Seizure detection and automatic magnet mode performance study

In 1994 and 1997, Cyberonics* VNS Therapy was approved by KEMA and FDA,
respectively, as an adjunctive therapy for the treatment of epilepsy. VNS
Therapy is available both as a scheduled stimulation, for example, a 30-second
burst every 5 minutes, and as on-demand
stimulation, for example, when a magnet is introduced briefly over the
implanted device.
Some VNS patients find that magnet-initiated stimulation delivered during an
aura or at the beginning of a seizure serves to terminate or dissipate the
effects of the seizure. (Boon 2001, Morris 2003, Murphy 2003) However, it is
not always possible for patients to use the magnet
mode feature immediately before a seizure due to multiple reasons including
cognitive impairment, sleep, or the effects of the seizure itself. Cyberonics
has developed an algorithm to identify seizure onset based on cardiac rhythm
disturbances. The advantage of such an algorithm
would not only be the potential to deliver stimulation at the onset of a
seizure, but also to provide a consistent assessment of the patient*s seizures.
Patient seizure diaries are notoriously inaccurate (Poochikian-Sarkissian
2009), having been documented as inaccurate due to many
different issues including cognitive impairment, memory alteration due to the
seizure itself, and not taking time to record the seizures every day as they
occur. As patient diaries are frequently used when adjusting patients*
treatment including titrating antiepileptic drug (AED) doses, an
accurate assessment of patients* seizure activity can be an invaluable tool for
the neurologist.
Epileptic seizures are often accompanied by changes in autonomic function
including heart rate, respiratory rate, gastrointestinal/abdominal function,
cutaneous appearance, and secretory function, among others. Research supports
the occurrence of ictal tachycardia using
electroencephalogram (EEG) and electrocardiogram (ECG) recordings in 64-93% of
complex partial seizures, primarily of temporal lobe origin (Van Buren 1967,
Marshall 1983, Keilson 1989, Smith 1989, O*Donovan 1995, Schernthaner 1999,
Leutmezer 2003, DiGennaro 2004).
Time frequency analysis of R-R intervals revealed that autonomic activation may
precede clinical seizure onset by several minutes (Novak, 1999). Heart rate
changes were observed by Schernthanaer et al. (1999) by an average of 18.7
seconds prior to EEG seizure onset on scalp-
EEG in 76.1% of focal seizures. Similarly, sinus tachycardia preceded EEG
seizure onset in 75.9% of all seizures by 0.7-49.3 seconds in patients with
focal epilepsy in results reported by Leutmezer et al. (2003). The time lag
between changes in heart rate and seizure onset was longer in patients with
mesial temporal lobe epilepsy compared to those with extratemporal epilepsy
(Leutmezer, 2003). Presence or absence of heart rate changes are dependent upon
the site of seizure origin as well as the volume of cerebral structures that
were recruited into the seizure (Epstein, 1992).
In a study including both focal and generalized epilepsies, although increases
in heart rate were reported in 73% of seizures, the rate increase preceded
electrographic and clinical onset of seizure in only 23% of seizures (Zijlmans,
2002). Moreover, heart rate increases are observed
earlier in mesial versus lateral temporal lobe seizures. Mesial seizures had
ictal heart rate changes before both EEG and clinical ictal onset, whereas in
lateral temporal lobe seizures, the start of the heart rate increase coincided
with the EEG ictal onset (DiGennaro, 2004). Finally, it
has been reported that small myoclonic seizures may not manifest in heart rate
changes (van Elmpt, 2006). This indicates that differences in ECG patterns may
exist by origin of the epileptic focus, seizure type, as well as the volume of
cerebral structures that were recruited into the
seizure. Return to baseline may also be complicated by a number of swings into
bradycardia (van Elmpt, 2006). Note additional information regarding
preclinical device testing is available in the E-36 Investigator Brochure.
Cyberonics has developed an algorithm designed to detect changes in heart rate
that may be indicative of a seizure. The algorithm has been implemented in a
VNS device capable of detecting heart rate, and when a potential seizure is
detected, the device is capable of delivering
VNS stimulation, much the same way that VNS devices currently deliver
stimulation when a magnet is passed over the device.

The primary study objective is to evaluate the Model 106 VNS Therapy® System
cardiac-based seizure detection and Automated Magnet Mode (AMM) stimulation in
epilepsy patients. These two new features will be evaluated using continuous
observational video electroencephalogram (vEEG) and electrocardiogram (ECG)
data in ictal and non-ictal events during an Epilepsy Monitoring Unit (EMU)
stay.
Secondary objectives include:
1. Validation of cardiac R-wave detection by the Model 106 device.
2. Assess stimulation-related adverse event rates to outline the tolerability
profile of Model 106 device.
3. Characterization of latency period.
4. Characterization of post-stimulation heart rate changes.
5. Evaluation of human factors and usability of the system.
6. Assess changes from baseline in seizure frequency.
7. Assess changes from baseline in seizure severity.
8. Assess changes from baseline in seizure duration.
9. Assess changes from baseline in seizure intensity.
10. Assess changes from baseline in post-ictal duration.
11. Assess changes from baseline in quality of life.

Prospective, observational, unblinded, multi-site study designed to collect
data on up to 35 patients implanted with a Model 106 VNS Therapy System from
baseline through an EMU stay of up to 5 days. Good Clinical Practices shall be
followed during the study including the Declaration of Helsinki, ISO
14155:2011, WMO, and MEDDEV 2.7.1. Note that each patient undergoes activation
of the Seizure Detection Algorithm (SDA)
feature during implantation of the device for intra-operative evaluation
(detection only) and throughout the EMU stay (detection and automated
stimulation). After the EMU stay, patients continue follow-up for safety until
final regulatory approval of the product.

Not applicable

Surgery and treatment with VNS Therapy have possible risks, complications, and
side effects. Most of these are known from previous clinical studies. However,
there may be side effects from surgery and stimulation by the VNS Therapy
System that are unknown at this time.

Stimulation may cause some health problems to get worse. Asthma and bronchitis
may get worse with stimulation. Swallowing problems, hoarseness, and lung
conditions also may get worse with stimulation. Stimulation may also irritate
the windpipe (larynx), especially in case of smokers. It is not known how safe
or effective the VNS Therapy System is in patients with:

* History of previous therapeutic brain surgery or brain injury
* Progressive neurological diseases other than epilepsy
* Irregular heartbeats or other heart abnormalities
* History of dysautonomias (problems caused by the autonomic nervous system)
* History of lung diseases or disorders, including shortness of breath and
asthma
* History of ulcers (gastric, duodenal, or other)
* History of fainting
* Other forms of brain stimulation
* Preexisting hoarseness
* Difficulty swallowing
* Shortness of breath
* Episodes where breathing stops for short periods of time while sleeping.
Because of the vagus nerve*s effect on the stomach, duodenal or gastric ulcers
could develop. If the patient has stomach problems already, treatment could
make them worse. However, these stomach problems have not yet been reported in
clinical studies of VNS Therapy.

Surgery that requires general anesthesia may carry the risk of death. Surgery
to implant VNS Therapy requires local, regional, or general anesthetic. Rarely,
heart rate slowing occurs during a short stimulation of the vagus nerve that is
done in the operating room. All patients have recovered from this slowing of
heart rate. Vagus nerve damage, if it occurs, can cause vocal cord paralysis.
It can also cause Horner*s syndrome. Horner*s syndrome has symptoms that
involve the eye and eyelid, and an absence of sweating on one side of the face.
Nerve damage during surgery may also occur. It could result in permanent
hoarseness, difficulty in swallowing, or gastrointestinal problems

Mechanical problems:

The following problems could occur:
* The Pulse Generator could malfunction and not deliver stimulation or deliver
too much stimulation.
* The Pulse Generator or Bipolar Lead could move or push through the skin.
* The Bipolar Lead could break or become corroded, disconnected, or dislodged.
* The battery in the Pulse Generator could become depleted (reach its
end-of-life).
Any mechanical problems could require surgery. Surgery could cause the nerve to
swell because of the Lead placement around the vagus nerve. If the Pulse
Generator and Lead are removed, it is possible the nerve may swell due to the
handling of the nerve. Moving the VNS Therapy System could cause damage to the
vagus nerve or other parts of the body.
A malfunction of the Pulse Generator could damage the vagus nerve. This
malfunction could lead to permanent hoarseness or other complications. A
malfunction could cause the Pulse Generator battery to last a shorter time
period than expected.

In addition, there are risks specific to the Model 106 VNS Therapy System,
which is designed to detect an increase in heart rate associated with some
seizures. These include:
* The device might stimulate too often because it is not accurately sensing the
heart rate. The device is programmed to ignore heart rates that are too high or
too unlikely to be real, and therefore, will not trigger the stimulation. To
further guard against excessive stimulation, each stimulation is followed by a
waiting period that lasts at least as long as the period of stimulation. Also,
the device cannot be programmed to give stimulation at a higher frequency than
what is given with the currently approved VNS Therapy System.
* Stimulation can sometimes cause a temporary increase in heart rate which
could trigger AMM. The device is programmed to wait for a minimum period after
stimulation before it can again sense increases in heart rate and trigger
another stimulation.
* If the device is set to be too sensitive, it might trigger stimulation too
often. Cyberonics has instructed the physician on how to choose initial
settings to avoid having the device be too sensitive.
* If the device is set to have too little sensitivity, it might not trigger
stimulation often enough. Cyberonics has instructed the physician on how to
choose initial settings to avoid having the device have too little
sensitivity. In addition, if the patient begins to experience a seizure and
the device does not trigger stimulation, he/she can manually trigger the
stimulation by passing the magnet over the generator (Magnet Mode
Stimulation).
* If the heart rate does not increase at the beginning of a seizure, or if the
device does not detect an increase in the heart rate at the beginning of a
seizure, the patient will not receive AMM. He/she will still be receiving
Normal Mode VNS therapy at the programmed settings and will also be able to
initiate Magnet Mode stimulation by passing the magnet over the generator.
* MRI or other sources of electromagnetic interference might make the device
trigger unwanted stimulation. The patient should avoid MRI and other sources
of electromagnetic interference. The physician has been instructed to
re-program the device for medical procedures that may involve MRI or may
involve other equipment that could involve electromagnetic interference.
* With standard VNS therapy, stimulation always happens at regular intervals.
However with AMM, the stimulation does not happen at regular intervals, and can
therefore seem to be unpredictable.


BENEFITS:

Each patient has different results from using the magnet. Some patients say
that the magnet stops all or most seizures, shortens them, or lessens their
intensity or their recovery period. For other people, the magnet has little or
no effect. This study is not powered to confirm any benefits of magnet mode;
however, it has been shown that for some patients, magnet use may terminate or
shorten seizures or post-ictal symptoms. (Morris, 2003)
Many patients and caregivers desire to use the magnet mode to trigger
stimulation at the beginning of an aura or seizure but are unable to do so, for
example because of the patient*s physical or cognitive disabilities or because
the seizure occurs during the patient*s or caregiver*s
sleep. The potential benefit of the Model 106 VNS Therapy System is that the
AMM feature has the ability to detect seizures and respond automatically with
VNS stimulation.