Postembolization neurological deficits in cerebral arteriovenous malformations: Stratification by arteriovenous malformation grade

OBJECTIVE: To stratify the risk of embolization during the treatment of cerebral arteriovenous malformations (AVMs) by grade and to assess its impact on the overall treatment risk. METHODS: Patients with cerebral AVMs treated with embolization between 1995 and 2004 were retrospectively reviewed. Age, sex; AVM grade, location of lesion, number and location of embolized arteries, and number of embolization sessions were analyzed with respect to neurological or vascular complications after embolization. RESULTS: Embolization was performed in 153 patients: 508 vessels were embolized during 203 sessions (mean, 3.3 vessels per patient). The mean angiographic and clinical follow-up periods were 1.7 and 2.1 years, respectively (range, 3-60 mo). The periprocedural morbidity and mortality rate was 11.8%; but at the last follow-up examination, only 2% of survivors were significantly disabled (modified Rankin score 2). One (0.7%) patient died, and 17 patients experienced unexpected neurological deficits immediately after embolization. Five of these patients demonstrated near or total recovery during follow-up. The number of branches embolized was the only variable significantly related to neurological deficit (P < 0.017). The long-term rates of neurological deficits after embolization were 0, 5, 7, 10, and 18%, respectively, for AVM Grades I through V. Among 114 patients who underwent preoperative embolization, follow-up deficit rates of Grades I through V were 0, 5, 6, 6, and 25%, respectively. Long-term permanent deficits from embolization occurred in 8.6% of patients. CONCLUSION: Endovascular treatment carries a procedural risk related to AVM grade and number of branches treated. This risk should be weighed carefully in the context of overall treatment morbidity and mortality.