Development of a prognostic gene model for myocardial infarction

Ischemic cardiomyopathy after myocardial infarction frequently causes adverse cardiac events. The absence of well-validated predictive models limits clinical risk-stratification practice. We aimed to construct a gene-based prognostic model for post-myocardial-infarction MACE integrating gene-expression signatures and clinical factors. This multicenter prospective cohort study enrolled AMI patients from June 2021 to January 2024, including a recruited model‑construction cohort ( n = 291) and an external validation cohort ( n = 100). After excluding cases with missing data, 283 and 99 patients were finally analyzed, respectively. LASSO-penalized and multivariate Cox regression were used to screen prognostic genes and clinical predictors. Model performance was assessed via C-index, time-dependent ROC curves, calibration curves, and decision curve analysis. Kaplan–Meier survival analysis was performed for risk stratification. Three hub genes (FADS2, TMEM176A, RPS4Y1) and clinical predictors were integrated into a prognostic nomogram. In the development cohort, this clinical-gene combined model attained an optimism-corrected C-index of 0.843, with a 12-month time-dependent AUC of 0.901. In the external validation cohort, the model yielded a C-index of 0.807 and 12-month time-dependent AUC of 0.819 (95% CI: 0.691–0.906). Incorporation of RPS4Y1 improved predictive performance (ΔC-index = 0.0419). We observed no significant sex × RPS4Y1 interaction. Time-dependent ROC, calibration curves and DCA demonstrated favorable discrimination, acceptable calibration and promising clinical net benefit. The established clinical-gene nomogram provides a promising and externally validated prognostic prediction for post-AMI MACE. It offers a promising tool for early risk stratification of AMI patients and supports further functional exploration of these hub genes.