Complete seizure-free rates following interventional treatment of intracranial arteriovenous malformations: a systematic review and meta-analysis

Seizures are common presenting symptoms of intracranial arteriovenous malformations (AVMs). This systematic review and meta-analysis aims to assess the current evidence regarding complete seizure freedom rates following surgical resection, stereotactic radiosurgery (SRS), and/or endovascular embolization of intracranial AVMs. A systematic review of PubMed, Ovid MEDLINE, and Ovid EMBASE was conducted following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. Included manuscripts were methodically scrutinized for quality, spontaneous AVM-associated or hemorrhage-associated seizures, complete seizure-free rates following each interventional treatment, follow-up duration; determination methods of seizure outcomes, and average time-to-onset of recurrent seizures after each treatment. Manuscripts that described patients with nondisabling seizures or reduced seizure frequency in their seizure-free calculations were excluded. Seizure freedom rates following surgical resection, SRS, and endovascular embolization were compared via random-effect analysis. Thirty-four studies with a total of 1765 intracranial AVM patients presenting with spontaneous AVM-associated seizures and 408 patients presenting with hemorrhage-associated seizures were qualitatively analyzed. For patients presenting with AVM-associated seizures, the complete seizure-free rates were 73.0% (321/440 patients; 95% CI 68.8–77.1%) following surgical resection, 60.5% (376/622 patients; 95% CI 56.6–64.3%) following SRS, and 44.6% (29/65 patients; 95% CI 32.5–56.7%) following endovascular embolization alone. For patients presenting with either AVM-associated or hemorrhage-associated seizures, the complete seizure-free rates were 73.0% (584/800 patients; 95% CI 69.9–76.1%) following surgical resection, 46.4% (572/1233 patients; 95% CI 43.6–49.2%) following SRS, and 44.6% (29/65 patients; 95% CI 32.5–56.7%) following embolization. For patients presenting with either AVM-associated or hemorrhage-associated seizures, the overall improvements in seizure outcomes regardless of complete seizure freedom were 82.6% (661/800 patients; 95% CI 80.0–85.3%), 70.6% (870/1233 patients; 95% CI 68.0–73.1%), and 70.8% (46/65 patients; 95% CI 59.7–81.1%) following surgical resection, SRS, and embolization, respectively. No study reported information about the time-to-onset for recurrent seizures in any patient following treatment, as seizure outcomes were only described at the last follow-up visit. The available data suggests that surgical resection results in the highest rate of complete seizure freedom. The rate of seizure improvement following surgery increased further to 82.3% when including patients who had improved seizure frequency without achieving true seizure freedom. Complete seizure-free rates following SRS or embolization were more ambiguous and lower when compared to surgical resection. There is a need for high quality studies evaluating AVM treatment modalities and clearly defined seizure outcomes, as the current literature consists mostly of heterogenous patient populations.