When is vagus nerve stimulation (VNS), RNS, or DBS considered for epilepsy?

Published by Unseen Progress, an independent publisher of caregiver research. Last reviewed 2026-05-10. Part of the epilepsy caregiver research overview.

Short answer. Vagus nerve stimulation (VNS), responsive neurostimulation (RNS), and deep brain stimulation (DBS) are three FDA-approved implantable devices for drug-resistant epilepsy that is not a candidate for resective surgery, or where prior resection has been incomplete. They differ in what they stimulate, when they stimulate, and who qualifies. Across the pivotal trials, all three produce a roughly 40–75% median seizure-frequency reduction at long-term follow-up, with seizure freedom rates around 5–15% and meaningful responder rates (≥50% reduction) around 50–65% by 3–5 years (Ben-Menachem et al., 2015; Bergey et al., 2015; Fisher et al., 2010; Salanova et al., 2021). They are options for the third of epilepsy patients who do not become seizure-free on medication, especially those who are not surgical candidates.

Where these devices sit in the treatment hierarchy

The decision tree for drug-resistant epilepsy, simplified:

1. Establish drug-resistance under the ILAE two-drug rule (Kwan et al., 2010). 2. Refer to a comprehensive epilepsy center for presurgical evaluation: video-EEG, MRI, neuropsychology, sometimes intracranial monitoring. 3. If a seizure focus is localised and resectable in a non-eloquent area, resective surgery is typically first-line — it has the highest seizure-freedom rate of any drug-resistant option (Wiebe et al., 2001; Engel, 2008). 4. If resective surgery is not appropriate — multifocal seizures, focus in eloquent cortex, bilateral involvement, or patient/family preference — neurostimulation enters the discussion. 5. Dietary therapy (ketogenic, modified Atkins) is also part of the discussion, with different indications.

VNS, RNS, and DBS are not in competition with resective surgery in most candidacy decisions. They occupy the space where resection cannot do the job.

Vagus nerve stimulation (VNS)

What it is. A pulse generator implanted under the skin of the upper chest, connected by a lead wire to the left vagus nerve in the neck. It delivers programmed electrical stimulation to the vagus nerve at intervals (e.g., 30 seconds on, 5 minutes off), with the mechanism of action incompletely understood but thought to involve modulation of brainstem and thalamic networks.

Approved indications. FDA-approved for drug-resistant focal epilepsy in patients aged 4 and older. The longest clinical track record of the three (approved in 1997).

Trial data. Pooled long-term follow-up across multiple studies shows mean seizure reduction of roughly 40–50% at 2 years, with responder rates (≥50% reduction) around 50% by 3–5 years and continued incremental improvement with longer use (Ben-Menachem et al., 2015; Englot et al., 2016). Seizure freedom rates are low — typically under 10%. The device often produces a gradual response rather than a binary on/off effect.

Candidacy and considerations.

  • Most permissive candidacy of the three — does not require focus localisation.
  • Newer models include cardiac-based seizure detection (AutoStim) that delivers extra stimulation in response to ictal tachycardia.
  • Common side effects: voice change, cough, throat discomfort during stimulation cycles — typically tolerable, often abating over time.
  • Lower up-front procedural risk than intracranial devices.
  • Used as a long-term adjunct to medication, not a replacement.

Responsive neurostimulation (RNS)

What it is. A small generator implanted in the skull, connected to one or two depth or strip electrodes placed directly at the seizure focus or foci. It continuously monitors local brain activity, detects patterns of abnormal electrical activity that precede the patient's seizures, and delivers targeted electrical stimulation to interrupt them — a closed-loop system.

Approved indications. FDA-approved (2013) for drug-resistant focal epilepsy with up to two well-localised seizure foci in patients aged 18 and older.

Trial data. Long-term follow-up of the pivotal trial cohort showed median seizure reduction of 75% at 9 years, with 18% of patients seizure-free for at least one period of six months or longer at long-term follow-up (Nair et al., 2020; Bergey et al., 2015). Notably, the response improves over years rather than declining — likely reflecting both device optimisation and neuromodulatory effects.

Candidacy and considerations.

  • Requires focus localisation — best for patients with one or two identifiable seizure foci, often after intracranial EEG monitoring.
  • Particularly useful when foci are in eloquent cortex where resection would cause unacceptable deficits.
  • Provides longitudinal electrocorticography data that informs medication and stimulation decisions — the device is also a chronic monitoring instrument.
  • Intracranial procedure with associated risks (haemorrhage, infection) — lower than open resection but higher than VNS.

Deep brain stimulation (DBS)

What it is. A generator implanted under the skin of the chest, connected to deep electrodes placed in specific brain targets — most commonly the anterior nucleus of the thalamus (ANT) for epilepsy. Stimulation is open-loop (continuous or duty-cycled), not responsive to detected events.

Approved indications. FDA-approved (2018) for drug-resistant focal epilepsy in patients aged 18 and older, including those with multifocal seizures or seizures not amenable to RNS.

Trial data. The pivotal SANTE trial (Fisher et al., 2010) and its long-term follow-up (Salanova et al., 2021) showed median seizure reduction of 41% at 1 year, rising to 75% at 7 years, with responder rates around 75% at 7 years and seizure freedom in roughly 18% at some point during follow-up.

Candidacy and considerations.

  • Appropriate when foci are multifocal, bilateral, or not amenable to RNS or resection.
  • Long established in other neurological conditions (Parkinson's disease, essential tremor), with well-characterised surgical and stimulation expertise.
  • Side effects include depression and memory complaints in some patients; ongoing monitoring is important.
  • Like RNS, requires intracranial implantation with associated procedural risk.

Choosing among them

The choice is not a free shopping menu — it is a candidacy assessment, run at a comprehensive epilepsy center, that maps the patient's seizure pattern, focus localisation, and prior surgical history to the device the evidence supports.

Simplified candidacy heuristic:

  • One or two well-localised foci, in or near eloquent cortex, not resectable: RNS is often the strongest fit.
  • Multifocal or bilateral seizures, not localisable to one or two foci: DBS (ANT) or VNS, depending on profile.
  • Focal epilepsy without clear localisation, or paediatric patient, or wanting the most permissive candidacy: VNS.
  • Prior incomplete resective surgery with residual seizures from the same region: RNS often considered.

These are not rigid rules; they are starting points. The decision is made by an epileptologist with full presurgical workup data.

What the data does and does not promise

Across the three devices, the consistent picture is:

  • Seizure freedom is uncommon but not impossible. Roughly 5–18% of patients achieve some sustained seizure freedom, with the higher rates seen in the longest follow-ups of RNS and DBS.
  • Responder rates are substantial. Roughly half of patients achieve at least a 50% reduction in seizures by 3–5 years, with the proportion growing in long-term follow-up.
  • Response is gradual. Unlike a successful medication that often shows its effect within months, neurostimulation typically improves over years. Early non-response is not the same as ultimate non-response.
  • Quality of life improves even where seizure freedom is not achieved. Reduction in severity, in injury rate, and in the unpredictability of seizures matters for daily function even when complete control is not reached.

What the devices do not promise: medication freedom (most patients remain on ASMs), zero side effects, or an immediate response. The realistic frame for the family is a meaningful incremental tool that compounds over time, not a binary cure.

What the research suggests doing

1. If two adequate ASM trials have failed, ask for referral to a comprehensive epilepsy center. 2. Complete a full presurgical evaluation — video-EEG monitoring, MRI, neuropsychology, and intracranial monitoring if indicated. The data shapes which option fits. 3. Ask explicitly whether resective surgery is on the table. If yes, that is typically the first-line discussion. If no, neurostimulation enters the conversation. 4. Discuss the three devices with the epileptologist in terms of this patient's focus pattern and candidacy, not abstract comparison. 5. Set expectations on timeline. Neurostimulation response typically grows over 1–3 years. Plan the diary and follow-up cadence accordingly. 6. Continue medication adherence and seizure logging. Devices and drugs work together; the diary remains the input the optimisation runs on.

Related questions

References

  • Ben-Menachem, E., Revesz, D., Simon, B. J., & Silberstein, S. (2015). Surgically implanted and non-invasive vagus nerve stimulation: a review of efficacy, safety and tolerability. European Journal of Neurology, 22(9), 1260–1268.
  • Englot, D. J., Rolston, J. D., Wright, C. W., Hassnain, K. H., & Chang, E. F. (2016). Rates and predictors of seizure freedom with vagus nerve stimulation for intractable epilepsy. Neurosurgery, 79(3), 345–353.
  • Bergey, G. K., Morrell, M. J., Mizrahi, E. M., et al. (2015). Long-term treatment with responsive brain stimulation in adults with refractory partial seizures. Neurology, 84(8), 810–817.
  • Nair, D. R., Laxer, K. D., Weber, P. B., et al. (2020). Nine-year prospective efficacy and safety of brain-responsive neurostimulation for focal epilepsy. Neurology, 95(9), e1244–e1256.
  • Fisher, R., Salanova, V., Witt, T., et al. (2010). Electrical stimulation of the anterior nucleus of thalamus for treatment of refractory epilepsy. Epilepsia, 51(5), 899–908.
  • Salanova, V., Sperling, M. R., Gross, R. E., et al. (2021). The SANTE study at 10 years of follow-up: effectiveness, safety, and sudden unexpected death in epilepsy. Epilepsia, 62(6), 1306–1317.
  • Kwan, P., Arzimanoglou, A., Berg, A. T., et al. (2010). Definition of drug resistant epilepsy: ILAE consensus. Epilepsia, 51(6), 1069–1077.

---

Unseen Progress publishes long-form caregiver research and builds research-backed daily trackers for the families covered. See the full epilepsy caregiver research overview for the complete framework.