Abstracts

Rationale: In recent years deep brain stimulation (DBS) has been growing in popularity as a treatment option for improving seizure control in people with epilepsy. The mechanism of reduced seizures is unknown, but may be related to changes in arousal state, which could in turn modulate seizure threshold. To further investigate the hypothesis of increased arousal with DBS we investigated physiological EEG markers of cortical arousal during stimulation of the thalamic centro-median nucleus (CM) in patients with chronic neurostimulators at the Sao Paulo Epilepsy Center. Methods: Conventional scalp EEG recordings were performed with 24 electrodes sampled at 600 Hz, with 0.1 Hz low frequency and 200 Hz high frequency filter settings during outpatient visits following DBS implants. No task was performed during recordings but patients were required to remain awake at all times. Prior to the CM stimulator first being turned on, we collected 10 minutes of baseline EEG data. Stimulation was next initiated of the bilateral thalamic CM with biphasic 300 µs per phase square wave pulses at 130 Hz, and voltage was gradually titrated upward as tolerated. Five minutes of EEG data were collected at 1v of stimulation followed by a 5-minute period with no stimulation. This process was repeated with increasing voltages at 1 V increments until 5v of stimulation was reached. Periods of EEG with artifact or epileptiform activity were excluded from analysis. We performed time-frequency analysis on EEG data in MATLAB using a short-time Fourier transform to quantify changes in EEG power across all electrodes during each stimulation epoch. Results: Data were collected from 5 patients, 3 male, 2 female, age 7 to 18 years. Time-frequency plots for each individual patient and the group average spectrogram demonstrated an increase in broadband EEG power in gamma frequency range (>25 Hz), and a decrease in theta and alpha frequency (4-13Hz) power during each stimulation epoch, and these effects increased with stimulus voltage. These observations were confirmed by plotting the difference between baseline power and average power in each stimulation period at each frequency range across stimulus voltage increments. The differences in power behaved in a dose dependent manner, increasing progressively with the stimulation voltage. Conclusions: The observed increase in high frequency EEG power accompanied by a decrease in lower frequency power has been well-established in prior work to be associated with increased physiological arousal. These findings support the hypothesis that DBS stimulation may reduce seizures by acting on thalamo-cortical arousal networks. Increased cortical physiological arousal could also contribute to improved attention function, which has been previously reported with thalamic CM stimulation. These results have implications for the therapeutic mechanisms of thalamic DBS, and with further investigation may help provide improved methods for determining optimal DBS stimulation parameters to improve clinical outcome. Funding: Supported by the Betsy and Jonathan Blattmachr Family and the Loughridge-Williams Foundation