Abstracts

RATIONALE: It has been frequently hypothesized that the dentate gyrus (DG) acts as a gate modulating neuronal activity through the hippocampus (Hip). Hip atrophy (HA), identified on MRI, is associated with neuronal cell loss which may alter the DG gating of Hip neuronal activity. If HA impairs DG gating, then epileptic patients with HA may have enhanced Hip neuronal discharge compared to epileptic patients without HA. We measured interictal Hip activity during waking and sleep to assess DG gating. METHODS: Epileptic patients (n=19) were chronically implanted with depth electrodes to localize area of seizure onset. Spontaneous interictal unit activity was recorded from microwires in the Hip. A sleep record consisting of EOG, EMG and cortical EEG was used to stage sleep. Concurrent unit activity was continuously recorded overnight. Ten minute epochs of Hip unit activity were selected for analysis during waking, SWS and REM sleep. RESULTS: MRI identified 7 patients with HA and 12 without atrophy. Analysis of 70 Hip units showed the greatest differences in firing rate occurred during sleep. In patients with HA, mean discharge rate of Hip units ipsilateral to side of seizure onset was significantly higher as compared to Hip units on the contralateral side (P<.036), with the largest difference occurring during REM sleep. Conversely, in patients without HA, mean discharge rate of Hip units ipsilateral to seizure onset was decreased across all states as compared to the contralateral side, especially during REM sleep (P<.026). Significant unit analyses also included the ratio of discharge/suppression and burst discharge properties of Hip neurons. CONCLUSIONS: These results suggest that in patients with HA the ability of the DG to limit Hip throughput ipsilateral to the side of damage is diminished as compared to the DG on the less damaged side. In contrast, patients without HA demonstrated suppressed Hip unit activity ipsilateral to the side of seizure onset, suggesting enhancement of DG gating. The differences in DG gating between the two groups of patients is present across all states, but most pronounced during REM sleep. Supported by NIH grants NS02808 and NS33310.