Current issues of ACP Journal Club are published in Annals of Internal Medicine


Carotid endarterectomy reduced death and strokes in patients with ipsilateral high-grade stenosis and recent hemispheric transient ischemic attacks or nondisabling strokes

ACP J Club. 1991 Sep-Oct;115:34. doi:10.7326/ACPJC-1991-115-2-034

Related Content in this Issue
• Companion Abstract and Commentary: Carotid endarterectomy reduced strokes in patients with recent, nondisabling cerebral ischemic events and severe carotid stenosis

Related Content in the Archives
Elderly patients had an increased 5-year risk for death, coronary events, and stroke after retinal infarction
Endarterectomy was not effective for moderate symptomatic carotid stenosis

Source Citation

North American Symptomatic Carotid Endarterectomy Trial Collaborators. Beneficial effect of carotid endarterectomy in symptomatic patients with high-grade carotid stenosis. N Engl J Med. 1991 Aug 15;325:445-53. [PubMed ID: 1674060]



To determine whether carotid endarterectomy reduces the risk for fatal and nonfatal ipsilateral carotid stroke among patients with recent cerebrovascular events and ipsilateral, surgically accessible carotid stenosis (70% to 99%).


Multicenter, randomized, controlled trial with 2 predetermined strata based on carotid stenosis: 30% to 69% and 70% to 99%. Interim results for the 70% to 99% stratum are in this report, with an average follow-up of 18 months.


50 clinical centers in the United States and Canada.


659 eligible patients were entered with a hemispheric transient ischemic attack or nondisabling stroke or retinal infarction within the previous 120 days, with 70% to 99% stenosis in the appropriate carotid artery, determined by angiography and accessible to endarterectomy. No patient was lost to follow-up.


All patients were provided with optimal medical care, including antiplatelet treatment. Patients allocated to surgery had carotid endarterectomy by neurosurgeons or vascular surgeons (average delay, 2 days after randomization).

Main outcome measures

All patients were followed by neurologists. End points were adjudicated by blinded, independent case review. Strokes ipsilateral to the randomized carotid lesion and all deaths in the first month after randomization were counted for both groups.

Main results

Only 1 patient refused surgery. 21 medical patients (6.3%) crossed over to surgery during follow-up. By 2 years, life-table estimates of the cumulative risk for ipsilateral stroke (including perioperative events) were 9% of the 328 surgical patients and 26% of the 331 medical patients, an absolute risk reduction (ARR) of 17% (P < 0.001); the relative risk reduction (RRR) was 65%. For major or fatal ipsilateral stroke, the corresponding estimates were 2.5% and 13.1%, an ARR of 10.6% (P < 0.001); the RRR was 81%. Carotid endarterectomy remained beneficial when the outcome analysis included all strokes of any severity in any territory (P < 0.001); major or fatal stroke in any territory (P < 0.001); and any major stroke or death from any cause (P < 0.01). A lesser degree of carotid stenosis before surgery was associated with a lesser risk reduction after surgery. Total mortality did not differ significantly between treatment groups {4.6% vs 6.3%; ARR 2%, CI -2% to 5%, P = 0.3171}*.


Carotid endarterectomy was beneficial for patients with recent hemispheric transient ischemic attacks or nondisabling strokes and ipsilateral, high-grade (70% to 99%) stenosis.

Source of funding: National Institute of Neurological Disorders and Stroke.

Address for article reprint: Professor D.W. Taylor, Department of Clinical Epidemiology and Biostatistics, McMaster University Faculty of Health Sciences, Chedoke Division #74, 1200 Main Street West, Hamilton, Ontario L8N 3Z5, Canada.

*Numbers calculated from data in article.


Carotid endarterectomy reduced strokes in patients with recent, nondisabling cerebral ischemic events and severe carotid stenosis

The ECST and the NASCET are landmark studies. Many clinicians believed that persons with cerebral ischemic symptoms associated with high-grade carotid stenosis benefited from carotid endarterectomy. However, considerable doubt arose about a net benefit to society, given the high risk of surgery reported from some communities and the reported high frequency of inappropriate indications for carotid endarterectomy. These studies do not completely allay the doubt. In both studies, the surgeons were carefully chosen and lower perioperative morbidity and mortality was shown to be the case in community surveys.

The major surprise in these studies is the high risk for stroke in the patients who did not have surgery; 22% in 3 years in ECST and 28% in 2 years in NASCET. The apparently higher risk in NASCET may be at least partly an artifact of their method of measuring stenosis: luminal diameter at greatest stenosis divided by diameter of normal artery beyond the bulb. Previous estimates from population-based studies of 5% per year (1) and of 6% per year among patients who had angiograms but not surgery (2) are misleading. The population-based studies did not separate the patients with high-grade stenosis by angiography, and the patients who had angiograms in the latter study included those with mild stenosis. These are presumed reasons for the underestimates. It would indicate, however, that the prevalence of high-grade stenosis in symptomatic patients in a community would probably be less than 15%.

These 2 trials provide unequivocal evidence of the benefit of carotid endarterectomy by qualified surgeons in patients with severe carotid stenosis. It is less certain where the cut point may be. In NASCET, the patients with 70% to 79% stenosis appear to benefit less than those with 90% to 99% stenosis, and there may not be a net benefit in those with 70% to 79% stenosis. Neither trial has stopped the entry of patients with 30% to 69% stenosis even though there are about 1100 such patients in ECST. This raises questions about the potential benefit in this group.

These 2 trials differ from each other in some ways. ECST used an expedient selection procedure based on the doctors' "substantial uncertainty." This probably provided a wider variety of risk in the patients randomized. In fact, the perioperative mortality and stroke morbidity were higher in ECST than in NASCET. There was also less benefit in ECST. ECST had a less precise design and a simpler format but still required 10 years to produce useful information. NASCET, on the other hand, produced a somewhat greater difference in the treatment groups in 3 years with less opportunity for slippage because the surgical patients had surgery sooner after their qualifying events and after randomization than was the case in ECST.

One of the most important messages of these 2 trials is that it is possible to assess carefully the benefits and risks of a surgical procedure and show that it is favorable. Surgeons have often resisted such assessments in favor of judgments about individual patients.

It is important to emphasize that these results do not apply to asymptomatic patients or to patients who have had symptoms for longer than 4 to 6 months. They also do not apply to the patients of all surgeons. A surgeon must show low surgical mortality and morbidity and careful selection of patients to make the case that the results of these trials apply to his or her patients.

Jack P. Whisnant, MD
Mayo ClinicRochester, Minnesota, USA


1. Whisnant JP, Wiebers DO. In: Sundt TM Jr, ed. Occlusive Cerebrovascular Disease: Diagnosis and Surgical Management. Philadelphia: WB Saunders; 1987:60-5.

2. Fields WS, Lemak NA. Joint study of extracranial arterial occlusion. IX. Transient ischemic attacks in the carotid territory. JAMA. 1976;235:2608-10.