Unexpected frequencies of HLA antibody specificities present in sera of multitransfused patients.

1. The resolution of single antigen beads is high enough to distinguish between a patient's own HLA phenotype and antibody specificity down to the allele level. 2. Mismatched antigens generally produce an antibody specific to that antigen. Therefore, one would naturally expect antibody specificity frequency to correlate with phenotype frequencies for any given population (e.g., rare phenotypes would rarely have an antibody generated against it). We called this the antigen mismatch model. 3. When testing the sera of 367 highly sensitized patients with single antigen beads, we found results which did not correlate with the antigen mismatch model. Antibodies specific for rare antigens occurred at surprisingly high levels within the study population. 4. We postulate that these high frequencies are due to mismatches on epitopes as defined by absorption/elution experiments (9) or NEpis. Thus, a mismatch on a common antigen could produce an antibody specific for a rare antigen as long as the two antigens share a NEpi (unique position/amino acid combination). 5. The probability of mismatching on 58 NEpis based on published phenotype frequencies was calculated. We called this the NEpi mismatch model. 6. Both the antigen mismatch model and the NEpi mismatch model were compared to observed antibody specificity frequencies in our study population. Overall, the NEpi mismatch model correlated with observed frequencies much more accurately than the traditional antigen mismatch model. 7. In conclusion, we present a probable explanation for higher-than-expected antibody specificity frequencies and propose that mismatching on NEpis has further potential clinical implications.