Extent of collateralization predicting symptomatic cerebral vasospasm among pediatric patients: correlations among angiography, transcranial Doppler ultrasonography, and clinical findings

Abstract

OBJECT Although the development and prevalence of cerebral vasospasm (CV) has been extensively investigated in adults, little data exist on the development of CV in children. The authors hypothesized that even though children have highly vasoreactive arteries, because of a robust cerebral collateral blood flow, they rarely develop symptomatic CV. METHODS The authors retrospectively reviewed their university hospital's neurointerventional database for children (that is, patients ≤ 18 years) who were examined or treated for aneurysmal or traumatic subarachnoid hemorrhage (SAH) during the period 1990–2013. Images from digital subtraction angiography (DSA) were analyzed for the extent of CV and collateralization of the cerebral circulation. Results from transcranial Doppler (TCD) ultrasonography were correlated with those from DSA. Cerebral vasospasm on TCD ultrasonography was defined according to criteria developed for adults. Clinical outcomes of CV were assessed with the pediatric modified Rankin Scale (mRS). RESULTS Among 37 children (21 boys and 16 girls ranging in age from 8 months to 18 years) showing symptoms of an aneurysmal SAH (comprising 32 aneurysms and 5 traumatic pseudoaneurysms), 17 (46%) had CV confirmed by DSA; CV was mild in 21% of these children, moderate in 50%, and severe in 29%. Only 3 children exhibited symptomatic CV, all of whom had poor collateralization of cerebral vessels. Among the 14 asymptomatic children, 10 (71%) showed some degree of vessel collateralization. Among 16 children for whom TCD data were available that could be correlated with the DSA findings, 13 (81%) had CV according to TCD criteria. The sensitivity and specificity of TCD ultrasonography for diagnosing CV were 95% and 59%, respectively. The time to CV onset detected by TCD ultrasonography was 5 ± 3 days (range 2–10 days). Twenty-five (68%) of the children had good long-term outcomes (that is, had mRS scores of 0–2). CONCLUSIONS Children have a relatively high incidence of angiographically detectable, moderate-to-severe CV. Children rarely develop symptomatic CV and have good long-term outcomes, perhaps due to robust cerebral collateral blood flow. Criteria developed for detecting CV with TCD ultrasonography in adults overestimate the prevalence of CV in children. Larger studies are needed to define TCD ultrasonography–based CV criteria for children.