WASHINGTON -- Technologies to provide instant at-home detection of life-threatening seizures and that could allow epilepsy patients to monitor their own drug levels are on the horizon, researchers said here.
A , worn on the upper arm to alert patients and others nearby that a generalized tonic-clonic seizure is underway, is expected to conclude next month and an FDA marketing application will follow immediately, said Michael Girouard, RN, president and CEO of , the San Antonio company developing the device.
Action Points
- Note that this study was published as an abstract and presented at a conference. These data and conclusions should be considered to be preliminary until published in a peer-reviewed journal.
Results of a pilot study, reported at the American Epilepsy Society (AES) annual meeting, indicated that the electromyography (EMG) based detection algorithm successfully identified 21 of 22 generalized tonic-clonic seizures in patients who simultaneously underwent conventional EEG monitoring.
One false-positive alert was signalled in the 33 patients participating in the study, Girouard said. The patients also experienced 84 myoclonic-only seizures, 34 tonic-only seizures, 38 focal seizures with impairment, and four psychogenic non-epileptic spells.
The technology for measuring anti-epileptic drug (AED) levels in oral fluids was also the subject of a small pilot study reported at the AES meeting. , of Rice University in Houston, presented results suggesting that the two biggest scientific hurdles had been overcome.
First, it showed that levels of two important AEDs, phenytoin and phenobarbital, in oral fluids corresponded closely with blood levels. Second, the "nanochip" multiplex array for analyzing oral fluid samples performed as accurately as conventional gas chromatography-mass spectrometry in measuring drug levels.
Both approaches could make life easier for epilepsy patients, Girouard and Christodoulides said.
In the case of generalized tonic-clonic seizures, such episodes can be fatal. Patients prone to such events (sometimes called grand mal seizures) are at high risk for sudden unexplained death in epilepsy (SUDEP); they may also cause patients to suffocate in bed or suffer fatal falls or other injuries.
Girouard said that many patients and their families live in fear of such events and resort to extraordinary measures to avoid them -- not letting patients sleep alone (generalized seizures occur disproportionately at night) or to go out in public.
Drug-level monitoring as currently practiced is also a significant burden for patients. Because many AEDs have narrow therapeutic indices, patients must undergo frequent testing to ensure that blood levels have not strayed out of the safe-and-effective range. Moreover, results are not immediately available and often take many days, by which time they may be obsolete, Christodoulides said.
Near-Instant Drug Level Monitoring
The oral fluid technology relies on nanoscale beads, coated with drug-binding antibodies -- 2o in all, placed on a 4×5 grid in the current two-drug application. Twelve of the beads bind the drugs of interest; the others provide calibration and positive and negative controls. The quantitative signal is delivered via fluorescence and analyzed optically to provide a quick readout.
Similar technologies are already used in multiplex diagnostics for cardiac risk biomarkers, drugs of abuse, and allergen panels, Christodoulides said.
In the pilot study, oral fluid specimens were collected from 10 patients via swab and preprocessed to provide standardized samples. Blood samples were collected as well for testing with a conventional quantitative immunoassay.
The first set of tests compared drug levels measured in the two types of samples. Christodoulides reported very high correlation, with R2 values of 0.96 for phenobarbital and 0.93 for phenytoin.
The 10 oral fluid samples were also tested with gas chromatography-mass spectrometry as well as the nanochip array. The R2 value for correlation with phenobarbital was also very high (0.93); it was less so for phenytoin (0.48), with the nanochip providing low values relative to the standard assay for several samples.
Nevertheless, Christodoulides told AES meeting attendees that the technology was "promising" and would be developed further. The next step is to package the analytical instrumentation into a toaster-sized device with replaceable cards containing the nanochip, he said.
At a press briefing, he acknowledged that placing such devices in patients' homes may not be financially practical, since it would only be used intermittently. He said that, at least initially, it might be better used in physician offices where the rapid results could inform drug management on the spot.
Christodoulides also noted that the system had originally been called the Electronic Tongue, until late-night TV host Jay Leno made fun of it on "The Tonight Show." He and his team have now settled on "Bio-Nano-Chip" as a duller but less risible name.
Home Seizure Detection
Girouard, a former trauma/burn center nurse turned entrepreneur and inventor, said he conceived the idea that the physical shaking involved in a generalized tonic-clonic seizure would have a unique EMG signature. By simultaneously recording EEG and EMG signals in patients prone to such seizures, he and other experts were eventually able to pick out the pattern that distinguishes such seizures from others and from normal movements.
In the pilot study reported at the AES meeting, the 33 participants, age 14-64, underwent upper-arm EMG and video EEG scalp monitoring for an average of 42.4 hours (total 1,400) at the .
During this period, a total of 197 seizures were recorded with the video EEG, including 22 generalized tonic-clonic events. Using the EMG-analysis algorithm that had been developed from earlier studies, 21 of these events were detected within 30 seconds of the beginning of arm movement, Girouard reported.
The key to distinguishing the generalized tonic-clonic events from myoclonic-only and tonic-only seizures is that the generalized events show a sharp transition in the EMG signal from tonic to clonic.
He said he was particularly pleased that the system did not falsely alarm in any of the four psychogenic non-epileptic spells, because the EMG signals from those events also show a similar type of transition.
As configured in the ongoing confirmation trial, the detector is about the size of a cigarette pack, curved to conform to the upper arm. Patients wear it around the clock; it's sufficiently waterproof that they can shower with it, although it should not be immersed, Girouard said.
When it detects that a generalized seizure is underway, it emits a loud auditory alarm and also sends a signal to a "base station" (a small netbook-type computer) that also sounds a loud alarm. The base station can be configured to send additional alarm messages over the Internet to virtually anywhere.
Girouard told 51˶ that Brain Sentinel planned to file a 510(k) marketing application with the FDA in January, although he expects that the agency will require the firm to go through either the premarket approval or de novo classification process. Hence, he said, approval will not be quick.
Disclosures
Brain Sentinel funded the seizure detection device study. The study of the oral fluid system for monitoring drug levels was funded by the John Dunn Foundation; future development will have commercial funding.
No other potential conflicts of interest were reported.
Primary Source
American Epilepsy Society
Girouard M, et al "EMG-based seizure detector: interim results comparing a generalized tonic-clonic seizure detection algorithm to video-EEG recordings" AES 2013; Abstract 2.037.
Secondary Source
American Epilepsy Society
Kalamangalam G, et al "Towards home monitoring of anticonvulsant levels: a pilot study using Bio-Nano-Chip technology" AES 2013; Abstract A.09.