Vagus Nerve Stimulation (VNS): Drug-Resistant Epilepsy, Depression & Transcutaneous Devices — Cost, Top Hospitals & Success Rates | MyMedicPlus

Vagus nerve stimulation (VNS) is a neuromodulation therapy that delivers intermittent electrical pulses to the left vagus nerve via a surgically implanted pulse generator, modulating neural activity across multiple brain regions to reduce seizure frequency, improve mood, and potentially provide benefit in a growing range of neurological and inflammatory conditions.

The vagus nerve (cranial nerve X) is the primary parasympathetic outflow pathway, carrying both efferent (80%) signals from brain to viscera and afferent (20%) signals from viscera to brain. VNS exploits the afferent projections — which ascend to the nucleus tractus solitarius (NTS) in the brainstem, then project broadly to the locus coeruleus (norepinephrine), dorsal raphe (serotonin), and diffusely to the cerebral cortex via the thalamus — to modulate cortical excitability and neurotransmitter dynamics without focally ablating any brain structure.

The first human VNS implantation for epilepsy was performed in 1988 by neurosurgeon John Penry and engineer Jake Dean at Wake Forest University. The device was granted FDA approval for refractory partial-onset seizures in 1997 — the first non-pharmacological, non-surgical brain stimulation device approved for epilepsy. A second FDA approval for treatment-resistant depression (TRD) followed in 2005, after studies demonstrated antidepressant effects independent of seizure control.

The current generation devices — LivaNova AspireSR (Model 106) and LivaNova SenTiva (Model 1000) — deliver programmable electrical pulses (output current 0–3.5 mA, pulse width 130–1000 μs, frequency 1–145 Hz, on-time 7–270 seconds, off-time 0.3–180 minutes) via a helical electrode coiled around the left cervical vagus nerve. Battery life ranges from 5 to 10 years depending on stimulation parameters, after which the pulse generator is replaced under local anaesthesia.

Transcutaneous VNS (tVNS) devices — most notably the gammaCore (electroCore) system — offer non-invasive vagal stimulation through the cervical skin without surgical implantation, approved for acute and preventive treatment of episodic cluster headache and migraine.

VNS therapy has regulatory approval and published clinical evidence across several neurological and psychiatric indications:

• Drug-resistant focal (partial-onset) epilepsy: The primary indication. Defined as failure to achieve sustained seizure freedom despite adequate trials of two or more appropriate antiepileptic drugs (AEDs) at tolerated doses. Approximately 30% of all people with epilepsy have drug-resistant disease; VNS is indicated when surgical resection is not feasible, has failed, or is declined by the patient.
• Lennox-Gastaut syndrome (LGS): A severe, childhood-onset epileptic encephalopathy characterised by multiple seizure types, cognitive impairment, and slow spike-wave EEG pattern. VNS is approved and effective for reducing drop attacks and tonic seizures in LGS patients who have failed multiple AEDs.
• Treatment-resistant depression (TRD): Defined as inadequate response to two or more adequate antidepressant trials (different pharmacological classes) of appropriate dose and duration over at least 2 years. VNS is used as an adjunctive treatment alongside ongoing antidepressant pharmacotherapy. Antidepressant effects develop progressively over 12–24 months — a slower time course than epilepsy benefits.
• Cluster headache — transcutaneous VNS (gammaCore): gammaCore is FDA-approved for acute treatment of episodic cluster headache (self-administered to the cervical skin) and for prophylaxis of episodic cluster headache. The ACT1 and ACT2 randomised trials demonstrated superiority over sham stimulation for acute cluster headache abort rates.
• Migraine — transcutaneous VNS: gammaCore is FDA-cleared for prevention of migraine in adults. Trial evidence includes the PRESTO and EVENT studies demonstrating reduction in migraine days per month vs sham.
• Acute ischaemic stroke — investigational: The MicroTransponder Vivistim system (paired VNS during upper limb rehabilitation) has shown benefit for chronic upper limb motor deficit after ischaemic stroke (FDA breakthrough device designation 2021). The TREVO trial and related studies are evaluating broader acute stroke applications of cervical VNS.
• Inflammatory conditions — investigational: Rheumatoid arthritis and inflammatory bowel disease are being evaluated in early-phase trials using implanted VNS (SetPoint Medical bioelectronic implant). Not yet approved for inflammatory indications.

Careful patient selection for implanted VNS therapy is essential to maximise benefit and ensure safety:

• Epilepsy eligibility — FDA and clinical criteria: Patients must have partial-onset or generalised epilepsy refractory to at least 2 appropriate AEDs (different mechanisms of action) at adequate doses for adequate duration. FDA approval is for patients aged 4 years and above; in clinical practice, VNS is used in younger children in selected centres under specialist guidance. Patients should have been evaluated by a comprehensive epilepsy centre, with the possibility of resective surgery considered and either excluded or declined.
• Epilepsy surgery evaluation: Before VNS is recommended, patients with focal epilepsy should undergo comprehensive presurgical evaluation including video-EEG monitoring, MRI, neuropsychological testing, and functional mapping (PET, SPECT, or fMRI as indicated) to determine whether resective surgery — which offers the possibility of seizure freedom — is a viable option. VNS does not offer seizure freedom and is inferior to successful epilepsy surgery in that regard.
• Treatment-resistant depression eligibility: Diagnosis of major depressive disorder or bipolar I/II depression in a depressive episode, with inadequate response to at least 4 adequate antidepressant treatments (two or more different pharmacological classes) over at least 2 years. Patients must be maintained on stable antidepressant regimen during VNS therapy.
• Contraindications: Prior left cervical vagotomy (destroys the target nerve), bilateral or left cervical surgery that may have damaged the vagus nerve, known cardiac arrhythmia (particularly bradyarrhythmia — relative contraindication given cardiac efferent vagal fibers), active pharyngeal or laryngeal pathology, and certain implantable cardiac devices that may interact electromagnetically. Pregnancy is a relative contraindication due to limited safety data.
• MRI considerations: Current LivaNova VNS systems are MRI-conditional under specific protocols (transmit-receive head coil only for the AspireSR; broader MRI compatibility for the SenTiva). Full-body MRI is generally contraindicated with older VNS models. Patients requiring frequent body MRI should discuss this with their neurologist before implantation.

VNS therapy is available in implanted and non-invasive (transcutaneous) forms, with several programming approaches:

• Implanted VNS — LivaNova AspireSR (Model 106): The leading closed-loop VNS device. The AspireSR incorporates a cardiac-based seizure detection algorithm (AutoStim) that uses a photodiode in the pulse generator to continuously monitor heart rate. During a seizure, ictal tachycardia (the rapid heart rate increase occurring in over 80% of complex partial and generalised tonic-clonic seizures) triggers an automatic burst of stimulation — delivering therapy at seizure onset without patient activation. This closed-loop feature is in addition to scheduled background stimulation and patient-activated magnet use.
• Implanted VNS — LivaNova SenTiva (Model 1000): Next-generation device offering smaller footprint, improved battery longevity, day programming (scheduled stimulation restricted to waking hours to improve tolerability during sleep), and enhanced physician and patient interaction through the SenTiva app ecosystem.
• Standard (open-loop) stimulation programming: Baseline VNS parameters are set at implant and titrated over 3–6 months. Typical clinical targets: output current 1.25–3.0 mA; pulse width 250–500 μs; frequency 20–30 Hz; on-time 30 seconds; off-time 5 minutes. Parameters are adjusted at outpatient visits using a programming wand communicating with the implanted device.
• Patient-activated magnet stimulation: All VNS patients receive a handheld magnet. Swiping the magnet over the pulse generator delivers an immediate extra stimulation burst — used by patients to abort seizures at onset (if there is an aura) or by caregivers during seizures to attempt early termination.
• Transcutaneous VNS (tVNS) — gammaCore (electroCore): Handheld, non-invasive device applied to the neck overlying the vagus nerve. Delivers electrical stimulation transcutaneously for 2-minute treatment cycles. Self-administered; no surgery required. Approved for episodic cluster headache (acute and preventive) and migraine prevention. Not equivalent in efficacy to implanted VNS for epilepsy.
• Transcutaneous auricular VNS (taVNS): Stimulation of the auricular branch of the vagus nerve (cymba conchae of the ear pinna). Research device used in clinical trials for depression, tinnitus, and atrial fibrillation. Not commercially approved in most markets.

VNS therapy provides clinically meaningful benefits across its approved indications, with evidence demonstrating progressive improvement over time:

• Seizure frequency reduction in epilepsy: The defining efficacy metric is the responder rate — defined as 50% or greater reduction in seizure frequency. Large registry studies and meta-analyses consistently report approximately 50% of patients achieving responder status (the "50/50 rule"). Importantly, outcomes improve progressively over time: 2-year responder rates (approximately 55–60%) are higher than 6-month rates (approximately 35–45%). Complete seizure freedom is rare (1–5%) but does occur.
• Progressive benefit over time: Unlike many pharmacological treatments where the effect is immediate or rapidly plateau, VNS benefits continue to accumulate over 2–3 years of therapy. Patients are counselled not to judge VNS efficacy by early response alone.
• Mood improvement — independent antidepressant effect: Patients with epilepsy receiving VNS consistently report mood improvement independent of seizure control. This observation led to the TRD indication; in TRD patients, antidepressant response rates improve progressively over 12–24 months of treatment, with remission rates of 20–30% at 2 years compared with 12% in the treatment-as-usual comparator arm (D-23 study).
• Potential SUDEP risk reduction: Sudden unexpected death in epilepsy (SUDEP) is a devastating complication of drug-resistant epilepsy. The cardiac monitoring function of the AspireSR AutoStim — detecting ictal tachycardia and delivering stimulation — may reduce SUDEP risk through combined anti-ictal and cardiac rhythm modulation effects. Registry data suggest lower SUDEP rates in VNS users vs matched controls, though prospective evidence is not yet definitive.
• Cognitive and quality-of-life benefits: Unlike many AEDs, VNS does not cause cognitive sedation, ataxia, or behavioral side effects. Patients report improved alertness and quality of life independent of seizure reduction, with significant gains in attention and processing speed in some studies.
• Non-pharmacological, reversible therapy: VNS is reversible — the device can be deactivated at any time without neurological consequence, and the electrode can be removed or replaced. It adds no hepatic metabolic burden and has no teratogenic risk profile (relevant for women of childbearing age).

VNS therapy has a well-characterised safety profile. Risks fall into surgical, stimulation-related, and device-related categories:

• Surgical risks of implantation: As with any cervical surgical procedure, risks include wound infection (approximately 1%), haematoma, inadvertent vagal nerve injury (rare — producing persistent hoarseness or dysphagia), and anaesthetic complications. Horner syndrome from surgical trauma to the cervical sympathetic chain is a rare but reported complication. Operative mortality is not a reported risk in published series.
• Voice change and hoarseness: The most common stimulation-related side effect — reported by 25–40% of patients during the stimulation ON phase. Caused by stimulation of the recurrent laryngeal nerve branch. Typically intermittent and limited to the stimulation cycle (30 seconds every 5 minutes). Often diminishes with dose reduction; persistent hoarseness affecting professional voice users (singers, teachers) should be discussed pre-implantation.
• Cough and throat discomfort: Cough and paresthesia in the throat or neck during stimulation are reported by 10–20% of patients. Generally mild and well-tolerated; may diminish over time as the patient habituates.
• Dyspnoea and respiratory effects: Rare; risk is higher at higher output currents. Reduces airway protective reflexes during stimulation in some patients with pre-existing obstructive sleep apnoea (OSA). Screening for OSA before implantation is recommended; if present, sleep apnoea should be treated before or alongside VNS.
• Bradycardia and asystole: Rare, occurring most commonly during intraoperative testing of the lead before generator connection. The left vagus was chosen (rather than right) specifically because it carries fewer cardiac efferent fibres, but cardiac effects during system testing are monitored with continuous ECG. Post-implant bradycardia from chronic stimulation is very rare at therapeutic parameters.
• Lead fracture and device malfunction: Over the device lifespan (5–10 years), lead wire fracture at the electrode-cable junction occurs in a small percentage of patients and may require surgical lead replacement. Device impedance checks at each programming visit detect early lead integrity issues.
• MRI restrictions: Older VNS models (models 100, 101, 102, 103) are MRI-conditional only with strict protocols. SenTiva offers improved MRI compatibility but still requires specialist assessment before any MRI. Full-body MRI is contraindicated with many older generators.

Structured long-term follow-up is essential to optimise VNS programming, monitor battery status, and assess clinical response:

• Initial programming visits: The VNS system is activated 2 weeks post-surgery, after wound healing, at an initial low output current (typically 0.25 mA) to minimise early stimulation side effects. Parameters are then titrated at 2-week to monthly intervals over the first 3–6 months, increasing output current by 0.25 mA increments toward the target therapeutic range (typically 1.25–2.5 mA).
• AutoStim threshold setting (AspireSR): The ictal tachycardia detection threshold is set by programming a heart rate increment above the patient's resting baseline that will trigger AutoStim delivery. This requires review of baseline heart rate data and calibration against the patient's own documented ictal tachycardia pattern from prior video-EEG monitoring where available.
• Response assessment: Seizure diaries (paper or app-based) are maintained by patients and reviewed at each visit. A reliable 3-month seizure baseline before implantation is established to allow post-implant comparison. Response is formally assessed at 6 and 12 months, and annually thereafter.
• Battery status monitoring: At each programming visit, the pulse generator battery status is automatically read by the programming wand. Battery end-of-service predictions are provided to allow timely planning of generator replacement surgery. Generator replacement is a minor surgical procedure under local anaesthesia, replacing the pulse generator while leaving the existing vagal electrode in place.
• Magnet use education: Patients and caregivers are educated on magnet use — including how to swipe the magnet during a seizure aura to trigger additional stimulation, and how to hold the magnet over the device to temporarily suspend stimulation (useful if voice change is prominent at an important social or professional moment).
• Long-term monitoring for depression indication: TRD patients are assessed using validated rating scales (HDRS-17, MADRS) at baseline and every 3 months. Antidepressant medication is continued and may be adjusted by the treating psychiatrist in collaboration with the VNS team. Response is not expected until 12–24 months post-implant.
• Annual neurological review: Comprehensive annual review encompasses seizure frequency and severity, AED status and tolerability, quality of life (QOLIE-89 or similar), side-effect profile, and device integrity assessment.

VNS therapy involves significant upfront device and surgical costs, with variable funding landscape across healthcare systems:

• Device cost: The LivaNova VNS pulse generator costs approximately $15,000–$25,000 USD at list price. The vagal lead and accessories add a further $2,000–$5,000. Generator replacement every 5–10 years incurs recurring device costs. The transcutaneous gammaCore device uses a subscription-based model with per-treatment dose pricing (approximately $600 per month in the US).
• Surgical implantation fees: Neurosurgical implantation requires a 60–90 minute procedure under general anaesthesia, typically as an overnight hospital admission. In the United States, total procedural costs (surgeon, anaesthesia, facility) range from $20,000–$50,000 at private rates. UK NHS implantation is funded for eligible patients through specialist neuroscience centres.
• Ongoing programming visits: Monthly visits in the first year (approximately 6–8 visits), then quarterly, then annually. These specialist neurology outpatient appointments add ongoing cost — estimated $2,000–$5,000 annually in the US private sector, covered by NHS in the UK.
• Insurance and public funding: United Kingdom: NHS funds VNS implantation for drug-resistant epilepsy meeting NICE criteria (TA631, 2022). Scotland: SMC guidance also supports funding. United States: Medicare covers VNS for epilepsy; since 2019, Medicare also covers VNS for TRD under specific criteria. Private insurance coverage for TRD varies widely and may require extensive prior authorisation documentation.
• Country-specific availability: VNS therapy is available in most high-income countries through specialist neurology and neurosurgery centres. In lower-middle income countries, device cost limits access — medical tourism for VNS implantation (to India, Thailand, or Eastern Europe) is an emerging pathway for patients from countries without public funding.
• Cost-effectiveness: Published health economic analyses from the UK, Netherlands, and Australia consistently demonstrate that VNS therapy is cost-effective compared with continued AED treatment alone in drug-resistant epilepsy, due to reduced seizure-related emergency admissions, reduced seizure-related injury costs, and improved caregiver burden.

Several neuromodulation, surgical, and pharmacological alternatives are considered when VNS is not appropriate, not tolerated, or has been inadequately effective:

• Resective epilepsy surgery: For patients with well-localised focal epilepsy and a resectable epileptogenic zone, surgery (temporal lobectomy, selective amygdalohippocampectomy, extratemporal resection) offers the possibility of seizure freedom (60–70% for temporal lobe surgery) — an outcome that VNS does not achieve. Resective surgery should be considered and evaluated before VNS is offered as an alternative.
• Responsive neurostimulation (RNS — NeuroPace): A closed-loop cortical stimulation device that continuously monitors EEG from depth or cortical strip electrodes placed at the seizure focus and delivers stimulation only when seizure activity is detected. Unlike VNS, RNS targets the seizure focus directly and is particularly suited to patients with well-defined focal onset. RNS is complementary rather than competing with VNS in multifocal disease.
• Deep brain stimulation (DBS) for epilepsy: Thalamic stimulation (anterior nucleus of the thalamus — ANT-DBS, SANTE trial; centromedian nucleus — CM-DBS for LGS) provides another neuromodulation option for drug-resistant epilepsy. DBS requires accurate stereotactic electrode placement and is offered in specialist centres with DBS programs.
• Dietary therapy — ketogenic diet and variants: The ketogenic diet (high fat, very low carbohydrate) and its modified variants (modified Atkins diet, low glycaemic index diet) are established non-pharmacological seizure reduction approaches, particularly effective in children. They can be combined with VNS therapy.
• Electroconvulsive therapy (ECT) for TRD: ECT remains the most effective treatment for severe treatment-resistant depression, achieving response rates of 50–70% even in patients who have failed multiple pharmacological treatments. ECT acts more rapidly than VNS (weeks vs months) and is preferred in acutely suicidal or severely unwell patients. The two treatments address different timescales and severity spectra of TRD.
• Transcranial magnetic stimulation (TMS) for TRD: Repetitive TMS (rTMS) to the left dorsolateral prefrontal cortex is FDA-cleared for TRD. Non-invasive, requires no surgery, but involves daily treatments for 4–6 weeks. Response rates (30–40%) are lower than VNS at 2 years but the non-invasive profile and faster onset may make it preferable for certain TRD patients.
• Pharmacological escalation and augmentation: Lithium augmentation, MAO inhibitor trials, ketamine or esketamine (Spravato, intranasal — FDA-approved for TRD), and clozapine in bipolar TRD are pharmacological alternatives or adjuncts to VNS in the TRD pathway.