The Impact of HLA-DPB1 Mismatch in T-Cell Replete Unrelated Donor Allogeneic Stem Cell Transplantation

Abstract

High-resolution matching of donor-recipient HLA improves outcome in allogeneic stem cell transplants. Matching for HLA-A, -B, -C, –DRB1 and –DQ is mandatory in our transplant centre, to identify 10/10 or 9/10 matched unrelated donors. High resolution matching for DPB1 has been added over the last 10-15 years. However, the role of DPB1 matching is not yet clearly defined. In this study, we retrospectively analyzed the impact of HLA-DPB1 matching on the outcome of T cell replete allogeneic hematopoietic stem cell transplants with Cya/Mtx- and without ATG as GVHD prophylaxis in patients with hematological malignancies at Oslo University Hospital between 2005 and 2017. 301 patients with an unrelated donor fully matched (10/10) at HLA-A, -B, -C, -DRB1, and –DQB1 loci were included. Further, 87 patient-recipient pairs were also fully matched on DPB1 (12/12); 118 had permissive and 96 had non-permissive mismatches of one or two DPB1 alleles. The three groups were comparable with respect to diagnosis, gender, age, cytomegalovirus serostatus and conditioning regimen (Table 1). Cumulative incidence of relapse at 5 years were significantly higher in the DPB1 matched pairs compared with the permissive and non-permissive mismatched ones, at 40% vs 22% and 13% (p Figure 1-2 ). No difference in frequency of acute GVHD grade II-IV between the three groups were found; DP match 43%, permissive mismatch 40% and non-permissive mismatched 48% (p=0.49). Neither was there a difference seen in GVHD grade III-IV; 12% versus 17% versus 11%, respectively. Finally, there were similar outcomes between the three groups regarding chronic GVHD and TRM (Figure 3). In corrected multivariate analyses, only DP matching had significant influence on mortality and survival. Our results show a favorable relapse free and overall survival following a MUD allotransplant with a DBP1 permissive or non-permissive mismatched donor compared to a fully DPB1 matched. This is likely due to an increased GvL-effect in the DPB1 mismatched groups without the counterbalance of increased acute GVHD and TRM.