Introduction

Cerebrovascular accident (CVA) is the third leading cause of death in the United States, surpassed only by heart disease and malignancy.¹ Stroke accounts for 10% to 12% of all deaths in industrialized countries. Almost one in four men and one in five women aged 45 years can expect to have a stroke if they live till 85 years of age. In a population of 1 million, 1600 people will have a stroke each year. Only 55% of these will survive 6 months, and a third of the survivors will have significant problems caring for themselves. As our population ages, the total number of people afflicted with stroke will continue to rise unless historic stroke rates decline in the future.²

The etiology of stroke is multifactorial. Ischemic stroke accounts for approximately 80% of all first-ever strokes, while intracerebral hemorrhage and subarachnoid hemorrhage are responsible for 10% and 5%, respectively. Of those strokes which are ischemic in nature, the majority are linked to complications of atheromatous plaques. The most frequent site of such an atheroma is the carotid bifurcation. Although the prevention of stroke in the general population has largely focused on the control of hypertension, a substantial number of strokes are preventable by the identification and treatment of carotid disease especially as the population ages.

Surgical endarterectomy of high-grade carotid lesions, both symptomatic and asymptomatic, has been identified as the treatment of choice for stroke prophylaxis in most patients when compared to “best medical therapy” (risk factor reduction and antiplatelet agents), as shown convincingly by the NASCET and ACAS studies.^3,4 More careful inspection of their respective results suggest that the risk of disabling stroke or death was 1.9% in NASCET, with a 3.9% risk of minor stroke. In ACAS, the risk of major stroke or death was 0.6% when one excludes the 1.2% risk of stroke caused by diagnostic arteriography (Table 24-1). Subsequently, carotid endarterectomy (CEA) has been performed in increasing numbers of patients, and now represents the most frequent surgical procedure performed by vascular surgeons. Despite the proven efficacy of CEA in the prevention of ischemic stroke, great interest has been generated in carotid angioplasty and stenting (CAS) as an alternative to surgical therapy. This chapter will examine the current role of CEA in the treatment of patients with stenosis of the cervical carotid arteries, will analyze the concept of “high-risk” CEA, and will discuss the evolving role of CAS.

Carotid Endarterectomy

The era of CEA was ushered in by the report of Eastcott, Pickering, and Robb in 1954. They treated a 66-year-old woman who had 33 episodes of transient cerebral ischemia (TIA); following operation, her symptoms resolved.⁵ Despite publication of satisfactory surgical results, the efficacy of CEA came into question during the 1970s and 1980s. Among Medicare beneficiaries, the frequency of CEA declined from 1985 (61 273 per annum) to 1989 (46 571 per annum).⁶ It was not until the 1990s that randomized trials of best medical therapy vs. CEA were undertaken. Following publication of the NASCET and ACAS trial results, the volume of CEA in the United States rose dramatically; again, among Medicare recipients, the incidence rose to 108 275 in 1996 following release of data from these trials. The Dartmouth Atlas of Vascular Health Care reported that the number of carotid endarterectomies performed between 1995 and 1997 nearly doubled, from 62 000 to 114 000 annually.⁷

The current indication for CEA include the following:

1. 
   

   Asymptomatic high-grade stenosis (typically >70% by duplex ultrasound)

   
   
2. 
   

   Symptomatic high-grade stenosis (typically >70% by duplex ultrasound)

   
   
3. 
   

   Symptomatic patients with moderate (>50%) carotid stenosis. These lesions are typically associated with deep ulceration and failure of medical (antiplatelet) therapy.

With the advent of CAS, CEA has again come “under attack,” despite its proven efficacy and durability in stroke prevention in patients with high-grade stenosis of the internal carotid artery (ICA).⁸ Proponents of CAS have suggested that the results of NASCET and ACAS are not achievable in general practice outside selected centers of excellence. The question is a reasonable one; if the combined stroke and death rate of CEA in asymptomatic patients were more than 3%, there would be little benefit of operation in the asymptomatic population.⁹ Both ACAS and NASCET included good-risk patients on the basis of reasonable life expectancy (so as to be available for follow-up) and exclusion of other potential causes of stroke (such as atrial fibrillation). Exclusion criteria included previous carotid surgery, prior myocardial infarction (MI), congestive heart failure (CHF), renal failure, unstable angina, and those requiring combined CEA and coronary bypass procedures. Tables 24-2 and 24-3 list inclusion and exclusion criteria for several important CEA and CAS trials. A review of 25 CEA studies reporting 30-day stroke and death rates by Rothwell et al.¹⁰ found a mortality rate of 1.3% in asymptomatic patients and 1.8% in symptomatic patients. The combined stroke and death rates were 3% in asymptomatic patients and 5.2% in those presenting with symptomatic carotid stenosis.

A number of studies have focused on NASCET and ACAS trial eligibility as they relate to the results of CEA in the general population. Lepore et al.¹¹ from the Ochsner Clinic reviewed 366 CEAs performed at their institution over a 2-year period. Surprisingly, 46% were found to be high risk based on NASCET and ACAS trial-ineligibility. Their cohort included 60% who presented with asymptomatic carotid stenosis; the remaining 40% had focal ipsilateral symptoms at presentation. The overall stroke and death rate (combined stroke and mortality, [CSM]) was 2.5%; trial-eligible “good risk” patients had a CSM of 1.5%, and the remainder (trial-ineligible) had a CSM of 3.6%. While there was a trend toward higher neurologic morbidity in trial-ineligible patients, this difference did not reach statistical significance (p = 0.17). These authors concluded that ineligibility for NASCET or ACAS should not be employed as a de novo indication for CAS. Illig et al.¹² examined the results of CEA at the University of Rochester in 857 patients. Stroke or death at 30 days occurred in 2.1%. Rates were similar in patients excluded from (2.7%) or included in (1.6%) NASCET and ACAS and in patients eligible (3.1%) or ineligible (2.1%) for ARCHeR, a CAS registry in high-risk patients. These rates did not differ according to whether exclusion or inclusion was based on anatomic risk, medical risk, or protocol exclusion; there was a trend, however, toward worse outcome in the high medical risk subgroup. Stroke and death rates were similar according to age, gender, repeat procedure, or the presence of contralateral occlusion.

Mozes et al.¹³ examined the results of 776 consecutive CEAs from the Division of Vascular Surgery at the Mayo Clinic in Rochester, MN. Patients were categorized as “high risk” based on the inclusion and exclusion criteria for the SAPPHIRE trial of CAS with cerebral embolic protection. Of 776 CEAs, 323 (42%) were considered high risk based on the criteria listed in Table 24-4. Clinical presentation was similar in the high- and low-risk groups (Table 24-5). The overall postoperative stroke rate was 1.4% (symptomatic: 2.9%, asymptomatic: 0.9%). When comparing high- and low-risk CEAs, there was no statistical difference in stroke rate. Factors associated with significantly increased stroke risk were cervical radiation therapy, class III/IV angina, symptomatic presentation and age ≤60 years. Overall mortality was 0.3% (symptomatic: 0.5%; asymptomatic: 0.2%), not significantly different between the high- (0.6%) and low-risk groups (0.0%). Non-Q MI was more frequent in the high-risk group (3.1 vs. 0.9%; p < 0.05). Of note, the only MIs that occurred in the entire series were nontransmural (non-Q). A composite cluster of adverse clinical events (death, stroke, and MI) was more frequent in the symptomatic high-risk group (9.3% vs. 1.6%; p < 0.005), but not in the asymptomatic cohort. There was a trend for more major cranial nerve injuries in patients with local risk factors, such as high carotid bifurcation, reoperation, and cervical radiation therapy (4.6% vs. 1.7%; p < 0.13). In 121 patients, excluded on the base of synchronous or immediate subsequent operations (who would have also been excluded from SAPPHIRE), the overall stroke (1.65%; p = 0.69), death (1.65%; p = 0.09), and MI (0.83%; p = 0.71) rates were not significantly different from the study population. The authors concluded that SAPPHIRE-eligible high-risk patients could undergo CEA with stroke and death rates well within accepted standards, and that patients with local risk factors were at higher risk for cranial nerve injuries, not necessarily stroke. These data bring into question the application of CAS as an alternative to CEA, even in high-risk patients.

While the previously cited studies do not support the premise that operative risk is higher in patients excluded from NASCET and ACAS, or trials of CAS in high-risk patients, there may in fact be categories of patients in whom CEA may not be optimal therapy. Hertzer et al.¹⁴ described the Cleveland Clinic experience for 2228 consecutive CEA procedures in 2046 patients from 1989 to 1995. The stroke and mortality rates for CEA as an isolated procedure were exemplary at 1.8% and 0.5%, respectively, for a combined rate of 2.3%. In addition, no statistical difference was found in stroke and mortality rates for asymptomatic patients, those presenting with hemispheric TIA, or those operated for stroke with minimal residua. Those patients having combined CEA and CABG had higher rates of perioperative stroke (4.3%) and death (5.3%) than those patients having isolated CEA. Carotid reoperations were also associated with higher stroke (4.6%) and death rates (2.0%). These data again lend credence to the idea that CEA can be performed safely in large groups of unselected patients, but may give some insight into categories of patients who are at increased risk for operative intervention.

A follow-up study from the Cleveland Clinic by Ouriel et al.¹⁵ attempted to identify a subgroup of patients who, upon retrospective analysis, were at increased risk for CEA, and therefore might be better served by CAS. From a prospective database over 10-year period, 3061 carotid endarterectomies were examined. A high-risk cohort was identified, based on the presence of severe coronary artery disease (requiring angioplasty or bypass surgery within the 6 months prior to CEA), history of CHF, severe chronic obstructive pulmonary disease (COPD), or renal insufficiency (serum creatinine greater than 3 mg/dL). (Figures 24-1 and 24-2) The rate of the composite endpoint of stroke/death/MI was 3.8% for the entire group (stroke 2.1%; MI 1.2%; and death 1.1%). This composite endpoint occurred in 7.4% of those considered high risk (n = 594; 19.4%), significantly higher than in those in the low-risk (n = 2467; 80.6%) category (2.9%; p = 0.008). Patients in the high-risk group were further subdivided into those who had CEA alone and those in whom CEA was combined with CABG. Not surprisingly, the incidence of the composite endpoint was greater in those having combined CEA/CABG than those having CEA as an isolated procedure. In those having CEA alone, the risk of death was significantly greater in the high-risk group (p < 0.001). Importantly, however, while the risk of the combined endpoint stroke/death/MI was greater in the high-risk group, this difference did not reach statistical significance (p = 0.078). In addition, the rates of the individual endpoints of MI and stroke did not differ statistically between the high- and low-risk groups. These data from the Cleveland Clinic vascular surgery registry seem to support the notion that patients enrolled in the multicenter trials of CEA (NASCET and ACAS) were likely similar to the low-risk group, while those in the high-risk group may not in fact have had such stellar outcomes if included in multicenter trials. Other authors have called into question the very idea of high-risk CEA; conflicting data exist as to factors such as high lesions, reoperations, cervical radiation, and contralateral carotid occlusion.^{16,17,18,19,20,21,22} Subsequent trials have therefore focused on medically compromised, high-risk patients as those who may benefit from an alternative procedure such as CAS.

High-Risk CEA

Significant disagreement exists regarding the definition of high-risk CEA. Based on data presented previously in this chapter, there is in fact a group of patients who are probably not best served by CEA—and should therefore be treated with CAS or with best medical therapy (especially asymptomatic patients felt to be at very high risk for any procedure). Tables 24-6, 24-7, 24-8 and 24-9 present data, principally from large single-center experiences, regarding the performance of CEA in various subgroups typically deemed high risk. Several points deserve emphasis. In Table 24-6, more than 5000 carotid endarterectomies are compiled from six centers, with a composite risk of stroke and death of less than 2%; note that approximately 40% of patients in these series were symptomatic, the remaining 60% performed in asymptomatic patients. Nevertheless, the results are similar to those reported for the ACAS trial, which was an entirely asymptomatic cohort of patients. Elderly patients (>80 years) represent a particular challenge with both open surgery (CEA) and CAS; conflicting data exists in the peer-reviewed literature regarding which treatment is optimal. The data presented in Table 24-7 suggests that octogenarians are well treated with CEA, and that results compare favorably with those in younger patients. Those patients with contralateral internal carotid occlusion can also be safely treated with CEA compared to those with patent contralateral ICAs, as reflected in Table 24-8. Finally, patients having restenosis following prior CEA and those with radiation-induced carotid stenosis also have acceptable risks of stroke and death following operation; the major risk in this group of patients is that of cranial nerve injury, which can be debilitating (Table 24-9). These data suggest that the recommendation for CEA or CAS must be individualized, and that the traditional concept of “high-risk for CEA” may be challenged, especially in centers that excel in endarterectomy.