Voxel-based morphometry in mesial temporal lobe epilepsy surgery: an exploratory Moroccan pilot study.

INTRODUCTION

Mesial temporal lobe epilepsy (MTLE) is a common focal epilepsy syndrome that may require surgery when seizures persist despite appropriate antiseizure medications. Although widespread gray matter (GM) abnormalities have been reported in MTLE, quantitative neuroimaging has not yet been implemented in the newly established epilepsy surgery programme in Fez, Morocco. This prospective observational exploratory pilot study evaluated the feasibility of implementing a whole-brain voxel-based morphometry (VBM) workflow using the Computational Anatomy Toolbox (CAT12), Statistical Parametric Mapping (SPM12), and ResectVol, and descriptively characterized GM differences before and after unilateral temporal lobectomy.

METHODS

Four patients with unilateral drug-resistant MTLE and eight age- and sex-matched healthy controls were prospectively recruited. Patients underwent preoperative and postoperative magnetic resonance imaging (MRI), whereas controls underwent one research MRI examination. Right-sided patient images were left-right flipped before tissue segmentation so that the epileptogenic and operated hemisphere corresponded to the left side in all patients. Exploratory whole-brain VBM maps were generated using general linear models for the contrasts controls > presurgical patients and presurgical > postsurgical patients; age, sex, and total intracranial volume (TIV) were entered as covariates in each model (voxel-level p < 0.001, uncorrected; cluster extent k > 50 voxels). These maps were used to identify anatomical regions for post hoc extraction of mean GM values. Regional values were compared using the Mann-Whitney-U-test and Wilcoxon signed-rank test.

RESULTS

The exploratory controls > presurgical contrast showed clusters in bilateral temporal, occipital, parietal, frontal, and sensorimotor regions. The exploratory presurgical > postsurgical contrast showed clusters predominantly in the ipsilateral temporal region. Post hoc paired comparisons showed lower postoperative mean GM values in the four studied regions. No postoperative GM increases were observed at the exploratory mapping threshold.

CONCLUSION

Implementing this VBM workflow was feasible within a newly established Moroccan epilepsy surgery programme. The findings are preliminary, descriptive, and hypothesis-generating; they do not establish disease mechanisms or postoperative reorganization. Larger, adequately powered longitudinal studies using dedicated repeated-measures models and multimodal quantitative MRI are required.