Abstract
OBJECTIVE: To evaluate the role of preradiosurgical embolization on obliteration rate, reduction of size, irradiation dose, and neurological outcome, in 90 patients presenting large arteriovenous malformations (AVMs). METHODS: Between October 1993 and October 2006, 90 radiosurgical procedures were performed to treat brain AVMs Spetzler-Martin (SM) grades III, IV, and V at the Department of Radiosurgery and Radiology of the Real e Benemérita Associação Portuguesa de Beneficência de São Paulo, São Paulo, Brazil. Fifty-nine patients had embolization before radiosurgery and complete clinical and radiologic follow-up for at least 3 years. Inclusion criteria were as follow: SM grades III, IV, and V AVMs, no previous treatment, and clinical and radiological (angiogram and MRI) follow-up for at least 3 years. Obliteration rate, reduction of size, irradiation dose, and neurological outcome were compared in these two cohorts of patients. Mann-Whitney test, "Student's t-test," and χ(2) tests were used for statistical analysis, as appropriate. The level of significance was determined at p < 0.05. RESULTS: The mean size of the AVMs that underwent embolization was significantly greater when compared with non-embolized group (p < 0.05). Embolization significantly reduced the AVM diameter. Irradiation dose was significantly smaller in the embolized group (p < 0.05). No significant differences in final clinical outcomes, postprocedural radiological findings, rate of occlusion, and need for additional procedures were observed between the two groups (p < 0.05). CONCLUSION: Preradiosurgical embolization of large AVMs does not result in impaired obliteration rate compared with cases treated with radiosurgery alone. It did not add further morbidity and presented benefits of reducing size of the AVMs. Preradiosurgical embolization may facilitate the coverage of the AVM with the effective irradiation dose. Combined management may be effective for selected large lesions considered unsuitable for radiosurgery and otherwise untreatable.