Closed-loop, responsive focal brain stimulation provides a new treatment option for patients with refractory partial onset seizures who are not good candidates for potentially curative epilepsy surgery. The first responsive brain neurostimulator (RNS® System, NeuroPace), provides stimulation directly to the seizure focus when abnormal electrocorticographic is detected. Seizure reductions of 44% at one year increase to 60 to 66% at years 3 to 6 of treatment. There is no negative impact on cognition and mood. Risks are similar to other implanted medical devices and therapeutic stimulation is not perceived.
Key points
- •
Closed-loop responsive stimulation of the seizure focus reduces the frequency of medically intractable partial onset seizures from 44% at 1 year to 60% to 66% over 3 to 6 years of treatment.
- •
Risks of responsive stimulation as provided by the first responsive neurostimulator (RNS System) are similar to other implanted medical devices, and the neurostimulator can be programmed so that therapeutic stimulation is not perceived.
- •
There are improvements in quality of life with responsive stimulation and no negative effects on mood or cognition, and some patients experience improvements in aspects of language and memory.
- •
Quantitative electrophysiological data combined with clinical seizure counts are used to establish optimal detection and stimulation settings for each patient.
- •
Chronic ambulatory electrocorticographic monitoring provides information regarding the location(s) of the seizure focus and may provide biomarkers to measure disease activity.
Introduction
Neurostimulation is an increasingly important treatment modality for disorders of the nervous system. Most neurostimulation devices are open-loop; stimulation settings are preprogrammed and do not automatically respond to changes in electrophysiological signals or the patient’s clinical symptoms. Open-loop stimulation is effective in several clinical applications, including Parkinson’s disease, essential tremor, dystonia, pain, and more recently, epilepsy. However, because many neurologic disorders are not static conditions, there is increasing interest in neurostimulation approaches that adapt to changes in clinical symptoms or quantitative biomarkers.
In contrast to open-loop stimulation devices, responsive (or closed-loop) neurostimulation devices modulate or adapt therapy in response to physiologic signals, in addition to clinically overt symptoms, and may be more efficient, effective, and better tolerated than open-loop stimulation. Responsive stimulation approaches are being explored for Parkinson’s disease. Investigational closed-loop stimulation for Parkinson’s disease provides stimulation in the subthalamic nucleus in response to changes in beta amplitude, with reports of improvements in motor function, speech, gait, and balance.
This review focuses on targeted cortical responsive stimulation, which has been approved by the US Food and Drug Administration (FDA) for the adjunctive treatment of epilepsy with partial onset seizures.
Introduction
Neurostimulation is an increasingly important treatment modality for disorders of the nervous system. Most neurostimulation devices are open-loop; stimulation settings are preprogrammed and do not automatically respond to changes in electrophysiological signals or the patient’s clinical symptoms. Open-loop stimulation is effective in several clinical applications, including Parkinson’s disease, essential tremor, dystonia, pain, and more recently, epilepsy. However, because many neurologic disorders are not static conditions, there is increasing interest in neurostimulation approaches that adapt to changes in clinical symptoms or quantitative biomarkers.
In contrast to open-loop stimulation devices, responsive (or closed-loop) neurostimulation devices modulate or adapt therapy in response to physiologic signals, in addition to clinically overt symptoms, and may be more efficient, effective, and better tolerated than open-loop stimulation. Responsive stimulation approaches are being explored for Parkinson’s disease. Investigational closed-loop stimulation for Parkinson’s disease provides stimulation in the subthalamic nucleus in response to changes in beta amplitude, with reports of improvements in motor function, speech, gait, and balance.
This review focuses on targeted cortical responsive stimulation, which has been approved by the US Food and Drug Administration (FDA) for the adjunctive treatment of epilepsy with partial onset seizures.
Neurostimulation for epilepsy
Epilepsy comprises a group of neurologic conditions characterized by recurrent seizures. The partial epilepsies (also known as localization-related epilepsy) are the most common type of epilepsy in adults, and at least one-third cannot achieve seizure control with antiepileptic medications. One option for these patients is to resect the seizure focus. However, many patients with partial onset seizures are not candidates for a cortical resection because the chance of a substantial seizure reduction is too low or the risk of a neurologic morbidity is too high. Moreover, not all patients who undergo resective procedures achieve seizure freedom. Approximately 30% to 40% of patients who undergo temporal lobectomies, which is the most common and most successful type of resective surgery, continue to have disabling seizures at 1 year after surgery.
Neurostimulation is an option for some patients with medically intractable partial seizures who have either failed or are not candidates for resective surgery. There are currently two FDA-approved neurostimulation therapies for adjunctive treatment of medically intractable partial-onset epilepsy: vagus nerve stimulation (VNS) and responsive cortical stimulation. A third neurostimulation modality, open-loop deep brain stimulation of the anterior thalamic nucleus, is not approved in the United States as of the date this is written, but is approved in other countries.
The VNS Therapy System (Cyberonics, Houston, TX, USA) provides open-loop scheduled stimulation to the vagus nerve using a pectorally implanted pulse generator and electrodes wrapped around the left vagus nerve. Stimulation is typically delivered for 30 seconds every 5 minutes. External application of a magnet over the pulse generator triggers additional stimulation.
Average seizure reductions in patients with intractable partial onset seizures during the blinded period of randomized controlled trials of the VNS Therapy System were 24% to 28% and were 35% to 44% at 2 years in open-label prospective studies. Side effects related to stimulation of the vagus and recurrent laryngeal nerve include voice alteration (50%), increased coughing (41%), pharyngitis (27%), dyspnea (18%), dyspepsia (12%), nausea (19%), and laryngismus (3.2%). A recently FDA-approved VNS model (Aspire SR; Cyberonics, Inc) activates stimulation when the heart rate exceeds a prespecified threshold in order to provide additional treatment for seizures that are accompanied by tachycardia (Aspire SR, Cyberonics, Inc). It is unclear at this time if this device will prove to be more efficacious or better tolerated than its open-loop precursor.
A reduction in seizures has been reported in several small and uncontrolled studies of open-loop continuous or scheduled neurostimulation as well as in one randomized controlled trial. Stimulation targets have included the cerebellum, caudate nucleus, centromedian nucleus, subthalamic nucleus, and hippocampus. In the only randomized controlled trial of open-loop stimulation, 110 adults with medically intractable partial onset seizures were randomized to scheduled or sham stimulation in the anterior nuclei of the thalamus on a schedule of 1 minute on and 5 minutes off. Patients treated with stimulation had a significantly greater reduction in seizures compared with the sham stimulation patients with an overall adjusted percent difference of −17% ( P <.039). The stimulated group had significantly more adverse events related to depression, memory, and concentration, but there were not differences between the groups by neuropsychological testing. The seizure reduction at 1 year was −41% and reached 57% to 65% in years 3 through 5.
Closed-Loop Responsive Direct Brain Stimulation
Closed-loop responsive direct brain stimulation represents a new treatment paradigm for epilepsy. Seizures and epileptiform activity are sporadic and relatively infrequent, making an episodic treatment approach such as responsive stimulation an attractive option. Clinical observations and small studies have shown that electrographic seizures induced by electrical stimulation during brain mapping can be shortened or even terminated by immediately delivering a brief burst of electrical stimulation at the site of the discharge. Other small studies in persons being evaluated with intracranial electrodes for epilepsy surgery have indicated that automated stimulation to the seizure focus delivered by investigational external devices was well tolerated and seemed to suppress electrographic and perhaps clinical seizures.
The Responsive Neurostimulator System
The first implantable closed-loop responsive direct brain neurostimulator was approved in the United States in late 2013 as an adjunctive therapy in adults with medically uncontrolled partial onset seizures localized to 1 or 2 epileptogenic foci. The RNS System technology builds on decades-long experience in cardiac rhythm management devices. A neurostimulator continuously senses and monitors electrical activity from the brain (electrocorticographic activity) through 1 or 2 leads placed at the seizure focus and provides responsive electrical stimulation to the seizure focus when abnormal electrocorticographic activity prespecified by the physician occurs. The physician tailors detection and responsive stimulation for each patient according to the patient’s report of their clinical response as well as the quantitative electrographic data that the device provides.
Components of the RNS System are illustrated in Fig. 1 . The cranially seated neurostimulator is connected to 1 or 2 depth or cortical strip leads that are surgically placed in the brain at 1 or 2 seizure foci ( Fig. 2 ). Each lead contains 4 electrodes, each of which can be used for both sensing and stimulating. The physician programs detection and stimulation settings and retrieves and reviews data provided by the neurostimulator, such as battery measurements, lead impedances, programmed settings, time and date of detection and stimulations, and samples of electrocorticograms. Up to 6 minutes of electrocorticogram samples can be stored at any one time. In order to clear neurostimulator memory for additional electrocorticogram (ECoG) storage, the patient transfers data from the neurostimulator to a home-use remote monitor daily to weekly. Patient data acquired by the physician’s programmer or the patient’s remote monitor are then transmitted over the Internet to the Patient Data Management System (PDMS), where it is securely stored and available for physician review.
Which electrocorticogram samples are stored is customizable and may include times of day when the patient swipes a magnet over the neurostimulator to indicate a clinical seizure, spike and slow waves, rhythmic changes in frequency, or other electrocorticographic patterns that are typical of electrographic seizures in that patient. This quantitative data may be used by the physician as one means by which to assess the response to therapy ( Fig. 3 ).
The neurostimulator delivers current-controlled, charge-balanced biphasic pulses through any or all of the 8 electrodes. The physician programs stimulation frequency (1–333 Hz), current (0.5–12 mA), pulse-width (40–1000 μs), and duration (10–5000 ms). The most common stimulation settings in the clinical trials were a current of 1.5 to 3 mA, a pulse width of 160 μs, a stimulation burst duration of 100 to 200 ms, and a pulse frequency between 100 and 200 Hz. Usually, patients had 600 to 2000 detections and stimulations a day. At a typical burst duration of 100 to 200 ms, this is a total of less than 5 minutes of stimulation a day.
The stimulation pathway is selected by the physician. Any combination of the 8 electrodes can be used for stimulation, and the neurostimulator housing (case) can be incorporated into the pathway as an indifferent electrode. Once an initial detection and stimulation occur, the neurostimulator can provide an additional 4 stimulations if the detected activity does not resolve. Each of these stimulations contains 2 bursts that are individually programmed, providing the opportunity for several different types of stimulation within a given electrocorticographic discharge.
Clinical Experience
The RNS System is approved in the United States as an adjunctive therapy in individuals 18 years of age or older with partial onset seizures localized to 1 or 2 foci who are refractory to 2 or more antiepileptic medications and currently have frequent and disabling seizures (motor partial seizures, complex partial seizures, or secondarily generalized seizures).
Safety and efficacy of the RNS System was established in 3 clinical trials: a 2-year primarily open-label safety study (Feasibility study, N = 65), a 2-year randomized controlled trial (Pivotal study, N = 191), and a 7-year long-term extension study (Long-term Treatment study, N = 230) for patients completing the Feasibility or Pivotal studies. In total, 256 patients were implanted within the RNS System trials.
The multicenter double-blinded randomized and sham-stimulation controlled Pivotal study demonstrated the safety and effectiveness of the RNS System for its indicated use. Demographics of the patients in the randomized controlled trial are presented in Table 1 . Nearly one-third (32%) had prior therapeutic epilepsy surgery, and more than one-third (34%) had been treated with VNS.
| Characteristic | All Implanted (N = 191) | Active Stimulation (N = 97) | Sham Stimulation (N = 94) |
|---|---|---|---|
| Mean ± SD (min-max) or % (n) | |||
| Age (years) | 34.9 ± 11.6 (18–66) | 34.0 ± 11.5 (18–60) | 35.9 ± 11.6 (18–66) |
| Female | 48% (91) | 48% (47) | 47% (44) |
| Duration of epilepsy (years) | 20.5 ± 11.6 (2–57) | 20.0 ± 11.2 (2–57) | 21.0 ± 12.2 (2–54) |
| Number of AEDs at enrollment | 2.8 ± 1.2 (0–8) | 2.8 ± 1.3 (1–8) | 2.9 ± 1.1 (0–6) |
| Mean seizure frequency during preimplant period (seizures/mo) | 34.2 ± 61.9 (3–338) median = 9.7 | 33.5 ± 56.8 (3–295) median = 8.7 | 34.9 ± 67.1 (3–338) median = 11.6 |
| Seizure onset location–mesial temporal lobe only (vs other) | 50% (95) | 49% (48) | 50% (47) |
| Number of seizure foci–two (vs one) a | 55% (106) | 49% (48) | 62% (58) |
| Prior therapeutic surgery for epilepsy a | 32% (62) | 35% (34) | 30% (28) |
| Prior EEG monitoring with intracranial electrodes | 59% (113) | 65% (63) | 53% (50) |
| Prior VNS | 34% (64) | 31% (30) | 36% (34) |
Related posts:
Hippocampal Transections for Epilepsy
Minimally Invasive Neurosurgery for Epilepsy Using Stereotactic MRI Guidance
The Role of Stereotactic Laser Amygdalohippocampotomy in Mesial Temporal Lobe Epilepsy
The Stereo-Electroencephalography Methodology
Stereo-Encephalography Versus Subdural Electrodes for Seizure Localization
Neuromodulation for Epilepsy
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
