Symptom Status

The most important indicator of future stroke risk is the presence of recent (within 6 months) ipsilateral neurologic symptoms. Such plaques are already clinically “vulnerable” or “embologenic.” The stroke risk in the medical arms of two major symptomatic carotid trials^1–3 was much higher than that seen in the medical arms of two major asymptomatic trials,^4,5 as shown in Table 99-1. Further, the North American Symptomatic Carotid Endarterectomy Trial (NASCET) and European Carotid Surgery Trial (ECST) both demonstrated that stroke risk is highest within the first month after an initial neurologic event and drops off to approach that of an “asymptomatic” lesion by approximately 6 months.⁶ The observed reduction in stroke risk as the initial event becomes more remote likely reflects “stabilization” or healing of the plaque over time, either spontaneously or in response to medical therapy.

In symptomatic patients with high grade (70%-99%) stenosis, the relationship of the stenosis to future stroke risk is so strong that a causal relationship can be assumed. However, in symptomatic patients with more moderate degrees of stenosis (50%-69%), the benefit of carotid intervention is not as great, implying that the relationship between the bifurcation lesion and stroke is not as clear, and therefore, other potential sources of neurologic symptoms should be excluded before carotid intervention is undertaken. This includes a search for a cardiac source, interrogation of the intrathoracic and intracranial circulation, and brain imaging to evaluate both the presence of ipsilateral infarction (supporting carotid bifurcation intervention) and the existence of intracranial lesions (interdicting carotid bifurcation intervention).

Although ipsilateral neurologic symptoms are the strongest predictors of stroke risk, history of stroke of any kind is also associated with increased risk of subsequent stroke. In a large contemporary natural history study, the Asymptomatic Carotid Stenosis and the Risk of Stroke Study (ACSRS), any previous stroke or a previous contralateral transient ischemic attack (TIA) was an independent predictor of stroke risk in a patient with asymptomatic carotid stenosis.⁷ The reasons for this are unclear; these patients may represent a group with more aggressive atherosclerosis, or they may have intracranial lesions that compromise the ability to compensate for a subsequent event. Many years ago, Imparato⁸ observed that an individual patient tends to have the same type of plaque in both carotid arteries. This implies that patients with previous symptomatic carotid plaque are more likely to have vulnerable plaque in the contralateral artery. Whatever the explanation, it is clear that a totally asymptomatic patient and a patient with any previous neurologic events are not at equal risk of stroke.

Degree of Stenosis

In the major symptomatic trials, as well as in natural history studies,^1–7 the degree of stenosis was directly related to stroke risk. In NASCET, each decile of stenosis was correlated with an increase in stroke risk, and lesions from 70% to 99% were much more likely to progress to stroke on medical therapy than were lesions of 50% to 69% diameter stenosis, whereas lesions of less than 50% stenosis were unlikely to result in subsequent neurologic symptoms.^1,2 This finding of increasing stroke risk with increasing diameter stenosis was also demonstrated in the ECST trial.³ In asymptomatic patients, no linear correlation between diameter stenosis and increased stroke risk was seen in the medical arms of the Asymptomatic Carotid Atherosclerosis Study (ACAS) trial⁴ or Asymptomatic Carotid Surgery Trial (ACST).⁵ There are a number of potential explanations for this seeming inconsistency. First, the use of duplex ultrasound (DUS) for diagnosis may not have permitted as accurate a distinction between degrees of stenosis as was possible in ECST and NASCET; these trials quantified stenosis angiographically. Second, the low number of events in the asymptomatic trials may have resulted in inadequate sample size to demonstrate a difference between 60% and 79% and 80% and 99% stenosis. In the subgroup analysis of the ACST study, the mean diameter reduction of the more than 60% stenosis group was 69%, suggesting that the lesions in the trial were not evenly distributed across deciles of stenosis, but rather were weighted toward the “60% to 79%” category.⁵

Data on plaque character were not recorded in the asymptomatic trials. However, in ASCRS trial, 17% of patients had carotid plaques of “type 4 or 5” (primarily fibrous or homogeneous)¹² that were associated with an annual stroke risk of 0.7%, despite greater than 60% diameter stenosis.⁷ It is reasonable to assume a similar number of “benign” plaques were present in ACAS and ACST. Taken together, this may have resulted in insufficient numbers of truly high risk lesions in the medical arm of the asymptomatic studies to show gradation in risk between deciles of stenosis. Natural history studies of asymptomatic plaques with stenosis ranging from minimal to severe confirmed the increased risk of stroke associated with plaques with more severe stenosis.^9–11 Nicolaides et al¹² studied the natural history of asymptomatic carotid stenosis in 1121 asymptomatic patients who did not undergo carotid endarterectomy. In these patients, the carotid plaques were carefully measured by standardized and routine ultrasound techniques, analyzed in a core laboratory. Multiple clinical and anatomic characteristics, including degree of stenosis, plaque surface irregularity, and echolucency were recorded. Plaque diameter stenosis was calculated using the “ECST” method, which compared the diameter of the lumen with that of the entire artery at the point of maximal stenosis. Using this methodology, degree of stenosis proved to be a predictor of future stroke risk after multivariate analysis. It is likely that the rigorous standardization of duplex imaging and use of a core laboratory to evaluate images accounts for the differences between the ASCRS study and ACAS or ACST. It is of interest to note that analysis of asymptomatic contralateral stenoses in the NASCET study did find a correlation between degree of stenosis and stroke risk.¹³ The total 5-year stroke risk in the contralateral artery was 8.0% in patients with contralateral stenosis of less than 60% and 16.2% in those with contralateral 60% to 99% stenosis. The risk of “large artery” ipsilateral stroke for an asymptomatic 60% to 99% stenosis was 9.9% in addition to a 6.0% risk of lacunar stroke (many of which are now thought to result from artery to artery emboli). The overall 5-year risk of stroke was higher in the 75% to 94% stenosis category (18.5%) than it was in the 60% to 74% (14.8%) or the 95% to 99% (14.7%) groups. Although these patients were not identical to neurologically asymptomatic patients, these findings confirmed the relationship of angiographically determined stenosis and increased late stroke risk.

Because the diameter of most carotid arteries fall within the range of a few millimeters, diameter stenosis is a surrogate measure of plaque size or volume. Plaque size has been correlated with plaque composition; larger plaques are less likely to be fibrotic and more likely to contain lipid, intraplaque hemorrhage with a necrotic core, and inflammatory cells.^14,15 It is likely then that diameter stenosis is a surrogate measure for plaque size, and in balance, larger plaques are more often ulcerated and contain lipid or intramural hemorrhage. It is clear then that although stenosis is an important feature used to assess stroke risk, on a hierarchal basis, it is not as important as the presence or absence of symptoms.

Plaque Progression

Plaque progression has also been found to be associated with increased risk of subsequent stroke. The preceding natural history studies demonstrated that plaques that progressed from less than 50% to more than 50% stenosis while under observation were at increased risk of stroke, and this was even more marked when the progression was more than 80% stenosis.^16,17 Bertges et al,¹⁶ studied 1701 asymptomatic carotid arteries over a 9-year period and demonstrated that plaque progression increased the odds risk of subsequent stroke or TIA by 1.68 (P < .001). They found a 9.3% annualized risk of progression, influenced by the degree of initial carotid stenosis and blood pressure. More recently, Sabeti et al¹⁷ followed 1065 asymptomatic patients with carotid plaque for a mean of 3.2 years. They documented major adverse cardiovascular events (MACEs) in 40% of patients, progression of carotid diseases in 9.3%, and a correlation between carotid plaque progression and MACEs (odds ratio [OR], 2.01; P < .001), stroke (OR, 2.0: P > .035), and cardiovascular death (OR, 1.75; P = .039).¹⁷ It is likely that plaque progression is a manifestation of dynamic plaque events such as inflammation, intraplaque hemorrhage, and/or rupture of the fibrous cap and consequent exposure of necrotic plaque elements to the flow lumen. Thus, patients with plaque progression are at particular risk for subsequent cardiovascular risks, including stroke.

Presence of Contralateral Disease

The presence of contralateral carotid occlusion (CCO) may be associated with increased stroke risk in patients with significant carotid stenosis. Although a post hoc analysis of patients in the ACAS study with CCO¹⁸ demonstrated an decreased risk of stroke (3.5% at 5 years) in 168 patients from the medical arm of the trial, the explanation for this is unclear and has not been substantiated by other studies, including ACST.⁵ AbuRahma et al¹⁹ followed 82 patients with more than 60% stenosis and CCO and documented a combined neurologic event rate of 60%, including a stroke rate of 33% at a mean 5-year follow-up, with events distributed between ipsilateral and contralateral arteries. It is likely that the degree of intracranial collateral circulation supplied by the contralateral carotid and vertebral arteries is related to late stroke risk. The adequacy of collateral cerebral circulation can be measured by imaging collateral intracranial circulation or functional assessment of collateral flow by measuring cerebrovascular vasomotor reactivity. Several studies employing this technique have shown that patients with CCO and impaired collateral circulation are at an increased risk of late stroke, whereas CCO patients with relatively normal intracranial hemodynamics have a lower risk of late stroke.^20,21

Plaque Character

As noted previously (see Chapter 97), features associated with vulnerable plaque include plaque heterogeneity with increased echolucency, evidence of surface irregularity as manifested by ulcerations on angiography, a thin or disrupted fibrous cap with an adjacent echolucent plaque area, and increased inflammatory infiltrate as seen on magnetic resonance imaging (MRI) or a positron emission tomographic (PET) scan. Moore et al²² first described the relationship between surface irregularity (ulceration) and stroke risk more than 3 decades ago. Asymptomatic lesions that were ulcerated had an annual stroke risk of up to 12%. Fisher et al,²³ who studied plaques from both the NASCET and ACS trials, found that ulceration was more common in symptomatic patients, as was the presence of thrombus. The Oxford Plaque study confirmed an association between ulceration and histologic plaque rupture, large lipid core, macrophage infiltration, and general evidence of inflammation.²⁴ More recent studies of plaque from patients with remote symptoms suggested they have a higher incidence of unstable plaque characteristics than do patients who have never been symptomatic.²⁵ Many authors correlated echolucent areas of plaque seen by DUS with symptomatic potential, and this was quantified by Nicolaides et al,²⁶ who used a normalized grey scale to quantify echolucency. Nicolaides et al²⁶ also used normalized DUS to correlate plaque echolucency with the frequency of ipsilateral neurologic symptoms.^12,26 Plaque echogenicity was also correlated with the frequency of recurrent ischemic events in symptomatic patients²⁷ and the amount of embolic material recovered after carotid stenting.²⁸ The most comprehensive data on identifying “stroke prone” patients with asymptomatic carotid stenosis was from the ACSRS study.⁷ This study followed 1121 neurologically asymptomatic patients with known carotid plaque between 50% and 99% stenosis found on ultrasound. By using a combination of clinical and anatomic risk factors, the investigators were able to identify patients at very low (<5%), low (5%-9.9%), moderate (10%-19.9%), and high (>20%) 5-year risk of stroke. The risk factors they found associated with increased risk of stroke for patients with asymptomatic carotid stenosis included percentage of stenosis, smoking history, renal insufficiency, previous contralateral events, plaque area, and plaque character (grey scale median and presence of “discrete white areas” within the plaque).

Evidence of Clinically Silent Emboli

Clinically silent cerebral infarction can manifest in one of two ways: (1) evidence of previous ipsilateral cerebral infarction on computed tomography (CT) or MRI imaging of the brain, or (2) detection of embolic material in the ipsilateral middle cerebral artery using transcranial Doppler (TCD) techniques. In either case, the purpose is to identify emboli that originate from a carotid bifurcation stenosis but do not result in clinically detectable neurologic symptoms.

Routine CT scans taken before operation in asymptomatic patients demonstrated an incidence of silent cortical infarction in up to 20% of cases.²⁹ Subsequent studies correlated the presence of such infarcts with increased risk of subsequent ipsilateral hemispheric symptoms.^14,30 A systematic review of 19 studies of TIA patients demonstrated that evidence of acute ischemic lesions on diffusion weighted MRI was associated with an early increased risk of ischemic stroke.³¹ In addition, a prospective cohort study of asymptomatic individuals correlated the presence of microembolic “hits” detected by TCD with increased risk of subsequent clinical stroke.³² A review of 58 articles investigating the impact of TCD microemboli on subsequent neurologic events was recently reported by Jayasooriya et al.³³ The duration of TCD monitoring ranged from 30 minutes to 8 hours, with the most common intervals being 30 or 60 minutes. Several of the studies used multiple monitoring sessions. The prevalence of microemboli was higher in patients with neurologic symptoms. In the asymptomatic patients, between 4% and 30% of patients had microemboli, with a frequency of microemboli of 0.35 to 4 embolic signal events per hour. They found a positive correlation between an increased rate of subsequent stroke or TIA in patients with silent microemboli detected by TCD compared with patients in whom no microemboli were identified (28% versus 2%; P = .001). The authors concluded that there were level one data to support using TCD evidence of microembolization as a risk stratification tool, and that it might be used with other techniques, such as sophisticated carotid plaque imaging to identify increased stroke risk associated with asymptomatic carotid stenosis. This technique is both patient- and operator-dependent. Currently, these factors, along with events in asymptomatic patients and the time required for the study, have limited its clinical application.

Although plaque characteristics and evidence of silent emboli may help predict patients at highest risk for subsequent events related to carotid bifurcation stenosis, their absence does not infer minimal stroke risk in patients with otherwise clinically significant carotid bifurcation disease. There are currently no data to support these endpoints as superior to the more straightforward measurements of percent stenosis and plaque progression for selecting patients for carotid intervention. The major clinical trials did not consistently collect data on either plaque characteristics or evidence of silent emboli; therefore, it is impossible to determine what effect these may have had on their outcomes. The ability to reliably characterize a vulnerable plaque using techniques that are readily reproducible and widely available has not been established and is an area of active investigation. Similarly, although detection of previous embolic plaque activity in a neurologically asymptomatic patient using brain imaging or TCD monitoring has been associated with increased stroke risk, absence of these features has not been shown to imply minimal stroke risk in patients who have a high grade stenosis. It is hoped that use of these modalities in future clinical trials will put their clinical value in perspective. At present, however, the most reliable imaging predictor of stroke risk remains the accurate measurement of diameter stenosis. Factors associated with increased stroke risk are listed in Table 99-2.

The preponderance of data has shown that results with CAS are inferior to CEA in neurologically symptomatic patients. This has been true in the randomized trials of symptomatic patients comparing CAS and CEA^63–65 and in the subanalysis of symptomatic patients in the CREST trial.⁶⁶ This position is reflected in all of the societal guidelines published to date (see the following). Of note, a recent publication combining the results of the three large European trials comparing CEA and CAS in symptomatic patients suggested that the benefit of CEA over CAS (stroke rate 2.8% versus 9.4%; hazard ratio [HR], 3.4) is particularly high in the first week after symptom onset.⁶⁷ This is particularly relevant because most guidelines currently recommend early carotid intervention after stroke (see the following).

Hostile Neck

This is defined by neck anatomy with poor tissue planes or increased risk of postoperative infection. Such circumstances occur most frequently after major neck surgery involving dissection within the carotid sheath, radiation to the neck, or presence of a stoma in the cervical region. Each of these circumstances can increase the complication rate after CEA. It is important to note that a history of radiation, stoma, or surgery alone does not, by itself, define a “hostile neck”; rather, it is the degree of local change in the tissues that surround the carotid artery that is relevant. There are series reporting excellent results with CEA for recurrent carotid stenosis⁶⁸ and after cervical radiation.⁶⁹ These series likely include patients who are appropriately selected and operated upon in centers of excellence. It is generally accepted that cranial nerve injury is increased when CEA is performed for restenotic lesions, and this is likely also true with radiated fields. Increased wound complications after radiation and/or in the presence of a tracheotomy are well known. In general, these conditions represent a relative increased risk for CEA.

Risk Factor Reduction and Medical Management

A variety of clinical conditions have been associated with increased stroke risk: hypertension, diabetes, hyperlipidemia, smoking, and alcoholism. Efforts to control these conditions have been shown to reduce stroke risk in large population studies. In addition to pharmacologic interventions, treatment with antiplatelet agents and use of statins has been shown to reduce long-term stroke risk. These efforts form the cornerstone of the medical management of all patients at risk for stroke. Many of these interventions not only reduce stroke risk, but overall cardiovascular morbidity and mortality. The current data on each of these are briefly summarized and have been recently re-capitulated in a set of guidelines by the AHA.⁷⁶ BMT should be provided to all patients with carotid stenosis, stroke, or other manifestations of cardiovascular risk. The cornerstones of BMT are listed in Table 99-3.