Voxel-based T2 relaxation rate measurements in temporal lobe epilepsy (TLE) with and without mesial temporal sclerosis

Abstract

Introduction: Quantitative measurements of T(2) relaxation in the hippocampus for focus lateralization in mesial temporal lobe epilepsy (mTLE) are well established. Less is known to what degree such relaxation abnormalities also affect regions beyond the ipsilateral hippocampus. Therefore, the aim of this study was to characterize extent and distribution pattern of extrahippocampal relaxation abnormalities in TLE with (TLE-MTS) and without MRI evidence of mesial-temporal sclerosis (TLE-no).

Methods: Double spin echo images (TE1/2: 20/80 ms) acquired in 24 TLE-MTS and 18 TLE-no were used to calculate relaxation rate maps. These maps were analyzed by SPM2 and by selecting regions of interest (ROI) in the hippocampus and several extrahippocampal brain regions.

Results: In TLE-MTS, the results of the SPM and ROI analysis were in good agreement and showed the most severe relaxation rate decreases in the ipsilateral hippocampus but also in other ipsilateral temporal regions, orbitofrontal, and parietal regions and to a lesser degree in contralateral frontal regions. The relaxation rate decreases in TLE-no were confined to small regions in the ipsilateral anterior inferior and medial temporal lobe in the SPM analysis while ROI analysis showed additional regions in the ipsilateral hippocampus, amygdala, and anterior cingulate.

Conclusion: TLE-MTS showed extensive, widespread but predominantly ipsilateral temporal and also extratemporal T(2) relaxation rate decreases. In contrast, the findings of the SPM and ROI analyses in TLE-no suggested that if relaxation rate decreases are present, they are less uniform and generally milder than in TLE-MTS. This further supports the hypothesis that TLE-no is a distinct clinicopathological entity from TLE-MTS and probably heterogeneous in itself.

FIG. 1

Postprocessing steps (see also Methods). The T₁ was coregistered to T2/PD and the resulting image was used for segmentation with SPM2. The gray matter map was then normalized to the gray matter prior and the deformation parameters were applied to the coregistered T₁, T₂, and PD. The normalized T₂ and PD were used to create T₂ relaxation rate maps, which were masked with a tissue mask to reduce partial volume effects. For the analysis in SPM, the masked T₂ relaxation maps were smoothed with a Gaussian kernel of 4 mm FWH.

FIG. 2

Placement of the regions of interest (ROI) in the hippocampus (size: 44 pixels, Talairach coordinates: −24, −14, −17 and 27,−14,−17) and, amygdala (size: 44 pixels, Talairach coordinates: −23,−4,−17 and 23,−4,−17) (a), temporal stem (size: 24 pixels, Talairach coordinates: −44,−30,−7 and 46,−30,−7) (b), orbitofrontal white matter (size: 61 pixels, Talairach coordinates: −21,42,−7 and 21,42,−7) (c), thalamus (size: 68 pixels, Talairach coordinates: −14,−23,8 and 18,−23,8) (d), posterior (size: 24 pixels, Talairach coordinates: −11,−46,23 and −7,−48,23) and anterior cingulate (size: 24 pixels, Talairach coordinates −10,29,19 and 10,25,19) (e) and parietal white matter (size: 68 pixels, Talairach coordinates: −28,−50,25 and 31,−50,25) (f).

FIG. 3

FIG. 3a. Regions with reduced-relaxation rates in TLE-MTS (a) and TLE-no (b) and in TLE-MTS compared with TLE-no (c). TLE-no had no regions with reduced-relaxation rate when compared with TLE-MTS, indicating that T₂ relaxation abnormalities are more severe in the former group. The regions are displayed on the averaged symmetrical T₁ image of the control group.

FIG. 3

FIG. 3a. Regions with reduced-relaxation rates in TLE-MTS (a) and TLE-no (b) and in TLE-MTS compared with TLE-no (c). TLE-no had no regions with reduced-relaxation rate when compared with TLE-MTS, indicating that T₂ relaxation abnormalities are more severe in the former group. The regions are displayed on the averaged symmetrical T₁ image of the control group.