Loading ...

Optimized medical therapy (OMT) is as effective as surgery for carotid revascularization in reducing stroke risk in the majority of patients with carotid artery stenosis, results of a planned interim 2-year analysis showed.

Using a recalibrated model to estimate stroke risk — the carotid artery risk (CAR) score — investigators determined that as many as 75% of patients currently receiving carotid endarterectomy (CEA) could instead be treated with OMT alone.

The CAR score model uses contemporary data on the risk for recurrent stroke in patients on OMT that reflects improvements in medical therapy and includes some characteristics not used in the original score.

“We recommend using the CAR score to identify patients with carotid narrowing that can be managed with optimized medical therapy alone,” study author Martin M. Brown, MD, emeritus professor of stroke medicine, University College London Queen Square Institute of Neurology, London, England, told Medscape Medical News.

Investigators said the findings could lead to updated practice guidelines, although Brown stressed that further follow-up and additional trials are needed to confirm the new results.

The findings were published online on April 16, 2025, in The Lancet Neurology.

Time for an Updated Approach

Current guidelines recommend CEA in selected patients with recently symptomatic carotid stenosis of 50%-99% or asymptomatic stenosis of 60%-99%, Brown said.

But those recommendations were based on landmark trials carried out 30 years ago when medical therapy “was basically just aspirin,” said Brown. “Medical prevention of stroke has since improved dramatically.”

He estimates nonsurgical treatment today is about 80% better at preventing stroke than it was 30 or 40 years ago. “We reckoned these older trials were out of date and effectively needed to be redone,” he said.

Also, there was a need to revaluate the ability of the predictive risk model used in older trials — the European Carotid Surgery Trial (ECST) model — to reliably identify low-risk patients for whom medical treatment alone might be appropriate.

The CAR score predicts the risk for stroke in patients with carotid stenosis treated with OMT. This involves using much more intensive and effective therapy not available 30 years ago when the original score was used, Brown said.

In addition to stenosis severity, the CAR score considers older age, male sex, history of myocardial infarction (MI), hypertension, diabetes, and peripheral vascular disease, as well as symptoms of carotid territory ischemia and shorter time since symptom onset, among other risk factors.

The current 5-year Second European Carotid Surgery Trial (ECST-2) trial included 429 adult patients (median age, 72 years; 69% men) at 30 centers in Europe and Canada who had 50% or greater carotid stenosis. Symptomatic patients were required to have a predicted 5-year risk for stroke < 20% based on the CAR score.

The study also included patients with stenosis who had been asymptomatic for at least 180 days. They were assumed to have a low 5-year risk for stroke (≤ 5%) with OMT alone. About 40% of the study population were symptomatic and 60% asymptomatic.

Participants received OMT alone or OMT plus revascularization with CEA. OMT included a low cholesterol diet, target-adjusted cholesterol-lowering medication, antihypertensive medication according to blood pressure readings, and guideline-based antithrombotic therapy. Where applicable, patients were also encouraged to stop smoking, lose weight, and manage diabetes.

Follow-up visits were scheduled at 4-6 weeks, 6 months, and then annually. Brain MRI or CT was done at the time of randomization, before any revascularization procedure, and with follow-up imaging at 2 years.

The primary outcome for this 2-year interim analysis was a hierarchical composite of: Periprocedural death (within 90 days), fatal stroke, or fatal MI; nonfatal stroke; nonfatal MI; or new silent cerebral infarction on MRI or CT.

Wins and Ties

For the primary analysis, researchers used the win ratio method, which compares outcomes of random pairs of patients (one patient from OMT alone and one from OMT plus revascularization. The patient in the pair with the better outcome was deemed a “win.” If patients had similar outcomes, researchers recorded a “tie.” 

The win ratio was the number of wins with OMT alone divided by the number of wins with OMT plus CEA.

The win ratio is a better way of accounting not only for the severity of components of the composite endpoint but also the time to an event, Brown said.

“Using older methodology, if one person had a stroke after a week and another after 2 years, you’d say they both had a stroke,” he said. “But using the win ratio method, you’d say the patient of the pair who had the stroke later had the better outcome.”

He and others believed this method provides a more powerful way of analyzing the various primary outcome events than conventional noninferiority comparisons.

Complete 2-year follow-up data were available for 94% of participants (n = 404). The authors noted compliance to medical therapy was “remarkably high” for both groups.

The majority of pairs (77.3%) were recorded as a tie, meaning outcomes were similar regardless of treatment approach. Of the remaining pairs, the number of wins was similar between therapies, with 11.4% doing better with OMT alone vs 11.3% with OMT plus CEA (win ratio, 1.01; 95% CI, 0.60-1.70; P = .97).

As for components of the primary outcome, the OMT alone group reported four periprocedural deaths, fatal stroke, or fatal MI; 11 nonfatal strokes; seven nonfatal MIs, and 12 new silent cerebral infarctions. The numbers for the OMT plus CEA group were three periprocedural deaths, fatal stroke, or fatal MI; 16 nonfatal strokes; five non-fatal MIs; and seven new silent cerebral infarctions.

Prespecified subgroup analyses of the primary hierarchical composite looked at outcomes by symptomatic status, age, sex, CAR score, diabetes, hypertension, stenosis severity, contralateral stenosis or occlusion, and size of the medical center. These analyses found no significant group differences although the statistical power to detect potential differences in subgroups was low.

While the results suggest patients whose CAR score is < 20% can forego an operation, Brown believed additional risk factors can be added to the score — such as imaging of the carotid clot itself — to improve patient selection for OMT-only therapy.

Patients with higher risk for stroke — those with a CAR score over 20% — would still likely benefit from the surgical intervention, said Brown.

Updating Guidelines

Discussions are underway with vascular surgeons about recommending OMT alone in low or intermediate risk patients with the aim of potentially updating relevant guidelines, “but we’re not there yet,” study coauthor Paul J. Nederkoorn, MD, PhD, Department of Neurology, Amsterdam UMC, Amsterdam, the Netherlands, told Medscape Medical News.

An updated guideline would have a major impact in the stroke field, he noted. Treatment with OMT alone would also allow patients to avoid the risks for CEA, which include stroke, death, and wound infection and cranial nerve injury from the surgery.

In the Netherlands, about 2000 patients with carotid artery stenosis undergo revascularization. That figure is much higher in the United States because “a large number of patients there are operated on for asymptomatic stenosis,” said Brown. He estimated that more than 120,000 CEA or carotid artery stenting procedures are performed each year in the United States for asymptomatic stenosis.

“For the vast majority of these patients, in the Netherlands perhaps up to about 75%, medication alone is probably sufficient to reduce the risk of another stroke. It is a lot nicer for the patient not to have to have surgery and it saves a lot of healthcare costs,” Nederkoorn said in a press release.

The current study shows risks appear to plateau within the first 2 years, “suggesting that further follow-up might not favor revascularization,” the authors wrote, adding this needs to be verified.

That’s why the study follow-up has continued. The 5-year results of ECST-2 have actually been collected but have not yet been analyzed, said Brown.

In the current analysis, the COVID-19 pandemic often caused follow-ups to limit the number of patients with brain imaging at the 2-year follow-up and led to follow-up visits being postponed or conducted by telephone rather than in person. Another limitation was the accuracy of the recalibrated CAR score hasn’t been independently tested.

A ‘Pertinent Question’

In an accompanying commentary, Jan K. Ho, MBBS, Perron Institute for Neurological and Translational Science, Perth, Australia, and Graeme J. Hankey, MD, also of the Perron Institute and the University of Western Australia Medical School, noted several strengths of this new trial, including the investigators’ selection of patients according to their predicted risk for future stroke rather than only according to symptoms and degree of stenosis and the use of the win ratio.

But they also noted that the estimated win ratio of 1.01 is “imprecise” as reflected by its wide CI and the high proportion of ties between groups, due to the small number of outcome events.

“The estimated win ratio also incorporates the early perioperative risks of carotid revascularization without incorporating possible longer-term benefits of revascularization” beyond 2 years, they wrote.

Also weighing in, Steven Messe, MD, professor of neurology, Hospital of the University of Pennsylvania, Philadelphia, called the study “an important contribution” to the literature, but also noted the relatively small study size, which reduced the power to rule out clinically meaningful effects. In addition, 60% of patients had an asymptomatic stenosis, who typically have a lower risk for stroke with medical therapy than those with symptomatic stenosis.

As well, study subjects with symptomatic stenosis were included based on a relatively low 5-year expected stroke risk, which may not apply to many patients with symptomatic carotid stenosis, said Messe.

But the most important limitation is that patients were only followed for 2 years, he noted.

“Prior studies of asymptomatic carotid stenosis demonstrated that it takes time, at least 3-5 years, for the benefit of long-term stroke prevention to outweigh the upfront risk of performing the revascularization,” Messe said.

This study was funded by the National Institute for Health and Care Research (NIHR); the Stroke Association; the Swiss National Science Foundation; the Dutch Organization for Knowledge and Innovation in Health, Healthcare and Well-being; and the Leeds Neurology Foundation. Brown reported receiving grant funding for ECST-2 from the NIHR, Stroke Association, and Leeds Neurology Foundation, and reported getting a receipt of fees for the Courts on Claimants with carotid artery disease. Nederkoorn reported receiving grant funding for ECST-2 and a Dutch cost-effectiveness substudy from the Dutch Organization for Knowledge and Innovation in Health, Healthcare, and Well-Being. Ho and Messe reported no relevant conflicts of interests.