Outcomes after use of two standard ablative regimens in patients with refractory acute myeloid leukaemia: a retrospective, multicentre, registry analysis.

BACKGROUND: Cyclophosphamide plus intravenous busulfan has not been compared with cyclophosphamide plus total body irradiation (TBI) in adults with advanced refractory acute myeloid leukaemia before allogeneic haemopoietic stem-cell transplantation (HCT). We aimed to assess whether survival of patients receiving ablative intravenous busulfan-based conditioning regimens before a related or volunteer-unrelated donor HCT for refractory acute myeloid leukaemia is not inferior to that of patients receiving an ablative TBI-based regimen.

METHODS: In this retrospective, multicentre, registry-based study, we obtained data for patients (aged >18 years) with refractory acute myeloid leukaemia in active phase of disease, who had received HCT from an HLA-identical sibling or an unrelated donor after intravenous busulfan plus cyclophosphamide or cyclophosphamide plus TBI conditioning between 2000 and 2012. Data was obtained from the European Group for Blood and Marrow Transplantation registry. The primary endpoints of the study were overall survival and leukaemia-free survival.

FINDINGS: We obtained data for 514 patients who had received intravenous busulfan plus cyclophosphamide and 338 patients who had received cyclophosphamide plus TBI. The median percentage of blasts before HCT did not differ significantly between groups (20% [range 5-100; IQR 10-32] in the intravenous busulfan plus cyclophosphamide group vs 16% [5-95; 9-33] in the cyclophosphamide plus TBI group; p=0·16). Overall survival at 2 years did not differ between the groups in the univariate analysis (31·2% [95% CI 26·8-35·5] with intravenous busulfan plus cyclophosphamide vs 33·4% [28·1-38·7] wth cyclophosphamide plus TBI; p=0·65). Leukaemia-free survival at 2 years also did not differ between groups (25·0% [95% CI 21·0-29·0] vs 28·4% [23·4-33·5]; p=0·47). In multivariable analysis adjusting for differences between both groups, no difference was noted between the two groups in terms of overall survival (hazard ratio [HR] 0·99 [95% CI 0·83-1·20]; p=0·95) or leukaemia-free survival (HR 0·97 [0·81-1·16]; p=0·71). Main causes of non-relapse mortality were graft-versus-host disease (49 [10%] in the intravenous busulfan plus cyclophosphamide group vs 25 [7%] in the cyclophosphamide plus TBI group) and infection (36 [7%] vs 18 [5%]).

INTERPRETATION: From a practical standpoint, the use of intravenous busulfan plus cyclophosphamide is likely to be a valid and efficient alternative to cyclophosphamide plus TBI conditioning regimen for patients with refractory acute myeloid leukaemia, especially for those transplant centres without access to radiation facilities.

FUNDING: None.