Prognostic Value of Composite Biomarkers Integrating Insulin Resistance and Coronary Inflammation for Major Adverse Cardiovascular Events in Post-PCI Patients with Coronary Artery Disease Across Glycemic Status: A Prospective Cohort Study
Abstract
Background The study aimed to assess the prognostic utility of composite Insulin resistance (IR)- various inflammation biomarkers for predicting poor outcomes in coronary artery disease (CAD) patients stratified by glycemic status. Methods The prospective cohort included 1947 CAD patients undergoing PCI with pre-procedural coronary computed tomography angiography. Novel composite biomarkers integrating insulin resistance and inflammation (TyG-hsCRP, CTI, and TyG-fat attenuation index [FAI]) were constructed and calculated. Patients were categorized by glycemic status (normoglycemia, prediabetes, diabetes). Three-year follow-up tracked MACE (all-cause death, non-fatal myocardial infarction, unplanned revascularization). Analyses involved Cox regression, Kaplan-Meier curves, restricted cubic splines, receiver operating characteristic curves, and subgroup analyses. Results MACE occurred in 171 patients (8.8%). After multivariable adjustment, high versus low biomarker levels were associated with increased MACE risk: hazard ratio (HR) 1.23 (95% CI, 1.03–1.74; P=0.025) for TyG-hsCRP, 1.39 (95% CI, 1.04–2.04; P=0.032) for CTI, and 2.32 (95% CI, 1.67–3.25; P<0.001) for TyG-FAI. TyG-FAI demonstrated superior discrimination (AUC 0.641) compared with TyG-hsCRP (0.594), CTI (0.604). Joint stratification revealed the highest risk in patients with high TyG-FAI and diabetes (HR, 3.21 [95% CI, 1.65–6.30]; P=0.001 versus low TyG-FAI and normoglycemia). Similar gradients were observed for high CTI-diabetes (HR, 1.88 [95% CI, 1.13–3.08]; P=0.014) and high TyG-hsCRP-diabetes (HR, 1.88 [95% CI, 1.13–3.08]; P=0.014). Associations were consistent across subgroups (P for interaction >0.05). Conclusion Composite biomarkers integrating IR and inflammation provide modest incremental value in predicting MACE in CAD patients. The predictive utility of these biomarkers is modulated by underlying glycemic status, highlighting the role of local coronary inflammation and supporting their potential integration into risk stratification.