Background: The coronary revascularization techniques of coronary artery bypass grafting (CABG) and percutaneous coronary intervention (PCI), with or without stents, are among the most common major procedures performed in the United States and Europe. Together, more than 900,000 of these procedures were performed in the United States in 2005.

Randomized controlled trials (RCTs) have shown that CABG improves survival over medical therapy in patients with left-main or triple-vessel coronary artery disease with decreased left ventricular function. PCI is preferred in patients with single-vessel disease because of its lower risks. However, the comparative advantage of each procedure is not clear in patients whose coronary artery disease is between these two extremes. Bravata and colleagues systematically reviewed the literature comparing CABG and PCI in patients with single-vessel, proximal left anterior descending disease; most forms of double-vessel disease; or less extensive forms of triple-vessel disease to evaluate survival, cardiovascular complications (e.g., stroke, myocardial infarction), freedom from angina, and need for revascularization.

The Study: Studies between January 1966 and August 2006 that directly compared the health outcomes of CABG and PCI were reviewed. Trials of balloon angioplasty and coronary stents, as well as standard or minimally invasive CABG techniques were included. Two authors independently reviewed abstracted data from these studies and graded the quality of each (A to C), with A meaning it was a high-quality trial, B meaning it was a randomized trial with incomplete information about methods, and C meaning it was a trial that had evident flaws. Risk differences between PCI and CABG were assessed and heterogeneity among studies was accounted for. Observational studies from databases that included at least 1,000 patients in each procedure group and that provided sufficient information about the patient populations were also included.

Of 1,695 potential articles, 113 that reported on 23 separate RCTs met inclusion criteria. These RCTs included 9,963 patients, with 5,019 randomly assigned to PCI and 4,944 to CABG. Only three studies were conducted in the United States; the remainder were conducted in the United Kingdom and the rest of Europe. Fourteen trials enrolled patients with multivessel disease (11 exclusively and three predominantly). The remaining nine studies limited the patients to those with single-vessel disease of the proximal left anterior descending artery. Trial quality was essentially good, with 21 trials graded as A, one trial as B, and one trial as C. The results did not change significantly with removal of the studies of poorer quality.

The trial participants were mostly of European ancestry, and had an average age of 61 years. Approximately 30 percent of the participants were women. Few of the participants had heart failure or poor left ventricular function, but 20 percent had diabetes, 40 percent had previous myocardial infarction, 50 percent had hypertension, and 50 percent had hyperlipidemia.

Results: Short-term survival (within 30 days of the procedure) was high in both groups, and was not significantly different. Long-term survival (at one and five years after the procedure) was not statistically different between the two procedure groups, with an absolute survival difference of less than 1 percent. This difference did not change in the trials that reported outcomes after five years.

The rate of stroke during the procedure was higher in the CABG group than in the PCI group (1.2 versus 0.6 percent). Relief from angina was more common after CABG than PCI at one, three, and five years following the procedure, with a statistically significant risk difference that ranged from 5 to 8 percent. Procedural myocardial infarction was not assessed consistently among the studies.

The need for revascularization was significantly higher across all studies for patients undergoing PCI than for those undergoing CABG, but the rate was lower in trials that used coronary stents. Although the trials favored CABG, the study results were statistically heterogeneous.

Conclusion: Although long-term survival is not different between CABG and PCI, CABG has a lower need for revascularization and is more likely to relieve angina, but it also has a higher risk of procedural stroke. The authors conclude that there are relative benefits and risks to each procedure, and that these risks and benefits may change as surgical and catheterization techniques continue to evolve.