Abstracts

Rationale: Diffusion tensor imaging (DTI) enables quantification of directional diffusion, which is particularly informative in cerebral white matter. DTI of patients with epilepsy has shown diffuse white matter abnormalities, but the relationship with their epilepsy remains unknown. The abnormalities detected by DTI have also proven difficult to validate with histopathology. We aim to longitudinally characterize white matter changes during epileptogenesis in the lithium pilocarpine rat model of TLE by combining DTI with dedicated histological stains for axonal and myelin integrity. Methods: SE was induced in 21-day old male Wistar rats by injection of lithium and pilocarpine, known to induce spontaneous seizures in 75% of animals after 3 months. DTI was performed at 4 weeks (n=5) and 8 weeks (n=5) after pilocarpine (+SE) or saline injection on a 9.4 T MR system under halothane anesthesia. For DTI 21 slices of 0.5 mm were acquired (TR 3.5 s, 128x128, FOV 2.56x2.56 cm, 8 acquisitions, 6 directions, b values 0, 1200 s/mm2). For each direction the apparent diffusion coefficient was calculated, the construction of a tensor for each voxel. Data was coregistered and analyzed by ROI selection. For histology, animals were perfused with paraformaldehyde at 4 (n=6) and 8 weeks (n=6). A Klver-Barrera myelin stain and a Bielschowsky stain for axonal integrity were performed on 7 m-thick paraffin coronal sections of rat brain. The severity of white matter lesions in the corpus callosum was evaluated by two blinded examiners. Results: A typical ADC map is shown in Fig 1. Preliminary ROI analysis shows a significant decrease in mean diffusivity in the dentate gyrus (DG), CA3 and corpus callosum at 8 weeks compared to controls, with concomitant significant decreases in the principal eigen-vector (p<0.05). Fig 2 shows a sections stained for myelin of a control animal on the left side and a +SE animal at the right side at 4 weeks It is evident that there is decreased staining of the corpus callosum at 4 weeks. A preliminary analysis confirms this for the 4 week group, but less so for the 8 week group. Conclusions: In this model, as well as in human TLE, the regions that are most severely affected by sclerosis are the DG and CA3 regions. The decrease in diffusivity in these areas indicates more restricted diffusion as can be caused by gliosis. The DTI results also show that the corpus callosum is progressively affected, particularly by a decrease in axial diffusivity. We will further investigate whether this is due to a decrease in axonal density with a Bielchowsky stain. The preliminary results of the myelin stain indicates that SE leads to a delay in myelination of the corpus callosum. Detailed analysis of the DTI data by tract based spatial statistics will be performed to investigate whether this has an effect on the radial diffusivity. We conclude that our study shows white matter changes during epileptogenesis both in MR and histology before spontaneous seizures occur. Further analysis and additional histological stains will be performed to understand the observed changes.