Abstracts

Rationale: Focal hypometabolism on fluoro-deoxyglucose positron emission tomography (FDG-PET) is a common finding in patients with temporal lobe epilepsy (TLE). Several studies have proposed that FDG-PET detected hypometabolism is caused by a loss of hippocampal neurons associated with hippocampal sclerosis, however recent studies have demonstrated significant hypometabolism in the absence of structural MRI abnormality. Further, no studies have investigated FDG-PET hypometabolism prior to the development of epilepsy. The underlying pathophysiology of hypometabolism, and whether it reflects a primary epileptogenic process or a secondary effect of chronic seizures therefore remains unknown. Such studies are difficult to perform in humans, so we have for the first time investigated the temporal ontogenesis of cerebral hypometabolism and limbic atrophy following an initial epileptogenic insult through to the development of epilepsy in the post kainic acid (KA) induced status epilepticus (SE) model of TLE using in-vivo imaging. Methods: Five serial FDG-PET and MRI imaging acquisitions were acquired in 19 animals over a period of six weeks; (i) one week prior to the induction of SE, (ii) 24 hours, (iii) one week, (iv) three weeks, and (v) five weeks following SE. Twelve of these animals received KA induced SE, one week following the initial scan, while the remaining seven control animals received saline only. Hippocampal histological correlates including stereological cell counts and immunohistochemical analysis of protein expression were also conducted. Results: Repeated measures ANOVA showed a significant effect of time (p>0.01), treatment (p>0.01) and brain region (p>0.01) in post KA induced SE animals. Post-hoc planned comparison analysis revealed markedly reduced cerebral FDG uptake significantly decreased when compared to control (21%) and baseline uptake (13%) at the earliest timepoint (24 hours) following SE (p<0.05). The hypometabolism most severely affected the limbic regions of the brain, and persisted (to a lesser degree) through the latent period before becoming more marked again at the epileptic time-point (5 weeks post-SE). MRI revealed a significant gradual reduction in limbic volume with a corresponding increase in ventricular volume from one week post SE which did not correlate with the hypometabolism. Histopathological correlations revealed a significant negative correlation between the degree of hippocampal hypometabolism in the chronic phase of epilepsy and the amount of synaptogenesis and glucose transporter expression. Conclusions: Limbic hypometabolism occurs very early in the epileptogenic process in this rat model of TLE and correlates with histological outcomes of epileptogenesis. These findings suggest that hypometabolism likely reflects an early functional change within the brain post SE rather than a secondary consequence of chronic repeated seizures or cell loss. This may represent an important initial process involved in the development of TLE.