ASXL1 and NOTCH1 mutations independently predict TKI treatment failure in chronic myeloid leukemia

Abstract Background: While tyrosine kinase inhibitors (TKIs) have improved outcomes in chronic myeloid leukemia (CML), a subset of patients still fail to achieve an optimal treatment response. ABL1kinase domain mutations confer TKI resistance, but ABL1mutations are found in only about half of the patients with TKI resistance. Emerging evidence suggests that genomic alterations other than mutations in the ABL1 kinase domain may influence therapeutic response. However, limited studies have specifically addressed the role of such mutations in TKI resistance. In this study, we investigated the impact of high-risk mutations identified by next-generation sequencing (NGS) on treatment outcomes in CML patients. Methods:We retrospectively analyzed 94 CML patients from the First Affiliated Hospital, Zhejiang University School of Medicine, who underwent NGS with an 84-gene myeloid panel at diagnosis. TKI treatment failure was defined according to ELN 2020 guidelines. ASXL1 mutational burden and gene function categories (e.g., histone methylation, transcription factors) were evaluated. Associations between clinical/genomic features and TKI failure were assessed using univariable and multivariable logistic regression. Results:The median age of the cohort (n = 94) was 45.5 years, with 52.13% male. Patients in the TKI-failure group had significantly lower hemoglobin (105.0 vs. 123.0 g/L, p < 0.001) and platelet counts (401.0 vs. 542.0×10⁹/L, p = 0.003) compared to non-failures. According to the Sokal risk score, a higher proportion of patients in the failure group were classified as intermediate/high-risk (76.74% vs 66.67%, p = 0.033). The total number of mutations was significantly higher in the failure group (median: 3 vs. 2, p < 0.001). Excluding ABL1mutations, the most frequent mutations in the failure group were ASXL1(30.23%), NOTCH1(16.28%), WT1(13.95%), CSMD1(11.63%), KMT2C(9.30%), and TCF3(9.30%). ASXL1, NOTCH1, and WT1mutations were significantly enriched among TKI-failure patients, and ASXL1variant allele frequency (VAF) was also significantly higher in this group. Functional annotation revealed a higher prevalence of histone methylation–related mutations in the failure group (20% vs. 13%, p= 0.033). We further investigated mutational burden across different clinical subgroups. Patients over 50 years of age harbored more mutations than younger patients (median 3 vs 2, p= 0.031). Those treated initially with first-generation TKIs had a significantly higher number of mutations than those receiving second-generation TKIs (median 3 vs 1, p < 0.001). Mutation burden was also higher among patients who failed to achieve major molecular response (MMR) within 12 months (median 3 vs 2, p = 0.014). In univariable logistic regression, ASXL1(OR 3.99, P = 0.016) and NOTCH1(OR 9.72, P = 0.037) mutations were significantly associated with TKI failure. In multivariable analysis adjusting for Sokal risk and ABL1status, ASXL1(OR 4.17, P = 0.025) and NOTCH1(OR 12.61, P = 0.029) remained independent predictors. When grouped by functional annotation, histone methylation–related mutations were predictive in both univariable (OR 2.54, p = 0.035) and multivariable analyses (OR 2.74, p = 0.037). Transcription factor mutations also showed significance in univariable analysis (OR 3.56, p = 0.029). Conclusions: ASXL1and NOTCH1mutations independently predict TKI treatment failure in CML, underscoring their potential role in resistance beyond ABL1. Additionally, higher mutational burden was associated with older age, use of first-generation TKIs, lack of 12-month MMR, and eventual TKI failure. Acknowledgement: This research was funded by the Zhejiang Provincial Health High-level Innovative Talent Project (2022-2026). *Correspondence to: Jian Huang, M.D., Ph.D., Department of Hematology, The First Affiliated Hospital of Zhejiang University School of Medicine, Hangzhou 310003, Zhejiang, China. E-mail:[email protected]