Peritransplantation Glucocorticoid Haploidentical Stem Cell Transplantation Is a Promising Strategy for AML Patients With High Leukemic Burden: Comparison With Transplantations Using Other Donor Types

Abstract

Using a murine haploidentical bone marrow transplantation (BMT) model, we recently showed that peritransplantation administration of glucocorticoid (GC) redistributed donor T cells from the gastrointestinal tract to bone marrow, which resulted in a significant reduction of graft-versus-host disease (GVHD) while promoting graft-versus-leukemia effects. Furthermore, in a retrospective clinical study of patients with acute myelogenous leukemia (AML) undergoing transplantation in non-remission, we also showed that haploidentical stem cell transplantation (haplo-SCT) using peritransplantation GC administration led to a significantly lower relapse rate and better overall survival rate compared with haplo-SCT using post-transplantation cyclophosphamide. In the present study, using the same dataset of patients undergoing GC haplo-SCT, we retrospectively compared with patients with AML undergoing transplantation in non-remission using 3 other donor types: matched sibling donor (MSD), matched unrelated donor (MUD), and umbilical cord blood (UCB). For GC haplo-SCT, 44 patients underwent peripheral blood stem cell transplantation in a single center (Hyogo College of Medicine), with the conditioning treatment consisting of fludarabine, melphalan, anti-thymocyte globulin (2.5 mg/kg), and TBI 3 Gy. Methylprednisolone was given from the start of conditioning treatment, and the GVHD prophylaxis consisted of tacrolimus and methylprednisolone (1 mg/kg). The transplantation outcomes were compared with data of 1889 patients undergoing MSD-SCT (n = 449), MUD-BMT (n = 493), or UCB transplantation (UCBT) (n = 947) in non-CR, which were extracted from the Transplant Registry Unified Management Program data, the largest data registry in Japan. For donor engraftment, significantly faster neutrophil and platelet engraftment was achieved with GC haplo-SCT compared with allo-SCT using the 3 other donor types. Neutrophil engraftment was achieved at a median of 10 days for GC haplo-SCT, and 20 days for MSD-, MUD-, and UCB-transplants. Platelet engraftment was achieved at a median of 19.5 days for GC haplo-SCT, 42 days for MSD-SCT and MUD-BMT, and 43 days for UCBT, respectively. The incidence of grade II-IV acute GVHD was lower after allo-SCTs using MSD (hazard ratio [HR] = 0.465, P = .003), MUD (HR = 0.524, P = .010), and UCB (HR = 0.647, P = .067) compared with GC haplo-SCT. There was no significant difference in the incidence of chronic GVHD between GC haplo-SCT and allo-SCT using the other 3 donor types. Regarding relapse, GC haplo-SCT was associated with a significantly lower risk compared with MSD-SCT (P < .001) or MUD-BMT (P = .004). GC haplo-SCT tended to have a lower risk compared with UCBT (P = .063). Especially, all the 43 evaluable GC haplo-SCT recipients achieved CR after transplantation, whereas 23.9%, 22.8%, and 27.0% of patients who underwent MSD-SCT, MUD-BMT, and UCBT could not achieve CR after transplantation, respectively. Regarding non-relapse mortality, GC haplo-SCT was associated with a significantly higher risk compared with MUD-BMT (P = .014), and tended to have a higher risk compared with MSD-SCT (P = .061). There was no significant difference between GC haplo-SCT and UCBT (P = .600). Allo-SCTs using MSD (HR = 2.548, P < .001), MUD (HR = 2.134, P = .005), and UCB (HR = 2.376, P = .001) lead to significantly higher overall mortality compared with GC haplo-SCT; the adjusted overall survival at 3 years was 19.8% for MSD, 26.1% for MUD, 28.0% for UCB, and 65.1% for GC haplo. Thus GC haplo-SCT is a promising treatment option for patients with AML with a high leukemic burden.