[Comparison on the long-term outcomes post percutaneous coronary intervention or coronary artery bypass grafting for bifurcation lesions in unprotected left main coronary artery].

To compare the long-term clinical outcomes of consecutive patients treated with coronary artery bypass grafting (CABG) or percutaneous coronary intervention(PCI) with drug-eluting stents (DES) for bifurcation lesions in unprotected left main coronary artery (ULMCA). A total of 663 consecutively patients with unprotected left main bifurcation disease (defined as stenosis≥50%) who received either drug-eluting stents (DES) implantation (=316) or underwent CABG (=347) between January 2003 and July 2009 in Beijing Anzhen Hospital were enrolled retrospectively in this study.The endpoints of the study were death, repeat revascularization, myocardial infarction (MI), stroke, the composite of cardiac death, MI or stroke and MACCE (major adverse cardiac and cerebrovascular events, the composite of cardiac death, MI, stroke or repeat revascularization). To reduce the bias between two compared groups , our study applid propensity score matching by logistic regression.Cumulative survival rate was calculated by Kaplan-Meier method.Multiple regression analyses, hazard ratio() and 95% were tested by Cox proportional hazard models with the CABG group as the reference category. The median follow-up was 7.2 years (interquartile range 5.1 to 8.3 years) in the overall patients.Unadjusted analysis showed that MACCE rate (PCI 44.5% vs. CABG 45.7%, =0.036), the rate of repeat revascularization (PCI 33.7% vs. CABG 19.5%, <0.001), the composite rate of serious outcomes (cardiac death, MI or stroke) (PCI 20.4% vs. CABG 34.3%, =0.032), stroke rate (PCI 3.7% vs. CABG 23.5%, <0.001) were significantly different for patients undergoing PCI or CABG. Death rate (PCI 13.0% vs. CABG 18.9%, =0.12) and MI rate (PCI 11.3% vs. CABG 6.1%, =0.108) were similar between PCI and CABG group.After adjusting covariates with multivariate Cox hazard regression model, there was no significant difference in rates of death (=0.286), MI(=0.165) and the composite rate of serious outcomes (cardiac death, MI or stroke) (=0.305) between the two groups. Patients in PCI group suffered significant higher rates of MACCE(=0.011), mainly driven by the significant higher rates of repeat revascularization(=2.46, 95% 1.662-3.642, <0.001). However, stroke rate was still significantly higher in CABG group than in PCI group(=0.001)after multivariate adjusting. After propensity score matching (PSM), there was no more significant difference for all covariates in the matched cohorts (202 pairs). Further PSM analysis showed that overall findings were consistent with multivariate Cox hazard regression model except for MI (PCI 12.7% vs. CABG 3.8%, =0.039). During a follow-up up to 8.3 years, the survival rate is similar between the PCI and the CABG group in patients with unprotected left main bifurcation disease.The rate of repeat revascularization is significantly higher whereas stroke rate is significantly lower in the PCI group compared to CABG group.After propensity score matching, PCI group not only had a significant higher rate of repeat revascularization, but also had significantly higher risk of MI in the matched cohorts; while they did not seem to translate into any disparity of mortality in ULMCA bifurcation disease patients.Accordingly, PCI for ULMCA disease can be used as a reasonable treatment option alternative to CABG.