ASXL1 mutations are associated with adverse outcomes in european leukemianet (ELN) 2024 favorable risk patients with Acute Myeloid Leukemia (AML) treated with venetoclax (VEN) containing low-intensity therapy (LIT)

Abstract

Abstract Background ASXL1 mutation (mut) is frequent in myeloid malignancies, and has adverse prognostic implications in AML treated with intensive chemotherapy. However, its prognostic significance is controversial with venetoclax (VEN) containing lower intensity (LIT) therapy. The four-gene genetic risk classifier by the ELN (ELN2024) for patients with AML receiving LIT ± VEN, categorizes ASXL1 mut/s as favorable risk in the absence of FLT3-ITD, RAS, and TP53 mut/s. Herein we analyzed the prognostic impact of ASXL1 mut/s in newly diagnosed (ND) AML treated with LIT+VEN. Methods We retrospectively analyzed adults (≥18 years) with ND AML treated with LIT+VEN at MDACC between Nov 2012 and Dec 2023. Patients with AML from an antecedent treated (with hypomethylating agent and/or chemotherapy) hematologic neoplasm were excluded. Mutational profiling was performed using an in-house 81-gene next-generation sequencing (NGS) panel with a variant allele frequency (VAF) sensitivity of ≥2%. Outcomes included composite complete remission (CRc), overall response rate (CRc+MLFS), median relapse-free survival (mRFS), and median overall survival (mOS). Stratified Cox multivariate analysis (MVA) was done to assess factors associated with survival. Patients with AML with pathogenic and likely pathogenic ASXL1 mut/s with VAF ≥5% were considered to have ASXL1 mut AML. Results Among 554 patients with ND AML treated with LIT+VEN, 73 (13%) had an ASXL1 mut/s. ASXL1 mut/s were more common in older individuals (median age 74 vs. 71 years, p=0.005), males (70% vs. 57%, p=0.032), and clinical secondary AML (34% vs. 16%, p<0.001) but rarer in therapy-related AML compared to the ASXL1 wildtype (WT) patients (8% vs 28%, p<0.001). ASXL1 was most frequently co-mut with SRSF2 (42%), TET2 (30%), and N/KRAS (30%), while NPM1 and FLT3-ITD co-muts were rare (5% and 4% respectively). Adverse risk cytogenetics were present in 14 (33%) of the ASXL1 mut, similar to ASXL1 WT disease (25%, p=0.33). Using the ELN2024 risk stratification system, patients with ASXL1 mut/s were more likely to be classified as favorable risk (62% vs. 40%) and less likely to be adverse (8% vs. 35%, p<0.001). In the entire cohort, 31% (n=170) received cladribine/low-dose cytarabine/VEN (CLAD-LDAC-VEN), and 69% (n=384) received a hypomethylating agent (HMA)/VEN backbone which was not statistically different between patients with and without ASXL1 mut AML (p=0.231). Among patients with FLT3-ITD, RAS, and TP53 WT disease (i.e. ELN2024 favorable), those with ASXL1 mut/s had significantly lower CRc rates (69% vs. 90%, p<0.001) and shorter mRFS and mOS (14.9 vs. 26.2 months [mos], p=0.09 and 16.2 vs. 35.4 mos, p=0.003, respectively) and their outcomes were comparable to patients classified as ELN2024 intermediate risk (mRFS 14.9 vs. 11.7 mos, p=0.95; and mOS 16.2 vs. 13.3 mos, p=0.89). The mOS was not different in this ELN2024 favorable group based on the presence or absence of concurrent DNMT3A+TET2 mutation with ASXL1 mut (14.8 vs. 17.0 mos, p=0.80). ASXL1 mut-ELN2024 favorable risk patients treated with CLAD-LDAC-Ven had a similar mOS of 17.0 mos compared to 14.8 mos with HMA+VEN doublet/triplets (p= 0.87), but both fared worse than their respective ASXL1 WT ELN2024 favorable risk patients (mOS NR vs. 17.0 mos, p=0.02 in CLV treated; 22.8 vs 14.8 mos, p=0.15 in HMA+VEN treated). In ELN2024 intermediate or adverse risk disease, there was no significant difference in mRFS or mOS between ASXL1 WT and ASXL1 mut, although this comparison was limited by smaller patient numbers in the ASXL1 mut group. With a median follow up of 34.5 mos, among patients with ASXL1 mut AML, survival outcomes did not differ based on ASXL1 VAF (mOS 17.0 mos with VAF 5-20% vs. 14.8 mos with VAF >20%, p=0.87). On MVA limited to ELN2024 favorable-risk patients, ASXL1 mut/s were independently associated with inferior OS (HR 1.65; 95% CI: 1.06–2.56; p=0.03), after adjusting for age, secondary and therapy-related disease, adverse cytogenetics, co-muts in IDH1, IDH2, or NPM1, and allogeneic SCT. Conclusions In our large retrospective analysis, ASXL1 mut/s were associated with lower response rates and worse survival among patients in the ELN2024 favorable risk strata treated with LIT+ VEN and was comparable to ELN2024 intermediate risk disease. Our data suggest that patients with ASXL1 mut AML may be best regarded as having intermediate risk disease even in the absence of RAS and FLT3-ITD mut/s when treated with LIT+VEN.