood, the number of
6 months (mean duration of follow up, 33 months; mean activated cells remains high, although their proportion is no
number of transfusions, 73 units). Those treated on demand longer significantly increased. In adults with SCA, we also
had a mean intensity of 1.4 units per 6 months (mean dura- observed an increase in both number and proportion of acti-
tion of follow up, 27 months; mean number of transfusions, vated B cells. We suggest that this B-cell activation reflects
6.4 units). immune stimulation secondary to tissue damage.

Table 4. Comparison of Hematologic Values and Peripheral Blood Mononuclear Cell Subsots in SCA Among Frequently Transfused
(ProphylaxisGroup). Occasionally Transfused ( A s Needed), and Untrandused Patients
Mean Value/pL (95% Cl1 for Blood Transfusion Category'
Regression Across
Index Untransfused (n = 62) As Needed (n = 421 Prophylaxis (n = 691 Categories ( P )

Hematologic values
Hemoglobin (g/dL) 8.3 (7.8, 8.7) 8.2 (7.7, 8.6) 9.6) (8.9, 9.2 ,0001
Platelets ( x i05/pLf 4.56 (4.14, 5.03) 4.86 (4.33, 5.45) 4.34 (3.98, 4.74) NS L29)
WBCS ( / / L ) 11,906 (10,801, 13,125) 12,077 (10,818, 13,482) 14,728 (13,503, 16,065) ,0010
Granulocytes (/pL) 6,967 (6,103, 7,889) 7,250 (6,258, 8,315) 9,637
(8,724, ,0001
Monocytes (/pL) 589 (483, 704) 794 (646, 957) 789 (674, 912) .03
Lymphocytes (/@L) 4,393 (3,902, 4.913) 4,056 (3,495, 4,657) 4,362 (3,906, 4,845) NS (.65)
PBMC subpopulationst
CD14' (/pL) 984 (785, 1,233) 835 (658, 1,061) 850 (709, 1.020) NS (.42)
CD4' UpL) 1,673 (1,473, 1,898) 1,558 (1,334, 1,820) 1,541 (1,369, 1,736) NS l.62)
CD8' UpLI 916 (801, 1,047) 898 (762, 1,058) 891 (786, 1,010) NS (.g61
CD4'/CD8' 1.86 (1.68, 2.04) 1.81 (1.61, 2.03) 1.77 11.61, 1.94) NS ( 5 7 )
CD2'CDllb' (/@L) 588 (491, 705) 521 (416, 653) 538
(454, NS C.82)
CD2TD26' (/pL) 1 , 1 1 1 (928, 1,311) 1,045 (830, 1.286) 1,218 (1,038, 1,414) NS (.70)
CD4+CD29+ (/pL) 874 (746, 1,013) 828 (687, 981) 847
(734, NS (.88)
CD8'HLA-DR' (/pL) 174 (131, 230) 137 (102, 185) 190 ( 152, 237) NS (.19)
CD20+CD21- VpL) 246 (194, 313) 282 (214, 371) 339 (274, 419) NS (.l11
CD56' (/pL) 520 (423, 638) 382 (300, 486) 383 (318, 463) NS (.06)
Age-adjusted to 18 years.
t CD14' counts based on PBMC gate: all others on the lymphocyte gate.

Monocytes and NKcells in SCA were increased compared increased WBC and granulocyte counts. Demeter et all6 ob-
with controls in both percent and number (Table 3). A three- served monocytosis and lymphocytosis but no granulocyto-
fold increase in the percent and number of monocytes has sis in patients after posttraumatic splenectomy.
been previously reported in SCA.20-22 It has been hypothe- In this report we present evidence that a broad-based leu-
sized that this monocytosis arises as an increased tissue kocytosis is characteristic in SCA at all ages, regardless of
phagocytic capacity in response to hemolysis or tissue dam- transfusion history. Granulocytes and monocytes participate
age.22It is also possible that the monocytosis and increase in the increase relatively more than lymphocytes. All T-
in NK cells in SCA augments host defense. Both monocytes and B-cell subsets participate in the lymphocytosis in SCA,
and NK cells participate in antibody-dependent cellular cyto- although primed, activated T cells are relatively prominent
toxicity mechanisms that provide important and effective in childhood SCA and activated B cells in adult SCA. Our
responses to viral infections.21 findings do not distinguish the relative contributions to in-
Neither the extent nor frequency of transfusion altered the creased cell numbers in SCA of increased production, altered
generalized lymphocytosis seen in SCA patients. Grady et trafficking, immune stimulation, and the infections and tissue
a17made a similar observation in splenectomized thalassemic damage present. Our findings suggest that transfusions in
individuals. We observed no transfusion-associated effect on SCA do not significantly alter immune competence, as mea-
NK cell numbers or CD4:CDS ratios. These findings are in sured by the peripheral blood phenotypic profile.
contrast to those of Escalona et a12' and other^^,^^ who re-
ported in SCA patients a decrease in CD4+:CD8' ratios ACKNOWLEDGMENT
attributed to increased CD8+ cells. We observed an increase We appreciate the contributions of the many people who partici-
in CD8+ cells balanced by an equal increase in CD4+ cells. pated at various times since the study began in May 1984 and who
Wang et aiz5reported that CD4:CD8 ratios were unaffected have been acknowledged in earlier publications.
by the numbers of units transfused in SCA patients undergo-
ing chronic transfusions. Similarly, Hassett et alZ found no
significant effect on CD4' or CD56' cells in patients with The following persons either have responsibility at present
various clotting factor deficiencies receiving factor therapy. for the conduct of the Transfusion Safety Study Group or
The extent of transfusion therapy is closely linked to dis- contributed particularly to the present report: J.W. Mosley,
ease progression and severity. Rather than being a transfu- J. Buckley, M. Hanis, C.K. Kasper, H. Lee, M.J. Nowicki,
sion effect, the higher number of WBCs, granulocytes, and E.A. Operskalski, M.C. Pike, and Y. Zhou (University of
monocytes in childhood patients receiving prophylactic Southern California, Los Angeles, CA); C. Hyman (Cedars-
transfusions may arise from progression of severe ischemic Sinai Medical Center, Los Angeles, CA); S.H.Kleinman
microvasculopathy beginning at an early age. We observed (University of California, Los Angeles, CA); S.L. Dietrich
a similar increase in WBCs, granulocytes, and monocytes in (Huntington Memorial Hemophilia Center, Los Angeles,
untransfused children with SCA compared with controls. CA); J.M. Lusher, J. Kaplan, and S. Sarnaik (Wayne State
The underlying disease process in SCA, therefore, may cause University, Detroit, MI); E.R. Schiff, M.A. Retcher, C.H.

Pegelow, J.D. Temple, and S. Toledano (University of Mi- 11. Horsburgh CR Jr, OuCY, Jason J, Holmberg SD, Longini
ami, Miami, FL); L.M. Aledort and J. Hassett (Mount Sinai IM Jr, Schable C, Mayer KH, Lifson AR, Schochetmen G, Ward
Medical Center, New York, NY); M. W. Hilgartner (Cornel1 JW, Rutherford GW, Evatt BL, Seage GR III, Jaffe H W : Duration
of human immunodeficiency virus infection before detection of anti-
University Medical Center, New York, W ) ;M.F. Kaufman
body. Lancet 2:637, 1989
and Y. Kulpa (Long Island College Hospital, New York, 12. Fletcher MA, Baron GC, Ashman MR. Fischl MA, Klimas
NY); A.R. Brown and S.T. Miller (State University of New NG: Use of whole blood methods in assessment of immune parame-
York Downstate Medical Center, New York, NY); C.E. Ste- ters in immunodeficiency states. Diag Clin Immunol 5:69, 1987
vens and P.E. Taylor (Greater New York Blood Program, 13. Parker JW, Adelsberg B, Azen SP, Boone D, Fletcher MA,
New York, NY); E. Donegan and M.A. Koerper (University Gjerset GF, Hassett J, Kaplan J, Niland JC, Odom-Maryon T, Oper-
of California at San Francisco, San Francisco, CA); M.P. skalski EA, Prince H, Scott D, Stites DP, Mosley J W , Transfusion
Busch (Irwin Memorial Blood Centers, San Francisco, CA); Safety Study Group: Leukocyte immunophenotyping by flow cytom-
R.A. Johnson and B.H. Lewis (Alta Bates Hospital, San etry in a multisite study: Standardization, quality control, and normal
Francisco, CA); G.F. Gjerset (Puget Sound Blood Center, values in the Transfusion Safety Study. Clin Immunol Immunopathol
Seattle, WA); C.G. Hollingsworth, G.J. Nemo, and J. Hoak 55:187, 199
14. Fletcher MA, Azen SP, Adelsberg B, Gjerset GF, Hassett J,
(National Heart, Lung, and Blood Institute, Bethesda, MD);
Kaplan J, Niland JC, Odom-Maryon T, Parker JW, Stites DP, Mosley
and H.J. Alter, D J . D’Alessio, G.F. Grady, P.V. Holland, JW, Transfusion Safety Study Group: Immunophenotyping in a
N.R. Rose, L.B. Seeff, and J. Wittes (Protocol and Data multicenter study: The Transfusion Safety Study experience. Clin
Monitoring Committee, Bethesda, MD). Immunol Immunopathol 5 1:38, 1989
15. Zar JH (ed): Biostatistical Analysis. Englewood Cliffs, NJ,
REFERENCES Prentice-Hall, 1984
1. Bjornson AB, Lobel JS, Harr KS: Relation between serum 16. Demeter J, Mihalik R, Benczlir M, Lehoczky D, PAI6czi K
opsonic activity for Streptococcuspneumoniue and complement Peripheral blood leukocyte subpopulations a long time after posttrau-
function in sickle cell disease. J Infect Dis 152:701, 1985 matic splenectomy. Haematologia 24: 139, 1991
2. Luo NP, Rowland HAK: The serum bactericidal and opsoniz- 17. Groom DA, Kunkel LA, Brynes RK, Parker JW, Johnson CS,
ing defect in sickle cell anaemia: Restoration of activity by control Endres D: Transient stress lymphocytosis during crisis of sickle cell
serum. Trans R SOCTrop Med Hyg 80:317, 1986 anemia and emergency trauma and medical conditions. Arch Pathol
3. Styrt B: Infection associated with asplenia: Risks, mechanisms, Lab Med 114570, 1990
and prevention. Am J Med 88:5, 1990 18. Venkataraman M, Westerman M P B-cell changes occur in
4. Wong WY, Overturf GD, Powars DR: Infection caused by patients with sickle cell anemia. Am J Clin Pathol 84:153, 1985
Streptococcus pneumoniue in children with sickle cell disease: Epi- 19. Donadi EA, Falcao R P Are the changes of lymphocyte sub-
demiology, immunologic mechanisms, prophylaxis, and vaccination. sets in sickle cell anemia due to the loss of splenic function? Acta
Clin Infect Dis 14:1124, 1992 Haematol 80:91, 1988
5. Platt OS, Dover GJ: Sickle cell disease, in Nathan DC, Oski 20. Hendriks J, De Cuelaer K, Williams E, Serjeant G R Mononu-
FA, (eds): Hematology of Infancy and Childhood, 4th ed. Philadel- clear cells in sickle cell disease: Subpopulations and in vitro response
phia, PA, W.B. Saunders, 1993, p 732 to mitogens. J Clin Immunol 13:129, 1984
6. Kaplan J, Samaik S , Gitlin J, Lusher J: Diminished helper/ 21. Escalona E, Malave I, Rodriguez E, Araujo Z, Inati J, Arends
suppressor lymphocyte ratios and natural killer activity in recipients A, Perdomo Y: Mitogen induced lymphoproliferative responses and
of repeated blood transfusions. Blood 64:308, 1984 lymphocyte sub-populations in patients with sickle cell disease. J
7. Grady RW, Akbar AN, Giardina PJ, Hilgartner MW, De Sousa Clin Lab Immunol 22:191, 1987
M: Disproportionate lymphoid cell subsets in thalassaemia major: 22. De Ceulaer K, Pagliuca A, Forbes M, Maude GH, Serjeant
The relative contributions of transfusion and splenectomy. Br J BE, Serjeant GR: Recurrent infections in sickle cell disease: Haema-
Haematol 59:7 13, 1985 tological and immune studies. Clinica Chim Acta 148:161, 1985
8. Pahwa S , Sia C, Harper R, Pahwa R: T lymphocyte subpopula- 23. Robertson MJ, Ritz J: Biology and clinical relevance of hu-
tions in high-risk infants: Influence of age and blood transfusions. man natural killer cells. Blood 76:2421, 1990
Pediatrics 76:914, 1985 24. Ballester OF, Abdallah JM, Prasad AS: Lymphocyte subpop
9. Fletcher MA, Mosley JW, Hassett J, Gjerset GF, Kaplan J, ulation abnormalities in sickle cell anemia: A distinctive pattern
Parker JW, Donegan E, Lusher JM, Lee H, Transfusion Safety Study from that of AIDS. Am J Hematol 21:23, 1986
Group: Effect of age on human immunodeficiency virus type 1- 25. Wang W, Herrod H, Presbury G, Wilimas J: Lymphocyte
induced changes in lymphocyte populations among persons with phenotype and function in chronically transfused children with sickle
congenital clotting disorders. Blood 80851, 1992 cell disease. Am J Hematol 20:31, 1985
10. Lusher JM, Operskalski EA, Aledort LM, Hilgartner M, 26. Hassett J, Gjerset GF, Mosley JW, Fletcher A, Donegan E,
Dietrich SL, Gjerset GJ, Donegan E, Mosley JW, Transfusion Safety Parker JW, Counts RB, Aledort LM, Lee H, Pike MC, Transfusion
Study Group: Transmission of human immunodeficiency virus infec- Safety Study Group: Effect on lymphocyte subsets of clotting factor
tionto sexual and household contacts of persons with congenital therapy in human immunodeficiency virus- 1 -negative congenital
hematologic disorders. Pediatrics 88:242, 1991 clotting disorders. Blood 82:1351, 1993