Transsylvian amygdalohippocampectomy for mesial temporal lobe epilepsy: Comparison of three different approaches

Objective This study's objective was to compare the transinsular (TI‐AH), transuncus (TU‐AH), and temporopolar (TP‐AH) amygdalohippocampectomy approaches regarding seizure control, temporal stem (TS) damage, and neurocognitive decline. Methods We included 114 consecutive patients with unilateral hippocampal sclerosis (HS) who underwent TI‐AH, TU‐AH, or TP‐AH between 2002 and 2017. We evaluated seizure control using Engel classification. We used diffusion tensor imaging and postoperative Humphrey perimetry to assess the damage of the TS. We also performed pre‐ and postoperative memory performance and intelligence quotient (IQ). Results There were no significant differences in the proportion of patients free of disabling seizures (Engel IA+IB) among the three surgical approaches in the survival analysis. However, more patients were free of disabling seizures (Engel IA+IB) at 2 years of postsurgical follow‐up with TP‐AH (69.5%) and TI‐AH (76.7%) as compared to the TU‐AH (43.5%) approach (p = .03). The number of fibers of the inferior fronto‐occipital fasciculus postoperatively was reduced in the TI‐AH group compared with the TU‐AH and TP‐AH groups (p = .001). The rate of visual field defects was significantly higher with TI‐AH (14/19, 74%) in comparison to the TU‐AH (5/15, 33%) and TP‐AH (13/40, 32.5%) approaches (p = .008). Finally, there was a significant postoperative decline in verbal memory in left‐sided surgeries (p = .019) and delayed recall for both sides (p