1. Vagus Nerve Stimulation for the Treatment of Epilepsy
Craig Watson, M.D., Ph.D. Professor of Neurology Wayne State University School of Medicine Founding Director, WSU/DMC Comprehensive Epilepsy Program

2. Ideal Epilepsy Therapy
Long-term antiseizure effect
Long-term antiepileptogenic effect
Safe
No AEs that impair QOL
Treatment effects that improve QOL
100% compliance
No interactions with other therapies

3. Goals of Epilepsy Therapy
Optimal Quality of Life (QOL)
No seizures
No adverse effects (AEs) from therapy
No depression

4. Success With Antiepileptic Drug (AED) Regimens
Previously Untreated Epilepsy Patients (N=470)
Kwan P, Brodie MJ. N EJM 2000;342:

5. Identifying Medically Refractory Epilepsy
Response to first AED is a powerful prognosticator of refractory epilepsy
Only 11% of patients whose first AED failed because of inadequate seizure control ever achieve seizure freedom
Compared with patients whose first AED failed because of intolerable side effects (41% seizure-free) or idiosyncratic reaction (55% seizure-free)
Objective: Kwan and Brodie studied the response to AEDs in patients with newly diagnosed epilepsy to identify factors associated with poor seizure control (ie, pharmacoresistant epilepsy).
Methods: Five hundred twenty-five patients, ages 9 to 93, were studied prospectively at a single clinical site between 1984 and Patients were diagnosed, treated, and followed up during the study. Epilepsy syndromes were classified as idiopathic (genetic basis for disease), symptomatic (structural abnormality), or cryptogenic (unknown cause). Seizure freedom was defined as the absence of seizure for 1 year.
Results: Three hundred thirty-three patients (63%) remained seizure-free during AED therapy. Some patients required AED polytherapy to reach seizure freedom. Thirty-six percent of patients had epilepsy that was refractory to AED treatment (pharmacoresistant epilepsy). The prevalence of persistent seizures was higher in patients with symptomatic or cryptogenic epilepsy. Seizure-free rates were similar whether a conventional AED or newer AED was used. Patients who had more than 20 seizures before initial therapy were less likely to achieve seizure freedom. Patients who had an inadequate response to the first AED tried were also less likely to ever achieve seizure freedom. For patients who did not respond to the first AED, only 11% ever became seizure-free when treatment failure was due to lack of efficacy, versus 41% when first failure was due to intolerable side effects or 55% when failure was caused by an idiosyncratic reaction.
Conclusions: Patients who have many seizures prior to AED therapy or who do not respond well to the initial drug used are likely to have refractory, or pharmacoresistant, epilepsy.
Kwan P, Brodie MJ. N Engl J Med. 2000;342:
Kwan P, Brodie MJ. N EJM 2000;342:

6. Medically Refractory Epilepsy
Persistent seizures despite “appropriate” trials of antiepileptic medications
Failure of 3 drug trials Or
Intolerable AEs
36% of all newly treated patients with epilepsy

7. Impact of Refractory Epilepsy
Cognitive and memory impairment1
Excessive AED burden2
Increased mortality (accidental, SUDEP)3
Higher depression rates4
Psychosocial dysfunction2
Reduced lifetime income5
Increased healthcare utilization6
1 Meador KJ. Neurology 2002;58(suppl 5):S21-S26.
2 Kwan P, Brodie MJ. NEJM 2000;342:
3 Annegers JF et al. Epilepsia 1998;39:
4 Harden CL. Neurology 2002;59(suppl 4):S48-S55.
5 Van Ness PC. Arch Neurology 2002;59:
6 Griffiths R, et al. Epilepsia 1999;40:

8. Potential AED Adverse Effects
Impaired cognition
Somnolence
Dizziness
Ataxia
Diplopia
Gastrointestinal effects
Weight gain
Alopecia
Gingival hyperplasia
Hepatic failure
Aplastic anemia, agranulocytosis, thrombocytopenia
Rash, Stevens-Johnson
Teratogenicity
Pancreatitis

9. Vagus Nerve Stimulation (VNS)

10. Origin of VNS Hypothesis

11. Historical Overview of VNS
1985 First animal studies (J. Zabara, Temple University)
1988 First human implant (K. Penry, B.J. Wilder, E. Ramsay)
1994 European community approval
completed controlled studies (N=454)
1997 USA (FDA) commercial approval
,000+ patients treated worldwide

12. VNS Indication for Use
“...indicated for use as an adjunctive therapy in reducing the frequency of seizures in adults and adolescents over 12 years of age with partial onset seizures, which are refractory to antiepileptic medications.”
FDA Physician’s Manual.

13. Mechanisms of Action (MOA) of VNS

14. Mechanisms of VNS Therapy
Exact MOA is unknown
Multiple actions of VNS therapy are supported by research in:
Human brain anatomy
Animal models of epilepsy
Human electroencephalogram (EEG), cerebrospinal fluid (CSF), functional brain imaging
No single mechanism explains the antiseizure effects and other beneficial effects of vagus nerve stimulation (VNS) therapy.

15. Vagus Nerve: Cranial Nerve X
Left cervical vagus nerve
80% afferent fibers, mostly myelinated
20% efferent fibers, mostly unmyelinated parasympathetic fibers to viscera, with myelinated fibers to vocal muscles
The vagus nerve is the 10th cranial nerve. Traditionally, the vagus nerve has been considered a parasympathetic efferent nerve (controlling and regulating autonomic functions, such as heart rate and gastric tone). However, more than 80% of the fibers of the vagus nerve are afferent sensory fibers, terminating in the nucleus of the solitary tract and carrying information to the brain from the neck, thorax, and abdomen.
The stimulation parameters of VNS therapy preferentially affect myelinated fibers. This explains effects on vocalization (hoarseness) during VNS therapy, and partly explains the almost complete absence of effects on the heart and other visceral organs during VNS therapy.
The right vagus nerve sends parasympathetic fibers to cardiac tissue controlling heart rate, but the left vagus nerve sends only a small number of fibers to the ventricular muscle itself.
Henry TR. Neurology. 2002;59(suppl 4):S3-S14.
Henry TR. Neurology 2002;59(suppl 4):S3-S14.

16. Vagus Nerve and Central Anatomy
Left cervical vagus nerve has 80% afferent fibers, which are myelinated and activated by VNS therapy
Vagus nerve’s parasympathetic efferents are unmyelinated and not activated much by VNS therapy
Left vagus nerve synapses bilaterally on nucleus of the solitary tract (NTS) in medulla
NTS projects to brain stem nuclei that supply serotonin and norepinephrine to the entire brain
NTS has widespread projections to limbic, reticular, and autonomic cerebral structures
The widespread, bilateral, multisystem, polysynaptic projections of the left vagus nerve account for the multiple therapeutic mechanisms of VNS therapy.

17. Animal Studies in Epilepsy Models

18. Locus Ceruleus Lesions Suppress the Seizure-Attenuating Effects of VNS
VNS demonstrated an anticonvulsant effect in rats against maximal electroshock
Chronic and acute chemical lesioning of the locus ceruleus (LC) was then performed
After lesioning LC, VNS was no longer effective
Conclusions
LC is involved in anticonvulsant effect of VNS
Effect of VNS may require norepinephrine release, a neuromodulator that has anticonvulsant effects
Objective: To chemically lesion the locus coeruleus (LC) to determine whether it is a critical structure involved in the anticonvulsive effects of VNS.
Methods: Rats were depleted of norepinephrine (NE) by a bilateral infusion of 6-hydroxydopamine (6-OHDA) in the LC. By lesioning the noradrenergic system with 6-OHDA and thus depleting the noradrenergic pathways, the VNS loses its antimaximal electric shock effects. Control rats had maximal electroshock but were not depleted of 6-OHDA. After 2 weeks, the maximal electroshock seizures were induced in rats and VNS was applied.
Results: When maximal electroshock seizures in rats were treated with VNS, VNS demonstrated an antiseizure effect in control rats but not in rats with chronic or acute LC lesions.
Conclusions: This study demonstrates some contribution of noradrenergic systems to the antiepileptic effect of VNS. The LC and NE are particularly important in the mechanism of action of VNS. Seizure suppression by VNS may therefore depend on the release of NE, which is a neuromodulator that has anticonvulsant effects. These data suggest that noradrenergic agonist might enhance VNS-induced seizure suppression.
Krahl S, et al. Epilepsia. 1998;39:
Krahl S, et al. Epilepsia 1998;39:

19. Conclusions from VNS Studies in Animal Models
Acute, abortive effects1
VNS terminates seizures when applied after seizure onset
Acute, prophylactic effects2
Seizure frequency and severity are reduced between trains of VNS
Chronic, progressive prophylactic effects3
Seizure frequency and severity are further reduced after chronic, long-term VNS
Data from McLachlan’s studies, in which focal penicillin, cortical injection, and then PTZ were used to trigger secondarily generalized seizures in rats, indicate that VNS has acute abortive effects when applied after seizures begin (McLachlan et al, 1993).
Takaya et al demonstrated that VNS exerts a persistent anticonvulsant effect when administered immediately before induced seizures. Seizures in rats that had received VNS were less severe compared with control rats. The greatest seizure-reducing effects occurred after 60 minutes of continuous VNS. Intermittent VNS was less effective than continuous VNS but more effective than when VNS was applied for only 1 minute prior to seizure induction. Effectiveness decreased with time after discontinuation of VNS, establishing the acute, prophylactic effects of VNS (Takaya et al, 1996).
Chronic prophylactic effects were observed by Lockard et al when VNS was administered over a period of 2 to 6 weeks in an alumina gel monkey model (Lockard, 1990).
These animal studies offered a foundation for future mechanism of action studies with VNS therapy in humans and contribute to a better understanding of how VNS therapy achieves anticonvulsant effects.
McLachlan RS, et al. Epilepsia. 1993;34:
Takaya M, et al. Epilepsia. 1996;37:
Lockard JS. Epilepsia. 1990;31(suppl 2):S20-S26.
1McLachlan RS, et al. Epilepsia 1993;34:
2Takaya M, et al. Epilepsia 1996;37:
3Lockard JS. Epilepsia 1990;31(suppl 2):S20-S26.

20. Effect of VNS Therapy on Human EEG
VNS therapy induces progressive EEG changes in the form of clustering of epileptiform activity followed by progressively increased periods of spike-free intervals
Objective: To determine the long-term effect of VNS on EEG in humans.
Methods: Patients (n=21) ages 4 to 31 receiving VNS were studied for a mean duration of 16.8 months with EEGs performed at baseline, 3, 6, and 12 months after VNS implant.
Results: Five patients with spike and wave activity at baseline showed synchronization of epileptiform activity and progressive increases in spike-free intervals (P<0.05). EEGs also revealed a progressive decrease in duration and frequency of spikes/spike and wave activity in these patients (P<0.01). The remaining 16 patients, who had less-active EEGs at baseline, did not have EEGs showing synchronization or spike clustering, but EEGs did show statistically significant progressive decreases in the number of spikes on EEG over time (P<0.004 at 3 months; P<0.008 at 6 months; P<0.004 at 1 year).
Conclusion: VNS results in progressive EEG changes (ie, clustering of epileptiform activity followed by longer spike-free intervals). These data may be an expression of the mechanism of action of VNS in controlling seizures.
Note: The graphic on the slide shows the inverse relationship between decreasing paroxysms and increased spike-free periods over time.
Koo B. J Clin Neurophysiol. 2001;18:
Koo B. J Clin Neurophysiol 2001;18:

21. VNS Therapy Modulates Blood Flow in Key Brain Structures
Significant bilateral changes in blood flow observed during VNS therapy1
Increased blood flow in the thalamus has been shown to have significant correlation with long-term seizure control (P<0.001)2
In 10 patients with epilepsy who received VNS therapy, positron emission tomography (PET) measurements were taken 3 times before and then during VNS therapy. The results demonstrated bilateral changes in brain blood flow from the resting state to during VNS therapy. Bilateral increased blood flow was seen in the thalamus, the hypothalamus, and the insular cortex. Bilateral decreased blood flow was seen in the amygdala, the hippocampus, and the posterior cingulate gyri (Henry et al, 1998).
These scans demonstrate statistical mapping of the regions where blood flow increases and decreases as a result of VNS therapy. Volumes of blood flow increases appear in yellow and decreases appear in blue. The areas that have t-values greater than or equal to 5 are superimposed on the gray-scale magnetic resonance images. The findings suggest that left cervical VNS therapy acutely increases synaptic activity in structures directly innervated by central vagal structures and areas related to left-sided somatosensory information. In addition, VNS therapy appears to acutely and bilaterally alter synaptic activity in several limbic system structures (Henry et al, 1998).
In a follow-up study, Henry et al reported that decreased seizures correlated with increased right and left thalamic blood flow (P<0.001) during VNS therapy. The time-course of PET imaging in this study analyzes the period of time during a train of VNS therapy (Henry et al, 1999).
Henry TR, et al. Epilepsia. 1998;39:
Henry TR, et al. Neurology. 1999;52:
1Henry TR, et al. Epilepsia 1998;39:
2Henry TR, et al. Neurology 1999;52:

22. Efficacy of VNS

23. VNS Clinical Studies

24. 1st and 2nd Controlled VNS Studies: Hypothesis & Design
“High” level stimulation would reduce overall seizure frequency to a greater degree than “Low” level stimulation.

25. High vs. Low (Active Control) Stimulation Parameters
Salinsky MC et al. Neurology 1995;45:

26. 1st Controlled VNS Study Results (E03)
p = 0.01
Salinsky MC et al. Neurology 1995;45:

27. VNS Long-Term Responder Rates (E01-E05)
Morris GL, Mueller WM. Neurology 1999;53:

28. VNS Long-Term Efficacy: % of Patients with >75% Seizure Reduction (E05)
(High Stim n=94)
(n=195)
DeGiorgio CM et al. Epilepsia 2000;41:

29. VNS Long-Term Efficacy: Seizure Frequency Reduction vs. Baseline
26% at 1 year, 30% at 5 years, 52% at 12 years using “last visit carried forward” analysis (n=25 of 47)1
28% at 1 year, 72% at 5-7 years (n=26 of 28)2
1 Uthman BM, et al. Neurology 2004;63:
2 Spanaki MV, et al. Seizure 2004;13:

30. AAN Position Statement
“…sufficient evidence exists to rank VNS for epilepsy as safe and effective, based on a preponderance of Class 1 evidence*.”
Fisher RS, Handforth A. Neurology 1999;53:
*Class 1 evidence: Provided by one or more well-designed randomized, controlled clinical trials.

31. Adverse Effects of VNS

32. VNS Adverse Effects Typically during stimulation “on time”
May lessen over time
May be reduced or eliminated with parameter adjustments
Similar across all age groups

33. Common Side Effects of VNS Therapy
Temporary hoarseness/changes in voice tone
Cough
Tickling in the throat
Shortness of breath
Most common adverse effect from implant surgery is infection

34. VNS Long-Term Adverse Effects (E01-E05)
*3-month results (High stimulation only, n=152). FDA Physician’s Manual. **Year 1,2 and 3 results (all study patients, N=440). Morris GL, Mueller WM. Neurology 1999;53:

35. VNS Retention Rates (E01-E05)
n=426 of 440
n=254 of 300
n=124 of 172
Morris GL, Mueller WM. Neurology 1999;53:

36. Long-Term Retention Rates for AEDs in Patients with Refractory Epilepsy
Lhatoo SD, Wong, et al. Epilepsia 2000;41:

37. Long-term Retention Rates with Levetiracetam
100 80 60 40 20
Patients (%)
This graph represents retention rates from a recent study of levetiracetam by Krakow et al. As with topiramate, lamotrigine, and gabapentin in the Lhatoo study, continuation rates with levetiracetam were demonstrated to decline substantially over time. The estimated continuation rates were 60% at 1 year, 37% at 3 years, and 32% at 5 years.
Krakow K, et al. Neurology. 2001;56: 1 3 5
Time (years)
Retention rates estimated by Kaplan-Meier method.
Krakow K, et al. Neurology 2001;56:

38. Safety of VNS

39. VNS Safety Profile No idiosyncratic reactions
No deaths attributed to VNS
No increase in Sudden, Unexpected Death in Epilepsy (SUDEP)

40. Quality of Life (QOL) with VNS

41. Quality of Life

42. Controlled VNS Studies: Effect on Overall Well-Being
Considerably
Worse
Considerably
Improved
No Change
E03
E05
* Between-group analysis P<0.05.
In Press. Dodrill C, Morris GL et al. Epilepsy & Behavior.

43. Technical Aspects of VNS Therapy

44. Vagus Nerve Stimulation
Intermittent electrical stimulation of the left cervical vagus nerve
Why the cervical vagus nerve?
Easy access
Few pain fibers
80% afferent fibers
Widespread anatomic projections in the brain
Left vagus has less cardiac innervation than the right

45. VNS Pulse Generator & Lead
Pacemaker-like pulse generator
Bipolar lead with two stimulating electrodes
Intermittent stimulation
30 sec on/5 min off
24 hours/day
On-demand therapy mode
6.9 mm thick
Weighs 25 grams
6 to 11-year battery life

46. VNS Implant Procedure Approximately 1 hour procedure
General anesthesia
Chest/axillary border incision for pulse generator
Neck incision for lead
Outpatient procedure

47. VNS Surgical Complications
Clinical Studies (N=454)1
Infection without explant
1.8%
Infection with explant
1.1%
Hoarseness/temporary vocal cord paralysis
0.7%
Hypesthesia/lower left facial paresis
All Patients (N=10,000+)
Asystole during routine lead test
0.1%
Mortality
0.0%
1Bruce et al. Epilepsia 1998;39(Suppl 6):92-93.

48. VNS Implant: Post-Op Scars

49. Programming Components
Computer, Software & Wand communicate transcutaneously to the pulse generator
Easy to use:
Training & equipment provided
Used for:
Surgical implant
Routine office visits

50. VNS Programmable Parameters
Pulse Generator cycle is 24 hours per day.

51. On-Demand Stimulation: Magnet Activation
Pass magnet over the pulse generator to start on-demand mode
Potential benefits:
Stop or shorten seizures/clusters
Decrease seizure severity
Improve post-ictal period
Sense of empowerment
Tape magnet over pulse generator to stop stimulation

52. Summary and Conclusions

53. VNS in Medically Refractory Epilepsy
As adjunctive therapy in refractory epilepsy patients with a confirmed diagnosis
Consider VNS after use of 3 appropriate AEDs along with the risk/benefit profile of all adjunctive therapies
As an adjunct for patients experiencing intolerable AEs

54. VNS in Relation to Epilepsy Surgery
Not before ideal resective surgery candidates
MTLE with hippocampal sclerosis, lesional epilepsy
Before more invasive palliative procedures such as callosotomy
Before invasive monitoring and resections in nonlesional, extratemporal partial epilepsies??

55. VNS Conclusions
Offers seizure control which is maintained and, for some, improves over time
Some patients report improvement in QOL
May be useful in treating depression associated with medically refractory epilepsy
No “Black Box Warning” regarding potential life threatening AEs

56. VNS Conclusions
AEs are well-tolerated and, for some, decrease with time
Unique advantages
100% compliance
“On-demand” therapy results in increased patient and family control
No drug interactions