Racial/ ethnic disparities in availability of volunteer unrelated donors for allogeneic transplantation

Despite the global unrelated donor (URD) registry size, the degree to which URD availability is a transplant barrier is not established. We evaluated the availability of 3,843 URDs requested for 455 diverse adult patients (predominantly with acute leukemia). URDs for non-Europeans were more likely to be domestic and had markedly lower Donor Readiness Scores. Of URDs requested for confirmatory HLA-typing (CT) alone (i.e. without simultaneous workup), 1,894/3,529 (54%) were available. Availability of domestic URDs was 45%. Donor Readiness Score was highly predictive of CT availability. Compared with Europeans (n=335), more non-European patients (n=120) had >10 URDs requested and <5 available. Of workup requests (after CT or CT-workup), <70% (604/889, 68%) were available. More non-Europeans had <2 URDs available. URD availability for CT was markedly worse for non-Europeans, with availabilities for African, non-Black Hispanic, and Asian patients of 150/458 (33%), 120/258 (47%) and 119/270 (44%), respectively, with further decrements in URD workup availability. Our data suggest the functional size of the URD pool is much smaller than appreciated, mandating major operational changes for transplant Centers and donor registries. Likelihood of donor availability should have a high priority in donor selection. Considering patient ancestry and URD Donor Readiness Scores, Centers should pursue, and registries permit, simultaneous pursuit of many URDs, and abandon futile searches. Patients should be informed about their likelihood of donor availability and alternative options. Finally, while registries should address high URD attrition and speed procurement, use of all HLA-disparate graft types is needed to facilitate timely transplantation for all.

An Optimized Search Prognosis Tool to Predict 8/8 HLA Allele-Matched Unrelated Donor Procurement

For patients in need of allogeneic transplantation who lack an HLA-identical sibling, an 8/8 HLA allele-matched unrelated donor (URD) is a standard alternative. However, delays in URD procurement can adversely impact patient care. Recipient genotype and search assessment (MSKv1.0)-based tools can predict search prognosis for many, but both tools have lower performance in non-European ancestry patients. Using the MSKv1.0 tool, we analyzed searches from 1530 potential allograft recipients (including 863 who underwent transplantation) with the aim of creating an optimized MSKv2.0 search prognosis tool that can classify a URD search as either Good or Poor with a high level of accuracy while also limiting an ambiguous Fair search prognosis regardless of patient ancestry. By MSKv2.0, the 8/8 URD search prognosis distribution was 57% Good, 21% Fair, and 22% Poor in Europeans and 15% Good, 21% Fair, and 63% Poor in non-Europeans. Importantly, compared to MSKv1.0, the likelihood of Fair categorization was reduced to <25% with comparable Fair rates (P = .847) in both European and non-European groups. Moreover, all patients with an MSKv2.0 Good prognosis had an 8/8 URD identified, and almost all of those who underwent transplantation had an 8/8 URD (Europeans, 99%; non-Europeans, 98%; P = .504). The MSKv2.0 tool also was highly accurate at classifying Poor searches, with <10% identifying an 8/8 URD, and almost all patients who underwent transplantation (Europeans, 95%; non-Europeans, 96%) receiving an alternative donor. Using preliminary search results, MSKv2.0 accurately classifies patients by likelihood of 8/8 URD procurement, greatly facilitating triage to 8/8 URD (Good prognosis) or alternative donor (Poor prognosis) transplantations.