REPORT FROM THE HLA CLASS-II TYPING BY PCR-SSP MULTICENTER STUDY

Results from 360 HLA-DR and -DQ 'low-resolution' typings with polymera se chain reaction sequence-specific primers (PCR-SSP), performed by ni ne laboratories, were analysed for their overall utility in routinely defining the HLA-DRI-DR18, DR51-DR53 and DQ1-DQ9 specificities in less than 2.5 h. Thirty EDTA blood samples and 10 DNA samples were distrib uted and analysed by each laboratory. DNA was extracted using a rapid bromide salt extraction protocol. Complete HLA-DR and -DQ typings were performed, three by three, on pre-aliquoted 96-tube PCR trays. When c ompared with reference typing, 351/360 (98%) correct DR typings were o btained, whereas 320/360 (89%) of the DQ phenotypes were correctly ass igned. The time for three complete KLA-DR and -DQ 'low-resolution' typ ings, including DNA extraction, ranged from 2.0 h to 2.3 h. Unfortunat ely, an unusually high level of PCR amplification failures was observe d (3%), probably due to diffusion and a significant volume loss from s ome of the pre-aliquoted primer mixes. Consequently, only 52% of the t ypings were without any amplification failure, and 0-2 amplification f ailures where found in 88% of the PCR-SSP typings performed. The numbe r of HLA-DR-DQ retypings needed was 7 and 8%, respectively, reflecting the low number of typings where allelic identification was directly a ffected by the relatively high level of amplification failures in this study. Thus, a 91-98% success rate of correctly identified HLA-DR and -DQ alleles could be maintained, even under suboptimal typing conditi ons.