Abstracts

Rationale: Responsive brain stimulation, delivered by the RNS System, has recently been investigated as an adjunctive therapy for medically refractory partial onset seizures in adults with 1 or 2 seizure foci. Data from open label periods of an initial feasibility study, a randomized double blinded controlled pivotal study and from an ongoing 7 year open label follow-up study suggests that concomitant use of the antiepileptic medication (AED) levetiracetam (LEV) may influence the clinical response to stimulation. Methods: Subjects were implanted with the RNS System Neurostimulator and Leads. The cranially implanted neurostimulator continuously senses electrocorticographic (ECoG) activity and, when abnormal ECoG activity is detected, provides brief bursts of biphasic stimulation. All subjects who started LEV and remained on a stable dose 3 months during the open label periods of the clinical trials were identified. Subjects were categorized by the location of seizure onset. Clinical seizure rates during 3 months immediately prior to initiating LEV (PRE) were compared to rates during 3 months after reaching a stable dose of LEV (POST). A percent change in seizure frequency was calculated ((POST-PRE)/PRE)*100. Subjects were considered to have an additional improvement with the initiation of LEV if the percent reduction in seizures (POST vs PRE) was 50% or greater. Measures of ECoG activity were also compared, including the rate of detections of abnormal ECoG activity recorded by the Neurostimulator. Results: Twenty-seven subjects, who were receiving responsive neurostimulation, initiated and maintained treatment with LEV as of 11/1/2012. Fourteen of these subjects were identified as having mesial temporal lobe onsets (MTL), 8 as having neocortical onsets (NEO) and 5 as having both MTL and non-MTL onsets (MTL+). After initiating LEV therapy, 79% (11/14) of the MTL subjects had an additional reduction in seizures of 50% or greater. The median percent reduction in this group was 63.4%, ranging from 100% reduction (seizure free) to a 6.3% increase. Twenty-nine percent (4/14) of these subjects became seizure free. In contrast, only 25% (2/8) of NEO subjects had an additional improvement ( 50% reduction) after initiating LEV and only 12.5% (1/8) became seizure free. For the MTL+ group, only 1 of the 5 subjects had an additional improvement with LEV. Conclusions: A potential synergistic interaction between LEV and responsive neurostimulation was observed for subjects with MTL onsets. This effect is consistent with preclinical studies indicating that LEV is effective in the kindling model of epilepsy and with clinical studies that have suggested that MTL patients are particularly likely to respond to LEV. The mechanism by which LEV may interact with neurostimulation to improve seizure control is unknown and warrants further investigation.