Background: In highly sensitized patients (HSP) awaiting renal transplantation, accurate delineation of acceptable human leukocyte antigen (HLA) mismatches (AMM) aids identification of suitable crossmatch negative donors. Comparison of differences in polymorphic triplet amino acid sequences in antibody accessible regions of HLA may predict immunogenicity. We have examined the ability of the HLAMatchmaker computer algorithm to predict AMM determined by antibody screening using the full repertoire of single-antigen HLA-A and -B specificities.
Methods: The HLA types of 24 HSP awaiting kidney transplantation were analyzed using HLAMatchmaker to determine the number of triplet amino acid (TAA) mismatches for each of 64 mismatched HLA-A and -B specificities. Patient sera with the highest immunoglobulin (Ig)G HLA-specific antibody reactivity were tested against the 64 individual HLA-A and -B specificities using single-antigen HLA antibody detection beads. Logistic regression analysis was performed to determine the association between AMM and the number of TAA mismatches.
Results: There was a strong positive association between the number of TAA mismatches and the presence of HLA-specific antibody. HLA specificities with zero TAA mismatches were antibody positive in only 4 of 47 (9%) cases. A single TAA mismatches was sufficient to invoke an antibody response in 40 (41%) of 98 cases, increasing to 97 (87%) of 112 cases with 9 or more TAA mismatches. However, there was considerable heterogeneity between individual patients, and only 16 (67%) of the 24 HSP studied fitted the logistic regression model for TAA mismatches and HLA-specific antibody.
Conclusions: Identification of TAA mismatches using HLAMatchmaker is a helpful tool for predicting potential donors with an acceptable HLA mismatch in HSP.