Stereotactic radiosurgery for intermediate- and high-grade arteriovenous malformations: outcomes stratified by the supplemented Spetzler-Martin grading system

Abstract

OBJECTIVE The supplemented Spetzler-Martin (Supp-SM) grading system was developed to improve the predictive accuracy of surgical risk for patients with brain arteriovenous malformations (AVMs). The aim of this study was to apply the Supp-SM grading system to patients having stereotactic radiosurgery (SRS) for Spetzler-Martin (SM) intermediate- (grade III) or high-grade (grade IV–V) AVMs to enable comparison with published microsurgical series. METHODS In 219 patients who underwent SRS during the period from 1990 to 2016, the Supp-SM grade was calculated for SM grade III (n = 154) or SM grade IV–V (n = 65) AVMs. The Supp-SM grades in these patients were 4 (n = 14, 6%), 5 (n = 36, 16%), 6 (n = 67, 31%), 7 (n = 76, 35%), and 8–9 (n = 26, 12%). Sixty patients (27%) had deep AVMs (basal ganglia, thalamus, or brainstem). Thirty-nine patients (18%) had volume-staged SRS; 41 patients (19%) underwent repeat SRS. The median follow-up period was 69 months for SM grade III AVMs and 113 months for SM grade IV–V AVMs. RESULTS AVM obliteration was confirmed in 163 patients (74%) at a median of 38 months after initial SRS. The obliteration rates at 4 and 8 years were 59% and 76%, respectively. Thirty-one patients (14%) had post-SRS deficits from hemorrhage (n = 7, 3%) or radiation injury (n = 24, 11%). Six patients (3%) died after SRS (hemorrhage, n = 5; radiation injury, n = 1). The rates of neurological decline or death at 4 and 8 years were 11% and 18%, respectively. Factors predictive of nonobliteration were deep location (HR 0.57, 95% CI 0.39–0.82, p = 0.003) and increasing AVM volume (HR 0.96, 95% CI 0.93–0.99, p = 0.002). Increasing AVM volume was the only factor associated with neurological decline (HR 1.05, 95% CI 1.02–1.08, p = 0.002). The Supp-SM grading score did not correlate with either obliteration (HR 0.94, 95% CI 0.82–1.09, p = 0.43) or neurological decline (HR 1.15, 95% CI 0.84–1.56, p = 0.38). CONCLUSIONS The Supp-SM grading system was not predictive of outcomes after SRS of intermediate- or high-grade AVM. In a cohort that included a high percentage (47%) of "inoperable" AVMs according to Supp-SM grade (≥ 7), most patients had obliteration after SRS, although there was a high risk of neurological decline.