Vagus nerve stimulation (VNS) spans a range of procedures from surgically implanted pulse generators to handheld transcutaneous devices available without prescription. The field has evolved substantially since the first FDA approval in 1997, with newer devices offering closed-loop feedback, smaller form factors, and expanding indications beyond epilepsy. This guide covers the specific devices, surgical procedures, programming parameters, outcome data, and practical considerations that matter for anyone evaluating VNS — whether as a patient, carer, or clinician.

Implanted VNS: Devices and Indications

The implanted VNS system — manufactured by LivaNova (formerly Cyberonics) — consists of a titanium-encased pulse generator implanted in the left chest wall and a helical lead electrode wrapped around the left cervical vagus nerve. The system has been through several generations. The VNS Therapy System received FDA approval for drug-resistant epilepsy in 1997 and for treatment-resistant depression (TRD) in 2005. The AspireSR model (2015) added automatic stimulation triggered by heart rate changes associated with seizures. The SenTiva model (2017) reduced generator size by 42 percent, added scheduled dosing changes and a patient-facing programming interface. The most recent models incorporate closed-loop algorithms that detect ictal tachycardia and deliver responsive stimulation within seconds of seizure onset — a significant advance over the fixed-schedule stimulation of earlier devices.

FDA-approved indications are epilepsy (adjunctive therapy for patients 4 years and older with partial-onset seizures refractory to medication) and treatment-resistant depression (adjunctive long-term therapy for patients 18 and older who have failed four or more adequate antidepressant treatments). Off-label use in clinical trials and specialist centres extends to cluster headache, post-stroke rehabilitation, rheumatoid arthritis (via the inflammatory reflex), heart failure, tinnitus, and post-traumatic stress disorder.

The Implantation Procedure

Implanted VNS is a same-day or overnight surgical procedure performed under general anaesthesia, typically lasting 60 to 90 minutes. Two incisions are made: one horizontal incision (approximately 5 cm) along a skin crease in the left neck to expose the carotid sheath, and one incision (approximately 5 cm) below the left clavicle to create a subcutaneous pocket for the generator.

Surgical steps: The surgeon identifies the left vagus nerve within the carotid sheath, between the internal jugular vein and common carotid artery. The nerve is carefully dissected free over a 3 to 4 cm length without damaging it. Three helical coils from the lead electrode are wrapped around the nerve — a positive contact, a negative contact, and an anchor coil. The lead is tunnelled subcutaneously from the neck incision to the chest pocket, connected to the generator, and the generator is seated in the pocket. Impedance testing confirms electrode-nerve contact and system integrity before closure.

The left vagus nerve is chosen specifically because the right vagus has stronger cardiac efferent innervation to the sinoatrial node. Stimulating the right side carries higher risk of bradycardia and cardiac arrhythmia. The left vagus still carries cardiac fibres (primarily to the atrioventricular node), but the cardiac effects of left-sided stimulation are clinically manageable — transient bradycardia occurs in less than 0.1 percent of implantations.

The generator is typically not activated during surgery. Activation occurs at a follow-up visit two to four weeks post-implantation, allowing surgical wounds to heal and reducing the risk of lead displacement from early stimulation.

Programming Parameters

VNS programming is performed by a trained clinician using a proprietary programming wand and tablet interface. The key parameters are output current (milliamps), signal frequency (hertz), pulse width (microseconds), and duty cycle (on-time and off-time in seconds or minutes).

Standard epilepsy protocol: Initial settings are typically conservative — 0.25 mA output current, 30 Hz frequency, 250 microsecond pulse width, 30 seconds on / 5 minutes off duty cycle. Over subsequent visits (every two to four weeks), current is gradually titrated upward in 0.25 mA increments to a target range of 1.0 to 2.0 mA or the maximum tolerated dose. Most patients settle between 1.5 and 2.5 mA. Higher currents (up to 3.5 mA) are occasionally used but limited by side effects — primarily voice alteration and throat discomfort during stimulation periods.

Depression protocol: Parameters for TRD differ slightly. Current is typically lower (0.5 to 1.5 mA), but the duty cycle is more aggressive — 30 seconds on / 3 minutes off rather than the 5-minute off-time used in epilepsy. This reflects the hypothesis that antidepressant effects depend more on cumulative stimulation time per day than on individual pulse intensity. The depression protocol delivers approximately 20 to 25 percent more total stimulation per 24-hour period.

Responsive stimulation (AutoStim): In AspireSR and later models, the device monitors heart rate via the vagus nerve lead itself. When heart rate increases by a preset threshold (typically 20 to 40 percent above baseline, configurable per patient), the device delivers an additional stimulation burst outside the normal duty cycle. This responsive stimulation can shorten seizure duration and reduce severity even when the scheduled stimulation does not prevent seizure onset.

Patients also receive a handheld magnet that, when swiped over the generator, triggers an on-demand stimulation burst. This allows patients or carers to initiate extra stimulation at seizure onset — essentially a non-pharmacological rescue therapy. The magnet can also be used to temporarily inhibit stimulation if side effects become problematic.

Outcomes: What the Evidence Shows

Epilepsy: The pivotal E05 trial (randomised, controlled, 316 patients) demonstrated a median seizure reduction of 28 percent at high stimulation versus 15 percent for low stimulation at 3 months. However, VNS efficacy improves substantially over time — a finding unique to this therapy. Registry data from over 5,000 patients show median seizure reduction of 50 percent or greater in approximately 50 to 60 percent of patients at 2 to 3 years, with some patients continuing to improve through year 5. Complete seizure freedom is uncommon (approximately 5 to 8 percent) but stable when achieved. The time-dependent improvement is thought to reflect progressive neuroplastic remodelling of central autonomic circuits.

Depression: The D-21 registry (795 patients, 5-year follow-up) showed cumulative response rates of 68 percent and remission rates of 43 percent at 5 years — substantially better than the 1-year rates (approximately 30 percent response). As with epilepsy, the antidepressant effect builds over months to years, which has complicated trial design (short trials underestimate efficacy) but represents a genuine and durable therapeutic effect for patients who persist. The mechanism likely involves ascending vagal afferent modulation of the locus coeruleus-noradrenaline system and the raphe nuclei-serotonin system.

Inflammatory conditions: The most intriguing emerging application is in autoimmune inflammation. A landmark 2016 study by Koopman and colleagues (Proceedings of the National Academy of Sciences) demonstrated that implanted VNS significantly reduced disease activity in 17 patients with treatment-resistant rheumatoid arthritis, working through the cholinergic anti-inflammatory pathway (vagal efferents to the spleen via the celiac ganglion, suppressing TNF-alpha production by splenic macrophages). Follow-up studies confirmed persistence of benefit at 2 years. Trials in Crohn’s disease have shown similar promise.

Non-Invasive VNS Devices

Transcutaneous VNS (tVNS) devices deliver electrical stimulation through the skin to vagal afferent fibres without surgery. Two anatomical targets are used: the auricular branch of the vagus nerve (ABVN) in the ear, and the cervical vagus nerve through the neck.

gammaCore (electroCore): A handheld cervical tVNS device — FDA-cleared for acute treatment and prevention of migraine (2018) and cluster headache (2017, 2018). The device is held against the neck over the vagus nerve and delivers 2-minute stimulation doses. For cluster headache prevention, the protocol is three stimulations three times daily. The PREVA trial (randomised, 97 patients) showed a mean reduction of 5.9 cluster attacks per week in the active group versus 2.1 in sham. For migraine, the PREMIUM trial (randomised, 477 patients) showed a modest but significant reduction in monthly migraine days (2.3 day reduction versus 1.8 for sham). gammaCore requires a prescription and uses a per-dose payment model.

NEMOS/tVNS Technologies (various): Auricular tVNS targets the cymba conchae of the outer ear — the only external skin area innervated by the ABVN (mapped by Peuker and Filler, 2002). The NEMOS device (tVNS Technologies, now Parasym) delivers stimulation via a clip electrode in the ear. Auricular tVNS has been studied primarily in epilepsy and depression, with mixed but generally positive results. A 2023 meta-analysis of auricular tVNS for depression (12 RCTs, 600+ participants) found a moderate effect size favouring active stimulation over sham. The FDA has cleared several auricular tVNS devices for anxiety and insomnia.

The non-invasive devices carry minimal risk — mild skin irritation and occasional dizziness are the most commonly reported adverse effects. They lack the continuous, programmed delivery of implanted VNS but offer accessibility and reversibility that make them appropriate for conditions where surgical implantation is not warranted.

Risks and Complications of Implanted VNS

Implanted VNS has a well-characterised safety profile over nearly three decades of clinical use. Surgical complications include infection at the generator or lead site (2 to 3 percent), vocal cord paresis from intraoperative nerve manipulation (1 percent, usually temporary), and Horner syndrome (rare, from sympathetic chain proximity). Lead fracture requiring surgical revision occurs in approximately 3 to 5 percent of patients over 10 years.

Stimulation-related side effects are the most common complaints and include hoarseness or voice alteration during on-periods (66 percent initially, often diminishing), cough (45 percent), throat tingling or pain (30 percent), and dyspnoea sensation (25 percent). These effects are current-dependent and occur only during active stimulation periods. Most patients tolerate them at therapeutic current levels, and they typically diminish over the first 6 to 12 months. Obstructive sleep apnoea can worsen in some patients due to pharyngeal muscle effects during nocturnal stimulation — sleep studies may be warranted if daytime somnolence increases.

Device-specific considerations: The generator battery lasts 3 to 8 years depending on stimulation parameters (higher current and shorter off-times drain faster). Battery replacement requires a brief surgical procedure to swap the generator — the lead electrode remains in place and is not routinely replaced unless damaged. MRI compatibility varies by model and sequence; newer generators are MRI-conditional under specific protocols (1.5T head-only with normal mode stimulation off), but full-body MRI remains contraindicated for most implanted VNS systems.

Patient Selection and Evaluation

For epilepsy, VNS is typically considered after failure of two or more appropriately chosen anti-seizure medications at adequate doses, when resective surgery is not an option (non-localised seizure focus, eloquent cortex, patient preference). Pre-surgical evaluation includes video-EEG monitoring, brain MRI, neuropsychological testing, and cardiac screening (ECG, cardiac history review). Left vocal cord function should be confirmed pre-operatively via laryngoscopy, as unilateral vocal cord paresis is a relative contraindication — stimulation could affect the remaining functioning cord.

For depression, VNS is reserved for truly treatment-resistant cases — typically defined as failure to achieve adequate response after four or more antidepressant trials from at least two different classes, plus failure of psychotherapy and, in many protocols, ECT. The slow onset of VNS antidepressant effect (months to years) means it is not appropriate for acute suicidality without concurrent acute interventions.

Absolute contraindications include bilateral or left cervical vagotomy (no nerve to stimulate), and conditions where general anaesthesia poses unacceptable risk. Relative contraindications include progressive neurological disease, significant cardiac conduction abnormalities, and active peptic ulcer disease (vagal stimulation increases gastric acid secretion).

Emerging Research and Future Directions

The VNS field is advancing on multiple fronts. Bioelectronic medicine — the concept of treating disease with targeted electrical impulses rather than drugs — positions VNS at the centre of a paradigm shift. Current research frontiers include: closed-loop systems that read neural biomarkers (EEG patterns, heart rate variability, cortisol surges) and adjust stimulation in real-time; miniaturised implantable devices that can be injected rather than surgically placed; organ-specific stimulation protocols targeting individual vagal fascicles (the vagus contains approximately 100,000 fibres serving different organs, and selective fascicular stimulation could treat the target organ without affecting others); and combination therapy pairing VNS with rehabilitation exercises for post-stroke motor recovery, where a 2021 Lancet trial demonstrated that VNS paired with intensive upper-limb therapy produced significantly greater improvement than therapy alone in chronic stroke patients.

Vagus nerve stimulation has matured from an experimental curiosity into a validated therapeutic platform with expanding applications. Whether considering implanted VNS for drug-resistant epilepsy, non-invasive devices for migraine, or following the inflammatory reflex research, understanding the specific devices, procedures, and evidence is essential. For a broader understanding of vagal anatomy and the science that underpins these interventions, explore our comprehensive resource. Get My Vagus Nerve & Me eBook and deepen your knowledge of this remarkable nerve.

Keep reading

The Impact of Vagus Nerve Stimulation on Patients5 min read

How to Assess the Vagus Nerve: A Comprehensive Guide14 min read

The Vagus Nerve Stimulation Procedure: A Comprehensive Guide14 min read