Prognostic impact of co-occurring mutations in NPM1-mutated Acute Myeloid Leukemia

Abstract Background: NPM1 is the most commonly mutated gene in adult acute myeloid leukemia (AML). Per ELN 2022, NPM1-mutated AML is classified as favorable risk, except when co-occurring with FLT3-ITD. However, the prognostic impact of other co-occurring mutations in this group remains uncertain. Methods: We retrospectively reviewed adult AML patients (≥18 years) diagnosed and treated at the University of Kentucky Markey Cancer Center between January 2015 and June 2025. Patients with acute promyelocytic leukemia were excluded. Demographic, clinical, laboratory, and pathologic data were abstracted from medical records. NPM1 and co-mutation analysis were performed using an in-house validated, CLIA-certified 97-gene next-generation sequencing (NGS) panel with a 5% variant allele frequency detection threshold. AML was diagnosed per 2016 WHO criteria. Treatment response was assessed per ELN 2022 criteria. Primary endpoints were complete remission (CR) or CR with incomplete hematologic recovery (CRi), relapse-free survival (RFS), and overall survival (OS). RFS and OS were estimated using the Kaplan–Meier method. Multivariate analysis using logistic regression identified predictors of CR/CRi and relapse, while Cox proportional hazards modeling was used to evaluate factors associated with overall survival. Survival analyses were censored at the time of allogeneic stem cell transplantation. A p-value <0.05 was considered significant. Statistical analyses were conducted using R version 4.4.1. Results: Among 457 patients with newly diagnosed AML, 104 (22.6%) had NPM1 mutations and were included in the analysis. The median age was 57 years (range, 20–85), and 39% were male (n=41). At diagnosis, 77% had de novo AML (n=84), 12% were therapy-related (n=13), and 11% were secondary AML (n=11). The majority had a normal karyotype (n=87, 88%). Co-mutations were highly prevalent: 97 patients (93%) carried at least one additional mutation, with a median of 2 (range, 0–6). The most frequent were FLT3 (n=48, 46%; FLT3-ITD: n=30, 29%; FLT3-TKD: n=18, 17%), DNMT3A (n=34, 33%), IDH2 (n=28, 27%), IDH1 (n=17, 16%), TET2 (n=17, 16%), NRAS (n=15, 15%), PTPN11 (n=14, 14%), SRSF2 (n=8, 8%), SF3B1 (n=7, 7%), and ASXL1 (n=6, 6%). Induction regimens included intensive chemotherapy (with FLT3 inhibitors when indicated) in 61 patients (59%), HMA plus venetoclax in 21 (20%), single-agent HMA in 9 (9%), and Vyxeos in 6 (6%). Overall, 82 patients (79%) achieved CR/CRi. At a median follow-up of 49.9 months, 26 patients (32%) relapsed. Twenty-eight patients (27%) proceeded to allogeneic stem cell transplant after achieving remission with salvage therapy. For the entire cohort, the median RFS was 29.5 months (95% CI: 17–89.7), and the median OS was 51.8 months (95% CI: 18.0– 102.3). At last follow-up, 55 patients (53%) remained alive. All patients without co-mutations (n=7, 7%) achieved CR/CRi, with no deaths observed at last follow-up. Each additional co-mutation was associated with significantly lower odds of achieving CR/CRi (OR 0.20, p=0.03). ASXL1 was specifically associated with reduced CR/CRi rates (OR 0.3, p=0.02). For RFS, worse outcomes were linked to higher co-mutation burden (HR 1.28, p=0.05), TET2 mutations (HR 2.52, p=0.01), and ASXL1 mutations (HR 3.01, p=0.02). For OS, adverse prognostic factors included higher co-mutation burden (HR 1.55, p=0.0004), ASXL1 (HR 3.65, p=0.003), and TET2 (HR 2.89, p=0.001). Conclusion: In this large single-center cohort, while NPM1 mutation is generally considered a favorable prognostic marker, the presence of co-mutations, particularly ASXL1 and TET2, was associated with inferior remission rates, shorter relapse-free survival, and reduced overall survival. These findings highlight the importance of incorporating co-mutation profiles into risk stratification and treatment decision-making for patients with NPM1-mutated AML.