Results of Venetoclax and Azacitidine Combination in Chemotherapy Ineligible Untreated Patients with Acute Myeloid Leukemia with IDH 1/2 Mutations

Introduction

Isocitrate dehydrogenase (IDH) mutations occur in ~20% of acute myeloid leukemia (AML) patients (pts) and are frequently identified in older pts. Pre-clinical data showed that cells with IDH1/2 mutation (mut) were susceptible to Venetoclax (Ven) therapy (Chan et.al., 2015). Clinical studies have demonstrated that pts with IDH1/2 ^mut treated with Ven and azacitidine (Aza) achieved high response rates that were durable, and had longer median overall survival (mOS) (DiNardo et. al., 2020; DiNardo et. al., 2019). Herein, we further evaluated the efficacy and safety of Ven+Aza among treatment-naïve AML pts with IDH1/2 ^mut unfit for intensive treatment either due to comorbidities and/or age ≥ 75 yrs.

Methods

Data were pooled from pts enrolled in an ongoing phase 3 study (NCT02993523, data cut-off: 04Jan2020) comparing pts treated with Ven+Aza or placebo (Pbo)+Aza, and a phase 1b study (NCT02203773, data cut-off: 19Jun2019) where pts were treated with Ven+Aza. Pts on Ven+Aza received Ven 400 mg daily orally (days 1–28) and Aza (75 mg/m²; days 1-7/28-day cycle). Disease assessments were performed per modified International Working Group response criteria for AML.

DNA was isolated from bone marrow aspirates collected from pts prior to the first dose of study drug and analyzed centrally. Pts with positive test results for IDH1 and 2 [RealTime IDH1 or IDH2 assay (Abbott) for phase 3 study, MyAML panel (Invivoscribe) for phase 1b study] were counted as mutation “detected”; pts with a negative test result were counted as mutation “not detected.” Pts without a result either due to an inconclusive test or missing specimen were excluded from the analyses.

Results

IDH1/2 ^mut was detected in 79 pts treated with Ven+Aza and 28 pts treated with Pbo+Aza. At baseline (Ven+Aza/Pbo+Aza), median age was 76 (range: 64-90)/78 (62-90) yrs. Cytogenetic risks were intermediate: 79%/68%, and poor:21%/32%. ECOG scores were: 0-1: 56%/68% and 2-3: 44%/32%. 73%/86% had de novo AML, and 27%/14% had secondary AML. IDH1 was detected in 32/11, IDH2 in 49/18, IDH2 ^R140 in 36/15, and IDH2 ^R172 were detected in 13/3 pts, respectively.

The median number of treatment cycles received (Ven+Aza/Pbo+Aza) by IDH1/2 pts was 8 (range: 1-37)/2.5 (1-18). Complete response (CR)+CR with partial hematologic recovery (CRh) in Ven+Aza/Pbo+Aza were 72% (95% CI: 61%-82%)/7% (1%-24%) (Table). Median time to first CR/CRh response was 1.0/2.6 months (mos), and 51%/0% achieved CR+CRh by initiation of cycle (C) 2. Median duration of response (mDoR) was 29.5 [95% CI: 16.7-not reached(NR)]/15.5 (NR-NR) mos and mOS was 24.5 [15.2- NR/6.2 (2.3-12.7) mos. The separation of 95% CIs for mOS indicate superior treatment effect of Ven+Aza (Figure).

In pts with IDH1, CR+CRh (Ven+Aza/Pbo+Aza) was 59%/9%. Median time to first CR/CRh response was 2.3/3.1 mos, and 25%/0% achieved CR+CRh by initiation of C2. Median DoR and OS were 21.9 (7.8-29.5)/NR and 17.5 (6.3-32.7)/2.2(1.1-5.6) mos. In pts with IDH2, CR+CRh rates were 80%/6%. Median time to first CR/CRh response was 1.0/2.1 mos and 67%/0% achieved CR+CRh by initiation of C2. Median DoR was NR (16.7-NR)/15.5 (NR-NR), and mOS was NR (17.6-NR)/13.0 (95% CI: 3.8-15.8) mos.

In pts with IDH2 ^R140, CR+CRh were 75%/7%. Median time to first CR/CRh response was 1.0/2.1 mos, and 58%/0% achieved CR+CRh by initiation of C2. Median DoR and mOS were NR (17.8-NR)/15.5 (NR-NR) and NR (15.0-NR)/12.7 (1.7-15.8) mos, respectively. Response rates in pts with IDH2 ^R172 were 92%/0%. Median time to first CR/CRh response was 1.0/NR mos, and 92%/0% achieved response by initiation of C2. Median DoR and mOS were 16.7 (7.5-NR)/3.5 (NR-NR) and NR (12.2-NR)/13.7 (10.6-NR) mos.

IDH1/2 ^mut pts achieved higher CR+CRh rates with Ven+Aza treatment as compared to pts with IDH not detected (72%/60%). Median DoR and mOS were 29.5 (16.7-NR)/17.5 (10.6-23.5) and 24.5 (15.2-NR)/12.3 (9.7-14.8) mos, respectively.

In pts treated with Ven+Aza, grade 3/4 hematologic adverse events (AEs) among pts with IDH1/2 detected/IDH not detected were similar (81%/74%) and included grade 3/4 thrombocytopenia (46%/34%), febrile neutropenia (43%/41%), neutropenia (35%/33%) and anemia (29%/24%).

Conclusion

Ven+Aza compared to Aza monotherapy resulted in higher response rates, longer DoR, and mOS among treatment-naïve pts with IDH1/2 ^mut ineligible for intensive chemotherapy. The safety profile was acceptable. No unexpected toxicities were noted with Ven+Aza combination.