Metabolic and Electrophysiological Alterations in Subtypes of Temporal Lobe Epilepsy: A Combined Proton Magnetic Resonance Spectroscopic Imaging and Depth Electrodes Study
Guye M, Le Fur Y, Confort-Gouny S, Ranjeva J-P, Bartolomei F, Régis J, Raybaud CA, et al.;
Commented by , 22 Nov 2002
Background
Temporal lope epilepsy (TLE) is often resistant to medication and surgical treatment is used in selected cases to stop or reduce seizures. Proton magnetic resonance spectroscopic imaging (H-MRSI) has been used to locate the side and place of the epileptogenic zone (EZ) in the pre-surgical evaluation of patients with mesial TLE.
This was done by measuring absolute or relative decreases of N-acetyl aspartate (NAA), which is considered to be a marker of neuronal loss/dysfunction. It has been shown that relative decrease of NAA extends beyond the EZ, although the pathophysiological relation to epileptogenesis remains unclear.
Aim
To compare the metabolic regional changes measured by H-MRSI, with electrophysiological abnormalities recorded by depth electrodes in patients with subtypes of TLE.
Methods
Ten patients with drug resistant TLE underwent thorough noninvasive evaluation (video-EEG, interictal EEG spike localization, interictal/ictal SPECT and structural MRI).
Furthermore, a precise definition of the anatomical localization of the epileptogenic processes was obtained using intracerebral electrode examination, stereo-electro-encephalography (SEEG). An integrated MRI/MRSI protocol was conducted on a 1,5 T system.
Two H-MRSI acquisitions associated with two single-voxel acquisitions in the temporal poles was obtained, and the NAA/(choline+creatinine) ratio was chosen as a metabolic index. Fifteen normal subjects was explored with H-MRSI and used as controls.
The metabolic profiles in regions of interest (ROI’s) were compared among all 25 subjects, divided into four groups according to SEEG findings (abnormal, contralateral, normal, control). Nonparametric Mann-Whitney U test was used for statistical analysis.
Results
The metabolic index was significantly lower in all regions involved in SEEG abnormalities compared to controls (p < 0,05). This included regions with ictal discharges as well as regions with interictal paroxymal events. ROI’s that had normal SEEG was not different from controls except in one ROI.
The metabolic reduction included the structural changes seen on MRI and extended beyond. The metabolic changes in mesial structures were not specific for the mesial TLE subtype and generally extended outside the mesial structures. The metabolic index of ROI’s in the contralateral hemisphere was reduced in the temporal pole and hippocampus.
Discussion
In line with other studies the authors demonstrate a reduction of the NAA ratio in areas with structural lesions as well as in the EZ. The SEEG recordings furthermore showed that the metabolic changes correspond to a more widespread area characterized by interictal paroxymal events.
Additionally, bilateral metabolic abnormalities were observed in patients with seizure onset in only one temporal pole. This corresponds with findings in positron emission tomography (PET) studies although the sensitivity in H-MRSI seems to be higher.
The study has many interesting aspects and provides results of pathophysiological interest that can be obtained noninvasively at any time in patients with TLE. This may have possible implications with regard to clinical evaluation and treatment in patients with TLE in the future although further data are required.