Abstracts

Rationale: A number of patients with medically intractable seizures have periventricular nodular heterotopias (PNH). However, it is not well established if these dysplasias can generate seizures and, if they can, whether responsive stimulation into the dysplasias will be effective in controlling seizures.Methods: Data was obtained from patients with a PNH participating in a clinical trial of responsive stimulation (RNS® System, NeuroPace) for adjunctive treatment of medically intractable partial onset seizures. Chronic ambulatory electrocorticographic recordings (ECoGs) were analyzed using a first experimental automated electrographic seizure detection algorithm to determine if electrographic seizures could occur or begin in the heterotopia. To confirm the algorithm’s performance, the ECoGs were manually reviewed to verify that the electrographic patterns detected by the algorithm were electrographic seizures. A second algorithm to detect the electrographic seizure onset was run on all of the ECoGs detected by the first algorithm to identify the timing and channel of the earliest onset. In addition, to assess effectiveness of responsive stimulation in patients with PNH, the rate of clinical seizures (as reported in seizure diaries) was computed for the most recent 3 months of treatment compared to the 3 month pre-treatment baseline.Results: Nine patients in the clinical trials for responsive neurostimulation had a PNH. Five of the 9 had 1 lead implanted in a PNH and 1 lead implanted in the neocortex or hippocampus, and 1 patient had leads implanted bilaterally in heterotopias. In the 5 patients who had 1 PNH lead and 1 non-PNH lead, 12,915 ECoG records collected over 5.5 to 11 years were analyzed, and electrographic seizures were detected in 495 ECoG records. In 3 of the 5 patients, electrographic seizures were more likely to involve the PNH lead than the non-PNH lead. The ratio of electrographic seizures (where an electrographic seizure is defined as at least 10 continuous seconds of high amplitude or high frequency activity) on the PNH lead compared to the non-PNH lead was 85.7%, 52.3%, 51.5%, 36.8%, and 16.1%. In 4 of 5 patients, the majority of seizure onsets were detected earliest in the PNH lead (although onsets were also seen in the non-PNH leads); in the 5th patient, earliest onsets occurred equally on the PNH and non-PNH leads. With responsive neurostimulation that included the PNH lead (n=5), the median clinical seizure rate change was -90% (range -30.43% to -100%) in the most recent 3 months compared to baseline. There were no device related serious adverse events.Conclusions: Chronic ambulatory ECoG recordings obtained by the RNS® System demonstrate that seizures can occur in a PNH and that responsive stimulation delivered into a PNH can effect a seizure reduction.