Only FLT3-ITD co-mutation did not have a deleterious effect on acute myeloid leukemia patients with NPM1 mutation, but concomitant with DNMT3A co-mutation or a < 3log reduction of MRD2 predicted poor survival.

Abstract

Co-occurring mutations are frequently observed in acute myeloid leukemia (AML) with NPM1 mutation, and NPM1 measurable residual disease (MRD) is an effective prognostic biomarker. This retrospective study investigated the impact of gene co-mutations and NPM1 MRD on outcomes in these patients. Among 234 patients, 11.5% carried the rare type NPM1 mutation (NPM1RT). The median age was 49years (IQR 36-58), with a median follow-up of 30.4months (IQR 12.1-55.7). Nine genes were mutated in > 10%, with DNMT3A (53.8%) and FLT3-ITD (44.4%) being most prevalent. Univariable analysis in 137 patients showed FLT3-ITD, DNMT3A co-mutations, and MRD2 < 3 log reduction predicted poorer survival. FLT3-ITD and DNMT3A co-mutations correlated with the lowest event-free (EFS) and overall survival (OS) (3-year EFS 30.0%; 3-year OS 34.4%; both p < 0.001). FLT3-ITD alone did not worsen survival compared to patients without FLT3-ITD. Multivariable analysis identified DNMT3A co-mutation [EFS, HR = 1.9, p = 0.021; OS, HR = 2.2, p = 0.023] and MRD2 ≥ 3 log reduction (EFS, HR = 0.2; OS, HR = 0.1, both p < 0.001) as independent survival predictors. Patients with FLT3-ITD and DNMT3A co-mutations or a MRD2 < 3 log reduction were identified as high risk, but allogeneic hematopoietic stem cell transplantation (allo-HSCT) improved survival significantly compared to chemotherapy only (3-year EFS, 57.9% vs. 30.0%, p = 0.012; 3-year OS, 72.9% vs. 34.4%, p = 0.001). In AML patients with NPM1 mutation, the detrimental impact of FLT3-ITD mutation was exacerbated by DNMT3A co-mutation. Poor-risk younger patients identified by FLT3-ITD and DNMT3A co-mutations or MRD2 < 3 log reduction benefit from allo-HSCT.