Modeling alloantibody formation to high-incidence red blood cell antigens in immune responders using genotypic data

Alloimmunization to red blood cell antigens is unpredictable and poorly understood. Patients who are negative for high-incidence antigens (HIAs) are at risk for developing the corresponding antibodies. Molecular methods can easily predict the lack of an antigen and thus, the risk of an individual to become immunized. We examined the prevalence and risk factors for HIA alloimmunization in patients at risk based on genotyping results. Genotyping using a molecular method (HEA BeadChip™, Immucor, Warren, NJ) was performed on all patient specimens referred for molecular testing over 45 months; serologic and clinical data were analyzed. We used simple and multiple logistic regression to model the risk factors for alloimmunization to an HIA. Of the 2591 patients genotyped, 32 (1.2%) were homozygous for at least one variant predicting absence of an HIA. Of these 32 patients, prior transfusion or pregnancy history was available for 29 (91%). Four susceptible patients made an antibody to an HIA (12.5% of all, 13.8% of those with a documented exposure). Two of these four patients (50%) had made an alloantibody to another antigen. The odds of forming an antibody to an HIA were not related to the total number of transfusions (p = 0.47), the total number of alloantibodies (p = 0.61), or diagnosis of sickle cell disease (p = 0.77) in simple logistic regression. Adjustment for the other two variables in a multiple logistic regression was also not significant for each variable (p = 0.6, p = 0.7, and p = 0.7, respectively). Although they had a known exposure to alloantigens through transfusion or pregnancy, 86.2 percent of patients (25 of 29) at risk for alloantibody formation to an HIA in fact did not mount an immune response to that antigen. Possible risk factors including the number of transfused units or the total number of alloantibodies made were not predictors of making an alloantibody to an HIA in our sampling. Our results suggest that other patient-specific risk factors for alloimmunization exist.