High remission rates with FLAG-ida-venetoclax in acute myeloid leukemia: “Less is More” on days of venetoclax in the intensive regimen

Abstract Background: Despite significant advances in treatment of AML over the past decade, long-term survival remains poor. The addition of venetoclax (Ven) to intensive chemotherapy backbones has been evaluated in multiple early-phase clinical trials demonstrating high response rates. Based on encouraging clinical activity, the FLAG-Ida-Ven regimen is being increasingly adopted in practice outside the context of clinical trials. However, the real-world outcomes have not been well documented. Methods: Medical records of adult patients (pts) with AML (n=32) treated with FLAG-Ida-Ven in the frontline (F) or relapsed/refractory (RR) setting at Vanderbilt University Medical Center between Sept 2021 and Apr 2025 were reviewed. Median overall survival (mOS) and median event-free survival (mEFS) were assessed using Kaplan-Meier method. Results: Ten pts were treated in the frontline setting (F1), 7 received FLAG-Ida-Ven as second induction (F2) for residual disease following initial induction chemotherapy (IC), and 15 pts were treated in the RR setting. In the F1 cohort (n=10), median age was 34 years (y). Eight pts (80%) had ELN'22 adverse risk disease and 2 (20%) were classified as Acute Undifferentiated/Mixed Phenotype Acute Leukemia (AUL/MPAL). Eight pts (80%) received 7 days (Ven7d) and 2 (20%) received 14 days (Ven14d) of Ven during induction. In the F2 cohort (n=7), median age was 58y. Six pts (86%) were ELN'22 adverse risk. Prior to the start of FLAG-Ida-Ven, 3 pts (43%) had >20% residual blasts, 2 (29%) had 10-19%, and 2 (29%) had 5-9% at a median of 13d from start of initial IC. Six pts (86%) received Ven14d. In the RR cohort (n=15), median age was 56y. One pt (7%) had AUL/MPAL. Seven cases (47%) were classified as ELN'22 adverse risk, 4 (27%) as intermediate, and 4 (27%) as favorable risk at the time of diagnosis. Five pts (33%) had prior Ven exposure and 2 (13%) had RR AML post allo-transplant (SCT). Eight pts (53%) received Ven14d and 7 (47%) received Ven7d. All pts experienced G4 neutropenia and thrombocytopenia. Median time to neutrophil recovery was significantly prolonged in F2/RR compared to F1 (34d vs. 23d; p= 0.025), and in Ven14d vs. Ven7d cohorts (35d vs. 23d; p <0.001); 6/7 pts who did not recover ANC to >1000 were in F2 or RR cohorts. Febrile neutropenia was common in all cohorts (F1: 80%, F2: 86%, and RR: 85%). We did not find any significant differences in the rates of ICU admission for infection among F1 vs. F2/RR and Ven7d vs. Ven14d cohorts. Median time to platelet recovery was similar in F1 vs. F2/RR settings (30d vs. 32d; p= 0.483), with no significant difference between Ven7d vs. Ven14d cohorts (35d vs. 31d; p= 0.504). However, we noted a trend towards incomplete recovery of platelets to >100k in F2/RR and Ven14d cohorts (F2/RR vs. F1: 46% vs. 10%, p= 0.050; Ven14d vs. Ven7d: 50% vs. 19%; p=0.063). Two patients in the Ven14d cohort experienced fatal intracranial hemorrhage during induction. Overall response rates (ORR: CR/CRi/MLFS/PR) were 90% in F1, 86% in F2, and 67% in RR cohorts; CR/CRi rates were 80%, 71%, 53%, respectively. No significant difference in ORR was noted between Ven7d and Ven14d cohorts (75% vs. 81%; p=1.00). Two of 3 pts with AUL/MPAL experienced CR and all 4 pts with KMT2A-r AML achieved CR/CRi with frontline (F1/F2) FLAG-Ida-Ven. Notably, 3 R/R pts (20%) were non-evaluable for response due to early death. Overall, 6 (35%) pts in F1/F2 cohort and 2 (13%) pts in RR cohort underwent SCT in CR post FLAG-Ida-Ven. Relapses occurred in 4 (40%) in F1, 2 (29%) in F2, and 9 pts (53%) in RR cohorts; mEFS was 7.7, 7.3, and 3.8 months (mo) in F1, F2, and RR cohorts, respectively (p=0.371). After a median follow up of 18 mo, mOS was 13.4, 30.5, and 5.9 mo in F1, F2, and RR cohorts (p=0.166), respectively. Among pts with KMT2A-r AML, those with t(9;11) (n=4) had numerically improved mOS compared to those with other translocation partners (n=2) (not reached vs. 2.8 mo; p=0.364). Conclusion: This real-world analysis of FLAG-Ida-Ven demonstrates high response rates in both frontline (first or second induction) and RR settings. Remission rates were high even in high-risk subtypes including ELN adverse risk and AUL/MPAL. Response rates and mOS were comparable between Ven7d and Ven14d. However, prolonged neutropenia, trend of incomplete platelet recovery in pre-treated patients, and fatal bleeding events underscore the need for close monitoring and limiting Ven to 7 days in this regimen.