Even with a vast number of antiepileptic drugs (AEDs), many patients continue to experience disabling seizure. At the December annual meeting of the American Epilepsy Society (AES), investigators offered the latest data on efforts to develop more efficient ways of detecting, monitoring and preventing them.
Jose E. Cavazos, MD, PhD, chair of the AES Science Committee, and an associate professor of neurology, pharmacology and physiology at the University of Texas Health Science Center in San Antonio, said that advances in understanding brain circuits that underlie seizure activity — and modern technology to short-circuit abnormal electrical signals — are among the more exciting developments in the field. He selected as most noteworthy several papers from the meeting that deliver on that promise.
Of particular note, he said, were trials to test devices that deliver electrical activity to the brain to control intractable seizures in patients who have failed every available treatment. Results from NeuroPace's multicenter double-blind, randomized control trial of its RNS (responsive neural stimulation) system to treat intractable partial epilepsy in adults suggested that it works to identify a seizure in the making and delivers electrical stimulation to block the seizure activity.
The company, which began this latest safety and efficacy study in 2005, moved forward with the device after a favorable two-year safety study, according to Martha Morrell, MD, chief medical officer at NeuroPace and a clinical professor of neurology at Stanford University. As of May 2009, 191 patients from 31 study sites have had the device implanted. The neurostimulator is connected to leads at the focal site of the seizure activity and is programmed to deliver electrical stimulation only when the brain “senses” that a seizure is in the offing. It continually monitors brain wave activity.
Patients accepted into the study had an average of three disabling partial seizures a month despite a mix of AEDs. They had localized seizures in one or two regions. Going into the study the patients knew that they would undergo surgery to implant the device but that there was a 50 percent chance that the machine would be turned off so that investigators could compare its benefits in those who would receive the active stimulation.
Seizure frequency was monitored beginning two months after the device was implanted and extended for almost three months. After this study period, all subjects had the device turned on to receive stimulation. The study period continued for two years post-implant and patients are now in a five-year open label treatment trial.
According to Dr. Morrell, the RNS System helped reduce the frequency of the seizures for 191 patients, many of whom had refractory seizures after taking an average of 2.8 anti-seizure medications and having tried surgery. One third of them had been implanted with a vagus nerve stimulator and a third had previous epilepsy surgery. Sixteen percent had both. Nearly half of them had seizures originating in the mesial temporal structures of the brain, most bilaterally.
During the last two months of the three-month blinded evaluation, 96 people randomized to receive active stimulation had 29 percent fewer seizures compared to a 14 percent reduction for the 93 people in the sham stimulation group. In the long term, open label period of the trial, at least 12 weeks of data were available for 171 study participants; 47 percent of these subjects experienced a 50 percent or greater reduction in their seizure frequency based on their most recent 12 weeks of data, as compared to their baseline. Of the subjects completing the two-year follow-up, more than 50 percent had a 50 percent or greater reduction in seizures.
The trial results showed that the responsive neural stimulation device worked better than the sham arm of the study, Dr. Cavazos said. “This can stop seizures in eloquent cortex or in multiple areas of the brain that could not have been done with traditional surgical approaches.” The limitations, he added, are that neurologists and neurosurgeons need to know where the lesion is emanating from. This is a major breakthrough for epilepsy patients.”
NEUROIMAGING FOR SURGERY
Harry Chugani, MD, chief of the division of pediatric neurology and professor of pediatrics and neurology at Children's Hospital of Michigan and his colleagues have been testing the benefits of a PET radioactive tracer called 11C-alpha-methyl-L-Tryptophan (AMT) to identify abnormal tissue in patients with tuberous sclerosis complex. They found that the imaging tracer could identify the tuber that was causing the seizure. But what about other malformations associated with epilepsy? Could scanning help reveal something about the source of the seizure?
Dr. Chugani reviewed 30 cases of children with malformations of cortical development (MCD) who underwent cortical resection for intractable epilepsy between 1998 and 2008 to see whether AMT-PET abnormalities provided insight in finding the exact location for surgical resection. They had information from the AMT-PET scan and they also had tissue that was excised in the surgery. They also knew the outcomes, whether the children continued to have seizures following the surgery.
Dr. Chugani reported at the meeting that PET was useful in determining a specific kind of pathology that leads to seizures: those with cortical dysplasia with type 2 balloon cells, unusually large degenerated cells. Dr. Chugani found that those with cortical dysplasia diagnosed from an MRI scan or from a pathology exam of the tissue and who had a positive AMT-PET scan had greater success following surgery than those with negative MRI scans. The hope is that this PET scan could be used in conjunction with MRI or pathology to better predict the outcome in surgery for these young patients with intractable seizures.
“The bottom line is that molecular PET imaging more and more is able to pinpoint the epileptic focus and even the pathology,” said Dr. Chugani. “We have never been able to do that before we head into surgery. These clever scans give us a huge amount of information.”
“It is fascinating,” said Dr. Cavazos. “This non-invasive scan may help predict the children with cortical malformations who would have good surgical outcomes or not.”
Philippe Ryvlin, MD, and his neurosurgical colleagues at the Federal Institute of Neurosciences at the Neurological Hospital in Lyon, France, have analyzed clinical trials in patients with intractable epilepsy that included a placebo arm and found that the risk of a sudden and unexplained death in epilepsy (SUDEP) was substantially lower in those patients who were getting active medication in addition to the other medicines they had been taking before the study. These are people in whom medications have not worked.
Study patients who received the experimental medicine plus whatever other AEDs they were on had an incidence five times lower for SUDEP than those on a placebo dose who were also taking the same number of their routine medicines.
Dr. Cavazos said these results are critically important because they support the idea that being on multiple medicines — even if they are not fully controlling the frequency of seizures — could dramatically reduce the risk for sudden death.
“Treating with anti-convulsants can work to reduce SUDEP,” Dr. Cavazos said. “So what is the risk of not having surgery in patients who have focal intractable epilepsy? SUDEP. We will use this argument to tell people who are reluctant to add medications when it doesn't help reduce the frequency of their seizures that it may ultimately save their life.”
EEG ANALYSIS: SEIZURE DETECTION
Kevin Kelly, MD, J. Chris Sackellares, MD, and their colleagues from the University of Florida in Gainesville have been working on using scalp EEG recordings to predict seizures associated with temporal lobe epilepsy. Intracranial EEG recordings have shown that seizures are preceded by a measurable change in the region's electrical pattern but identifying such activity via the skull poses problems with the distance from the source, recording artifacts and rhythmic signal patterns due to normal physiological functions such as sleep, even chewing. A successful EEG recording must be able to separate normal EEG patterns from epileptic-related activities.
The scientists developed a mathematical formula to test the power of the scalp EEG test. The algorithm was tested over a two-day recording period in 51 patients with temporal lobe epilepsy. At the end of the two days, there were a total of 159 seizures. They evaluated the performance of the algorithm to see whether it could predict a seizure within 2.5 hours of the event. Among findings, the scalp recording worked and on average it issued a warning within an hour of the seizure. They concluded that scalp EEG “can serve as a reliable pattern for seizure prediction,” according to the abstract from the meeting. “
“The ability to warn of impending seizures from analysis of scalp EEG recordings could have broad clinical applications in routine in-patient monitoring units and ambulatory EEG recordings,” according to the study authors.
Added Dr. Cavazos: “If we could predict from the scalp it would be much safer and could be used more widely than intracranial EEG recordings. One day, people with epilepsy may be able to wear a hat that records the electrical activity inside the brain and alerts them that they are about to have a seizure.”