Neurologists are generally reluctant to recommend surgical or interventional procedures to patients and recommend endarterectomy and sometimes stenting to treat suitable carotid artery stenosis. In the USA, carotid endarterectomy is one of the most common major surgical procedures performed on any group of patients. However, this enthusiasm for carotid revascularisation is increasingly being questioned and randomised clinical trials are being set up to test the hypothesis that modern medical treatment renders intervention unnecessary writes Fiona Kennedy, Roland L Featherstone and Martin M Brown
Surgery for symptomatic carotid stenosis
Carotid endarterectomy has been used as an effective treatment to prevent recurrent stroke in patients with carotid atherosclerosis since the 1950s. Two major trials, NASCET (The North American symptomatic carotid endarterectomy trial and ECST (European carotid surgery trial) confirmed that performing carotid endarterectomy in selected patients with recently symptomatic carotid artery stenosis of at least 50% stenosis measured by the NASCET technique was beneficial in the secondary prevention of stroke. In NASCET, the overall two-year risk of stroke in patients with stenosis measured between 70% and 99% was reduced from 26% in patients allocated medical treatment alone, down to 9% in those allocated surgery, with lesser but significant reductions in those with 50% to 69% stenosis. These results led to guidelines recommending carotid endarterectomy for patients matching the inclusion criteria for the trials, which still stand today.
In the last five years, large randomised clinical trials have been published assessing the safety and efficacy of carotid artery stenting compared with carotid endarterectomy. These have shown that for symptomatic patients, carotid endarterectomy is the safer procedure in the short term, but has similar long-term efficacy in the prevention of stroke. The largest trial for only symptomatic patients between endarterectomy and stenting was the ICSS (International carotid stenting study) which included 1713 patients.
The short term results of ICSS favoured endarterectomy over stenting. The 30-day per protocol analysis showed that the risk of stroke, death or procedural myocardial infarction was higher in the stenting group compared with the endarterectomy group (RR 1.83, 95% CI 1.83, 2.77, p=0.003). The long-term results (median follow up four years, maximum 10 years) presented at the European Stroke conference in 2012, however did not show any long-term difference in the primary outcome event of disabling or fatal stroke between carotid endarterectomy and carotid artery stenting.
Many neurologists have taken these results as indicating that stenting should never be offered to patients as an alternative to surgery. However, there is strong evidence that the excess risks of stenting are confined to older patients, and in patients younger than 70 years of age, the risks of perioperative stroke or death are similar after stenting and endarterectomy. The CSTC analysis showed that in patients younger than 70 years, the 120-day risk of stroke or death was 5.8% in patients in the carotid artery stenting group and 5.7% in the carotid endarterectomy group (RR 1.00, 95% CI 0.68 to 1.47); while in patients 70 years or older, the risk with stenting was twice that after endarterectomy (12.0% vs. 5.9%, RR 2.04, 95%CI 1.48 to 2.82, interaction p=0.0053, p=0.0014 for trend).
The best evidence concerning the benefits of surgery for asymptomatic stenosis comes from ACST (Asymptomatic carotid surgery trial), which randomised 3120 asymptomatic patients with 60%–99% carotid stenosis and also showed a significant benefit of endarterectomy over medical therapy alone. Over 10 years follow-up, the risk of stroke or perioperative death was significantly reduced in those allocated carotid endarterectomy by approximately 50% compared to those allocated deferral of any carotid procedure, but the absolute risks were low (13.4% vs. 17.9%) with a net gain over 10 years of only 4.6% (95% CI 1.2 to 7.9). These data have been widely interpreted as justifying a policy of routine carotid endarterectomy for asymptomatic stenosis, although enthusiasm for this approach varies from country to country.
There are very little data to support carotid artery stenting as an alternative to endarterectomy for asymptomatic stenosis. Most of the randomised evidence comes from CREST (Carotid revasularization endarterectomy vs. stenting trial), but it does not include sufficient numbers of asymptomatic patients to come to a firm conclusion. Certainly, the data do not justify the widespread adoption of stenting in some parts of the world for asymptomatic stenosis.
Part of the evidence challenging the need for carotid revascularisation comes from post-hoc analysis of the data from the original carotid surgery trials. Rothwell et al used the data from ECST to develop a model that predicted the risk of recurrent stroke in the trial patients on medical treatment and compared this to a model that predicted 30-day risk of surgery. The model was validated on the NASCET data set and showed that only patients with a high predicted risk of stroke during follow-up were likely to benefit from carotid endarterectomy. Rothwell showed that surgery appeared to be harmful or at least not beneficial in patients included in the trials whose characteristics predicted a low risk of recurrent stroke (which was about two thirds of the patients), yet it almost halved the absolute risk in patients who were at high risk. However, the main evidence challenging the need for revascularisation in patients at lower risk of stroke is the fact that medical therapy for the secondary prevention of stroke has improved.
Observational studies have shown that patients who take statins have a 30% to 50% risk reduction in recurrent stroke rate. With the new medical therapy available one has to question the validity of the older trials. No previous trials have specified targets for blood pressure and cholesterol levels and taking into consideration the new evidence regarding statin use, we could assume that the risk of recurrent stroke on medical therapy as quoted from ECST and NASCET should realistically be halved. Therefore new trials are required to investigate the effect of modern medical therapy on patients at lower risk of stroke from carotid stenosis.
Current trials of carotid revascularisation
The ECST-2 (Second European carotid surgery trial) is testing the hypothesis that patients who have clinical characteristics that predict a five-year risk of future stroke of <15% when treated with optimised medical thearpy alone will not benefit from early revascularisation by surgery or stenting due to the associated procedural risk. The trial will include patients with asymptomatic and symptomatic stenosis at low to intermediate risk of future stroke, who will then be randomly allocated between a policy of immediate revascularisation plus optimised medical therapy or optimised medical thearpy alone. Revascularisation can be done by either carotid endarterectomy or carotid artery stenting. In ECST-2, optimised medical therapy incorporates measures to encourage control of cholesterol and blood pressure levels with specified targets consistent with national guidelines, in addition to antiplatelet therapy. The results should eventually determine whether asymptomatic and low risk carotid stenosis should receive revascularisation therapy, and which patient characteristics are most likely to predict recurrent events.
Another ongoing investigator-led carotid trial is SPACE-2, based in German speaking countries in Europe. This trial initially randomised patients with asymptomatic stenosis between carotid artery stenting vs. carotid endarterectomy. The protocol has recently been modified to include a parallel arm in which patients will receive optimal medical treatment alone. In North America, CREST-2 is currently seeking funding for a trial of revascularisation vs. contemporary medical management alone.
In conclusion, our neurological view is that the body of evidence suggests that carotid endarterectomy is indicated for patients with recently symptomatic carotid stenosis and clinical features indicating a high early risk of recurrence, and the operation should be carried out between three and 14 days after the onset of symptoms. In comparison with stenting, endarterectomy is safer in the short-term in the majority of patients, but the two procedures have a similar long-term risk of disabling stroke or perioperative death. Stenting might be preferred because of its lower risk of complications related to the incision in patients younger than 70 years of age. There is also a cohort of patients unsuitable for carotid endarterectomy that might benefit from stenting so long as the interventionist maintains a low rate of procedural events. Regardless of the need for intervention, all patients should be on ‘optimised’ secondary medical prevention with good control of modifiable stroke risk factors. There may also be a proportion of patients with symptomatic stenosis in whom revascularisation is currently recommended, who will not benefit from surgery because the perioperative risk of stroke may equal or exceed the risk of recurrent stroke on optimised medical treatment. To facilitate decisions on the appropriate management of a patient with carotid stenosis, a multi-disciplinary approach is required and decisions should be made on an individual patient basis using the clinical trial evidence as guidance.
Fiona Kennedy, Roland L Featherstone, Martin M Brown are all with the UCL Institute of Neurology, London, UK