Transcatheter versus surgical intervention: lessons from trials of coronary revascularisation

Abstract

In this paper, a cardiac surgeon and an interventional cardiologist draw lessons from the trials of percutaneous coronary intervention/coronary artery bypass graft surgery (PCI/CABG) for the transcatheter aortic valve implantation/surgical aortic valve replacement (TAVI/SAVR) era. Both PCI and CABG are effective treatments, but do the trials ask the right questions? They dwell on the ‘either/or’ decision for selected patients suitable for both treatments, but provide little guidance for the majority of ‘real world’ patients with comorbidities precluding CABG, or complex coronary heart disease precluding PCI. The control group must be meaningful and relevant. The pace of technological and therapeutic change causes trials to date rapidly. Procedures often do not reflect everyday practice, such as performing multivessel PCI, copious stenting and full arterial grafting. Composite endpoints such as major adverse cardiovascular and cerebrovascular events (MACCE) provide statistical significance but little insight into patient-orientated needs. There is a variety of temporal, safety, symptomatic and prognostic endpoints, provoking debate over their relative and absolute magnitude and importance; and there are issues of interpretation and inappropriate extrapolation. Trial interpretation, crystallised in the Kaplan-Meier curve, focuses on the relative benefit of one treatment over another, but deserves careful scrutiny. Subgroup analysis tends to exceed its role in dealing with issues such as poor left ventricular function, diabetes, multivessel disease and proximal left anterior descending coronary artery stenosis. Meta-analysis is controversial and guidelines date rapidly, lacking robust evidence in some domains, yet assuming considerable importance. Measures of frailty, physiological measures of blood flow, ‘real world’ activity levels, and predictions of benefit rarely feature. The multidisciplinary Heart Team meeting, now integral to study design, is challenging to deliver in practice. The PCI/CABG trials, and the issues arising from them, provide salutary lessons in the TAVI/SAVR era.