Aims Using surface based morphometry we have examined the area, thickness and volume of the cortex in the same groups of patients and explored the relationship between these cortical parameters and cognitive performance.

Methods 22 chronic interictal psychosis IP (9 males; mean age 38.9 years; duration of epilepsy 31.9 years; duration of psychosis 12.1 years), 23 non-psychotic TLE patients NIP (11 males; mean age 38.7 years; duration of epilepsy 25.7 years) and 20 healthy controls HC (11 males; mean age 36.0 years). Patients were excluded if they had lesions other than hippocampal sclerosis. Diagnosis was made using DSM-IV criteria. Imaging was performed on a 1.5T GE Signa MRI scanner. A T1-weighted axial dataset was obtained using an IR-SPGR echo sequence. Images were processed using surface based morphometry (FreeSurfer version 4.3.0), a method that allows the measurement of cortical parameters. Premorbid IQ, current IQ, executive function (working memory span, working memory manipulation and planning) and verbal episodic memory were measured. Associations between cortical parameters and cognitive measures were examined using linear mixed models adjusted by age, gender and total brain volume.

Results There were no significant differences in age, gender or handedness across the three groups. The total brain volume was largest in HC and smallest in the IP group (p<0.001). Inferior frontal thinning was observed in IP compared to NIP patients (pars opercularis p=0.027, pars triangularis p=0.028) and in IP compared to HC (superior frontal p=0.013, pars opercularis p=0.009, pars triangularis p=0.016). IP patients had lower current IQ and worse scores in the working memory span and working memory manipulation (p<0.05) tests compared to NIP. Higher current IQ was associated with larger frontal (p=0.021), and temporal cortical areas (p=0.035), and with the thickness of the parietal cortex (p=0.010) in the IP group. Higher current IQ was also associated with larger cortical volume of frontal (p=0.014), temporal (p=0.011) and parietal (p<0.001) regions. No such associations were present in NIP patients.

Conclusions The most salient findings in our study were the thinning of the frontal cortex in IP patients but not in NIP, and the association in the IP group between greater area of the fronto-temporal cortex and the thickness of the parietal cortex with higher current IQ. These findings demonstrate the presence of extra-lesional cortical abnormalities in IP patients and suggest that there are important neurobiological differences between IP and NIP. Our findings also suggest a similar relationship between cognitive changes and cortical parameters in IP and schizophrenia.