Chronic red blood cell (RBC) transfusion therapy is the predominant treatment modality in children with sickle cell disease (SCD) at high risk of first or recurrent strokes. RBC alloimmunization develops in some patients receiving chronic transfusion therapy, due in part to genetic differences in the prevalence of blood group antigens between the patient population and the blood donor pool. Many children’s hospitals have developed designated donor or “buddy” programs to recruit African-American blood donors and assign them to specific SCD patients with matched phenotypes, particularly in the Rh and Kell antigen groups, to reduce the likelihood of RBC alloimmunization. However, the practical constraints of such programs may make it difficult to ensure that patients’ transfusions always come from designated donors. Moreover, it is unclear whether such programs result in a lower risk of RBC alloimmunization when compared to the use of non-designated-donor but phenotype-matched RBC units. We aimed to determine the proportion of transfusions from designated donors at our institution, hypothesizing that the development of new RBC alloantibodies is associated with a lower proportion of transfused units from designated donors.


This is a single-institution retrospective cohort study of 38 patients with SCD who received chronic transfusion therapy (manual exchange or erythrocytapheresis) for primary or secondary stroke prevention from 1/1/2008 through 12/31/2012. Patients on transfusion therapy for 6 or more months were included. Subjects were censored at last date of follow-up or date of hematopoietic stem cell transplant. The local designated donor program was started in 1999. Designated donors are selected to be ABO/RhD compatible and phenotype-matched to patients for the C, E, and K antigens. When units from designated donors are not available, compatible units phenotype-matched for C, E, and K are issued from general inventory. The number and percentage of units transfused from either designated or non-designated donors, and the identification of new RBC alloantibodies during the study period, were evaluated. The rates of alloimmunization were compared between patients who received a “high” (above the median) or “low” (below the median) proportion of designated donor units. Categorical variables were compared with Fisher’s exact test and medians with the Mann-Whitney U-test in SPSS version 21 (IBM, Armonk, NY). A p-value below 0.05 was statistically significant.


During the study period, 38 subjects (42% male) met all inclusion criteria. A median of 120 units (IQR 60-186) was transfused to each subject, and each subject received a median of 63% (IQR 45%-77%) of units from designated donors. Of the 38 subjects, 18 (47%) produced at least one newly identified RBC alloantibody during the study period. Among these 18 antibody producers, a total of 29 new alloantibodies were detected, with a range of 1-3 per subject. Ten of the newly identified alloantibodies were directed against C, D, E, or K. No statistically significant difference between antibody producers and non-producers was identified for total number of RBC units transfused (median 161 vs. 96, p = 0.067), number of units transfused from designated donors (median 107 vs. 49, p = 0.099), number of non-designated-donor, phenotype-matched units transfused from general inventory (median 38 vs. 26, p = 0.059), or proportion of units transfused from designated donors (median 68% vs. 49%, p = 0.28). Although there was a trend toward a higher incidence of alloimmunization in patients who received a high proportion of designated donor units (OR 2.4, CI 0.6-8.7), it was not statistically significant (p= 0.33).


Despite receiving phenotypically matched RBC units, almost half of the children with SCD on chronic transfusion therapy in this cohort developed new RBC alloantibodies during a five-year period. The number of units transfused from a designated donor did not significantly affect alloimmunization rate. One-third of the new alloantibodies were directed against antigens specifically matched for in the designated donor program. Patient-specific factors, such as genetic variation in the Rh locus, may be responsible for the risk of alloimmunization. Alternative matching strategies, such as genotypic matching of RBC donors and recipients, should be explored in prospective studies.


Jackups:Immucor: Consultancy.

Author notes


Asterisk with author names denotes non-ASH members.