Ruptured brain arteriovenous malformations in children: correlation of clinical outcome with admission parameters

To gain a better understanding of how clinical outcome in children with ruptured brain arteriovenous malformations (bAVMs) correlates with clinical and imaging parameters on admission.

The authors retrospectively reviewed patients with bAVMs, aged 18 or below, managed at their hospital between January 1992 and December 2008. Clinical outcome was assessed using the modified Rankin Scale (mRS). Patients with ruptured bAVMs were analyzed; their clinical parameters and computerized tomography findings on admission were recorded. Clinical outcome was then evaluated against admission scores using the Glasgow Coma Scale (GCS), the World Federation of Neurosurgical Societies Grading System of Subarachnoid Hemorrhage (WFNS-SAH), the Spetzler-Martin grade, the intracerebral hemorrhage score and 2 other independent parameters, namely pupillary response and significant focal neurological injuries. Spearman's correlation coefficient, linear regression analysis and multivariate logistic regression analysis were used for data analysis.

40 pediatric patients with bAVMs were found and 32 of them presented with hemorrhage (80%). In the 32 children with ruptured bAVMS, follow-up ranged between 7 and 204 months (median 100.5 months). The complete excision/obliteration rate as confirmed by digital subtraction angiography was 73.3%. The mRS scores at the last follow-up were: grades 0-II in 87.5%; grades IV-V in 6.25%, and grade VI in 6.25%. The Spetzler-Martin grade, the intracerebral hemorrhage score, the WFNS-SAH grade, the GCS scores, and a combined scoring scale consisting of the GCS, pupillary response and significant focal neurological injuries correlated significantly with clinical outcome 6 months after hemorrhage. However, on the scatter diagrams, it appeared that only the combined scoring scale might be valid for clinical practice. Multivariate logistic regression analysis showed that the combined scoring scale was a statistically significant independent predictor of clinical outcome 6 months after hemorrhage.

In this series of pediatric patients with ruptured bAVMs, although various grading scales correlated significantly with clinical outcome 6 months after hemorrhage, only the combined scoring scale might have the potential to be applied to predict clinical outcome in these children.