Abstract
BACKGROUND: The Global Registry of Acute Coronary Events (GRACE) score derived from clinical parameters at the time of hospital discharge is a powerful predictor of long-term mortality and reinfarction after acute coronary syndrome (ACS). The triglyceride glucose (TyG) index, which is a simple and reliable surrogate marker of insulin resistance, has been demonstrated to be an independent predictor of long-term adverse major adverse cardiac events, irrespective of diabetes mellitus. We investigate whether the addition of the TyG index improves the predictive ability of the GRACE score after percutaneous coronary intervention (PCI) in ACS patients regardless of diabetes mellitus. METHOD: A retrospective cohort of 986 ACS patients undergoing PCI was enrolled in the present analyses. The GRACE score for discharge to 6 months and the TyG index were calculated. The primary endpoint was the composite of MACEs, including all-cause death and nonfatal myocardial infarction. Patients were stratified according to the primary endpoint and the tertiles of the TyG index. Cumulative curves were calculated using the Kaplan-Meier method. Multivariate Cox regression was adopted to identify predictors of MACEs. The predictive value of the GRACE score alone and combined with the TyG index or fasting blood glucose (FBG) was estimated by the area under the receiver‑operating characteristic curve, likelihood ratio test, Akaike's information criteria, continuous net reclassification improvement (NRI), and integrated discrimination improvement (IDI). Internal validation was assessed using the means of bootstrap method with 1000 bootstrapped samples. RESULTS: During a median follow-up of 30.72 months ((interquartile range, 26.13 to 35.07 months), 90 patients developed MACEs, more frequently in the patients with a higher TyG index. Multivariate Cox hazards regression analysis found that the TyG index, but not FBG was an independent predictor of MACEs (hazard ratio 1.6542; 95% CI 1.1555-2.3681; P = 0.006) in all types of ACS regardless of diabetes mellitus when included in the same model as GRACE score. Furthermore, Kaplan-Meier analysis revealed that the incidence of the primary endpoint rose with increasing TyG index tertiles (log-rank, P < 0.01). Adjustment the GRACE score by the TyG index improved the predictive ability for MACEs (increase in C-statistic value from 0.735 to 0.744; NRI, 0.282, 95% CI 0.028-0.426, P = 0.02; IDI, 0.019, 95% CI 0.004-0.046, P = 0.01). Likelihood ratio test showed that the TyG index significantly improved the prognostic ability of the GRACE score (χ(2) = 12.37, 1 df; P < 0.001). The results remained consistent when the models were confirmed by internal bootstrap validation method. CONCLUSION: The TyG index, but not FBG is an independent predictor of long-term MACEs after PCI in all types of ACS patients regardless of diabetes mellitus after adjusting for the GRACE score, and improves the ability of the GRACE score to stratify risk and predict prognosis of ACS patients undergoing PCI.