Related Reading: This commentary accompanies the Currents NEWS summary, “CREST-2: Intensive Medical Management with or without Revascularization for Asymptomatic ≥70% Carotid Stenosis.”

CREST-2 comprises two parallel randomized trials comparing intensive medical management (IMM) alone versus IMM plus revascularization in patients with asymptomatic carotid stenosis ≥70%. The stenting trial met statistical significance for the primary endpoint (2.8% vs 6.0% at 4 years; p = .02, NNT=31), whereas the endarterectomy trial did not (3.7% vs 5.3%; p = .24). [1] That split result—two revascularization strategies aimed at the same mechanism in a similar population—should make us pause before translating a single “positive” trial into a default procedural pathway.

Several methodological choices shape how I read these data. Chimowitz (2026) highlights design features that may have favored revascularization—exclusion of periprocedural myocardial infarction and major hemorrhage from the primary endpoint, operator-team involvement in neurologic assessments, selective and stringent credentialing (particularly for stenting), and equal weighting of early procedural events and late ipsilateral ischemic strokes. [2] The trial’s own sensitivity analyses reinforce how narrow the margin is: shifting outcomes in only 3-4 patients would have changed statistical significance in the stenting trial, a >10% relative change in event counts. [1]

The medical-therapy comparator is also a moving target. As Brown et al. (2026) point out, targets were not universally achieved during follow-up. Only about 60–70% met systolic blood pressure goals, < 80% met low density lipoprotein (LDL) goals, and approximately 50% of patients with diabetes met A1c targets, leaving room for more intensive prevention to further lower event rates. [3] By the time the trial was completed, PCSK9 inhibitors, modern diabetes/weight-loss agents (e.g., GLP-1 receptor agonists), and transcarotid revascularization were more widely used, none of which CREST-2 was positioned to test as part of baseline strategy. [1] Newer dyslipidemia guidance endorses LDL-C targets as low as <55 mg/dL for very high-risk atherosclerotic cardiovascular disease, underscoring that “optimal medical therapy” continues to evolve beyond the CREST-2 protocol. [4]

External validity matters as much as statistical significance. CREST-2 procedures were performed by highly credentialed operators, and outcome ascertainment relied heavily on MRI for cerebral infarctions (82%), which may accentuate differences relative to CT-first real-world pathways. [1] In that context, the practical question becomes reproducibility: can typical practice settings match CREST-2’s procedural safety and follow-up intensity, and if not, how should we counsel asymptomatic patients about a modest absolute benefit?

For neurointensivists counseling patients and advising teams, CREST-2 supports a conservative posture: start with IMM, reinforce close follow-up, and reserve revascularization for patients who remain higher risk despite optimized prevention (or who strongly prioritize even a small absolute stroke-risk reduction after understanding the early procedural tradeoff). [1–4] CREST-2 does not justify “revascularize all asymptomatic ≥70% stenosis,” and it does not establish a uniform preference for endovascular treatment over endarterectomy. The next step is better risk stratification—identifying the smaller subgroup that truly fails medical therapy—rather than expanding procedures for all-comers.

Author: