Stroke recurrences in patients with symptomatic vs asymptomatic middle cerebral artery disease

Obstructive intracranial artery disease is the etiology of cerebral ischemic events in approximately 5% to 10% of white patients.^1–3 In Asian acute stroke patients, it is the most common vascular lesion.^4,5 Because of the wide availability of transcranial Doppler ultrasonography (TCD), transcranial color-coded duplex ultrasonography (TCCD), and magnetic resonance angiography (MRA), obstructive disease of the middle cerebral artery (MCA) and of other major intracranial brain–supplying arteries is increasingly detected in stroke patients as well as in asymptomatic individuals.^6–8

In contrast to symptomatic and asymptomatic carotid artery disease, where the natural history has been illuminated in several large carotid endarterectomy trials and other prospective studies,^9–14 corresponding data on intracranial vessel disease are limited. Previous studies, including the prospective extracranial-intracranial arterial bypass study, suggested an annual stroke risk in patients with symptomatic MCA disease between 4% and 15%.^15–17 In asymptomatic patients, however, the risk of cerebrovascular events is unknown because most studies excluded these patients.

In contrast to retrospective studies,^16,18,19 the final results of the Warfarin-Aspirin Symptomatic Intracranial Disease study (WASID) showed no advantage of anticoagulation to prevent recurrent ischemic events in patients with symptomatic intracranial arterial stenosis.²⁰ Additionally, intracranial angioplasty has been proposed for symptomatic patients who do not respond to medical treatment.^18,21 Considering the ongoing controversy on optimal therapeutic strategies for these patients, it becomes of increasing interest to determine their long-term risk of stroke recurrence.

We therefore investigated the natural history and prognosis of both symptomatic and asymptomatic MCA disease with regard to the occurrence of associated cerebral ischemia. We also sought to identify potential predictors for cerebrovascular events and to compare the efficacy of platelet inhibitors and anticoagulants for stroke prevention in these two patient groups.

Methods.

We recruited and consecutively followed symptomatic and asymptomatic patients with evidence of obstruction of one or both MCAs to assess their clinical and vascular course. Patients were included if they met the following criteria: 1) evidence of stenosis or occlusion of one or both main stems of the MCA or of one of its major branches at study entry proven by both TCD and TCCD; 2) sufficient transtemporal bone windows for consecutive bilateral insonation of the MCA; 3) attendance of at least one follow-up visit after study entry, resulting in a minimal observation period of 6 months; and 4) continuous treatment with either platelet inhibitors (acetylsalicylic acid, ticlopidine, or clopidogrel) or anticoagulants (intended international normalized ratio 2–3) according to the individual physician’s decision.

We excluded patients with other potential causes of stroke or TIA, such as ipsilateral high-grade carotid artery stenosis; cardiac sources of embolism; patients with uncertain diagnosis of MCA obstruction or inconsistent findings from TCD and TCCD; patients with intracranial stenosis due to vasculitis, vasospasm, dissection, aneurysm, basilar meningitis, or moyamoya disease; and patients who were unable or unwilling to attend the study.

One hundred ninety-two patients with sufficient temporal bone windows and obstruction of at least one of the major intracranial arteries were screened in our center within 1 year. Eighteen had an isolated obstruction in the anterior cerebral artery, 11 had an isolated obstruction in the posterior cerebral artery, 31 had an isolated obstruction in the basilar artery, and 12 had an isolated obstruction in the intracranial segment of the internal carotid artery, leaving 120 patients with stenosis or occlusion of one or both MCAs. Eleven patients were excluded: 3 patients with high-grade ipsilateral extracranial carotid artery stenosis, 4 patients with atrial fibrillation, 1 patient with a prosthetic aortic valve, 1 patient with known intracranial aneurysm, 1 patient with suspected intracranial vasculitis, and 1 patient with moyamoya disease. The remaining 109 patients were considered for inclusion, gave consent to attend neurologic and sonographic follow-up examinations, and were enrolled in the study. Four patients were lost during follow-up because they refused further participation. One patient independently stopped his treatment with platelet inhibitors. In two patients, findings of TCD and TCCD were inconsistent (MCA stenosis in TCD was not confirmed by TCCD in either case).

One hundred two patients aged 61.9 ± 13.6 years (67 men, mean age 60.7 ± 13.4 years; 35 women, mean age 64.1 ± 14 years) completed the study protocol and qualified for statistical data analysis. At study entry, patients were considered primarily symptomatic if a cerebrovascular event, including certainly diagnosed TIA or ischemic stroke, had occurred in the vascular territory of MCA disease. If there was no history of cerebrovascular events or if a TIA or a stroke occurred in a vascular territory outside the affected MCA, the patient was assigned to the asymptomatic group.

Patients were followed up for a mean period of 30.7 months (range 6 to 117 months). At the first patient visit at inclusion and at every visit during follow-up performed at regular 6- to 12-month intervals, medical history in terms of a structured interview and physical and neurologic examination, including the National Institutes of Health Stroke Scale, was obtained. The patients were routinely investigated for signs of cerebrovascular events, including TIA and stroke, risk factors, current medical treatment, side effects, and minor or major hemorrhagic complications. Risk factors included history of hypertension, diabetes, hypercholesterolemia, history of coronary heart disease, and cigarette smoking of at least 10 cigarettes per day during the past 5 years or more. All patients underwent a routine laboratory test and EKG during follow-up; a transthoracic echocardiogram was obtained in 58.9%.

Transcranial Doppler ultrasonography was performed using a 2-MHz pulse-wave probe (DWL Multi-Dop T2, DWL Elektronische Systeme GmbH, Germany) insonating the intracranial segments of the internal carotid artery, middle cerebral arteries, anterior cerebral arteries, and posterior cerebral arteries from a transtemporal approach. Intracranial parts of the vertebral artery and the basilar artery were investigated routinely by transforaminal insonation. TCCD was performed using a Philips HDI 5000 system (Philips Medical Ultrasound, The Netherlands) with a 2- to 4-MHz phased array transducer including frequency- and power-based color coding focusing on the arteries of the circle of Willis. Intracranial vessels were assessed by two experienced investigators who were unaware of each other’s findings using TCD and TCCD at study entry and at each follow-up visit. Details of transcranial ultrasound investigations and analyses are reported elsewhere.^22,23 The degree of stenosis was classified according to criteria based on the peak systolic flow velocity. MCA stenosis was graded as one of three categories: low-grade (140 to 209 cm/s), moderate (210 to 280 cm/s), and high-grade (>280 cm/s). MCA occlusion was diagnosed when the Doppler signal was lacking and other ipsilateral basal cerebral arteries were identified.²⁴

During follow-up, the diagnosis of obstructive MCA disease was confirmed by a complementary method for vascular imaging in 74% of the patients. Fifty-three percent of the patients were examined with MRA alone; 34% underwent conventional intra-arterial angiography alone; and in 13% of the patients, both MRA and conventional angiography were performed at least once during follow-up.

Middle cerebral artery disease progression or regression was recorded by TCD and TCCD data at every patient visit. Disease progression was documented if the MCA stenosis had upgraded in severity verified at two repeat visits or if stenosis progressed to occlusion. We defined disease regression as development from occlusion to stenosis, decrease of the degree of stenosis, or normalization of flow velocities confirmed at two repeat visits.

Statistical data analysis was performed with SPSS version 11.5. Univariate analysis was used to compare variables in patient groups using χ² tests for discrete variables and t tests for contiguous variables. Ipsilateral and overall annual incidences of cerebrovascular events during the observation period were calculated. To determine independent predictors for the occurrence of ipsilateral and overall cerebrovascular events, a Cox proportional hazards model was applied that incorporated the duration of follow-up. The hazard ratio (HR) of each variable included in the model was given with 95% CI. Survival analysis was performed to compare cumulative event-free rates for the time to a cerebrovascular event between symptomatic and asymptomatic patients at study entry and between patients treated with platelet inhibitors and patients treated with oral anticoagulants. Survival curves were calculated from the Cox proportional hazards model; the difference of two survival curves was estimated by the −2 log likelihood method. For all statistical tests, p < 0.05 was considered significant.

Results.

At study entry, 56 of patients (54.9%) were classified asymptomatic; 46 of patients (45.1%) entered the study after TIA (n = 17) or stroke (n = 29) in the vascular territory ipsilateral to MCA disease. Seventy-five patients (73.5%) were continuously treated with platelet inhibitors, and 27 (26.5%) were treated with anticoagulants (table 1).

Nineteen cerebrovascular events occurred during follow-up, resulting in an overall rate of 7.3% per year. Eight accidents were TIAs with complete regression of clinical symptoms within 24 hours or less, and 11 were classified as ischemic stroke. Thirteen events (8 strokes, 5 TIAs) occurred in the vascular territory ipsilateral to MCA disease (ipsilateral annual incidence 5.0%). Three patients had recurrent events: 1 patient had a stroke in the contralateral MCA territory followed by an ipsilateral event 4 months later, and 2 patients had recurrent ipsilateral TIAs. Patients with MCA occlusions exclusively had ipsilateral events (n = 3). Cerebrovascular events ipsilateral to MCA disease occurred at a mean of 16.3 months after study entry, and any cerebrovascular event occurred after 20.7 months. Statistical analysis of potential predictors for the occurrence of ipsilateral or overall cerebrovascular events is summarized in table 2 (univariate statistics) and table E-1 (available on the Neurology Web site at www.neurology.org) (Cox proportional hazards model).

Six patients died during the observation period (annual mortality 2.3%); death was classified as of vascular cause in one patient with a fatal stroke. One patient treated with anticoagulants had a major nonfatal intracerebral hemorrhage. Two other patients treated with anticoagulants experienced minor hemorrhagic complications (gingival bleeding and rectal bleeding from hemorrhoids). No hemorrhagic complications occurred during treatment with platelet inhibitors.

Comparison of symptomatic and asymptomatic MCA disease.

In the symptomatic group, patients were younger, patients had higher grades of MCA stenosis, and MCA occlusions were more frequent as compared with the asymptomatic group. Eighty-two percent of patients with MCA occlusion were symptomatic at study entry, compared with only 40.7% of patients with MCA stenosis. Regression of MCA obstruction was associated with symptomatic MCA disease (p = 0.04).

Symptomatic MCA disease was associated with occurrence of ipsilateral and overall cerebral ischemic events in the observation period (p < 0.01, χ² test). Of the patients with symptomatic MCA disease, 32.6% experienced events in any vascular territory, compared with only 7.1% of the patients with asymptomatic MCA. All cerebrovascular events in asymptomatic MCA disease occurred during therapy with platelet inhibitors. In the subgroup of 13 patients with ipsilateral events, 11 had already been symptomatic at study entry. Patients with symptomatic MCA disease at study entry had an annual risk of 12.5% (ipsilateral: 9.1%), whereas the incidence in asymptomatic patients was 2.8% per year (ipsilateral: 1.4%). In the Cox proportional hazards model, symptomatic MCA disease remained an independent risk factor for recurrent TIA or stroke (HR for ipsilateral events 9.66, 95% CI 1.5 to 62.25, p = 0.02; HR for any event 7.91, 95% CI 2.03 to 30.79, p < 0.01). Cumulative event-free rates estimated in the Cox proportional hazards model (figure, A) were different between symptomatic and asymptomatic MCA disease (p = 0.03).

• Download figure
• Open in new tab
• Download powerpoint

Figure. Survival curves estimated with the Cox proportional hazards model show the cumulative rate of patients remaining free of cerebrovascular events during the observation period. (A) Comparison of patients with asymptomatic MCA disease (upper curve) and with symptomatic MCA disease at study entry (lower curve; p = 0.02). (B) Comparison of patients treated with anticoagulants (upper curve) and of those treated with platelet inhibitors (lower curve; p = 0.02).

Discussion.

Despite the widespread availability of valid diagnostic methods, few data on the incidence, prognosis, and treatment of intracranial arterial obstructive disease are available. Our study provides information on the differences between symptomatic and asymptomatic MCA obstructions. It demonstrates that patients with symptomatic stenoses or occlusions are a high-risk group for recurrent cerebrovascular events. Our ipsilateral recurrence rate of 9.1% per year corresponds to previous long-term observations in white patients with symptomatic MCA disease.^{15–18,25–27} The cumulative recurrence rate of symptomatic MCA disease seems to be even higher compared with data of symptomatic extracranial carotid artery disease²⁸ because it continuously remains stable beyond the first year. In contrast, asymptomatic MCA disease represents a relatively benign condition with low annual incidences of TIA and stroke. The overall annual risk for TIA and stroke of 2.8% in asymptomatic patients is comparable to that of asymptomatic carotid artery disease.^12,13

Intracranial arterial stenoses have been described as dynamic lesions demonstrating both progression and regression.²⁷ In particular, progression of stenosis was identified as potential predictor of recurrent ischemic events.¹⁷ Among our population, disease progression was not associated with an increased risk of TIA and stroke. This lack of association may be due to the low incidences of TIA and stroke and the relatively small number of progressive stenoses in the current study. Another limitation of the study may be that the diagnosis of MCA obstruction was confirmed by a complementary method in only 74% and based on TCD and TCCD alone in a small minority of asymptomatic patients. Although conventional angiography is still considered the accepted standard for the detection and quantification of intracranial vessel obstruction, we relied on these noninvasive techniques because of their good sensitivity and specificity.^29,30

An increasing degree of arterial obstruction can be explained by progression of an atherosclerotic lesion, whereas regressive obstructions are contrary both to current concepts of atherosclerosis and to conditions known from carotid artery disease. Although a decreasing degree of MCA obstruction was not a predictor for recurrent stroke or TIA in our study, we found a remarkably high regression rate. Furthermore, there was a significant association between regressive obstructions and symptomatic MCA disease. Lower rates of regressive stenosis have also been reported in asymptomatic compared with symptomatic MCA disease.³¹ This observation may be due to different pathomechanisms accounting for obstructive intracranial disease: asymptomatic stenoses of atherosclerotic nature as stable or slowly progressive condition, and symptomatic stenoses or occlusions that may be either atherosclerotic or embolic obstructions with the potential for spontaneous regression or recanalization. This hypothesis is further supported by findings that regression in symptomatic MCA disease is associated with occurrence of microembolic signals detected during the acute phase of ischemic stroke,³² whereas microembolic signals were not detected in asymptomatic or symptomatic MCA disease during the chronic phase after stroke.³³

Because interventional techniques such as intracranial angioplasty and stenting develop fast as alternatives to medical treatment,^21,34–36 knowledge about the natural course of intracranial obstructive disease and about the impact of conservative therapy becomes of increasing interest. However, which medical treatment should be evaluated in a randomized trial comparing conservative therapy with interventional procedures is still controversial.³⁷ Patients with symptomatic intracranial atherosclerosis who did not respond to antithrombotic therapy were reported to be at high risk for cerebral ischemic events¹⁸ and may therefore be a target group for the medical arm of such a randomized angioplasty trial. Because of the high recurrence rate under anticoagulation in this study, one may conclude that this therapy might not be effective in intracranial artery disease. Also, in patients with previous noncardioembolic ischemic stroke, no advantage of anticoagulants for preventing stroke recurrence was demonstrated.³⁸ Correspondingly, the final results of the WASID trial did not show a benefit of anticoagulation over aspirin to prevent recurrent ischemic events in patients with symptomatic intracranial arterial stenosis, and there was also a high rate of adverse events under treatment with anticoagulants.²⁰

In contrast to the WASID trial, we observed low incidences of stroke and TIA in patients with MCA disease under treatment with anticoagulants, similar to previous observational studies.^16,39,40 Considering our results, one may conclude that, because of the benign natural course, platelet inhibitors are preferable in asymptomatic obstructions of the MCA, whereas anticoagulants may be beneficial in selected patients with symptomatic, high-grade MCA stenoses. The different study design, inclusion of patients with stenosis of any intracranial artery, the use of higher doses of aspirin, and the shorter observation time in the WASID trial may explain this discrepancy. However, their high stroke rate of more than 17% within 2 years corresponds well with our findings in symptomatic patients.

Acknowledgment

The authors thank Joachim Brade from the Department of Medical Biostatistics in Mannheim, Germany, for help with statistics.