Asymptomatic Carotid Stenosis: Mainly a Medical Condition

The recent explosion of carotid stenting and endarterectomy for asymptomatic carotid stenosis is an alarming consequence, one hopes, of misguided beneficence and perhaps misguided enthusiasm for intervention. (The alternative, that cardiologists and interventionalists do this because they are venal, is invidious.)

My mentor, Henry Barnett (principal investigator of the North American Symptomatic Carotid Endarterectomy Trial [NASCET] and a giant in stroke prevention), called the presence of a carotid stenosis the “Mount Everest” indication for endarterectomy or stenting. (When Mallory was asked why he climbed Mount Everest, he famously replied, “Because it was there.”)

The use of stenting and carotid endarterectomy for asymptomatic stenosis is entirely based on mythology. Some of the myths are discussed below.

The notion that it is a good idea to open up an asymptomatic carotid stenosis is historical: it is based on the risk of medically treated patients in the Asymptomatic Carotid Atherosclerosis Study (ACAS) ¹ and the Asymptomatic Carotid Surgery Trial (ACST). ² Seldom do interventionalists consider the number of patients who would need to be treated (number needed to treat [NNT]) ³ to prevent an adverse outcome. Table 1 shows the NNT for various categories of patients with carotid stenosis. The NNT of 83 to prevent one stroke in 2 years by carotid endarterectomy in asymptomatic stenosis ⁴ is entirely predicated on the low (3%) risk of endarterectomy in clinical trials; in the real world, the risk is considerably higher (Table 2). ⁵ As Rothwell recently showed, ⁶ outcomes with stenting are even worse than those with endarterectomy.

OPEN IN VIEWER

Table 1.

Number Needed to Treat by Endarterectomy to Prevent One Stroke in 2 Years*

Category	NNT
Symptomatic severe > 70% age < 75 yr	6
Symptomatic severe > 70% age ≥ 75 yr	3
Symptomatic moderate 50–69%	15
Asymptomatic	83

Adapted from Barnett HJ. ⁴

NNT = number needed to treat.

*Predicated entirely on the low surgical risk seen in clinical trials.

OPEN IN VIEWER

Table 2.

Surgical Risk of Carotid Endarterectomy

3% risk of stroke or death in ACAS 1 and ACST 2

6% in NASCET 21 (1% death, 1% disabling stroke)

9.9% in CAVATAS 22 (6% death or disabling)

11.4% in an academic medical center 23

30-day mortality 4.5%

Medicare average 5.2% 30-day stroke/death 5

ACAS = Asymptomatic Carotid Atherosclerosis Study; ACST = Asymptomatic Carotid Surgery Trial; CAVATAS = Carotid and Vertebral Artery Transluminal Angioplasty Study; NASCET = North American Symptomatic Carotid Endarterectomy Trial.

It may not be obvious to modern-day physicians that statins were seldom used before 1994; that was the year the Scandinavian Simvastatin Survival Study (4S) ⁷ was published. It was the first study to show that statins reduced clinical events. That the ACST was published 9 years later has often been cited as a reason to expect that statins would have been more widely used at that time, but the study was done mainly in the United Kingdom (a bastion of therapeutic nihilism), and, in fact, intensive statin therapy was not in widespread use among patients in the ACST: at the beginning of the trial, only 40% of patients were on any statin, and by the end of the trial, only 70% were on any statin, and virtually all of the patients were on low doses.

In recent years, with more intensive medical therapy, the risk of asymptomatic carotid stenosis has declined remarkably. Two recent reviews showed that the risk has declined in a linear fashion and is now lower than the risk of endarterectomy or stenting. ^8,9 Abbott showed that by 2005, the risk of stroke ipsilateral to an asymptomatic stenosis within 1 year was below 1.5%, and by 2007, it was below 1%. ⁸ In 2007, in the Second Manifestations of ARTerial disease (SMART) study, the risk of stroke during 3.6 years of follow-up was only 3% among patients with asymptomatic stenosis of > 50% versus 2% for those with no stenosis. ¹⁰ As before, among patients with asymptomatic stenosis, there was no difference in stroke risk according to the degree of stenosis. Patients were three times as likely to have a myocardial infarction and 4.5 times as likely to have vascular death than they were to have a stroke.

In 2000, we began to study the question of how to identify among patients with asymptomatic carotid stenosis which ones would be at high enough risk to warrant endarterectomy or stenting. We based our study on the work of Molloy and Markus, who had shown that among a mixed population of symptomatic and asymptomatic patients, microemboli on transcranial Doppler ultrasonography identified a group with an eightfold risk of stroke. ¹¹

In 2005, we reported that among 319 patients with asymptomatic stenosis, 10% had microemboli on transcranial Doppler ultrasonography and a 1-year stroke risk of 15.6%. ¹² On the other hand, the 90% of patients with no microemboli had an annual stroke risk of only 1%, with very tight 95% confidence limits (1.01–1.36), so they could not possibly benefit from a procedure with a risk of 3% (or, worse, for stenting).

Based on this study, we obtained from the Heart and Stroke Foundation of Canada a grant to study the biology of patients with or without microemboli and with or without ulceration on three-dimensional ultrasonography. In the interim, we had implemented in 2003 in our clinics a new paradigm of therapy: treating arteries instead of treating risk factors. ^13,14 After 2 years, we had to report to the funding agency that we were no longer able to identify enough patients with microemboli to achieve the planned statistical power. In an effort to understand this, we analyzed the occurrence of microemboli and clinical events among patients with asymptomatic carotid stenosis before and since 2003. We found that microemboli, which strongly predict the risk of stroke, declined from 12.6% of patients before 2003 to 3.7% since 2003. More importantly, the risk of cardiovascular events also declined dramatically: strokes in the first 2 years declined from 8% to 1% of patients, as did the risk of myocardial infarction. The combined risk of stroke, death, or endarterectomy for the development of symptoms (transient ischemic attack [TIA] or stroke) declined from 17% to 5%. The risk of having a stroke with microemboli persisted in the face of more intensive therapy; patients without microemboli had a 1-year stroke risk of 1% versus 14% with microemboli.

These results have been criticized as unbelievable, but Marquardt and colleagues reported in 2009 from a prospective study in Oxfordshire, United Kingdom, that the average annual risk of ipsilateral stroke in patients with asymptomatic carotid stenosis was only 0.34%. ⁹

It is now apparent, therefore, that among patients with asymptomatic carotid stenosis treated with intensive medical therapy, very few (less than 5%) can possibly benefit from endarterectomy or stenting. The widespread practice of stenting or endarterectomy for such patients should therefore be regarded as inappropriate. ^8,15 Several of the myths used to justify inappropriate revascularization of asymptomatic carotid stenosis are discussed below.

Myths

Severe stenosis identifies patients at high risk, who will benefit from revascularization.

Although widely stated, this is simply not true for asymptomatic stenosis. In symptomatic stenosis, both the NASCET and the European Carotid Surgery Trial (ECST) showed that the benefit of endarterectomy was greater for more severe stenosis up to 95%. However, in patients with near-occlusion (with relative collapse of the internal carotid artery distal to the stenosis), the risk of stroke was similar to that of patients with moderate stenosis, and the benefit of surgery was less than for patients with severe stenosis. ¹⁶ This is explained by reduced flow with less likelihood of emboli and by the fact that patients who remain asymptomatic with near-occlusion probably have better collateral circulation, which develops as the stenosis progresses. Neither ACAS ¹ nor ACST ² showed any greater benefit of surgery in asymptomatic patients by degree of stenosis.

The notion that endarterectomy or stenting for asymptomatic carotid stenosis might be beneficial for patients scheduled for coronary bypass is also a myth; this was recently reviewed by Caplan. ¹⁷ Only 5% of perioperative strokes are related to large artery disease, and even in these cases, the causal relationship is not established.

Increasing blood flow to the brain is a good thing.

For the vast majority of patients, endarterectomy and stenting are not about increasing blood flow; they are about preventing embolization. I have a busy stroke prevention practice; for the past few years, I have been seeing over 900 new patients per year with TIA or nondisabling stroke; before that, it was about 500 per year. Without using toes, I can count on the fingers of both hands the number of patients I have seen with convincing hemodynamic TIAs, and most of these had carotid occlusion, not stenosis. It must be understood that a patient with carotid stenosis whose TIAs are in the vertebrobasilar territory has an asymptomatic carotid stenosis! This is why endarterectomy or stenting should be done only by neurosurgeons or on referral from a neurologist. Most other physicians are pathetically uninformed about neurologic diagnosis.

My skills are so good I can get away with it. Complications are for other operators.

This is pure hubris. In studies in which surgical results are audited with examination of patients postoperatively by neurologists, complication rates are always higher than surgeons think they are. Even if your skills are better than those of other operators, you cannot claim a complication rate of less than 1%, which is what it would take now to do better than intensive medical therapy.

Stenting is less invasive than surgery.

This sounds obvious and certainly appeals to patients. But shoving a catheter or a protective device up through a severe carotid stenosis breaks off emboli, which are readily detected by transcranial Doppler ultrasonography. ¹⁸ The recent meta-analysis by Rothwell (analyzing data from the SAPPHIRE [Stenting and Angioplasty with Protection in Patients at High Risk for Endarterectomy], EVA-3S [Endarterectomy versus Angioplasty in Patients with Symptomatic Severe Carotid Stenosis] SPACE [Stent-Protected Angioplasty versus Carotid Endarterectomy], and CAVATAS [Carotid and Vertebral Artery Transluminal Angioplasty Study] trials) showed that in symptomatic carotid stenosis, stenting is significantly worse than endarterectomy: stroke or procedural death occurred in 183 of 1,173 (15.6%) patients randomized to stenting versus 140 of 1,150 (12.2%) randomized to endarterectomy (relative risk 1.35; 95% confidence interval 1.06–1.71; p = .02). ¹⁹ The SAPPHIRE trial is widely cited as showing that stenting was safer than surgery in high-risk patients, but there were two crucial problems with that study: (1) the difference was significant only if reduction in myocardial infarction was included, and no thinking person would imagine that a carotid stent would prevent myocardial infarction, and (2) almost two-thirds of the patients had asymptomatic stenosis, so they would have been better off with medical therapy! Furthermore, at 5%, the risk of stenting in SAPPHIRE was well above the risk of modern medical therapy. Derdeyn in 2007 called for a moratorium on stenting of asymptomatic stenosis outside of trials and pointed out that such trials must include a medical arm. ¹⁵

Conclusions

With modern intensive medical therapy, the risk of stroke among patients with asymptomatic carotid stenosis is so low that, on average, such patients cannot benefit from endarterectomy or stenting. Only patients with microemboli on transcranial Doppler ultrasonography have a high enough risk to be considered for intervention. No clinical trial of stenting versus endarterectomy should now be done without a medical arm. Future imaging methods may offer additional approaches to identifying patients at high enough risk to warrant endarterectomy or stenting. Similarly, evidence of retinal embolism ²⁰ (Hollenhorst plaques) or computed tomography or magnetic resonance imaging evidence of silent infarction in the territory of a severe carotid stenosis might be regarded as justification for intervention. Routine stenting for patients with asymptomatic carotid stenosis should now be regarded as inappropriate.