P4733Risk of myocardial infarction with coronary artery bypass grafting versus percutaneous coronary intervention: a systematic review and meta-analysis of fifteen randomised trials

Abstract

Abstract Background It remains controversial whether coronary artery bypass grafting (CABG) or percutaneous coronary intervention (PCI) is more effective in the prevention of myocardial infarction (MI). MI has been evaluated only as a secondary endpoint without a focused systematic review in multiple meta-analyses. Purpose To compare the risk of MI at the latest follow-up available between CABG versus PCI with stents in patients with multivessel or left main coronary artery disease in a pairwise meta-analysis of randomised controlled trials (RCT). Methods We searched EMBASE, Cochrane, and Pubmed databases for articles comparing CABG versus PCI for the treatment of multivessel or left main disease. We utilised random-effects model to calculate pooled risk ratio (RR) and 95% confidence interval (CI). Fifteen trials with a total of 13,592 patients treated with either CABG (n=6,596) or PCI (n=6,996) were eligible and included. A multivariable random-effects meta-regression model, including variables such as age, sex, diabetes mellitus, publication year, follow-up duration, type of stent used, and type of coronary artery disease, was used to explore the source of potential heterogeneity of the primary result. Results After a weighted follow-up of 4.3 years, patients treated with CABG had a significantly lower risk of MI than patients treated with PCI (RR 0.75, 95% CI 0.58–0.96, P=0.024, I2=66%). The lower risk of MI with CABG as compared to PCI was more evident during a longer duration of follow-up (≥3 years, RR 0.69, 95% CI 0.52–0.91, P=0.008; ≥5 years, RR 0.64, 95% CI 0.48–0.86, P=0.003) and in diabetic population (RR 0.55, 95% CI 0.44–0.70, P<0.001). There was a statistically meaningful trend toward fewer MIs with CABG with a similar magnitude of risk reduction across patients with left main disease (RR 0.74, 95% CI 0.47–1.15) and multivessel disease (RR 0.72, 95% CI 0.53–0.99). Moderate inter-study heterogeneity could not be explained by the clinical and trial-based variables tested in meta-regression, and is likely because of differences in definitions of MI, risk profile of enrolled patients, and procedural specifics. Forest plots Conclusions In patients undergoing revascularization for multivessel or left main disease, the risk of MI was lower with CABG compared to PCI. The quality assurance for MI definition and treatment-specific procedures should be emphasized for future RCTs.