Two recent studies in the New England Journal of Medicine attempted to cast some light on the question how percutaneous coronary intervention (PCI) compares to coronary artery bypass grafting (CABG) but arrived at somewhat different conclusions. As Robert Harrigton pointed out in the accompanying editorial, randomized trials such as the Randomized Comparison of Coronary Artery Bypass Surgery and Everolimus-Eluting Stent Implantation in the Treatment of Patients with Multivessel Coronary Artery Disease (BEST) remain the gold standard to test a hypothesis although they are not entirely free of pitfalls. Indeed, the BEST trial exemplifies some issues incurring with noninferiority trials and composite outcomes combining safety and efficacy. From a narrow statistical perspective, BEST was a futile attempt in establishing noninferiority of PCI compared with CABG in patients with multivessel disease on the primary composite of death, myocardial infarction (MI), or revascularization (notably excluding stroke). The difference was driven by the established advantage of CABG preventing revascularizations, the magnitude of which made it unlikely that PCI could be on par. From a broader clinical perspective, the question how PCI compares with CABG in terms of death, MI, or stroke remained unanswered. The premature termination of BEST resulted in estimates that remain imprecise and compatible with both, a relevant benefit of PCI or harm.

The propensity score–matched comparison of PCI with everolimus-eluting stents and CABG in a considerably larger cohort by Bangalore et al in the same issue suggested that modern PCI is associated with a lower risk of stroke, a similar risk of death and higher risk of MI than CABG—the latter being observed only when complete revascularization could not be achieved. Remarkably, both studies had somewhat concordant results for individual end points. There was no difference in death between PCI versus CABG and an increase in MI after PCI occurred in both studies. However, in the study of Bangalore et al, the increased MI rate with PCI was no longer present in patients who were completely revascularized. The take-home message of both studies is that inpatients with multivessel disease, short- and long-term risks, and benefits of both approaches must be carefully weighted as is recommend in current guidelines . As to BEST, it certainly was a good study but in answering the question of PCI versus CABG, it clearly does not live up to its acronym.