Validity assessment of grading scales predicting complications from embolization of cerebral arteriovenous malformations

Objective Endovascular embolization, though initially approved as an adjunctive therapy for surgical excision of cerebral arteriovenous malformations (AVMs), has found extensive use in the management of these lesions. A number of systems have been proposed to stratify AVMs by the procedural risk of embolization, including the Buffalo score and AVM Neuroendovascular grade. An external validity assessment of these systems has not been performed. Patients and methods A retrospective review of all patients who underwent embolization of cerebral AVMs at a single institution, between 2010 and 2016, was performed. Data including patient demographics, AVM characteristics, procedural details, complications, and outcomes were collected. Results Fifty-five embolization procedures in 39 patients (median age 53.1 years) were identified. Ten (25.6%) patients underwent more than 1 embolization procedure. A triaxial catheter system for support was used in 48 (87.3%) of the embolization procedures and a detachable tip microcatheter was used in 28 (50.9%). Complete obliteration of the AVM was achieved in 10.9% of the cases. There was one (2.6%) mortality unrelated to the procedure. Three minor (5.5%) and 2 major (3.6%) clinical complications occurred. Neither Spetzler-Martin grade, Buffalo score, or AVM Neuroendovascular grade correlated with complications. Conclusion Neither Buffalo score nor AVM Neuroendovascular grade predicted complications from embolization in the present study. Given the relative infrequency of complications, the number of factors that may influence AVM treatment, recent advancements in endovascular technologies, and the subjectivity inherent in these grading systems, the relative utility of risk stratification scales in the embolization of AVMs remains largely unknown.