Post-transplant cyclophosphamide (PTCy) is a commonly used graft-vs-host disease (GVHD) prophylaxis, particularly in the setting of haploidentical (haplo) hematopoietic cell transplantation (HCT). The rate of graft failure has been reported to be as high as 12% to 20% in haplo-HCT recipients using PTCy. The objective of this study was to determine whether donor type influenced the risk of late graft failure following reduced-intensity conditioning (RIC) HCT using PTCy-based GVHD prophylaxis. A retrospective cohort analysis using the Center for International Blood and Marrow Transplant Research (CIBMTR) database among adult patients who underwent first RIC haplo or 8/8 matched unrelated donor (MUD) HCT between 2011 and 2018 for acute myeloblastic leukemia (AML), acute lymphoblastic leukemia (ALL) or myelodysplastic syndrome (MDS) with PTCy GVHD prophylaxis. The primary outcome was incidence of late graft failure, defined as secondary graft loss in the absence of relapse or poor graft function requiring a cellular therapy intervention. A total of 1336 patients met the eligibility criteria (1151 haplo, 185 MUD). Patients in the MUD group were older (65 vs. 61 years), less ethnically diverse (95% vs. 72% White), received fewer bone marrow grafts (45% vs. 16%), and had younger donors (median age, 28 vs. 37 years old). Conditioning regimens were predominately fludarabine, cyclophosphamide, and total body irradiation (TBI; 87% haplo and 38% MUD). At 2 years, the adjusted probabilities of late graft failure for the haplo group was 6.5% (95% confidence interval [CI], 5.2-8.0) versus 5.9% (95% CI, 2.7%-10.9%) for the MUD group (P = .79). Multivariate analysis for risk factors associated with late graft failure found associations with a diagnosis of MDS (HR, 1.98; 95% CI, 1.22-3.20; P = .005), and earlier year of HCT (2015-2018 vs. 2011-2014; HR, 0.39; 95% CI, 0.24-0.64; P = .0002). A post-hoc sensitivity analysis was performed to evaluate the effect of donor age and use of peripheral blood stem cell (PBSC) grafts. Graft failure did not differ between haplo and MUD HCT (HR, 1.19; P = .67) when adjusted for donor age nor when restricted to PBSC grafts only (HR, 0.85; P = .70). In this registry-based analysis of patients undergoing RIC HCT for AML, ALL, or MDS using GVHD prophylaxis with PTCy, there was no significant difference in late graft failure rates between haplo and MUD donors. Overall rates of late graft failure were high.
Read full abstract