### Clinical Question
In patients with multivessel coronary artery disease, is PCI with everolimus-eluting stents noninferior to CABG in terms of major adverse cardiovascular events?

### Bottom Line
In patients with multivessel coronary artery disease, PCI with everolimus-eluting stents was not noninferior to CABG with respect to major adverse cardiovascular events at 2 years. Over longer-term follow-up, CABG has a lower rate of major adverse cardiovascular events compared to PCI.

### Major Points


### Guidelines
Current clinical guidelines recommend CABG as the preferred revascularization strategy for patients with complex coronary lesions and without excessive operative risk.

### Design
- Multicenter, prospective, open-label, randomized, noninferiority trial
- N=880 patients with multivessel coronary artery disease
- PCI with everolimus-eluting stents (n=438) vs. CABG (n=442)
- Setting: 27 centers in East Asia
- Enrollment: 2008-2013
- Median follow-up: 4.6 years
- Primary outcome: composite of death, myocardial infarction, or target-vessel revascularization

### Population
- Age ≥18 years
- Angiographically confirmed multivessel coronary artery disease suitable for PCI or CABG
- Excluded if had significant left main coronary artery disease

### Interventions
- PCI with everolimus-eluting stents
- CABG with preference for using the internal thoracic artery for revascularization of the left anterior descending coronary artery

### Outcomes
#### Primary Outcomes
- At 2 years: primary end point occurred in 11.0% of PCI group and 7.9% of CABG group (absolute risk difference, 3.1 percentage points; P=0.32 for noninferiority)
- At median 4.6 years follow-up: primary end point occurred in 15.3% of PCI group and 10.6% of CABG group (hazard ratio, 1.47; P=0.04)

#### Secondary Outcomes
- Composite of death, myocardial infarction, stroke, or any repeat revascularization was higher after PCI than after CABG (19.9% vs. 13.3%; P=0.01)
- No significant difference in composite safety end point of death, myocardial infarction, or stroke
- Higher rates of any repeat revascularization and spontaneous myocardial infarction after PCI
- Stent thrombosis occurred in 1.6% of patients within the PCI group

### Criticisms
- Early termination of the trial might have reduced the statistical power.
- Crossovers from CABG to PCI may have introduced a bias.
- Results from the subgroup analyses should be considered exploratory due to restrictions in sample size.
- Use of some medications varied significantly between groups.
- Includes only patients of Asian race which could affect generalizability.

### Funding
The trial was supported by the CardioVascular Research Foundation, Abbott Vascular, and a grant from Korea Healthcare Technology R&D Project, Ministry for Health and Welfare Affairs, South Korea.

### Further Reading
Additional details and investigators can be found in the Supplementary Appendix at NEJM.org.