Comparative Results of Azacitidine and Decitabine from a Large Prospective Phase 3 Study in Treatment Naive Patients with Acute Myeloid Leukemia Not Eligible for Intensive Chemotherapy

Background: Prognosis of elderly (≥65 years of age) patients (pts) with acute myeloid leukemia (AML) remains dismal with a substantial proportion being deemed unfit for intensive chemotherapy. Monotherapy with the hypomethylating agents azacitidine (AZA) or decitabine (DEC) has been the de facto standard of care for the treatment of chemotherapy-ineligible AML pts although both AZA and DEC did not improve median OS compared to low-dose cytarabine (LDAC) or physician choice, respectively, in phase III trials. No clinical trials comparing AZA and DEC head-to-head in AML exist. Here, we present a subgroup analysis of pts enrolled in the phase III ASTRAL-1 trial (NCT02348489) who were randomized to the AZA or DEC control arm.

Methods: ASTRAL-1 randomized 815 treatment-naïve AML pts ineligible for intensive chemotherapy in a 1:1 ratio to either guadecitabine or treatment-choice (TC) of AZA, DEC, LDAC (NCT02348489). Study protocol and results have been presented previously (Fenaux, EHA 2019). Briefly, adult (≥18 years of age) pts with newly-diagnosed AML ineligible for intensive chemotherapy based on age of 75 years or older, major organ comorbidities, and Eastern Cooperative Oncology Group performance status (ECOG PS) of 2-3 were eligible for enrollment. Exclusion criteria included prior treatment with AZA or DEC, extramedullary central nervous system AML, inability to tolerate treatment in the TC arm, or refractory congestive heart failure, uncontrolled active infection, or advanced pulmonary disease. Pts were pre-selected to receive either AZA, DEC, or LDAC with subsequent 1:1 randomization to either guadecitabine or TC in the respective arm. Patients treated with standard doses and schedules of AZA or DEC within the TC arm were included in the subgroup analysis presented here. Co-primary outcomes were rates of complete response (CR) and median, 1-year, and 2-year overall survival (OS) as defined by the International Working Group response criteria for AML. Composite CR (CRc) was defined as the composite of CR, CR with incomplete platelet count recovery (CRp), and CR with incomplete cell count recovery (CRi).

Rates of CR among pts treated with AZA and DEC were compared using Fisher’s exact test. Survival outcomes were compared using log-rank tests to compare the hazard ratio for death among the AZA and DEC treated pts. Subgroup analyses for OS stratified by patient and disease characteristics were performed.

Results: 815 patients were enrolled in the ASTRAL-1 trial across 144 sites in 24 countries with 171 and 167 pts being randomized to and treated with AZA and DEC in the TC arm of the trial, respectively. Baseline patient and disease characteristics were well-balanced between the AZA and DEC-treated pts (Table 1). The median number of treatment cycles was 6 (range [R]: 1-31) in the AZA arm and 5 (R: 1-34) in the DEC arm. There was no statistically significant difference in the co-primary endpoint of CR with 30 pts (17.5%) in the AZA and 32 pts (19.2%) in the DEC arm achieving CR (p=0.78). The rate of CRc (CR + CRp + CRi) was comparable among AZA and DEC-treated patients with 22.2% (38 out of 171 pts) and 25.1% (42 out of 167 pts), respectively (Table 2). Median OS between AZA and DEC-treated pts was similar with 8.7 months and 8.2 months in the two arms, respectively (hazard ratio [HR] for death: 0.97; 95% CI: 0.77-1.23; p=0.81). One-year and 2-year OS was comparable in both groups with 39% and 15% in the AZA arm and 32% and 14% in the DEC arm, respectively. Median OS estimates in clinically or genetically-defined patient subgroups were similar between AZA and DEC-treated pts. Serious adverse events leading to death occurred more frequently in the AZA arm compared with DEC (AZA: 38% vs 26% with DEC; p=0.02).

Conclusion: Outcomes in treatment-naïve AML pts ineligible for intensive chemotherapy treated with AZA or DEC in the randomized phase III ASTRAL-1 trial are comparable with CR rates of 17.5% and 19.2% and median OS of 8.7 months and 8.2 months, respectively. No patient, disease, or molecular characteristics predicted a higher likelihood of response to either AZA or DEC. Safety in this frail patient population was comparable to prior trails of HMAs in AML and no major safety differences between AZA and DEC were detected although fatal serious adverse events tended to be higher in the AZA-treated cohort.