The FREEDOM trial: a definitive answer to coronary artery bypass grafting or stents in patients with diabetes and multivessel coronary artery disease

Abstract

For the last two decades, there has been intense debate between cardiologists and surgeons regarding what the most effective mode is, of revascularization in patients with diabetes already receiving optimal medical therapy, who additionally require invasive coronary artery intervention on symptomatic and/or prognostic grounds. The debate began in earnest with the original interim analyses of the BARI trial, suggesting a survival advantage for the subset of patients with diabetes receiving coronary artery bypass grafting (CABG) rather than angioplasty [1], and was confirmed in the final 10-year follow-up with respective survival rates of 58 vs 46% (P=0.025) [2]. Since then, the recent publications of the 5-year outcomes of the SYNTAX Trial [3] and the ASCERT registry [4 ]h ave provided strong evidence that CABG, in comparison with percutaneous coronary intervention (PCI), offers a strong survival benefi ta s well as a marked reduction in myocardial infarction (MI) and repeat revascularization in patients with intermediate and more severe coronary artery disease (CAD) as judged by SYNTAX scores of >22. Because of the perception that patients with diabetes often have more severe and aggressive forms of CAD, intuitively there is an impression that they are also therefore likely to derive greater clinical benefit from CABG than from PCI. Others have argued, however, that significant advances in both medical therapy and stent technology would have eliminated the former—if any— benefit of CABG over PCI. Consequently, until the recent publication of the FREEDOM Trial [5], the relative merits of both interventions in patients with diabetes have also been fiercely debated. The main reason for the controversy was that, until the FREEDOM Trial, there was not a single trial in patients with diabetes, adequately powered, to give a definitive answer regarding whether CABG offered a survival or any other clinical benefit over PCI. It is therefore informative not only to examine the results of the FREEDOM trial, but also in the context of whether its findings are consistent with other evidence. The FREEDOM trial randomized 1900 patients with diabetes and multivessel CAD, already receiving aggressive medical therapy, to CABG or PCI with drug-eluting stents. This trial therefore included more patients with diabetes than the total from all the subgroups from the other randomized trials. The primary 5-year composite outcome of death from any cause, non-fatal MI or stroke, occurred in 26.6% of the PCI group and 18.7% of the CABG group (P=0.005). The benefit of CABG was driven by superior outcomes in both rates of death from any cause (10.6 vs 14.9%; P= 0.049) and MI (6.0 vs 13.9%; P< 0.001), but at the cost of a higher risk of stroke in the CABG group (5.2 vs 2.4%; P=0.03). And it is important to appreciate that this benefit of CABG over PCI was still evident although all patients were receiving aggressive medical therapy. It is alsoworth noting the relative risk of non-fatal stroke between CABG and PCI was 2.41 at 1 year, narrowing to 1.72 at 5 years [5]. The absolute clinical difference in stroke rates between CABG and PCI was actually small, being 2.1% (30 of 1444) and 0.82% (12 of 1471), respectively, at 1 year and 4.8% (65 of 1344) and 2.8% (42 of 1500) at 5 years. The early difference in stroke is almost certainly a consequence of the very well-recognized perioperative risk of stroke with CABG, with subsequent rates being more similar between the interventions. The same phenomenon was observed in the SYNTAX trial with CABG patients receiving less in the way of single and dual anti-platelet medication in the postoperative period [5]. Are the results of the FREEDOM trial consistent with other evidence? The answer is overwhelmingly, yes, from several other sources. Probably, the second-most definitive piece of evidence regarding the relative benefits of interventions in patients with diabetes is a subset analysis of the collaborative analysis of individual patient data from 7812 patients from 10 trials, of whom 1243 had diabetes, by Hlatky and co-workers. At a median follow-up of 5.9 years, mortality was substantially lower in the CABG group than in the PCI group (hazard ratio [HR]: 0.70, 95% confidence interval [CI]: 0.56–0.87) [6]. Since this analysis, several other pieces of data have emerged supporting the use of CABG rather than PCI in patients with diabetes. In the BARI 2D trial, 2368 patients with both type 2 diabetes and CAD were randomized to intensive medical therapy alone or to additionally undergo either prompt revascularization with PCI or CABG, according to the preference of the treating physician [7]. At 5 years, there was no significant difference in the composite primary end point between the PCI and medical therapy groups. In contrast, in the CABG group, the rate of major cardiovascular