Molecular Responses Are Observed across Mutational Spectrum in Treatment-Naïve Higher-Risk Myelodysplastic Syndrome Patients Treated with Venetoclax Plus Azacitidine

Introduction: Patients (pts) with higher-risk myelodysplastic syndromes (MDS) are typically treated with azacitidine (Aza). Venetoclax (Ven) is a selective, potent, oral BCL-2 inhibitor that has demonstrated synergy with Aza in preclinical studies of myeloid malignancies. Higher-risk MDS is associated with mutations in genes involved in RNA splicing, epigenetic regulation, transcription, and cellular signaling. Assessing the dynamics of genetic variants during treatment of higher-risk MDS enables understanding of the molecular determinants of response. This phase 1b study (NCT02942290) evaluates Ven + Aza for treatment-naïve higher-risk MDS, and we report efficacy among mutationally defined subgroups as well as the depth of molecular response.

Methods: Pts (≥18 years) with higher-risk MDS enrolled in the study had International Prognostic Scoring System intermediate-2 or high-risk MDS, bone marrow (BM) blasts <20% at baseline, and ECOG ≤2 performance status. Aza 75 mg/m ² was administered on Days (d) 1-7 of each 28-d cycle. The Ven RP2D was 400 mg x 14 d of each 28-d cycle. Primary objectives were to assess the Ven+Aza safety profile and to establish the RP2D. Key secondary objectives were to assess the overall response rate (ORR), defined as complete remission (CR), marrow CR and partial remission (PR), and overall survival. Analyses were carried out on all pts who received ≥1 dose of study drug and efficacy was evaluated per IWG 2006 response criteria.

Molecular responses were quantified by mutation analysis of baseline and serial BM aspirate (BMA) or peripheral blood (PB) samples collected at protocol-specified time points. Mutations were identified in BMA using Archer VariantPlex Myeloid 75-gene panel [limit of detection (LOD) 5%; time points: pre-treatment (n=43), on-treatment (n=32), and treatment completion visit (TCV) (n=25)] or TruSight Myeloid 54-gene Sequencing Panel in PB [LOD 2%; time points: pre-treatment (n=21), on-treatment (n=19), and TCV (n=8)] to assess molecular dynamics during Ven + Aza treatment. Molecular responses were only compared within the same tissue type.

Results: At the Dec 15, 2020 cutoff, 78 pts had received Ven+Aza, including 51 who received Ven at the RP2D of 400 mg x 14 d, with median follow up time of 23 mos (range 0.1-44.2). Median age was 70 years (range 26-87); 72% male; and 91% had excess BM blasts (>5 to ≤10%, n=21; >10 to ≤20%, n=49; >20%, n=1). For the entire population, mORR was 80% (CR 40% and mCR 40%; no PR). 42% with mCR also had hematologic improvement [HI].

Screening mutational profiling was performed on 46/51 pts who received the RP2D of Ven + Aza. The most common mutations were TP53 (26%), ASXL1 (24%), U2AF1 (17%), and RUNX1 (15%), consistent with a higher-risk MDS population. Clinical responses (CR + mCR) were observed across the mutational spectrum, including in pts with poor prognostic mutations in TP53 (83%), ASXL1 (82%), and RUNX1 (71%).

Sixty-four pts treated with Ven + Aza (all Ven dosing cohorts) had paired BMA or PB pre-treatment and on-treatment and/or end of study samples available for serial analysis at the Dec 15, 2020 data cutoff. Ven + Aza resulted in robust and rapid molecular responses across the mutational spectrum. Pts who achieved CR at the time of serial sample acquisition had more significant reduction of variant allele frequencies (VAFs) (n=18 pts, mean VAF pre-treatment = 38.3; mean VAF at CR = 11.8) compared to pts who achieved stable disease or HI (n=11 pts, mean VAF pre-treatment = 29.8; mean VAF at SD or HI = 24.0) or progressive disease (n=10 pts, mean VAF pre-treatment = 27.4; mean VAF at PD = 26.9).

Reductions of VAFs below the LOD were observed across the mutational spectrum, including in genes that are considered poor prognostic in higher-risk MDS (e.g., TP53, ASXL1 and RUNX1), demonstrating the broad molecular activity of Ven + Aza (Figure). Finally, molecular responses were observed quickly, with VAF reductions below the LOD observed as early as end of Cycle 1, consistent with the mechanism of action of Ven directly activating the mitochondrial apoptotic pathway in cells.

Conclusions: Pts with higher-risk MDS treated with Ven + Aza had rapid, durable responses and high remission rates. Pts across key mutational profiles achieved meaningful clinical and molecular responses, supporting an all-comers approach. Updated data will be presented, which will provide more follow up time and durability of responses among mutational subsets.

Figure 1

View largeDownload slide

Figure 1

View largeDownload slide

Close modal