Are Asymptomatic Patients Likely To Benefit From Carotid Interventions?

Medicare has not approved stenting or surgical treatment of asymptomatic patients with carotid stenosis except in clinical trials. Most insurance companies cover treatment of asymptomatic patients. However, many of these patients also are in clinical studies. So asymptomatic patients are being treated.

Like many interventional radiologists, I am fairly conservative in my own practice but believe carotid revascularization is appropriate in asymptomatic patients with carotid stenosis over 80%.

Dr. Cleveland

Some arguments against intervention in asymptomatic carotid patients are based on historic data from registries. Any interpretation of benefit has to consider how good we are at treating patients today.

Results from all the carotid stenting trials to date show continuous improvement over time. We have gotten better at what we do--and there are data to support intervention. But before we look at the numbers, we need to understand what we are trying to treat. There are two issues. One is stroke; the other is carotid stenosis. We must determine whether observation is the best approach.

We know that coronary artery disease and carotid artery stenosis are linked. Increasingly, trials have looked at them as comorbidities. We also know that as the carotid artery wall increases in thickness, the risk of heart attack and stroke goes up (N. Engl. J. Med. 1999;340:14-22).

In symptomatic cases, if carotid stenosis is greater than 70%, the risk of stroke goes up 15 times, compared with the normal population. If a patient has greater than 90% stenosis along with other risk factors, the risk of stroke goes up 25 times (Neurology 2000;54:1022-8).

Dr. Katzen

Does the risk of stroke really increase based on the degree of stenosis? Or is stroke strictly an embologenic disease, in which case endarterectomy might be treating a totally different process?

The potential for stroke coming from other causes confounds the issue. About 40% of strokes in asymptomatic patients are not due to extracranial stenosis (JAMA 2000;283:1429-36). Hypertension is a risk factor, and the prevalence of high blood pressure increases with age. Diabetes and smoking also must be taken into account.

Now let's look at those numbers. The Asymptomatic Carotid Surgery Trial (ACST) enrolled 3,120 asymptomatic patients with stenosis greater than 70% (Lancet 2004;363:1491-502). Conducted in 126 hospitals in 30 countries, it compared early vs. deferred endarterectomy, with up to 5 years' follow-up.

The risk of stroke and death within 30 days of endarterectomy was 3.1%, which compares with the 3% maximum risk accepted in the American Heart Association guidelines for surgeons treating carotid stenosis.

The 5-year risk of stroke was 3.8% with immediate endarterectomy and 11% if treatment was deferred until stenosis advanced. Combining perioperative and nonperioperative strokes, the 5-year risk became 6.4% with immediate endarterectomy vs. 11.8% with delayed treatment.

Deferral of carotid endarterectomy until stenosis becomes severe actually increases the patient's risk of stroke. This is not insignificant. Half of all strokes in the ACST study were disabling.

In another major trial, the Asymptomatic Carotid Atherosclerosis Study (ACAS), investigators similarly showed a reduction in 5-year stroke risk from 11% to 5.1% in patients who had carotid artery surgery (JAMA 1995;273:1421-8). Though treatment halved the risk of stroke, some would argue that the risk is quite low, even among untreated patients, and that we would have to treat quite a few patients to prevent one stroke. If you are the patient with 80%-90% carotid artery stenosis, however, you look at your individual risk.

Cost is a big issue, but we need to consider the cost of stroke as well as the price of intervention. When a person has a stroke, there is a huge cost to society as well as to the patient and the patient's family, financially and otherwise.

Putting this all together, I believe the data are conclusive. As stenosis increases, the risk of subsequent stroke rises. Therefore, advanced stenosis should be treated, particularly in light of the low rate of stroke following carotid endarterectomy.

If you believe the literature in recent years, half to two-thirds of patients treated for carotid stenosis have been asymptomatic. Though many physicians clearly think there is justification for treating asymptomatic patients, I would argue there is little to be gained by the patient or the community.

Asymptomatic patients are healthy when you start treating them. The whole thing is about preventing stroke. If you have asymptomatic patients with 50%-99% stenosis of the carotid artery bed, they are two to three times more likely to die of other causes.

Stroke is a relatively small risk to this patient. And it does not really matter whether stenosis reaches 50% or 99%; increasing levels of stenosis do not necessarily relate to atherosclerosis. In 2001, a review of surgical literature by the Cochrane Review concluded that there was little benefit in surgery for asymptomatic patients.

Let's take another look at that Asymptomatic Carotid Atherosclerosis Study (ACAS). There was no benefit to women or to patients with increasing levels of stenosis. You would have to operate on 20 patients to prevent one stroke. That means 19 persons would have an unnecessary operation, with all the risks that surgery entails.

The Asymptomatic Carotid Surgery Trial (ACST) ended up with very similar results. You've got to keep the surgical event rate down to less than 3% to come out ahead. Even if you do that, you don't provide any benefit to women. Only men benefited in a subgroup analysis.

Turning to the literature on protected stent series from 1996 to 2003, we see 1,756 lesions, some of which may have been double reported. Admittedly, this is retrospective data. Nonetheless, the neurological event rate was 1.5%. With the Carotid Revascularization Using Endarterectomy or Stenting Systems (CARESS) trial, the event rate was 2% (J. Endovasc. Ther. 2003;10:1021-30).

Let's apply all this to 100 asymptomatic patients. Looking at a 5-year period, we provide benefit to six patients by intervening to reduce their risk of stroke. That means 94 patients will not benefit, or they will suffer actual harm. Six will have a stroke despite having a carotid stent to prevent stroke, and three will be hurt by their procedure.

If you delay the intervention, six will have mild strokes, which you will treat as you would a symptomatic patient, and six to eight patients will have major strokes. What this means is whether or not you intervene (and treat patients should they become symptomatic), six patients are going to benefit.

If you are thinking about carotid stenting as a cost-effective procedure, then consider that hospital charges are higher with carotid stenting than with surgery. If the ACAS data were applied across Australia (Med. J. Aust. 1995;163:197-200), preventing 3% of all first strokes would cost $1.5 million Australian per stroke prevented per year.

I am not alone in my reservations. Look at what the guidelines say about asymptomatic patients:

PIThe American Heart Association finds surgery is acceptable only if you keep periprocedural strokes and deaths to less than 3%.

PIIn 2002, the Intercollegiate Working Party for Stroke concluded there was "no indication in asymptomatic disease."

PIA year later, the European Stroke Initiative said carotid artery stenting is "not routinely recommended" for patients with asymptomatic carotid stenosis. It should be considered only in randomized controlled trials, according to that body.

Why would you want to do it? You are not helping the patient. You are not helping the community. It is not cost effective, and you introduce a substantial risk of harm.

At my institution, we consider stenting in asymptomatic patients who are going on to coronary artery bypass grafting. We do know some stenoses remain progressive, but at the moment, we cannot predict which stenoses will progress.

If interventional radiologists are going to do carotid artery stenting in asymptomatic patients, we need to generate data to show that it is the correct thing to do.