Asymptomatic carotid artery stenosis—it's time to stop operating

Anne Abbott  About the author

Asymptomatic severe (>50%) atherosclerotic stenosis of the proximal internal carotid artery is common in older individuals and is responsible for about 9–18% of all ischemic strokes in the anterior circulation.¹ It is, therefore, important to determine how best to prevent stroke caused by this lesion. Most clinicians are likely to advise management of vascular disease risk factors for patients with asymptomatic severe carotid artery stenosis (CAS) because many of these patients are identified through symptoms of atherosclerosis. The role of additional surgery (carotid endarterectomy [CEA]) has been controversial. I believe, however, that a surgical approach to stroke prevention in these patients is not warranted, for reasons discussed in this Viewpoint.

In some countries, surgery on asymptomatic severe CAS is common.¹ Recommendations to perform surgery (or stenting) for asymptomatic patients have been made largely on the basis of a small overall stroke-prevention benefit of about 1% per year² with CEA (combined with medical intervention) seen in three randomized trials conducted during the period 1983–2004.^{3, 4, 5} The surgical stroke-prevention benefit in the two larger trials, ACAS and ACST, was estimated to begin about 1–2 years after surgery, becoming significant by 3–5 years.^{4, 5} No data on long-term outcomes are available. Recommendations to perform surgery have usually come with the caveat that the perioperative risk of stroke or death at the operating center should be less than 3% (the average risk observed in the randomized trials).¹

Conversely, CEA for asymptomatic patients has been discouraged by some authors because the benefit seen in the randomized trials was small and highly dependent on operative stroke and death rates.¹ There is no evidence that general surgical standards have improved since these randomized trials were conducted. On the contrary, concern has been raised over higher stroke and death rates reported elsewhere, especially when relatively few procedures are performed.¹ Of particular concern to clinicians advising patients is that the rates of major surgical complications of local service providers are not usually known, making it impossible to decide if a patient should undergo CEA. Much effort would be required to set up accurate measures of surgical outcomes in routine clinical practice.

In addition, the randomized trial results indicate that CEA is inefficient for stroke prevention. In ACAS, for example, the average annual absolute risk reduction of ipsilateral stroke with surgery was 1.18% and the perioperative stroke or death rate was 2.3%.⁴ Assuming the results of ACAS are generalizable, therefore, about 85 operations would be required to be 'ahead' by one ipsilateral stroke annually. This means that three ipsilateral strokes would be prevented by these 85 operations at the expense of two strokes (or deaths) caused by surgery.

Furthermore, the overall stroke-prevention benefit in the randomized trials was expensive. Using ACAS results once again, it would cost about AU$600,000 (US$550,000) just to perform the 85 operations required to be ahead by one ipsilateral stroke (assuming a cost of AU$7,000 [US$6,400 per operation⁶). The cost of preventing one stroke (from unspecified causes in patients with previous stroke or transient ischemic attack [TIA]) using either aspirin, cholesterol, blood-pressure-lowering agents or smoking cessation⁷ would be up to 300 times lower than that of performing 85 CEAs. For a given health budget, considerably more strokes would be prevented by medical intervention than by surgery.

The cost-effectiveness of surgery for asymptomatic severe CAS might be improved if methods could be found to identify patients who—although receiving medical intervention—have a sufficiently high risk of ipsilateral stroke to warrant surgery. Despite much research,¹ this goal has not yet been achieved. Even if such patients are one day reliably identified, they might also face higher surgical risks. Thus, randomized trials would be required in this patient subgroup before surgical intervention could be justified in routine practice.

The case against surgery for asymptomatic severe CAS is further strengthened by recent evidence indicating that the risk of stroke or TIA in unoperated patients has fallen considerably over the last 25 years.¹ In addition, there are downward trends in average annual stroke rates. Recently, two trials observed the average annual rates of ipsilateral stroke in unoperated patients receiving medical intervention alone to be 1.7% (95% CI 0.9–2.5%)⁸ and 1.0% (95% CI 0–2.4%),⁹ while average annual rates of any-territory stroke were 2.1% (95% CI 1.3–2.9%)⁸ and 2.2% (95% CI 0.2–4.2%).⁹ Statistically, these rates are no different from the 2.2% average annual incidence of ipsilateral stroke seen in ACAS⁴ and the 2.4% average annual incidence of any-territory stroke seen in ACST⁵ among operated patients. Data reported from the SMART study this year show continuation of the downward trend in average annual stroke rates among patients receiving medical intervention for vascular disease.¹⁰

This 25-year decrease in stroke and TIA risk in unoperated patients is evidence that more-widespread use of aspirin, statins and, possibly, newer agents for blood pressure control is effective at preventing stroke caused by (or otherwise associated with) asymptomatic severe CAS. It should be added that the available data probably underestimate what can now be achieved with medical intervention alone, because the impact of initiating comprehensive vascular disease risk factor therapy—which includes ongoing patient education, and evaluation of efficacy using evidence-based treatment targets—has not yet been described. Furthermore, the recently observed low risk of stroke among medically managed, unoperated patients implies that CEA (even of the standard achieved in randomized trials) is no longer helpful in this setting.

The way ahead for patients with asymptomatic severe CAS is to measure the impact of comprehensively applied and combined medical (nonsurgical and nonendovascular) interventions that are known to be effective in the prevention of vascular disease complications in general. These interventions include nonpharmacological and pharmacological management of established risk factors and prescription of aspirin (if tolerated). The impact of these interventions needs to be measured in patients stratified by risk-factor load, to enable appreciation of, for example, outcomes for patients with and without symptoms of vascular disease. Vascular disease specialists could be consulted on a regular basis to provide education and assist with compliance matters. Some authors have suggested further randomized trials to identify the best strategy for managing asymptomatic CAS. Given the known limitations of a surgical approach in asymptomatic patients (discussed above) and the recent evidence supporting the effectiveness of current medical intervention alone, however, a strong case could be made for starting with well-designed and well-conducted observational studies, reserving randomized trials for patient subgroups with sufficiently high ipsilateral stroke risk, if such subgroups are identified.

In conclusion, the surgical stroke-prevention benefit in the randomized trials of asymptomatic severe CAS was small, expensive and highly dependent on both the operative risk of stroke or death and the efficacy of the medical intervention employed. This benefit has not been clearly replicated in the wider community. Medically managed patients with particularly high stroke risk who benefit from CEA have not been identified. Furthermore, recent evidence indicates that medical intervention alone, targeting vascular disease, has improved sufficiently over the past 25 years such that additional surgery is no longer helpful and might be harmful. Considering all these factors, I believe that, currently, surgery for asymptomatic severe CAS is not warranted.

Competing interests

The author declared no competing interests.

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