Abstract
BACKGROUND: The Triglyceride-glucose (TyG) index, as a surrogate marker of insulin resistance, is independently associated with the severity of coronary artery lesions and the prognosis of coronary heart disease. The investigation aimed to explore the relationship between the TyG index and recurrent revascularization in individuals with type 2 diabetes mellitus (T2DM) resulting from the progression of lesions or in-stent restenosis (ISR) after percutaneous coronary intervention (PCI). METHOD: A total of 633 patients who met the inclusion and exclusion criteria were enrolled and divided into three groups based on the tertiles of the TyG index. The primary endpoint was recurrent revascularization resulting from the progression of lesions or ISR. All-cause death was considered as the competing risk event. Competing risk analysis and Cox regression analysis for predicting recurrent revascularization after PCI were conducted stepwise. Variables were standardized to make the hazard ratio (HR), subdistribution hazard ratio (SHR) and corresponding 95% CI more consistent prior to being used for fitting the multivariate risk model. The predictive ability of the TyG index was evaluated using several measures, including the ROC curve, likelihood ratio test, Akaike's information criteria, category-free continuous net reclassification improvement (cNRI > 0), and integrated discrimination improvement (IDI). Internal validation was conducted through bootstrapping with 1000 resamples. RESULTS: During a median follow-up period of 18.33 months, a total of 64 (10.11%) patients experienced recurrent revascularization, including 55 cases of lesion progression and 9 cases of in-stent restenosis. After controlling for competitive risk events, the TyG index was independently associated with a higher risk of recurrent revascularization [SHR:1.4345, (95% CI 1.1458-1.7959), P = 0.002]. The likelihood ratio test and Akaike's information criteria showed that the TyG index significantly improves the prognostic ability. Additionally, adding the TyG index improved the ability of the established risk model in predicting recurrent revascularization, indicated by a C-index of 0.759 (95% CI 0.724-0.792, P < 0.01), with a cNRI > 0 of 0.170 (95% CI 0.023-0.287, P < 0.05), and an IDI of 0.024 (95% CI 0.009-0.039, P = 0.002). These results remained consistent when the models containing TyG index were confirmed using an internal bootstrap validation method. CONCLUSION: The findings highlight the potential of the TyG index as a predictor of recurrent revascularization. Lesion progression emerged as the primary contributor to recurrent revascularization instead of in-stent restenosis. The incorporation of the TyG index into risk prediction models is likely to be beneficial for accurate risk stratification in order to improve prognosis.