Gamma Knife® stereotactic radiosurgery for intracranial cavernous malformations

Introduction: Gamma Knife (R) stereotactic radiosurgery (GKSRS) is a non-invasive alternative to surgical resection for cerebral cavernous malformations (CCMs), especially in eloquent locations. Methodology: A retrospective review was performed on an Australian cohort of patients receiving GKSRS for CCMs at a single institution. All patients exhibited symptoms and/or radiological evidence of haemorrhage before therapy. The minimum follow-up was 1-year post-GKSRS. McNemar's test was used for differences in matched-pair outcomes pre- and post-GKSRS with an alpha = 0.05. A systematic review and meta-analysis was additionally performed to synthesise the current published evidence on the clinical efficacy of stereotactic radiosurgery in reducing haemorrhage risk in CCMs using a DerSimonian and Laird random effects model. Results: Thirty-five patients (39cavernomas) underwent GKSRS. 87.2 % of patients had evidence of at least one haemorrhage before GKSRS and the remainder exhibited seizures. The median dose was 12.5 Gy in a single fraction (IQR 12-13). The median follow-up duration from GKSRS was 809 days (IQR 536-960). There was a significant reduction in matched annual bleed rate from pre-GKSRS (52.1 %) compared to after SRS (12.3 %) (p < 0.001) [OR = 0.07, 95 % 0.008-0.283] There was no statistically significant difference in seizure incidence pre- (30.7 %) versus post-GKSRS (17.9 %) (p = 0.13) [OR = 0.167, 95 %CI 0.004-1.37]. One patient (3 %) with a brainstem lesion experienced long-term treatment-related oedema with persistent ipsilateral weakness and tremors. On meta-analysis of 25 pooled studies, radiosurgery for the treatment of CCMs was associated with a statistically significantly relative risk (RR) reduction in haemorrhage events [random effects RR 0.12 (95 % CI 0.074-0.198), p < 0.001)], with most of the proportionate risk reduction occurring in the initial 2 years following SRS. Conclusion: GKSRS significantly reduces the annual rate of haemorrhage for intracranial cavernomas in this cohort and on meta-analysis, particularly in the first 2 years following treatment. The overall risk of treatmentrelated morbidity is low.