Long-term outcome of endovascular stenting for symptomatic basilar artery stenosis

Basilar artery (BA) stenosis portends a high risk of TIAs, stroke, or death.^1–3 In a prospective study of posterior circulation stroke, 30% of patients with ≥50% BA stenosis had mortality or severe disability at 30 days after hospital discharge.² In other retrospective studies, 1-year stroke-free survival was 67% and the risk of recurrent stroke was approximately 11% per year in the stenotic BA territory despite medical therapy.^4,5

Endovascular angioplasty of the cerebral vasculature was introduced as a promising therapy for high-grade BA stenosis in 1980.⁶ However, case series at selected centers revealed periprocedural morbidities of 7 to 50%, with mortalities as high as 16.7%.⁷ The common complications of balloon angioplasty include elastic recoil, acute closure, plaque embolization, arterial dissection or perforation, and occlusion of the perforating branches during balloon inflation. Some of these limitations may be overcome with use of a compact and flexible stent.⁸ Although a number of case series have described the feasibility and short-term results of stenting for symptomatic BA stenosis,^8,9 the long-term outcome of such therapy has not been previously reported.

Methods.

We have searched the databases of the Neurointerventional Radiology and Neurovascular Service at our institute to identify patients who underwent endovascular stenting for symptomatic BA stenosis from 1999 to August 2003. Patients who were treated only with angioplasty or those who presented with BA occlusion, artery dissection, or angiitis were excluded. In all eligible cases, patient demographics and clinical characteristics were reviewed and abstracted by an experienced stroke neurologist. The stenosis rate, (1 − N/D) × 100%, was estimated from measurements of the diameters at the narrowest point (N) and normal artery distal to the lesion (D) on digital subtraction angiography.

All patients presented with either TIAs or ischemic stroke referable to a high-grade BA stenosis. They were evaluated by stroke neurologists and pretreated with aspirin 325 mg and clopidogrel 75 mg daily. Stenting was performed under general anesthesia by one of the neurointerventional radiologists. Percutaneous access was obtained using a 6- or 7-French catheter sheath via one of the femoral arteries. Complete diagnostic angiography was performed to identify areas of stenosis and collateral circulation. A guiding catheter was then coaxially placed in the cervical vertebral artery (VA) providing the best access to the BA stenosis. The stenoses were then primarily crossed with a 0.035-in. microwire, and a premounted balloon expandable coronary stent was delivered across the lesion and deployed, with or without predilatation. IV heparin was administered during the procedure to maintain an activated clotting time between 2 and 2.5 times normal for adequate systemic anticoagulation.

After stenting, all patients were admitted to the Neurologic Intensive Care Unit for observation and management under the care of the Stroke and Neurocritical Care Service. The Modified Rankin Scale (MRS) was used to evaluate functional outcome. All patients had either clinic or telephone follow-up.

Results.

Eighteen patients were identified to have undergone endovascular stenting for symptomatic BA stenosis. The clinical and radiologic features of these patients are summarized in table 1. Sixteen patients (89%) had TIAs or stroke despite medical therapy with antiplatelet agents, anticoagulants, or both. All the patients had more than 60% BA stenosis. There were eight focal (<5 mm in length), five tubular (5 to 10 mm), and five segmental (>10 mm) lesions. Six patients had no collateral circulation, and three of them (50%) presented with TIAs. Twelve patients had one or two posterior communicating arteries. Patients with poor collateral circulation appeared to be more likely to have TIAs. Endovascular stenting was performed successfully in all patients with excellent flow restoration (figure). The residual stenosis ranged from 0 to 30%. Four patients had additional stenting for concomitant VA and ICA lesions. Two patients had balloon angioplasty for tandem VA stenoses.

• Download figure
• Open in new tab
• Download powerpoint

Figure. Cerebral angiograms show focal (A), tubular, preocclusive (B), and diffuse segmental (C) lesions before and after stent placement. A tandem distal left vertebral artery stenosis was also stented (C). All the lesions were revascularized with excellent angiographic results.

Major periprocedural complications occurred in three patients. One patient developed severe dysarthria and right-sided hemiparesis approximately 24 hours after stenting of severe tandem stenoses in BA, distal left vertebral artery (VA), and right petrous internal carotid artery (ICA). CT scan of the head showed a left pontine infarct and multifocal intracerebral hemorrhages. Another patient had oculomotor nerve palsy and depressed level of consciousness after stenting for tandem BA and left VA stenoses. He was found to have subarachnoid and intraventricular hemorrhages, which required an external ventricular drainage for 10 days. The third patient experienced profuse bleeding from a femoral artery, the oropharynx, and the urethra that required transfusion with two units of packed red blood cells. The overall rate of major periprocedural complications was 16.7% without fatality, and the rate of cerebrovascular complications was 11.1%.

Minor complications were noted in three patients with one cerebral microembolism and two asymptomatic VA dissections. The patients with VA dissection were treated with an oral anticoagulant for 6 months without neurologic sequelae.

The long-term follow-up results are summarized in table 2. Fifteen patients (83.3%) had excellent long-term functional outcome (MRS score ≤1) without disability. Five of them reported mild transient symptoms, including dizzy spells, sensation of head congestion, neck pain, and hand incoordination. Only one (5.6%) patient had a recurrent pontine infarct due to in-stent restenosis 9 months after the stenting. The lesion was treated with angioplasty without further events. Neither of the two deaths (11.1%) was related to the stenting or stroke.

Discussion.

Endovascular stenting of symptomatic BA stenosis remains controversial due to the high risk of complications and lack of proven efficacy. Although emergent revascularization in unstable patients was associated with very high periprocedural morbility and mortality,⁷ elective stenting of BA stenosis appeared to be safe.⁹ In this case series, 18 patients had stenting for symptomatic BA stenosis. Four had additional stent placement for concomitant VA and ICA stenoses. Of the three patients with major periprocedural complications, two initially presented with acute stroke and were stented for BA lesions and additional tandem stenosis. Therefore, consistent with the results of a recent report,⁷ patients with acute stroke and tandem stenoses tended to have a higher risk of periprocedural complications and poorer outcome.

There have been limited data on the long-term outcome of stenting for symptomatic BA stenosis. Over the past 5 years, only two case series reported the outcome of stenting for up to 15 months.^9,10 One of the two case series described elective stenting of symptomatic BA stenosis in 12 patients.⁹ Clinical follow-up was available for 0.5 to 16 months. During a mean 5.9-month follow-up after stenting, 10 patients (83%) were asymptomatic. One patient had a TIA and the other had nonspecific symptoms. In the other case series, 10 patients with BA stenting were followed for 2 to 42 months.¹⁰ At a mean 15.6-month follow-up, six (60%) of the patients were asymptomatic. Two patients (20%) had TIA or stroke. Of the other two patients, one died of metastatic lung carcinoma and the other died of coronary artery bypass. We demonstrate that, at a mean 26.7-month follow-up, 83.3% of patients had excellent long-term outcome. There were only one recurrent stroke (5.6%) and two deaths of medical illness. Therefore, endovascular stenting for symptomatic BA stenosis appeared to be effective in preventing stroke and reducing death.

Of note, there are a few limitations in this case series. First, it is a retrospective study. The patients were stented empirically without standard inclusion and exclusion criteria. Possible selection bias is one of the major limitations. Second, not every patient received maximal medical therapy before stenting. Last, these are single-center data and may not be generalizable for reasons of referral and selection bias, neurointerventional physicians' expertise, and multidisciplinary care. It should also be pointed out that although all stented patients were treated with aspirin and clopidogrel based on percutaneous coronary intervention data, we advise against using the combination therapy for stroke prevention in general patient population.

Because of the poor prognosis of symptomatic BA stenosis found in previous studies, prospective multicenter randomized trials of endovascular BA stenting are warranted despite the risk of major procedural complications.