Abstracts

Rationale: Pathological changes to the hippocampus and other medial temporal lobe structures are well-documented in patients with temporal lobe epilepsy (TLE). However, few studies have examined the effects of long-standing TLE on neocortical deterioration, including possible reductions in cortical thickness. Methods: High-resolution volumetric MRI scans were obtained on 21 patients with TLE and 21 healthy controls. In 15 patients, a diagnosis of TLE was supported by the presence of mesial temporal sclerosis, with no evidence of dual pathology in any patient. Mean cortical thickness was determined within regions of interest (ROIs) and point-by-point across the entire neocortex using automated cortical parcellation software (FreeSurfer). Group differences in mean cortical thickness for lateral temporal, medial temporal, frontal, parietal, and occipital lobe regions were tested using repeated measures analysis of covariance, controlling for age. Follow-up comparisons of individual gyral regions were performed when the overall ROI was significant. Discriminant function analysis was used to determine which linear combination of regions provided optimal classification rates of individual participants. Results: Decreased cortical thickness was observed bilaterally in frontal and lateral temporal lobe regions in patients with TLE relative to controls. Group differences were not observed within medial temporal, parietal, or occipital lobe regions (see Figure). Follow-up analyses within frontal and lateral temporal regions revealed bilateral reductions in cortical thickness within transverse temporal, paracentral, precentral, and pars opercularis cortex. Within only the left hemisphere, cortical thinning was observed in the superior temporal gyrus and the bank of the superior temporal sulcus, as well as in the middle frontal and medial orbital cortices in patients relative to controls. Within only the right hemisphere, cortical thinning was observed in the middle temporal gyrus, superior frontal gyrus, lateral orbitofrontal cortex, and pars triangularis. Discriminant function analysis revealed that average thickness of the left lateral temporal cortex was the strongest predictor of group membership, correctly classifying 74% of participants (16 controls, 15 TLE) in both the original and cross-validated samples. Conclusions: Patients with TLE show widespread pathology in neocortical regions that is not appreciated on standard imaging. Decreased cortical thickness in frontal and lateral temporal lobe regions may explain why many patients with TLE exhibit generalized cognitive dysfunction, despite previous suggestions of focal disease on clinical MRI. Detailed analysis of cortical thickness and other cortical features can help clinicians evaluate the total burden of disease in patients with TLE and perhaps increase diagnostic accuracy in patients with normal MRIs on clinical reading. Supported by: K23 NS056091 & GE.