Clinical manifestations

Presentation and course

The neurologic syndromes associated with atherosclerotic occlusive disease of the major intracranial arteries are not unique. Other pathologies, such as cardioembolism or artery-to-artery embolism, that cause occlusions of these same arteries produce similar neurologic syndromes. Review of a patient's risk-factor profile, history of transient ischemic attacks, and temporal course of the neurologic deficit, however, may help the clinician distinguish between these different pathologies.

The presence of atherosclerotic risk factors such as hypertension, diabetes, smoking, or hyperlipidemia increases the a priori odds that a transient ischemic attack or stroke may be due to large artery cerebrovascular occlusive disease. Recurrent, stereotypical transient ischemic attacks should always suggest intrinsic large artery disease (extracranial large vessel disease or intracranial large vessel disease) rather than cardioembolism. Transient ischemic attacks tend to be more common in patients with extracranial carotid disease than in those with carotid siphon or middle cerebral artery disease, although there are frequent exceptions. In 1 study of patients with infarction in the territory supplied by the middle cerebral artery, the rate of transient ischemic attacks preceding stroke in patients with middle cerebral artery occlusive disease was 20% compared to 64% in patients with extracranial carotid occlusive disease (14). Transient ischemic attacks in patients with middle cerebral artery disease also occur over a shorter period than transient ischemic attacks associated with extracranial carotid disease (14).

The rate of transient ischemic attacks preceding stroke in patients with carotid siphon stenosis has not been well studied. In 3 retrospective studies, transient ischemic attacks were the presenting complaint in 28% to 47% of patients with carotid siphon stenosis (27; 55; 75). However, only 1 of these studies provided data on the rate of transient ischemic attacks preceding stroke—1 of 5 patients presenting with stroke had a previous transient ischemic attack (75). The rate of transient ischemic attacks preceding stroke in patients with vertebral or basilar stenosis has not been systematically evaluated in a large number of patients; however, a few small series suggests that at least 50% of patients with basilar artery disease have transient ischemic attacks preceding stroke (71; 63).

An understanding of the neurologic syndromes associated with occlusive disease of the major intracranial arteries requires knowledge of the territories of the brain supplied by each of these vessels. The reader is referred to detailed descriptions of these vascular territories in the monograph by Osborn (62). A brief summary of the common neurologic syndromes associated with occlusive disease of the major intracranial arteries follows.

Atherosclerosis of the carotid siphon typically involves the cavernous section. Infarcts in patients with stenosis or occlusion of the intracranial carotid artery usually involve the middle cerebral artery territory. These patients typically present with hemiparesis, aphasia (dominant hemisphere), anosognosia and neglect (nondominant hemisphere), hemianopia, and hemisensory loss. Partial syndromes are common depending on which division of the middle cerebral artery is primarily involved. Carotid siphon disease may also cause infarction in the anterior cerebral artery territory (18). Patients with an infarct in this territory often have leg weakness or leg and face weakness with relative sparing of the arm.

Atherosclerosis of the middle cerebral artery typically involves the stem (M1 segment) but occasionally affects the superior division alone (45). Infarction of the entire middle cerebral artery territory causes hemiplegia, hemisensory loss, hemianopia, gaze preference, and global aphasia (left hemisphere) or neglect syndromes (right hemisphere). Infarction in the territory supplied by the superior division of the middle cerebral artery causes hemiplegia and hemisensory loss (sometimes sparing the lower extremity), conjugate gaze preference, contralateral neglect (nondominant hemisphere), and a Broca-type of aphasia (dominant hemisphere). Vision is spared with these infarcts. Infarctions in the territory supplied by the inferior division of the middle cerebral artery in the dominant hemisphere cause Wernicke aphasia, hemianopia, and agitation, whereas the mirror image infarct in the nondominant hemisphere causes confusion, agitation, hemianopia, and poor drawing and copying (16).

Nonstenotic atherosclerotic disease of the middle cerebral artery stem that involves the origins of several lenticulostriate arteries causes a moderate-sized (2 to 5 cm) infarction involving the internal capsule and basal ganglia (striatocapsular infarction) with sparing of the cortical areas supplied by the middle cerebral artery. Typically, these patients present with pure motor hemiparesis, suggesting small vessel disease as the cause, but the clue that the vascular pathology involves the middle cerebral artery—and not a single small penetrating artery—is the relatively large size of the subcortical infarct (23).

The anterior cerebral artery is uncommonly affected by hemodynamically significant atherosclerosis. The majority of infarcts in the anterior cerebral artery territory are caused by emboli from the extracranial carotid artery, heart, or carotid siphon (10). When atherosclerosis is present, it is usually distal to the stem of the anterior cerebral artery (A1 segment) and involves the territories supplied by the pericallosal or callosal marginal branches (47). Infarcts caused by atherosclerotic occlusive disease of these 2 main branches of the anterior cerebral artery cause weakness, typically involving the foot and thigh and occasionally involving the shoulder, and cortical sensory disturbances. Language dysfunction is also common after left anterior cerebral artery infarction, particularly mutism in the acute stage, which is sometimes followed by a mild mixed transcortical aphasia (10). When the anterior corpus callosum is involved, the following disconnection syndromes may be present: left arm apraxia, poor naming of objects in the left hand, and aphasic writing with the left hand.

When a large infarct involves most of the anterior cerebral artery territory because of an occlusive A1 lesion, there may be abulia and a gaze preference. The caudate and anterior internal capsule are supplied by the anterior cerebral artery through penetrating arteries that arise from the A1 segment. Most infarcts of the caudate or anterior capsule are caused by intrinsic disease of these penetrating arteries and not by intrinsic disease of the anterior cerebral artery stem (17).

Atherosclerosis usually involves the intracranial vertebral artery at the level of the posterior inferior cerebellar artery. The clinical syndromes produced by intracranial vertebral occlusive disease depend on the location of the obstruction, whether the vertebral artery is the source of distal embolism, and whether 1 or both vertebral arteries are involved. Atherothrombosis of the vertebral artery at the origin of posterior inferior cerebellar artery causes lateral medullary infarction (Wallenberg syndrome) that is often associated with infarction of the inferior cerebellar hemisphere (38). If the thrombus extends from 1 vertebral artery to the proximal basilar artery, or if both vertebrals are occluded, the resultant clinical syndrome may resemble that of intrinsic basilar occlusion. If the vertebral artery is a source of distal embolism, the embolus most commonly lodges at the top of the basilar artery or in the posterior cerebral arteries (19).

The clinical syndromes associated with basilar artery occlusive disease depend on the location of the vascular lesion, whether unilateral or bilateral brainstem infarction has occurred, and whether distal embolism has occurred. Transient ischemic attacks commonly precede stroke in patients with basilar occlusive disease and typically consist of diplopia, dizziness, dysarthria, perioral numbness, paraplegia, or alternating hemiplegia (05; 71; 63). In patients with a proximal basilar occlusion whose distal basilar artery and superior cerebellar arteries are patent (ie, supplied by retrograde flow from the anterior circulation through the posterior communicating arteries), the infarct is usually limited to the midline and paramedian structures in the pons (18). The pontine tegmentum and cerebellum are usually spared because these structures are supplied by circumferential branches arising from the segment of the basilar artery that is patent and from the posterior inferior cerebellar arteries. Consequently, patients with proximal basilar occlusion usually have combinations of the following signs: quadriparesis (bilateral infarction), hemiparesis (unilateral infarction), pseudobulbar palsy, abnormalities of eye movements (unilateral gaze palsy, internuclear ophthalmoplegia, the one-and-a-half syndrome, skew deviation, ocular bobbing), pupillary abnormalities (bilateral small pupils, Horner syndrome), and reduced level of consciousness, with sparing of sensory and cerebellar function. Occipital headache occurs frequently as well (18).

In some patients with bilateral infarction of the basis pontis from proximal basilar occlusion, severe weakness of the limbs and horizontal eye movements permits only eye blinking or vertical gaze (ie, the "locked-in" syndrome). Nonstenotic atherosclerotic disease of the basilar artery may occasionally cause bilateral basis pontis infarcts by occluding the orifices of paramedian penetrating branches of the basilar artery (36). The deficit in these patients may be similar to that of patients with proximal basilar artery occlusion.

When the distal basilar artery is occluded, infarction may involve the midbrain, thalamus, medial temporal lobes, and occipital lobes. Caplan has described in detail the signs associated with "top of the basilar syndrome" (13). Midbrain involvement causes pupillary abnormalities (decreased reactivity, eccentric shape, altered size), oculomotor abnormalities (vertical gaze palsies, skew deviation, third nerve palsy), and behavioral abnormalities (peduncular hallucinosis). Thalamic involvement causes decreased alertness, amnesia, and sensory abnormalities depending on the size and location of the infarct. The signs associated with temporal lobe and occipital lesions are described below.

Atherosclerosis of the posterior cerebral artery most often affects the proximal (perimesencephalic) segment near the origins of the thalamogeniculate branches (64). Therefore, the ventroposterolateral thalamus, medial temporal lobes, and occipital lobes are at risk of infarction from occlusive disease in this segment of the posterior cerebral artery. Hemianopia or hemisensory transient ischemic attacks are the most common presentation (64). Signs in patients with infarction in the territory supplied by the posterior cerebral artery include hemisensory loss (infarction of ventroposterolateral), memory disturbance and naming difficulties (medial temporal infarction), hemianopia, hemiachromatopsia, alexia without agraphia, and visual agnosias (occipital infarction) (65).

Headache associated with posterior cerebral artery infarction is typically felt around the ipsilateral orbit or forehead (18). Paralysis occurs only when the orifices of the perforating branches to the cerebral peduncle are involved by an atherosclerotic lesion in the proximal part of the posterior cerebral artery. Involvement of the medial thalamoperforating branches of the posterior cerebral artery results in behavioral abnormalities and memory loss (11). Bilateral posterior cerebral artery infarction is associated with cortical blindness, memory loss (due to bilateral medial temporal involvement), and agitation.

Clinical vignette

A 55-year-old African-American woman with a history of hypertension, hyperlipidemia, ischemic heart disease, and smoking presented with 2 episodes of transient quadriparesis and dysarthria. These events lasted about 5 minutes each. She had a prior history of carotid endarterectomy also. MRI of the brain was negative for stroke, but MRA showed signal drop out in the proximal third of the basilar artery. Transcranial Doppler also showed elevated basilar artery flow velocity. The patient was placed on aspirin but had an additional transient ischemic attack with vertigo and dysarthria lasting 4 minutes. She was then switched to warfarin and has been stable for 6 months. She was then placed on aspirin monotherapy with no further transient ischemic attacks. In addition to antithrombotic therapy, the patient’s initial LDL was 127 mg/dl, and she was placed on atorvastatin 80 mg per day with the aim of achieving an LDL of less than 100 mg/dl and preferably less than 70 mg/dl. A target blood pressure of less than 140/90 mm Hg is also being sought.

Biological basis

Etiology and pathogenesis

This topic discusses stroke caused by atherosclerotic occlusive disease of the major intracranial arteries. Other less common causes of intrinsic intracranial occlusive disease (eg, dissection, moyamoya disease) are discussed in other topics.

The pathogenesis and pathophysiology of intracranial atherosclerosis is similar to that of atherosclerosis affecting other major arteries, such as the extracranial carotid arteries, the coronary arteries, and the peripheral arteries. A discussion of the pathogenesis and pathophysiology of atherosclerosis is beyond the scope of this review. The reader is referred to a review article by Ross for further information on this topic (67).

There are 2 major mechanisms of ischemic stroke in patients with atherosclerotic intracranial occlusive disease. The most common is artery-to-artery embolism, which occurs when a thrombus forms at the site of the stenotic or occluded intracranial artery. The thrombus subsequently embolizes downstream to occlude a more distal intracranial artery. The other possible mechanism of ischemia in patients with atherosclerotic intracranial occlusive disease is hypoperfusion of the most distal part of the brain that is supplied by the diseased artery. Predisposing factors for this mechanism of ischemia are poor collateral circulation and hypotension.

Epidemiology

Atherosclerotic narrowing of the large intracranial arteries is an important cause of ischemic stroke in the United States and worldwide (42). In a prospective study of 405 patients evaluated for stroke, intracranial large artery disease was the cause of 8% of ischemic strokes (69).

Race has an important influence on the likelihood of developing intracranial large artery disease. In postmortem studies where most patients were white, it has been shown that atherosclerosis usually involves different arterial trees at different ages. The aorta is involved first followed by the peripheral arteries, coronary arteries, extracranial carotid and vertebral arteries, and finally the intracranial arteries (37). In contrast to whites, blacks with symptomatic anterior circulation ischemia are more likely to have middle cerebral artery or carotid siphon disease than extracranial carotid disease (44; 40). In a study of patients with anterior circulation ischemia, the rates of middle cerebral artery occlusion were 90% in black males and 12% in white males (44).

Asymptomatic atherosclerotic lesions also tend to occur intracranially in blacks and extracranially in whites (41). Generally, whites are more prone to large vessel and embolic disease than blacks (04). However, blacks have a higher frequency of intracranial large artery occlusive disease in the posterior circulation. In a study of 27 white and 24 black patients, distal basilar artery and intracranial branch lesions were significantly more frequent in blacks, whereas severe proximal vertebral artery disease was 5 times more common in whites (39).

Studies of Japanese and Chinese patients have shown that these racial groups are also more likely to develop intracranial large artery disease than extracranial disease (78). In another study comparing Chinese living in the United States with American whites, the Chinese patients had significantly higher rates of intracranial carotid artery and middle cerebral artery stenosis (32). A multicenter study from China found evidence of intracranial stenosis in 46.6% of patients hospitalized with stroke (74). Patients with intracranial stenosis had more severe strokes and a longer length of stay in the hospital.

The explanation for the variance in the distribution of cerebral atherosclerosis in different races is uncertain. Although genetic susceptibility to intracranial large artery disease in blacks and Asians may be important, the influence of lifestyle and risk-factor profiles on the distribution of atherosclerosis may also be significant. Support for the latter theory is provided by the results of studies showing a lower frequency of intracranial large artery disease in black Africans than in black Americans (77) and a strong correlation between the high rate of intracranial large artery disease in Hispanics and the presence of diabetes (69). Most studies, however, that have controlled for the presence of traditional vascular risk factors have found that race is independently associated with site of atherosclerosis in the cerebrovascular circulation (41; 32).

The influence of sex on the risk of developing intracranial large artery disease has not been systematically studied. Data from a few small series suggest that women may be more likely to develop intracranial large artery disease than men, who tend to develop extracranial large artery disease (15).

Prevention

In general, atherosclerosis of the major intracranial arteries and the extracranial arteries share traditional risk factors such as hypertension, diabetes, hypercholesterolemia, and smoking. Of these, diabetes and hypertension appear to be the most important risk factors for intracranial large artery disease (39; 41). Homocysteine has not been well studied in this type of cerebrovascular disease.

Differential diagnosis

The clinical syndromes associated with atherosclerotic intracranial large artery occlusive disease are not specific. Other vascular pathologies that need to be considered in the differential diagnosis are cardioembolism, extracranial atherosclerotic carotid or vertebral occlusive disease, other extracranial vasculopathies (eg, dissection, fibromuscular dysplasia), nonatherosclerotic intracranial large artery vasculopathies (eg, moyamoya disease, dissection, vasospasm, radiation arteriopathy, infectious or noninfectious vasculitis), intracranial penetrating artery (ie, small vessel) disease, and procoagulant states (eg, antiphospholipid antibody syndrome, oral contraception use, cancer, abnormalities of the circulating anticoagulant system or the fibrinolytic system).

Diagnostic workup

Brain and vascular imaging studies coupled with data obtained from the history and physical examination usually enable the clinician to establish the diagnosis of atherosclerotic intracranial large artery occlusive disease. The initial imaging study in patients presenting with one of the clinical stroke syndromes described above is usually a CT of the brain. This enables differentiation between ischemic and hemorrhagic stroke and, in the case of ischemic stroke, provides data on the size and location of the infarct. MRI of the brain is more sensitive than CT at detecting cerebral infarction and is commonly ordered when CT does not reveal the suspected stroke.

In patients presenting with stereotypical transient ischemic attacks, a large (greater than 2 cm) subcortical infarct, or a cortical infarct in the territory of a single intracranial artery, atherosclerotic intracranial large artery disease should be considered in the differential diagnosis. When the transient ischemic attacks or infarct are in the territory supplied by the middle cerebral artery or anterior cerebral artery, carotid ultrasound should be performed to exclude extracranial carotid occlusive disease. Patients with multiple patterns of ischemic lesions on MRI, including concomitant perforating artery infarcts, pial infarcts, and borderzone infarcts, are more likely to have combined extracranial and intracranial stenoses (53). Echocardiography should also be considered to exclude a cardiac source of embolus, particularly if brain imaging shows multiple cerebral infarcts in different vascular territories. If an extracranial carotid or cardiac source for the patient's transient ischemic attacks or infarct is excluded, vascular imaging studies of the intracranial vessels should be considered to determine whether the patient has large artery intracranial occlusive disease.

The gold standard for establishing the diagnosis of intracranial large artery disease is conventional cerebral angiography. Angiography enables differentiation between atherosclerotic large artery occlusive disease and other intracranial vasculopathies such as moyamoya disease, dissection, and vasculitis. Additionally, angiography enables accurate measurement of the degree of stenosis of the diseased artery, diagnosis of arterial occlusion, evaluation of collateral flow patterns, and evaluation of other intracranial and extracranial arteries. The major drawback of angiography is the risk of stroke, which is 0.5-1% in patients investigated for cerebrovascular disease.

The development of transcranial Doppler in 1982 by Aaslid and associates (01) and the development of MRA have enabled noninvasive diagnosis of intracranial large artery occlusive disease (52; 56; 66; 68; 28). Transcranial Doppler and angiographic correlative studies show that transcranial Doppler has a sensitivity of 85% to 90% and a specificity of 96% to 98% for detecting stenosis (greater than 30%) or occlusion of the carotid siphon or middle cerebral artery stem when performed by experienced ultrasonographers (51). Transcranial Doppler is less reliable for detecting occlusive lesions in the vertebral or basilar arteries (sensitivity 76%, specificity 99%), especially in the distal basilar region (68). Further prospective evaluations with transcranial Doppler have identified other reliable, specific indicators of stenosis in stroke patients (28).

MRA is also used commonly to evaluate patients with suspected intracranial occlusive disease. One study, using 3D time-of-flight MRA versus 3D phase contrast MRA, looked at 18 patients with stenosis and was able to diagnose stenosis with at least 90% sensitivity and correctly grade stenosis with 80% accuracy using 3D time-of-flight MRA (61). In the same group of patients, the sensitivity of transcranial Doppler was 76% and the specificity was 99%. Although transcranial Doppler had a lower sensitivity than MRA, transcranial Doppler provided more complete hemodynamic data (collateral supply, estimations of degree of stenosis) than MRA (68). With the increased sophistication of MRA and transcranial Doppler, the combination of the 2 may overcome individual shortcomings--MRA’s overexaggeration of stenosis and transcranial Doppler’s insonation window limitations—obviating the need for digital subtraction angiography (09). With the development of high-strength magnets such as 3T MRI, direct thrombus imaging and visualization of intraplaque hemorrhage in symptomatic intracranial stenoses has been demonstrated (06).

A multicenter study has been conducted to assess transcranial Doppler and MRA in comparison to conventional angiography for the identification of 50% to 99% stenosis (33). This study, the SONIA trial, found relatively low positive predictive values (PPV) for these 2 modalities but higher negative predictive values (NPV). The PPV and NPV for transcranial Doppler were 36% and 86%, and for MRA, the figures were 59% and 91%. Further information is also needed on the reliability of CT angiography in the grading of intracranial stenoses. In general, if identification of intracranial atherosclerosis will affect management, it is best to perform conventional angiography to confirm the results of a noninvasive test.

Intracranial vessel wall imaging using MRI can be useful in some circumstances. It can potentially help with differentiation between atherosclerotic lesions and inflammatory vasculopathies (54). Advantages of vessel wall imaging include being able to visualize the vessel lumen and identify plaque characteristics. Additional imaging features that are being examined as part of research studies are perfusion status distal to the stenosis, the effect of collateral vessels, and fractional flow beyond the area of stenosis (50).

Management

Therapeutic options for managing patients with acute stroke caused by intracranial occlusive disease are beyond the scope of this section and are reviewed in the acute stroke treatment topics.

The main options for antiplatelet therapy in patients with intracranial atherosclerosis are aspirin, clopidogrel, or dual antiplatelet therapy (either aspirin plus clopidogrel or aspirin plus dipyridamole).

Aspirin was used preferentially in the ECIC bypass study. In the past, warfarin was frequently used for the treatment of intracranial large artery disease based on the results of nonrandomized studies of patients with symptoms suggestive of vertebrobasilar disease (59). However, these were small, frequently retrospective studies.

The Warfarin Aspirin Symptomatic Intracranial Disease study mentioned above is a prospective, randomized study in patients with recent transient ischemic attack or stroke and angiographically verified 50% to 99% stenosis (25). Patients were randomly assigned to treatment with warfarin (INR 2-3) or aspirin 1300 mg per day between 1999 and 2003. The study was originally planning to enroll 806 patients, but it was halted prematurely due to an increased rate of death and bleeding events in the warfarin arm. At the time of its termination, the Warfarin Aspirin Symptomatic Intracranial Disease study had enrolled 569 patients. The mean follow-up was 1.8 years. The primary outcome was stroke or vascular death, and this outcome was reached in 22% of aspirin treated patients and 22% of warfarin patients (log rank test, p value 0.82). The rates of systemic bleeding, cardiac ischemic events, and death were higher in the warfarin group. Based on these prospective data, antiplatelet therapy is the preferred antithrombotic treatment for patients with intracranial atherosclerosis.

Modification of stroke risk factors is important in patients with symptomatic intracranial atherosclerosis. Some clinicians have advocated a “relaxed blood pressure” policy for patients with intracranial atherosclerosis, but in WASID, systolic blood pressure > 140 mm Hg was associated with a higher risk of overall stroke and stroke in the territory of the stenotic vessel (73). In data from the WASID study, it was also found that patients with total cholesterol >200 mg/dl had a higher rate of future stroke compared to patients with better risk factor control (21). The importance of regular physical activity has also been demonstrated for patients with intracranial atherosclerosis. In a clinical trial, it was noted that greater physical activity was associated with a 40% decrease in stroke, myocardial infarction, or vascular death (72).

The initial enthusiasm for bypass surgery in patients with intracranial large artery disease has waned since the extracranial-intracranial bypass study (31). In that study, patients with extracranial carotid occlusion, distal carotid occlusive disease, or middle cerebral artery occlusive disease were randomized to medical therapy alone (risk-factor management and antithrombotic therapy, usually aspirin 325 mg 4 times daily) versus medical therapy and extracranial-intracranial bypass. The results demonstrated that extracranial-intracranial bypass was ineffective for preventing stroke in these patients. Subgroup analyses of the stroke rates in patients with distal carotid or middle cerebral artery disease showed that extracranial-intracranial bypass was also ineffective in these groups. In fact, patients with severe middle artery stenosis who underwent extracranial-intracranial bypass had a higher rate of stroke than the medically treated patients with severe middle cerebral artery stenosis (31). Bypass surgery or endarterectomy has also been used for intracranial vertebral artery occlusive disease (70; 08), but the efficacy of these procedures has not been systematically evaluated. Bypass surgery in the posterior circulation usually involves anastomosing the occipital branch of the external carotid artery to the posterior inferior cerebellar artery or the anterior inferior cerebellar artery. These procedures are rarely used today for the treatment of intracranial vertebral disease.

Transluminal angioplasty is another therapeutic option for the treatment of intracranial large artery stenosis. A large retrospective study of 50 patients reported that 49 of 50 patients had an improved clinical or angiographic outcome (26). However, 14% of patients had a vessel dissection and 16% had residual stenosis of more than 50%. It must be established that angioplasty and stenting results are better than those that can be achieved with medical therapy (20).

A multicenter registry with the Wingspan stent has been reported as well (80). In this study, 129 patients with symptomatic stenosis of 70% to 99% were treated with intracranial stenting. The technical success rate was high at 97%, and the primary endpoint of any stroke or death within 30 days plus ipsilateral stroke between 30 days and 6 months was 14.0%. The restenosis rate was relatively high at 25%. A restenosis rate of approximately 30% was seen in another multicenter registry using the Wingspan stent, with restenosis being more common in the anterior circulation (34).

Another multicenter study from Germany reported results in 372 patients (49). In this analysis, the disabling stroke rate was 4.8%, and there was a 2.2% death rate. Hemorrhagic events were more frequently seen after middle cerebral artery stenting, and compromise of perforating vessels was more common in the posterior circulation.

Restenosis can be a problem in small caliber vessels subjected to angioplasty. A small series of 8 patients who received drug-eluting stents found no cases of restenosis over an average 11-month period of follow-up (02). These results require larger studies to confirm the initially encouraging findings.

With the development of intracranial stenting, the Stenting and Aggressive Medical Management for Preventing Recurrent stroke in Intracranial Stenosis (SAMMPRIS) trial was launched (24). Patients who had stroke or transient ischemic attack in the previous 30 days and symptomatic stenosis of 70% to 99% as confirmed by angiography were enrolled and randomly assigned to either aggressive medical therapy (AMT) alone or AMT plus stenting. AMT included a target low-density lipoprotein (LDL) of less than 70 mg/dl and treatment of systolic blood pressure to achieve a target of less than 140 mm Hg in nondiabetics and less than 130 mm Hg in diabetics. Patients in both groups received aspirin and clopidogrel for 90 days followed by aspirin monotherapy.

The original study hypothesis was that AMT plus stenting would be superior to AMT alone. Four hundred and fifty-one subjects were recruited from 50 sites in the United States. The study was stopped prematurely, however, due to the hazards of stenting. The 30-day stroke/death rate was 14.7% in the stenting arm and 5.8% in the AMT group (p=0.002). Beyond 30 days, the stroke rate in the territory of the stenosed vessel was comparable in the 2 groups. The rate of the primary endpoint (stroke or death within 30 days or stroke in the territory of the qualifying artery) at 1 year was 20.0% in the stenting group and 12.2% in the AMT group (p=0.009).

Three-year results of SAMMPRIS have confirmed that the benefits of AMT were persistent (29). During a median follow-up period of 32.4 months, the rate of the primary endpoint (including stroke or death within 30 days of enrollment and stroke in the territory of the stenotic vessel beyond 30 days) was 15% in the AMT alone and 23% in the AMT + stenting group (p=0.025). For individual adverse events, each stroke was higher in the AMT + stenting group (26% vs. 19%, p=0.047), and major hemorrhage was higher in the AMT + stenting group (13% vs. 4%, p=0.0009). An additional analysis comparing WASID and SAMMPRIS patients treated medically found that the SAMMPRIS AMT regimen reduced major vascular events by about 42% (21.9% vs. 12.6%, hazard ratio 1.9) (22). Therefore, AMT alone is preferred for patients with severe, symptomatic intracranial disease.

With regard to patients in SAMMPRIS who received stenting, a detailed analysis of the strokes related to the endovascular procedure has been published (30). The 3 most common mechanisms of stroke were perforating artery occlusion, subarachnoid hemorrhage (due to presumed wire perforation), and delayed cerebral hemorrhage.

A second multicenter interventional trial compared a balloon-expandable intracranial stent versus medical therapy in patients with symptomatic, 70% to 99% intracranial stenosis (VISSIT trial) (79). The primary safety measure was a composite of stroke, death, or intracranial hemorrhage within 30 days of study entry or definite transient ischemic attack. The study was halted early due to futility. After 112 patients were enrolled, the primary safety endpoint occurred in 24.1% of the stent group, compared to 9.4% in the medical arm (p=0.05). Intracranial hemorrhage results and 12-month outcomes were also worse in the intracranial stenting group. Along with SAMMPRIS, these results show that aggressive medical therapy is the preferred treatment strategy for patients with symptomatic intracranial atherosclerosis.

After the results of SAMMPRIS, a refined strategy of stenting has been proposed. This involves more careful patient selection of submaximal angioplasty and avoiding procedures within 7 days of a stroke. A multicenter registry of 152 patients reported a periprocedural stroke, bleeding, and death rate of 2.6% (03). However, the study recruited very select patients and it needs replication in a larger study to assure clinicians that improved results with intracranial stenting are possible.

A novel treatment strategy that was evaluated in a small study involves the concept of ischemic preconditioning (57). In this study of 68 patients with symptomatic intracranial disease, patients in the intervention group received bilateral arm ischemia and reperfusion with a novel device for 300 days. After 300 days, the stroke rate was 27% in the control group and 8% in the intervention group (p< 0.01). This treatment strategy appears worthy of a larger study in the future.

Outcomes

The most recent data concerning prognosis for patients with intracranial atherosclerosis come from the Warfarin Aspirin Symptomatic Intracranial Disease (WASID) study (25). In this trial, patients were enrolled with 50% to 99% stenosis of the distal internal carotid artery, the middle cerebral artery, the intracranial vertebral artery, or the basilar artery. During a mean follow-up period of 1.8 years, the rate of ischemic stroke was 18.7% (combining the aspirin and warfarin arms). Patients with more severe stenosis (70% to 99% compared to 50% to 69%) and patients with a stroke as opposed to a transient ischemic attack are at higher risk for future stroke (46). In WASID, patients with a stroke and 70% to 99% stenosis had a 23% risk of stroke at 1 year, compared to a 3% one-year risk of stroke in patients with a transient ischemic attack and 50% to 69% narrowing.

Data are also limited on the outcome of patients with posterior cerebral artery stenosis. Pessin and associates studied 6 patients with symptomatic posterior cerebral artery stenosis of 50% to 80% (64). Five patients had unilateral disease and 1 had bilateral disease. Four patients presented with hemianopic or hemisensory transient ischemic attacks and 2 patients presented with stroke (1 had a hemianopia; 1 had a superior quadrantanopia). All patients were treated with warfarin and the range of follow-up was 4 months to 4 years. None of the patients had a stroke in the territory of the stenotic posterior cerebral artery, but 3 patients died (1 from a traumatic intracerebral hemorrhage, 1 from a middle cerebral artery infarct, 1 sudden death).