Background: Acute myeloid leukemia (AML) remains a prognostically heterogeneous malignancy despite advances in molecular risk stratification. While the 2022 European Leukemia Network (ELN) guidelines refine risk classification, their accuracy in predicting survival outcomes across genetic requires validation.

Methods: We conducted a retrospective analysis of 154 newly diagnosed AML patients at Daping Hospital, integrating next-generation sequencing (NGS)-based genetic profiling, 2022 ELN risk classification, and clinical outcomes.

Results: The most frequent mutations were FLT3 (26.6%), DNMT3A (21.4%), NPM1 (18.2%), CEBPA (17.5%), and TET2 (15.6%). Median overall survival (OS) and progression-free survival (PFS) were 22.9 months and 14.1 months, respectively. Hematopoietic stem cell transplantation (HSCT), female sex, age <60 years, and normal karyotype emerged as favorable prognostic factors. No significant differences observed between allogeneic (Allo-HSCT) and autologous (ASCT). IA±E chemotherapy yielded superior survival, while AZA+BCL2 regimens underperformed. Conversely, TP53 and KIT mutations correlated with inferior survival, while NPM1/CEBPA mutations predicted longer survival. Notably, significant survival heterogeneity existed within ELN 2022 risk groups, particularly among patients with FLT3, CEBPA, or TET2 mutations.

Conclusions: The ELN risk classification demonstrate limitations in prognostication, particularly for patients with FLT3, CEBPA, or TET2 mutations. Our findings highlight the necessity for refined risk models incorporating additional molecular markers (KIT) and mutation interactions to enhance personalized prognostication. Gene coexistence is also a factor that needs to be considered when determining patient prognosis.

Trial registration: The study was registered on the ChiCTR platform (No. ChiCTR2500096484).

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