Design Matters: How Methodological Decisions May Have Shaped the Findings of CREST-2.

Abstract

CREST-2 (Carotid Revascularization and Medical Management for Asymptomatic Carotid Stenosis Trials) comprised 2 parallel randomized trials in asymptomatic carotid stenosis comparing medical management alone with medical management plus carotid artery stenting or carotid endarterectomy. Carotid stenting achieved a modest but statistically significant 4-year reduction in primary events, whereas carotid endarterectomy did not. This commentary examines methodological features that may have favored the revascularization arms, including omission of peri-procedural myocardial infarction and major hemorrhage as end points; allowing operator-team members to perform neurological assessments; highly selective credentialing of carotid stenting operators; and a primary analysis that weighted early and late events equally. Taken together, these methodological features and ongoing advances in contemporary medical therapy suggest that carotid stenting, carotid endarterectomy, and medical management may yield similar outcomes in clinical practice.