♦ Initial Evaluation

Carotid stenosis, like all atherosclerotic disease, is the result of a combination of genetics and lifestyle choices. Similarly, the natural history of extracranial carotid atherosclerosis depends on the same factors. Patients who are at a higher risk for developing carotid stenosis are also at an increased risk for stroke from that stenosis. The initial evaluation of patients with carotid disease must be done bearing in mind the risk factors for both carotid stenosis as well as for stroke from carotid stenosis. For the most part, patients are referred to a carotid surgeon after the diagnostic workup has begun. For those patients who are asymptomatic, a bruit is usually heard by a primary care physician, prompting a carotid ultrasound, which in turn leads to either a referral or further imaging. Symptomatic patients either present with a transient ischemic attack (TIA) that is retinal or hemispheric or have had a stroke, and then the imaging protocol varies. Suffice it to say, carotid duplex ultrasonography remains the current standard for the initial, noninvasive, assessment of the extracranial circulation. The place of ultrasound in the imaging workup of ischemic stroke is covered extensively in Chapter 11, but a brief review is necessary to define its role when deciding to operate.

The sensitivity and specificity for spectral analysis of carotid duplex–derived waveforms for detecting carotid stenosis ranging from ≥50% to 99% varies from 90% to 95% when compared with conventional angiography.4 The spectral criteria published by Strandness5 at the University of Washington is probably the most commonly used by ultrasound laboratories throughout the United States to grade carotid stenosis and classifies the stenosis as normal; 1 to 15%, 16 to 49%, 50 to 79%, 80 to 99% stenosis; or complete occlusion. The University of Washington criteria focus on the peak systolic velocity (PSV) of the internal carotid artery (ICA), the amount of spectral broadening during the deceleration phase of systole, and the amount of plaque present, and these criteria are usually found at the bottom of a carotid duplex report. When using these criteria, duplex ultrasonography is extremely sensitive in detecting occlusive disease, but several short-comings make the University of Washington criteria less than ideal in the evaluation of carotid disease.

The technique for degree of stenosis measurement as used in NASCET has become the standard method for determining carotid stenosis ( Fig. 17.1 ).1 The University of Washington criteria used the carotid bulb instead of the distal internal carotid as the reference diameter; therefore, these criteria tend to overestimate the degree of stenosis. The range of stenosis provided by these criteria are too broad and do not fit with indications published by NASCET and ACAS. Symptomatic ICA stenosis ranging from ≥70 to 99% is best treated with endarterectomy and medical management, whereas ACAS demonstrated benefit from endarterectomy in patients with asymptomatic ICA stenosis ranging from ≥60 to 99%.1 ^, 2 Because of this shortcoming, Moneta’s group6 ^– 8 sought to establish other criteria for detecting ICA stenosis with a greater degree of accuracy. These criteria include the end diastolic velocity (EDV) and the ICA/common carotid artery (CCA) PSV ratio ( Table 17.1 ).

These criteria do increase the diagnostic accuracy of carotid ultrasound, but ultrasound leaves much to be desired when it comes to evaluating the level/location of the stenosis, proximal and distal anatomy, as well as the status of any collateral flow. Because of these shortcomings, it is the author’s view that performing a CEA based on carotid ultrasound as the sole imaging modality is, at best, a poor decision. Patients with an ultrasound that suggests a high-grade stenosis should have at least one other confirmatory study before the decision to intervene is made. Adjuvant imaging modalities include computed tomography angiography (CTA), magnetic resonance angiography (MRA), and digital subtraction angiography (DSA).

Digital subtraction angiography has long been considered the gold standard in the evaluation of the cerebrovasculature, but it does have two significant drawbacks: it is invasive and therefore carries the risk of arterial injury, and the contrast dye is nephrotoxic. Conventional catheter-based angiography exposes the patient to the risk of a disabling stroke, and the incidence of transient or permanent neurologic deficits ranges from 0.17 to 2.63%.9 ^– 11 A recent study of 2243 patients undergoing cerebrovascular DSA for surveillance after intracranial aneurysm treatment found a very low (0.43%) complication rate.11 Although DSA does carry some risk, the benefits include the ability to measure the degree of stenosis exactly, evaluate proximal and distal anatomy, as well as evaluate the status of collateral flow ( Fig. 17.2 ). Both CTA and MRA can provide similar information regarding proximal and distal anatomy, but collateral flow can only be inferred as both tests are not truly dynamic evaluations. CTA does carry the risks of radiation exposure and nephrotoxicity, and some patients may not be able to have an MRA due to metallic implants or foreign bodies. That being said, both modalities have been compared with DSA extensively, and the pooled sensitivity and specificity for the diagnosis of a carotid stenosis that is ≥70% using CTA were found to be 85% and 93%, respectively.12 For the diagnosis of the same degree of carotid stenosis, MRA had a pooled sensitivity of 95% and a pooled specificity of 90%.13 The accuracy of either of these modalities is good enough to be used as a confirmatory test following a carotid ultrasound that suggests a high-grade carotid stenosis. Catheter-based angiography should be used when there is a discrepancy between the carotid ultrasound findings and those of the confirmatory test.

♦ Initial Decisions: Symptomatic Carotid Stenosis

Many patients are referred to carotid surgeons with the diagnosis of “symptomatic carotid stenosis,” when in reality they are asymptomatic. The confusion exists because patients experience a “symptom” that is not referable to the carotid. Examples include dizziness, generalized weakness, syncope or near-syncope, and positive visual changes (floaters or scotoma). These do not qualify as symptomatic carotid ischemic events. The symptom has to be a transient or permanent focal neurologic deficit involving the ipsilateral hemisphere or retina in order for the carotid stenosis to be labeled symptomatic. The distinct separation between symptomatic and asymptomatic carotid stenosis is crucial to assessing a patient’s risk because any extrapolation on the natural history with medical management or on the benefits of revascularization is based on clear-cut definitions of the terms symptomatic and asymptomatic.

The degree of stenosis along with the presence or absence of symptoms, referable to the carotid stenosis, are the main factors to consider when deciding to intervene on a patient with ICA stenosis. A brief review of just how they shape that decision is also in order. The results of NASCET and ECST demonstrated that those patients with symptomatic carotid stenosis in the range of 70 to 99% benefited from CEA when compared with medical therapy.1 ^, 3 Those patients with a 50 to 69% stenosis also benefited, but there was only a 4.6% absolute risk reduction in ipsilateral ischemic stroke at 5 years compared with the 16% absolute risk reduction for stroke in the 70 to 99% subgroup. Those patients with near occlusions had no substantial benefit at 5 years. Rothwell et al14 performed a meta-analysis using patient-level data from NASCET and ECST and found that the greatest benefit from endarterectomy was in this high-grade stenosis group, with a 16% absolute risk reduction for stroke at 5 years (p <.001). Observations from NASCET also suggested that the benefit from CEA was greater in those patients who had a stroke as opposed to a TIA and if the ischemic event was hemispheric versus retinal.1 The findings of the NASCET and ECST studies led to the conclusion that the immediate risk of surgery is outweighed by the long-term reduction in stroke as a result of the revascularization of high-grade stenotic lesions. This conclusion, of course, is predicated on the achievement of an acceptably low surgical morbidity and mortality rate by the operating surgeon.

The American Heart Association Stroke Council ad hoc committees have published guidelines on the acceptable risk of CEA.15 ^– 17 They recommend that the combined risk of perioperative stroke and death should not exceed 3% for asymptomatic patients, 5% for patients with TIA, 7% for patients with stroke, and 10% for patients with recurrent stenosis if the benefits of CEA are to outweigh its inherent risks. In certain patient subgroups the risk of CEA is higher than 6%, and risk/benefit analysis becomes much more complicated than simply assuming that symptomatic high-grade ICA stenosis equals a CEA.

Factors such as surgeon experience, presenting symptoms, patient age and sex, medical comorbidities, anatomy, and plaque morphology all play a role in making the decision to perform a CEA. It is difficult to sort out whether the operative risk for CEA differs much among surgeons because of variables such as case mix and chance effects caused by the relatively small numbers in each surgeon’s experience. It is known, however, that surgeons who perform a large number of endarterectomies tend to have better outcomes than do those surgeons who perform CEA infrequently.18 ^, 19 There is no hard-and-fast rule as to how many CEAs one needs to perform in a given year to achieve satisfactory outcomes, and these criteria are usually left to the credentialing committee at each individual institution. Competency with regard to CEA for symptomatic ICA stenosis is usually based on the benchmark of a complication rate of ≤6%.17 Any complication rate higher than this should prompt a thorough review of the surgeon’s experience. The risks of CEA depend on the patients’ clinical presentation, their overall medical condition, and the surgeon’s personal experience and outcome data.

The multiple analyses by Rothwell et al14 ^, 20 ^– 23 regarding CEA for symptomatic carotid stenosis found that the nature of the presenting event had a statistically significant effect on the operative risk associated with CEA ( Table 17.2 ). Urgent CEA for crescendo TIAs or stroke in evolution is associated with a much higher risk than elective CEA, and it is felt that this discrepancy is due to several factors: the plaque in the carotid is considered acutely unstable with overlying thrombus; the risk of acute cardiac complications is increased from either systemic inflammation or atherosclerotic plaque instability; the patients lack the normal preoperative evaluation with correction of medical comorbidities; and there already might be irreversible ischemic damage.22 ^, 24 ^– 26 The combined risk of neurologic and cardiovascular complications following urgent CEA is high but not prohibitively so given the natural history of patients with unstable symptoms. When deciding to perform a CEA in an urgent/emergent fashion, the surgeon should be cognizant of the heightened risk associated with this scenario and take whatever steps possible to optimize the patient prior to the operation. Surgery for stroke is associated with a higher risk than surgery for TIA, and surgery for TIA is associated with a higher risk than that for ocular events only. The benefit of CEA, however, is directly related to the risk of the operation in these patients.

Both NASCET and ECST included very small numbers of patients in their 80s or older because of the perceived risks of CEA, risks of general anesthesia, and the presumed shorter life expectancy.1 ^, 3 The general thinking is that these patients are more at risk for stroke and have an increased morbidity and mortality secondary to stroke than does the younger population. Subgroup analysis of these studies have subsequently shown that the risk seen in these patients is not so much related to chronologic age as it is “physiologic” age.27 ^, 28 The elderly population tends to have more medical comorbidities, and this increases their risk of perioperative mortality but not the risk of perioperative stroke. When evaluating an elderly patient for a CEA, life expectancy and the yearly risk of stroke must be compared with the risk of CEA. Currently, there is no real justification for withholding CEA from elderly patients who are otherwise physiologically fit to undergo the operation.

The subgroup analyses of both NASCET and ECST demonstrated a decreased benefit of CEA in women due in part to the higher operative risk.1 ^, 3 ^, 23 ^, 29 In a meta-analysis, the overall odds of stroke and death were actually increased twofold in women, and one of the main reasons offered for this dramatic difference is that female carotid arteries are smaller than those in their male counterparts.23 ^, 29 The smaller size of the artery may create a more technically demanding operation for the surgeon, and one may theorize that a smaller artery is more prone to hemodynamic changes. Also, population studies have shown that women at any age have a lower risk for stroke compared with men of the same age.30 The results of NASCET and ECST support the principle that patients with lower risk profiles for stroke benefited less from CEA than those with higher risk profiles. The only problem with that notion is that although the perioperative risks for women undergoing CEA were higher than those for men, they experienced a similar long-term benefit from CEA.29 So it goes without saying that the surgeon should factor in these gender differences during the risk assessment for female patients.

The benefit of cervicocerebral angiography over carotid ultrasound is that it yields a full evaluation of the vessels above and below the lesion as well as the level of the stenosis. Lesions that are at or above the C2 level are often difficult to expose, as are lower-lying lesions. Extremely proximal or distal lesions are usually better suited for carotid angioplasty and stenting, as the risk for cranial nerve injury is high with CEA and virtually nonexistent for CAS. Patients with isolated hemispheres by angiography need intraoperative shunting, which can increase the risk of CEA in the form of vessel dissection. Patients with contralateral ICA occlusions are also at risk of ischemia during CEA if they do not have a patent contralateral posterior communicating artery. Finally, patients with intracranial aneurysms found incidentally during the workup for the carotid stenosis are theoretically at increased risk of rupture once the pressure gradient across the proximal stenosis is removed.31 ^, 32 Other factors, though not strictly “anatomic,” affect a patient’s anatomy in ways that definitely increase the risk of neurovascular injury during CEA. The scar tissue associated with previous neck surgery or irradiation increases the technical demands of CEA. These factors and others are listed in Table 17.3 . Although none of the above anatomic findings are absolute contraindications for CEA, their associated risks must be considered, especially in light of advances in CAS.

Prabhakaran et al33 found that plaque surface irregularity portended a threefold increase in ischemic stroke. They found that plaque surface irregularity, even after adjusting for degree of stenosis and plaque thickness, is an independent predictor of ischemic stroke. Patients with irregular or ulcerative symptomatic high-grade stenosis are at an increased risk of ischemic stroke when compared with those with nonulcerative lesions, but plaque irregularity is not clearly associated with an increased risk of stroke perioperatively with CEA.23 This subgroup of patients benefits greatly from CEA.