Major Finding: Seizure frequency declined by a mean of 29% during active stimulation with the device over the first 12 weeks, compared with a 14% reduction during sham activation.
Data Source: Multicenter, randomized, sham-controlled clinical trial of 191 patients with medically intractable, partial onset seizures.
Disclosures: Dr. Morrell is the chief medical officer of NeuroPace, which developed the system and funded the trial.
BOSTON — Patients with treatment-resistant epilepsy can reduce seizure frequency with the use of an implantable device that detects preseizure electrical activity and aborts seizures, the results of a multicenter, randomized trial suggest.
In 191 patients with medically intractable, partial onset seizures who were implanted with the neurostimulator, seizures declined by a mean of 29% during active stimulation with the device, compared with a 14% reduction during sham activation, Dr. Martha J. Morrell reported at the annual meeting of the American Epilepsy Society.
In the later, open-label phase of the study in which all patients received active stimulation, nearly half of the 171 patients for whom 12 weeks of data were available had at least a 50% reduction in seizure frequency relative to baseline, said Dr. Morrell, clinical professor of neurology at Stanford (Calif.) University and chief medical officer of NeuroPace, developer of the Responsive Neurostimulator System (RNS).
The cranially implanted RNS device differs from conventional, “open loop” brain stimulation technologies that involve the scheduled delivery of electrical stimulation to specific brain regions independent of brain activity.
“The RNS delivers stimulation in response to a detected event,” Dr. Morrell said. The “individualized and dynamic” treatment responds to patterns of brain activity specific to a patient's seizure pattern. The electrodes are implanted in epileptic regions of the brain and connected to the battery-powered neurostimulator embedded in the skull.
“The programming is done wirelessly by the physician via a laptop computer,” Dr. Morrell said. “It's highly modifiable in that the physician can view the patient's electrocorticographic activity in real-time and change the [signal-detection] criteria at any time based on individual patient characteristics.”
Up to two leads, each containing four electrodes, can be connected to the neurostimulator, so the system can monitor and stimulate two distinct epileptogenic zones independently, she noted.
Because the neurostimulation occurs in response to aberrant electrical activity, fewer electrical impulses are delivered than would occur with continuous stimulation. This diminishes the risk of treatment-related adverse events, Dr. Morrell explained.
In an initial feasibility study of 65 patients, the responsive neurostimulation system demonstrated excellent safety, tolerability, and preliminary evidence of efficacy, Dr. Morrell said. “There were no serious device-related adverse events, and stimulation-related symptoms experienced by several subjects were addressed by adjusting the stimulation settings.”
A minimum 50% reduction in seizure frequency was achieved in 43% of the patients with complex partial seizures and 35% of those with total disabling seizures (Neurotherapeutics 2008;5:68-74).
In the double-blind pivotal trial, the 191 patients were randomized to active or sham therapy. All were 18-70 years of age (median 35 years), and all had partial onset epilepsy localized to one or two foci and had failed at least two antiepileptic medications.
The patients were taking an average of three antiepileptic medications to attempt seizure control; 34% had been treated previously with vagus nerve stimulation, 33% had prior surgical resection, and 16% had been treated with both.
“These patients tended to be very ill. Most of them had epilepsy for more than 20 years, and many were having at least three seizures per 28-day period—often many more than that,” Dr. Morrell said.
Of the 191 patients implanted with the device, 50% had mesial temporal seizure onset, 42% had neocortical seizure onset, and 8% had both, Dr. Morrell said at a press briefing during the meeting.
An initial, 12-week period prior to system implantation, during which baseline seizure activity was collected, was followed by a 12-week blinded period when participants were randomly assigned to have the responsive stimulation activated or left inactive, she said.
At each of the 31 trial sites, the patients and one neurologist were blinded to stimulation status, while a separate neurologist programmed the devices in order to maintain the study blinding. The responsive stimulation was optimized in the treatment over the next 4 weeks, followed by 84 days of data collection. At the end of the blinded efficacy period, stimulation was activated for all of the study participants for 2 years.
There were no serious, unanticipated device-related adverse events during the trial, nor was there a difference between the two groups with respect to the rate of adverse events, including depression, memory impairment, and anxiety, Dr. Morrell reported.