Enasidenib (ENA) Monotherapy with Addition of Azacitidine in Non-Responders Is Effective in Older Patients with Newly Diagnosed IDH2 Mutated Acute Myeloid Leukemia (AML): A Completed Phase 2/1b Sub-Study of the Beat AML Master Trial

Background: ENA is an oral, selective inhibitor of IDH2 approved for the treatment (Tx) of patients (pts) with relapsed/refractory IDH2 mutated (IDH2m) AML. Here we report the results of a Phase 2 expansion and Phase 1b of the Beat AML Master Trial Phase 2/1b sub-study to assess the efficacy of Tx of newly diagnosed (ND) IDH2m AML pts ≥ 60 years of age with ENA monotherapy (ENAm) and subsequent response-driven addition of AZA Tx. (ClinicalTrials.gov NCT03013998).

Methods: The study initiated with a 3-outcome, 2-stage Phase 2 design, which enrolled patients on ENAm for up to 5 cycles. Pts without CR/CRi after 5 cycles of ENAm, or progression/intolerance prior to this time, were transferred to Phase 1b to receive ENA + AZA (Figure 1). Key eligibility included ND IDH2m AML pts with age ≥ 60 years and ECOG performance status 0-2. Pts received ENAm 100 mg/day in continuous 28-day cycles and ENA + AZA (75 mg/m² days 1-7 every 28 days) for Phase 1b. Response was assessed using 2017 ELN AML criteria. The primary endpoint was CR/CRi rate. The 2-stage design required 24 pts and tested the null hypothesis (H₀) that CR/CRi rate equaled 20% vs 50% and then expanded to test a revised H₀ of 30% vs 50% in 60 pts (conditional alpha=0.025, power=77%). Expansion also allowed further assessment of safety of this treatment regimen.

Results: At data cut off (06/18/2020), 60 pts enrolled, received ENAm, and were evaluable for the primary endpoint. Median age was 75 years and 52% were female (Table 1). Median time on ENAm was 4.7 months (mos). At data cut off, 12 pts were still on ENAm Tx. Most common reasons for discontinuing ENAm were Tx failure (defined as no response to treatment) (23 or 38%), disease progression (loss of response to treatment) (7 or 12%) and adverse events (AEs; 6 or 10%). Five pts (8%) went to transplant. CR/CRi was achieved in 28 pts (47%; adjusted 95% CI 28-59, unadjusted exact 95% CI 34-60) (Table 2). Responses were higher (p=0.04) among the 44 pts with IDH2 R140 (55%) as compared to the 16 with IDH2 R172 mutation (25%) further supporting distinct biology between these subsets. After a median follow up of 14.6 mos, the median overall survival (mOS) was 24.4 mos (95% CI 10.6-not reached). The median duration of response was not reached with 12 mos estimation of 57% (95% CI 34-75). Overall, 20 ENA-related serious adverse events (SAEs) occurred in 15 pts, the most common was differentiation syndrome (12 or 20%) and 1 had ENA-related SAE of tumor lysis syndrome (1.7%). One pt had ENA-related Grade 5 AE (renal failure/death). Most common AEs of any grade (in ≥20%) were nausea, anemia, and low potassium (Table 3). The 7-day/30-day/60-day deaths observed with ENAm were 2%/5%/11%, respectively.

Phase 1b: Seventeen pts had inadequate response to ENAm and transferred to Phase 1b to receive ENA + AZA. Median time on Tx (including ENAm) was 6.2 mos and median time on Tx after pts started ENA + AZA was 2.1 mos (Table 2). Most common reasons for discontinuing ENA + AZA included Tx failure (5 or 29%), disease progression (2 or 12%), transplant, death and AEs (each 2 or 12%). CR/CRi was 41% (exact 95% CI 18-67). After a median follow up of 12.7 mos, the mOS from start of ENA + AZA combination Tx was 8.9 mos. Four ENA-related SAEs occurred in 3 pts on ENA + AZA Tx and the most common was differentiation syndrome (2 or 12.5%). One dose-limiting toxicity (Grade 3 nausea) related to both Txs was observed. Most common AEs (≥20%) of any grade were anemia, low albumin and vomiting (Table 3). One death occurred at day 13 of ENA + AZA.

Conclusions: In newly diagnosed pts ≥60 years old with IDH2m AML, ENA had a low early death rate, high CR/CRi rate (47%, adjusted 95% CI 28-59), and yielded durable remissions. The most common unique toxicity with ENA was differentiation syndrome that occurred in 20% of patients. In pts who did not achieve CR/CRi with ENAm, a subset of patients achieved CR/CRi with addition of AZA. This combined approach of serial therapy with ENA monotherapy followed by AZA addition in pts with sub-optimal response resulted in a mOS exceeding 2 years for pts enrolled on study. Further focus on improving response among patients with IDH2 R172 mutations, identifying subsets of pts not responding to ENA monotherapy, and integrating new targeted agents into this treatment regimen are warranted.