Hasan GÖZE, Tuna Eren ESEN, İsa YALÇINKAYA, Gökhan BURUL, Tahir Alper CİNLİ, Mesut AYER, İstemi SERİN
Experimental Biomedical Research - 2026;9(4):251-262
Aim: To determine whether next-generation sequencing (NGS) improves 2022 European LeukemiaNet (ELN) risk stratification by identifying myelodysplasia-related (MDS-related) mutations in acute myeloid leukemia (AML). Methods: We analyzed 111 consecutive adults with newly diagnosed non-promyelocytic acute myeloid leukemia (AML) who underwent diagnostic NGS with a myeloid gene panel. For each patient, an ELN 2022 risk category was assigned using four diagnostic modalities -metaphase karyotyping, fluorescence in situ hybridization (FISH), targeted molecular (PCR) testing, and NGS. A conventional category was defined as the most adverse assignment obtained from karyotyping, FISH, and PCR, whereas the NGS-inclusive category was defined as the most adverse assignment obtained when NGS was added to these modalities. The primary outcome was within-patient reclassification attributable to the incorporation of NGS. Reporting followed the STROBE guidelines. Results: The median age was 56 years, and 65.8% of patients were male. The most frequently mutated genes were FLT3 (25.2%), NPM1 (19.8%), TET2 (17.1%), TP53 (16.2%), and RUNX1 (16.2%); 38.7% of patients harbored at least one MDS-related mutation. Among the 110 patients with both classifications available, incorporation of NGS upgraded the ELN 2022 category in 33 patients (30.0%; 95% CI, 22.2-39.1) -29 from intermediate and 4 from favorable to adverse -with no downgrades. The proportion of patients classified as adverse risk more than doubled, from 30 (27.3%) to 63 (57.3%) (McNemar p < 0.001). Upgrades were driven predominantly by MDS-related mutations: 29 of 33 upgraded patients (87.9%) carried at least one such mutation, most commonly in ASXL1 (45.5%), RUNX1 (39.4%), SRSF2 (30.3%), or STAG2 (24.2%). Conclusions: Even among patients selected for NGS testing, an ELN 2022 category based on conventional cytogenetic and PCR testing alone was inaccurate, almost invariably too favorable, in approximately one in three patients, and no clinical feature reliably identified them. NGS is therefore indispensable for accurate risk stratification in AML.
