Abstracts

Rationale: Quantitative MRI has consistently shown that drug-resistant temporal lobe epilepsy (TLE) relates to mesiotemporal and widespread neocortical atrophy (Bernhardt BC, Front Hum Neurosci, 2013). While some studies have suggested more marked atrophy in left compared to right TLE (Bonilha L, J Neurol Neurosurg Psych, 2007; Keller SS, PLoS One, 2012), others did not (Pail M, Epilepsia, 2010; Dabbs, Epilepsia, 2012). In all reports, TLE diagnosis and focus lateralization were based on pre-operative criteria alone. Here, we compared MRI markers of structural integrity in LTLE versus RTLE patients with histologically-confirmed hippocampal sclerosis and long-term seizure-freedom after surgery. Methods: From a cohort of 92 TLE patients who had undergone selective amygdalo-hippocampectomy, we selected those 60 (LTLE: n=25, RTLE: n=35) with histologically-proven hippocampal sclerosis who became seizure-free (Engel I) at a follow up of 7±2 years. There were no statistical differences in age, gender, age at seizure-onset, disease duration, history of febrile convulsions, and diagnosis of hippocampal atrophy on MRI volumetry. Based on isotropic T1-weighted MRI, we measured mesiotemporal subregional volumes and neocortical thickness (Kim H, MICCAI 2008; Bernhardt BC Neurology 2010), both at individual surface-points (vertex level) and in anatomical parcels. Measurements were normalized with respect to the corresponding distribution in 46 healthy controls, thereby controlling for normal inter-hemispheric variations. In each patient, normalized measures were sorted with respect to the seizure focus into ipsi- and contralateral. We carried out t-tests to detect differences between LTLE and RTLE. We systematically evaluated findings after false discovery rate (FDR) correction at p=0.05. Results: We did not find significant vertex-wise or parcellation-based differences in mesiotemporal and neocortical structures between LTLE and RTLE. Maximal absolute Cohen's d of a between-group difference was 0.43 across mesiotemporal parcels (ipsilateral hippocampal CA1) and 0.47 across neocortical parcels (ipsilateral medial occipital cortex), indicative of weak to moderate effects. Power analysis based on these effects revealed that samples with more than 75 patients per group would have been necessary to detect a significant effect in these areas. Conclusions: Our retrospective study in rigorously defined and closely matched TLE cohorts demonstrates equivalent structural compromise of mesiotemporal volume and neocortical thickness regardless of the side of the seizure focus.