RESUMEN
To enhance selection of appropriate deceased donors for pancreas transplants, we sought to determine whether HLA matching improved posttransplantation outcomes. In this single-center study of 1219 pancreas transplants, we correlated posttransplantation outcomes with HLA-A, -B, -C, -DR, and -DQ matches and mismatches. Rejection was linearly correlated with the number of mismatches. The individual number of HLA mismatches reached significance at four or more with a 2.3- to 2.9-fold increase in rejection. The effect was most predominant with HLA-B (1.8-fold with one mismatch and 2.0-fold with two mismatches) and -DR (1.9-fold with two mismatches) loci, whereas HLA-A, -C, and -DQ matches or mismatches did not independently predict acute rejection. The affect was strongest in solitary pancreas transplants, with little impact for simultaneous pancreas and kidney (SPK). In contrast, HLA matching did not affect graft or patient survival rates but was associated with a reduced risk of opportunistic infection. Avoidance of acute rejection saved an estimated $32 000 for solitary pancreas recipients and $52 000 for SPK recipients in hospital costs. Our data do not support the use of HLA matching for predicting pancreas graft survival but do support its significance for the reduction of acute rejection, particularly for solitary pancreas recipients.