Modeling the optimal ethnic composition of an adult stem cell registry

Abstract For patients suffering from a blood related disease, a stem cell transplant represents the best, and sometimes the only, possible treatment. Registries have been created throughout the world to match patients with stem cell donors. Canada's adult registry, OneMatch, was formed to meet the needs of Canadian patients. However, only 20–30 percent of unrelated adult stem cell transplants in Canada are sourced from Canadian donors. Self-sufficiency has proven difficult for OneMatch, in part, because the Canadian registry is dwarfed by other international registries with similar donor populations. In this paper, we present a study to evaluate changes the Canadian registry designed to promote ethnic diversity while meeting the needs of the Canadian patient population. We formulate the composition problem as a linear optimization model and solve it using a combination of exact and heuristic methods for a registry of 1 M donors. We conclude that when registry size is constrained, there are advantages to increasing ethnic diversity over self-sufficiency. However, results show that some communities cannot be easily accommodated within an adult registry of a fixed size. Thus, our results highlight the need for stem cells derived from cord blood for hard to match populations.