IDH1 mutation predicts response and survival in treatment-naïve Acute Myeloid Leukemia patients receiving with venetoclax with a hypomethylating agent

Abstract Introduction Venetoclax + hypomethylating agent (Ven-HMA) is currently the standard frontline therapy for older/unfit patients (pts) with newly diagnosed acute myeloid leukemia (ND-AML) based on VIALE-A trial results. Post-hoc analyses of the VIALE-A trial showed that the incremental benefit of adding VEN to HMA-based chemotherapy was not uniform and varied based on the presence of distinct mutations, such as FLT3, IDH1/2 and TP53. However, real world studies have shown less robust outcomes and data on the genomic landscape of HMA-VEN therapy is limited. The objective of the current study was to examine Ven+HMA treatment outcomes including the impact of mutations on survival in ND-AML pts in routine clinical practice. Methods We conducted a retrospective, single-center study of ND-AML pts treated with Ven-HMA at the John Theurer Cancer Center. To be included, pts required an available baseline next-generation sequencing (NGS) panel. Cytogenetic and molecular profiling were performed using conventional karyotyping and NGS, respectively. All pts received at least one cycle of HMA therapy, consisting of either azacitidine (75 mg/m²) or decitabine (20 mg/m²), combined with venetoclax, typically dosed at 100mg daily with posaconazole prophylaxis. Genetic risk and treatment response were defined by the 2022 European LeukemiaNet (ELN) criteria. The timing of response assessment was at the discretion of the treating physician. Key outcomes analyzed were composite complete remission (CR/CRi), event-free survival (EFS), and overall survival (OS). Logistic regression and Cox proportional hazards models were used to identify predictors of response and survival. Results A total of 78 ND-AML pts (median age 77 years, 53% male, 53% de novo, 47% secondary or therapy-related) received a median of 4 cycles of venetoclax and hypomethylating agent. All pts were cytogenetically annotated and ELN 2022 cytogenetic risk included favorable (9%, n = 7), intermediate (27%, n = 21) or adverse (64%, n = 49). A total of 7pts (10%) underwent transplantation. CR/CRi was documented in 62 (79%) pts and partial remission in 2 (3%) pts. 4 (5%) pts were refractory or had progressive disease while 10 (13%) pts could not be assessed for response due to death or other event. The median time to CR/CRi was 50 days (16-202). The CR/CRi rate did not differ significantly based on the HMA used or by ELN 2022 genetic risk stratification. In a univariate analysis of response predictors, the presence of a TP53 mutation was associated with an inferior response rate (60% vs. 86%; p=0.016), while venetoclax dose adjustment predicted a superior response (89% vs. 36%; p=0.001). In the multivariable model for response, TP53 status lost significance, whereas venetoclax dose adjustment remained a significant predictor. After a median follow-up of 27.89 months, 54 (69%) deaths were documented. Median OS was 11.47 months (95% CI 7.39-15.31) and median EFS was 7.62 months (95%CI 5.78-11.47). Based on ELN 2022 genetic risk, median OS was not statistically significant with 17.28 in favorable risk, 15.31 in intermediate risk while shortened in adverse risk patients with 9.43 months (p=0.197). In univariate analysis for survival, TP53 mutation (HR 2.41; p=0.003) and secondary AML (HR 1.83; p=0.030) predicted inferior OS, while achieving CR/CRi (HR 0.14; p0.001), venetoclax adjustment (HR 0.38, p=0.003), IDH1 mutation (HR 0.15; p=0.008), and NPM1 mutation (HR 0.41; p=0.029) predicted superior OS. In the multivariable analysis for survival, only achieving CR/CRi (HR 0.15, p0.001) and the presence of an IDH1 mutation (HR 0.21, p=0.037) remained independent predictors of superior survival. Discussion In this real-world analysis of older/unfit ND-AML pts, Ven-HMA therapy yielded high remission rates. Our study identifies achieving a deep clinical response (CR/CRi) and the presence of an IDH1 mutation as independent predictors of superior survival. This finding further deepens the treatment sequencing questions that remain in elderly and unfit individuals with IDH1-mutated AML and underscores the need for a randomized trial to address the best frontline regimen for IDH1 mutant individuals. The primary limitations of this study include its retrospective design, sample size, and variability in the timing of response assessment. Despite these limitations, this study supports further development of treatment-specific, genetics-enhanced, and response-based prognostic tools in AML.