Validation of the modified radiosurgery-based arteriovenous malformation score in a linear accelerator radiosurgery experience in Hong Kong.

The modified radiosurgery-based arteriovenous malformation (AVM) score (modified AVM score or Pollock-Flickinger AVM score [PFAS]) is a simplified grading system developed to predict outcome after gamma knife radiosurgery for cerebral AVM. The purpose of this study was to test the PFAS in a cohort of patients managed with linear accelerator (LINAC) radiosurgery. We analyzed 70 consecutive patients with cerebral AVM treated with LINAC radiosurgery in Hong Kong. The scores were determined by the following equation: Modified AVM score=(0.1×volume [cm(3)])+(0.02×age [years])+(0.5×location). The location values are as follows: hemispheric/corpus callosum/cerebellar=0; basal ganglia/thalamus/brainstem=1. A total of 74% of patients presented with ruptured AVM before radiosurgery. The overall obliteration rate was 86%. Five (7%) patients developed new permanent neurological deficits from delayed bleeding or radiation-induced complications. Modified AVM score correlated with the percentage of patients with AVM obliteration without new neurological deficits (≤1, 96%; 1.01-1.50, 78%; 1.51-2.00, 90%; >2, 50%; Spearman's rho 0.354, p=0.003). In conclusion, the modified AVM score is a good predictor of patient outcome after LINAC radiosurgery in our cohort. The modified AVM score can be used to guide treatment selection for cerebral AVM and stratify patients for future comparative analyses.