Vagal Nerve Stimulation in the Management of Epilepsy - Recent Concepts
Correspondence Address: Source of Support: None, Conflict of Interest: None DOI: 10.4103/0028-3886.302475
Source of Support: None, Conflict of Interest: None
Keywords: Drug refractory epilepsy, idiopathic generalized epilepsy, Lennox Gastaut syndrome, neuromodulation, partial-onset seizure, symptomatic generalized epilepsy, transcutaneous VNS
Vagal nerve stimulation (VNS) for aborting seizures in humans was first demonstrated by Corning in the nineteenth century. The idea was not well taken till 1985 when Zabara showed the anticonvulsant action of VNS in dogs.,, This was followed by the pilot study of Penry and Dean (1988), implanting the first commercially available VNS device in four patients. The US Food and Drug Administration (FDA) approved VNS in 1997 for patients aged twelve years and above with drug-resistant partial-onset seizures. Recently (2017), it has been extended to include pediatric epilepsy, as young as 4 years and above. Globally over 100,000 VNS devices have been implanted currently for the treatment of DRE.
Stimulation of afferent nerve fibers of the vagus (comprising 60-80% of the nerve) results in the therapeutic effects of this therapy., Majority of these afferent fibers synapse intracranially on the nucleus of tractus solitarius (NTS), which in turn projects to multiple regions, the most important of which in the context of VNS is the locus coeruleus (LC) and dorsal raphae nucleus (DRN). Efferent fibers from the LC (noradrenergic) and DRN (serotonergic) innervate multiple areas of the hypothalamus, parts of the neocortex, cerebellum, and spinal cord., VNS leads to an increase in the activity of these nuclei, which in turn may be responsible for the anti-seizure effects., Iatrogenic lesioning of the LC in rats has been reported to lead to decreased responsiveness of VNS. The extensive connections of the NTS, LC, and DRN may be responsible for the widespread network changes observed after VNS, most notably in the limbic circuit. Additional mechanisms may be responsible but till date remain unknown.
Approximately 30–40% of newly diagnosed patients of epilepsy are refractory to initial treatment, with only 5–10% of these achieving seizure freedom, with additional antiepileptic drugs (AEDs) and less than half attain 50% reduction in seizure frequency. Of the 40% of patients with DRE, only half are candidates for resective surgery. We find the best long-term seizure outcomes with temporal lobe surgeries (60–70% seizure freedom) followed by frontal with up to 50% seizure freedom.,, However, not all patients with DRE have a localizable focus nor all the foci amenable to focal resection despite a comprehensive workup. Neuromodulation, especially VNS, offers an attractive alternative to corpus callosotomy in these cases. It also has the advantage of being an extracranial procedure. The seizure outcomes following VNS are inferior when compared to the resective epilepsy surgery. Hence, we should exclude patients suitable for resective surgery before considering VNS.
A comprehensive preoperative epilepsy workup is the essential prerequisite in all the cases deemed refractory to medical treatment. We should make all efforts to localize the epileptogenic foci. This is achieved by maximizing noninvasive investigations. These include an epilepsy protocol MRI, video electroencephalography (VEEG), Positron emission tomography (PET), single-photon emission tomography (SPECT), subtraction ictal SPECT coregistered with interictal SPECT/MRI brain (SISCOS/SISCOM). Besides, magnetoencephalography (MEG) and invasive intracranial (subdural grid/stereotactic electroencephalography placement) evaluation should also be considered in suitable cases [Figure 1]. These patients should, therefore, be managed at a specialized epilepsy care center comprising a robust epilepsy surgery program.
Indications for VNS
Suitable candidates for VNS can be categorized as follows:
This form of epilepsy is usually secondary to diffuse brain insult, usually sustained during childhood (birth-related injuries, hypoxic damage). Most of these patients suffer from mental retardation, neurocognitive deficits, and abnormal EEG showing multiple seizure patterns, along with epileptiform discharges. Lennox Gastaut Syndrome (LGS) is a typical example of this group of patients.,
Patients characterize this group without any neurological deficits indicating the absence of brain insult, commonly have genetic association along with normal EEG except for the epileptiform discharges and often normal brain imaging. Nearly 80-90% of patients with IGE respond to AEDs. The unique problem in this group of patients is pseudo intractability of seizure progression. This occurs due to inappropriate usage of AEDs, specifically carbamazepine and phenytoin, which may aggravate the seizures. Hence, this differentiation is imperative before considering VNS in these patients.,,
VNS may be used as salvage therapy in patients with failed resective surgery or following corpus callosotomy (CC). The seizure reduction following resective surgery has been modest in the reported literature. Nonetheless, the subjective quality of life improves markedly, also termed as a positive psychotropic effect of VNS.,, Similarly, VNS has been successfully used for managing persistent epilepsy following CC, and one of the studies analyzed the combination of CC and VNS in LGS with severe encephalopathy with promising results.,,,
Prior neck surgery, active peptic ulcer disease, chronic pulmonary diseases, asthma, preexisting hoarseness, cardiac arrhythmias, dysautonomias, pregnancy and insulin-dependent diabetes mellitus are relative contraindications, and these candidates may not be suitable for VNS therapy.
An 18 years male with seizure onset at four years of age, presented with multiple seizure types: predominant drop attacks, bilateral tonic, and occasional clonic seizures. He had sustained fracture involving right humerus, multiple healed scars over forehead due to frequent drop attack and also with a history of delayed cry 24 hours after birth. Neuropsychological examination revealed moderate mental retardation. The patient was started on five antiepileptic drugs (optimum doses) throughout illness without respite in seizures. On referral to our institution, we evaluated with MRI brain, which showed bilateral parietooccipital gliosis suggesting perinatal insult.
Further evaluation with VEEG revealed bilaterally symmetrical generalized seizure onset seen during head drops without any lateralization. Additional evaluation with fluro-deoxyglucose SPECT showed bilateral parietooccipital localization of the dye. MEG also revealed bilateral seizure onsets with tight clusters in left temporal and also in the right parietooccipital region [Figure 2].
We diagnosed the patient as Lennox gastaut syndrome, given no lateralization of the seizure focus to any particular hemisphere. We performed a VNS implantation for the patient. At 12 months follow-up, this patient has >50% reduction in the seizure frequency with >90% relief in drop attacks with significant subjective improvement in cognition and behavior as reported by the parents.
A 24-year-old male presented with seizure onset at the age of 6 years, with a seizure frequency of 3–4/week and two types of seizure semiology. One of the seizures began with flashes of lights, followed by turning off the head to the right and tonic posturing of the right upper and lower limb and occasionally progressing to generalized tonic-clonic along with urinary incontinence. The patient also recalled an occasional history of postictal vomiting. The other variety was a complex partial seizure variety. The patient was on five AEDs at optimum doses started throughout the disease.
We performed VEEG, which revealed T3, T5 ictal onsets. MRI brain showed bilateral occipital gliosis (L > R) and subtle changes in the left anterior temporal lobe suggestive of focal cortical dysplasia (FCD). The visual field evaluation, despite bilateral occipital gliosis, showed normal vision without any field defects. We further evaluated with PET scan, which showed mild hypometabolic area localized to the left anterior temporal lobe. SISCOM image showed a focus localized to left temporal and frontal lobes. We then performed MEG, which showed tight clusters localized to the left temporal on DANA sequence, while LORETA localized strongly to the left occipital region [Figure 3] and [Figure 4].
The ictal onset zones were lateralized to the left anterior temporal (Semiology, VEEG, SISCOM, MRI, PET, MEG) and left occipital lobes (Semiology, MRI, MEG).
Hence, to confirm our hypothesis, we performed SEEG implantation of left temporal and bilateral occipital and sentinel electrodes in the right temporal lobe. We placed a total of 13 SEEG electrodes. The SEEG recordings revealed seizure onset consistently localized to left mesial occipital lobe in the vicinity of the visual cortex, both on spontaneous and on electrode stimulation-induced seizures [Figure 5].
After analyzing the SEEG, we felt that the patient was an ideal candidate for VNS therapy, owing to the location of the epileptogenic zone in the visual cortex. This case reiterates the fact that despite localization of the epileptogenic zone by maximizing the available tools, curative resection may not always be suitable. VNS offers the best option of therapy in this group of patients.
Illustrative case 3
A 27-year-old male had seizures for 25 years and presented two types of seizures. These included an aura of fear with oral and right upper limb automatisms, suggestive of complex partial seizures without secondary generalization. This type of seizures occurred once/week. Occasionally, he also had seizures starting with a head turn to the right, followed by tonic movements of the right hand followed by secondary generalization. There was a history of a minor head injury at two years of age. MRI brain under standard epilepsy protocol revealed bilateral MTS along with dysplasia of the left anterior temporal lobe. The VEEG localization was F7, T3 and T5 suggesting bilateral onset. He was started on three AEDs throughout the disease without benefit. We further evaluated with Positron emission tomography (PET) and SPECT, which showed bilateral hypometabolism (although Left > right). Magnetoencephalography (MEG) showed bilateral temporal and right inferior frontal lobe localization [Figure 6] and [Figure 7].
The hypothesis formed after a comprehensive epilepsy surgery meeting was bilateral MTS with left anterior temporal FCD. Although MRI, SPECT, PET, and MEG revealed bilateral temporal foci, the findings were more on the left compared to the right temporal lobe.
Because of the bilateral seizure onset, we planned for SEEG placement to localize the epileptogenic zone. We performed standard bilateral temporal and left frontal exploration with SEEG. We placed a total of 12 SEEG electrodes under robotic guidance (ROSA, Schiller). The SEEG recordings lateralized to the left temporal lobe (left amygdala and left hippocampal) on both spontaneous and stimulation-induced seizures [Figure 8].
We were able to localize the epileptogenic networks here due to an extensive preoperative workup and SEEG. Hence, we planned for a standard left temporal lobectomy, after a detailed discussion with the patient and family members with a guarded prognosis. An additional option of VNS was provided, in case of failure of the resective surgery. This patient underwent a standard left temporal lobectomy and is doing well with no seizures at two years follow-up (ILAE class-1). Resection of the epileptic focus yields better results with greater chances of curing epilepsy compared to palliative neuromodulation. We included this case here to reiterate the fact that, by maximizing the utilization of available investigations, a curative resection may be performed. This case demonstrates the need for extensive workup before considering VNS and the necessity to rule out the possibility of resective surgery.
The programming is performed after a lag period of two weeks following the VNS device implantation. This avoids the irritating sensation of the neck and also allows for wound healing. The device once activated stimulates the nerve intermittently at regular intervals. The stimulation is of two types: Normal mode (Ongoing as per the set parameters) and the Magnet mode (used to abort an acute attack).
The initial parameters (normal mode) are started with a current strength of 0.25 mA, frequency of 30 Hz, on-time of 30 s, off time of 5 min, a pulse width of 500 μs. The magnet mode is on-demand stimulation, which is activated during an acute attack of seizure. Here, the caregiver/patient himself passes the magnet provided by the company over the pulse generator implanted in the chest for 1 second. This magnet may be worn over the wrist in case of adolescent/adult patients. The stimulation parameters can vary across institutions, based on the discretion of the treating physician, customized to patient response. The VNS therapy dosing guidelines for standard dosing and duty cycles are presented (LeVaNova PLC, London, UK. Currently, the newer devices available have an auto-stimulation function. They are capable of automatically delivering the stimulation current by identifying preictal tachycardia, seen in the majority of the patients with DRE. With these devices, it is even possible to schedule future programming, thereby reducing the number of visits to outpatient clinics.
The E03 and E05, two pivotal multicenter, blinded randomized controlled trial by the VNS study group compared the efficacy of HIGH stimulation and LOW stimulation. The high stimulation group received: 30 Hz, 30 s on, 5 min off, 500 μs pulse width, while the low stimulation group received: 1 Hz, 30 s on, 90–180 min off, 130 μsec pulse width. They found the statistically significant seizure reduction rate in the HIGH compared to the LOW stimulation at three months. Hence, a HIGH stimulation strategy should be applied in non-responders, especially to the LOW stimulation parameters.,
Vagal nerve stimulation is nonpharmacological therapy for the treatment of DRE. Seizure reduction remains the primary goal of any epilepsy surgery. However, VNS produces other beneficial effects, apart from long-term seizure reduction like mood elevation, cognition, improvement in memory.
There have been five randomized controlled trials reporting on the efficacy of the VNS in reducing the seizure frequency. All these trials recruited patients with partial-onset epilepsy.
The vagal nerve stimulation group conducted the short-term multicenter randomized, blinded controlled trials E03 (n = 114) and the E05 (n = 199). These studies aimed to compare the effect of HIGH versus LOW-frequency stimulation strategy in patients implanted with the VNS device. The HIGH-frequency stimulation group fared better than the LOW stimulation (control) group achieving statistical significance in terms of seizure reduction rate. These studies were, however, were limited by a short follow-up of 3 months.,
To overcome the limitation of E03 and E05 trials, the open-label prospective study by the VNS Study group (E01-E05) was performed, to assess the long-term efficacy, safety and tolerability. This study reported ≥50% seizure reduction post-implantation in 36.8% of patients at 1 year, in 43.2% at 2 years, and in 42.7% at 3 years.
Englot et al. analyzed the largest cohort of patients (n = 5554) drawn from the VNS therapy registry and systematic review (n = 2869) across 78 studies published in the literature. The overall response rate (≥50% seizure reduction) was 49% at four months, with 5% of patients achieving remission. The response rate improved to an impressive 60% at 2-4 years follow-up steadily with a remission rate of 8%. Age at onset 12 years and generalized type of epilepsy predicted seizure freedom, while non-lesional epilepsy predicted an overall good response to VNS therapy.
A retrospective study, spanning over 11 years of consecutive VNS implanted patients (n = 436) with a mean follow-up of 4.94 ± 3.24 years, reported ≥50% seizure reduction in 63.75% patients. A subgroup analysis of this study included consecutively treated patients (n = 65) with ≥10 years follow-up. The mean reduction in seizures at 6 months and years 1, 2, 4, 6, 8, and 10 years was 35.7, 52.1, 58.3, 60.4, 65.7, 75.5, and 75.5%, respectively suggesting an improving efficacy of VNS with time. However, the indication of VNS implantation was not restricted to partial epilepsy and included patients with multiple seizure types (Symptomatic and Idiopathic generalized epilepsies).
It is evident that the beneficial effects of the VNS progressively improves with time beginning from the point of VNS implantation in the ‘responders’ (defined as >50% reduction in the seizure rates compared to presurgical status).
The only randomized controlled trial in the pediatric population studied the efficacy of VNS in seizure reduction and behavioral aspects like cognition, mood, psychosocial parameters, and also epilepsy-based restrictions. Here, children were divided into HIGH and LOW stimulation groups. The patient population included both focal and generalized epilepsies, including Lennox gastaut syndrome. Patients treated in the LOW stimulation group achieved higher than expected seizure reduction, although it was not statistically significant. Overall 26.47% patients had ≥50% seizure reduction in this study.
The FDA approved VNS therapy in the treatment of major depression, owing to its essential action of mood elevation. As the majority of the patients with chronic epilepsy are depressed, VNS induces a sense of well-being in these patients apart from seizure reduction. It is also reported to improve cognitive function, improve quality of sleep. One of the studies showed an improvement in the retention capacity of memory.,
The PuLsE study, an open-label prospective randomized trial, compared the efficacy of adjunctive VNS to best medical practice (BMP) versus BMP alone. The VNS and BMP group was found to be superior in attaining significant improvement of health-related quality of life compared to BMP alone group. The study also demonstrated the superiority of the VNS in seizure rate reduction compared to BMP.
In the RCT by Klinkenberg and colleagues, the authors noted an overall beneficial effect of VNS in terms of cognition, behavior, mood and overall quality of life. The authors found no adverse effects of VNS on these parameters compared to the AEDs and that the overall improvement may be attributed to this indirect effect. Another essential aspect noted was an improvement in sleep quality and reduction in the seizure rate, which also indirectly contribute to the positive effects of VNS.
Sudden unexpected death in epilepsy (SUDEP) is among the common causes of mortality associated with chronic epilepsy, especially in DRE patients. The incidence of SUDEP is 1-2/1000 patient-years with chronic epilepsy and 2-10/1000 patient-years in patients with DRE. There has been a lack of studies analyzing the role of VNS in the prevention of premature deaths. The available studies show mixed results, with few studies concluding no effect of VNS on SUDEP.,, The most extensive registry-based patient data (n = 40,443) with up to 10 years follow-up revealed a significant reduction in the occurrence of SUDEP. The extent of seizure reduction, associated comorbidities, severe cognitive deficits, and encephalopathy might contribute to the premature deaths. However, long-term prospective follow-up studies in the future might add more to our limited understanding currently.
Most of the studies, including the randomized trials, have shown tolerable adverse events following VNS implantation. The E01-E05 study reporting hoarseness (28%) and paraesthesias (12%) at 1 year, hoarseness (19.8%) and headache (4.5%) at 2 years, and shortness of breath in 3.2% at the end of 3 years.
Elliot et al. reported varying degrees of permanent vagus nerve injury (hoarseness, dysphagia, unilateral vocal cord palsy) in 2.8% cases. Device removal (2.6%) was required due to: infection (1.6%), persistent, severe neck pain synchronous with the duty cycle resistant to alteration in parameters (0.7%); and pneumothorax (0.2%).
The adverse events associated with VNS are mostly as a result of the surgery and can be minimized with meticulous surgical technique. The other nonsurgical complications are rarely disabling and tend to improve with time.
Closed loop VNS
The recent technological advancement is the introduction of the closed-loop version of VNS – Aspire SR (seizure response). This was based on the observation that more than two-thirds of the patients had ictal onset tachycardia. An Automatic Stimulation (AutoStim) mode of the device can detect the increase in heart rate, compare it to the normal baseline of the individual patient and deliver a stimulation based upon a threshold decided by the physician., Initial trials reported that approximately 60% of simple and complex partial seizures could be aborted by stimulation. Responder rate (≥50% reduction in seizure frequency) ranged from 30 to 50%. Long-term analysis at 12 months showed sustainable specificity of seizure detection and improved seizure frequency and quality of life. Adverse effects were comparable to previous VNS devices., With gradually increasing experience with the device, authors have reported improved outcomes. In a recent series of 46 patients, epilepsy etiology, age of the patient and type of seizure were found to have no bearing on the outcome. 10.9% (n = 5) of the patients attained seizure freedom. In another study, twenty patients with generalized epilepsy were followed up with regular VNS for six months and the results compared with closed-loop VNS for a further six months. Although no statistically significant difference in outcome was observed, 40% of the patients were observed to have a further reduction in seizure frequency after closed-loop VNS.
The Sentiva-1000 is the most recent modification, which is small in size (can be used even pediatric patients) along with the Auto-stim function. This device also enables dose customization and schedule future titrations in a single visit, thereby avoiding multiple visits to the hospital. This device also offers tailored programming for a specific period of the day suited for the patient needs. Randomized controlled trials comparing the closed-loop VNS over regular VNS should be planned in future to provide concrete answers regarding the efficacy of the closed-loop VNS devices.
Noninvasive vagal nerve stimulation
Transcutaneous VNS (tVNS) was developed to overcome the potential complications and high costs associated with invasive device implantation. Additionally, there is no need to exhaust all other investigations and treatment modalities before using this therapy, and it may also be implemented as a possible first-line treatment modality. It is a noninvasive technique that utilizes surface electrodes to target the afferent fibers the vagus nerve passing through the auricular branch. Another area targeted is the cervical branch of the vagus in the neck, which is relatively non-specific, in terms of both the efferent and afferent fibers getting stimulated indiscriminately. The success of invasive VNS led to a pilot study in 2012, using tVNS on ten patients with DRE. The authors showed a reduction in seizure frequency in 50% of the patients. A controlled trial (n = 60) was then conducted. It demonstrated a significant reduction (P = 0.003) in seizure frequency in the treatment group at twelve months follow-up. Mood and QoL were also noted to improve. Reversible stimulation related side effects in the form of daytime somnolence were recorded in three patients and dizziness in one case. A more recent study reported an average of 64.4% decreased frequency of seizures in 16 out of 17 patients at six months follow-up, and ten patients showed an improved EEG.
The paper has been partially funded by the Department of Biotechnology, Ministry of Science and Technology, India as a part of MEG Resource facility.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
[Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6], [Figure 7], [Figure 8]