Abstract
OBJECTIVES: This study aimed to determine whether the anatomical burden of prior percutaneous coronary intervention(PCI) influences long-term outcomes after coronary artery bypass grafting, beyond the impact of intervention presence alone. METHODS: This retrospective study analyzed consecutive patients undergoing coronary artery bypass grafting at a single institution between 2000 and 2024. The inclusion criteria comprised isolated, non-emergent surgery. Patient categorization was based on prior PCI-treated lesions: none, single, or multiple. The primary endpoint was long-term overall survival. The secondary endpoints included cardiac death, myocardial infarction, stroke, heart failure hospitalization, and repeat revascularization. Long-term outcomes were assessed using Kaplan-Meier analysis and Cox multivariable models, adjusting for 26 clinical factors. RESULTS: Of 2,442 patients, 1,205 met the inclusion criteria (755 none, 227 single-lesion, 223 multiple-lesion intervention). Over a median follow-up of 12.0 (interquartile range, 11.3-12.9; maximum: 24.2) years, the multiple-lesion intervention group had higher rates of in-hospital acute kidney injury (34.1% vs. 21.1% vs. 24.2%, P = 0.003). Overall survival differed significantly between groups over the follow-up period (log-rank P = 0.004), with 15-year survival rates of 35.8%, 46.0%, and 48.0% for multiple-lesion, single-lesion, and no prior PCI groups, respectively. After adjustment, multiple-lesion intervention was associated with increased risks of cardiac death (adjusted subdistribution hazard ratio: 1.91), myocardial infarction (2.26), and repeat revascularization (1.92) compared with no prior intervention. CONCLUSIONS: Multiple-lesion PCI was associated with higher long-term risks of cardiac death, myocardial infarction, and repeat revascularization, while stroke risk was similar. Single-lesion PCI showed outcomes comparable to no prior PCI except for higher heart failure hospitalization. These findings require confirmation in larger, multicenter comparative studies to address residual confounding.