Abstract
Background: High dose post-transplant cyclophosphamide has improved and expanded the use of haploidentical hematopoietic cell transplantation (haplo-HCT). The impact of HLA disparity in this setting, however, is unclear and traditionally measured at the antigen or allele level. The HLA Matchmaker tool more precisely quantifies HLA disparity by calculating the number of mismatched three-dimensional amino acid patches, or epitopes, on the surface of two mismatched HLA antigens. Previous research has shown that class II epitope mismatch (EM), which includes the combination of DRB1 and DQB1 EM, is associated with reduced relapse and delayed engraftment in haplo-HCT (Rimando et al, ASBMT 2018 Abstract ID#53). The individual impact of DRB1 and DQB1 EM on clinical outcomes is currently unclear.
Methods: 148 patients who received a peripheral blood T cell-replete haplo-HCT at a single center between July 2009 and November 2017 were retrospectively analyzed. All patients age ≥ 18 were included regardless of diagnosis. The HLA EM load for total class II disparity, DRB1, and DQB1 was quantified using the HLA Matchmaker software and a python script in a dose-dependent and vector-stratified fashion. The primary outcome was the incidence of relapse. The secondary outcomes included relapse-free survival (RFS), time to neutrophil engraftment, and time to platelet engraftment. The association between HLA EM and outcome was analyzed using the Cox proportional hazard model or Gray's sub-distribution method for competing risk as appropriate. The hazard ratios calculated report the hazard rate for a single unit increase in EM. This study was approved by the Institutional Review Board.
Results: In this updated cohort, class II graft-versus-host (GvH) EM was again associated with reduced incidence of relapse (HR 0.966; 95% CI 0.938-0.995; p=0.023) and improved RFS (HR 0.978; 95% CI 0.957-0.999; p=0.037) (Table 1). Class II host-versus-graft (HvG) EM was again associated with delayed neutrophil (HR 0.976; 95% CI 0.955-0.997; p=0.027) and platelet (HR 0.971; 95% CI 0.950-0.993; p=0.010) engraftment. Neither DRB1 nor DQB1 GvH EM was independently associated with relapse or RFS. DRB1 HvG EM was associated with worse relapse-free survival (HR 1.038; 95% CI 1.002-1.076; p=0.040). DRB1 HvG EM was associated with delayed time to neutrophil engraftment (HR 0.965; 95% CI 0.934-0.997; p=0.033), while DQB1 HvG EM was not. DQB1 HvG EM was associated with delayed time to platelet engraftment (HR 0.967; 95% CI 0.938-0.997; p=0.032), while DRB1 HvG EM was not.
Summary: HLA Matchmaker software enables the quantification of HLA EM in haplo-HCT patients. Neither DRB1 nor DQB1 GvH EM was independently associated with relapse or RFS. DRB1 HvG EM was associated with neutrophil but not platelet engraftment, while DQB1 HvG EM was associated with platelet but not neutrophil engraftment. HLA EM represents a novel strategy to predict clinical outcome in haplo-HCT. The mechanism for the divergent roles of DRB1 EM and DQB1 EM on neutrophil and platelet engraftment are currently unknown and warrant further investigation.
Disclosures
No relevant conflicts of interest to declare.
Untreated donor specific antibodies against HLA are associated with poor outcomes in peripheral blood haploidentical hematopoietic cell transplantation
