Introduction: HLA alloimmunziation has been described in patients, including pediatric patients with sickle cell disease (SCD), who received only leukocyte-reduced red blood cell (RBC) transfusions. While HLA alloimmunization has significant implications in bone marrow transplant (BMT) with an HLA-disparate donor (donor-specific HLA antibodies increase the risk of graft rejection), the clinical significance of this alloimmunization in HLA-identical BMT is unclear. In particular, it is unknown whether HLA class I alloimmunization affects platelet transfusion support during BMT. Platelet transfusions are especially important in BMT for SCD because of the historic high incidence of intracranial bleeding and the subsequent adoption of a relatively high platelet transfusion threshold.

Methods: We tested for HLA class I alloimmunization using pre-BMT serum or plasma from 32 pediatric patients with SCD undergoing BMT with an HLA-identical sibling donor. HLA class I alloimmunization was evaluated using the FlowPRA® (panel reactive antibody) screening test which detects IgG antibodies to the majority of HLA class I antigens. Pre-BMT, patients had received leukocyte-reduced RBC transfusions. All patients undergoing BMT received leukocyte-reduced, irradiated, apheresis platelet transfusions to maintain a platelet count greater than 50 x 109/L. The number of platelet transfusions received in the first 45 days post-BMT was compared between those with a positive PRA versus those with a negative PRA using a two-sample t-test.

Results: Among the 32 patients studied, the mean age at BMT was 9.4 years (range 1-21 years). Pre-BMT, 40% had received greater than 10 RBC transfusions and 32% of the entire cohort was receiving chronic RBC transfusions. BMT conditioning regimens included: traditional myeloablative busulfan, cyclophosphamide, antithymocyte globulin (BU-CY-ATG) (n=10); dose-reduced BU-CY-ATG + fludarabine (flu) (n=16); reduced intensity conditioning (RIC) alemtuzumab, flu, melphalan +/- thiotepa (n=5); and RIC BU-flu-ATG (n=1). All patients received HLA-identical sibling donor bone marrow, with 3 patients receiving a combined bone marrow and cord blood graft. All patients engrafted with 100% survival at day +45. Six patients developed acute graft-versus-host disease (GVHD): 1 grade I, 3 grade II, 1 grade 3, 1 grade 4. No patient had an intracranial bleed or other serious bleeding.

From testing of the pre-BMT sample, 11/32 (34%) of patients had a positive HLA class I antibody screen. HLA class I alloimmunized patients received significantly more platelet transfusions than non-alloimmunized patients: mean 19.2 (SD 13.9) vs. 11.2 (SD 8.0) transfusions, p=0.049 (Figure). On univariate analysis, age, sex, conditioning regimen, and GVHD were each not significantly associated with the number of platelet transfusions (p>0.1). Only two patients developed hepatic sinusoidal obstruction syndrome (one HLA alloimmunized patient, one non-alloimmunized patient). Mean day of platelet engraftment was not significantly different between alloimmunized compared to non-alloimmunized patients: day +34.6 (SD 10.6) vs. day +30.2 (SD 11.3), p=0.28. Among HLA class I alloimmunized patients, the mean positive PRA was 19.7% (range 4-55%); there was no association with this PRA value and the number of platelet transfusions (r2=0.0064, p=0.81).

Conclusion: In pediatric HLA-identical sibling BMT for SCD, pre-BMT HLA class I alloimmunization (but not the absolute PRA value) appears to be associated with more platelet transfusions. It is unclear if this increased platelet transfusion requirement is due directly to HLA antibody immune-mediated clearance of transfused platelets or other inherent differences of HLA class I alloimmunized patients.


Qayed:Novartis: Consultancy.

Author notes


Asterisk with author names denotes non-ASH members.