In response: The RNS System multicenter randomized double-blinded controlled trial of responsive cortical stimulation for adjunctive treatment of intractable partial epilepsy: knowledge and insights gained.

Abstract

We thank Dr. Osorio for the opportunity to highlight the objectives of the RNS System Pivotal trial, to describe the electrocorticographic (ECoG) data obtained by the RNS System, and to provide additional information on the primary effectiveness analysis method. Dr. Osorio correctly identifies the limitations of assessing the severity of epilepsy with seizure diaries and proposes that efficacy of the RNS System should have been assessed with quantitative ECoG measures of subclinical as well as clinical seizures. We would like to clarify that this was not possible. The RNS System is not designed to function as a precise seizure counter. The neurostimulator is programmed by the physician to detect and respond to specific ECoG patterns. In some cases these will be electrographic seizures, but in most cases these are interictal discharges. Therefore, detected events cannot be used to provide a reliable seizure count. In addition, although the RNS Neurostimulator continuously senses the ECoG, it stores a limited quantity of recordings. Therefore, stored ECoG events provide a sample of the total number of events. Finally, the RNS System senses and provides responsive focal stimulation to one or two seizure foci via two quadripolar leads (each containing four electrodes). Determining the duration and extent of seizure spread requires greater spatial coverage than is possible with the RNS System. The intent of the Pivotal trial was to demonstrate safety and effectiveness of the RNS System and to provide information necessary for U.S. Food and Drug Administration (FDA) approval for a specific indication for use. The trial design was agreed upon with the FDA and, as in all epilepsy trials leading to FDA approval, efficacy was evaluated with seizure diaries. Based on his own work, Dr. Osorio asserts that responsive electrical stimulation does not decrease the total frequency of clinical and subclinical seizures. The RNS System Pivotal trial demonstrated that responsive stimulation significantly reduced the frequency of clinical seizures compared to sham stimulation. The effect of responsive stimulation improves over time and is associated with significant improvements in quality of life. The effect on features of the ECoG is arguably less relevant. As we learned in our neurology training, we treat the patient, not the electroencephalogram (EEG). Dr. Osorio is concerned that we omitted details about the application of the generalized estimating question, which was the analysis method requested by FDA for the primary efficacy endpoint in order to properly account for over-dispersed and highly variable longitudinal seizure count data. In fact, copious information about the analysis is easily accessible. Morrell et al. (referenced in Heck et al.) provide detail, and additional highly technical information, including the degree of dispersion, is available in the RNS System product labeling.1, 2 For drugs or devices, a single trial is not sufficient to define the entirety of a therapy's clinical effects and mechanism(s) of action. Dr. Osorio lauds this study as a landmark in the annals of epilepsy therapeutics and we take satisfaction that its success will spur additional research to contribute to development of stimulation therapy and to provide fundamental information about epilepsy. Dr. Morrell is an employee of Stanford and of NeuroPace and holds an equity interest in NeuroPace. I confirm that I have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.