The natural history of alloimmunization to platelets

Sixty-three patients have provided evidence that platelets are highly immunogenic even in recipients of potentially immunosuppressive therapy for malignant diseases. Approximately 70 per cent of patients who receive repeated transfusions of platelets from random donors over a prolonged period can be expected to develop lymphocytotoxic antibodies. Antibodies became detectable in one patient ten days after his first exposure to HLA antigens in the form of platelet concentrates, and as early as four days in two patients with prior exposure to HLA antigens. In the most heavily immunized patients, the presence of antibody correlated with poor increments of platelets after transfusion. Patients with prior exposure to HLA antigens are more likely to have antibodies resulting in poor platelet survival. On the other hand, 30 per cent of recipients of repeated platelet transfusions show no tendency to form cytotoxic antibodies.     

Lymphocytotoxins in sera from highly sensitized multiparous dialysis patients: antibody class, relationship with the HLA and with paternal antigens.

1. Sera from 11 highly sensitized multiparous dialysis patients were studied in order to define the target antigens, antibody class and relationship with paternal HLA class I antigens of the underlying lymphocytotoxic antibodies. All sera contained lymphocytotoxic antibodies to over 70% of a panel of lymphocytes from 24 donors (panel reactivity greater than 70%). 2. Inhibition of cytotoxic activity against paternal lymphocytes by monoclonal antibodies to HLA framework determinants indicated that all 11 sera contained lymphocytotoxic antibodies to paternal class I antigens. In addition, five sera contained lymphocytotoxic antibodies to paternal class II antigens. 3. In order to determine the extent to which lymphocytotoxic antibodies were directed to paternal antigens, the panel reactivity of sera was compared before and after absorption with paternal peripheral blood lymphocytes. Over 50% of panel reactivity was absorbed from eight out of 11 sera, and in three of these 11 over 80% was absorbed. In the majority of patients this change in panel reactivity could be ascribed to binding of lymphocytotoxic antibodies to specific paternal class I antigens. 4. Digestion of sera with dithiothreitol had no significant effect on panel reactivity, indicating that the lymphocytotoxic antibodies were of immunoglobulin G class. 5. No sera reacted with either autologous lymphocytes or K562 cells, indicating an absence of autoantibodies. 6. These studies imply that panel-reactive lymphocytotoxic antibodies in the sera of highly sensitized multiparous patients are those which mediate hyperacute renal allograft rejection. Their development may be related to secondary humoral responses to antigens in blood transfusions from donors who share paternal class I specificities.     

The detection of platelet alloantibodies by flow cytometry

A flow cytometric procedure was investigated for its ability to detect antibodies directed against blood group A, HLA, and PlA1 (HPA-1a) antigens. When type O sera were tested against platelets from blood group A donors, only 9 of 14 positive reactions were observed. Furthermore, the expression of blood group A varied more than 100-fold on platelets derived from individual donors. When anti-HLA-A2 and -B7 were evaluated, 11 of 11 individuals with HLA-A2 and -B7 antigens reacted. In contrast, when platelets from donors whose HLA antigens included HLA-B8 or -B12 were tested with anti-HLA-B8 or -HLA-B12, respectively, positive reactions were observed in only 3 of 7 instances, despite the fact that the lymphocytes reacted strongly. Platelets from 10 HLA-A2-positive donors, which had been stored for up to 20 months at -70 degrees C, were studied. In all cases, frozen-stored platelets reacted well with an anti-HLA-A2. Limited testing with an anti-PlA1 (anti-HPA-1a) showed equal reactivity with fresh and frozen platelets. Finally, the method was compared to a visual immunofluorescence assay using sera from patients who were refractory to platelet transfusions. The results agreed in 30 of 37 comparisons, and most discrepancies were resolved in favor of flow cytometry. It is concluded that flow cytometry is useful for detecting platelet alloantibodies and possibly for prospective platelet crossmatching, as HLA- and platelet-specific antibodies can be identified by using platelets stored frozen for several months.     

Guidelines for platelet transfusions

Recommendations for the optimal transfusion support of patients likely to receive repeated platelet transfusions. 1. Determine policy for prophylactic platelet support, and select the platelet count below which platelet transfusions will be used. 2. Consider using leucocyte depletion of red cell and platelet concentrates to prevent HLA alloimmunization from the outset. 3. Type patients for HLA-A and B antigens at an early stage. 4. Use random donor platelet concentrates for initial platelet support (either single or multiple donor, depending on availability). 5. If refractoriness occurs, determine whether clinical factors, which may be associated with non-immune consumption of platelets, are present and test the patient's serum for HLA antibodies. 6. Use HLA-matched platelet transfusions if HLA alloimmunization is the most likely cause of refractoriness. 7. If there is no improvement with HLA-matched transfusions, platelet crossmatching may identify the cause of the problem and help with the selection of compatible donors. 8. Discontinue prophylactic platelet support if a compatible donor cannot be found. Use platelet transfusions from random donors to control bleeding and increase the dose, if necessary.     

Selection of compatible platelet donors: a prospective evaluation of three cross-matching techniques

A prospective study was undertaken to assess the values of platelet aggregometry, lymphocytotoxicity, and mixed lymphocyte cultures in selecting compatible donors for patients refractory to random platelet transfusions. Donors were selected at random without regard to HLA types. Concurrent with each platelet transfusion, platelet aggregometry and lymphocytotoxicity were performed using patient serum and donor cells. The results were compared with HLA types, MLC, and increments in platelet counts. Forty-one transfusions were given to 21 patients; 27 were from related and 14 from unrelated donors. Platelet aggregometry was used successfully in selecting compatible donors in 37 cases (90%) with three false negative and one false positive results. Lymphocytotoxicity was useful in 73 per cent of cases with eight false negative and three false positive results. The response to platelet transfusions correlated poorly with HLA matches or MLC reactions. These data suggest platelet aggregometry and lymphocytotoxicity are useful cross matching techniques in selection of compatible platelet donors.     

HLA class I-eluted platelets as an alternative to HLA-matched platelets

BACKGROUND: Alloimmunized refractory thrombocytopenic patients often require HLA-matched platelet transfusions. As the HLA system is very polymorphic, sufficient HLA-matched donors are not available for every patient. STUDY DESIGN AND METHODS: In vitro elution techniques with citric acid incubation of platelets at pH 3.0 showed that platelets lose expression of HLA, whereas platelet-specific glycoproteins are preserved. This technique was modified for clinical use. Random-donor platelet concentrates were incubated with citric acid, subsequently washed, and transfused to two patients. RESULTS: Platelet-specific glycoproteins were unaffected, and HLA expression decreased generally to below 25 percent of the initial expression. One alloimmunized patient who was without compatible donors because of a rare HLA type underwent repeated transfusions with acid-treated platelets. In contrast to the results with random-donor platelet transfusions, posttransfusion increments up to 47 × 10(9) per L were obtained with acid-treated platelets, and profuse gastrointestinal bleeding was stopped, while multiple skin hemorrhages were resolved. No side effects were observed. A second patient developed a severe transfusion reaction without platelet increment after one transfusion with acid-treated platelets expressing 30 percent of the original HLA antigens. Further transfusions were not given. CONCLUSION: Standardization of the acid elution technique and validation of the technique in patients is necessary. The results suggest, however, that HLA-eluted platelets prepared under specified conditions may gain a place in platelet transfusion therapy.     

Selecting donors of platelets for refractory patients on the basis of HLA antibody specificity

BACKGROUND: Patients who are refractory to platelet transfusion as a result of HLA alloimmunization are generally given HLA-matched or crossmatched platelets. However, HLA-matched platelets that are matched at HLA-A and -B loci (A-matched) or those without any mismatched or cross-reactive antigens (BU-matched) are frequently unavailable. A disadvantage of crossmatching is that crossmatched platelets have a shelf life of only 5 days, so that crossmatch tests must be performed frequently for patients requiring long-term platelet transfusions. An alternative method is the selection of platelets according to the patient's HLA antibody specificity, called the antibody specificity prediction (ASP) method. STUDY DESIGN AND METHODS: An anti-human globulin-enhanced microlymphocytotoxicity test modified by a double addition of serum and a computer program were used to determine the specificity of patients' HLA antibodies. Platelet crossmatching was performed with a solid-phase adherence assay. The percentage of platelet recovery (PPR) was determined in 1621 platelet transfusions in an observational study in 114 patients, and the PPR of platelets selected by the ASP method was compared with the PPR of those that were HLA-matched, crossmatched, or randomly selected. The numbers of potential donors in files of HLA-typed donors as identified by HLA matching vs. the ASP method were determined. RESULTS: After adjustments for covariates, the mean +/- SEM PPR was similar for HLA-matched (21 +/-4%), cross-matched (23+/-4%), and ASP-selected (24+/-3%) platelets and was significantly lower for randomly selected (15+/-1.4%) platelets. For 29 alloimmunized HLA-typed patients, the mean number of potential donors found in a file of 7247 HLA-typed donors was 6 who were an HLA-A match (median = 1), 33 who were an HLA-BU match (median = 20), and 1426 who were identified by the ASP method (median = 1365). CONCLUSION: The ASP method of donor selection for refractory alloimmunized patients appears as effective as HLA matching or crossmatching. Far more donors are identified in a file of HLA-typed donors by the ASP method than by HLA matching, and this indicates that the ASP method provides important advantages regarding the availability of compatible platelet components.     

Donor selection for alloimmunized patients by platelet crossmatching of random-donor platelet concentrates

Samples obtained from bag segments of stored platelet concentrates (PCs) were crossmatched with sera from alloimmunized patients to identify potentially compatible donors. Sera from 41 patients who were refractory to random-donor platelet transfusions were tested. All sera had lymphocytotoxic antibody (LCTAb), and 37 patients had LCTAb that reacted strongly against greater than 75 percent of the cells in the panel. At least 18 units of PCs were tested per patient. Sera from 33 patients were tested by microenzyme-linked immunosorbent assay (microELISA). Crossmatches with 703 (74 percent) of 944 PCs were positive, 23 percent were negative, 3 percent were indeterminant. Eight additional patients were also tested with a solid-phase system; 352 (85 percent) of 417 crossmatches were positive and 15 percent were negative. Patients with more reactive LCTAb had very few (often 0-2) in vitro compatible units. Thirty-four transfusions of pooled compatible PCs, all stored less than 3 days, were administered to clinically evaluable patients. Six of 20 transfusions selected by microELISA produced satisfactory increases (corrected count increments [CCIs] greater than 7500) 10 minutes after transfusion in 4 of 14 patients. Eight of 14 transfusions selected by solid-phase produced CCIs greater than 7500 in 5 of 8 patients. Although it is difficult to select large numbers of compatible units for heavily alloimmunized patients by these techniques, some transfusions were successful in patients who had very few HLA-compatible donors. This approach may be of particular value in emergency situations.     

Platelets do not adsorb HLA class I molecules during storage of pooled platelet concentrates

The origin of HLA class I molecules on platelets is still under discussion. Adsorption of HLA molecules on platelets using specific experimental conditions has been described. The study presented investigates whether there is a significant elution and adsorption of HLA class I molecules on platelets during storage of pooled random platelet concentrates (PRPC) under routine conditions. Platelet concentrates (PCs) from whole blood were prepared from HLA-A2-positive and HLA-A2-negative donors, pooled and stored under routine conditions. In addition, platelets from HLA-A2-negative donors were pelleted and resuspended in cell-free plasma from HLA-A2-positive donors. HLA-A2-positive PCs (positive control), HLA-A2-negative PCs (negative control) and HLA-A2-negative platelets in plasma from HLA-A2-negative donors were stored simultaneously. Binding of FITC-conjugated monoclonal murine antihuman HLA-A2 antibodies (anti-HLA-A2-mab) was measured during 5-day storage by flow cytometry. An increased binding of anti-HLA-A2-mab during storage was found on HLA-A2-negative platelets (P < 0.005) independently whether they were incubated with cell-free plasma or platelets from HLA-A2-positive donors or autologous HLA-A2-negative cell-free plasma. However, non-specific binding of IgG controls increased equally, whereas anti-HLA-A2-mab binding to platelets from HLA-A2-positive donors did not decrease during storage. This study suggests that there is no significant elution and adsorption of HLA class I antigens of platelets in pooled PCs during storage under the usual conditions for platelet storage. Increased anti-HLA-A2-mab signal was due to non-specific binding. Therefore, HLA class I compatible platelets should maintain their compatibility for an immunized patient when stored in a pool with HLA incompatible platelets and shortened survival after transfusion should not be expected.     

Clinical importance of platelet antigens

HLA antibodies are the most important cause for immunological refractorines to platelet transfusions, and their formation can be avoided in the majority of patients by exclusive use of leukocyte filtered blood products with a residual white cell count < 1 × 10⁶ per unit.Chemotherapy reduces the risk of platelet antibody formation; patients under cytostatic treatment for malignancies often demonstrate transient immunization and they can also lose antibodies present at start of the treatment.In the management of patients with immunologic refractoriness, treatment should first be attempted with HLA compatible platelets from single donors. If a transfusion effect is not obtained, cross match (lymphocytotoxicity test + platelet radioactive antiglobulin test) negative platelets from single donors should be attempted. As the next step related donors can be evaluated, and finally treatment with plasma apheresis or high dose intravenous IgG therapy should be considered.     

Serious adverse events after HPA-incompatible platelet transfusions in an alloimmunized patient with leukaemia

We describe an alloimmunized female patient who developed serious adverse reactions when receiving HPA-incompatible platelet transfusions. She had received 13 transfusions with random platelets before the first allergic reactions. Antibodies against both the human leucocyte antigens (HLA) and several human platelet antigens (HPA) were detected at the time of transfusions. When the patient received HLA- and HPA-compatible platelets, no reactions followed the transfusions and platelet count increments were good. When she was transfused with platelets from donors with one foreign HLA antigen, her reactions were fever, chills and headache and the response to platelet transfusions was poor. When the platelets were HLA compatible but HPA incompatible, the reactions were repeatedly rapid pulse, shortness of breath, tightness of chest and wheezing interpretable as anaphylactoid reactions. Platelet count increments were satisfactory. When rare side-effects occur after transfusion, detailed immunohaematological studies are indicated.     

Detection of HLA antibodies by platelet crossmatching techniques

Thirty-seven monospecific HLA antibodies directed against all common HLA-A and -B loci and reactive by the microlymphocytotoxicity assay (LCT) were tested against platelets carrying the corresponding antigen by three platelet crossmatch methods, the platelet immunofluorescence test (PIFT), platelet enzyme-linked immunosorbent assay (P-ELISA), and platelet radioimmunoassay (P-RIA). Positive reactions were obtained with the PIFT in 67 percent, the P-ELISA in 41 percent, and the P-RIA in 49 percent of 85 cell-serum pairs. The same cell-serum combinations gave 49 percent positive reactions in the lymphocyte immunofluorescence test. Three multispecific HLA antisera were positive in nine of nine cell-serum combinations by all four methods. Thirteen transfusions were given to eight patients with known HLA antibodies. All donor-recipient pairs were LCT positive, six were PIFT positive, and seven were PIFT negative. Three of seven PIFT-negative and none of six PIFT-positive transfusions were successful. Thus, platelet crossmatching is less sensitive than the LCT for the detection of complement-binding monospecific HLA antibodies. The platelet crossmatch, however, is able to identify some potentially successful HLA-incompatible donors for patients with multispecific HLA antibodies and limited access to HLA-identical donors.     

Activation of human platelets by antibodies to thymocytes and beta 2-microglobulin. III. Effect of cell absorptions on the platelet aggregating and lymphocytotoxic activities of HATG and SA beta 2mG

Horse anti-human thymocyte globulin (HATG) and sheep anti-human beta-2-microglobulin globulin (SA beta 2-mG), both known to have immunosuppressive potency, were quantitatively absorbed by a number of human cells, then, tested for platelet aggregating and lymphocytotoxic reactivities. Peripheral blood lymphocytes, lymphoblastoid T- and B-cell lines, platelets and spermatozoa could remove varying amounts of both reactivities from HATG and SA beta 2mG. Daudi cells were effective in reducing the lymphocyte and some platelet reactivity of HATG but did not affect those of SA beta 2mG. Erythrocytes had no absorption capacity for either antibody. The qualitative and quantitative differences in absorbing with various cells on the platelet aggregating and lymphotoxic activities of HATG and SA beta 2mG indicate the high specificity and direct action of the antibodies on platelets and lymphocytes.     

Flow cytometric platelet cross-matching to predict platelet transfusion in acute leukemia

A great variety of patient- and product-related factors influence the outcome of platelet transfusions. Our study assessed the predictive value of a flow cytometric platelet cross match test for the outcome of HLA matched and unmatched platelet transfusions in patients with acute leukemia. Thirty nine patients (26 adults and 13 children) received 60 ABO compatible apheresis platelet unites ranging from 1 to 4 per patient (mean = 1.54; median = 2). We performed flowcytometric platelet cross-matching, HLA Class I typing by sequence-specific primer (SSP) for patients and complement-dependent cytotoxicity (CDC) for donors and screening of HLA Class I antibodies by ELISA. Effectiveness of platelet transfusion was evaluated using the corrected count increment which was calculated at 60 min and 18- to 24-h posttransfusion. Multivariate analysis was performed to detect which variable can predict transfusion response more than others. Cross-matched platelet transfusions associated with good response in 51.4% of transfusion events in adults and 73.3% in children. The noncrossmatched platelet transfusions associated with poor response in 83.3% in adults and 100% in children (P-values 0.143, 0.041, respectively). In the presence of clinical factors or HLA alloimmunization in adults, cross-matched platelets were associated with good response in 29.6 and 22.2% respectively. In children this occurred in 81.8 and 66.7%, respectively. In presence or absence of HLA matching, flow cytometry platelet cross-matching was the most predictor for transfusion response (P = 0.05). Because of the difficulties to find frequent HLA matched donors for acute leukemia patients; flow cytometric platelet cross-matching may provide the most useful way for selecting donors. It is useful even in the presence of alloimunization in children.     

Mechanism of Transfusion-induced HLA alloimmunization

<正>Platelet transfusion therapy has allowed successful prevention and treatment of hemorrhagic complications associated with thrombocytopenia since its first use in the 1960's.Consequently, the use of platelets has increased at an exponential rate.Since platelet     

血小板输注无效患者体内血小板反应性抗体特异性调查研究

目的探讨血小板输注无效患者体内血小板反应性抗体的发生频率及其特异性。方法对48名血小板输注无效患者,利用MACE法筛查其血清中的血小板反应性抗体,进一步用PAK12诊断试剂MAIPA法鉴定血小板HPA抗体特异性。结果血小板输注无效患者中血小板反应性抗体的总检出率为50%(24/48),其中同种HLA抗体发生率为39.6%(19/48),占所有抗体检出的79.2%,同种HPA抗体的检出率为29.2%(14/48),而其中有64.3%(9/14)是与HLA抗体同时存在。检出的HPA抗体特异性有抗-HPA-3a(n=2),-5a(n=1),-5b(n=1),-2b(n=1),-4b(n=1);同时HPA抗体中有78.6%(11/14)是针对GPⅡb/Ⅲa,Ⅰa/Ⅱa和/或Ⅰb/Ⅸ的多反应性抗体。女性患者中检出抗体的比例55.2%(16/29)高于男性患者42.1%(8/19),但差异没有显著性统计学意义。结论血小板输注无效患者中血小板同种抗体以HLA抗体多见,但HPA抗体并不像之前报道的罕见。与高加索人群PTR患者血小板同种反应性抗体以抗-HPA-5b,-1b为主不同,研究发现中国人群PTR患者HPA同种抗体以HPA-3,-5系统抗体多见,同时检出了HPA-4b,-2b抗体。     

Establishment of platelet donor registry improves the treatment of platelet transfusion refractoriness in Guangzhou region of China

Platelet transfusion refractoriness (PTR) is the major complication of long‐term platelet supportive care. To improve the effectiveness of platelet transfusion therapy in PTR patients, we aimed to establish a platelet donor registry in our region (Guangzhou, China) by typing the human leukocyte antigen (HLA) and human platelet antigen (HPA). Blood donors (n = 864) from our population were genotyped for HLA‐A, HLA‐B and HPA systems by polymerase chain reaction amplification with sequence‐specific primer(PCR‐SSP) techniques. Using this cohort, we compared the results of platelet transfusions (matched vs. random) in 23 patients with PTR. Matched platelets were selected either by HLA antigen matching or by HLA antibody matching, as predicted by antibody specificity prediction (ASP) analysis. Significantly higher platelet recovery (PPR) values were obtained with HLA‐matched platelets in comparison with random platelets. No significant difference in PPR was observed between HLA matching and ASP methods. In two patients, platelet‐specific alloantibodies (alloabs) (anti‐HPA‐3b and anti‐HPA‐5b) were detected besides HLA class I alloabs. Transfusion with HLA‐ and HPA‐compatible platelets in both the patients resulted in significantly higher PPR when compared with HLA‐compatible platelet transfusion alone. In this study, we demonstrated that the establishment of an HLA‐ and HPA‐typed platelet aphaeresis donor registry is useful to improve the treatment outcome of PTR patients and to maintain a long‐term platelet transfusion strategy.     

Platelet-specific antibodies in HLA-immunized patients receiving chronic platelet support

BACKGROUND: In HLA-alloimmunized patients, the unexpected failure of HLA-matched platelet transfusions usually raises the suspicion about concomitant platelet-specific antibodies. As the reported frequency of platelet-specific antibodies in multitransfused patients varies widely, the aim of this study was to determine the prevalence of such antibodies in a population of chronic thrombocytopenic patients with HLA antibodies. STUDY DESIGN AND METHODS: From 1985 to 1997, 11,777 determinations of HLA antibodies were performed in 1330 hematologic patients receiving chronic platelet support. Fifty-two patients with HLA alloimmunization that lasted more than 1 month were selected. The search for platelet-specific antibodies was performed by using a monoclonal antibody immobilization of platelet antigens assay, thus allowing the identification of platelet-specific antibodies directed against the platelet glycoproteins (GP) Ib/IX, GPIIb/IIIa, and GPIa/IIa. Specificity of the platelet-specific antibodies was further investigated by using a solid-phase assay with chloroquine-treated platelets. RESULTS: Only 2 (3.8%) of the 52 patients had platelet-specific antibodies. One antibody reacted with an epitope of the GPIIb/IIIa that was present in all the panel platelets, and that probably was an autoantibody. The other was an anti-HPA-5b. CONCLUSIONS: The prevalence of platelet-specific antibodies in patients with HLA alloimmunization is very small. The search for concomitant platelet-specific antibodies would be indicated only when other causes of refractoriness to HLA-matched platelets are ruled out.     

Statewide support of thrombocytopenic patients with ABO matched single donor platelets

Use of ABO matched, HLA nonmatched platelet units obtained from single donors by pheresis, using the Haemonetics 30 Cell Separator, has a lowered hepatitis risk, and possibly delays the onset of the refractory state. Pheresis also offers a method of obtaining HLA matched platelets for the already refractory patient. A mean of 4.2 × 10(11) platelets are collected in about 1 1/2 hours. ABO matched, HLA nonmatched platelets produced corrected increments of more than 2500/mm3 per 7 × 10(10) platelets infused in 67 per cent of transfusions to nonselected recipients. The procedure is simple, safe, and requires little time. It is a worthwhile large-scale program for a regional blood center to undertake in order to provide optimal therapy for thrombocytopenic patients.     

Antibodies to private and public HLA class I epitopes in platelet recipients

BACKGROUND: Transfusions or pregnancies can cause immunization against private HLA determinants and public epitopes shared by more than one private HLA antigen. HLA antibodies are correlated with febrile transfusion reactions, lower platelet response following platelet transfusion, and an increased rate of renal transplant rejection. Until now, antibody specificities in alloantisera from platelet recipients have been poorly characterized. STUDY DESIGN AND METHODS: Consecutive serum screens from platelet recipients were analyzed for antibodies against private HLA class I antigens and public HLA epitopes using a serum analysis program based on the 2 x 2 table analysis of correlations. Serum screens of highly immunized patients and of patients with new alloimmunization events were reviewed separately. RESULTS: Of the serum screens from 566 platelet recipients, 1577 indicated alloimmunization (panel-reactive antibodies >5%). The program assigned a specificity in 1024 of these screens (64.9%) and at least once in 522 of 566 patients (92.2%). In 267 patients, antibodies detecting public epitopes in the combined A- or B-locus cross-reacting groups were found; other public markers were detected in 39 patients. Patterns of reactivity were remarkably less stable than in patient groups previously studied. In many patients, antibodies with apparent private epitope specificity preceded the identification of antibodies against a shared marker of the same cross-reactive group. However, the disappearance of antibodies (whether or not this was followed by a new antibody against a private or public marker belonging to another cross-reacting group) was also observed. CONCLUSION: The computerized analysis of microlymphocytotoxicity tests enhances the rate of antibody specification in sera from platelet recipients with lymphocytotoxic antibodies. The identified antibodies should be taken into account in the selection of platelet donors. The data confirm and extend previous observations on HLA class I antibodies and elucidate new alloimmunization events.