Carotid endarterectomy for asymptomatic carotid stenosis: asymptomatic carotid surgery trial.

Effective prevention is inarguably the best option for reducing the individual and societal burden of stroke. For each patient, clinicians balance the benefits of a given preventive therapy against its associated risks and costs. Where possible, these assessments should be based on the results of randomized clinical trials. Carotid endarterectomy (CEA), the most-commonly used surgical procedure to prevent stroke, has been subjected to several randomized trials. These underlie evidence-based guideline and consensus statements providing recommendations for its use.1–7 The evidence base for endarterectomy for symptomatic stenosis is considerable,8,9 but guidelines on surgery for asymptomatic stenosis have been largely based on the results of the Asymptomatic Carotid Atherosclerosis Study (ACAS)10 in conjunction with other smaller trials.11,12 Guidance differs from endorsement of the operation for selected patients (eg, based on patient age, life expectancy, concomitant illnesses, etc.) with varying degrees of asymptomatic stenosis (generally either 60% to 99% or 80% to 99%) in whom the procedure can be performed with low (ie, 3%) complication rates to advising that endarterectomy not be performed in patients without referable symptoms. ACAS reported a 47% relative reduction in the risk of ipsilateral stroke and perioperative death in patients randomized to surgery despite a 5-year risk of ipsilateral stroke without the operation of only 11%.10 The results led to major increases in rates of endarterectomy for asymptomatic stenosis in some countries, most notably the United States. Of the approximate 150 000 endarterectomies performed in the United States each year, at least half are done for stenoses that have never been symptomatic.13 In contrast, the ACAS results had little effect on endarterectomy rates in other countries such as the United Kingdom, where it was felt that the benefit (it was estimated that 40 operations were needed to prevent 1 disabling or fatal stroke after 5 years) did not justify the cost. There was also concern that the very low operative risks in ACAS (excluding complications of angiography: 1.5%, 95% CI, 0.6% to 2.4% for stroke and death; and 0.14%, 95% CI, 0% to 0.4%, for death) could not be matched in routine clinical practice. ACAS only accepted surgeons with an excellent safety record, rejecting 40% of initial applicants and subsequently barring from further participation some surgeons who had adverse operative outcomes during the trial.14 Figure 1 compares the operative risks in ACAS with the results of a meta-analysis of the 46 surgical case series that published operative risks for asymptomatic stenosis during ACAS and the 5 years after publication.15 Operative mortality was 8 higher than in ACAS (1.11% versus 0.14%; P 0.01), and the risk of stroke and death was 3 higher among comparable studies in which outcome was assessed by a neurologist (4.3% versus 1.5%; P 0.001). Even after community-wide performance measurement and feedback, the overall risk for stroke or death after endarterectomy performed for asymptomatic stenosis in 10 US states was 3.8% (including 1% mortality).16 Therefore, the degree to which the ACAS results can be generalized to routine clinical practice remained uncertain. Results of the largest randomized trial of endarterectomy for asymptomatic stenosis, the Medical Research Council Asymptomatic Carotid Surgery Trial (ACST), have now been published.17 How will these results affect current practice recommendations? Between 1993 and 2003, ACST randomized 3120 patients with 60% mainly asymptomatic carotid stenosis (12% had symptoms at least 6 months previously) to immediate endarterectomy plus medical treatment versus medical treatment alone or until the operation became necessary.17 Surgeons were required to provide evidence of an operative risk of 6% for their last 50 patients having an endarterectomy for asymptomatic stenosis, but none were excluded on the basis of his/her operative risk during the trial. Selection of patients was based on the “Uncertainty Principle,” with very few