PDF: NASCET
(1991)Objective
To evaluate the benefit of CEA in patients with symptomatic moderate or severe stenosis.
Study Summary
• Patients with severe stenosis (≥70%) had a durable benefit with a 65% risk reduction. • Patients with moderate stenosis (50-69%) had a moderate risk reduction of 30%..
Intervention
Carotid endarterectomy (CEA) versus medical management.
Inclusion Criteria
Patients with symptomatic carotid stenosis identified using angiography and a jeweler’s lens graded in mm.
Study Design
Arms: CEA vs. Medical Management
Patients per Arm: 2267 patients (split between CEA and medical management)
Outcome
• Five-year ipsilateral stroke risk: ≥70% stenosis: 9% in CEA vs. 26% in controls (RR reduction 65%, NNT 7-8). <br>• 50-69% stenosis: 15.7% in CEA vs. 22.2% in controls, P=0.045 (RR reduction 30%, NNT 15). <br>• <50% stenosis: 14.9% in CEA vs. 18.7%, P=0.16..
Bottom Line
Carotid endarterectomy performed in NASCET had an overall perioperative stroke and death rate of 6.5%. While the procedure carried risks, particularly in older patients and those with specific angiographic features, its benefits in preventing future disabling strokes for symptomatic patients with severe stenosis were demonstrated in the broader NASCET trial.
Major Points
- Landmark RCT that established carotid endarterectomy (CEA) as standard of care for symptomatic high-grade carotid stenosis; 1415 patients underwent CEA across 106 North American centers.
- For severe stenosis (70–99%): 2-year risk of ipsilateral stroke was 26% medical vs 9% surgical (ARR 17%, NNT 6). Trial was stopped early for this group in 1991.
- For moderate stenosis (50–69%): 5-year ipsilateral stroke rate was 22.2% medical vs 15.7% surgical (ARR 6.5%, p=0.045). Benefit was marginal and depended on subgroup characteristics.
- For mild stenosis (<50%): no benefit from CEA — surgery was equivalent or inferior to medical therapy.
- Perioperative stroke and death rate at 30 days was 6.5% overall (death 1.1%, disabling stroke 1.8%, nondisabling stroke 3.7%). At 90 days, disabling stroke improved to 1.1% due to neurological recovery.
- 88% of perioperative strokes were ischemic (not hemorrhagic) and 68% were nondisabling — supporting that CEA complications were generally manageable.
- Risk factors for perioperative complications: age >75 years, contralateral ICA occlusion (14.2% of patients), ipsilateral carotid siphon stenosis (29.9%), history of prior stroke, poor collateral flow, and low surgeon volume.
- Surgeon volume mattered: perioperative stroke/death was 3.5% for surgeons doing >10 cases/year vs 7.4% for ≤10 cases/year — establishing minimum volume standards.
- Stenosis measured by NASCET method: (1 − [minimal residual lumen / normal distal ICA]) × 100. This became the standard measurement method worldwide, replacing the ECST method.
- Restenosis ≥50% occurred in 8.3% at mean 15-month follow-up. Long-term follow-up showed CEA benefit was durable, with ipsilateral stroke risk remaining low for years.
- NASCET defined the NNT framework for CEA: NNT 6 for severe stenosis over 2 years, NNT 15 for moderate stenosis over 5 years — guiding clinical decision-making.
- Complementary to ECST (European Carotid Surgery Trial) and VA cooperative study — all three confirmed benefit for severe stenosis, establishing Level 1 evidence.
Study Design
- Study Type
- Prospective, multicenter, randomized controlled trial (surgical results report)
- Randomization
- Yes
- Sample Size
- 1415
- Follow-up
- 30 days (perioperative outcomes) and 90 days (stroke severity assessment); mean 15 months for durability of endarterectomy.
- Countries
- North America
Primary Outcome
Definition: Perioperative stroke and death within 30 days after carotid endarterectomy.
| Control | Intervention | HR/OR | P-value |
|---|---|---|---|
| NA | 6.5% (92 events in 1415 patients) | - | - |
Limitations & Criticisms
- The surgical results paper reports only on CEA patients, not the randomized comparison — the full trial results were published separately.
- Medical therapy of the 1990s (aspirin only, no statins, less aggressive BP/glucose control) may overestimate CEA benefit vs modern best medical therapy.
- Conventional catheter angiography was required for stenosis measurement — now largely replaced by non-invasive CTA/MRA, raising questions about stenosis measurement concordance.
- The NASCET stenosis measurement method systematically underestimates stenosis compared to the ECST method, creating confusion in clinical practice (70% NASCET ≈ 82% ECST).
- Surgeon selection and case volume thresholds may not reflect real-world practice — perioperative complication rates in community settings have been reported higher than NASCET's 6.5%.
- Women were underrepresented (~30%) and showed less benefit from CEA, particularly in the moderate stenosis group — raising equity concerns.
- The trial excluded patients with near-occlusion (string sign), who were later shown to have a more benign natural history, complicating clinical decisions at the extreme end of stenosis.
- No comparison with carotid artery stenting (CAS), which emerged as an alternative during the trial era — addressed later by CREST and other trials.
- Long-term follow-up showed restenosis in 8.3% at 15 months, but the clinical significance of asymptomatic restenosis was not well characterized.
Citation
Stroke. 1991;22:711-720.