Factors Influencing Postoperative Seizure Persistence in Patients with Supratentorial Meningioma

Introduction: To identify factors that increase risk of persistent seizures after resection in patients with supratentorial meningioma and preoperative seizures.

Methods: A retrospective review of patients with supratentorial meningiomas who underwent surgery between 1980 and 2019 at the Cleveland Clinic was conducted. Patients were excluded if they did not experience preoperative seizures or if they had an existing seizure disorder diagnosis prior to meningioma resection. Data was collected on demographics, tumor characteristics, pre- and postoperative management and seizure features, and radiographic features. Persistent seizures were defined as preoperative seizures plus seizures after post-operative day 30.

Results: Among 219 patients that were reviewed, 176 met inclusion criteria. Of those, 100 patients experienced preoperative seizures (average 57.8 years, 64% female), and 76 patients experienced pre- and post-operative seizures (average 56.5 years, 66% female). There was no significant difference in preoperative seizure semiology (43% vs 49% focal, p=0.51) and tumor laterality (42% vs 39% right-sided, p=0.63) between cohorts. 59 patients with pre-operative only seizures and 47 patients with pre- and post-operative seizures had radiographic data: tumor volume was greater for patients with preoperative only seizures (41.9cm3 vs 24.9cm3, p=0.003) and Rolandic involvement was more likely for patients with persistent seizures (44.1% vs 66.7%, p=0.05). When comparing seizure-free patients at last follow-up by Engel classification to those with ongoing seizures (classes II-IV), absent peritumoral brain edema was a significant predictor of seizure freedom (35 vs 19 patients, p=0.03). Extent of resection was not significantly different between the cohorts (75% vs 75% subtotal resection, p=0.97).

Conclusion : Preliminary findings demonstrate that tumor volume was a significant predictor of preoperative seizures alone, but Rolandic involvement was a significant predictor of persistent seizures. Furthermore, persistent seizures may be influenced by tumor location, motor strip involvement, and peritumoral brain edema.