Remote effects of temporal lobe epilepsy surgery: long-term morphological changes after surgical resection

Abstract

Objective: We present a semi-automated method for quantifying structural changes after epilepsy surgery that accounts for tissue deformation caused by resection. We demonstrate its utility by comparing the remote structural effects of two surgical approaches, the anterior temporal lobectomy (ATL) and the selective amygdalohippocampectomy (SAH). Methods: We studied 37 temporal lobe epilepsy (TLE) patients who underwent resective surgery. Patients were treated with either an anterior temporal lobectomy (ATL, N=21) or a selective amygdalohippocampectomy (SAH, N=16). All patients received same-scanner MR imaging preoperatively and postoperatively (5+ months after surgery). To analyze structural changes in remote brain regions, we (1) implemented an automated method for segmenting resections with manual review, (2) applied cost function masking to the resection zone, and (3) estimated longitudinal cortical thickness changes using Advanced Normalization Tools (ANTs). We then compared post-operative changes in cortical thickness between the two surgical groups in brain regions outside the resected area. Results: Patients treated with ATL exhibited significantly greater cortical thinning globally when compared to patients treated with SAH (p = 0.049). There were significant focal differences between the two treatment groups in the ipsilateral frontal lobe (superior medial and medial orbital regions) and insula (p < 0.001, alpha = 0.05 Bonferroni corrected). No significant effects were seen in the contralateral hemisphere. Significance: We present and share a semi-automated pipeline for quantifying remote longitudinal changes in cortical thickness after neurosurgery. The technique is applicable to a broad array of applications, including surgical planning and mapping neuropsychological function to brain structure. Using this tool, we demonstrate that patients treated with SAH for refractory temporal lobe epilepsy have less postoperative cortical thinning in remote brain regions than those treated with ATL. We share all algorithm code and results to accelerate collaboration and clinical translation of our work.