CREST
Stenting versus Endarterectomy for Treatment of Carotid-Artery Stenosis
Clinical Question
Is carotid-artery stenting non-inferior to carotid endarterectomy in preventing stroke, MI, or death in patients with symptomatic or asymptomatic carotid stenosis?
Bottom Line
Carotid-artery stenting and endarterectomy had similar rates of the composite endpoint, but stenting had higher periprocedural stroke risk and lower MI risk. Age influenced relative outcomes.
Major Points
- Largest randomized comparison of carotid artery stenting (CAS) vs carotid endarterectomy (CEA). 2,502 patients with symptomatic (≥50% stenosis, TIA/stroke within 180 days) or asymptomatic (≥60% stenosis) carotid disease at 117 U.S./Canadian centers.
- Composite of 47% symptomatic and 53% asymptomatic patients — mixed population is a key design feature.
- Primary composite endpoint (stroke, MI, or death periprocedurally OR ipsilateral stroke within 4 years): 7.2% CAS vs 6.8% CEA (HR 1.11, 95% CI 0.81–1.51, P=0.51) — no significant difference.
- Periprocedural stroke was higher with CAS: 4.1% vs 2.3% (P=0.01). Periprocedural MI was higher with CEA: 2.3% vs 1.1% (P=0.03). Cranial nerve palsy: 0.3% CAS vs 4.7% CEA.
- Critical age interaction (P=0.02): CAS favored in patients <70 years (HR 0.69), CEA favored in patients ≥70 years (HR 1.66). Crossover age was ~70 years.
- Devices: RX Acculink nitinol stent + Accunet embolic protection device (Abbott Vascular). Operators required lead-in phase of ≥10 cases with ≤6% complication rate.
- CREST-2 (ongoing) is testing CAS vs intensive medical management alone (without CEA comparator) and CEA vs intensive medical management alone for asymptomatic carotid stenosis.
- Long-term follow-up (10 years, published 2016): composite endpoint remained similar. Restenosis was higher after CAS (6.0% vs 3.3%) but did not translate to more ipsilateral strokes.
Design
Study Type: Randomized, controlled, open-label, blinded-endpoint trial
Randomization: 1
Blinding: Blinded endpoint adjudication
Enrollment Period: December 2000 – July 2008
Follow-up Duration: Median 2.5 years (up to 4 years)
Centers: 117
Countries: USA, Canada
Sample Size: 2502
Analysis: Intention-to-treat, Kaplan-Meier, proportional hazards modeling with subgroup analyses
Inclusion Criteria
- Age ≥40 years with symptomatic (TIA/stroke ≤180 days) or asymptomatic carotid stenosis
- ≥50% stenosis (symptomatic) or ≥60% (asymptomatic) by angiography; or ≥70% by ultrasound
- Clinical and anatomical suitability for either procedure
Exclusion Criteria
- Major disabling stroke precluding endpoint assessment
- Recent MI (<30 days), unstable angina
- Chronic or paroxysmal atrial fibrillation requiring anticoagulation
- Severe comorbidities or contraindications to revascularization
Arms
| Field | Carotid Artery Stenting (CAS) | Control |
|---|---|---|
| Intervention | Carotid artery stenting with the RX Acculink nitinol self-expanding stent and Accunet embolic protection device (Abbott Vascular). Pre-procedure: aspirin 325 mg + clopidogrel 75 mg daily for ≥48 hours before procedure. Post-procedure: clopidogrel for 4 weeks + aspirin indefinitely. Heparin administered during procedure. Stenting operators required lead-in phase: ≥10 cervical CAS procedures with complication rate ≤6%. | Standard carotid endarterectomy with plaque removal. Surgical technique at operator discretion (conventional or eversion). General or regional anesthesia per center protocol. Shunting at surgeon discretion. Aspirin 325 mg daily started before surgery and continued indefinitely. Surgeons required ≥12 CEA procedures per year and complication rate ≤3% for asymptomatic and ≤5% for symptomatic patients. |
| Duration | Up to 4 years | Up to 4 years |
Outcomes
| Outcome | Type | Control | Intervention | HR / OR / RR | P-value |
|---|---|---|---|---|---|
| Any stroke, MI, or death during the periprocedural period or ipsilateral stroke within 4 years | Primary | 6.8% | 7.2% | 1.11 | 0.51 |
| Periprocedural stroke | Secondary | 2.3% | 4.1% | p=0.01 | |
| Periprocedural MI | Secondary | 2.3% | 1.1% | p=0.03 | |
| Stroke or death at 4 years | Secondary | 4.7% | 6.4% | p=0.03 | |
| Stroke more frequent with CAS | Adverse | 4.1% | |||
| MI more frequent with CEA | Adverse | 2.3% | |||
| Cranial nerve palsy: 0.3% (CAS) vs 4.7% (CEA) | Adverse | 4.7% | 0.3% | HR 0.07 (95% CI 0.02-0.18) |
Subgroup Analysis
Most important interaction: age (P=0.02) — CAS favored for patients <70 years (HR ~0.69), CEA favored for ≥70 years (HR ~1.66). No significant interaction by sex, symptom status (symptomatic vs asymptomatic), or degree of stenosis. Symptomatic subgroup: composite 6.0% CAS vs 5.4% CEA (HR 1.13, NS). Asymptomatic subgroup: composite 7.9% CAS vs 5.6% CEA (not powered). Contralateral occlusion did not significantly modify treatment effect. Race/ethnicity (26% non-White) did not significantly interact with treatment. Periprocedural MI was included in the composite but its long-term impact on disability was debated — post hoc analyses suggested periprocedural strokes had more impact on quality of life than periprocedural MIs.
Criticisms
- Single stent system (RX Acculink + Accunet) — newer-generation stents with mesh protection (CGuard, GORE Flow Reversal) may have better outcomes.
- Highly credentialed operators with lead-in phase — community results may be worse, especially for CAS.
- Mixed symptomatic/asymptomatic cohort (47%/53%) dilutes power for each subgroup — cannot draw firm conclusions for either group alone.
- No medical therapy-only comparator arm — cannot assess whether either procedure is superior to optimal medical therapy alone (being addressed by CREST-2).
- Inclusion of MI in the composite endpoint is controversial — strokes had greater impact on quality of life than MIs in post hoc analyses.
- Long enrollment period (2000–2008) — medical therapy and stent technology evolved significantly during this time.
- Age interaction (P=0.02) is the key finding but was based on a continuous interaction test — the 70-year cutoff is approximate.
Funding
NINDS (NIH) and Abbott Vascular (device donation, <15% of cost)
Based on: CREST (New England Journal of Medicine, 2010)
Authors: Brott TG, Hobson RW, Howard G, ..., Meschia JF
Citation: N Engl J Med 2010; 363:11–23. doi:10.1056/NEJMoa0912321
Content summarized and formatted by NeuroTrials.ai.