Natural History

The risk of stroke related to intracranial stenosis has often gone unrecognized and underreported in the past. Focus on the more easily treatable extracranial carotid and cardiac or aortic sources of stroke in fact may detract attention from coexistent intracranial disease. Although atherosclerosis of the extracranial cerebral arteries is more common, intracranial stenosis occurs in 20 to 40% of patients with atherosclerotic disease.3 ^, 4 The most common sites of intracranial disease include the intracranial internal carotid artery (ICA), the main trunk or M1 segment of the middle cerebral artery (MCA), the distal vertebral artery or vertebrobasilar junction, and the midbasilar artery. Thus, intracranial stenosis must be considered in patients with cerebral ischemic signs and symptoms. Population studies indicate that the incidence of intracranial stenosis is higher in blacks, Asians, and Hispanics than in whites.

After intracranial stenosis is identified, estimating the risk of the lesion is difficult. Specifically, estimates of the annual risk for stroke depend on the extent of stenosis and the location of the disease. However, calculating the percent of stenosis is more reliably achieved in the extracranial ICA than in the intracranial vessels. For example, given that the diameter of certain normal intracranial arteries is 2.5 mm, stenosis of 80% is a residual lumen of only 0.5 mm—a diameter difficult to accurately measure. As most clinicians consider stenosis exceeding 50% as significant intracranial stenosis, these patients are grouped together for clinical studies. Yet, by this measure, significant intracranial stenosis poses a high risk of stroke in virtually all locations.

Intracranial ICA stenosis has a high risk of ipsilateral stroke—reported to be as high as 7.6% per year.5 In the medical arm of the EC-IC bypass trial, the annual risk of stroke with ICA siphon or MCA stenosis was 8 to 10%.1 Furthermore, risk of stroke with posterior circulation disease may be even higher. In the Warfarin Aspirin Symptomatic Intracranial Disease (WASID) study, patients with significant basilar artery stenosis treated with aspirin had an annual risk of stroke of >10% despite medical therapy.6

Although most clinicians agree that there is a high incidence of stroke with intracranial stenosis, it is often difficult to predict which patients are at highest risk and what extent of stenosis poses a clinically significant degree of risk worthy of consideration for intervention. Although most clinicians consider intracranial stenosis exceeding 50% as significant, Borozan et al7 found that the onset of new symptoms occurred when stenosis averaged (± standard deviation) 35.4% ± 14.4%. The significance of presenting symptoms in the setting of intracranial stenosis is also unclear. In the EC-IC bypass trial, only one third of all major strokes were preceded by a transient ischemic attack.1 Furthermore Borozan et al found no difference in stroke-free survival between symptomatic and asymptomatic patients. Collectively, these results convey the difficulty that many clinicians face in predicting which patients are at highest risk for stroke.

Medical Therapy

Appropriate medical therapy for intracranial stenosis remains a matter of debate. Several retrospective studies as well as the observational arm of the WASID trial6 have suggested a reduced risk of stroke with warfarin therapy when compared with aspirin.8 However, in the 1995 prospective WASID trial showing fewer strokes in patients treated with warfarin (3.6%) than with aspirin (10.4%), the incidence of major hemorrhage was 8.3% with warfarin and 3.2% with aspirin, ultimately resulting in higher mortality rates as well (9.7% versus 4.3%, respectively).

Surgical Therapy

When the EC-IC bypass trial failed to demonstrate a benefit of surgical bypass, use of this procedure all but halted. Since its publication, detractors of the EC-IC bypass trial have argued that patient selection bias may have doomed the study because the investigators made no effort to differentiate between patient symptoms resulting from either embolic events or hypoperfusion.

New imaging techniques now provide valuable data regarding brain perfusion and metabolism in specific vascular territories. These techniques include xenon computed tomography (CT), single photon emission computed tomography (SPECT), positron emission tomography (PET), and CT perfusion. Using cerebral oxygen extraction ratios to augment cerebral blood flow data, Grubb et al9 classified cervical ICA occlusion and cerebral hypoperfusion into three stages: stage 1, vasodilatory compensation; stage 2, maximized vasodilatation with increased cerebral oxygen extraction; and stage 3, exceeding compensatory mechanisms. They reported that patients with stage 2 hypoperfusion faced a dramatically higher risk of stroke. Buoyed by these findings, a new, prospective, randomized trial for the treatment of cerebral hypoperfusion called the Carotid Occlusion Surgery Study (COSS) will examine EC-IC bypass.

Intracranial Angioplasty

Considering that the risk of stroke remains high after the best medical therapy, many endovascular surgeons opt for angioplasty, with or without stenting, to repair intracranial stenosis. Since the 1980s, angioplasty as a primary treatment of arterial atherosclerotic disease became widely accepted for coronary artery disease.10 Its application to intracranial disease first gained popularity in the treatment of arterial vasospasm following aneurysmal subarachnoid hemorrhage.11 Its popularity for intracranial atherosclerosis has been tempered by high procedural risks, including initial studies in which the rates of periprocedural neurologic events were as high as 33%.12 ^– 14 However, more recent series have demonstrated complication rates of <10%,15 ^, 16 which has been attributed to the use of smaller balloon dilation relative to the normal size of the target artery, the acceptance of some residual stenosis, and slower inflation times.

Several factors that affect angioplasty risk and success can be considered before treatment. Factors include the angiographic appearance of the lesion, the severity of stenosis, and the lesion’s location, length, and eccentricity. Mori et al17 evaluated several lesion characteristics (e.g., lesion length and eccentricity) and then assessed procedural results and angiographic follow-up. Their grading scale for intracranial stenosis was used to predict outcome after angioplasty based on lesion characteristics ( Table 31.1 and Fig. 31.1 ). In their 1998 follow-up study,15 the overall success of angioplasty, defined as uncomplicated reversal of >70% stenosis to <50%, was 79%, with clinical success in 76%. However, when lesion-specific characteristics were considered for types A, B, and C as defined in Table 31.1 , the clinical success rates were 92%, 86%, and 33%, respectively. For lesions of types A, B, and C, angiographic restenosis at 1 year was appreciated in 0%, 33%, and 100%, respectively, and the cumulative risk of fatal or nonfatal ipsilateral ischemic stroke was 8%, 12%, and 56%, respectively, at 1 and 2 years. The authors concluded that type A lesions were more often successfully treated with the lowest incidence of restenosis. More importantly, they demonstrated that preprocedural anatomic characteristics can be used to predict the success of angioplasty.

Accessibility of the lesion is another important factor. For example, highly tortuous proximal anatomy makes navigation of endovascular devices quite difficult and risks proximal dissection. In evaluating our success rates, we observed significant differences in the proximal anatomy. In measuring the radius of curvature of the proximal vascular loop (carotid siphon or vertebral C1 loop), we found a significantly smaller radius for unsuccessful cases (unpublished data) ( Fig. 31.2 ). Although stent devices are clearly more difficult to navigate than angioplasty balloons, greater difficulties can be encountered with virtually any device in more tortuous anatomy.