Predictors of recurrent stroke in patients with symptomatic intracranial arterial stenosis.

BACKGROUND AND PURPOSE Our goal was to investigate whether initial ischemic lesion pattern can predict stroke recurrence in patients with symptomatic intracranial arterial stenosis. METHODS Of the Trial of Cilostazol in Symptomatic Intracranial Arterial Stenosis (TOSS)-2 trial participants, we included patients who underwent diffusion-weighted imaging and fluid attenuation inversion recovery imaging at baseline with a follow-up fluid attenuation inversion recovery imaging at 7 months. Based on the diffusion-weighted imaging findings, we classified the initial ischemic lesion patterns according to location (subcortical versus cortical versus subcorticocortical) and multiplicity (single versus multiple). We also evaluated the occurrence of new ischemic lesions on follow-up fluid attenuation inversion recovery as well as clinical stroke in the symptomatic intracranial arterial stenosis territory. RESULTS Of 353 patients included in this study, 44 (12.5%) and 13 (3.7%) patients had new ischemic lesions and clinical recurrent stroke in the initial symptomatic intracranial arterial stenosis territory, respectively. On multivariable analysis, the initial lesion patterns of subcorticocortical and multiple lesions were independent predictors of new ischemic lesions in the symptomatic intracranial arterial stenosis territory (OR, 3.01; 95% CI, 1.33-7.01; P=0.03; OR, 2.81; 95% CI, 1.34-5.9; P=0.006). These patterns also predicted clinical recurrent stroke. CONCLUSIONS Subcorticocortical lesions and multiple lesions are radiological predictors of recurrent ischemic stroke in symptomatic patients with intracranial arterial stenosis. Clinical Trial Registration- URL: www.clinicaltrials.gov. Unique identifier: NCT00130039.