Background and purpose Intracranial stenting and angioplasty has been increasingly utilized for the treatment of symptomatic intracranial atherosclerotic disease (ICAD). The use of self-expanding stents (SES) has been reported to have a high technical success rate but there are concerns with restenosis. We sought to determine if the use of balloon mounted stents (BMS) was associated with higher technical complication rates compared with self-expanding stents.

Methods Three academic institutions contributed data on patients treated for symptomatic ICAD over a 3 year period (2006–2009). 259 consecutive patients were identified. Demographics, type of stent placed, periprocedural complications, per cent stenosis prior to treatment and after treatment, if the treatment was within 2 weeks of clinical symptomatology, lesion classification with the Mori scale and follow-up imaging for restenosis were collected retrospectively. A univariate analysis was performed to determine predictors of periprocedural stroke and also for restenosis at follow-up. Variables with a p value <0.20 were entered into a binary logistic regression model to indentify independent predictors.

Results 259 patients with a mean age of 61±13 years with 168 (65%) patients presenting with strokes and the rest with transient ischemic attacks. The distribution of the lesions were as follows: vertebrobasilar 109 (42%), middle cerebral artery 85 (33%) and intracranial internal carotid artery 65 (25%). 79 (31%) BMS were implanted and 180 (69%) SES. A total of 17 (6.5%) patients developed a periprocedural stroke with no difference noted between use of type of stent. Patients in whom a BMS was implanted had a significant reduction in stenosis compared with SES (81±10% to 7±11% vs 80±12% to 24±9%; p<0.0001). The independent predictors of periprocedural stroke were as follows: treatment within 2 weeks (OR 2.5 (1.2 to 5.6); p<0.01), Mori B or C lesion (OR 6.7 (3.3 to 10.1); p<0.001) and patients with an index event of a stroke (OR 1.5 (1.2 to 3.4); p<0.018). The independent predictors of restenosis at follow-up were as follows: use of SES (OR 3.22 (1.40 to 7.41); p<0.006) and Mori B or C lesion (OR 3.97 (2.05 to 7.71); p<0.0001).

Conclusions The use of BMS was not associated with higher complication rates but was associated with lower restenosis rates. Placement of stents in patients with more complex lesions is associated with higher periprocedural stroke rates and restenosis. Further prospective studies comparing BMS with SES are warranted.