HLA-CLASS I AMBIGUITY RESOLUTION IN THE NATIONAL MARROW DONOR PROGRAM® DONOR/RECIPIENT PAIR PROJECT: WHEN IS ENOUGH ENOUGH?

4.1
#4-OR
HLA-CLASS I AMBIGUITY RESOLUTION IN THE NATIONAL MARROW DONOR PROGRAM® DONOR/RECIPIENT PAIR PROJECT: WHEN IS ENOUGH ENOUGH?
S. Spellman , R. Cullen , M. Maiers , M. Fernandez-Vina Ph.D. , T. M. Williams, M.D. , L. Baxter-Lowe Ph.D. and C. K. Hurley, Ph.D. . Minneapolis MN, National Marrow Donor Program, 55413, Coordinating Center ; Kensington MD, C.W. Bill Young Marrow Donor Program, 20895, Oncology ; Albuquerque NM, University of New Mexico School of Medicine, 87131, Pathology ; San Francisco CA, University of Califorinia, San Francisco, 94143, Surgery and Washington DC, Georgetown University Medical Center, 2007, Oncology .

Allele level DNA based typing of the class I loci (HLA-A, B, C) is complicated by allele combinations that yield multiple possible pairs of alleles that cannot be distinguished without additional testing increasing the time and expense for typing. Availability of a reliable validated source of allele frequency and ambiguity resolution data could allow assignment of alleles with relative confidence without additional testing. The NMDP Donor/Recipient Pair Project has retrospectively typed >4500 stem cell donors for HLA-A, B, C at the allele level resolving all known ambiguities. All data were validated and quality controlled. The relative frequency for each ambiguous combination of alleles was determined and a threshold set for discontinuing ambiguity resolution when the likelihood of one pair of alleles was at least 20,000-fold the likelihood of a second possible pair of alleles. An example is HLA-B*0702/0801 vs. *0705/0807. The estimated frequencies of B*0702/0801 and B*0705/0807 are 0.014 and 3x10-6, respectively, with a ratio of 46,667:1. Thus, the typing B*0702/0801 can be assigned with little risk of error. This approach will reduce the proportion of typings requiring ambiguity resolution from 29%, 33% and 4% to 4.6%, 9.9% and 0.7% for the HLA-A, B, C loci, respectively. Although the dataset is likely biased towards common HLA types, the described ambiguity resolution analysis is a valuable resource for laboratories performing allele level typing on similar populations in low risk situations. Statistical analysis of typing ambiguities can reduce the time and expense of allele level tests without sacrificing accuracy and consistency.