Paper Details

ABSTRACT

Introduction:

The adoption of post-transplant cyclophosphamide (PTCy) based graft vs host disease (GVHD) prophylaxis has expanded the use of mismatched unrelated donors (MMUD) in allogeneic hematopoietic cell transplantation (HCT). A prior CIBMTR study noted equivalent outcomes between 8/8 matched unrelated donors (MUD) and 7/8 MMUD (considering matching at HLA-A, -B, -C and -DRB1) when PTCy is used (Shaffer et al., JCO 2024). HLA Class I and II MMUD recipients had similar outcomes, whereas a report from the EBMT (Arrieta-Bolaños et al., JCO 2024) reported worse outcomes with HLA class I mismatched donors in the PTCy setting. We therefore examined whether the mismatched HLA locus affects outcomes after 7/8 MMUD HCT with PTCy in a large, contemporary U.S. cohort to inform donor selection practices.

Methods:

We conducted a retrospective CIBMTR registry analysis of adults (≥18 years) undergoing first allogeneic HCT in the US from 2017-2024 using a 7/8 MMUD and PTCy-based GVHD prophylaxis plus tacrolimus or cyclosporine with or without mycophenolate mofetil. The primary objective was to compare overall survival (OS) and GVHD-free, relapse-free survival (GRFS) at 1 year post-HCT between MMUDs based on mismatched HLA locus. Secondary objectives included evaluating relapse, relapse-free survival (RFS), non-relapse mortality (NRM), acute GVHD (aGVHD) grades 2-4 and 3-4, moderate/severe chronic GVHD (cGVHD), and primary graft failure (PGF) by mismatched locus. Eligible diagnoses included acute myeloid leukemia (AML), acute lymphoblastic leukemia (ALL), and myelodysplastic syndromes (MDS). To adjust for multiple testing, false discovery rate (FDR) control was applied using the Benjamini-Hochberg procedure across the set of Type 3 omnibus p-values, one per outcome. Statistical significance was declared at an FDR-adjusted p-value threshold of 0.05.

Results:

Included were 1,834 patients from 123 centers with AML (n=1022), ALL (n=317), and MDS (n=495). Median (range) recipient and donor age was 60.2 (18.7-81.0) and 27.6 years (18.4-61.2), respectively. Fifty percent of recipients were female, 35.7% were ethnically diverse, and most received peripheral blood grafts (94.5%). The median follow-up of survivors in the most recent year of HCT included (2024, 30% of total cohort) was 12.4 months, and accordingly outcome assessments were limited to 12 months post-transplant. Most patients received non-myeloablative or reduced-intensity conditioning (61.7%) and had an HCT-CI ≥3 (50.8%). The distribution of single-locus mismatches was HLA-A (n=844), HLA-B (n=376), HLA-C (n=185), and HLA-DRB1 (n=429). HLA-DQB1 and -DPB1 T cell epitope (TCE) data were available for 99.6% and 97.8% of recipients, respectively. DQB1 mismatches were more common among DRB1-mismatched pairs (55.5%) versus HLA-A (6.6%), -B (8.2%), and -C (6.5%) MMUDs. HLA-DPB1 TCE matching was similar across groups (15.4-21.8% matched). Neither HLA-DQB1 mismatching nor -DPB1 TCE match status was associated with any clinical outcomes in multivariable models.

OS and GRFS at 12 months were not significantly different by locus of mismatch: HLA-A – OS 78.4% (75.4-81.0)/GRFS 46.2% (42.7-49.5), HLA-B – OS 79.6% (75.1-83.3)/GRFS 46.1% (40.9-51.2), HLA-C – OS 76.8% (69.9-82.3)/GRFS 42.9% (35.5-50.1), and HLA-DRB1 – OS 72.8% (68.3-76.8)/GRFS 47.6% (42.7-52.3). In multivariable analysis using the largest group (HLA-A) as the reference, no significant differences were observed in any outcomes: OS (overall FDR-adjusted p=0.291), GRFS (p=0.906), relapse (p=0.906), RFS (p=0.769), NRM (p=0.592), aGVHD 2-4 (p=0.275), aGVHD 3-4 (p=0.769), moderate-severe cGVHD (p=0.769), and PGF (p=0.275) across mismatched loci. The only donor factor that entered any of the multivariable models was younger donor age, which was associated with improved OS, GRFS, and NRM.

Conclusions:

In this large contemporary cohort of 7/8 MMUD HCT with PTCy-based GVHD prophylaxis, we did not observe significant differences in OS, GRFS, relapse, RFS, NRM, aGVHD 2-4, aGVHD 3-4, moderate/severe cGVHD, and PGF according to the mismatched HLA locus. Based on the current available cohort, these results support MMUD donor selection strategies that prioritize non-HLA donor characteristics, such as younger donor age, rather than the specific HLA mismatched locus considering HLA-A, -B, -C, and -DRB1 (-DQB1 and -DPB1). Further evaluation in larger cohorts is warranted.