Abstracts

Adults with medically intractable partial onset seizures were implanted as part of a multicenter safety/feasability study with a NeuroPace responsive neurostimulator system (RNS^TM) to continuously monitor and record Electrocorticography (ECoG) from depth and/or subdural strip electrodes and deliver electrical stimulation upon detection of ictal activity. Delivery of responsive stimulation is controlled by selection of various stimulation parameters including: 1. pulse width, 2. frequency, 3. current and 4. burst duration. Stimulation may also be controlled by adjustment of detection sensivity; detectors may be configured to respond to interictal and/or ictal onset activity. We sought to determine whether the amount of responsive stimulation correlates with a reduction in seizure frequency. Subjects ages 18-65 years [underline][gt] [/underline]4 disabling seizures per month for 3 consecutive months were implanted with the RNS at 3 Mayo sites (Arizona, Rochester and Jacksonville). Clinical seizure counts (Simple partial motor, complex partial and secondarily generalized) were collected for 3 months pre-implant and 4 months post-implant. Detection and responsive stimulation were enabled for a minimum of 3 months. The number of episodes treated with stimulation was recorded by each subject[apos]s implanted RNS device. Responsive stimulation settings (current, pulse width, frequency, and burst duration) and detection parameters were adjusted per clinical judgement during the study period and varied among subjects. No burst durations of [gt] 3000 ms were delivered in response to a detection. The total amount of responsive stimulation delivered for each subject was quantified in terms of total charge delivered over all electrodes in the leading phase of charge-balanced biphasic stimulation (The product of current, leading phase pulse width, pulses per stimulation burst, and number of stimulations delivered). As of May 3, 2005, data are presented for a three month analysis spanning days 28 through 111 post-implant. Nine subjects were implanted and had responsive stimulation enabled for a minimum of 3 months. Monthly averages of episodes-treated-per-day ranged from 1.17 to 3750 per day (mean 1580). Monthly averages of charge-delivered-per-day (current x puslewidth x number of pulses x stimulations per day) ranged from 0.114 millicoulombs (mC) to 224 mC (mean 74.8 mC). The total charge delivered over the analysis period was found to be related to a reduction in seizure frequency with a Pearsons r correlation of -0.83 (significance p[lt] 0.01) An implanted RNS system delivered stimulation for at least 3 months in 9 patients. Increased delivery of responsive stimulation, as measured by total charge delivered, was found to be significantly correlated with reduction in clinical seizure frequency. (Supported by Neuropace, Inc., Mountain View, CA.)