Molecular correlates of response and survival with venetoclax plus hypomethylating agent in treatment-Naïve high-risk myelodysplastic syndrome Free

Introduction High-risk myelodysplastic syndrome (HR-MDS) remains therapeutically challenging and is associated with poor overall survival (OS). While potentially curative, many elderly patients are not eligible for Allogeneic stem cell transplantation (allo-SCT). For these patients, hypomethylating agents (HMAs) remain standard of care but yield modest responses. The addition of venetoclax (Ven) to HMA has demonstrated promise, but the prognostic implications of molecular features in this setting remain poorly defined. We evaluated treatment outcomes and the impact of somatic mutations on response and survival in a cohort of treatment-naïve HR-MDS patients treated with Ven-HMA.

Methods

We retrospectively analyzed 43 treatment-naïve patients with HR-MDS (IPSS-R/M high or very-high risk) who received Ven-HMA at our institution. HMA backbones included azacitidine (75 mg/m² IV/SC, days 1–7), IV decitabine (20 mg/m², days 1–5), or oral decitabine-cedazuridine (35 mg/100 mg, days 1–5). Venetoclax was initiated at 100 mg daily for 14 days, with azole-adjusted dosing and subsequent modifications based on tolerability. Responses were assessed per IWG 2023 criteria, typically after 1–2 cycles. Primary outcomes included overall response rate (ORR), event-free survival (EFS), and OS. Logistic regression and Cox proportional hazards models were used to evaluate predictors of response and survival.

Results The cohort had a median age of 68 years; 55% were male. Baseline bone marrow blasts were <10% in 21 (49%) and ≥10% in 22 (51%) patients. Patients received a median of 3 cycles (range, 1–26); 40% received azacitidine and 60% a decitabine formulation. Venetoclax dose modifications or interruptions were required in 70%. Nineteen patients (44%) proceeded to allo-SCT. Common somatic mutations included TP53 (56%), spliceosome genes (SRSF2, SF3B1, U2AF1, ZRSR2; 27% total), DNMT3A (18%), TET2 (16%), ASXL1 (14%), RAS pathway (NRAS/KRAS; 12%), and RUNX1 (11%).

ORR which included complete remission (CR) or CR with incomplete count recovery (CRi) was achieved in 30 (70%) patients; partial remission (PR) in 3 (7%). Four patients (9%) had stable disease or no response, 2 (5%) had progressive disease, and 4 (9%) were not evaluable due to early death or other events. Median time to CR/CRi was 52 days (range, 21–102). Response rates did not differ between HMA backbone. With median follow-up of 34.0 months, median EFS was 8.0 months (95% CI, 5.7–13.2), and median OS was 11.2 months (95% CI, 7.8–13.2). OS was significantly longer in patients undergoing allo-SCT (16.9 months; 95% CI, 8.7–NR) vs. non-transplanted patients (9.0 months; 95% CI, 5.8–NR).

No individual mutation or baseline blast percentage was significantly associated with response. On univariate analysis for OS, older age predicted inferior survival (HR 1.04; 95% CI, 1.01–1.08; p=0.022), while achieving CR/CRi (HR 0.41; 95% CI, 0.19–0.90; p=0.026) and subsequent allo-SCT (HR 0.38; 95% CI, 0.17–0.93; p=0.015) were associated with improved OS. Although not statistically significant, mutations in TP53, ASXL1, and RAS pathway genes trended toward inferior OS, whereas RUNX1, TET2, and DNMT3A mutations trended toward favorable OS. In multivariable analysis, achieving CR/CRi remained the only independent predictor of superior OS (HR 0.39; 95% CI, 0.17–0.92; p=0.031). Baseline blast percentage was not predictive of survival.

Discussion In this single-center cohort of treatment-naïve HR-MDS, the Ven-HMA combination induced high rates of deep responses, with a 70% CR/CRi rate. Despite these responses, overall survival remained modest, which may in part reflect the high-risk molecular landscape of the cohort—particularly the high prevalence of TP53 mutations (56%)—though survival was significantly improved in patients bridged to allo-SCT. No individual mutations or baseline blast percentage significantly predicted response or survival, potentially due to sample size and co-mutation burden. However, we observed trends suggesting prognostic relevance for specific mutations (TP53, ASXL1, RAS vs. RUNX1, TET2, DNMT3A). These findings highlight the utility of Ven-HMA as a bridge to transplant, frequent need for dose modification, and underscore the complex interplay of molecular genetics in determining outcomes, which warrants validation in larger studies.