Coronary artery bypass grafting versus percutaneous coronary intervention in patients with noninsulin treated type 2 diabetes mellitus: A meta-analysis of randomized controlled trials.

Abstract

The outcomes and prognosis of revascularization by either coronary artery bypass grafting (CABG) surgery or percutaneous coronary intervention (PCI) in patients with noninsulin-treated type 2 diabetes mellitus (NITDM) have not yet been well established. Randomized controlled trials (RCTs) were identified by searching Pubmed, EMBASE, and Cochrane library from inception until May 2016. Heterogeneity was evaluated, and the pooled hazard ratio (HR) was calculated by using a fixed-effect model. A random-effect model was used when statistically significant heterogeneity was observed (I2 ≥50%). All data analyses were carried out by using RevMan 5.3 and STATA software 12.0. A total of 4 RCTs involving 5 studies, consisting of 2270 patients with noninsulin-treated type 2 diabetes mellitus, were identified. Compared with CABG-treated patients, PCI-treated patients had significantly higher all-cause mortality (HR 1.39; 95% CI 1.01 to 1.91; P=.04), myocardial infarction (HR 2.14; 95% CI 1.40 to 3.27; P=.0004), repeated revascularization (HR 2.52; 95% CI 1.77 to 3.57; P<.00001), and major adverse cardiovascular and cerebrovascular events (HR 1.50; 95% CI 1.20-1.87; P=.0004). However, PCI was associated with lower incidence of stoke (HR 0.47; 95% CI 0.24 to 0.90; P=.02). In NITDM patients, our study suggests that CABG surgery is associated with reduced risk of mortality and morbidity, although with increased incidence of stroke compared with percutaneous coronary intervention. The decision if to have percutaneous coronary intervention or CABG surgery should factor the risk for stroke of the patients when considering CABG over percutaneous coronary intervention. Adequately powered RCTs are needed to confirm the results of this meta-analysis.