Carotid Endarterectomy Data in Asymptomatic Stenosis

Outcome in Patients with Asymptomatic Neck Bruits Chambers/Norris NEJM 315:860-5, 10/2/86 OF/E/Ltr N=500 patients referred for carotid U/S with asymptomatic neck bruit followed for up to 4 yrs (mean F/U of 23 mo). Outcome: 24 TIA alone (16/24 [66%] in > 75% occlusion group; 113/500 [23%] had > 75% occlusion), 4 stroke preceded by TIA (2/4 [50%] with > 75% occlusion), 8 stroke without preceding TIA (4/8 [50%] with > 75% occlusion), 18 with myocardial ischemia (13/18 [72%] with > 75% occlusion), 16 with nonfatal MI (4/16 [25%] with > 75% occlusion), 27 with fatal MI (12/27 [44%] with > 75% occlusion), and 15 deaths from other causes. Of the 8 strokes without a preceding TIA, only 2 were in the hemisphere supplied by the stenotic carotid artery. The overall incidence of stroke at 1 yr was 1.7% (1% in patients without previous TIA), but 5.5% in patients with > 75% stenosis. In multivariate analysis, severe carotid stenosis, progressing stenosis, presence of heart disease and male sex were significant risk factors for stroke. The authors recommend that asymptomatic bruits be studied with Doppler U/S; the patients should be told to report TIA symptoms promptly, and then endarterectomy considered in patients with TIA. Note: carotid angiography was performed on 95 patients; complicated by minor stroke in 1 and major stroke in 1. Endarterectomy was performed on 19 asymptomatic patients (30 total); 1/19 died following surgery of an intracerebral bleed, and 4/30 had perioperative strokes. Letters criticize the high morbidity/mortality of angiography and surgery in this study, stating that other centers report angiographic morbidity of 0.5% (2% here) and endarterectomy morbidity of 2-3%. Comments: Although this study has problems (referral bias, selective attrition and surgical intervention), I think it clearly shows that the risk of an unheralded stroke in the distribution of the stenotic artery in patients with an asymptomatic bruit is quite low.

Risk of Stroke in Asymptomatic Persons with Cervical Arterial Bruits: A Population Study in Evans County, Georgia Heyman et al NEJM 302:838-41, 4/10/80 OF/E N=1620 persons > age 45 without previous stroke, TIA or other overt ischemic disease; 72/1620 (4.4%) with cervical bruit (carotid or supraclavicular). Of these 72 followed for 6 yrs, 10 had stroke (OR for men = 7.5, for women = 1.6). "Correlation between the location of the bruit and the type of subsequent stroke was poor"; however, only 3/10 strokes were spontaneous/ischemic strokes, and all of these were ipsilateral to a mid-carotid bruit. The other 7 strokes were: 2 massive intracranial bleeds, 1 stroke following an endarterectomy in a woman with bilateral carotid bruits, 2 men with supraclavicular bruits who had multiple cerebral infarcts, and 2 strokes that have poor documentation derived only from death certificates. The authors conclude that cervical bruits indicate only systemic vascular disease, and don't warrant an invasive W/U. Editorial makes the same point I made above. Comments: I am amazed that the authors make the claim they do; I hope and wish they are right, but their data doesn't support their conclusion.

Efficacy of Carotid Endarterectomy for Asymptomatic Carotid Stenosis Hobson et al NEJM 328:221-7, 1/28/93 OF/E (JW 11:26, 2/15/93) N=444 men with asymptomatic carotid stenosis with arterial lumen decreased by 50% or more by arteriogram randomized to aspirin with endarterectomy or aspirin alone. Follow-up for 4 years. The incidence of ipsilateral neurologic events was 8.0% in the surgical group vs 20.6% with the medical group (p< 0.001, giving a RR of event of 0.38 to the surgical group). The incidence of ipsilateral stroke was 4.7% vs 9.4%. The incidence of ipsilateral stroke and death during the first 30 postoperative days was not different between the groups; nor was there a difference between the groups in all strokes and deaths (surgical 41.2% vs 44.2% in the medical group). Most deaths were due to CAD. Comments: Basically this study says that people with > 50% asymptomatic carotid stenosis are likely to die of heart disease and have strokes; endarterectomy decreases the risk of ipsilateral stroke/TIA, but does not decrease the risk of total stroke/death. An editorial states that endarterectomy should have no place in the treatment of asymptomatic carotid disease.

Endarterectomy for Asymptomatic Carotid Artery Stenosis JAMA 273:1421-8, 5/10/95 OF/E N=1662 patients from 39 clinical sites with asymptomatic stenosis of > 60% randomized to endarterectomy or not (all patients also received aspirin and risk factor management). After median follow-up of 2.7 years, the aggregate risk over 5 years of ipsilateral stroke or any perioperative stroke or death was 5.1% for those treated surgically vs 11% for those treated medically (risk reduction of 53% [22-72%]). The risk of any major stroke or perioperative death was 9.1% for those treated medically vs 6.4% of those treated surgically (P = 0.26). The risk of any major stroke or death was 25.5% for those treated medically vs 20.7% for those treated surgically (P = 0.16). Comments: For me, the outcomes that are meaningful are major stroke or death (if I have a TIA, then I can address that issue and perhaps opt for surgery at that point). The risk of major stroke/death does not seem to be markedly different among the groups, and hence I would favor medical treatment until some other event identifies me at high risk for a major stroke.

Guidelines for Carotid Endarterectomy: A Statement for Healthcare Professionals from a Special Writing Group of the Stroke Council, American Heart Association Biller et al Circulation 97:501-9, 1998 OF For those with a surgical risk of < 3%, a proven indication for endarterectomy is stenosis > 60% (though many experts suggest holding off unless stenosis is > 80%). Whether an ulcerated plaque with stenosis is any more urgent is unclear. If surgical risk is 3-5%, it is acceptable to do endarterectomy if stenosis > 75% in the presence of contralateral ICA stenosis of 75-100%. If surgical risk is 5-10%, there is no proven indication for surgery, though it might be considered if someone is undergoing simultaneous CABG. In short, surgery is proven to provide benefit only when the surgeon and hospital have a good track record, the patient has a low perioperative risk, and there is stenosis > 60-80% (pick the number you want within that range).

Carotid endarterectomy for asymptomatic carotid stenosis: A meta-analysis Benavente O et al BMJ 1998 Nov 28; 317:1477-1480Several studies have shown that carotid endarterectomy improves the long-term prognosis of patients with symptomatic high-grade stenosis. This meta-analysis reviewed the evidence on the value of this procedure for *asymptomatic* carotid stenosis. Five randomized trials involving 2,440 asymptomatic patients with carotid stenosis of 50 percent or more were analyzed. Over a period of three years, the risk for ipsilateral stroke plus perioperative stroke or death was 6.4 percent for medically treated patients and 4.4 percent for surgical patients, a significant difference. The risk for stroke in any location was also significantly reduced (9.2 percent for the medically treated patients and 7.4 percent for the surgical patients). However, surgical patients had a 2.4 percent risk for stroke or death in the 30 days after randomization, compared with 0.4 percent for the medically treated group. Comment: It's now reasonably clear that surgery improves long-term prognosis in patients with asymptomatic carotid stenosis. However, the magnitude of this benefit is relatively small, and some patients will suffer significant complications of surgery early on. An accompanying editorial points out that 50 people would have to be operated on to prevent one stroke and that we need better ways of identifying which asymptomatic patients are at higher or lower stroke risk before recommending surgery. -- KI Marton

How Accurate Is Carotid Ultrasound? Clinicians rely increasingly on noninvasive carotid Doppler ultrasonography to make decisions about invasive carotid procedures. These 2 studies -- one from a veterans hospital and the other from an urban teaching hospital -- examined the accuracy of carotid ultrasound, using angiography as the gold standard. The first study included 334 vessels in 202 patients. Using peak systolic velocity to determine degree of stenosis, Doppler ultrasound correctly classified 91 percent of stenoses less than 50 percent, 33 percent of stenoses between 50 percent and 69 percent, and 87 percent of stenoses 70 percent or greater. Using the ratio of internal carotid velocity to common carotid velocity, the accuracies for the 3 degrees of stenosis were 89 percent, 46 percent, and 80 percent, respectively. The second study included 132 carotid arteries in 66 patients. Investigators determined the accuracy of ultrasound for classifying stenosis in 3 categories -- less than 60 percent, 60 percent to 79 percent, and 80 percent to 99 percent. There were discrepancies between ultrasound and angiography in 17 of 132 arteries (13 percent); in 13 of those discrepancies, ultrasound overestimated the degree of stenosis. Seven of the 17 discrepant readings were interpretive errors; in the other 10, there was no obvious cause for the discrepancy. Comment: These studies remind us that, though carotid Doppler ultrasound is reasonably accurate, it is not a perfect surrogate for angiography. Accurate assignment to the middle range of stenosis (50 percent to 69 percent in the first study) appears particularly problematic. -- AS Brett Ability to use duplex US to quantify internal carotid arterial stenoses: Fact or fiction? Grant EG et al, Radiology 2000 Jan; 214:247-252

Predictors of Progressive Carotid Stenosis Many patients undergo serial noninvasive carotid imaging to determine whether stenosis has progressed to a point at which surgery might be beneficial. These researchers from the Pittsburgh VA Medical Center determined predictors of progressive stenosis in 905 asymptomatic people (98 percent men). Each carotid artery was imaged by duplex scanning (mean, 3 times) during an average follow-up of 30 months. At baseline, 57 percent of arteries exhibited no stenosis, 18 percent showed mild stenosis (less than 50 percent), 14 percent showed moderate stenosis (50 percent to 79 percent), and 10 percent exhibited severe stenosis or preocclusive disease (80 percent or greater). About a third of the patients had progression (i.e., from less than 50 percent to 50 percent or greater stenosis, from moderate to severe stenosis, or from severe stenosis to occlusion). In a multivariate analysis, only 2 risk factors independently predicted progressive stenosis: lower HDL cholesterol, and higher pulse pressure (i.e., systolic minus diastolic blood pressure). For example, using cutoffs of 35 mg/dL for HDL and 80 mm Hg for pulse pressure, the 30-month probability of progression was about 35 percent for those with low HDL and high pulse pressure, but only about 15 percent for those with high HDL and low pulse pressure. Comment: These data eventually could prove useful in designing risk factor intervention studies for carotid stenosis and in guiding policy on the most cost-effective intervals for serial imaging. However, the findings should be validated in other settings and in women. -- AS Brett Etiologic factors in progression of carotid stenosis: A 10-year study in 905 patients Garvey L et al, J Vasc Surg 2000 Jan; 31:31-38

How Accurate Are Noninvasive Carotid Studies? Noninvasive carotid studies, especially Doppler ultrasound (DU) and magnetic resonance angiography (MRA), usually are performed to determine whether a patient is a suitable candidate for endarterectomy. Because the risks of the gold standard test -- intra-arterial angiography -- are not trivial, some surgeons operate based on noninvasive studies alone. But, how accurate are DU and MRA? Researchers from Duke University examined the records of 569 patients who had undergone conventional intra-arterial carotid angiography at 2 hospitals; 452 patients had previous DU, and 120 had previous MRA. When a noninvasive result agreed with the conventional angiogram in classifying a patient as suitable (50 percent to 99 percent stenosis for symptomatic patients; 60 percent to 99 percent stenosis for asymptomatic patients) or not suitable for surgery, the noninvasive result was considered to be accurate. For DU, the misclassification rate was 28 percent (22.5 percent false positives; 5.6 percent false negatives). For MRA, the misclassification rate was 18 percent (9.1 percent false positives; 9.1 percent false negatives). The misclassification rate was only 8 percent among the 40 patients who had undergone both DU and MRA with concordant results. Comment: These findings are only approximations, because the study was retrospective, and most patients had not undergone all 3 tests. However, the results raise questions about basing decisions about endarterectomy on noninvasive tests. The findings also suggest that clinicians should be prudent in ordering noninvasive carotid studies, because false-positive results ultimately will subject some patients to the risks associated with unnecessary angiograms. -- AS Brett Clinical carotid endarterectomy decision making: Noninvasive vascular imaging versus angiography. Johnston DCC and Goldstein LB. Neurology 2001 Apr 24; 56:1009-1015

Natural History of Asymptomatic Carotid Stenosis The Asymptomatic Carotid Atherosclerosis Study (ACAS) researchers projected that endarterectomy in asymptomatic patients with high-grade stenoses would reduce the 5-year incidence of stroke from about 11% to 5% (Journal Watch May 16 1995). However, the median follow-up in that study was only 2.7 years. For patients who do not wish to incur the small risk for immediate postoperative stroke, it is important to know the longer- term outlook without surgery. In this cohort study from Toronto, researchers followed 106 patients in whom asymptomatic carotid stenoses were diagnosed by Doppler examination. During a median follow-up of 10 years (range, 5 to 18 years), only 11 strokes occurred. Among patients with initial stenoses of 0% to 49%, the 10 - and 15-year actuarial risks for ipsilateral stroke were 5.7% and 8.7%, respectively. Among patients with 50% to 99% stenoses, 10- and 15-year actuarial risks were 9.3% and 16.6%, respectively. However, 95% confidence intervals around these estimates were very wide. Comment: Some experts have expressed concern that, because the ACAS was performed in experienced centers, its low perioperative stroke rate might not apply to the general medical community. Thus, the natural history of carotid stenosis without surgery is of great interest. Results from the current study suggest that only a minority of patients with asymptomatic carotid stenoses suffer strokes during a 10-year follow-up. However, these figures should be viewed cautiously, given the small sample size and the small number of strokes that occurred in the study. -- Allan S. Brett, MD Published in Journal Watch August 16, 2002 Source Nadareishvili ZG et al. Long-term risk of stroke and other vascular events in patients with asymptomatic carotid artery stenosis. Arch Neurol 2002 Jul; 59:1162-6