Background: ENA and AZA each induce overall response rates (ORR) of ~30% and complete remission (CR) rates of ~20% in ND-AML. In vitro, combining ENA + AZA enhances cell differentiation. We report results of the phase II portion of an open-label, randomized phase I/II study of ENA + AZA (“E+A”) vs. AZA monotherapy (“A”) in patients (pts) with mIDH2 ND-AML (NCT02677922). Methods: Pts age ≥ 18 years ineligible for intensive chemotherapy, with ECOG PS ≤ 2 and intermediate- or poor-risk cytogenetics, were randomized 2:1 to E+A or A in 28-day (d) cycles. All pts received SC AZA 75 mg/m²/d x 7 d/cycle; pts randomized to E+A also received ENA 100 mg QD. The primary endpoint was ORR (CR, CR with incomplete recovery, partial remission, morphologic leukemia-free state). Other endpoints include duration of response (DOR), overall and event-free survival (OS, EFS), safety, and mIDH2 VAF. Results: 101 pts received E+A (n = 68) or A (n = 33). Median age was 75 years (57–85); most pts (83%) had intermediate-risk cytogenetics. 21 pts in the E+A arm and 1 in the A arm were ongoing at data cutoff (Aug 2019). Most common reason for discontinuation was disease progression (E+A 31%, A 52%). Median number Tx cycles was 10 (1–26) in the E+A arm and 6 (1–28) in the A arm. 7 pts (21%) in the A arm received subsequent Tx with ENA. ORR, CR rate and DOR were significantly improved with E+A vs. A (Table). Median OS was 22 mo in both arms (HR 0.99 [95%CI 0.52, 1.87]; P = 0.97). Median EFS was 17.2 and 10.8 mo in the E+A and A arms, respectively (HR 0.59 [95%CI 0.30, 1.17]; P = 0.13). Maximal mIDH2 VAF change from BL was –83.4% with E+A vs. –17.7% with A (P < 0.01). No baseline co-mutation predicted primary resistance. Common Tx-related grade 3–4 AEs in the E+A arm were thrombocytopenia (37%), neutropenia (35%), anemia (19%), and febrile neutropenia (15%); these occurred in 19%, 22%, 22%, and 16% in the A arm. Grade 3–4 infections occurred in 18% of E+A pts and 31% of A pts. IDH differentiation syndrome occurred in 12 pts (18%) in the E+A arm. 5 E+A pts (7%) and 1 A pt (3%) died in the first 60 d. Conclusions: Combining ENA + AZA resulted in significantly improved response rates and durations, and was generally well-tolerated in older patients with mIDH2 ND-AML. The impact of subsequent Tx on OS/EFS and new translational data will be presented at the meeting. Clinical trial information: NCT02677922. ENA + AZA
n = 68AZA Only
n = 33ORR, n (%)48 (71)14 (42)P < 0.01    DOR, mo, median [95%CI]24.1 [11.1, NR]12.1 [2.8, 14.6]P = 0.05CR, n (%)36 (53)4 (12)P < 0.01Stable disease, n (%)13 (19)13 (39)