Nidal embolization of brain arteriovenous malformations: rates of cure, partial embolization, and clinical outcome.

OBJECT: Nidal embolization of brain arteriovenous malformations (bAVMs) has become an increasingly important component of bAVM treatment. However, controversy exists as to the relative efficacy and safety of single-stage versus multistage approaches to bAVM embolization, with recent literature favoring multistage strategies. The authors present a series of consecutive bAVMs embolized at their institution, demonstrating the safety and efficacy of a predominantly single-stage embolization strategy. The safety and efficacy of embolization are reported in the context of predetermined treatment strategies to provide more generalizable insight into treatment outcome.
METHODS: One hundred thirty consecutive patients with 131 bAVMs underwent endovascular embolization at a single center. Diagnostic angiography with superselective microcatheterizations was performed in all patients. Postembolization angiograms were reviewed by 3 neuroradiologists for degree of occlusion and angiographic evidence of procedural complications. Patients were divided into cohorts based on the prospectively determined treatment strategy, which included the following: global devascularization of the bAVM (Devasc); targeting of a focal angioarchitectural weakness (Target), typically as an adjunct to surgery or Gamma Knife treatment; and primary occlusion of the bAVM by embolization alone (Occlude). Safety and efficacy were evaluated in the context of these treatment groups.
RESULTS: The 131 bAVMs were treated over an average of 1.28 embolization sessions per bAVM; 105 bAVMs (80%) were treated in a single stage. The average percentage devascularization in the Devasc arm was 85.3%, which was statistically significantly greater than the 72% aggregate devascularization reported in 8 modern N-butyl cyanoacrylate and Onyx papers based on 1-sample Wilcoxon rank-sum testing (p<0.001). Focal angioarchitectural weaknesses were successfully embolized for all 24 bAVMs in the Target group, directly with the embolic agent in 23 bAVMs and indirectly in 1 bAVM with a venous aneurysm/pseudoaneurysm by reducing arterial inflow and inducing venous thrombosis. Lesions in all patients in the Occlude arm were 100% occluded with embolization alone. Overall, the bAVMs in the Occlude arm were significantly smaller and required embolization of fewer pedicles than those in the Devasc group. One patient (0.8%) experienced significant morbidity following embolization, and 1 patient in the cohort died (0.8%).
CONCLUSIONS: This research communicates the authors' experience in developing a largely single-stage strategy for embolization of bAVMs. The results suggest that an aggressive, single-stage embolization may be implemented with a margin of safety and effectiveness similar to the multistage approaches more commonly reported in the literature. This work additionally introduces the importance of prospective assignment to a treatment strategy in assessing procedural outcome in bAVM embolization, thereby improving generalizability of the results and allowing for more rigorous interpretation of efficacy and safety.