Research progress of vagus nerve stimulation in the treatment of epilepsy

Abstract The International League Against Epilepsy (ILAE) defined drug‐resistant epilepsy (DRE) that epilepsy seizure symptoms cannot be controlled with two well‐tolerated and appropriately chosen antiepileptic drugs, whether they are given as monotherapy or in combination. According to the WHO reports, there is about 30%‐40% of epilepsy patients belong to DRE. These patients need some treatments other than drugs, such as epilepsy surgery, and neuromodulation treatment. Traditional surgical approaches may be limited by the patient's clinical status, pathological tissue location, or overall prognosis. Thus, neuromodulation is an alternative choice to control their symptoms. Vagus nerve stimulation (VNS) is one of the neuromodulation methods clinically, which have been approved by the Food and Drug Administration (FDA). In this review, we systematically describe the clinical application, clinical effects, possible antiepileptic mechanisms, and future research directions of VNS for epilepsy.


| INTRODUC TI ON
Epilepsy is a chronic disease characterized by sudden abnormal discharge of brain neurons, which leads to transient brain dysfunction and is defined as the presence of spontaneous recurrent seizures.
The lifetime prevalence of epilepsy is estimated at 1%-5% globally. 1 It becomes the second most common disease after headache in neurology department all over the world. 2 The treatment of epilepsy includes drugs, surgery, neuromodulation, and ketogenic diet. According to Martin's research, 3 more than 30% of epilepsy patients are drug-resistant epilepsy (DRE). The International League Against Epilepsy (ILAE) defined that DRE is uncontrollable seizure within two well-tolerated and appropriately chosen antiepileptic drugs, whether they are given as monotherapy or in combination. 4 Especially, focal epilepsy such as temporal lobe epilepsy and some epilepsy syndromes such as Lennox-Gastaut syndrome, West syndrome, and O'Hara syndrome could not be wholly controlled by drugs. Surgical resection is one of the choices of epilepsy treatments, but not all patients meet the criteria for the surgical procedure. Even though the epileptogenic focus is removed surgically, 30%-50% of the patients cannot become seizure-free. 5 Therefore, in addition to drugs and surgery, there should be an alternative way for epilepsy patients to reduce their epilepsy syndrome and sufferings.

| THE VAG US NERVE S TIMUL ATI ON
In the past few years, neurostimulation has been proved to be a safe and effective method in many preclinical and clinical trials, which can be combined with traditional drug therapy to reduce seizures.
At present, the American FDA has approved vagus nerve stimulation (VNS) for the treatment of epilepsy. [6][7][8] In this study, we mainly focus on some evidence-based medical evidence related to VNS.
Vagus nerve stimulation is a relatively novel therapy method in the management of the neurological disease such as depression, epilepsy, tinnitus, and schizophrenia. [9][10][11] At present, there are many studies on the treatment of epilepsy and depression, and the effect is more affirmative. There is insufficient evidence for the treatment of schizophrenia and tinnitus by VNS. It was first proposed in the 1880s in Ref. 12 Nothnagle considers that "venous hyperaemia of the brain" caused by carotid artery pulsation leads to seizures. Electrical stimulation of the vagus nerve may reduce the "venous hyperaemia of the brain." According to the theory, James Corning developed several carotid artery compression devices for the treatment of seizure.
He combined the instrumented carotid artery compression with transcutaneous vagal nerve stimulation device to decrease cerebral blood flow. He was the first physician to report the usage of transcutaneous electrical stimulation of the vagus to interrupt convulsions in 1883. 13 After a century of exploration, in 1988, it was first reported that the VNS device was implanted into the human body for the treatment of drug-resistant epilepsy with chronic VNS. 14 Vagus nerve stimulation was approved as an adjunct therapy to reduce the frequency of seizure aged ≥12 years with DRE by the American FDA in 1997. 15,16 In 2017, the FDA approved of using VNS in patients over 4 years old that characterized partial seizure and intractable epilepsy. 17 Vagus nerve stimulation also has evidence for treatment of depression, which was approved by the FDA and may be useful for other comorbidities of epilepsy. 18

| APPLI C ATI ON
Indeed, many clinical studies have proved VNS is safe and effective in the treatment of epilepsy. A multicenter controlled study of VNS in the treatment of focal epilepsy (EO3) recruited 67 patients who had completed the blinded acute phase (14-weeks) and properly randomized analysis. One group (N = 31) was treated with highfrequency stimulation (0.25-3.0 mA, 20-50 Hz, 500 μs), and the control group (N = 36) was treated with low-frequency stimulation (0.25-3.0 mA, 1-2 Hz, 130 μs). After 14 weeks of VNS treatment, the result showed that the mean seizure frequency reduced about 30.9% in the high-frequency stimulation group. However, in the low-frequency stimulation group it was about 11.3%. There was a statistical difference (P = .029) in the change of seizure frequency between these two groups. In the high-frequency stimulation group, the frequency of seizure decreased by more than 50% is about 39%, while in the low-frequency stimulation VNS group, it is about 19%. In the aspect of 50% reduction, the comparison of these two groups did not reach the statistical significance requirement (P = .0704), but there is a big tendency between these two groups. Although it has been reported that the intensity and severity of seizures have been reduced in individual patients, no significant statistical difference has been found for the time being. 15 Despite recruited a small number of subjects and performed shorttime follow-up study, this study lays a foundation for the wide application of the VNS.
Some other randomized controlled clinical trials have also confirmed similar findings. 15,19-21 A retrospective study from our hospital also proved that 60 out of 94 patients achieved the therapeutic effect (50% frequency of seizure decreased) from November 2008 to April 2014, regardless of age and gender. 22 To confirm long-time treatment safety and understand the efficacy of time cumulation, a single-center study of VNS treatment for epilepsy followed up for 10-17 years (N = 74) showed that the rates of seizure frequency reduction between 50% and 90% were 38.4%, 51.4%, 63.6%, and 77.8% in the years of 1, 2, 10, and 17 after VNS treatment, respectively. The rates of seizure frequency reduction more than 90% were 1.4%, 5.6%, 15.1%, and 11.1% in corresponding treatment follow-up time. 23 With the prolongation of treatment, the VNS therapeutic effect is gradually enhanced.
Although the FDA has approved VNS for intractable epilepsy in children and adults, the application of VNS in epilepsy remains controversial due to its small sample size and short follow-up time.
To further demonstrate the efficacy of VNS and identify which patient populations respond best to treatment, Englot et al conducted a large retrospectively registry-based study. In this VNS treatment study, totally about 1285 physicians from 978 centers registered outcomes of 4483 patients treated with VNS. 24 Compared with the patients' ages, disease course, and seizures types and followed about 1 year, more than half of the patients responded to VNS treatment (with a 50% reduction in seizure frequency). First, the younger patients under 18 years old (60% fewer seizure frequency than baseline) respond better to VNS treatment than those over 18 years old (53% fewer seizure frequency than baseline). Studies have proved that children may get more benefit from the VNS treatment compared with adults group. Second, the course of epilepsy before VNS device implantation is a potential predictor for the response to VNS treatment. Less than 10 years' seizure history is predicted a somewhat higher clinical response (56% vs 52%) to VNS (OR, 1.19; 95% CI, 1.00-1.41). The result suggests that individuals with shorter epilepsy course are more favorable to respond to VNS. Therefore, the earlier the VNS treatment intervention, the better the efficacy may be. Third, according to the types of seizures, the individuals with predominantly focal seizures, including auras, achieved the most significant clinical benefit. These patients were more likely to respond to VNS than other types of seizures by 1 year's study (OR, 1.37; 95% CI, 1.04-1.81, P = .025).
The therapeutic effect of VNS may not immediately respond to the stimulation, and the frequency of seizure begins to decrease after implantation within a few months. During this period, some patients may change the dosage of antiepileptic drugs due to their displeasure for the seizure control, and these changes may have an impact on the evaluation of the efficacy of VNS. To clarify this controversy, Garcia-Pallero et al recruited 85 patients undergoing VNS operation who were included for prospective analysis and followed up after the operation. Among them, 43 patients were not allowed to change antiepileptic drugs during the follow-up period, and 42 patients could change antiepileptic drugs under the guidance of doctors. Comparing these two groups following within 18 months, the results showed that 54.1% of the patients had more than 50% frequency of seizure reduction. About 63% of the patients in drug unchanged group had a 50% frequency of seizure reduction, and 45.2% of the patients within drug change group had the 50% frequency of seizure reduction. Therefore, there is no statistical difference in frequency reduction whether there is any change of drugs. 25 Changing the dosage and type of antiepileptic drugs (AED) may not influence the treatment results of VNS.
In the treatment of epilepsy including children and adults, VNS treatment has been proved safe and effective. 26 Thus, the FDA has approved that VNS can be used in children over 4 years old recently. [27][28][29][30][31] The VNS treatments do not only reduce the frequency of seizures but also reduce the psychological burden of children. 32 There is another study has reported that VNS may have a therapeutic effect on epilepsy patients <3 years old. 33 In this clinical research, the authors concluded that the frequency of status epilepticus was also significantly reduced after VNS treatment. 33  The results show that VNS can also improve the comorbidity of epilepsy, such as mood disorders and cognitive deficits. These are two large pivotal clinical trials confirmed the effectiveness of VNS in the treatment of refractory partial epilepsy and also concluded that epilepsy patients who had comorbid depression improved. 15,19,20,41,42 Since then, more and more studies have been conducted on the treatment of depression with VNS treatment. It is effective in patients with treatment-resistant depression (TRD). Treatment-resistant depression is an American FDA-approved indication for VNS. 43 The overall effect of VNS on cognition is unclear, and many studies focus on memory. There is an evidence that VNS treatment can acutely improve memory in a short time, but there seems to be no significant improvement within the long-term treatment. 44

| THE MECHANIS MS ABOUT VN S REG UL ATION
Although the therapeutic effect of VNS is remarkable in the clinic, the mechanism of VNS treatment in epilepsy is not completely understood. Currently, some potential mechanisms are proposed.
Neuroelectrophysiology: Some preclinical studies are of great help in exploring VNS regulation mechanism. Early studies initially revealed that VNS treatment could cause desynchronization of cortical electrical activity in cats. 47 Zanchetti et al 48 54,67 The forebrain and limbic system also receive NTS projections, including the bed nucleus of the stria terminalis, paraventricular, dorsomedial, and arcuate hypothalamic nuclei, preoptic and periventricular thalamic nuclei, and central amygdala nucleus. 68,69 However, how these brain regions are connected and functioning is not very clear.
In the past, most of the electrophysiological methods were used to record the activities of local brain areas. Recently, many studies have applied functional neuroimaging to explore the effects of VNS which to be characterized throughout the brain. Some researchers used positron emission tomography (PET) or single-photon emission computerized tomography (SPECT) to detect increased metabolism or blood flow in thalamus, hippocampus, amygdala, lower cerebellum, and cingulate cortex during or after VNS treatment. [70][71][72][73] However, these techniques are difficult to capture the dynamics of poststimulus responses because of their poor temporal resolution.
Functional magnetic resonance imaging (fMRI) has a high temporal and spatial resolution, which can compensate for this shortcoming.
Our team traced the neural circuits of VNS with viruses and found that they were activated in the cortex, hippocampus, and lower part of nucleus accumbens of rats. Next, we will carry out functional exploration and verification.

| SAFE T Y AND COMPLI C ATI ON S
Many studies have shown that VNS implantation is a relatively safe operation. It has been reported that there is no definite teratogenicity in pregnant women treated with VNS because the sample size is too small; further research is needed. 79 The common complications include infection, postoperative hematoma, and vocal cord paralysis. The incidence rate is about 2%. Very few patients suffer from bradycardia or cardiac arrest, and most of them occur during operation. Some equipment-related complications may also occur, such as repetitive surgery, battery replacement, lead fracture, or malfunction. 80 Since neuromodulation has become an important treatment option of drug-resistant epilepsy, in this review we also briefly comment on the indication, effectiveness, and adverse effects of responsive nerve stimulation (RNS) and deep brain stimulation (DBS), as compared to VNS. Detail information could be found in the following Table 1.
Direct VNS is a kind of invasive surgery, which is relatively expensive, and the patients bear the risk of repetitive surgery such as battery replacement. Therefore, compared with VNS, transcutane-

| CON CLUS ION
Vagus nerve stimulation is an adjuvant treatment for drug-resistant epilepsy, which had been approved to treat focal epilepsy patients more than 4 years old by the US FDA. Randomized clinical trials have provided that VNS is also a safe and effective treatment for the younger infants and children. Some clinical cases reported that VNS is also helpful for some particular epilepsy types, includ-

CO N FLI C T O F I NTE R E S T
The authors declare no conflict of interest. TA B L E 1 Indication, effectiveness, and adverse effects in VNS, RNS, and ANT-DBS

VNS RNS ANT-DBS
Indication VNS is indicated for symptomatic localization-related epilepsy with multiple and bilateral independent foci, symptomatic generalized epilepsy with diffuse epileptogenic abnormalities, refractory idiopathic generalized epilepsy, failed intracranial epilepsy surgery, and other several reasons of contraindications to epilepsy surgery Adults with partial-onset seizures who have undergone diagnostic testing that localized no more than 2 epileptogenic foci and refractory to two or more antiepileptic medications, and currently have frequent and disabling seizures (motor partial seizures, complex partial seizures, and/or secondarily generalized seizures) Bilateral stimulation of the anterior nucleus of the thalamus (ANT) for epilepsy is indicated as an adjunctive therapy for reducing the frequency of seizures in individuals 18 y of age or older diagnosed with epilepsy characterized by partialonset seizures with or without secondary generalization that are refractory to three or more antiepileptic medications Effectiveness Seizure frequency was reduced by an average of 45%, with a 36% reduction in seizures at 3-12 mo after surgery and a 51% reduction after >1 y of therapy. At the last follow-up, seizures were reduced by 50% or more in approximately 50% of the patients, and VNS predicted a ≥50% reduction in seizures with a main-effects OR of 1.83 The average decrease in seizures was 44% after 1 y, 53% at 2 y, and up to 66% after 3-6 y of using RNS. The same trend was seen when some of these people were followed for 7 y. Seizures decreased by an average of 72%