Guidelines for Carotid Endarterectomy

The introduction of carotid endarterectomy as a strategy for stroke prevention has sparked prolonged controversy with respect to its safety and efficacy. In response to genuine concerns expressed by many investigators, considerable effort has resulted in publication of retrospective reviews, natural history studies, audits of community practice, position papers by various societies, expert opinion statements, and finally prospective randomized trials designed to address specific issues and indications for surgery.

The American Heart Association assembled a group of experts in the field of cerebrovascular disease at a multidisciplinary conference to determine indications for carotid endarterectomy about which there is agreement, to highlight areas that remain controversial, and to recommend further studies or clinical trials.

The conference, held July 16 to 18, 1993, in Park City, Utah, included a review of the available data and preparation of this consensus statement.

Consensus Process

A program of critical topics was developed, and experts in each topic were identified and invited to participate in the consensus process. Each participant presented a topic to the group and prepared a brief written statement summarizing the data from his or her presentation. These summary statements were collated by the conference chair as a first draft of the statement.

After the formal presentations, a list of 96 potential indications for carotid endarterectomy was circulated to the participants. The indications were based on symptom status, percent stenosis, plaque characteristics, status of opposite carotid artery, and various levels of surgical risk. Each participant was asked to rate the indication for operation with one of four options: proven; acceptable but not proven; uncertain; and proven inappropriate.

When the individual indications for carotid endarterectomy were tabulated, the first draft of the position statement was read to the entire group of participants for discussion and comment. After discussion by the consensus panel, the draft was rewritten to reflect the comments. The following day the second draft was read to the entire group for final comment, after which the manuscript was revised to this final form.

Symptomatic Patients

Natural History

Symptoms related to carotid artery lesions include transient or persistent monocular visual loss, hemispheric transient ischemic attacks (TIAs), and ischemic stroke. Retrospectively, it is difficult to assign mechanism-specific stroke risk based on currently available data since an increasing number of recognizable factors that may affect subsequent stroke risk were not considered in the original studies. These include type of event, frequency of events, percent stenosis of the carotid artery in question, and characteristics of the plaque producing the lesion. Despite these recognized limitations, several observations are possible.

Patients with TIAs related to severe carotid stenotic lesions are at risk of stroke at the rate of 12% to 13% within the first year after onset of symptoms, with a cumulative stroke risk of approximately 30% to 35% at the end of 5 years. Those patients with hemispheric TIAs, recent TIA, increasing frequency of TIA, or high-grade stenosis have stroke rates that are probably higher than those with a remote or single event or lesser stenosis.¹² Similar findings have been noted by D.O. Wiebers, MD, and H.J.M. Barnett, MD (unpublished data from the Rochester Epidemiological Project and NASCET).

Patients who have had a stroke continue to be at risk for subsequent strokes at the rate of 5% to 9% per year, with approximately 25% to 45% of patients having another stroke within 5 years of the original event.³⁴⁵

Plaque characteristics may significantly affect subsequent ischemic events. Echolucent and heterogeneous plaques have a high content of lipid or intraplaque hemorrhage that may produce plaque ulceration, leading to a greater embolic potential.⁶⁷⁸ In a study of asymptomatic patients with carotid artery disease (CAD) only 20% to 30% of these patients had echolucent plaques, in contrast to symptomatic patients, in whom echolucent plaques were present in 70% of the cohort.^9101112 Cranial computed tomography of patients with carotid artery plaques demonstrates a 36% frequency of cerebral infarction in patients with echolucent plaques but only a 6% frequency in patients with echogenic plaques, suggesting that patients with echolucent or heterogeneous plaques have a neurological event rate two through four times greater than those with echogenic plaques.^910111213

An examination of the control group of the North American Symptomatic Carotid Endarterectomy Trial (NASCET) study demonstrated that patients with high-grade carotid stenosis in the absence of angiographic evidence of ulceration had a 2-year stroke rate of 17% in contrast to a 2-year stroke rate of 30% when ulceration was present with similar degrees of stenosis.¹⁴ Finally, the percent stenosis present in the proximal internal carotid artery is the most important plaque characteristic for subsequent neurological events, including stroke. This is true in both asymptomatic and symptomatic patients. The NASCET study control group demonstrated that for every 10% increase in stenosis beyond 70%, there was an increased rate of subsequent stroke risk.¹⁴

Patient Evaluation

Atherosclerotic carotid artery occlusive disease is part of a systemic disease. Evaluation of patients with ischemic strokes, TIAs, and suspected CAD should include a thorough history for the presence of coronary and peripheral vascular occlusive disease and stroke risk factors such as hypertension, tobacco and other substance abuse, use of oral contraceptives, hyperlipidemia, and diabetes mellitus. Neurological examination, blood pressure measurement in both arms and, when appropriate, a test for postural hypotension, measurement of pulse rate and rhythm, cardiac auscultation, and peripheral vascular examination are essential. Arterial pulses and bruits should be described.

Laboratory evaluation must include tests that define the presence, location, and severity of CAD. Ultrasound using pulsed Doppler has been accepted by some investigators in qualified laboratories as a satisfactory means of determining the severity of carotid artery stenosis. Duplex examinations (combined B-mode ultrasound and pulsed Doppler), when performed in settings in which the results have been consistently well validated by comparison with standard angiography, is an accepted and accurate technique for determining the severity of carotid artery stenosis. However, there is a risk of calling a high-grade stenosis total occlusion. This risk has been reported to be as high as 14% (false-positive rate)¹⁵ or as low as 1%,¹⁶ with a spectrum in between.¹⁷¹⁸¹⁹ Therefore, it is premature to make a definitive statement, since these techniques are still in evolution. Magnetic resonance angiography can also be helpful in providing images of the carotid artery, but at the present level of development it can overestimate or underestimate the severity of stenosis. Standard contrast angiography remains the gold standard for providing accurate images of the carotid arteries, the proximal vessels, and the intracranial circulation. In some patients being considered for carotid endarterectomy, duplex examination (in a laboratory whose accuracy has been validated) or duplex examination combined with magnetic resonance angiography may suffice for determining the severity of the extracranial portion of CAD, although arteriography remains the most reliable method of assessing the precise degree of carotid artery stenosis.

It is important to define how the degree of stenosis is measured. For instance, a 75% diameter stenosis will result in a cross-sectional area stenosis of around 90%. Similarly, if percent stenosis is based on diameter, and if the denominator is an estimated carotid bulb diameter, the percent stenosis will be greater than if the diameter of the internal carotid artery above the stenosis is used as the denominator. The carotid bulb is defined as the dilated (bulbous) first portion of the internal carotid artery just as it leaves the bifurcation. This is to be distinguished from the carotid sinus, ie, the terminal enlarged portion of the common carotid artery, which gives rise to the bifurcation of the internal and external carotid arteries. For consistency we believe that diameter stenosis, using the diameter of the internal carotid artery above the stenosis as the denominator, should be used. This is the measurement used in NASCET and the Asymptomatic Carotid Atherosclerosis Study (ACAS).

Studies that provide information about flow and/or pressure in the ophthalmic and intracranial carotid artery branches, such as transcranial Doppler, oculoplethysmography, and single-photon emission-computed tomography are also helpful for evaluating the severity and significance of CAD. Blood tests including hematocrit and platelet count are essential; prothrombin time and activated partial thromboplastin time are desirable. Brain imaging with computed tomography or magnetic resonance imaging is essential for patients with strokes and may be useful in patients with TIAs or asymptomatic patients who are considered for surgery.

An electrocardiogram is mandatory. Additional cardiac evaluation and consultation should be considered to seek potential cardiac sources of embolism that might have caused the brain ischemia and to assess the presence and severity of coexistent coronary artery disease. Neurological consultation is important for patients with neurological symptoms or signs. Some patients with concurrent medical illnesses (eg, pulmonary, renal, hematologic, hepatic, and other diseases) will require an evaluation to assess surgical risk.

Medical Management of Symptomatic Patients With Carotid Bifurcation Disease

This topic is covered only briefly here because the main purpose of the conference was to discuss indications for carotid surgery. Factors associated with increased risk of stroke related to CAD include age, hypertension, ischemic heart disease, diabetes, hyperlipidemia, diabetes, cigarette smoking, and high hematocrit. While it may appear axiomatic that elimination or reduction of these risk factors would be helpful for reducing stroke risk, there is very little evidence to support this hypothesis, with the exception of treating hypertension and stopping smoking.²⁰

One of the most important approaches to the medical management of patients with CAD with respect to stroke risk reduction has been the use of antiplatelet drugs, principally aspirin. Newer drugs, such as ticlopidine, have shown some improvement over the effect of aspirin.²¹²² Combination therapies such as aspirin plus dipyridamole are no better than aspirin alone. Recent aspirin trials and meta-analyses further clarify the effect of aspirin and its effect on stroke risk reduction.^23242526 In the most recent meta-analysis,²⁷ data from 18 prospective randomized trials of patients with a history of prior stroke or TIA were analyzed. In the antiplatelet drug treatment group, there were 479/5837 (8.2%) new nonfatal strokes; in the control group, there were 600/5870 (10.2%) new nonfatal strokes. Thus, antiplatelet drugs gave a 2% absolute risk reduction, or a 23% odds reduction, with respect to the end point of nonfatal stroke.

Results of Endarterectomy for Symptomatic CAD

Retrospective Reviews

The cumulative (immediate and long-term) results following carotid endarterectomy with respect to the incidence of subsequent stroke are influenced by the initial 30-day operative stroke morbidity and mortality. These rates are also influenced by the indication for surgery. Patients experiencing TIAs are at somewhat lower operative risk than patients who have had a prior stroke with varying degrees of recovery. A review of four recent series reporting operative experience in the management of symptomatic patients demonstrates a range of combined mortality and stroke morbidity from a low of 2% to a high of 6.1%.^28293031 Aggregating the results of the four series (modified meta-analysis) provided the opportunity to look at the results of treating 1498 patients with symptoms of TIA or prior stroke with a combined 30-day operative mortality and stroke morbidity of 2.74%.^28293031 Further insight into the risk of surgery is gained from data provided by participants in the ongoing ACAS. A review of the experience of surgeons applying to participate in this study provided an experience with 5641 carotid endarterectomies performed for a variety of indications with a combined operative mortality and stroke morbidity rate of 2.23%.³² Thus, the overall effect of carotid endarterectomy on immediate and subsequent stroke risk will be materially influenced by both patient selection and selection of the operating surgeon.

Late Stroke Risk After Carotid Endarterectomy in Symptomatic Patients

Patients who have undergone successful carotid endarterectomy for the indication of TIA (without perioperative stroke) continue to be at risk for subsequent ipsilateral hemispheric stroke at the rate of 1% to 2% per year.^{33343536373839}

Patients who have had successful carotid endarterectomy for the indication of prior stroke, without neurological complication, are at risk for subsequent ipsilateral stroke at the rate of 2% to 3% per year.^{28404142434445}

Position Statements

A number of organizations^464748495051 have issued position statements on indications for carotid endarterectomy; for the purposes of this document, statements by the following three organizations were reviewed: the Society for Vascular Surgery/International Society for Cardiovascular Surgery (SVS/ISCVS),⁴⁹ the Rand Corporation,⁵⁰ and an ad hoc committee of the American Medical Association (AMA) (D.B. Matchar, MD, C.T. Huesgen, MD, W.S. Moore, MD, unpublished data). The AMA subcommittee that reviewed the SVS/ISCVS recommendations and compared them with the Rand panel found that the two statements were in substantial agreement.

Indications for Carotid Endarterectomy Based on Prospective Randomized Trials

To date, NASCET, the European Carotid Surgery Trial (ECST), and the Veterans Administration (VA) Symptomatic Trial have reported definitive results. In the NASCET Study, 3 years after entering the first patient and after 659 patients with stenoses 70% or greater diameter reduction had been randomly selected, the “stopping rules” were triggered by the Data and Safety Monitoring Board of the National Institute of Neurological Disorders and Stroke. This occurred because there was a clear benefit in favor of carotid endarterectomy. The life table estimate of cumulative stroke risk in the distribution of the study artery at 2 years was 26% in 331 medical patients in contrast with 9% in 328 surgical patients, providing an absolute risk reduction of 17±3.5% (mean±SEM) (P≤.001).¹⁹ Similar and confirmatory findings were noted in the European and VA trials. The European trial noted significant benefit in favor of surgery for stenosis >69% but no benefit of surgery in symptomatic patients with <30% stenosis.⁵¹ The VA trial noted the benefit for prevention of stroke and crescendo TIAs with stenoses as low as 50%.⁵² NASCET and ECST have not reached definitive conclusions about symptomatic patients with stenoses between 30% and 69%, and both trials continue to randomize such patients. (Because ECST calculations use an estimated carotid bulb diameter rather than the internal carotid artery diameter as the denominator, 70% in ECST is equivalent to 45% stenosis in NASCET.)

Carotid Endarterectomy in Special Circumstances

Acute Carotid Occlusion

Most patients who have a neurological deficit associated with acute carotid occlusion are not candidates for carotid endarterectomy because of the severity of their deficits and/or a delay in diagnosis. A few patients may benefit from emergency surgery by virtue of proximity to a medical center with documented expertise in the management of cerebrovascular disease, having had a mild neurological deficit associated with an acute carotid occlusion, and having undergone a rapid workup (within hours of the event). As there are insufficient data either to support or refute this concept, this is an important topic for future study and trials. In the meantime, decisions must be made on a case-by-case basis, keeping in mind that there may be an opportunity to save carotid artery patency and neurological function with acute surgical intervention in carefully selected patients.

Evolving Stroke

Reports on natural history and attempts at surgical management are now more than 10 years old and predate current imaging technology. This makes conclusions in this area impossible and provides an important area for new investigation.^5354555657

Combined Carotid/Coronary Surgery

Patients with either symptomatic or asymptomatic CAD in the presence of symptomatic coronary artery disease represent a difficult decision matrix. The options include operating on the carotid lesion first, with an increased risk of morbidity and mortality from myocardial infarction; operating on the coronary lesion first, with an increased risk of perioperative stroke; operating on both lesions during the same period of anesthesia; or operating on the coronary arteries alone. Of the 57 English-language papers on this topic, only 19 report on more than 50 patients. A meta-analysis of 56 reports reviewed three operative strategies: simultaneous carotid and coronary artery bypass grafting (CABG), carotid surgery followed by CABG, and CABG followed by carotid surgery. The meta-analysis indicates that the perioperative stroke rate was similar if carotid and coronary surgery were combined or if carotid surgery preceded coronary bypass grafting. The frequency of stroke was significantly greater if CABG preceded carotid surgery. However, the frequency of myocardial infarction (P=.01) and death (P=.02) were greater when carotid surgery preceded coronary bypass grafting (Table).^{585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121}

The optimal strategy for management of patients with combined coronary and carotid disease will be established only by a well-designed prospective randomized trial.

Acute Carotid Dissection

Acute dissections of the internal carotid artery can occur either spontaneously or after blunt trauma. The pathology is an intimal tear with an intramural hematoma of the internal carotid artery that can remain stable, extend along the artery, or expand to produce a dissecting aneurysm. Thrombi from this site can extend or embolize intracranially.

This condition is treated medically. In those rare instances when focal ischemic symptoms recur despite medical treatment, and there is an appropriate vascular lesion, surgery may represent an alternative approach.^{122123124125126}

Asymptomatic Patients

Natural History

Early attempts to study the natural history of asymptomatic patients were flawed by inability to define the carotid artery lesion. These studies focused on finding a bruit in the cervical region.¹²⁷¹²⁸ It has now been well documented that bruits do not define the presence of a critical carotid lesion, nor are critical carotid lesions always associated with a bruit.^129130131 With the advent of carotid duplex scanning and more precise definition of the percent stenosis, natural history studies based on the extent and character of the carotid lesion are now possible. Three specific factors affect outcome: (1) percent stenosis (Figure) (2) progression between examination intervals (3) presence or absence of ulceration Evidence of brain infarction or embolization may also be important.^132133134 Patients with stenoses of the internal carotid artery >75%, as measured by Doppler examination, are at risk for stroke in the range of 2% to 5% within the first year of observation. An additional group of patients will develop TIAs as a harbinger of stroke risk and then fall into the symptomatic patient category. For those asymptomatic patients who pass the first year of observation without symptoms, the risk of subsequent stroke falls significantly. Of patients who develop stroke during the observation of an asymptomatic lesion, 83% had no warning symptoms.^{135136137138139140141142143}

The presence of ulceration, as documented by angiography, usually performed during evaluation of a symptomatic contralateral carotid artery, has also been a marker for subsequent stroke risk. The size and extent of ulceration has been correlated with the neurological event rate. Using conventional cut-film angiography, the ulcer size can be defined by multiplying the length and width of the ulcer in millimeters. Thus, ulcers that measure <10 mm² are defined as “A” ulcers; ulcers that range from 10 to 40 mm² are defined as “B” ulcers; and ulcers that exceed 40 mm², as “C” ulcers.¹⁴⁴ The presence of a “C” ulcer, independent of associated carotid stenosis, identified a group of patients who were at risk of stroke at the rate of 7.5% per year.¹³⁹¹⁴⁰ The presence of the small “A” ulcers was not associated with an increased stroke risk.¹⁴⁴¹⁴⁵ Controversy exists concerning the natural history of the “B” ulcer, with some reports suggesting that the stroke risk in patients associated with “B” ulcers was 4.5% per year¹³⁹¹⁴⁰ and other reports showing no relation between the presence of a “B” ulcer and subsequent stroke risk.¹⁴⁶¹⁴⁷

Results of Carotid Endarterectomy for Asymptomatic Patients

The rationale of carotid endarterectomy for asymptomatic carotid bifurcation lesions is based on the assumption that there will be a reduction in long-term stroke risk in the distribution of the artery. To achieve this objective, the following criteria must be fulfilled: (1) a lesion must be associated with a demonstrable stroke risk (2) removal of the lesion must eliminate or reduce long term stroke risk (3) the surgeon who operates on the asymptomatic carotid lesion must have a low rate of perioperative neurological morbidity and mortality. Perioperative and long-term results of surgical management are readily available from retrospective reviews. Comparison of these reviews with the natural history reports has resulted in a series of position or consensus statements. Finally, prospective randomized trials either completed or in progress will provide the most definitive evidence either supporting or refuting the efficacy of prophylactic carotid endarterectomy in asymptomatic patients.

Recent publications have documented the risk of surgery as ranging from 0.0% to 3.8%.^148149150151 A survey of surgeons participating in ACAS yielded an experience of 1511 operations for asymptomatic stenosis with a combined operative mortality and neurological morbidity of 1.7%.³² A review of an experience in the community of Rochester, NY, during the sample years 1984 through 1985 identified 226 carotid endarterectomies performed for asymptomatic carotid stenosis with a combined operative morbidity and mortality of 3.9%.¹⁵² Finally, the operative risk in the VA prospective randomized trial demonstrated a combined risk of death and stroke of 4.3% in 211 operations.¹⁴³ The AHA Stroke Council has stated that for carotid endarterectomy to be efficacious in asymptomatic patients, the target for combined perioperative death and stroke rate should be <3%.¹⁵³

A review of six series in the literature documented the outcome of patients followed for prolonged intervals after surgery for asymptomatic carotid stenosis. The ipsilateral annual stroke rate, including perioperative stroke, ranged from 0.7% to 2.0% per year.^{153154155156157}

Position statements concerning the use of prophylactic carotid endarterectomy have been published by the Joint Council of the SVS/ISCVS⁴⁹ and the Rand Corporation.⁵⁰ An ad hoc committee of the AMA (D.B. Matchar, MD, C.T. Huesgen, MD, W.S. Moore, MD, unpublished data), and an international consensus conference (A. Nicolaides, MD, unpublished data) have also issued statements on prophylactic carotid endarterectomy. The International Consensus Conference on Asymptomatic Patients With Carotid Bifurcation Disease concluded that

In an effort to clarify the role of carotid endarterectomy in reducing stroke alone or stroke and death, this consensus group supports randomization of patients with severe asymptomatic stenosis to ongoing prospective clinical trials. However, for patients unable or unwilling to participate in these trials, some members of this panel believe that carotid endarterectomy may be considered in good risk patients free of life threatening coronary disease by surgeons with low surgical complications. (A. Nicolaides, MD, unpublished data.)

Five prospective randomized trials were designed to study the efficacy of prophylactic carotid endarterectomy for the treatment of patients with asymptomatic carotid stenosis. The first trial to publish its results, the Carotid Artery Surgery Asymptomatic Narrowing Operation Versus Aspirin (CASANOVA) trial, concluded that carotid endarterectomy was not efficacious when compared with their control group. Unfortunately the study was seriously flawed.¹⁵⁸

The Mayo Asymptomatic Carotid Endarterectomy (MACE)¹⁵⁹ trial included too few observations of cerebral ischemic events to allow the authors to judge between the medical and surgical treatment groups. This trial was stopped early because of the frequency of a secondary end point (myocardial infarction) in the randomized surgical group, an end point that was significantly greater than in the medical part of the trial. Surgical patients in this trial had been discouraged from taking aspirin unless other indications (eg, cardiac) existed for its use.

The VA trial¹⁴³ was designed to test the hypothesis that carotid endarterectomy plus aspirin antiplatelet therapy would be more effective than antiplatelet therapy alone in reducing the incidence of neurological events, including TIA and stroke. The results of this study proved the hypothesis and demonstrated that the combined incidence of ipsilateral neurological events in the surgical group was 8% in contrast to 20.6% in the medical group (P<.001). Unfortunately, when the study was designed, the proposed sample size was not large enough to provide statistical power to show a difference for stroke alone. Nonetheless, after a 4-year follow-up interval, the ipsilateral stroke rate in the surgical group was 4.7% (including perioperative strokes), in contrast to 9.4% in the medical group (P=.056). However, when perioperative mortality (1.9%) was added to the surgical stroke rate, the difference between the two groups with respect to stroke failed to reach statistical significance. As a consequence, this trial has failed to resolve the controversy.

ACAS, the largest of the five trials, was originally designed to randomly assign 1500 patients to best medical management or surgery plus best medical management. ACAS investigators received permission to expand the sample size to ensure that there are a sufficient number of patients to be able to consider differences in the end point of stroke alone. At this time entry to the trial is complete, with 1662 patients randomized, and the follow-up phase is in progress.¹⁶⁰

On September 28, 1994, the NINDS issued a clinical advisory indicating a clear benefit in favor of surgery and that the trial has been halted. There was a relative reduction in risk for stroke of 55% in the surgical group compared with the group of patients who did not undergo surgery. The manuscript documenting the results of the ACAS study is in preparation.

The Asymptomatic Carotid Surgery Trial (ACST) has started randomizing patients from approximately 100 centers around Europe. Details on plaque characterization are included to test the hypothesis that echolucent plaques are more hazardous.

Analysis of Surgical Risk for All Indications

Carotid endarterectomy is a prophylactic operation designed to reduce the risk of subsequent stroke and stroke-related death. As such, the effectiveness of prophylaxis is directly related to the perioperative risk of operation. Operative risk is affected by patient selection, selection of surgeon, and the institution in which the operation is performed.

Operative risk, as a function of patient selection, is determined not only by a patient’s general state of health but also by the indication for surgery. A review of the medical literature clearly shows that the lowest complication rates occur with operations on asymptomatic patients. The risk of operation is slightly higher in patients suffering transient cerebral ischemia and appears to be highest in patients who have had a prior stroke, particularly in those who have a greater residual neurological deficit.³²¹⁶¹

It is not possible to assign a generic risk for the operation since major factors associated with operative risk will include the expertise of the surgeon performing the operation and the quality of care available in the hospital in which the operation is performed. Morbidity and mortality data extracted from retrospective reviews tend to be of limited value, inasmuch as they represent the experience of the surgeon or surgeons preparing the report. Population-based community audits provide better insight into the average risk of perioperative complications. These numbers, however, are influenced by the patient selection process and surgical expertise in the community being audited. Results of five community-based audits have demonstrated that the percentage of patients disabled or dead following carotid endarterectomy ranges from a low of 4.8% to a high of 9.0%.^{140162163164165}

An ad hoc committee of the AHA Stroke Council reviewed available reports and made recommendations with regard to the upper acceptable level of risk for combined death and/or stroke associated with carotid endarterectomy as a function of indication for surgery. The limits set were 3% for asymptomatic patients, 5% for patients experiencing transient cerebral ischemia, 7% for patients who have suffered a prior stroke, and 10% for patients undergoing surgery for recurrent stenosis.¹⁵³ More recently, many series have been reported with complications lower than the upper acceptable limits defined.

It is clear that some surgeons perform this operation with low risk while others have an unacceptably high complication rate; thus, it would be desirable to develop methods for auditing the individual surgeon’s practice of carotid endarterectomy and to limit surgical privileges to those who can document that their results fall within an acceptable range. The design of an audit should ensure objective, unbiased, factual information regarding surgical morbidity and mortality. Such an audit should include the following safeguards:

(1) An institutionally based, computerized registry of the results of all carotid endarterectomies should be maintained on an ongoing basis.

(2) An external review should be conducted by an independent observer, eg, a PSRO (professional standards review organization) nurse, of selected records of patients undergoing carotid endarterectomy to ensure that all major complications are accurately reported to the registry.

(3) Performance standards should be established to define acceptable surgical volume as well as upper acceptable limits of morbidity/mortality before a chart review is begun. Large, retrospective, multi-institutional data as well as recommendations by the AHA can be used to help define acceptable norms for each indication for carotid endarterectomy.

(4) The registry should be regularly audited, and the results should be made available to each surgeon promptly. Safeguards should be instituted to provide rigorous confidentiality.

(5) When complications significantly exceed the acceptable limits, an institutional peer review committee should investigate. If extenuating circumstances cannot be identified, appropriate corrective action, which can range from careful monitoring of the surgeon’s subsequent endarterectomies to actual loss of hospital privileges to perform the procedure, should be implemented.^166167168 The results of the audit also can be used to identify problems with and improve care within the institution.

Reporting Standards

One of the major difficulties encountered in reviews of clinical series of patients undergoing carotid endarterectomy is the great variation in presentation and analysis of data. These variations frequently preclude a definitive comparative analysis. After studying this issue in depth, an ad hoc committee of the SVS/ISCVS has suggested standards for reports dealing with cerebrovascular disease.¹⁶⁹ That report recommends standards for grading risk factors, for stratifying by clinical classification, for measuring carotid stenosis and ulceration, and for gauging early and late outcome. This report also recommends that the CHAT classification (Clinical presentation, History, site of Arterial disease, Target organ [brain]) be used to categorize or define subsets of patients under consideration. The CHAT method includes uniform criteria for comparing results among institutions as well as identifying specific patient subsets that may derive unusual benefit or be at unacceptably high risk for operation.¹⁷⁰¹⁷¹

Cost-Effectiveness of Carotid Endarterectomy in Stroke Prevention

Stroke is devastating, not only from the perspective of patient morbidity and mortality, but also because it is financially devastating to the family and society. The cost of acute treatment as well as long-term disability has been estimated to range from 9 to 15 billion US dollars per year in direct medical costs. The burden of indirect costs, such as lost wages, may equal or surpass this figure.

It has been difficult to determine the exact cost to the healthcare system of treating an individual with stroke due to the paucity of comprehensive surveys of costs for patients. Also, it is difficult to extract the true costs of medical resources used from either charges or actual reimbursement. However, Medicare claims data as well as published non-US reports suggest that the estimated average direct lifetime medical cost attributable to an incident of stroke range from $30 000 to $60 000.^172173174 For a patient who may require substantial posthospital services, this amount can be significantly higher. Clearly, more data on direct and indirect costs of the care of the patient with stroke related to CAD are needed.

The cost of carotid endarterectomy, including patient evaluation, special consultation, surgeon’s and anesthesiologist’s fees, and hospitalization, also varies considerably. Based on 1991 Medicare claims data, the average estimated cost for a carotid endarterectomy is approximately $11 000. The cost may be much higher depending on the extent of preoperative evaluation and complications. Counterbalancing this expenditure is the cost savings that result from strokes avoided. The symptomatic trials suggest that operating on 100 patients with high-grade (>70%) carotid stenosis would lead to 17 fewer strokes over 2 years. The savings in direct costs attributable to this reduction in stroke risk can be broadly estimated to range from $5100 to $10 200 per patient undergoing surgery. Improved quality of life for the patient and family and savings in indirect costs by preventing stroke are inestimable but undoubtedly quite large.

Cost-effectiveness analysis provides another useful perspective on carotid endarterectomy. For symptomatic patients, the strategy of performing angiography and operating on individuals with high-grade carotid stenosis is both medically beneficial and cost-effective. The additional cost per healthcare benefit gained (marginal cost-effectiveness ratio) compares favorably with other accepted medical practices. The cost effectiveness of carotid endarterectomy for asymptomatic stenosis will depend on the frequency and severity of neurological deficits in surgically treated versus medically treated patients. Until more data from randomized studies are available, this aspect of cost effectiveness cannot be predicted. However, the lower frequency of stroke in patients with asymptomatic stenosis suggests that the cost effectiveness of endarterectomy for this group will be less than that for patients with symptomatic lesions. For the same reason, it appears that the cost of large-scale screening programs would likely negate any potential cost benefit of surgery for asymptomatic stenosis. Therefore, unless we can further identify subsets of patients who are at particularly high risk for stroke, the benefit of performing carotid endarterectomy indiscriminately on asymptomatic patients with carotid stenosis will likely be largely outweighed by the cost of screening.

Definition of Terms Used in Indications for Carotid Endarterectomy

A list of 96 potential common indications was circulated to each conference participant. This list was based on symptomatic status, percent stenosis, plaque characteristic, status of opposite carotid artery, and various levels of surgical risk. The terms used are defined below. Each participant was asked to rank each surgical indication into one of four options: proven (score=1); acceptable but not proven (score=2); uncertain (score=3); and proven inappropriate (score=4). The scores were averaged for each of the 96 indications. Finally, the indications were aggregated again to make the presentation more manageable. Since many of the indications generated a range of scores, some participants rated a given indication higher (or lower) than other participants. For this reason, an average score was selected rather than attempting to find a unanimously acceptable score.

Definitions of Ranks for Surgical Indication for Carotid Endarterectomy

Four choices were available for each indication as a function of surgical risk. For asymptomatic patients, the options for surgical risk for combined stroke and death as a consequence of operation were <3%, 3% to 5%, and 5% to 10%. For symptomatic patients, the surgical risk options were <6% and 6% to 10%.

Surgical risk is based on a combined estimate of the patient’s general medical fitness to undergo surgery and the individual surgeon’s risk of morbidity and mortality for patients with a specific surgical indication.

A surgical indication that carries a high benefit-to-risk ratio would be acceptable in patients who were at higher surgical risk, whereas a surgical indication that had a lower benefit-to-risk ratio might be acceptable in only the best-risk patients.

Proven (Score=1)

This designation constitutes the strongest indication for carotid endarterectomy and strongly implies that to withhold surgery in the presence of this indication would be inappropriate under normal circumstances. Indications classified as proven are generally supported by data from contemporary, prospective, randomized clinical trials.

Acceptable but Not Proven (Score=2)

There is general agreement that this represents a good indication for surgery, with the expectation that benefits outweigh the risks. This rank is supported by promising, but not scientifically certain, data. Indications in this category may be the subject of ongoing prospective randomized trials. In that case, it is expected that patients will be offered the opportunity to participate in the trial. However, when this is not possible, either by geography or patient preference, surgery would be an acceptable alternative at the present level of knowledge.

Uncertain (Score=3)

There are insufficient data to define the risk/benefit ratio. These potential indications should be evaluated in clinical trials.

Proven Inappropriate (Score=4)

The current database is adequate to indicate that the stated risks of carotid endarterectomy outweigh the benefits. In general, the database includes contemporary, prospective, randomized clinical trials.

Definitions of Stroke Categories

Mild Stroke

The residual neurological symptoms and signs of a mild stroke cause no important functional impairment.

Moderate Stroke

The residual neurological symptoms and signs of a moderate stroke result in a loss of function that may be complete in one domain (eg, arm or leg function, speech loss) and incomplete in others, but the total functional loss still allows independent existence.

Severe Stroke

Residual neurological signs of a severe stroke are directly responsible for the patient’s loss of independence.

Current Indications for Carotid Endarterectomy

Asymptomatic Patients With CAD

For Patients With a Surgical Risk of <3%

1. Proven indications: none¹

2. Acceptable but not proven indications: ipsilateral carotid endarterectomy for stenosis ≥75% with or without ulceration, irrespective of contralateral artery status, ranging from no disease to total occlusion¹

3. Uncertain indications

• Stenosis <50% with a “B” or “C” ulcer irrespective of contralateral internal carotid artery status

• Unilateral carotid endarterectomy with CABG, coronary bypass graft required with bilateral asymptomatic stenosis >70%

• Unilateral carotid stenosis >70%, CABG required, unilateral carotid endarterectomy with CABG

4. Proven inappropriate indications: none defined

For Patients With a Surgical Risk of 3% to 5%

1. Proven indications: none

2. Acceptable but not proven indications: ipsilateral carotid endarterectomy for stenosis ≥75% with or without ulceration but in the presence of contralateral internal carotid artery stenosis ranging from 75% to total occlusion

3. Uncertain indications

• Ipsilateral carotid endarterectomy for stenosis ≥75% with or without ulceration irrespective of contralateral artery status, ranging from no stenosis to occlusion

• Coronary bypass graft required, with bilateral asymptomatic stenosis >70%, unilateral carotid endarterectomy with CABG

• Unilateral carotid stenosis >70%, CABG required, ipsilateral carotid endarterectomy with CABG

4. Proven inappropriate indications: none defined

For Patients With a Surgical Risk of 5% to 10%

1. Proven indications: none

2. Acceptable but not proven indications: none

3. Uncertain indications

• Coronary bypass graft required with bilateral asymptomatic stenosis >70%, unilateral carotid endarterectomy with CABG

• Unilateral carotid stenosis >70%, CABG required, ipsilateral carotid endarterectomy with CABG

4. Proven inappropriate indications

• Ipsilateral carotid endarterectomy for stenosis ≥75% with or without ulceration irrespective of contralateral internal carotid artery status

• Stenosis ≤50% with or without ulceration irrespective of contralateral carotid artery status

Symptomatic Patients With CAD

For Patients With a Surgical Risk of <6%

1. Proven indications

• Single or multiple TIAs within a 6-month interval or crescendo TIAs in the presence of a stenosis ≥70%, with or without ulceration, with or without antiplatelet therapy

• Mild stroke within a 6-month interval, in the presence of a stenosis ≥70%, with or without ulceration, with or without antiplatelet therapy

2. Acceptable but not proven indications

• TIA (single, multiple, or recurrent) within a 6- month interval, in the presence of a stenosis ≥50%, with or without ulceration, with or without antiplatelet therapy

• Crescendo TIAs in the presence of a stenosis >50%, with or without ulceration, with or without antiplatelet therapy

• Progressive stroke in the presence of a stenosis ≥70%, with or without ulceration, with or without antiplatelet therapy

• Mild stroke in the presence of a stenosis ≥50%, with or without ulceration, with or without antiplatelet therapy

• Moderate stroke in the presence of a stenosis ≥50%, with or without ulceration, with or without antiplatelet therapy

• Ipsilateral carotid endarterectomy combined with CABG in a patient experiencing TIAs, in the presence of unilateral or bilateral stenoses ≥70%, coronary bypass grafting needed

3. Uncertain indications

• TIA (single, multiple, or recurrent) with stenosis <50% with or without ulceration, with or without antiplatelet therapy

• Crescendo TIAs, with or without ulceration, and a stenosis <50%

• TIAs in a patient who requires coronary bypass grafting and has a stenosis <70%

• Mild stroke with carotid stenosis <50%, with or without ulceration, with or without antiplatelet therapy

• Moderate stroke with carotid stenosis <69%, with or without ulceration, with or without antiplatelet therapy

• Evolving stroke with carotid stenosis <69%, with or without ulceration, with or without antiplatelet therapy

• Global ischemic symptoms with ipsilateral carotid stenosis >75% but contralateral stenosis <75%, with or without ulceration, with or without antiplatelet therapy

• Acute dissection of internal carotid artery with persistent symptoms while on heparin

• Acute carotid occlusion, diagnosed within 6 hours, producing transient ischemic events

• Acute carotid occlusion, diagnosed within 6 hours, producing a mild stroke

4. Proven inappropriate indications

• Moderate stroke with stenosis <50%, not on aspirin

• Evolving stroke with stenosis <50%, not on aspirin

• Acute internal carotid artery dissection, asymptomatic, on heparin

For Patients With a Surgical Risk of 6% to 10%

1. Proven indications: none

2. Acceptable but not proven indications

• Single or multiple TIAs within a 6-month interval, in the presence of a carotid stenosis ≥70%, with or without ulceration, with or without antiplatelet therapy

• Recurrent TIAs, while on antiplatelet drugs, for a carotid stenosis ≥50% in the presence of ulceration, or ≥70% with or without ulceration

• Crescendo TIAs with a stenosis ≥50%, with or without ulceration, with or without antiplatelet therapy

• Mild stroke in the presence of a stenosis >70%, with or without ulceration, with or without antiplatelet therapy

• Moderate stroke with a stenosis >70%, with or without ulceration, with or without antiplatelet therapy

• Evolving stroke in the presence of a >70% stenosis with large ulceration

3. Uncertain indications

•Single TIA with stenosis <70%, with or without ulceration, with or without antiplatelet therapy

• Multiple TIAs within 6 months with stenosis <70%, not on antiplatelet drugs, with or without ulceration

• Recurrent TIAs while on antiplatelet drugs with stenosis <70%, with or without ulceration

• Crescendo TIAs for stenosis <70%, with or without ulceration, with or without antiplatelet therapy.

• Acute carotid occlusion with transient cerebral ischemia

• Acute occlusion with mild stroke

• Acute carotid artery dissection with continued symptoms while on heparin

• Patient with transient cerebral ischemia secondary to a stenosis ≥70%, in need of CABG, with or without contralateral stenosis, use of combined operation

• Mild stroke with stenosis <70%, with or without ulceration, with or without antiplatelet therapy

• Moderate stroke with stenosis <70%, with or without ulceration, with or without antiplatelet therapy

• Evolving stroke with stenosis <70%, with or without ulceration, with or without antiplatelet therapy

• Global ischemic symptoms with an ipsilateral stenosis >75%, with or without symptoms, irrespective of contralateral artery status, with lesions up to and including contralateral occlusion

4. Proven inappropriate indications

• Single TIA, <50% stenosis, with or without ulceration, not on aspirin

• Multiple TIAs within 6 months, stenosis <50%, not on aspirin

• Mild stroke, stenosis <50%, not on aspirin

• Moderate stroke, stenosis <50%, with or without ulceration, not on aspirin

• Evolving stroke, stenosis <50%, with or without ulceration, not on aspirin

• Global ischemic symptoms with stenosis <50%, with or without ulceration

• Acute dissection of internal carotid artery, no symptoms while on heparin

• Asymptomatic unilateral carotid stenosis ≥70% in patient undergoing CABG

Summary

A consensus conference sponsored by the AHA reviewed the current potential indications for carotid endarterectomy. Twenty-two committee members, representing the disciplines of healthcare policy, neurology, neurosurgery, and vascular surgery reviewed the current medical literature and ranked 96 potential indications for surgery. The votes for each indication were averaged and ranked by four categories: proven; acceptable but not proven; uncertain; and proven inappropriate.

While this statement represents the best current recommendation by an expert committee, it should be noted that many indications are currently the subject of study in prospective randomized trials. As these trials reach completion, it is likely that the recommendations in this document will be subject to change. It is the hope of the committee that this document will undergo periodic revision.

Appendix

Program for Consensus Conference (Wesley S. Moore, MD, Chair)

Surgical Risk

Risk of operation as a function of indication: retrospective institution/individual surgeon review, Richard F. Kempczinsky, MD

Risk of operation based upon community audit, Tom Brott, MD

Upper acceptable risk for performance of carotid endarterectomy: physician statement of Stroke Council of the American Heart Association, Hugh G. Beebe, MD

Method and rationale for auditing the individual surgeon: hospital privileges, Richard F. Kempczinsky, MD

Identification of Population at Risk

Importance of symptoms (hemisphere or monocular TIA, prior stroke, global ischemic symptoms) in predicting risk for subsequent stroke, David O. Wiebers, MD

Importance of percent stenosis as an independent variable for predicting stroke risk (symptomatic and asymptomatic patients), Robert B. Rutherford, MD

Importance of plaque characteristics (consistency and surface morphology) as an independent variable for predicting stroke risk (symptomatic and asymptomatic patients), Andrew Nicolaides, MD

Reporting Standards

The importance of establishing reporting standards in presenting and analyzing published reports of carotid endarterectomy results, Robert B. Rutherford, MD

The CHAT classification: will it provide better insight in patient selection, data analysis, and reporting results of carotid endarterectomy? Eugene F. Bernstein, MD

Status of Carotid Endarterectomy in Asymptomatic Patients

The natural history of patients with asymptomatic carotid stenosis, John Norris, MD

The natural history of asymptomatic carotid ulceration, Wesley S. Moore, MD

Immediate and long-term outcome of surgery for asymptomatic carotid stenosis, John J. Ricotta, MD

Results of the VA Prospective Randomized Trial of Carotid Endarterectomy for Asymptomatic Carotid Stenosis, Robert W. Hobson, MD

Results of the European Trial of Carotid Endarterectomy: the CASANOVA study, Wesley S. Moore, MD

Design and current status of the ACAS study, James F. Toole, MD

The SVS/ISCVS published position regarding surgery for asymptomatic carotid stenosis, Wesley S. Moore, MD

The Rand Study: position concerning operations for asymptomatic carotid stenosis, Louis Caplan, MD

The AMA committee position on surgery for asymptomatic carotid stenosis, David Matchar, MD

The international consensus statement concerning diagnosis and management of the patient with asymptomatic carotid stenosis, Andrew Nicolaides, MD

Status of Carotid Endarterectomy in Symptomatic Patients

The effect of antiplatelet drugs and risk factor control in reducing stroke risk in patients with TIA and prior stroke, David Thomas, MD

Immediate and long-term results of carotid endarterectomy in reducing stroke risk in patients with hemispheric and monocular TIA: a retrospective review, James Robertson, MD

Immediate and long-term results of carotid endarterectomy in reducing recurrent stroke risk in patients with prior hemispheric stroke who have made a good recovery: a retrospective review, Hugh H. Trout III, MD

Prospective randomized trial of symptomatic patients with carotid artery disease: current status and results of the NASCET study, H.J.M. Barnett, MD

Prospective randomized trial of symptomatic patients with carotid artery disease: results of the ECST (European) study, David Thomas, MD

Prospective randomized trial of symptomatic patients with carotid artery disease: results of the VA trial, Marc R. Mayberg, MD

Natural history and current status of carotid endarterectomy for stroke secondary to acute carotid occlusion, Hugh G. Beebe, MD

Natural history and current status of carotid endarterectomy for treatment of stroke-in-evolution and crescendo TIA, Jerry Goldstone, MD

Current status of managing patients with carotid bifurcation disease who are about to undergo cardiac surgery, Bruce Brener, MD

Traumatic dissection of the internal carotid artery: conservative vs surgical management, Arthur Day, MD

The SVS/ISCVS position statement on carotid endarterectomy in symptomatic patients, Wesley S. Moore, MD

The Rand study/position on carotid endarterectomy for symptomatic patients, Louis Caplan, MD

The AMA committee position on carotid endarterectomy for symptomatic patients, David Matchar, MD

Cost/Benefit Analysis

Current annual cost of stroke secondary to carotid artery disease (individual and cumulative nationwide) compared with the cost of prevention by population screening and extended application of carotid endarterectomy: a surgical perspective, Hugh H. Trout III, MD

A healthcare policy perspective, David Matchar, MD

“Guidelines for Carotid Endarterectomy” was approved by the SAC/Steering Committee on June 16, 1994. It is being published simultaneously in Circulation and Stroke.

• Download figure
• Download PowerPoint