Does carotid endarterectomy (CEA) reduce the risk of a cerebrovascular accident (CVA) in patients with recent transient ischemic attacks (TIA) or CVA and ipsilateral carotid stenosis?

     multicenter randomized controlled trial in U.S. and Canada comparing CEA + best medical therapy vs. best medical therapy alone

     inclusion criteria: participating centers had <6% CVA/death within 30 days of CEA for at least 50 consecutive operations; patients younger than 80 years old with hemispheric TIA or nondisabling CVA within previous 120 days and 30-99% stenosis of ipsilateral internal carotid artery (ICA) suitable for CEA (stenosis measure = 1- (diameter at point of highest stenosis)/(normal diameter of ICA beyond stenosis once vessel walls parallel); exclusion criteria: life expectancy less than 5 years due to organ failure/cancer/etc.; cardioembolic symptoms

     The investigators arbitrarily divided degree of stenosis into 2 groups before the trial began: 30-69% (moderate stenosis) and 70-99% (high grade stenosis).  The trial was stopped early for high grade stenosis patients after an interim analysis revealed clear benefit of CEA in these patients.  At the time of termination, 3 years had elapsed, 662 high grade stenosis patients had been enrolled (331 medical management/328 surgical management) with an average follow-up period of 18 months.  CEA was performed, on average, 2 days after randomization.  At last follow-up, 94% of medical therapy patients and 98% of surgical therapy patients were on aspirin (ASA).  Interestingly, before the trial began, 85% of each group were already on an antithrombotic medication.  The perioperative morbidity and mortality (all perioperative CVA/death) was 5.8% in the surgical group vs. 3.3% in the medical group, with the perioperative period considered 2 days prior to surgery (post-randomization) and 30 days postoperatively with comparison of an equivalent period for medical therapy patients.  The primary outcome, measured at 2 years, was the rate of ipsilateral CVA: 26% in the medical group vs. 9% in the surgical group, representing an absolute risk reduction (ARR)=17%, relative risk reduction (RRR)=65%, and number needed to treat (NNT)=6.  There was no significant difference between groups in mortality: 6.3% in medical group vs. 4.6% in surgical group.  The early increased morbidity and mortality risk incurred by the surgical group in the perioperative period was quickly overcome at 3 months, when the benefits of CEA outweighed the risks (ie. Kaplan-Meier curves for adverse events in the 2 groups crossed at 3 months, after which the surgical arm was superior).

     Medical therapy for most patients was antiplatelet treatment with 1300 mg ASA daily.  In a secondary analysis, patients were stratified according to other risk factors (diabetes, hypertension, smoking, etc).  For patients in the medical arm, the number of risk factors correlated with likelihood of CVA and could be used to stratify risk, but this was not true in the CEA group.  Surgery was so effective that it abolished the significant difference in risk attributable to other factors.  When this study was published, the trial was still ongoing for the moderate stenosis cohort (30-69%).  A later publication based on this cohort would show that patients with 50-69% stenosis also benefit from CEA, but not those with stenosis less than 50%.