Radiosurgery for arteriovenous malformations: the University of Toronto experience

From July 1989 to February 1996, 130 patients underwent stereotactic radiosurgery. We report the results of the first 50 patients eligible for a minimum of three years of follow-up.

Twenty women and 30 men, (mean age: 37.5 years) were treated by dynamic rotation on a 6 MV linear accelerator. Prior treatment was embolization in seventeen, surgery in three and embolization and surgery in six. All had DSA and enhanced CT scanning, while some had MRI. Forty-seven treatments used a single isodose. Restricting eloquent normal tissue to 15 Gy, margin doses (at 50-90% isodose) were 12 Gy (one patient); 15 Gy (sixteen patients); 20 Gy (31 patients); 25 Gy (two patients). Maximum diameters were: < 1.5 cm (12 patients); < 2.0 cm (nine patients); < 2.5 cm (twelve patients); < 3.0 cm (thirteen patients; 3.0 cm (four patients).

Forty-five patients were evaluable at three years, with thirty-nine having angiography. Twenty-five had angiographically confirmed obliterations; two had parenchymal AVMs obliterated but with residual dural components; four had MRI evidence of obliteration (refused angiography). One patient acutely had a seizure; one patient (with hemorrhages, resection, and embolizations preceding two applications of radiosurgery, separated by 3.5 years) had worsening of memory.

Our uncorrected (five patients unevaluable at three years) and corrected angiographically confirmed obliteration rates are 54% and 60% respectively. Our follow-up (98% accounting of cohort; 78% angiographic rate) and explicit derivation of denominators help delineate the efficacy of radiosurgery at these doses.