Background: The efficacy of carotid endarterectomy in patients with asymptomatic carotid stenosis has not been confirmed in randomized clinical trials, despite the widespread use of operative intervention in such patients.
Methods: We conducted a multicenter clinical trial at 11 Veterans Affairs medical centers to determine the effect of carotid endarterectomy on the combined incidence of transient ischemic attack, transient monocular blindness, and stroke. We studied 444 men with asymptomatic carotid stenosis shown arteriographically to reduce the diameter of the arterial lumen by 50 percent or more. The patients were randomly assigned to optimal medical treatment including antiplatelet medication (aspirin) plus carotid endarterectomy (the surgical group; 211 patients) or optimal medical treatment alone (the medical group; 233 patients). All the patients at each center were followed independently by a vascular surgeon and a neurologist for a mean of 47.9 months.
Results: The combined incidence of ipsilateral neurologic events was 8.0 percent in the surgical group and 20.6 percent in the medical group (P < 0.001), giving a relative risk (for the surgical group vs. the medical group) of 0.38 (95 percent confidence interval, 0.22 to 0.67). The incidence of ipsilateral stroke alone was 4.7 percent in the surgical group and 9.4 percent in the medical group. An analysis of stroke and death combined within the first 30 postoperative days showed no significant differences. Nor were there significant differences between groups in an analysis of all strokes and deaths (surgical, 41.2 percent; medical, 44.2 percent; relative risk, 0.92; 95 percent confidence interval, 0.69 to 1.22). Overall mortality, including postoperative deaths, was primarily due to coronary atherosclerosis.
Conclusions: Carotid endarterectomy reduced the overall incidence of ipsilateral neurologic events in a selected group of male patients with asymptomatic carotid stenosis. We did not find a significant influence of carotid endarterectomy on the combined incidence of stroke and death, but because of the size of our sample, a modest effect could not be excluded.