Abstracts

Detection of Taylor type focal cortical dysplasias (TTFCDs) with 18 FDG-PET has been reported in early series, but its usefulness has not been evaluated in cases with subtle or even absent abnormalities at current optimal MRI.The aim of the study was to evaluate the contribution of PET in cases in which MRI failed to clearly demonstrate the dysplastic lesion.
We analyzed imaging data of 13 consecutive patients (8 M, 5 F aged from 12 to 40 yrs, mean 25) operated on for early onset severe partial epilepsy associated with TTFCDs during the 3 last years. High-resolution MRI scans were performed at 1.5T (Signa GE) with 3D SPGR acquisition (1.2 mm thick contiguous slices), T2-weighted and FLAIR images.Twelve patients underwent an 18 FDG-PET scan using a high-resolution PET camera (ECAT 953/31B Siemens: 5 cases, HR+ CTI Siemens: 7 cases). MRI and PET scans were superimposed with ANATOMIST software. All patients were operated on after stereo-EEG following the PET scans. All MRI were reviewed by a trained radiologist in order to describe the MRI abnormalities, without knowledge of PET results and localization of the lesion. Cases with typical MRI features of TTFCDs were excluded from the study (n=5).
Eight patients had either subtle or not specific MRI abnormalities (n=6) or normal MRI (n=2). Subtle findings consisted of isolated minor gyral abnormalities (unusual gyral orientation or organization) in 4 patients, slight blurring of grey and white matter in 1 and mild linear signal abnormality in white matter in 1. Location of the MRI abnormalities was frontal in 3, central in 1, parietal in 2 and concordant with electro-clinical data. The 2 patients with normal MRI had fronto-temporal and central focus. PET scans (performed in 7 out 8 patients) showed a focal and severe hypometabolism corresponding to a single gyrus or a part of gyrus, associated with a surrounding mild to moderate hypometabolism in the region of MRI abnormalities, when present. Metabolic data were concordant with stereo-EEG findings. A continuous rhythmic spike activity was recorded within the hypometabolic gyrus which was also the site of spontaneous seizure onset. Histological examination demonstrated the typical features of TTFCDs with presence of giant dysmorphic neurons and balloon cells clearly corresponding to the severely hypometabolic gyrus. All patients are seizure free after a limited resection (follow-up 2 months to 3 years, mean 22 months).
In a significant number of cases (8 on 13 in our series), TTFCDs remain difficult to detect even with optimal MRI. However, high resolution 18 FDG-PETcoregistred with MRI proved very sensitive and accurate in these patients operated on at the adult age but with epilepsy onset in childhood, contributing to improved surgical strategy and results.