Treatment with a 5-day versus a 10-day schedule of decitabine in older patients with newly diagnosed acute myeloid leukaemia: a randomised phase 2 trial

Background: Hypomethylating agents, such as decitabine, are the standard of care for older patients with newly diagnosed acute myeloid leukaemia. Single-arm studies have suggested that a 10-day schedule of decitabine cycles leads to better outcomes than the usual 5-day schedule. We compared the efficacy and safety of these two schedules. Methods: Eligible patients were aged 60 years or older with acute myeloid leukaemia but unsuitable for intensive chemotherapy (or <60 years if unsuitable for intensive chemotherapy with an anthracycline plus cytarabine). The first 40 patients were allocated equally to the two treatment groups by computer-generated block randomisation (block size 40), after which a response-adaptive randomisation algorithm used all previous patients' treatment and response data to decide the allocation of each following patient favouring the group with superior response. Patients were assigned to receive 20 mg/m ² decitabine intravenously for 5 or 10 consecutive days as induction therapy, every 4–8 weeks for up to three cycles. Responding patients received decitabine as consolidation therapy on a 5-day schedule for up to 24 cycles. We assessed a composite primary endpoint of complete remission, complete remission with incomplete platelet recovery (CRp), and complete remission with incomplete haematological recovery (CRi) achieved at any time and assessed by intention to treat. This trial is registered with ClinicalTrials.gov, number NCT01786343. Findings: Between Feb 28, 2013, and April 12, 2018, 71 patients were enrolled. 28 received decitabine for 5 days and 43 for 10 days, and all were assessable for efficacy and safety. The primary endpoint was achieved in similar proportions of patients in the two treatment groups (12 [43%] of 28 in the 5-day schedule group, 95% credible interval 26–60, and 17 [40%] of 43 in the 10-day schedule group, 26–54, p=0·78; difference 3%, −21 to 27). Total follow-up was 38·2 months, during which the median duration of overall survival was 5·5 months (IQR 2·1–11·7) in the 5-day group and 6·0 months (1·9–11·7) in the 10-day group. 1-year overall survival was 25% in both groups. Complete remission, CRp, CRi, and overall survival did not differ between groups when stratified by cytogenetics, de-novo versus secondary or therapy-related acute myeloid leukaemia, or TP53 ^mut status. The most common grade 3–4 adverse events were neutropenic fever (seven patients [25%] in the 5-day group and 14 [33%] in the 10-day group) and infection (five [18%] and 16 [37%], respectively). One patient (4%) died from sepsis in the context of neutropenic fever, infection, and haemorrhage in the 5-day group, and in the 10-day group six patients (14%) died from infection. Early mortality was similar in the two groups. Interpretation: In older patients with newly diagnosed acute myeloid leukaemia, efficacy and safety did not differ by the 5-day or the 10-day decitabine schedule. Funding: University of Texas MD Anderson Cancer Center and National Cancer Institute Specialized Programs of Research Excellence.