A Phase 2 Randomized Study Comparing Venetoclax and Azacitidine to Conventional Induction Chemotherapy for Newly Diagnosed Fit Adults with Acute Myeloid Leukemia

Background: The standard therapy for acute myeloid leukemia (AML) for younger, functionally fit patients has not substantially changed in the last 40 years. Intensive induction chemotherapy (IC), with various schedules of cytarabine and an anthracycline, remains the mainstay of therapy. Conventional IC results in complete remission rates of 60-75% of younger and 40-60% of older adults. Despite this, rates of long-term survival and cure remain sub-optimal. These regimens are also associated with significant risk and burden of morbidity, including substantial marrow suppression, mucosal toxicity, high rates of infectious and bleeding complications, potential cardiac injury, prolonged hospitalization, and long-term risk of secondary malignancies. Recently, a placebo-controlled randomized phase 3 trial of older AML patients, and those with comorbidities precluding intensive IC, compared azacitidine and venetoclax to azacitidine and revealed a significant overall survival (OS) advantage for the combination, impressive tolerability, with a markedly higher remission rate at 66% and median OS of 14.7 months, and improvements in other outcome parameters. These recent data compare favorably to outcomes expected from conventional IC, particularly considering their evaluation to date in older, frailer patients, in whom outcomes would be expected to be poorer than younger patients. Based on these considerations, the current randomized phase 2 study seeks to test the hypothesis that treatment with venetoclax and azacitidine will lead to an improved event-free survival (EFS) among fit patients with newly diagnosed AML, when compared to IC. We hypothesized that rates of composite remission for azacitidine and venetoclax will be similar to conventional IC, with an improved safety profile, fewer and less prolonged hospitalizations, lower cost of care, improved candidacy for stem cell transplantation (SCT), and fewer post-SCT complications. Study Design and Methods: This is an open-label, multicenter, phase II randomized clinical trial comparing the therapeutic activity of conventional IC (7+3 regimen or liposomal daunorubicin and cytarabine) to the combination of venetoclax and azacitidine among IC-eligible patients, excluding those with protocol-defined favorable risk genetics or FLT3 mutations. The primary outcome is EFS, where an event is defined as progressive disease (>50% increase in marrow blasts, >50% increase in peripheral blasts, new extramedullary disease), change in therapy due to persistent leukemic disease, relapse after remission, transition to hospice, or death. This study randomizes, in 1:1 fashion, to conventional IC (SOC) or venetoclax plus azacitidine (investigational). Patients will be stratified by age (≥ 65 versus < 65 years old). There is a preselection prior to randomization for either 7+3 or liposomal daunorubicin and cytarabine, to apply to those who are subsequently assigned to the SOC arm. With 172 patients, 86 allocated per arm, the study will achieve 85% power to detect a treatment difference at the one-sided 10% type I error rate, assuming the true hazard ratio is 0.635, based on assumptions of 1-year EFS of 55% on the investigational arm (azacitidine-venetoclax) and 39% on SOC (IC). Before study completion, futility analysis will be trigged after accruing 50% of the subjects and observing 40% of required EFS events. Assuming a failure rate of 30% on the SOC arm and 50% on the investigational arm, if 43 or more events are seen in the latter group, the trial will be stopped for futility. O'Brien-Fleming boundaries set on the p-value due to interim analysis are handled with the Lan-DeMets procedure with the interim p-value = 0.693, and the final p-value = 0.1. Assessments of differences in EFS between the arms will be made with the log rank test; modeling will employ the Cox proportional hazards regression. Differences in EFS at 1-year will also be estimated using the Kaplan Meier method, with standard deviations calculated by Greenwood's formula. Secondary endpoints include composite remission rates, OS, toxicity, rate of measurable residual disease, number and duration of hospitalizations, cost and quality of life measures. Accrual is assumed to be 24 months with a minimum follow up of 15 months, yielding an estimated study duration of 39 months. The study is currently enrolling patients and will be open to accrue at seven hospitals across the United States.