ELISA methods detect HLA antibodies with variable sensitivity

The accuracy of human leukocyte antigen (HLA) antibody screening results is dependent on the technique employed. In this study, a total of 612 serum samples from patients awaiting kidney transplantation were tested by two different enzyme-linked immunosorbent assay (ELISA) methods: LAT-M(R) (One Lambda/BMT) and AbScreen(R) (Biotest). The results of these assays were identical for HLA class I and II antibodies in 524 cases (85.6%), and discrepant in the remaining 88 cases (42 class I, 24 class II, 22 class I and II). More specifically, class I results were in agreement in 26 positive and 522 negative cases and in disagreement in 64 cases. Class II results were the same in 50 positive and 516 negative cases and different in 46 cases. Retesting the samples using LAT 1288 (One Lambda/BMT) and considering previous HLA antibody test results and the history of immunising events, the sensitivity and specificity, respectively, of the two methods were determined as follows: class I: 89.6% and 97.5% (LAT-M) vs. 70.6% and 96.8% (AbScreen), class II: 96.5% and 98.4% (LAT-M) vs. 88.3% and 95.5% (AbScreen). These results indicate that a single ELISA does not invariably prove or exclude the presence of HLA antibodies, and that additional testing is required in some cases.