Anesthetic Management of a Patient With a Vagal Nerve Stimulator

Abstract

Vagal nerve stimulation (VNS) is an established adjunctive treatment for patients with refractory epilepsy. VNS is effective in many cases, but few patients achieve complete elimination of seizures. Furthermore, VNS can cause respiratory complications, such as obstructive sleep apnea. This report describes the successful anesthetic management of a 28-year-old woman with a VNS device who underwent dental treatment under general anesthesia. She was morbidly obese and had undergone placement of a VNS device secondary to drug-resistant epilepsy 2 years prior but continued to experience daily epileptic seizures. Because of concerns about the risk of perioperative epileptic seizures and apneic events, use of the dedicated VNS device magnet was planned if such complications occurred. Total intravenous anesthesia was induced with propofol and remifentanil and a bispectral index sensor was used to help monitor brain wave activity for evidence of seizures along with the depth of anesthesia. Postoperatively, the patient received positional therapy and supplemental oxygen while being closely monitored in recovery. The anesthetic course was completed uneventfully without need of the VNS magnet. A thorough understanding of the mechanics of a VNS device, including proper use of the VNS magnet, is critical for an anesthesiologist during the perioperative period.

Vagal nerve stimulation (VNS) has been available for the treatment of refractory epilepsy since 1994 in Europe and since 1997 in the United States.¹ VNS relieves the frequency and extent of an epileptic seizure by intermittently applying electrical stimulation along a wire connecting an implantable pulse generator to the left cervical vagal nerve. There is substantial evidence in the current literature supporting the effectiveness of VNS in providing relief from seizures. However, 10–30% of patients fail to respond to treatment with VNS.² Furthermore, it has been shown that VNS not only is associated with respiratory complications, but also exacerbates obstructive sleep apnea (OSA).^3–5

The issues involved in anesthetic management of patients with a VNS device are refractory epilepsy as a primary disease and respiratory complications during the perioperative period. However, anesthetic management of these patients has not been described. In this report, the successful anesthetic management of a patient with a VNS device is described.

CASE REPORT

The patient was a 28-year-old woman with a height of 164 cm, a weight of 105 kg, and a body mass index of 39 kg/m². She was diagnosed with tuberous sclerosis at the age of 6 months and developed epilepsy with tonic seizures. She had been prescribed anticonvulsants since childhood and was currently taking valproate, zonisamide, and topiramate. Because of persistent daily epileptic seizures, she underwent placement of a VNS device under general anesthesia at 26 years of age, after which the frequency of her seizures initially decreased to once every few months. However, her epilepsy gradually worsened, and she was again experiencing daily epileptic seizures when she presented for dental treatment.

She was intellectually disabled, and therefore was scheduled to undergo restorative dental treatment because of dental caries, periodontal scaling, and a single extraction under general anesthesia because of her inability to receive treatment without anesthesia. Her laboratory data, including a basic metabolic panel and complete blood count, were within normal limits, and there were no significant abnormal findings on the preoperative chest radiograph or electrocardiogram. She was morbidly obese and, according to her mother, did not have a definitive diagnosis of OSA, but snored frequently. Preoperative physical assessment confirmed no impairment or restriction to her maximum interincisal distance, but her Mallampati classification could not be determined because of lack of compliance.

She took her 3 daily prescribed anticonvulsants on the morning of the surgery and no other premedication was administered. Standard anesthesia monitors (pulse oximetry, standard electrocardiography, bispectral index [BIS] monitoring, and noninvasive blood pressure cuff) were applied upon arrival to the operating room. A 22-gauge intravenous catheter was placed in the dorsum of the left hand. General anesthesia was induced with remifentanil (0.2 μg/kg/min, based on an ideal body weight of 59 kg) and propofol (3 μg/mL, the predicted effective site concentration) using a target-controlled infusion. Rocuronium (0.9 mg/kg, total dose 90 mg) was administered to help facilitate ease of nasotracheal intubation. Because of concerns about the potential difficulties of mask ventilation and nasotracheal intubation due to the patient's morbid obesity, a properly sized oropharyngeal airway and flexible fiberoptic scope were prepared for potential use. However, these devices were unnecessary, as mask ventilation and subsequent nasotracheal intubation (using a 7.0-mm ID Ivory nasal endotracheal tube) were successfully performed without difficulty. Anesthetic depth and brain wave activity were monitored throughout the case using the BIS sensor. Anesthesia was maintained using total intravenous anesthesia with propofol and remifentanil. Dosing adjustments were made as appropriate for propofol and remifentanil infusions (ranges of 2–3 μg/mL [target-controlled infusion] and 0.15–0.2 μg/kg/min) respectively. No other neuromuscular blocking agent was administered. Prior to the start of the surgical procedure, local anesthesia was delivered using 2% lidocaine (36 mg) with 1:80,000 epinephrine (0.0225 mg).

Use of ultrasonic equipment is contraindicated in patients with a VNS device because it may cause failure or heating of the VNS pulse generator.⁶ Therefore, air and hand scalers were used in this patient, rather than an ultrasonic scaler. A dental polymerization light irradiator was used during the restorative treatment of her dental caries, as it has been reported to have no impact on the VNS device.⁶

At the end of surgery, neuromuscular blockade was reversed with sugammadex (∼2.5 mg/kg, total dose 250 mg) and the patient demonstrated adequate return of neuromuscular function. Cough reflex was appreciated following intratracheal suctioning. The patient was extubated without complications after confirmation of spontaneous ventilation with appropriate tidal volumes. The operating time was 1 hour, 30 minutes, and the total anesthesia time was 2 hours, 15 minutes. After completion of anesthesia, the patient was transferred to the recovery room and placed in the lateral decubitus position to help reduce the risk of respiratory complications associated with a VNS device. Oxygen was administered at a rate of 4 L/min via a face mask with close monitoring by the recovery room staff. The patient's subsequent recovery was uneventful, and she was discharged from the hospital approximately 3 hours after entry into the recovery room without any noted seizure activity.

DISCUSSION

The issues involved in the anesthetic management of patients with a VNS device are refractory epilepsy as a primary disease and respiratory complications during the perioperative period. Perioperative management of these patients first requires consideration of the refractory epilepsy. The VNS device consists of a pulse generator, a lead wire, and an electrode wrapped around the left vagal nerve (Figure 1). Intermittent electrical stimulation of the left vagal nerve via the lead from the pulse generator attenuates epileptic seizures. The dedicated magnet for the VNS device can be used either to deliver a burst of vagal stimulation by placing the magnet transiently over the generator or to temporarily inhibit output by leaving the magnet in place. The design for most VNS magnets allows patients to easily carry it with them, often around the wrist, similar to a bracelet or watch (Figure 2). Following the onset of a seizure, application of the magnet by waving it over the generator for ∼1 second activates the VNS device, thereby helping to attenuate the seizure. Placement of the magnet over the device for more than 65 seconds temporarily inhibits the VNS while the magnet remains in place. The VNS will return to its normal programmed mode once the magnet is removed.³ The therapeutic effect of a VNS device has generally been found to gradually increase when treatment is continued for 2–3 years.² The frequency of seizures is often reduced by more than 50% in 35–45% of patients with refractory epilepsy.⁷ However, although VNS is effective for many patients, efficacy is not guaranteed, and approximately 10–30% of patients fail to respond at all to treatment with these devices.²

Figure 1. .

Vagal nerve stimulator seen on a chest radiograph. Black arrow indicates electrode; triangle, lead wire; and white arrow, pulse generator.

Figure 2. .

Dedicated magnet for the vagal nerve stimulator.

Initially following placement of the VNS device, this patient experienced a noted decrease in epileptic seizure frequency, with a reduction down to 1 seizure every few months. However, that initial benefit was lost, and the severity of her epilepsy gradually worsened over time, culminating with multiple epileptic seizures occurring daily by the time she presented for consultation. As such, the anesthetic plan was carefully developed, taking into consideration the patient's refractory epilepsy.

Intravenous propofol was selected for induction of anesthesia because the patient was uncooperative with dental treatment but tolerated placement of an intravenous (IV) cannula. If the patient had been unable to cooperate for placement of the IV, induction with volatile anesthetic agents via face mask would have been used as the primary alternative. If that route had been utilized, conversion from inhalational anesthetics to intravenous propofol would have occurred following successful insertion of the IV catheter.

Proconvulsant or anticonvulsant properties have been reported for almost all anesthetic agents.⁸ Although there are a few reports of potential epileptogenic effects,⁹ most studies have shown that isoflurane can be considered safe for use in patients with refractory epilepsy.^10–12 One drawback of isoflurane is that some institutions, including this one, may not have the specific vaporizer required to use isoflurane. Desflurane appears to have anticonvulsant properties, but there are insufficient data available to judge its efficacy.^13,14 Sevoflurane is widely used for mask induction and maintenance of anesthesia in children and adults primarily because of its general ease of acceptance and lack of causing respiratory irritability.^15,16 However, numerous reports have demonstrated sevoflurane's ability to cause epileptogenic electroencephalogram (EEG) effects in patients, with or without epilepsy, especially during induction.^17,18 Sevoflurane has also been suspected to have epileptogenic properties when administered at a steady-state concentration.¹⁷ Several risk factors for this epileptogenic activity have been proposed, ie, hypocapnia, rapid induction, and a high alveolar concentration of sevoflurane.^14,15,19 Concurrent use of nitrous oxide has been shown to help suppress sevoflurane-induced epileptiform activity, while simultaneously reducing the concentration of sevoflurane needed to achieve an adequate depth of anesthesia.²⁰ A previous report suggested not exceeding 2.5 times the minimum alveolar concentration of sevoflurane when it is administered along with nitrous oxide during induction to prevent epileptiform activity.²¹ Therefore, when using sevoflurane for mask induction of anesthesia in patients with a seizure history or with a VNS device, it may be ideal to use nitrous oxide while gradually increasing the sevoflurane concentration, while simultaneously avoiding hyperventilation and remaining below 2.5 times the minimum alveolar concentration of sevoflurane. Inhalational agents were not selected for use in this case for these reasons.

Instead, total intravenous anesthesia using propofol and remifentanil was selected as the anesthetic technique of choice for this patient. Some reports suggest that propofol can cause seizure-like phenomena in patients with and without epilepsy.^7,22 However, these seizure-like phenomena do not correlate with epileptiform activity seen on an EEG and are thought to be excitatory movements caused by enhancement of the activity of alpha motor neurons by propofol, most commonly associated with rapidly changing cerebral concentrations of propofol.^23,24 Experimental studies and clinical evidence have demonstrated that propofol elevates the seizure threshold and has antiepileptic effects.^25–28 Therefore, propofol was considered a safe anesthetic agent for use in this patient with refractory epilepsy and a VNS device.

Opioids, in contrast, have a more complex role in terms of inducing or inhibiting seizure activity. During induction of anesthesia, a 70-μg bolus infusion of remifentanil was reported to induce tonic-clonic seizures that were recorded using a BIS sensor.²⁹ High-dose continuous infusions of remifentanil at 1.0 μg/kg/min were also reported to induce tonic-clonic seizures in an adult.³⁰ Large doses of fentanyl (400 μg/kg) have been shown to produce epileptiform activity on an EEG in animal models, whereas smaller doses (200 μg/kg) did not.³¹ Several opioid agonists, at relatively low doses, have demonstrated antiepileptic effects that help to counter the increased activity often observed with sevoflurane. For example, injection of a low dose of alfentanil (20 μg/kg) was shown to help limit the epileptogenic activity of sevoflurane, which is in contrast with a previous study in which a large dose of alfentanil (50–75 μg/kg) had a proconvulsant effect.²⁰ In one report, injection of a low dose of fentanyl (100 μg) reduced the frequency of spike waves in a 36-year-old man with intractable epileptic seizures who was anesthetized with 1.5% sevoflurane.³² These reports seem to indicate that higher dosages of opioids are associated with an increased risk of inducing epileptic seizures whereas lower dosages can help to suppress that risk. In this case, a higher dosage of remifentanil or any other opioid was not needed, as the dental procedures rendered were minimally invasive and local anesthesia was used to minimize noxious afferent stimulation. However, if a higher opioid dosage is required for an invasive surgical procedure, it may be necessary to prepare for the increased possibility of epileptic seizures.

Previous reports have shown that not only refractory status epilepticus, but also a single epileptic seizure can produce neuronal damage and death of brain cells. Seizure-induced damage and cell death may adversely affect the functional properties of neural circuits and networks.^33,34 Therefore, a BIS sensor was used in this case to help detect epileptic seizures even if the tonic-clonic muscular activity was masked by the concurrent use of a muscle relaxant. There are reports of an association between a low-frequency, high-voltage EEG waveform with increased δ and θ power and epileptiform discharges on a BIS sensor.^29,35,36 However, using changes in BIS values to help diagnose seizure activity remains controversial. In one study, 10 patients with refractory status epilepticus who received propofol in the intensive care unit were found to have increased BIS values from <30 to an average of 64 (95% CI = 53–74) when the patient exhibited epileptiform activity on the EEG.³⁷ In another report, a 67-year-old man scheduled to undergo carotid endarterectomy under general anesthesia developed generalized tonic-clonic seizures when he received 70 μg of remifentanil during induction of anesthesia.²⁹ In that patient, the BIS value decreased from 95 to a minimum of 42, and the EEG waveform suddenly changed from low-voltage, high-frequency waves to high-voltage, low-frequency waves. These reports indicate that development of epileptiform discharges can result in unusual increases or decreases in BIS values, in addition to the presence of epileptiform EEG waves on the BIS monitor. Therefore, the EEG waveform should be assessed for epileptiform waves if BIS values change abruptly in a patient with a VNS device. In this patient, anesthesia was maintained using only intravenous agents, and no abnormal fluctuations in BIS values or EEG waveforms were observed during surgery. If the BIS sensor had detected epileptiform EEG activity in this patient, the plan was to utilize the dedicated magnet to activate the VNS device. It is possible that use of the magnet would have been ineffective in breaking the seizure activity, based on the recent increasing frequency of seizures over the last several years following placement of the VNS, which would have required an alternative approach, such as administration of a benzodiazepine.

A previous report recommended premedication with benzodiazepines for prevention of epileptic seizures because of the presence of epileptiform EEG during sevoflurane anesthesia.³⁸ Another study reported that epileptiform activity on the EEG was not noted with the use of sevoflurane following induction with midazolam and thiopental.³⁹ Furthermore, a few previous studies reported that the coadministration of midazolam and propofol not only is safe, but also offers several significant potential advantages, such as anxiolysis, amnesia, decrease in the pain associated with propofol injection,⁴⁰ and reduced hemodynamic impairment.⁴¹ Therefore, administration of a benzodiazepine, either preoperatively or intraoperatively, may be helpful in preventing epileptic seizures, even in patients with a VNS device. The decision to not use benzodiazepines for this case is one that likely warrants reconsideration, as upon further examination the literature suggests benzodiazepines would have been beneficial.

The second issue regarding perioperative management of patients with a VNS device is the potential for respiratory complications. Use of a VNS device can lead to a decrease in air flow, tidal volume, and oxygen saturation during sleep.³ Most patients with a VNS device have an increase in their apnea-hypopnea index immediately after initiation of therapy.⁴ Proposed mechanisms include effects on the respiratory center and peripheral stimulation of the vagal nerve, which activates motor efferents and results in altered neuromuscular transmission to the laryngeal and pharyngeal musculature.⁴² Additionally, a previous report demonstrated that patients with refractory seizures are at an increased risk of OSA (13/39 patients, 33%).⁴³ Therefore, the additive or synergistic effects of VNS, OSA, and concurrent use of anesthetic agents were likely to increase the risk of significant respiratory complications occurring during the perioperative period. It was unclear while reviewing this patient's medical history whether she had sufficient signs or symptoms to warrant high suspicion of OSA, because no screening tool, such as a STOP-Bang assessment, was used. However, given her morbid obesity, there were concerns about the possibility of perioperative respiratory complications, such as apnea or obstruction. The risk of respiratory complications associated with VNS devices can be reduced by positional therapy (lateral decubitus positioning or prone positioning) to help maintain airway patency and/or by adjusting the settings of the VNS device. In severe cases, use of the magnet to temporarily deactivate the VNS device has even been proposed.⁴⁴ However, more research is needed to address the question of adequacy of seizure control for patients in whom the VNS settings are altered or even deactivated to prevent respiratory complications. The decision to change the VNS settings or ultimately deactivate the device requires careful consideration.³ The anesthetic strategy planned for our patient during the postoperative period included positional therapy, supplemental oxygen, and close monitoring in the recovery room within the American Society of Anesthesiologists guidelines for OSA.⁴⁵ If severe respiratory complications had persisted after the procedure, deactivating the VNS device by using the magnet would have been considered. However, no such events occurred. In a patient with a VNS device with suspected severe or persistent respiratory complications, consideration should be given to deactivating the device using the magnet or altering the VNS settings prior to surgery, after weighing the risks and benefits of reducing the VNS activity versus the risk of persistent respiratory complications.

One limitation of this report is that a BIS sensor was used to detect the development of epileptiform activity on an EEG. Previous reports demonstrated that a BIS sensor was used successfully to identify the development of epileptiform EEG activity and its immediate resolution following administration of anticonvulsants.^35,44 A BIS sensor records frontal brain activity and does not necessarily capture epileptic discharges in other parts of the brain.²⁰ However, a BIS sensor may be able to detect abnormal EEG waves originating in other parts of the brain if that activity propagates to the frontal regions where the electrical activity can be captured and recorded.³⁵ It is possible that this patient experienced seizure activity that remained undetected by the BIS sensor.⁴⁶ Although a BIS sensor may be less than ideal for use specifically as a diagnostic tool, it may still be somewhat beneficial to help monitor for epileptiform EEG activity during an anesthetic.

CONCLUSION

This case report presents the safe anesthetic management of a patient with refractory epilepsy and a VNS device using propofol, remifentanil, and a BIS monitor. Additional areas of discussion included the mechanics of using the VNS device during the perioperative period as well as the increased potential for respiratory complications. Anesthesiologists must understand the impact of anesthetic drugs on the seizure threshold, the limitations of using a BIS monitor when identifying and interpreting epileptiform EEG activity, and the mechanics of the VNS device itself, particularly use of the VNS magnet.