Page 467 - ILAE_Lectures_2015

shown to reduce anhedonia in rats treated with kainic acid, lending support to the use of VNS
in depression6. However, the exact molecular mechanism by which VNS exerts its effect is
uncertain.

Some interesting insights into the action of VNS in humans have been explored with EEG
recording. A small study compared five patients who had responded well to VNS
implantation to five patients in whom VNS had not been beneficial. EEG recordings five
years after implantation showed that patients who had benefited had a decrease in gamma
band frequency desynchronisation7. More recently another small study has shown that
patients who respond to VNS therapy demonstrate a shift in EEG architecture towards a more
efficient configuration8 raising the possibility that chronic VNS offers some stabilisation of
neuronal networks. Studies have also begun to explore whether there are EEG signatures that
might better predict response to VNS9.

VNS  device and process

Originally only a single manufacturer (Cyberonics) had rights to produce the device but this
restriction has now expired and a number of devices are becoming available. In some
experimental animal models stimulating the right vagus nerve produces cardiac dysrhythmias
so VNS is only licensed for implantation on the left in humans. The generator is usually
implanted in the left upper chest with the electrodes placed around the left cervical vagus.
Afferent pain fibres may be activated, especially at higher levels of clinical stimulation,
producing discomfort in the throat, but even at normal therapeutic levels patients are usually
aware when the device activates due to a sensation in the throat.

An electrical test of the device is performed intra-operatively but the device is usually
activated some time post-operatively, often at the first outpatient follow-up. Continuous
electrical stimulation of the vagus nerve in animal models has been shown to produce fibrosis
and ultimately failure of the nerve, so stimulation is provided in an intermittent manner.
Typically the VNS device is initially set to provide 30 seconds’ stimulation every five
minutes. The device is programmed externally (output current, signal frequency, pulse width,
signal on- and off-times) and adjustments are made on the basis of tolerability of side effects
and clinical efficacy.

In addition to the continuing cycling on and off, it is possible to manually activate the device
by passing a magnet over the generator box. Patients or carers can use this when a seizure
starts, and in some the magnet seems to shorten or limit the extent of the attack. Commonly
the current delivered following magnet-induced activation is set slightly higher than the
baseline level. The magnet also provides a means for the patient to deactivate the device. If
the magnet is placed over the generator box and it remains there for more than a few seconds
the VNS switches off. When the magnet is subsequently removed, it reactivates at the
previous settings. This may be used by patients to investigate whether a symptom such as
cough is related to the device or to some unrelated cause. Similarly, the device can be
switched off for certain occasions when patients may wish, for example, to make a speech,
when they are keen to avoid any fluctuation in voice quality.

Regular follow-up is needed, with gradual current adjustment to achieve maximum benefit
in a similar way to adjustments of antiepileptic drug (AED) dose. Battery life, which depends
on output and magnet use, is likely to exceed six years even at higher output levels, after
which the pulse generator will need to be replaced. The device should be checked regularly
and an early replacement indicator (ERI) or ‘near end of service’ (NEOS) alert will warn the
clinician of impending battery exhaustion. A rechargeable battery is in development but is
not yet in mainstream use.