The Utility of Collaterals as a Biomarker in Pediatric Unilateral Intracranial Arteriopathy PEDIATRIC NEUROLOGY Elbers, J., Armstrong, D., Benseler, S. M., Dlamini, N., Steinberg, G. K., Yeom, K. W. 2018; 78: 27–34


Intracranial arteriopathies are frequent causes of pediatric stroke and important risk factors for stroke recurrence. Without tissue diagnosis, vascular imaging is relied upon to identify the underlying etiology and prognosis. We hypothesized that children with unilateral intracranial arteriopathy with lenticulostriate collaterals would demonstrate distinct vascular outcomes compared with children without collaterals.We retrospectively identified children with unilateral intracranial arteriopathy from two institutions. Two blinded raters from each institution reviewed magnetic resonance or digital subtraction angiography at baseline and =12 months. Patients were grouped according to presence or absence of lenticulostriate collaterals. Clinical features and vascular imaging outcomes were compared using univariate analysis and multivariate logistic regression.Forty-four children were included: 22 males, median age 8.2 years (range two to 16.9 years), and further stratified into the collateral group (n = 20) and non-collateral group (n = 24), with median follow-up of 25.5 months and 23 months, respectively. Both groups demonstrated similar rates of progression on vascular imaging at =12 months, 50% in the collateral group versus 37.5% in the non-collateral group (P > 0.05). The collateral group was associated with asymptomatic clinical presentation, normal brain MRI, border zone infarcts, and either vascular stabilization or new contralateral disease. The non-collateral group demonstrated either vascular improvement or discordant progression (combination of improved and progressive lesions). Using a multivariate model, collaterals continued to be an independent predictor of vascular outcome.This study suggests that lenticulostriate collaterals in children with unilateral intracranial arteriopathy may serve as a useful neuroimaging biomarker that helps to stratify patients with distinct clinical features and patterns of vascular evolution.

View details for PubMedID 29174857