1

Introduction

Epilepsy, a disorder characterized by recurrent seizures, is thought to be present in 0.5%–2% of the world’s population ; moreover, seizures are the most common neurologic condition encountered in the pediatric population . Vagus nerve stimulation, first investigated in 1938 and subsequently studied as a potential therapy for epilepsy . In 1985, Zabara et al. reported that electrical stimulation from the vagus nerve produces inhibition of the neural processes, which can alter brain electrical activity and terminate seizures in dogs. Multiple controlled trials have been implemented to investigate vagal nerve stimulator (VNS) safety and efficacy for seizure control, in patients with intractable epilepsy, and all of which showed positive results indicating efficacy and safety of VNS. The American Food and Drug Administration (FDA) approved the use of VNS in 1997 as an adjunctive nonpharmacologic symptomatic treatment option for refractory epilepsy for adults and adolescents over 12 years . VNS was later on approved by the FDA for the treatment of severe, recurrent unipolar and bipolar depression in 2005 .

Despite the fact that initial approval of VNS in the USA was only for adults and adolescents with partial epilepsy, children and patients with generalized epilepsy have also benefited significantly from VNS implantation . It was realized that many non-seizure parameters (for example, behavior, school performance, and mood) are more improved with earlier seizure control and the minimization of adverse drug effects. This led to a paradigm shift toward an earlier identification of pharmacologically resistant epilepsy and patients who would benefit from surgical intervention. This evolving conceptualization of epilepsy treatment in children has led to an increase in the use of VNS in children younger than 12 years of age . VNS has been found to be well tolerated and effective as add-on therapy for refractory seizures in children of all ages. Response is even more favorable in the younger group (< 12 years at implantation) .

Physiologically, the vagus nerve is composed of both afferent and efferent fibers. The vagus relays afferent impulses from the thoracic and abdominal visceral organs to the cell bodies of the vagal afferent sensory component in the nodose ganglion, which relays this input to the nucleus tractus solitarius. This nucleus has projections to many higher cerebral centers, including the reticular formation in the medulla, locus ceruleus, parabrachial nucleus, amygdala, hypothalamus, thalamus, and the dorsal motor nucleus of the vagus . These central projections represent important neuronal pathways that most likely play a vital role in VNS therapy . The locus ceruleus is the major norepinephrine secreting region of the brain; VNS activates the locus ceruleus and releases norepinephrine, which in turn increases seizure threshold, thus modulating seizure control . The release of norepinephrine also may play a role in mood regulation. In addition, the locus ceruleus and parabrachial nucleus have efferent connections to the amygdala, which controls mood and emotion.

Anatomically, branches from the right vagus nerve innervate the sinoatrial node while branches from the left vagus nerve innervate the atrioventricular (AV) node . Right-sided VNS caused more cardiac slowing than left-sided VNS in a canine model . To avoid cardiac slowing, the VNS electrode array is almost always placed surrounding the left vagus nerve.

Technically, a VNS is made of implantable components and two peripheral components. The implantable components are a programmable pulse generator, subcutaneous connecting wire and a bipolar electrode lead. VNS pulse generator is a pacemaker-like device that sends mild intermittent electrical impulses via the connecting wire through the lead to the vagus nerve, which then sends signals to the brain. The peripheral components are a computer-connected programming wand, which is used by neurologists to program the VNS trans-cutaneously, and a strong handheld magnetic piece that is used by the patient or the caregiver to activate the device on demand in case the patient senses a seizure is about to occur. This handheld magnetic piece can also be used to temporarily suspend activity of the VNS in patients in whom the VNS has created significant side-effects.

Functionally, the VNS mechanism of action is not yet well understood, but there are three major theories proposed to explain the effect of the VNS on seizure activity. The first theory states that the VNS might cause stimulation of the nucleus of the solitary tract, resulting in an alteration in the autonomic nervous system and abortion of seizure activity. The second theory suggests that the VNS alters the activity of the brainstem reticular system, thereby enhancing the level of arousal and inhibiting aberrant cortical activity . Finally, the third theory proposes that the VNS causes stimulation of the noradrenergic nuclei leading to disruption of seizure activity .

Surgically, the majority of VNS implantation procedures have been performed by neurosurgeons and only (5%) of the procedures have been performed by otolaryngologist although there are reports describing them as well trained for the procedure . This is despite the fact that it would be easier to follow the patients up and track them since the commonest post-operative side effects will need an otolaryngologist to assess them, as the commonest side effect is hoarseness of voice caused by stimulating laryngeal motor fibers, by the electrical impulses generated by the VNS, through the recurrent laryngeal nerve . The VNS can also cause vocal cord dysfunction, which typically presents as an immobile fold in the paramedian or median position, with higher stimulation parameter .

The aim of this study is to review our results of VNS implantation at the otolaryngology department of a major international referral university hospital in term of efficacy, complications and side effects.

2

Methodology

Thirty patients (21 males and 9 females) were VNS-implanted over a 7-year period. Complete pre- and post-VNS surgical data were available for all patients, neurological data of seizure frequency, severity and other neurological assessments were obtained, with a minimum post implantation follow up of 2-year duration. The median age at implantation was 14 years (ranging from 3 to 25 years) and the mean duration of epilepsy was 6 years.

All patients were referred by the neurology service to our ENT clinic for implantation due to their intractable epilepsy despite treatment with at least 3 anti-epileptic drugs. The list of vagal nerve stimulator implantees was obtained from the institutional database. Medical records were reviewed for operative note, progress notes, in-hospital stay duration, and seizure frequency diaries of both pre- and post-implantation periods. All devices were implanted on the left side, the surgical procedure involved two incisions for implanting the device ( Fig. 1 ).

The surgical procedure incisions for implanting the device.The first incision is a left-sided, paramedian, horizontal incision at the level of the cricoid cartilage over the sternocleidomastoid muscle. The second incision was made vertically in the left anterior axillary line.

The first incision was a left-sided, paramedian, horizontal incision that was made at the level of the cricoid cartilage over the sternocleidomastoid muscle. The platysma was opened and the plane medial to the sternocleidomastoid muscle was created, exposing the carotid sheath. The sheath was opened, the jugular vein was retracted laterally, and the vagus nerve was usually apparent in the groove between the jugular vein and carotid artery. Careful dissection was performed to expose 2–3 cm of the vagus nerve. The second incision was made vertically in the left anterior axillary line. A pocket was created just under the skin of the chest above the fascia of the pectoralis major muscle for the pulse generator (~ 5 cm in diameter, determined by the size of the model).

The lead was tunneled subcutaneously between the two incisions. The diameter of the vagus nerve was measured with a caliper to determine if a 2 or 3 mm coiled electrode should be used. Then the helical electrode was encircled around the exposed vagal nerve, and the lead was attached to the pulse generator.

Before the incisions were closed, intra-operative electrical impedance testing was done trans-cutaneously by means of a handheld programming wand to ensure the integrity of the implant. At the beginning of the test, the anesthesia team was notified as cardiac arrhythmia and asystole had been reported during such test (0.1%) . After successful testing, the pulse generator was made inactive for a period of 2 weeks post-operatively, allowing wound healing before its clinical use. A mean surgical duration of 80 min was needed for the procedure.

Most patients were discharged on day 1 post-operatively, sent home with one week course of oral amoxicillin-clavulanate (1 g, BID), and care providers were educated regarding wound care.

Post-operative follow-up was performed in both the otolaryngology clinic and at the referring neurology clinic. Patients were scheduled for follow-up with the otolaryngologist on the 7th postoperative day for wound check and suture removal, and then the 3rd, 6th, and 12th months after activation of the VNS by the neurologist. Patients were followed up by the neurologist 2 weeks post-VNS implantation for device activation. Initial settings were: output current 0.25 mA, frequency 30 Hz, pulse width of 250 μs, stimulation period 30 s at five minute intervals. Patients were followed up every 2 to 4 weeks for optimizing the amplitude and type of current delivered by the generator, output current was increased by 0.25 mA every visit to target output of 1.75–2.25 mA, according to the patient response and tolerance.

Minimum neurological follow up for all patients was 2 years. The McHugh VNS specific outcome scale was utilized for assessment of seizure frequency and severity after implantation . The frequency of emergency room visits, ward admissions, and intensive care unit admissions pre- and post-VNS implantation were obtained from patients’ medical records and from caregivers.

The study included a fiberoptic laryngoscopy assessment of vocal cords, Voice Handicap Index-10 (VHI-10) questionnaire , to identify the severity of the voice changes. The ten questions were answered on a scale of 0 to 4, resulting in a range of scores from 0 to 40 for each patient. In addition to obtaining the Maximum Phonation Time (MPT) which was also measured by asking the patients to pronounce a prolonged /α/ at a typical speaking volume as long as possible after taking a deep inspiration, for all feasible pre-VNS activation implantees.

After six months of VNS activation, patients who reported voice changes, cough, and shortness of breath Post-VNS-implant activation, vocal cord motion re-assessment using a fiberoptic laryngoscopy was done as well as repeating the (VHI-10) questionnaire and repeating the (MPT). The results before and 6 months after the VNS activation were compared.

2

Methodology

Thirty patients (21 males and 9 females) were VNS-implanted over a 7-year period. Complete pre- and post-VNS surgical data were available for all patients, neurological data of seizure frequency, severity and other neurological assessments were obtained, with a minimum post implantation follow up of 2-year duration. The median age at implantation was 14 years (ranging from 3 to 25 years) and the mean duration of epilepsy was 6 years.

All patients were referred by the neurology service to our ENT clinic for implantation due to their intractable epilepsy despite treatment with at least 3 anti-epileptic drugs. The list of vagal nerve stimulator implantees was obtained from the institutional database. Medical records were reviewed for operative note, progress notes, in-hospital stay duration, and seizure frequency diaries of both pre- and post-implantation periods. All devices were implanted on the left side, the surgical procedure involved two incisions for implanting the device ( Fig. 1 ).

The surgical procedure incisions for implanting the device.The first incision is a left-sided, paramedian, horizontal incision at the level of the cricoid cartilage over the sternocleidomastoid muscle. The second incision was made vertically in the left anterior axillary line.

The first incision was a left-sided, paramedian, horizontal incision that was made at the level of the cricoid cartilage over the sternocleidomastoid muscle. The platysma was opened and the plane medial to the sternocleidomastoid muscle was created, exposing the carotid sheath. The sheath was opened, the jugular vein was retracted laterally, and the vagus nerve was usually apparent in the groove between the jugular vein and carotid artery. Careful dissection was performed to expose 2–3 cm of the vagus nerve. The second incision was made vertically in the left anterior axillary line. A pocket was created just under the skin of the chest above the fascia of the pectoralis major muscle for the pulse generator (~ 5 cm in diameter, determined by the size of the model).

The lead was tunneled subcutaneously between the two incisions. The diameter of the vagus nerve was measured with a caliper to determine if a 2 or 3 mm coiled electrode should be used. Then the helical electrode was encircled around the exposed vagal nerve, and the lead was attached to the pulse generator.

Before the incisions were closed, intra-operative electrical impedance testing was done trans-cutaneously by means of a handheld programming wand to ensure the integrity of the implant. At the beginning of the test, the anesthesia team was notified as cardiac arrhythmia and asystole had been reported during such test (0.1%) . After successful testing, the pulse generator was made inactive for a period of 2 weeks post-operatively, allowing wound healing before its clinical use. A mean surgical duration of 80 min was needed for the procedure.

Most patients were discharged on day 1 post-operatively, sent home with one week course of oral amoxicillin-clavulanate (1 g, BID), and care providers were educated regarding wound care.

Post-operative follow-up was performed in both the otolaryngology clinic and at the referring neurology clinic. Patients were scheduled for follow-up with the otolaryngologist on the 7th postoperative day for wound check and suture removal, and then the 3rd, 6th, and 12th months after activation of the VNS by the neurologist. Patients were followed up by the neurologist 2 weeks post-VNS implantation for device activation. Initial settings were: output current 0.25 mA, frequency 30 Hz, pulse width of 250 μs, stimulation period 30 s at five minute intervals. Patients were followed up every 2 to 4 weeks for optimizing the amplitude and type of current delivered by the generator, output current was increased by 0.25 mA every visit to target output of 1.75–2.25 mA, according to the patient response and tolerance.

Minimum neurological follow up for all patients was 2 years. The McHugh VNS specific outcome scale was utilized for assessment of seizure frequency and severity after implantation . The frequency of emergency room visits, ward admissions, and intensive care unit admissions pre- and post-VNS implantation were obtained from patients’ medical records and from caregivers.

The study included a fiberoptic laryngoscopy assessment of vocal cords, Voice Handicap Index-10 (VHI-10) questionnaire , to identify the severity of the voice changes. The ten questions were answered on a scale of 0 to 4, resulting in a range of scores from 0 to 40 for each patient. In addition to obtaining the Maximum Phonation Time (MPT) which was also measured by asking the patients to pronounce a prolonged /α/ at a typical speaking volume as long as possible after taking a deep inspiration, for all feasible pre-VNS activation implantees.

After six months of VNS activation, patients who reported voice changes, cough, and shortness of breath Post-VNS-implant activation, vocal cord motion re-assessment using a fiberoptic laryngoscopy was done as well as repeating the (VHI-10) questionnaire and repeating the (MPT). The results before and 6 months after the VNS activation were compared.