Stemness-driven INTS7 overexpression links metabolic comorbidity to chemoresistance in acute myeloid leukemia with type 2 diabetes.

Acute myeloid leukemia (AML) often coincides with type 2 diabetes mellitus (T2D), yet the impact of metabolic comorbidity on leukaemic biology and prognosis remains underexplored. We integrated two transcriptome-derived stemness indices (mRNAsi, EREG.mRNAsi) with weighted gene-co-expression network analysis (WGCNA) in 137 TCGA-LAML samples. Higher stemness scores clustered with adverse cytogenetics, and a stemness-linked "brown" module yielded 13 candidate genes. We prospectively enrolled 152 newly diagnosed AML patients with confirmed T2D. Peripheral-blood RNA quantification placed patients into INTS7-high and INTS7-low groups by cohort median. INTS7 not only predicted overall survival in TCGA but also retained significance in four independent GEO cohorts (total n = 306). Importantly, GWASdb ranks INTS7 as a significant T2D locus (standardised score ≈ 0.57), suggesting shared pathogenic circuits between dysglycaemia and leukaemia. Baseline demographics, BMI, counts and ELN-2022 risk were balanced, yet complete-remission after induction therapy was 41% versus 96% (p < 0.001). Kaplan-Meier analysis showed that high INTS7 expression halved median overall (18.2 vs 41.3 months) and progression-free survival (10.1 vs 24.8 months); multivariable Cox models confirmed independent adverse impact (HR_OS = 2.05; HR_PFS = 2.33). The effect size exceeded that in unselected AML, indicating that diabetic physiology amplifies INTS7-driven chemoresistance. Our integrative approach-linking stemness metrics, network modelling and comorbidity-focused validation-identifies INTS7 as a biomarker at the nexus of metabolic disorder and leukaemic aggressiveness. Quantifying INTS7 could refine risk stratification and inspire metabolism-tailored therapeutic strategies for the growing AML-T2D population.