Abstracts

Rationale: The long-term fluctuations in seizure occurrence probability are well documented in both patients and various models of epilepsy. The most frequently observed type of the fluctuation is clustering. Previously, we have shown that seizures within the cluster follow specific dynamical path with increasing inter-seizure interval (ISI) towards the end of the cluster. The analysis of the cluster dynamics revealed that seizures occur by chance during early phases of the cluster. In contrast, dynamics of terminal parts of the cluster displays features of the process with memory. The main aim of this study was to explore the mechanisms which govern such a specific dynamical pattern of clusters. Methods: Epilepsy was induced in adult male Wistar rats (n = 7) by injection of 10 ng of tetanus toxin into the right dorsal hippocampus. Following the injection, electrodes were implanted bilaterally into dorsal hippocampus and motor cortex. After one week, animals were subjected to >2 weeks of continuous video-EEG monitoring. Seizures were detected semi-automatically and their behavioral correlates were classified according to Racine scale. Racine 3-5 seizures were classified as convulsive. Signal power during each seizure in each studied area was computed. Inter-ictal epileptiform discharges (IED) were detected in each channel using Hilbert envelope-based detector. For statistical analysis, each cluster was divided into two parts containing the same number of seizures. Results: Long-term fluctuation in seizure incidence was observed in all animals and we identified one cluster in each animal. On average, cluster lasted 2.3 0.2 days and contained 90 15 seizures. All clusters were characterized by progressive increase of ISI. During the first part of cluster, only 4 3 % of seizures were convulsive while during the second part, it was 46 8 % (p < 0.05). Power in motor cortex during seizures was 0.047 0.002 during the first and 0.082 0.003 mV2 during the second part of the cluster (p < 0.05). IED rate in the motor cortex increased from 0.018 0.003 Hz to 0.054 0.011 Hz (p < 0.05). In addition, analysis of 1 hour of signal from the second part of the cluster revealed the presence of independent IEDs in motor cortex. Conclusions: We have shown that during the course of a single cluster, the brain undergoes complex changes on local and global levels. They manifest by progressive increase of ISI which is paralleled by increase in behavioral severity of seizures and increased involvement of extrahippocampal structures, in our case of motor cortex, which was confirmed by increased signal power during seizures and IED rate. We hypothesize that the long-term seizure occurrence is governed by seizures themselves. The early seizures act as a positive feedback mechanism which increases the seizure severity and brain propensity to seize. In contrast, later convulsive seizures act as a negative feedback mechanism reducing the seizure occurrence and their cumulative effect promotes cluster termination. Funding: Supported by grants NFKJ 2012/10, GA CR 14-02634S, GA CR 15-08565S, AZV CR 15-29835A.