Extent of piriform cortex resection in children with temporal lobe epilepsy

Abstract

Objective: A greater extent of resection of the temporal portion of the piriform cortex (PC) has been shown to be associated with higher likelihood of seizure freedom in adults undergoing anterior temporal lobe resection (ATLR) for drug-resistant temporal lobe epilepsy (TLE). There have been no such studies in children, therefore this study aimed to investigate this association in a pediatric cohort.

Methods: A retrospective, neuroimaging cohort study of children with TLE who underwent ATLR between 2012 and 2021 was undertaken. The PC, hippocampal and amygdala volumes were measured on the preoperative and postoperative T1-weighted MRI. Using these volumes, the extent of resection per region was compared between the seizure-free and not seizure-free groups.

Results: In 50 children (median age 9.5 years) there was no significant difference between the extent of resection of the temporal PC in the seizure-free (median = 50%, n = 33/50) versus not seizure-free (median = 40%, n = 17/50) groups (p = 0.26). In a sub-group of 19 with ipsilateral hippocampal atrophy (quantitatively defined by ipsilateral-to-contralateral asymmetry), the median extent of temporal PC resection was greater in children who were seizure-free (53%) versus those not seizure-free (19%) (p = 0.009).

Interpretation: This is the first study demonstrating that, in children with TLE and hippocampal atrophy, more extensive temporal PC resection is associated with a greater chance of seizure freedom-compatible with an adult series in which 85% of patients had hippocampal sclerosis. In a combined group of children with and without hippocampal atrophy, the extent of PC resection was not associated with seizure outcome, suggesting different epileptogenic networks within this cohort.

Figure 1

Flowchart demonstrating the initial cohort of children who underwent any form of temporal lobe resection, number of exclusions per category, and the final cohort of children who had necessary data for the intended analysis. *Three children were excluded from the analysis since postoperatively they were found to have evidence of multifocal‐onset epilepsy. Ages are given as a median (and interquartile range). ATLR, anterior temporal lobe resection; F, female; L, left; M, male; R, right.

Figure 2

Image processing pipeline in one child who underwent a left anterior temporal lobe resection.

Figure 3

Boxplots showing the relationship between extent of resection of the piriform cortex (PC) (temporal and frontal parts), hippocampus and amygdala in (A) all children with TLE and (B) those children with TLE and hippocampal atrophy. *Shows statistical significance (p = 0.009). The colored boxplots show the seizure‐free (SF) group for each region and the gray boxplots the respective not seizure‐free (NSF) group.

Figure 4

Preoperative volumes for the (A) temporal part of the piriform cortex (PC), (B) frontal part of the PC, (C) amygdala, and (D) hippocampus in the entire TLE cohort (n = 50). The preoperative and postoperative PC ROIs could be parcellated in all 50 children, but in 5 out of 50 and 4 out of 50 parcellations of the amygdala and hippocampus could not be performed, respectively, due to lesions in these regions. There were no significant differences in the volumes of the piriform cortex (temporal or frontal parts) or amygdala between the ipsilateral and contralateral sides. The median hippocampal volume was smaller on the ipsilateral than contralateral side. *Shows statistical significance (p = 0.0035). Marks statistical significance in the differences detected between the ipsilateral (“Ipsi”) and contralateral (“Contra”) sides to the resection.

Figure 5

Inter‐rater agreement for segmentation of the piriform cortex volumes (mm³) (including the frontal and temporal parts) between two raters.