- Design
- Substudy of a phase 3 randomised trial (CREST-2)
- Population
- 51 adults with asymptomatic carotid stenosis of 70% or more and reduced baseline cognition, 61 North American sites
- Primary outcome
- Change in composite cognitive z score over 1 year
- Effect
- Adjusted difference -0.15 SD (95% CI -0.54 to 0.24); interaction p=0.40
CREST-H was a substudy of CREST-2, which randomised adults with asymptomatic carotid stenosis of 70% or more to endarterectomy or stenting plus intensive medical management, or intensive medical management alone. Perfusion imaging measured how delayed blood flow was in the hemisphere beyond the stenosis. The question was whether restoring flow improves cognition in patients whose brains were under-perfused.
Of 392 enrolled, 291 had complete data and 51 had reduced cognition at baseline. In those 51, cognitive scores changed from -1.36 to -1.17 with revascularisation and from -1.34 to -1.01 without it. The adjusted difference was -0.15 SD (95% CI -0.54 to 0.24), and the degree of perfusion delay did not modify the effect.
The key group was small, so a modest benefit cannot be excluded. But the study gives no support for referring a patient for carotid surgery or stenting to protect memory. Decisions about asymptomatic stenosis should rest on stroke prevention and the main CREST-2 results.
- Do not refer a patient with asymptomatic carotid stenosis for revascularisation to improve or protect cognition.
- Investigate cognitive decline in the usual way, looking for the common causes before attributing it to the carotid.
- Make sure every patient with carotid stenosis is on intensive medical management: statin, blood pressure control, antiplatelet therapy and smoking cessation.
- Base the decision about carotid intervention on stroke risk, not on perfusion imaging.
Why it matters
It removes a hopeful argument, that restoring flow might rescue memory, that has been used to justify intervention.
Don't overread it
The analysed group was 51 people, so this rules out a large cognitive benefit, not a small one.
The statistics, in plain English
A z score measures cognition in standard deviations from the average for age and education. The difference of -0.15 SD slightly favoured medical treatment, and the interval from -0.54 to 0.24 includes no effect. With only 51 people in the key group, the study could miss a small benefit, but it gives no sign of a large one.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for top clinical updates, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free