HLA-DPB1 Mismatches and the Impact on Unrelated Donor Transplant Outcomes with the Use of Post-Transplant Cyclophosphamide

Background

Disease progression, graft-versus-host disease (GvHD) and non-relapse mortality are the main causes of failure after allogeneic hematopoietic cell transplantation (HCT). There have been many studies focusing on the biological models of HLA-DPB1 mismatching, namely permissiveness according to T-cell epitope (TCE) groups to determine their association with the risks of non-relapse mortality (NRM) and GvHD/relapse. Most studies were conducted prior to the advent of post-transplant cyclophosphamide (PTCy) as standard GvHD prophylaxis for matched unrelated donor (MUD) transplants. Thus, in this analysis, we aimed to investigate the impact of HLA-DPB1 mismatching on MUD transplants with the use of PTCy as GvHD prophylaxis.

Methods

Our study cohort included 404 patients with myeloid malignancies who had HCT at our institution between 2011-2023 from an unrelated HLA-A, -B, -C,-DRB1 and DQ matched donor by high resolution typing (10/10 matched). The recipients had a median age of 64 (range, 18-77) while donor median age was 29 (range, 19-60). Disease risk index (DRI) at HCT was high or very high in 102 (25%) in the study population. 248 (61%) received myeloablative conditioning (MAC) while the rest received reduced intensity conditioning (RIC). Stem cell source was peripheral blood (PBSCs) in 342 (85%). 161 (40%) had a sex-mismatched transplant. All patients received PTCY at 50 mg/kg/day on days +3 and +4 after transplant with tacrolimus and mycophenolate mofetil.

Results

Of 404, 142 (35%) had HLA-DPB1 matched donors with the remaining being DPB1 mismatched recipient-donor pairs. 158 (39%) were categorized as permissive by TCE groups, 54 (13%) non-permissive at GvHD direction and 50 (12%) non-permissive at HvG direction. The median follow up was 33 months (range; 3-120).

On univariate analysis, we observed a significant reduction in relapse in DPB1 permissive mismatched pairs compared with DPB1 matched pairs (HR=0.60, 95%CI=0.40-0.90, p=0.04). HLA-DPB1 non-permissive mismatched pairs (at HvG direction and GvH direction respectively) also experienced lower risk of relapse (HR=0.6, 95%CI=0.30-1.30, HR=0.9, 95%CI=0.5-1.8) but that difference did not reach significance, (p=0.2, 0.9). The impact of DPB1 mismatch on disease relapse persisted in multivariate analysis, independent of DRI risk group.

DPB1 mismatching had no statistically significant impact on NRM at 1-year or 3-years, progression-free survival (PFS) or overall survival (OS) when compared with DPB1-matched pairs. OS at 3-years in DPB1 matched pairs was estimated at 54% and was not significantly different from OS in permissive mismatched (62%, HR=0.9, 95%CI 0.60-1.3, p=0.5), non-permissive mismatched at HvG (69%, HR=0.6, 95%CI 0.3-1.1, p=0.1) or non-permissive mismatched at GvH groups (56%, HR=1.05, 95%CI 0.6-1.7, p=0.8).

The incidence of grade II-IV and III-IV aGvHD at day 180 and cGVHD at 3-years post-HCT did not differ based on HLA-DPB1 mismatched versus matched status (non-permissive mismatched at GvH direction: grade II-IV aGvHD HR=1.1, 95%CI 0.7-1.8, p=0.7, grade III-IV aGvHD HR=0.5, 95%CI 0.1-1.19, p=0.3 and cGVHD HR=0.4, 95%CI 0.1-1.3, p=0.1).

Conclusion

When compared to DPB1 matched pairs, DPB1 permissive mismatching is associated with lower risk of disease relapse, independent of DRI risk, without increase in rates of acute or chronic GVHD. Within the limits of our analysis, PTCy appears to modify the impact of DPB1 matching on transplantation outcomes, and is not associated with difference in OS, PFS, NRM nor rates of acute or chronic GVHD. More work is required to examine the impact of immunologic donor factors in the PTCy era and inform PTCy-specific donor selection.