Time to response and survival in hypomethylating agent-treated acute myeloid leukemia

Azacitidine and decitabine are valuable therapeutic options in older patients with acute myeloid leukemia (AML) deemed unfit for induction chemotherapy. Mechanisms underlying response to hypomethylating agents (HMAs) include direct cytotoxic effects, terminal differentiation, and apoptosis [1]. Unlike with other induction chemotherapies, HMAs may require several treatment cycles before demonstrating anti-leukemic response [2,3]. In fact, a prolonged treatment approach is key to observing a survival advantage with HMAs. In addition, survival benefits are realized not only in complete and partial response but also in patients who obtain a hematological improvement (HI) or stable disease (SD) [4]. Data are lacking on the optimal duration of maintenance therapy, and it is currently recommended to continue treatment until evidence of overt progression. This becomes especially significant as outcomes after HMA failure are dismal with median residual survival of only 4.5–5.5 months [5]. It remains unclear as to what factors determine the kinetics of treatment response. Mechanistic studies suggest that higher doses of therapy are primarily cytotoxic while hypomethylation activity is more prominent at lower levels of exposure [6]. It may be that direct cytotoxic mechanisms dominate blast clearance responses in very early responders, while differentiation induction continues to exert its effects over later cycles of therapy. While time to complete disease remission independently predicts for residual and disease free survival in patients who receive standard induction [7], the same situation may not hold true for differentiation modifiers, in this case, hypomethylating therapies. With this background, we sought to determine the relationship between kinetics of response (time to first response beyond SD) on overall survival (OS). Patients with newly diagnosed AML treated between Jan 2003 and July 2016 were included in the study. Patients’ disease required to fulfill the WHO criteria for AML ( 20% blasts either in peripheral blood or bone marrow) to be included. All studies were approved by the Institutional Review Board and conducted in accordance with the Declaration of Helsinki. Remission criteria (complete remission with or without platelet recovery (CR/CRp), partial (PR)) were graded according to the international working group (IWG) response criteria [8]. Patients who did not meet the above criteria, were analyzed for persistence of SD with HI on the erythroid, platelet, or neutrophil lineages, as per IWG 2006 criteria [9]. SD was defined by failure to reach at least HI, without evidence of progressive disease for >2 months. Response was analyzed on an intent-to-treat analysis considering patients as treatment failure if there was no response after at least eight cycles of therapy regardless of treatment interruptions. Two regimens were included: azacitidine 75mg/m 7 d and decitabine 20mg/m 3–5 d. Response assessments were performed on the bone marrow after every cycle for the first 2–3 cycles, and periodically every 2–3 cycles thereafter, based on physician determination of improvement in response by peripheral blood counts and other relevant parameters, or by protocol requirements. Landmark time point assessments for survival included 3-month post-HMA and 6 month post HMA start of therapy, with patients classified into two groups based on whether they furthered their response beyond SD. Since the 10-day (d) decitabine is a far more aggressive approach that results in higher and faster complete remission rates, patients treated on 10-d decitabine regimens were not included to maintain homogeneity in regimen intensity [10]. Epigenetic modifier mutations identified in our 28-gene next generation sequencing panel included ASXL1, EZH2, IDH1/2, and DNMT3A [11]. The covariates abstracted from our database were: response, age, gender, ECOG-PS, regimen type (azacitidine vs decitabine), white blood cell count, marrow blast percentage, peripheral blasts, cytogenetics (intermediate versus adverse) [12], mutational status (epigenetic vs non-epigenetic mutation/no mutation),