HCT Donor Options for Patients without Matched Donors- Impact of Patient HLA and Genetic Ancestry in the NMDP-Sponsored, CIBMTR-Led Access Trial

The ACCESS trial studied safety and efficacy of peripheral blood allografts from mismatched unrelated donors (MMUD) in adults with hematologic malignancies receiving combination post-transplant cyclophosphamide , mycophenolate mofetil, and tacrolimus as graft-vs-host disease prophylaxis. As over half (51%) of enrolled patients were ethnically diverse (ED), we aimed to investigate trends in participants’ HLA and HLA-defined genetic ancestry and potential impact on donor selection . ACCESS required high-resolution matching at HLA-A, B, C, and DRB1 ranging from 4-7/8 MMUD (18-35y) and recipients without a suitable 8/8 related or matched unrelated donor (MUD) who lacked HLA-specific antibody (AB) to any MM allele/antigen with mean fluorescence intensity >3000. Search Prognosis (SP) scores were calculated based on participant HLA and self-identified race and ethnicity (SIRE). SP scores were classified as “Good,” “Fair,” or “Poor” based on SIRE-specific HLA genotype frequency (GF) ranges. Genetic ancestry was inferred from participants’ and donors’ HLA genotypes using population haplotype frequencies for U.S. race groups. The most likely race values were assigned. Among 268 adult participants 25% were Hispanic (HIS n=66), 13% African American (AFA n=35), 8% Asian or Pacific Islander (API n=22), 4% Multiracial (MLT n=11), and 1% Native American (NAM n=3). Most participants received HCT using 7/8 MMUD (68%, n=183). A larger proportion of AFA (57%) and MLT (45%) participants received HCT using ≤6/8 MMUD vs. other SIRE groups. Ten participants (4%) had Good SP while most had Fair (59%) or Poor SP (37%) ( Figure 1 ). Some participants with Good SP had at least one potential 8/8 MUD at the time of search, but eventually received MMUD HCT due to presence of HLA-DP and/or HLA-DQ ABs, older donor age, donor unavailability, or other exclusion criteria. Of the single SIRE reports, 80% (n=99 of 124) and 75% (n=199 of 265) of the donor and participant imputed races respectively aligned with SIRE while 65% (n=95 of 147) of the imputed races for donors with UNK or MLT SIRE aligned with NHW. About 16% of participants reporting NHW had an imputed race of HIS, AFA or API while 21% of participants reporting HIS had an imputed NHW race. Similar cases were observed in other groups indicating complex participant genetic admixtures beyond reported SIRE, which can impact their HLA GF and odds of finding a MUD. Overall, 184 (68%) participants in the ACCESS study had an HLA GF less than 1 in 40 million, reflecting a low likelihood of identifying a MUD from the global donor pool of 40 million donors. Using MMUD expands access to HCT as participants enrolled in the ACCESS study would not have found a MUD for multiple reasons, including rare HLA GF, complex genetic admixture and AB reactivity. Tools like SP provide decision support for timely donor selection for patients with lower odds of finding MUDs.