Abstracts

Rationale: Very fast ECoG activity, especially >250 Hz ("fast ripples"), has been implicated in epileptogenesis and seizure generation. Previous reports have concentrated on temporal lobe epilepsy with hippocampal sclerosis in both chronic experimental models and clinical cases. Here we determine whether fast ripples also provide a marker for the epileptic focus in an experimental temporal lobe epilepsy with no detectable neuronal loss. Methods: We induced epileptic foci in anaesthetised rats by unilateral intrahippocampal injection of tetanus toxin, and implanted electrodes into ipsi- and contralateral CA3 and either CA1 or dentate gyrus. Recordings started 3-6 days after the operation, with sessions of a few hours repeated up to 21 days. At the end of the recordings the rats were killed by anaesthetic overdose and their brains were prepared for histology. Results: A total of 60 electrographic seizures were recorded from 10 rats. The interictal periods were characterised by interictal discharges, typically lasting under 100 ms, and repeated epileptic discharges or polyspikes lasting a few seconds. 73% of seizures were first detected by the electrodes in the injected hippocampus, while interictal discharges could occur in either hippocampus alone or in both within a few ms. High-frequency activity was superimposed on both interictal and ictal activity and was analysed in both the frequency and time domains. The first spectral moment and the ratio between the powers of ripples (100-250 Hz) and fast ripples (251-600 Hz) both were higher in the injected hippocampus. These differences could be attributed to the more frequent occurrence of fast ripples in the injected hippocampus, and the similar rates of occurrence of ripples in the two hippocampi. Similar differences in high-frequency activity were recorded between the two sides during seizures. None of the 7 rats that were analysed histologically showed evidence of neuronal loss. Conclusions: Fast ripples can occur in epileptic foci lacking any discernible neuronal loss, and certainly without hippocampal sclerosis. In the tetanus toxin model fast ripples were always more common in the injected hippocampus, unlike interictal discharges and ripples, both of which were equally likely to occur in either hippocampus. We conclude that fast ripples are a more reliable marker for the primary epileptogenic zone than other kinds of interictal activity. Fast ripples should be considered for their potential contribution to pre-surgical work-up of non-lesional temporal lobe epilepsy. Funding: Wellcome Trust and Epilepsy Research UK