Platelet-Related Gene Signature Predicts Prognosis, Immune Landscape, and Drug Sensitivity in Acute Myeloid Leukemia.

Acute myeloid leukemia (AML) remains challenging to treat due to clinical heterogeneity and a lack of prognostic biomarkers. To address this, we developed a prognostic signature based on platelet-related genes (PRGs). By analyzing transcriptomic data from TCGA-LAML, GSE146173, and Beat AML 2.0 cohorts, we identified and validated an 11-gene signature (PSME2, PPIF, SYTL4, S100A4, CCND3, SMIM15, PARVB, STXBP5, KCNMB1, GABRE, SLC50A1) using LASSO-Cox regression. This model effectively stratified patients into high- and low-risk groups with distinct survival outcomes (p < 0.001) and demonstrated high predictive accuracy (1-/3-/5-year AUC: 0.832/0.782/0.880). High-risk patients exhibited immunosuppressive features, including upregulated immune checkpoints (CD274, CTLA4, HAVCR2, LAG3, PDCD1LG2, PDCD1), prominent monocyte infiltration, and reduced dendritic`11 cell activity. Drug sensitivity analysis suggested gefitinib, zebularine, and simvastatin as potential therapies for high-risk AML (p < 0.05). We further validated the signature's prognostic value using qPCR and clinical grouping. Notably, in vitro studies indicated that KCNMB1 facilitates AML progression. In conclusion, our robust PRG-based model elucidates the link between platelet biology, immune dysregulation, and therapeutic vulnerability in AML, offering clinical utility for risk stratification and treatment decisions.