Impact of venetoclax dosing on outcomes in AML: A single-center real-world study.

Abstract

e18511 Background: VIALE-A established Aza and Venetoclax (Ven) as first-line therapy (1L) for acute myeloid leukemia for patients (pts) ineligible for intensive chemotherapy. Real-world (RW) use of Ven often differs in dose, schedule, timing of bone marrow biopsy (Bmbx) to assess response, and use of concomitant medications. We examined RW outcomes with 1L Ven at our center. Methods: Our cohort was diagnosed from 2018-2023. Chart abstraction was completed for 160 pts (data cutoff (DCO) 1 Dec 2025) to capture Ven dosing and duration, mutations, concomitant medications, Bmbx, and hematologic adverse events (HAE). Duration of remission (DoR) was defined as time from 1st composite complete response (CRc, per IWG criteria) to relapse, death, or start of 2L, with censoring at transplant or DCO. Ven dose intensity was calculated from dose, duration, and CYP3A inhibitor use. Differences were assessed using the log-rank test, with p<0.05 considered significant. Results: 863 cycles were recorded for 160 pts, with median follow-up of 191 days (d) (range 3-1758) and a median of 2 Ven cycles (1-33). There were 317 Bmbx for 134 pts. Of this number, 88 pts achieved CRc at a median of 41.5d (20-565) and 11 remained in remission at DCO. Overall CRc rate was 55% (66% for pts with Bmbx), with a median DoR of 155.5d (5-1341). Of pts achieving CRc, time to CRc differed significantly across Ven exposure groups (p<0.01). Patients receiving 21 +/- 3 d of Ven in cycle 1 (C1) had the highest CRc rate (68%); cumulative incidence of remission (CIR) was 27% at 28d (p=0.02), with median time to CRc of 34d (20–132) and median DoR of 176.5d (5–975). DoR did not differ across Ven exposure groups (p=0.21). Differing Ven dose intensities in C1 did not predict differing times to CRc (p=0.61) or DoR (p=0.45). 45% of pts with TP53 mutations (mut) achieved CRc vs 67% with RAS mut. Pts with neither mut had significantly longer DoRs (median 176.5d; 9-1341; p<0.001). 92% of CRc occurred by end of C3, 99% by end of C5. Pts experiencing HAE achieved CRc more frequently (71%) than those without and had a higher 28d CIR (21% vs 7%, p=0.02) and trend towards longer DoRs (p=0.07). The proportion experiencing HAE per cycle plateaued by C2 (56%). The median effective Ven dose decreased with increasing Ven cycles (50% reduction by C3), while median cycle length remained stable at 42d (14-1168). Conclusions: In our RW study, Ven cycles averaged 42d with frequent delays due to HAE. Over half of pts achieved remission lasting ~5 months (155.5d), with >10% still in remission at DCO. Mutation status significantly influenced DoR, whereas Ven duration >21 d or C1 dose intensity did not influence time to CRc or occurrence. Despite RW treatment variability, outcomes were comparable to VIALE-A. These results support the use of 21d Ven duration during induction to optimize the balance between achievement of CRc and DoR. The observed association between HAE and CRc is interesting and warrants further study as a potential prognostic marker.