AUTHORS
Sunmin Park, MD PhD, Michael Carter2, Ni-Chun Tsai3, Dongyun Yang, PhD4, Ping Rao, PhD5, Idoroenyi Amanam, MD6, Nicole Karras, MD7, Geoffrey Shouse, DO PhD8, Haris Ali, MD9, Ketevan Gendzekhadze, PhD10, Monzr M. M Al Malki, MD11
AFFILIATIONS
1 City of Hope National Medical Center, Duarte, California, United States
2 City of Hope National Medical Center, Duarte, California, United States
3 City of Hope National Medical Center, Duarte, California, United States
4 City of Hope National Medical Center, Duarte, California, United States
5 City of Hope National Medical Center, Duarte, California, United States
6 City of Hope National Medical Center, Duarte, California, United States
7 City of Hope National Medical Center, Duarte, California, United States
8 City of Hope National Medical Center, Duarte, California, United States
9 City of Hope National Medical Center, Duarte, California, United States
10 City of Hope National Medical Center, Duarte, California, United States
11 City of Hope National Medical Center, Duarte, California, United States
ABSTRACT
Background
Post-transplant cyclophosphamide (PTCy) has improved the outcomes of patients undergoing HLA mismatched unrelated donors (MMUD) for allogeneic hematopoietic cell transplantation (HCT). However, conventional allele-level HLA matching may not fully capture the biologic heterogeneity of HLA disparity. Eplets are polymorphic amino acid configurations on HLA molecules that provide a more granular assessment of HLA disparity than allele matching alone. We evaluated eplet mismatch (EpMM) models to determine whether molecular HLA mismatch was associated with HCT outcomes.
Methods
We performed a single-center, retrospective study of patients with hematologic malignancies undergoing first ≤7/8 MMUD HCT with PTCy-based GVHD prophylaxis at City of Hope between 2016-2025. EpMM was calculated using version 2 Eplet Mismatch worksheets obtained (PMID: 27965660). EpMM was analyzed by median cutoffs. Directionality was defined by the presence of unique eplets present in the recipient but absent in the donor (GH) or present in the donor but absent in the recipient (HG). Outcomes included overall survival (OS), relapse-free survival (RFS), non-relapse mortality (NRM), cumulative incidence of relapse (CIR), graft-versus-host disease-free relapse-free survival (GRFS), and acute and chronic GVHD. OS, RFS, and GRFS were compared using log-rank test and adjusted using multivariable Cox proportional hazards models. Competing risk outcomes were analyzed using the Fine-Gray method.
Results
A total of 266 patients underwent first ≤7/8 MMUD HCT with PTCy (7/8: N=237, 6/8: N=25, 5/8: N=4). Median recipient age was 57 years (11–79), donor age 28 years (18–55), and median follow-up among survivors was 24.9 months. Diagnoses included AML (35.7%), ALL (25.6%), MDS (16.5%), lymphoma (7.9%), and other hematologic malignancies (14.3%). Peripheral blood stem cells were used in 98.5% of transplants and 38.7% received myeloablative conditioning. Majority received PTCy with mycophenolate with either tacrolimus (n= 236) or sirolimus (n= 20).
Allele mismatching rates were: 41.7% (HLA-A), 27.8% (HLA-B), 15.8% (HLA-C), 25.9% (HLA-DRB1), 13.9% (HLA-DQB1), and 79.3% (HLA-DPB1: 48.1% TCE-permissive and 31.2% TCE-nonpermissive). Fifty-five patients (20.7%) were double mismatched for DPB1. Class I and class II mismatches were present in 212 (79.7%) and 225 (84.6%), respectively.
Median class I and class II EpMM were 2 (range, 0-12) and 4 (range 0-63), respectively. Class I and class II GH EpMM and class II HG EpMM were not associated with outcomes; however, class I HG EpMM above median was associated with improved OS, GRFS, and RFS. We therefore evaluated the interaction between class I and class II HG EpMM. Patients were classified into three biologically defined groups: no HG EpMM (class I=0 and class II=0; n=22), high HG EpMM (class I>2 and class II>4, both above the cohort medians; n=50), and low HG EpMM (all remaining patients; n=194).
High HG EpMM was associated with significantly improved OS, RFS, and GRFS (p=0.0071, p=0.00046, p=0.0017, respectively), driven by reduced relapse (p=0.035) and a trend toward lower NRM (p=0.084). Grade III-IV acute GVHD and moderate-severe chronic GVHD did not differ significantly among groups. In exploratory subgroup analyses, HG EpMM was associated with RFS among recipients of reduced-intensity/non-myeloablative conditioning (p=0.0038). After adjustment for KPS, HCT-CI, DRI, recipient and donor ages, disease, conditioning intensity, high HG EpMM remained independently associated with improved OS (HR 0.42, 95% CI 0.21-0.84, p=0.015) and RFS (HR 0.37, 95% CI 0.19-0.75, p=0.005). Similar results were observed after restricting the analysis to 7/8 MMUD recipients.
Conclusions
Eplet mismatch modeling is feasible in HLA-mismatched unrelated donor (≤7/8 MMUD) HCT with PTCy. Neither total EpMM, class II EpMM, nor GH-direction EpMM predicted transplantation outcomes. In contrast, concurrent elevation of both class I and class II in HG EpMM identified a subgroup with reduced relapse and improved RFS, GRFS, and OS without increased GVHD. These findings suggest that the prognostic impact of molecular HLA disparity in the PTCy setting depends on the interaction between class I and class II HG EpMM rather than the overall EpMM burden. Validation in independent cohorts and mechanistic studies is warranted.
DISCLOSURES
Monzr M. M Al Malki, MD: incyte - Research Funding; gamida cell - Consultancy (Includes expert testimony); Ossium Health - Consultancy (Includes expert testimony); CareDx - Consultancy (Includes expert testimony); Gilead - Consultancy (Includes expert testimony); Jazz Pharma - Consultancy (Includes expert testimony); Tr1x - Consultancy (Includes expert testimony); Tscan Therapeutics - Consultancy (Includes expert testimony)