Abstracts

Rationale: We have previously proposed that recording neuronal firing can be a useful method for providing feed-back from the brain for implanted drug delivery devices, like the hybrid neuroprosthesis, to treat focal epilepsy (Ludvig and Kovacs, 2002, US Patent No. 6,497,699; www.uspto.gov ). In a recent pilot study we indeed observed that in the focal acetylcholine (Ach)-seizure model in rats increased multi-neuron activity can be recorded before the development of EEG seizures (Ludvig et al., 2007, Epilepsia 48 (Suppl.6):388). The present study was devoted to the systematic analysis of this cellular electrophysiological phenomenon. Methods: Rats were chronically implanted with an epidural cup placed over the right frontal/parietal cortex, with an array of microelectrodes (50 micrometer diameter each) introduced into the underlying cortex 1 mm below the cortical surface. Two days after recovery, EEG and extracellular neuronal recordings started. Recordings were made before and after the delivery of 50 microliter of either artificial cerebrospinal fluid (ACSF), 150 mM Ach or 10 mM kainic acid (KA) into the epidural cup, using a microperfusion pump system, while the rat was behaving freely in a test chamber. The behavoral, EEG and neuronal data were collected and partly analyzed with the SciWorks software by DataWave Technologies; the neuronal data were analyzed with our proprietary software. We developed algorithms for (a) action potential detection, (b) a range of action potential statistics, and (c) template-matching-based waveform clustering. Results: The epidural Ach and KA deliveries generated focal EEG seizures associated with clonic convulsions. Before Ach deliveries, the average multi-neuronal firing rate was 38.2 ± 5.1 Hz (mean ± S.E.M.) during movement and 31.2 ± 5.2 Hz during immobility, which significantly increased to 53.3 ± 6.3 Hz in the pre-ictal state before EEG seizure onset (p < 0.05; n = 41). Before KA deliveries, the average multi-neuronal firing rate was 22.5 ± 6.7 Hz during movement and 19.2 ± 6.6 Hz during immobility, which, again, significantly increased to 44.6 ± 10.8 Hz in the pre-ictal state before EEG seizure onset (p < 0.05; n = 10). No significant changes in average multi-neuron firing rates were observed following the similar epidural delivery of ACSF. Conclusions: This study showed that before the development of focal EEG seizures, at least in the rat neocortex, local neurons increase their firing rate. We propose that capturing such pre-ictal multi-neuron activity increase can be used to predict focal seizures. (Supported by NYU/FACES).