Meningiomas-Related Epilepsy After Surgery

Abstract

Background: Meningioma-related epilepsy (MRE) is observed in approximately 30% of patients. Although studies focus on identifying risk factors related to pre- and postoperative MRE, there is no clear evidence regarding the timing for discontinuing antiseizure medications (ASMs) after surgical resection.

Methods: We retrospectively collected data from a series of naïve supratentorial meningiomas treated with surgical resection. Preoperative MRI was used to calculate the meningioma and peritumoral edema (PE) volumes through a voxel-based system. We analyzed the frequency of pre- and postoperative epilepsy in the group of meningiomas with and without perilesional edema (with PE > 1 cm³ as the cut-off).

Results: From a clinical series of 507 patients, we included 128 who underwent surgical resection in our center between January 2020 and December 2022, with a mean follow-up of 30.1 ± 19.8 months. Surgical treatment had a curative effect on MRE (41.4% preoperative vs. 19.5% postoperative; p = 0.0001). We observed a statistically significant reduction in the seizure rate in cases with preoperative PE (45.3% preoperative vs. 18.9% postoperative; p = 0.0002) and a non-statistically significant reduction in cases without PE (32.5% preoperative vs. 21.4% postoperative; p = 0.24). We observed ASM continuation in 37.8% of Engel IA patients.

Conclusions: PE increases the likelihood of MRE resolution with surgery. Our results show that surgical resection directly impacts MRE and ASM discontinuation in the presence of preoperative PE. The PE is a reassuring factor in decision-making regarding the timing of ASM discontinuation after surgery.

Keywords: antiseizure medications; epilepsy; meningioma; peritumoral edema.

Figure 1

Sequences selected for the segmentation. (A–C) Coronal, sagittal, and axial FLAIR sequences to quantify PE. (D–F) Coronal, sagittal, and axial T1-weighted sequences with contrast enhancement to estimate meningioma volume. FLAIR: fluid-attenuated inversion recovery; PE: peritumoral edema.

Figure 2

Example of segmentation process with the Slices website. (A,E) Automatic thresholding technique to identify initial regions of interest based on intensity values, respectively, for the PE on the FLAIR sequence and the meningioma on the T1-weighted-with-gadolinium sequence. (B,F) Partially automatic erasure of redundant signal with “erase outside” tool. (C,G) Definition of boundaries of edema and tumor through manual erasure. (D,H) Final segmented volumes. FLARI: fluid-attenuated inversion recovery; PE: peritumoral edema.

Figure 3

Screenshot of 3D Slicer showing the completed segmentation of the meningioma and PE, with 3D reconstruction and statistical quantification: the segmented PE is shown in green; the segmented meningioma is shown in yellow; and at the bottom, the statistical quantification of the segment volumes is displayed. The MRIs include a T1-weighted sequence and a FLAIR sequences merged at 50%. PE: peritumoral edema; MRI: magnetic resonance image; FLARI: fluid-attenuated inversion recovery.

Figure 4

Segmentation process for PE < 1 cm³. (A,E) Automatic thresholding technique to identify initial regions of interest based on intensity values, respectively, for the PE and the meningioma. (B,F) Partially automatic erase of redundant signal with “erase outside” tool. (C) Precisely definition of boundaries of edema through manual erasure, cutting all around redundant signal. (G) Manual erasing boundaries for tumor volume. (D,H) Final segmented volumes. PE: peritumoral edema.

Figure 5

Segmentation process for PE > 1 cm³. (A,E) Automatic thresholding technique to identify initial regions of interest based on intensity values, respectively, for the PE and the meningioma. (B,F) Partially automatic erase of redundant signal with “erase outside” tool. (C) Precisely definition of boundaries of edema through manual erasure, cutting all around the redundant signal inside the tumor. (G) Manual erasing boundaries for tumor volume. (D,H) Final segmented volumes. PE: peritumoral edema.

Figure 6

Patients flow diagram.

Figure 7

Binomial logistic regression: only preoperative seizure significantly correlated to the dependent variable “seizure outcome” with p = 0.016. The other independent variables analyzed (the volume of meningioma, the volume of PE, and the volume ratio) are not statistically correlated to the seizure outcome. PE: peritumoral edema.

Figure 8

Binary logistic regression analysis of tumor location and its association with (A) preoperative seizures and (B) seizure outcome (Engel IA). None of the tumor locations (convexity, sphenoidal wings, falx cerebri) showed a statistically significant association with either preoperative seizure presence (p = 0.317) or seizure outcome (p = 0.485). While convexity meningiomas showed a trend toward lower odds of preoperative seizures (OR = 0.179; p = 0.132), the results were not statistically significant.

Figure 9

Binary logistic regression analysis in the subpopulation of patients with preoperative epilepsy. Among the variables included (seizure semiology, seizure frequency, and use of antiseizure medications), only seizure frequency > 1/month was significantly associated with worse seizure outcome (Engel > IA), with an odds ratio of 9.39 (p = 0.003). Seizure semiology and ASM use did not show statistically significant associations.

Figure 10

Comparison of proportion of the rate of pre- and postoperative seizures in the presence or absence of PE in the surgical group: the surgery significantly decreases the seizure rate in the presence of PE (45.3% vs. 18.5%; p = 0.0002), and does not in the absence of PE (32.5% vs. 21.4%; p = 0.24). PE: peritumoral edema.

Figure 11

Results of the generalized estimating equations show how the Engel class influences the decision to discontinue or not to discontinue the ASM in the postoperative period, while the timing and the PE do not. ASM: antiseizure medication; PE: peritumoral edema. Notes: ^a: Set to zero because this parameter is redundant.