Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Efficacy of vagal nerve stimulation in children with medically intractable epilepsy

Three Part Question

In [children with intractable or refractory or difficult epilepsy], does [vagal nerve stimulation] reduce [seizure frequency or rate]?

Clinical Scenario

A 12-year-old boy was admitted to our department as part of our epilepsy surgery protocol. He was admitted for video EEG and for ultimate consideration for vagal nerve stimulator (VNS) insertion. He had a background of intractable epilepsy which had not responded to multiple antiepileptic medications.

Search Strategy

The search was performed independently by two people.

An advanced search of National Library for Health, AMED, BNI, EMBASE, HMIC, MEDLINE, PsycINFO, CINAHL and Health Business Elite was carried out.

Key words were: children, intractable epilepsy, vagal nerve stimulation, seizure frequency and seizure rate.

Search Outcome

Fifteen papers were found, six were relevant to the question, and one was excluded due to small size.

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
60 children with pharmaco-resistant epilepsy Age range 3.5–18 years, median 15 yearsProspective study (level 2b)Seizure frequencyLength of follow-up: 18 months. Patients with complete remission: not known. Significant (>50%) reduction: 29.4%Reduction (30–50%): 73.3%No reduction: 0 Side effects/morbidities Voice alteration: 21.7% Infection: 18.3% Cough: 25.0% Reasonably large prospective study. These findings were similar to those reported for adults. Specific disorders such as absence epilepsy were not studied. A specific stimulation paradigm was used in these children, but other parameters might be more or less effective in improving seizure control. In no case did adversity cause withdrawal of this therapy
Patwardhan et al,
38 children with medically refractory epilepsyAge range 11 months to 16 years Median age 8 yearsRetrospective(level 2bSeizure frequencyLength of follow-up: 18 months; median: 12 monthsPatients with complete remission: not knownSignificant (>50%) reduction: 68%Reduction (30–50%): not knownNo reduction: 18%Side effects/morbidities Transient hoarseness: 54.3% Infection: 2.9% Cough: 14.3%It was not stated why children in this study had greater seizure reductions compared to other published reportsNo limitations were caused by the design of a prospective study regarding stimulation parameters. Not enough information given about maximal stimulation parameters, magnet usage and changing level of amplitude
Murphy et al,
100 patient under went vagal nerve stimulator (VNS) insertion at an average age of 10.4 yearsRetrospective(level 2b)Seizure frequencyLength of follow-up: mean 2.7 yearsPatients with complete remission: 17%Significant (>50%) reduction: 45%Reduction (30–50%): not knownNo reduction: 24%Side effects/morbidities Increased seizure frequency: 5% (1% had 150% increase in seizure frequency but improved performance) Abscesses requiring removal of VNS: 3% Voice changes: 1%Specifically selected a large paediatric group and included children younger than 12 and analysed outcomes by age category as well as overall. Retrospective study reviewing the first 100 patients to have had the implant inserted. Responder rate of 45% in keeping with other studies25% had device removed due to side effects, cosmetic reasons or lack of effect. 5 patients had increased seizure frequency as a result of VNS.Included patients who had had epilepsy surgery
Bremer et al,
60 children with pharmaco-resistant epilepsy underwent VNS insertion at 4–16 years of ageRetrospective study(level 2b)Seizure frequencyLength of follow-up: mean 2.5 yearsPatients with complete remission: not knownSignificant (>50%) reduction: 43%Reduction (30–50%): not knownNo reduction: not knownSide effects/morbidities Minor side effects: 50%Full text not English
Benifla et al,
41 children, mean age 13Retrospective review(level 2b)Seizure frequencyLength of follow-up: mean 31 monthsPatients with complete remission: 0%Significant (>50%) reduction: 43%Reduction (30–50%): 7%No reduction: 50%Side effects/morbidities Side effects: 50% Wound infections: 2 (5%) Complained of inability to sing: 1 (2.5%) Pain over generator and leads: 1 (2.5%) Electrode failure after 3.5 and 4 years: 2 (5%) Vocal alterations: 6 (15%) Excessive coughing: 6 (15%) Sleep problems: 1 (2.5%) Worsened behaviour: 3 (7.5%)Wide variation in follow-up length. Varying spectrum of seizure types and co-morbidities. Included patients who had already had epilepsy surgery. Retrospective review of case notes. Despite side effects only a 5% removal rate noted


As approximately one third of patients with epilepsy will have persistent seizures despite maximal antiepileptic drug therapy or will experience unacceptable side effects, alternative therapeutic measures are sought. Surgical resection for epilepsy has been shown to be effective only in selected patients and also carries significant morbidities (ILAE Commission).

Vagal nerve stimulation uses an implanted stimulator that sends electric impulses to the left vagus nerve in the neck via a lead wire implanted under the skin. The vagus nerve arises from the medulla and carries both afferent and efferent fibres. The afferent vagal fibres connect to the nucleus of the solitary tract which in turn projects connections to other locations in the central nervous system. Little is understood about exactly how vagal nerve stimulation modulates seizure control, but proposed mechanisms include alteration of norepinephrine release by projections of solitary tract to the locus cerulean, elevated levels of inhibitory GABA related to vagal stimulation and inhibition of aberrant cortical activity by reticular system activation (Ghanem).

VNS insertion takes place under a general anaesthetic and involves implantation of the generator device. Following the procedure the output current is slowly titrated upwards over a period of months. The most commonly reported complication is infection with rates of between 3% and 7% being reported. However, more minor side effects have also been cited, including cough, voice change and change in behaviour (Benifla)

Removal was relatively infrequent but occurred following infection, because of dissatisfaction with cosmetic appearance and in those for whom seizures worsened or showed no improvement. The risk of a general anaesthetic is a considerable factor as many of the VNS candidates have significant comorbidities.The children included in these studies differ widely in age. In addition, they vary greatly in terms of seizure subtype, which ranged from Lennox-Gastaut syndrome and Landau Kleffner syndrome to absence seizures to seizures occurring after encephalitis or after head injury. Children also differed widely with a range of disability and co-morbidity. Many of the studies included in their cohort a number of children who had already had epilepsy surgery previously. This wide variation means that the results from the various studies are very difficult to compare. In addition, the many differences between these children make confounding factors difficult to exclude. It also means that the success of VNS in epilepsy subtypes is hard to pin down. However, particular success has been reported in Lennox-Gastaut and tuberous sclerosis (Labar, Parain).

The children included are at the severe end of the spectrum with medically refractory seizures. There is a global description of improvement in the quality of life in the majority of participants. However, there is an inherent placebo effect in such a surgical study, particularly given the severe nature of the condition and its effect on everyday life. Furthermore, such single-armed uncontrolled studies create an unavoidable bias and make results difficult to extrapolate.

It should be noted that the natural history of some epilepsy encephalopathies means that some of the children may have shown spontaneous improvement irrespective of the VNS. This is particularly important in those cases where the children show improvement as a late response to VNS insertion.

Few children were able to reduce their anti-epileptic drugs. However, a large proportion of children in each study achieved at least a >50% reduction in seizure frequency. Given that, by definition, these children are on maximal medical therapy, and in view of the comparatively lower side effect profile and reported improvement in quality of life, this makes this therapeutic option an acceptable choice for both family and clinician alike.

Clinical Bottom Line

Vagal nerve stimulation may be an effective alternative treatment for patients with intractable epilepsy who are not candidates for epilepsy surgery. (Grade B)


  1. Murphy JV. Left vagal nerve stimulation in children with medically refractory epilepsy. The Pediatric VNS Study Group. J Pediatr 1999;134:563–6.
  2. Patwardhan RV, Stong B, Bebin EM, et al. . Efficacy of vagal nerve stimulation in children with medically refractory epilepsy. Neurosurgery 2000;47:1353–8.
  3. Murphy JV, Tokelson R, Dowler I, et al. Vagal nerve stimulation in refractory epilepsy. Arch Pediatr Adolesc Med 2003;157:560–4.
  4. Bremer A, Eriksson AS, Roste GK, et al. Vagal nerve stimulation in children with drug-resistant epilepsy . Tidsskr Nor Laegeforen 2006;126:896–8.
  5. ILAE Commission on Neurosurgery of Epilepsy. A global survey on epilepsy surgery, 1980-1990: a report by the Commission on Neurosurgery of Epilepsy, the International League against Epilepsy. Epilepsia 1997;38:249–55.
  6. Ghanem T, Early SV. Vagal nerve stimulator implantation: an otolaryngologist’s perspective. Otolaryngol Head Neck Surg 2006;135:46–51.
  7. Benifla M, Rutka JT, Logan W, et al. Vagal nerve stimulation for refractory epilepsy in children: indications and experience at The Hospital for Sick Children. Childs Nerv Syst 2006;22:1018–26
  8. Labar D, Murphy J, Tecoma E, for the EO4 VNS Study Group. Vagus nerve stimulation for medication-resistant generalised epilepsy. Neurology 1999;52:1510–12..
  9. Parain D, Menniello MJ, Berquen P, et al. Vagal nerve stimulation in tuberous sclerosis complex patients. Pediatr Neurol 2001;25:213–16