Final Results of a Prospective Randomized Multicenter Phase III Trial Comparing Treosulfan/Fludarabine to Reduced Intensity Conditioning with Busulfan/Fludarabine Prior to Allogeneic Hematopoietic Stem Cell Transplantation in Elderly or Comorbid Patients with Acute Myeloid Leukemia or Myelodysplastic Syndrome

Background

Allogeneic hematopoietic stem cell transplantation (aHSCT) is a potentially curative treatment option for patients with acute myeloid leukemia (AML) and myelodysplastic syndrome (MDS). A significant challenge in the management of elderly and comorbid patients with AML or MDS is the high transplantation-related mortality (TRM) associated with standard myeloablative preparative regimens preceding aHSCT. Reduced intensity conditioning (RIC) regimens are less toxic, but have often been associated with a higher relapse risk. Hence, there is a high unmet medical need for preparative regimens that reduce the risk of TRM without increasing the incidence of relapse.

Patients and Methods

In this prospective, randomized (1:1), group-sequential, open-label, multicenter, phase III trial, we enrolled patients with AML in complete remission (i.e. blast counts < 5 % in bone marrow) or MDS (blast counts < 20 % in bone marrow during disease history), indicated for aHSCT but considered to be at an increased risk for standard preparative regimens (i. e. patient age ≥ 50 years and / or hematopoietic cell transplantation co-morbidity index [HCT-CI] > 2). The trial was designed to demonstrate at least non-inferiority in event-free survival (EFS) of intravenous (i.v.) treosulfan (10 g/m²/day [d-4 to d-2]) to i.v. busulfan (3.2 mg/kg/day [d-4 to d-3]), both in combination with i.v. fludarabine (30 mg/m²/day [d-6 to d-2]), within 24 months after aHSCT. Events of EFS were defined as relapse of disease, graft failure, or death, whichever occurred first. Secondary objectives included overall survival (OS), cumulative incidence of relapse/progression, and transplant-related mortality (TRM). Engraftment as well as d+28 and d+100 incidence of complete donor-type chimerism (CC) were also evaluated. Further, occurrence of acute and chronic graft-versus-host disease as well as incidence and severity of adverse events between d-6 and d+28 were compared. Pre- and post-transplant immunosuppressive treatment was standardized in both arms and included ciclosporin and short-course methotrexate; anti-thymocyte globuline was used in case of matched unrelated donors (MUD).

A confirmatory interim analysis as pre-specified in the protocol was conducted and the results were reviewed by an independent data monitoring committee (DMC) in Nov 2016. The DMC recommended to stop further patient accrual due to statistically proven non-inferiority of the treosulfan/fludarabine regimen. All data presented here are derived from the 476 patients included in the confirmatory final analysis.

Results

Median patient age was 60 years with 94 % of patients aged 50 years or older; 64 % had AML and 76 % underwent MUD-aHSCT. HCT-CI scores were well-balanced between treatment regimens. Both regimens were comparable with regard to all safety evaluations between d-6 and d+28. Trilineage engraftment was also similar after both regimens, although the rate of CC at d+28 was significantly higher after treosulfan. The EFS at 24 months after aHSCT was significantly non-inferior (64.0 % vs. 50.4 %), corresponding to a hazard ratio (HR) of 0.65 in favor of treosulfan (p=0.0000164). The Kaplan-Meier estimate of OS at 24 months was also significantly superior for patients treated with the treosulfan as compared to the busulfan regimen (72.5 % vs. 56.4 %; p=0.0082). Estimates of 24-months EFS and OS were consistent within relevant prognostic subgroups. Furthermore, the treosulfan regimen significantly reduced 24-months TRM (11.3 % vs. 28.2 %; HR 0.54; p=0.0201) in this more vulnerable patient population. Notably, the cumulative incidence of relapse/progression at 24 months was comparable between the two regimens.

Conclusions

The results of this large randomized, multicenter, phase III trial clearly demonstrate at least non-inferiority of the treosulfan/fludarabine regimen compared to the RIC busulfan/fludarabine regimen for elderly or comorbid AML and MDS patients undergoing aHSCT. Moreover, the clinical results were consistently in favor of the treosulfan/fludarabine regimen regarding the most relevant endpoints of aHSCT and thus confirm a major clinical benefit of this regimen as compared to the RIC busulfan/fludarabine regimen in this selected patient population.

(Funded by medac GmbH; MC-FludT.14/L, EudraCT no: 2008-002356-18; ClinicalTrials.gov identifier: NCT00822393)