Published online Aug 26, 2012. doi: 10.4330/wjc.v4.i8.242
Revised: July 17, 2012
Accepted: July 24, 2012
Published online: August 26, 2012
In the past, comparative effectiveness trials evaluating percutaneous coronary interventions (PCI), using either balloon angioplasty or bare metal stent (BMS) implantation, versus coronary artery bypass surgery (CABG) found similar survival rates at long-term follow-up with both revascularization strategies. Two major meta-analyses of these trials reported 5- and 6-year comparative effectiveness between PCI and CABG: one included only four trials that compared PCI with BMS implantation versus CABG whereas the largest one also included trials using balloon angioplasty. In these studies, the authors observed no survival differences between groups although a significant survival advantage was seen in diabetics treated with CABG and this benefit was also perceived in elderly patients. In both reports, number of involved vessels, presence of left anterior descending artery stenosis or poor left ventricular ejection fraction were no predictors of poor survival with PCI. Therefore, extent of the coronary artery disease (CAD) was not associated with poor outcome after PCI in the pre-drug eluting stent (DES) era. Recently, the ASCERT (Database Collaboration on the Comparative Effectiveness of Revascularization Strategies) registry found higher mortality rate with PCI in patients ≥ 65 years old in comparison with CABG, and advantages of surgery were seen in all subgroups including those at low risk. In this registry, PCI was accomplished by implantation of the first type of DES designs in 78% of cases. The intriguing observation of high mortality rate with PCI, including for non-diabetics and patients with two-vessel CAD, meaning a lack of clinical benefit with DES implantation, had not been seen previously. The study was not randomized, although its results are largely strengthened by its sample size. In this manuscript, the authors describe other registries and randomized trials reporting similar results supporting the findings of the aforementioned study and explore the reasons for these results, while also searching for potential solutions.