Time to response with HMA/venetoclax as a predictor of survival in acute myeloid leukemia: Analysis of 1- vs 2-cycle responders.

Abstract

6548 Background: Hypomethylating agents (HMA) with venetoclax are the standard of care for patients with AML ineligible for intensive chemotherapy. However, the prognostic significance of time to response and the optimal duration of venetoclax, often modified in real-world practice due to myelosuppression, remain unclear. Methods: We conducted a retrospective analysis of adults with AML achieving CR/CRi with HMA/venetoclax at University of Kentucky from 2016-2025. Patients were stratified by cycle of response: 1 vs 2 cycles, using ELN 2022 response criteria. These groups were compared by demographics, cytogenetic and mutational profiles, and allogeneic HSCT. RFS and OS were estimated using Kaplan-Meier methods and compared with log-rank tests. Cox models evaluated associations between cycle of response, response type (CR vs CRi), and survival outcomes. Results: Eighty-six responding patients were included, with 69 (80%) achieving response after 1 cycle and 17 (20%) achieving response after 2 cycles. Median age was 66.7 vs. 67.1 years, male sex was 58% vs. 76%, and ELN risk distribution was similar: favorable 25% vs. 18% and adverse 64% vs. 65% (p=0.671). Frequent cytogenetic profiles included normal karyotype (44% vs. 41%) and complex karyotype (27% vs. 24%), and common mutations included IDH1/2 (23% vs. 18%), DNMT3A (19% vs. 24%), and NPM1 (22% vs. 12%). Overall, baseline demographics and cytogenetic/mutational profile were similar between 1-cycle and 2-cycle responders, supporting comparability of these groups for outcome analyses. With a median follow-up of 23.8 months, relapse occurred in 36% of 1-cycle responders vs. 82% of 2-cycle responders (p=0.002). Inferior RFS was noted in 2-cycle responders (HR 3.04, p=0.002), which remained after adjusting for total venetoclax exposure (HR 3.26, p=0.002). Median OS was 22.8 months (95% CI 14.0–NR) for 1-cycle responders and 12.7 months (95% CI 7.2–NR) for 2-cycle responders (p=0.01). In Cox models, 2-cycle response was associated with significantly inferior OS (HR 2.33, p=0.012), which persisted after adjusting for venetoclax exposure (HR 2.26, p=0.018). Hematologic recovery at time of CR was assessed and was not significantly associated with cycle of response or survival outcomes. Increased venetoclax exposure was associated with improved OS (HR 0.82 per day; p=0.049). When venetoclax exposure was dichotomized, receipt of ≤14 days was associated with inferior OS (HR 2.49; p=0.028). In this cohort, 6% of 1-cycle responders and 18% of 2-cycle responders proceeded to allogeneic HSCT (p=0.14). Conclusions: Delayed response beyond 1 cycle in patients receiving HMA/venetoclax identifies AML patients at higher risk of relapse and inferior survival. Time to response can inform early risk-adapted strategies, including venetoclax optimization.