164-P: CAN WE PREDICT HLA-DQA1 ALLELE ASSIGNMENTS BASED ON PUBLISHED DRB1/DQB ASSOCIATION DATA?

Aim: We wanted to see how accurately we can predict HLA-DQA1 assignments by using NMDP published DRB1/DQB1/DQA1 associations. Methods: We selected a total of 116 samples that were previously typed. Seventy two of these samples were proficiency testing samples and the remaining forty four were clinical samples typed in our Laboratory using Life Technologies DQA1-SSP kits. Using the NMDP association chart, we assigned DQA1 alleles for all the samples based on the DRB1/DQB1 association. We then compared our assignments with the actual typing. Results: By counting homozygous alleles as one, a total of two hundred seventeen alleles were defined. Nine of these DRB1/DQB1/DQA1 associations were not listed on the NMDP chart possibly due to uncommon DRB1/DQB1 associations. Out of the nine, eight could be assigned correctly based on the published association with DRB1 allele group. The remaining one sample could not be assigned correctly. Of the two hundred eight remaining allele assignments two hundred six agreed at least at the allele group level. Conclusions: Our data demonstrates that DQA1 assignment can be inferred on the basis of DRB1/DQB1 association at least at an allele group level with an accuracy rate of 99%. While this may not be a justifiable approach to stem cell transplantation, it certainly has application in solid organ transplantation. [Copyright &y& Elsevier]