Abstract Introduction: Venetoclax, a selective BCL-2 inhibitor combined with a hypomethylating agent (azacitidine or decitabine), has become the most commonly used standard of care first-line treatment approach for patients with AML who are not candidates for intensive chemotherapy. While this doublet therapy is a sizeable advance over single-agent induction therapy, response rates and duration can vary significantly, and there is little real-world data on outcomes. Furthermore, recent updates to the ELN risk stratification schema are based on specific resistance mutations, rather than the historical standard of cytogenetic and molecular risk stratification. We aimed to examine real-world outcomes in our registry of patients treated with venetoclax, to assess long-term outcomes, and predictors of outcomes, in the real-world setting. Methods: We performed a retrospective cohort study of patients diagnosed with AML between November 30, 2018 and December 31, 2023 who received venetoclax-based regimens at our center as first-line therapy. Our goal was to assess the following outcomes: overall survival (“OS”; defined as months between venetoclax initiation and death, loss to follow-up, or May 1, 2025 study censor date), event-free survival (“EFS”; defined as months between venetoclax initiation and disease progression, relapse, death, loss to follow-up, or study censor date), duration of first response, and duration of remission. We compared these outcomes between 3 groups stratified by 2022 European LeukemiaNet (ELN) genetic risk category (favorable, intermediate, or poor/adverse). We also compared outcomes based on the presence or absence of resistance-inducing mutations (TP53, FLT3-ITD, KRAS, or NRAS; “VEN-RES”). Finally, we compared outcomes based on the presence or absence of more venetoclax-sensitive mutations (NPM1, IDH1, IDH2, RUNX1, or bzip CEBPA; “VEN-SEN”). Results: Our registry included 160 patients (59% male, 71% Caucasian) with median age of 72 (range 28 to 91) at diagnosis. When comparing patients based on the 2022 ELN genetic risk categories (favorable (n=9), intermediate (n=23), or poor/adverse (n=128)), there was no significant difference in OS, EFS, duration of first response, or duration of remission. However, when comparing patients with (n=86) and without (n=63) a resistance-inducing mutation (VEN-RES), those with VEN-RES had lower OS compared to those without VEN-RES (4.05 vs 9.48; p0.01). Patients with VEN-RES also had lower EFS compared to those without VEN-RES (3.28 vs. 6.66; p0.01). There was no significant difference in duration of first response in those with and without VEN-RES (5.80 vs 7.84; p=0.18), or duration of remission (5.34 vs 7.25; p=0.23). In contrast, when comparing patients with (n=77) and without (n=65) at least one venetoclax-sensitive mutation (VEN-SEN), there was a significant difference in EFS (6.23 vs 2.95; p=0.046). However, there was no significant difference in OS (8.46 vs 3.05; p=0.14), duration of first response (7.57 vs 3.41; p=0.86), or duration of remission (7.31 vs 3.08; p=0.84), between those with and without VEN-SEN, respectively. We also found a statistically significant decrease in OS in patients with complex or monosomal karyotypes (n=61) compared to patients without these cytogenetic abnormalities (n=87) (3.44 vs 8.46; p0.01), along with decreased EFS (3.18 vs 6.23; p0.01). There was no significant difference in duration of remission or duration of best response, according to the presence or absence of complex or monosomal karyotype. Discussion: These real-world results support the recently updated 2024 ELN risk stratification schema for venetoclax-based therapy in AML over the historical standard of stratifying patients based on cytogenetic and molecular risk category. For this cohort of patients treated with venetoclax-based regimens in the first-line setting, we found no difference in outcomes by 2022 ELN risk strata but saw notable differences in outcomes (OS and EFS) based on the presence or absence of mutations known to confer resistance to venetoclax as well as differences in EFS based on the presence or absence of venetoclax-responsive mutations.
Mitchell et al. (Mon,) studied this question.