Background Azacitidine (AZA) combined with venetoclax (VEN) is the standard of care for patients with acute myeloid leukemia (AML) who are ineligible for intensive chemotherapy (1). However, the composite complete remission (CRc) rate remains modest at 66.4%, with a median overall survival (OS) of 14.7 months. Several prognostic models have been proposed to refine risk stratification in this setting, including ELN 2024 (2), refined-ELN 2024 (3), and the Mayo Clinic risk-prognostication model (4). Other functionally based approaches have been developed using immunophenotypic markers (5) or apoptosis mediator scores (6). The objectives of the VENETACIBLE study were to (i) correlate plasma VEN levels with BCL-2 inhibition, (ii) assess the predictive value of BH3 profiling for AZA+VEN response, and (iii) evaluate the relationship between treatment response, plasma VEN levels, and Bcl-2 inhibition.

Methods From February to September 2024, we prospectively collected peripheral blood samples on days 1, 5, and 9 of treatment, and at 1 year or relapse, from newly diagnosed AML patients treated with frontline AZA+VEN (per the VIALE-A protocol (1)) at two centers (Nice University Hospital and Marseille University Hospital) (VENETACIBLE; NCT06030089). Phospho-flow cytometry for BCL-2 phosphorylation was performed on days 1, 5, and 9; BH3 profiling on days 1 and 5; and plasma VEN levels were measured on days 1 and 9. Patient characteristics were analyzed descriptively, and treatment response was evaluated according to the ELN 2022 criteria (7).

Results Twenty patients with newly diagnosed AML ineligible for intensive chemotherapy were included. The majority were male (60%) with an ECOG performance status of 0 (15%) or 1 (60%), and a median age of 79 years (range: 69–90) at baseline.

Genetic risk according to ELN 2022 was favorable in 15%, intermediate in 10%, and adverse in 75%. According to ELN 2024, risk distribution was favorable (45%), intermediate (25%), and adverse (30%). The most frequent gene mutations at diagnosis were ASXL1 (50%), TP53 (30%), TET2 (30%), N/KRAS (25%), and NPM1 (15%). After the first two cycles of AZA+VEN, the CRc rate was 35%.

First, we observed inter-individual variability in VEN plasma levels, with a median concentration on day 9 of 1239 ng/mL [range, 149–4159]. There was a trend toward higher concentrations among responders, with a median of 1503 ng/mL compared to 1259 ng/mL in non-responders (p = ns).

Second, BH3 profiling predicted VEN resistance as early as day 5 of treatment. In the overall cohort, a significant increase in sensitivity to Mcl-1 (MS-1) (p < 0.01) and Bcl-xL (HRK) (p < 0.05) peptides was observed at day 5 post-treatment. Among responders, no significant changes were detected, whereas in non-responders, a significant increase in sensitivity to the Mcl-1 (MS-1) peptide was observed at day 5 post-treatment (p < 0.01).

Finally, the percentage of Bcl-2-expressing cells within the monocyte population was significantly higher in responders compared to non-responders (p = 0.03). A decrease in Bcl-2 phosphorylation levels was identified on days 5 and 9, correlating with clinical response. In contrast, no significant changes were observed within the neutrophil subpopulation.

Conclusion In conclusion, this pilot study identified inter-individual variability in VEN absorption and/or metabolism, which may impact its efficacy. We demonstrated that on-treatment BH3 profiling could predict response to AZA + VEN, and that baseline Bcl-2 expression may serve as a predictive biomarker of treatment response. These preliminary findings warrant validation in a larger cohort.

Bibliography (1). DiNardo et al. NEJM 2020

(2). Döhner et al. Blood 2024

(3). Lachowiez et al. Blood 2024

(4). Gangat et al. AJH 2025

(5). Zhao et al. Blood Adv 2025

(6). Waclawiczek et al. Cancer Discovery 2023

(7). Döhner et al. Blood 2022

This content is only available as a PDF.
2025
Sign in via your Institution