Background. The impact of allogeneic hematopoietic stem cells transplantation (allo-HSCT) for treating acute myeloid leukemia (AML) is as yet unclear. Discrepancies arise due to the heterogeneity of diagnostic and clinical features that individual patients present with, as well being affected ,by the chosen therapies and statistical methods used. The study is thus aimed to compare treatment outcomes in a patient group suffering from AML with first complete remission (CR1), aged 16–60 years, being either treated with or without allo-HSCT and a sub-group receiving DAC induction treatment according to cytogenetic risk; outcomes being assessed by relapse free survival (RFS), overall survival (OS), cumulative incidence relapse (CIR) and non relapse mortality (NRM).

Patients and Methods. Subjects were a cohort of 622 AML patients, median age 44 years (16–44), divided into those not receiving allo-HSCT (n = 775, 76%), and those that underwent allo-HSCT (n = 147, 24%). The patients had been recruited for two prospective, randomised and multicentre clinical trials, including PALG AML DAC vs. DA (1999–2002) and PALG AML 1/2004 DAC vs. DAF vs. DA (2004–2008), where therapeutic schedules were similar regarding induction of remission and post remission protocols. Survival analysis was performed by the Mantel-Byar method; being a modified log-rank test using the allo-HSCT as the time-dependent covariate.

Results. A univariate analyses demonstrated that allo-HSCT procedure significantly improved RFS and OS in all patients with AML in CR1. In addition, improvements were likewise observed in the subgroup of patients treated with DAC induction chemotherapy, those with unfavourable cytogenetics according SWOG (SWOG 2) and in the patient age groups of 41–60 and 41–50 years. Improvements in just RFS were seen only in either the patient subgroup aged 16–40 years or those with intermediate cytogenetic risk according to SWOG (SWOG 1). The allo-HSCT procedure did not however enhance RFS and OS in patients with favourable cytogenetics according SWOG (SWOG 0). Multivariate analyses demonstrated the allo-HSCT procedure to be an independent and favourable factor for RFS outcomes in all AML patients in CR1, in those subgroup patients treated using DAC induction, in age groups 16–40, 41–50, 41–60 years and in the SWOG 1 and SWOG 2 subgroups. Similarly, the allo-HSCT proved to be an independent and favourable factor for OS outcomes in patients aged 41–60 years and in the SWOG 2 subgroup. It was also observed that age, in a linear manner, independently and favourably affected RFS in patients treated using DAC induction. In contrast, having more than one induction cycle before CR1 assessment was found to independently and unfavourably affect the RFS in patients treated with DAC induction and in patients aged 41–60 years. At diagnosis and in linear fashion, the white blood count (WBC) unfavourably affected OS in patients aged 51–60 and 41–60 years. Previous diagnosis of myelodysplastic syndrome (MDS) before AML independently and favourably affected OS outcomes for all AML patients in CR1 within age groups 51–60 and 41–60, years together with patients in the SWOG1 subgroup. However, MDS was seen to independently and unfavourably affect RFS in patients from the SWOG 0 subgroup, but favourably in the age group 41–60 years. CIR outcomes showed that allo-HSCT reduced the risk of relapse in all AML patients with CR1 and in the SWOG 2 subgroup. Furthermore, NRM outcomes demonstrated that allo-HSCT significantly reduced death unrelated to relapse in all AML patients in CR1 as well as those in subgroups aged 41–50 and 41–60 years.

Summary. It was thereby proved that adopting allo-HSCT procedure is especially favourable for OS and RFS outcomes, but also for CIR in patients with AML in CR1 with unfavourable cytogenetics according to SWOG. It is however, not recommended that patients with favourable cytogenetics to undergo allo-HSCT. In patients treated with DAC induction, those belonging to the intermediate cytogenetic risk group and in all the age groups, employing allo-HSCT should be considered, but account should also be taken of additional factors that may influence outcome, such as patient age, previous MDS before AML was diagnosed, WBC at diagnosis and the number of induction chemotherapy cycles required to obtain CR1.