Does surgery have a role in the management of asymptomatic carotid artery stenosis? Yes, but...

Vincent Thijs  About the author

High-grade internal carotid artery stenosis (CAS) is a relatively common condition, occurring in 5–10% of the general population over the age of 65 years. The prevalence of asymptomatic CAS is increased in individuals with vascular risk factors and in those with other manifestations of atherosclerosis such as peripheral artery disease. Only a small fraction of patients with CAS will develop a cerebral or retinal vascular event ipsilateral to this stenosis. In a recent study in patients with asymptomatic CAS who had diabetes or clinical manifestations of arterial disease, the rate of ipsilateral stroke was less than 1% per year.¹ According to another study, the rate of ipsilateral stroke in individuals with asymptomatic CAS remained this low over extended periods of follow-up (up to 15 years).² Surprisingly, CAS seems to be a predictor of vascular death and coronary events, rather than of stroke. Once a transient ischemic attack or amaurosis fugax has occurred, there is little doubt that carotid endarterectomy (CEA) is required, especially if the stenosis is tight and symptoms have occurred recently, because the risk of recurrent ipsilateral stroke is high.³ Few treatments in preventive vascular medicine, however, generate as much controversy as does CEA performed in patients with CAS who do not have any related symptoms, even though large clinical trials have been performed in this setting. The discussion centers mainly around two issues: the clinical relevance of statistically significant results from clinical trials; and the applicability of the trial results to clinical practice. Most guidelines advocate cautious use of CEA for asymptomatic stenosis. Despite these controversies and prudent guidelines, rates of CEA are soaring, at least in some countries.

A recent Cochrane Review summarized the outcomes of CEA for asymptomatic CAS.⁴ Three trials were performed that collectively randomized 5,223 patients with a total follow-up period of 17,037 years. These trials demonstrated a significant benefit of CEA in reducing the risk of ipsilateral stroke or any subsequent stroke. The operation lowered the risk of an individual meeting the composite end point of perioperative stroke or death or subsequent ipsilateral stroke by 29% (relative risk 0.71, 95% CI 0.55–0.90). The absolute risk reduction was 3.0% over 2.7 years in the Asymptomatic Carotid Atherosclerosis Study (ACAS) and 3.1% over 3.4 years in the Asymptomatic Carotid Surgery Trial (ACST).^{5, 6} On the basis of these figures, about 100 patients would need to undergo the operation to prevent 5 strokes over 5 years, raising doubts about the cost-effectiveness of this procedure. Moreover, the small benefits of surgery are not distributed evenly over the entire group of patients with CAS. In women, evidence is not in favor of the operation. In ACAS, women had a 9% increase in the risk of perioperative stroke, death or subsequent ipsilateral stroke.⁵ In ACST, the same risk was reduced by 10%.⁶ Combining data from these clinical trials yields a small, nonsignificant benefit of a 4% relative risk reduction, suggesting no benefit of CEA for women.⁴ This gender difference might be a statistical artifact resulting from the relatively low proportion of women included in the trials. Other reasons for a reduced benefit might be the higher operative risk and the generally lower recurrent stroke risk in women as compared with men. Longer-term follow-up might clarify the uncertain benefits of surgical treatment in women. Doubt is also raised over the efficacy of the procedure in elderly patients. ACAS did not demonstrate an increased risk of stroke with surgery in patients older than 68 years,⁵ but in ACST no surgical benefit was observed in patients aged over 75 years.⁶ The 10-year results from ACST will help to identify whether the risk reduction with CEA persists over time, and whether benefits in women and the elderly might emerge at a later time after surgery.

The concern has been raised that participants in clinical trials might not be representative of all patients with CAS; trial participants are likely to have less comorbidity and a lower risk of perioperative stroke. Additionally, the setting in which these trials were carried out might be atypical, as trials are likely to be performed at academic medical centers with experienced surgeons. The question then is whether the benefits of surgery demonstrated in clinical trial settings can be translated to nonacademic settings, in which patients tend to have more comorbidity and procedures are usually performed by less-experienced surgeons.

When the benefits of a treatment are substantial, differences in the clinical setting and case mix are unlikely to offset the benefit of the intervention. When the risk–benefit margin is small, however, as is the case with surgery for asymptomatic CAS, such differences can tip the balance, turning a beneficial intervention into a harmful one. In ACAS, only surgeons with a particularly good track record were authorized to operate. During the trial, surgeons were barred from the trial if too many adverse outcomes occurred. As a result, only 0.14% of the patients died in the perioperative period in ACAS, compared with 1.11% in published registries and 1% in ACST.⁷ Hospitals should provide audited data on outcomes of CEA procedures to enable referring physicians to determine whether local surgical risks are comparable with the risks seen in randomized trials.

Another issue that affects the external validity of the CEA trials is whether results obtained in the 1980s or 1990s are still valid in the age of combination antiplatelet therapy, intensive statin regimens, inhibitors of angiotensin-converting enzyme and angiotensin-II-receptor antagonists. The rate of statin use in ACAS was probably very low, and in ACST only 17% of patients were receiving statins at study initiation. In the Stroke Prevention by Aggressive Reduction in Cholesterol Levels (SPARCL) trial, the risk of requiring revascularization of the carotid artery was reduced by more than 50% in patients with CAS who were treated with high-dose atorvastatin.⁸ On the other hand, it is unknown whether, in a real-life setting, reduced long-term compliance to treatment with multiple drugs might tip the balance in favor of surgery.

Carotid stenting with a cerebral protection device is increasingly used as an alternative to CEA, despite disappointing results from relatively small randomized trials.⁹ Trials are underway to evaluate the safety and the long-term outcomes of stenting versus CEA in patients with asymptomatic CAS. These trials will also provide more data on contemporary rates of surgical complications. Moreover, at least one trial is planning to recruit an additional medically treated arm.¹⁰ This trial will hopefully answer the important questions of whether modern medical management surpasses surgery, in terms of patient benefit, and whether the high rates of surgery for asymptomatic CAS are still warranted.

For the moment, clinicians should probably stick to the cautious guideline from the American Academy of Neurology, which states that it is reasonable to consider CEA for patients between the ages of 40 and 75 years who have asymptomatic stenosis of 60–99%, provided the patient has a life expectancy of at least 5 years and provided the perioperative rate of stroke or death can be reliably documented to be lower than 3% for the center that will perform the procedure.¹¹

Competing interests

The author has received speaker honoraria and grant or research support from Boehringer Ingelheim and Pfizer.

ARTICLE NAVIGATION -ARTICLE TOOLS

Full text
• Download PDF
• Send to a friend
• Export citation
• Rights and permissions
• Order commercial reprints

ISSUE NAVIGATION - THIS ISSUE

• Table of Contents
• Previous article
• Next article