ARTICLE IN BRIEF
In a narrow vote, an FDA panel recommended approval of a deep brain stimulator for epilepsy, despite marginal improvements in a three-month trial.
A deep-brain stimulator for intractable epilepsy produced marginal improvement during a three-month clinical trial, but showed enough promise to win approval from the FDA Neurological Devices Panel by a narrow 7-5 vote in March.
“The hard-core statisticians on the panel voted against approval because they just looked at the numbers, and if you just look at the numbers, the results weren't that impressive,” said Harry Chugani, MD, who served on the panel. “But if you look at the total picture, these patients have failed to improve after drug treatment, after surgery, after vagal nerve stimulation. There's nothing left for them. If you got any kind of benefit for them, that's worth pursuing. That goes beyond looking at the numbers.”
Dr. Chugani admits he came close to voting against approval himself, but decided that the improvement in the patients over time hinted at benefits that didn't emerge clearly during the brief three-month trial.
“The last third of the three-month trial produced the strongest results,” said Dr. Chugani, professor of pediatrics and Rosen Family Chair in Tourette's Syndrome, Neurology and Radiology at Wayne State University in Detroit, MI. “What that means, I believe, is that it takes time for this treatment to work. If they ran the blind period for six or nine months, which is what I proposed, the latter half could have become highly significant.”
Although only 40.4 percent of participants experienced a reduction in seizures during the three-month placebo phase of the trial, compared to 14.5 percent in the placebo group, 56 percent of participants showed a reduction in seizures after 25 months, and two participants were seizure free for at least six months.
The panel recommended approval on the condition that the device come with proper warnings, and that follow-up monitoring studies be performed.
The trial enrolled 110 people 18 to 65 who had experienced at least six partial-onset seizures per month, but not more than 10 per day. They experienced a median of about 20 seizures per month. All had tried at least three anti-epileptic drugs that failed to control their seizures.
Subjects whose devices were turned on received stimulation for one minute, followed by five minutes of no stimulation. The initial results were skewed by one patient — an “outlier” — who experienced a total of 210 brief partial seizures over three days that coincided with the stimulation. When the voltage was turned down 20 percent, however, the stimulation appeared to work normally for him, according to Robert Fisher, MD, PhD, Maslah Saul Professor in the Department of Neurology and Neurological Sciences at Stanford University School of Medicine, and director of the Stanford Epilepsy Center.
Dr. Fisher, the lead author of an online March 17 paper in Epilepsia that describes the results, presented the data from the trial before the FDA panel. The patient known as the outlier also testified before the panel about the improvement he has experienced through DBS.
With the outlier's experience removed, the data achieved statistical significance, but not until the third month of the trial. And the response depended in part on where the seizures originated. Patients with seizures that originated in one or both temporal lobes experienced seizure reduction of 44.2 percent, while those with seizures originating in the frontal, parietal or occipital lobes “did not demonstrate significant differences in seizure reduction,” according to the authors of the Epilepsia paper. Patients with multifocal or diffuse seizure activity showed a 35 percent reduction, as opposed to the 14.1 percent reduction in the control group, “but with only eight and nine subjects, respectively, in each group this difference did not achieve significance.”
As with deep brain stimulation for other disorders, the mechanism by which the electrical stimulation produces benefit remains unknown, but researchers assume it disrupts the seizure circuit.
“I doubt that the stimulation involves simple inhibition or excitation,” said Dr. Fisher. “More likely, stimulation produces changes in neuronal networks to disrupt synchrony, seizure formation, and seizure spread, and these changes seem to increase over time.”
One of the subinvestigators involved in the trial believes the ongoing benefit involves a neuromodulatory effect that becomes more pronounced over time. “You might not see it in initial stages of treatment,” said Andres M. Kanner, MD, professor of neurological sciences at Rush Medical College, director of the Laboratory of Electroencephalography and Video-EEG-Telemetry, and associate director of the Section of Epilepsy and Clinical Neurophysiology at the Rush Epilepsy Center. “But the third year you see a significant number of people achieve a 50 percent reduction in seizures. This neuromodulatory effect may also enhance the response to anti-epileptic drugs, which may account for the improvement in seizure frequency.”
The device that received the approval of the FDA panel — the Activa PC — is a smaller and more highly programmable version of the Medtronic Intercept stimulator, which was actually used in the study. The Activa PC received FDA approval in 2009 for Parkinson disease.
Mohamad Z. Koubeissi, MD, assistant professor of neurology at the University Hospitals Case Medical Center of Case Western Reserve University, who co-authored a 2007 article on cerebellar and thalamic stimulation for epilepsy, said the overall results of the trial encouraging, but he was not surprised by the narrow FDA vote, given the fact that there were not meaningful differences in seizure reduction between those who had the device turned on and the control group. He agreed with others, however, that the fact that after two years of using the thalamic stimulation, more than half of all patients had reduced seizure frequency by 50 percent or more and 14 patients were seizure-free was “somehow promising.”
He added: “Whether electrical stimulation induces slow, long-term beneficial changes in the seizure circuitry can only be discerned with wider use and more studies.”