Abstract

Aim The genes encoding Human Leukocyte Antigens (HLA) have diversified significantly within individual ethnic groups, and individuals are most likely to find a suitable HLA matched donor from a population with common ancestry. Consequently, registry self-reliance in an ethnically diverse nation such as Canada represents significant challenges compared to an ethnically homogeneous country. Since its inception, the Canadian Blood Services’ (CBS) OneMatch (OM) Stem Cell and Marrow Network, and the newly created cord blood bank has strived to create an ethnically diverse donor and cord blood repository to reflect this equally diverse population. However, HLA genetics and HLA match rates for the populations served by the OM registry have not yet been well characterized. Methods HLA A∼C∼B∼DRB1 haplotype frequencies were calculated from 141,405 OM donors typed by DNA methods inclusive of 2011, resolving both allelic and phase ambiguity using the Expectation Maximization algorithm. Based on frequencies from 13 Canadian sub-populations, HLA match rates were estimated for subsequent years based on OM registry size increase projections. Matching models considered availability for adult donors. The haplotype frequency estimation and match rate projection tools were developed by NMDP Bioinformatics Research and previously applied to the Be The Match® (USA) registry. Results 8/8 HLA Allele-level match rates for patients searching for adult donors ranged from 2% for Blacks to 46% for Caucasians. This was improved to 12% for Blacks to 86% for Caucasians when 7/8 match rates were included. Projections indicate that 8/8 and 7/8 match rates will improve by 1–2% per population per year through 2019. Interestingly, common haplotypes between Aboriginal populations differ between Eastern and Western provinces, and admixture is occurring between Caucasian and Aboriginal, Black, and Hispanic populations. Conclusions Continued recruitment of adult donors in Canada will improve HLA match rates at a steady pace. When analyzing indigenous populations, it was important to divide the population into geographical regions to avoid error in matching models due to population substructure.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call