Epilepsia, 33(6): 1021-1028, 1992 Raven Press, Ltd., New York 0 International League Against Epilepsy

Electrophysiologic Studies of Cervical Vagus Nerve Stimulation in Humans: 11. Evoked Potentials Edward J. Harnmond, Basim M. Uthman, *Steven A. Reid, and B. J. Wilder Neurology Service and *Neurosurgery Section, Department of Veterans Affairs Medical Center, Gainesville, Florida, U . S . A .

Summary: Evidence from studies of experimental animals indicates that electrical stimulation of the vagus nerve not only can alter the EEG but evokes activity in specific brain areas. We report effects of electrical stimulation of the vagus nerve in 9 patients with medically intractable seizures as part of a clinical trial of chronic vagal stimulation for control of epilepsy. The left vagus nerve in the neck was stimulated with a programmable implanted stimulator. Effects of stimulus amplitude, duration, and rate were studied. Noncephalic reference recording of the vagus nerve evoked potential showed some unusual properties: a scalp negative component occurred 0 with a latency of 12 ms, very high amplitude ( ~ 6 pV),

and widespread scalp distribution. Field distribution studies indicated that this potential was myogenic in origin and generated in the region of the stimulating electrodes in the neck area. Chemically induced muscle paralysis confirmed this observation. Bipolar scalp recording showed several small-amplitude topographically distinct potentials occurring in 30 ms. No effect, either acute or chronic, could be detected on pattern-reversal evoked potentials, auditory brainstem evoked potentials, auditory 40-Hz potentials, or cognitive evoked potentials. Key Words: Vagus nerve-Electrical stimulation-EpilepsyComplex partial seizures-Visual evoked potentialsAuditory evoked potentials.

Stimulation of the cervical vagus nerve in experimental animals can produce evoked activity in the cerebral cortex (Bailey and Bremer, 1938; Chase et al., 1967; O’Brien et al., 1971; Car et al., 1975), the hippocampus (Serkov and Bratus, 1970), the thalamus (Car et al., 1975), and the cerebellum (Hennemann and Rubia, 1978). Stimulation of vagal afferents can depress the activity of the spinal cord motor (Schweitzer and Wright, 1937; Theis and Foreman, 1981) and spinothalamic neurons (Ammons et al., 1983). Recent evidence indicates that vagus nerve stimulation has an antiepileptic effect (Uthman et al., 1990; Woodbury and Woodbury, 1990). To determine which brain regions are affected by vagus nerve stimulation in humans, we studied the scalp topography of the vagus nerve evoked potential (EP) in humans and the effects of vagus nerve stimulatiod-mvisual, auditory brainstem, auditory \ 40 Hz, and cognitive EPs. Among our findings, we showed that vagal stimulation evokes an unusually high-amplitude far-field potential at the scalp.

METHODS We studied 9 patients aged 21-58 years who had medically intractable complex partial seizures. All patients gave written informed consent for participation in the study. The Cyberonics (Houston, TX) model 100 prosthesis, an implantable, programmable pulse generator that delivers electrical signals to the vagus nerve, was implanted subcutaneously (s.c.) in the upper chest. A bipolar stimulation lead was tunneled S.C. from the pulse generator to the left vagus nerve (Reid, 1990). The stimulating electrode was described in detail in the previous article (Hammond et al., 1992). The system can be programmed with an IBM-compatible personal computer, interface unit, and software. The pulse width, output current, stimulation and no-stimulation periods, and signal frequency may be adjusted noninvasively . All EPs were recorded using standard techniques. EPs were recorded with a Cadwell (Kennewick, WA, U.S.A.) 8400 16-channel averaging system. Using standard techniques, we recorded visual, auditory, and cognitive EPs before, during, and after vagal stimulation. Certain stimulus parameters are given in the figure legends to Figs. 1-13 for particular EPs. All scalp electrode placement locations are according to the International 10-20 system.

~~

Received December 1991; revision accepted March 1992. Address correspondence and reprint requests to Dr. E. J . Hammond at Neurology Service (127), Department of Veterans Affairs Medical Center, Gainesville, FL 32608-1197, U.S.A.

1021

1022

E. J . HAMMOND ET AL. RESULTS

EPs evoked by vagus nerve stimulation Noncephalic reference recording The most prominent potential evoked by vagus nerve electrical stimulation was a scalp negative EP peaking at 8-12 ms (Fig. 1A). In 3 patients, this prominent EP was bilobed (Fig. 1B). Two patients showed a later and smaller negative EP at -22 ms (Fig. 2), and 1 patient occasionally showed a negative EP at 85 ms. The amplitude of the 8- to 12-ms component ranged from 10 to 60 pV across patients. The E P had a widespread scalp distribution: it was almost identical at each of the locations in the 10-20 system of electrode placement, suggesting a far-field EP, although with bipolar derivations some differences in scalp topography were evident (Fig. 1D-F and Discussion section). The EP had a much higher amplitude over the region of the stimulating electrode site in the neck (Fig. 3). Analysis times G400 ms were used, but longer latency EPs were not observed. No EP recorded on the scalp appeared to be different from a far-field EP from the neck (Fig. 4). Patients (n = 5 ) under general anesthesia showed the same EP. The EP was abolished in 1 patient by a muscle-paralyzing agent (Fig. 5), indicating that early EP is myogenic in origin and not a cortical or subcortical evoked response. Figure 6 shows the effect of different low- and highpass filter settings on this EP. Ear-reference recordings When scalp electrodes are referred to the left ear 10-20 system location A l , the 12-ms component is recorded with a positive polarity (Fig. 1B); when scalp electrodes are referred to the right ear, the 12-ms component is recorded with a negative polarity (Fig. 1C). Ear-hand recordings indicate that both ears are active for the 12-111s (Fig. 1F) component, but it is recorded with larger amplitude at the left ear, ipsilateral to the stimulation site (Fig. 1F). The amplitude of the EP recorded at A1 is larger than EPs recorded from the scalp; therefore when scalp electrodes are referenced to this negative EP, a positive EP results in the scalp-A1 derivation. Because the amplitude at A2 is less than scalp potential amplitudes, referencing scalp recordings to A2 results in a net negative potential difference over the right hemisphere (Fig. 1C). Scalp recordings: C, reference Lower amplitude EPs with a more complex scalp topography were evident using the Cz reference. On the left, a negative EP peaking at 10-12 ms (across subjects) was recorded at scalp location FP1. The peak latency of this EP increased by -2 Epilepsia, Vol. 33, NO. 6, 1992

ms as electrode locations progressed from anterior to posterior (Fig. 1D). At scalp location 01, there was a smaller negative EP (Fig. 1D). The amplitude of the EP recorded with bipolar derivations in temporal regions was typically 10-20 pV. Polarities were reversed on the right side of the head. In some subjects, a small component was evident before the 12-111s component. Scalp recordings: Anteroposterior bipolar chains For temporal chains, the same pattern was the same as that with the C, reference recordings, i.e., a negative 10- to 12-ms component that increased in latency by -2 ms as electrode derivations progressed from anterior to posterior (Fig. 1E). Potential polarity did not change over the other hemisphere as with the Cz reference. With parasagittal chains, a positive EP peaking at -10 ms was apparent and was maximum in posterior scalp regions.

Effects of stimulus intensity on vagus nerve EP The effects of stimulus intensity were studied on EPs recorded over the neck and from the scalp (Fig. 7). Generally, the EP is first recorded at the lowest intensity that the subject reports feeling (tingling sensation in throat). The amplitude of both neckand scalp-recorded EPs increases monotonically with increases in stimulus intensity ( ~ 2 - 3mA). The peak latency of the 12-ms component shortens only by -1 ms between the lowest and highest stimulus intensities. Effects of stimulus rate on vagus nerve EP Effects of stimulus rate are shown in Fig. 8. With stimulus intensities 3 2 mA, the amplitude of the 12-msec component was generally stable up to stimulus rates of lois; the amplitude then declined. Effects of stimulus pulse width on vagus nerve EP Effects of stimulus pulse widths of 130, 250, 500, 750, and 1,000 ps were studied (Fig. 9). Increasing stimulus duration to >250 p s had no effect on the EP. Effects on other EPs Two different types of effects must be considered: acute effects during actual stimulation of the vagus nerve and more chronic long-term effects. Vagus nerve stimulation appeared to have no acute or chronic effect on pattern reversal visual, auditory brainstem, 40-Hz, or cognitive (P300) potentials (Figs. 10-13). There was one exception: in 1 patient who had an abnormally prolonged P300 (latency 370 ms), the latency decreased by 30 ms after 2 months of chronic stimulation. DISCUSSION The vagus nerve innervates the superior and middle constrictors and muscles of the soft palate

VAGUS NERVE STIMULATION: EFFECTS ON EP

FP1 F7 T3

1023

-

T5 01 P3

c3

c3 FP2 F8 T4 T6 02 P4

c4 F4

A

r 0

60

38

+REFERENCE n-T3

FP1 F7 T3

WT5 T5-01 FPI-F3 F3c3 C*P3 PJ-01

w

c3 F3 FP2 F8 74 16 02 P4 c4 F4

B 30

8

PW2 F8-T4 T4-T6 T6-02

E

68

8

A1 REFERENCE

68

38

A-P BIPOLAR CHAINS -7-1

I

1

1

I

I

T

FP1 F7

1 c3 F3 FP2 FB T4 T6 02

c. 8

,

,

,

,

, 30

,

,

,

,

,

P4 c4 F4

-

NECK HAND

FZ - HAND

- HAND A2 - HAND A1

68

A2 REFERENCE B 3B FIG. 1. Scalp distribution of vagus nerve-evoked potentials (EPs). A: Noncephalic reference. Stimulus: 1 mA, 2 5 0 - k ~pulse delivered at 5/s. Amplifier bandpass 30-2,000 Hz. Reference electrode is on the right wrist. In all traces of the vagus nerve EP, the tracing starts with the synchronizing pulse to the averager, which is put out by the implanted pulse generator; this pulse lasts 16 ms and can be detected as a downgoing square wave in some of the tracings. Twenty-four milliseconds after onset of the synchronizing pulse, the stimulus to the nerve is delivered. This appears as a shock artifact with an onset latency of 24 ms. All discussed EP latencies are measured from onset of stimulus artifact. B: Left ear reference ( A l ) recording (amplifier bandpass 1-200 Hz); the major component has a positive polarity with this reference. C: Right ear reference (A2) recording; the major component now has a negative polarity with this reference. D: Scalp distribution of vagus nerve EPs using a bipolar scalp derivation (C, reference). Differences are apparent in scalp topography of some components. E: Scalp distribution of vagus nerve EPs using an anteroposterior bipolar derivation. F: Comparison of EPs recorded from neck, right and left ears, and scalp. Reference electrode is on right wrist. In Figs. 1-13, negativity is plotted upward, unless specifically stated otherwise. X-axis units are in milliseconds. All EPs in Fig. 1 were taken from 1 subject, but results are typical of all subjects.

through its pharyngeal branch and the inferior constrictor and cricothyroid muscles through the external laryngeal nerve. The recurrent laryngeal nerve sends a motor branch to the inferior constrictor and

all the laryngeal muscles (except cricothyroid). The muscle or muscles that generate the large myogenic potential we describe are not known, but studies are underway to determine them. Epilepsia, Vol. 33, No. 6 , 1992

1024

E . J . HAMMOND ET AL. '

synch pulse

'

' ,

1

I

,

,

I

I

NONCEPHALIC REFERENCE

A

stlmulus

30

0

.lg B

T 0

Previous studies in experimental animals As the diffuse projections of the nucleus of the solitary tract suggest, stimulation of vagal afferents can cause profound effects on CNS function. O'Brien et al. (1971) studied evoked cortical responses to vagal, laryngeal, and facial afferents in monkeys under chlorolose anesthesia. Animals were paralyzed with gallamine triethiode. Recording directly from the surface of exposed cortex, O'Brien et al. (1971) stimulated the cervical vagus nerve. They reported EPs that were most prominent in the prerolandic area. Only small contralatera1 EPs were observed in the postrolandic region. In the prerolandic region, the projection was bilateral and EPs were symmetric over both hemispheres. All EPs occurred with a latency 4 0 ms. Serkov and Bratus (1970) stimulated the vagus nerve in rabbits and noted an EP with a latency of 18-20 ms from the hippocampus. Hippocampal EPs in response to vagus nerve stimulation were similar to EPs evoked by stimulation of other afferent systems. Stimulation of the right or left vagus nerve produced bilateral hippocampal EPs. Car et al. (1975) stimulated the superior laryngeal nerve, which is a branch of the vagus. They recorded EPs in the nucleus solitarius with a latency Epilepsia, Vol. 33, N o . 6 , 1992

I100

flv

l*,,g;i,J .

.

.

ZB

FIG. 2. Vagus nerve evoked potentials (EPs) showing surface recorded components not recorded in all patients. A: Bilobed large negative component (observed in 3 patients). B: Possible early EP at 4 ms latency. This may be a recording of the peripheral whole nerve EP, but could also be myogenic in origin.

60

B

.

.

.

.

.

*

t

-

0

.:: 30

0

BIPOLRR

IMFl

-

0 0

,

:A 60

120 flv

2S0USEC

FIG. 3. Field distribution studies of vagus evoked potentials. Black dots on head indicate recording locations. A: Monopolar (hand reference) recording shows evoked potential largest over the neck. B: Bipolar recording shows phasereversal over neck incision.

of 2 ms. EPs were also recorded more rostrally in the pons (latency 2 ms) and the ipsilateral nucleus ventralis-posteromedialis of the thalamus (latency 4-5 ms). This secondary relay projected to the ipsilateral frontal cortex (latency 8-10 ms).

p-;7T7;j 0

1%

3Yla

c4,

+- 125 uv

FIG. 4. Vagal evoked potentials (EPs) recorded with long analysis time. Amplifier bandpass 1-100 Hz; stimulus rate O.l/s.Other than negative components over neck (top trace) no scalp recorded EPs were noted (subject C08).

VAGUS NERVE STIMULATION: EFFECTS ON EP

FPl F7 T3 T5 01

P3

:,"

c3

NON-PARALYZED

14 T6 02

P4 c4 F4 T

PV IB

28

FPl F7

13 T5 01

P3 c3 FPZ

FB

INTERMEDIATE

14

ie 02

P4 c4 F4

FP1 FT

---.

&

PARALYZED

FB T4 16 02

I025

ling et al. (1989) in the esophagus is obviously different than ours and might well involve stimulation of cutaneous afferent nerves, which are known to produce long-latency scalp recorded EPs. Similarly, the long-latency EPs observed by Upton et al. (1991) could also have been due to differences in stimulation site and/or current spread to cutaneous somatosensory afferents. Results from stimulation in other afferent systems might be relevant. Stimulation of the brachial plexus evokes cortical potentials over the contralatera1 sensory cortex. The latency of the initial negativity is 10 ms, followed by a positive component at 15 ms. These EPs are mediated through largemyelinated somatic fibers. The latency of these cortical EPs is comparable to the EPs we report but there are important differences. The vagus nerve potential has a much higher amplitude (30-60 vs. -1-3 pV for the near-field sensory cortex EP) and a widespread scalp distribution in comparison to the topographically distinct cortical EP in response to large-fiber stimulation. Bennett and Janetta (1980) studied trigeminal EPs in response to maxillary gum stimulation and reported scalp-recorded N20-P35-N40 waves, Lean-

P4 c4 F4

I

FIG. 5. Effect of muscle paralyzing agent (vecuronium) on vagus nerve evoked potentials (EPs). A: Sixteen-channel bipolar scalp recording under isoflurane anesthesia. B: After administration of 18 mg vecuronium, all components were markedly attenuated. C: After administration of 20 mg vecuronium, all components were abolished. EPs returned to normal a few minutes later.

AMPLIFIER BANDPASS ._

1-100 hz 1-500 hz

1-1ooO hz

1-2000 hZ

Previous studies in humans Frieling et al. (1989) reported cerebral responses to electrical stimulation of the esophagus. The stimulus was applied with a probe equipped with bipolar ring electrodes positioned in the middle and distal esophagus. The stimulus frequency was 0.1-1 Hz and intensity was 1.8-10 mA. Stimulating 20 cm from the incisors, Frieling et al. (1989) recorded multiphasic scalp EPs (at C,). An N93-P163-N221P281-N379 complex of waves was observed, with an amplitude of -10 pV for each recorded component. Frieling et al. (1989) believed that the EPs recorded in their study were mediated by slower conducting visceral afferents. Upton et al. (1991) stimulated the cervical vagus nerve (1 patient) and reported the appearance of two large negative components at 218 and 352 ms and a smaller component at 72 ms. We noted no long-latency EPs comparable to the EPs reported by Frieling et al. (1989) or Upton et al. (1991). The stimulation site used by Frie-

A a

3a

60

EFFECTS OF LOW PASS FILTERING

1-2000 hz 10-2000 hz 30-2OOU hz

SUBJECT C05

100-2000 hZ

B a

3a

6a

FIG. 6. A: Effects of high-pass filter settings on vagus nerve evoked potentials (EPs). B: Effects of low-pass filter settings on vagus nerve EPs. Epilepsia, Vol. 33, No. 6 , 1992

1026

E . J. HAMMOND ET AL.

’

I I I I In I I

. 1

I

I

NECK

n NECK n

NECK FZ

NECK

501s

R NECK R NECK FZ NECK R NECK

Fz NECK Fz

8

38

FIG. 8. Effects of stimulus rate on vagus nerve evoked potential (EP). Stimulus amplitude 2.0 mA, pulse width 250 ps. For traces corresponding to 10,20,40, and 50 Hz, the shock artifacts have been removed for readability; shock onset (triangle). These later traces are not averaged responses but represent single (representative) sweeps (subject C05).

of field distribution studies, disappearance with muscle paralysis, and high amplitude, the prominent 8- to 12-ms EP apparently is myogenic in origin. What is the neural pathway to this muscle? The latency of the vagus EP is similar to that of other scalp-recorded myogenic “microreflexes” (Bickford, 1972; Hammond and Wilder, 1985) in which

FIG. 7. Effects of stimulus intensity on vagus nerve evoked potential (EP). Recordings from both scalp (F,) and neck are shown. EPs from a subject who demonstrates the bilobed type of EP; the second EP began to appear at higher intensities and grew in amplitude. Very early EP appeared at high stimulus intensities (see also Fig. 2B).

130 us

250 us

dri et al. (1987) believed these EPs to be myogenic in origin; they stimulated the intraorbital nerve and recorded seven small components in the first 7 ms after the stimulus (at the infraorbital foramen). An N10 (amplitude of < I pV) was then recorded on the scalp contralateral to the stimulus. Leandri et al. (1987) believed that the origins of these waves were trigeminal nucleus, trigeminal lemniscus, thalamus, thalamic radiation, and cortical projection area. Later potentials were shown to have a myogenic and not a neurogenic origin. Again there are certain latency similarities with the EPs we report, but there are also distinct topographic and amplitude differences. The EPs with a latency of 8-12 ms recorded in this study are clearly different from EPs reported when the trigeminal or infraorbital nerves or the brachial plexus is stimulated-on the basis of their scalp topography and large amplitude. On the basis Epilepsiu. Vol. 33, No. 6 , 1992

500

US

750 us

lo00 us . SUBJECT CO8 0

.

.

.

.

1

1

1

1

.

.

,

,

30

.

.

. I

.

. ,

. ,

60

FIG. 9. Effects of stimulus pulse width on vagus nerve evoked potential. Stimulus amplitude 2.0 mA; stimulus rate 2 Hz (subject C07).

1027

VAGUS NERVE STIMULATION: EFFECTS ON EP 1

1

l

1

1

1

1

1

1

1

1

PRE VAGAL STlM

PRE-VAGUS STlM

CONCURRENT STlM

AFTER 1 YEAR CHRONIC STlM I

1

1

I

,

,

I

100

0

I

200

DURING VAGAL STlM

FIG. 10. Effects of vagus nerve stimulation (stim) pattern reversal visual evoked potential. Stimulus: check size 50 min; 0,-F, recording derivation (subject C02). Positivity is plotted upward.

the pathway is believed to ascend to the collicular level and then to descend to the particular scalp muscle generating the myogenic EP. Another more probable pathway is through the recurrent laryngeal nerve, a branch of the vagus that innervates the laryngeal muscles. This nerve has a high proportion of myelinated fibers (Hoffman and Kuntz, 1957). Stimulation of the vagus nerve in our patients had an obvious effect on our patients' voices, indicating activation of the recurrent laryngeal nerve.

Effects on other EPs Vagus nerve stimulation appeared to have no acute or chronic effect on pattern reversal, auditory brainstem, 40-Hz, or cognitive EPs. We did not study every combination of stimulus rate, pulse width, and stimulus amplitude on each type of EP in each of the 9 patients, but the effect of at least two different stimulus parameters (e.g., vagal stimula-

FIG. 12. Effects of acute and chronic vagus nerve stimulation on auditory brainstem evoked potential. Stimulus monaural click presented at 60 dBSL at ll/s; C, earlobe recording derivation. Positivity is plotted upward (subject C03).

tion at 1 and 30 Hz) was studied in each patient for each EP. If in initial testing some effect had been observed at a certain combination of stimulus parameters, we would have pursued this combination; but no effect on any EP was noted in any patient. 1.

I

1 .

1 .

1 p

.

1

,

.

1

.

1

.

1

.

!

I

I

.

.

.

.

.

I

. . . .

DURING VAGAL STlM

-3mm CHRONIC STlM

i

SUBJECTC03

LX---J

I

5u

1

.-

L1 VL8

FIG. 11. Acute and chronic effects of vagus nerve stimulation on 40-Hz potential. Auditory stimulus binaural 60 dBSL clicks at 40.7/s; C, earlobe recording derivation (subject C02).

0

375

750

FIG. 13. Effects of acute and chronic vagus nerve stimulation on cognitive (P300) evoked potential. Positivity is plotted upward. Target stimulus binaural, 2,000-Hz tone bursts presented at 60 dBHL (subject C03).

Epipilepsia, Vol. 33, N o . 6 , 1992

1028

E. J . HAMMOND ET AL.

Acknowledgment: This research was supported by the Epilepsy Research Foundation of Florida and the Medical Research Service of the Department of Veterans Affairs. We thank Janet Wootten and Anne Crawford for help in manuscript preparation and Charlotte Mixon for figure preparation.

REFERENCES Ammons WS, Blair RW, Foreman RD. Vagal afferent inhibition of primate thoracic spinothalamic neurons. J Neurophysiol 1983;50:92640. Bailey P, Bremer F. A sensory cortical representation of the vagus nerve. J Neurophysiol 1938;1:405-12. Bennett MH, Janetta PJ. Trigeminal evoked potentials in humans. Electroencephalogr Clin Neurophysiol 1980;48:51726. Bickford RG. Physiological and clinical studies of microreflexes. Electroencephalogr Clin Neurophysiol 1972;31:93-108. Car A, Jean A, Roman C. A pontine primary relay for ascending projections of the superior laryngeal nerve. Exp Brain Res 197.5;22: 197-2 10. Chase MH, Nakamura Y, Clements CD. Afferent vagal stimulation: neurographic correlates of induced EEG synchronization and desynchronization. Brain Res 1967;5:236-49. Frieling T , Enck P, Wienbeck M. Cerebral responses evoked by electrical stimulation of the esophagus in normal subjects. Gastroenterology 1989;97:475-8. Hammond EJ, Wilder BJ. Enhanced auditory postauricular evoked potentials in patients with corticobulbar lesions. Neurology 1985;35:278-81. Hammond EJ, Uthman BM, Reid SA, Wilder BJ. Electrophysiological studies of cervical vagus nerve stimulation in humans: I. EEG effects. Epilepsia 1992;33:1013-1020. Hennemann HE, Rubia FJ. Vagal representation in the cerebellum of the cat. P’ugers Arch 1978;375:119-23. Hoffman H H , Kuntz A. Vagus nerve components. Anat Rec 1957;1 2 7 : s 1-67. Leandri M, Parodi CI, Zattoni J, Favale E . Subcortical and cortical responses following infraorbital nerve stimulation in man. Electroencephalogr Clin Neurophysiol 1987;66:253-62. O’Brien JH, Pimpaneau A, Albe-Fessard D. Evoked cortical responses to vagal, laryngeal and facial afferents in monkeys under chloralose anaesthesia. Electroencephalogr Clin Neurophysiol 1971;31:7-20. Reid SA. Surgical technique for implantation of the neurocybernetic prosthesis. Epilepsia 1990;31(suppl 2):S38-9. Schweitzer A, Wright S. Effects on the knee jerk of stimulation of the central end of the vagus and of various changes in the circulation and respiration. J Physiol ( L o n 4 1937;88:459-75. Serkov FN, Bratus NV. Electrical responses of the hippocampus to stimulation of the vagus nerve. In: Rusinow VS, ed. Elecfrophysiology of the central nervous system. New York: Plenum Press, 1970:391-402. Theis R, Foreman RD. Inhibition and excitation of thoracic spinoreticular neurons by electrical stimulation of vagal afferent nerves. Brain Res 1981;207:178-83. Upton ARM, Tougas G , Talalla A, White A , Huboda P, Fitzpatrick D, Clark B, Hunt R. Neurophysiological effects of left vagal nerve stimulation in man. PACE 1991;14:70-6. Uthman BM, Wilder BJ, Hammond EJ, Reid SA. Efficacy and safety of vagus nerve stimulation in patients with complex partial seizures. Epilepsia 1990;3l(suppl 2):S44-50.

Epilepsia, Vol. 33, N o . 6, 1992

Woodbury DM, Woodbury JW. Effects of vagal stimulation on experimentally induced seizures in rats. Epilepsia 1990;31 (suppl 2):S7-19.

R6SUME Les etudes experimentales chez I’animal indiquent que la stimulation Clectrique du nerf vague non seulement modifie I’Clectroencephalogramme mais peut tgalement evoquer une activite au niveau de certaines regions cCrCbrales. Les auteurs rapportent les effets de la stimulation Clectrique du nerf vague chez 9 patients presentant des crises rebelles aux traitements, cette etude fait partie d’un essai clinique d’une stimulation vagale chronique pour le contrble de I’Cpilepsie. Une stimulation du nerf vague gauche au niveau du cou a ete effectuee au moyen d’un stimulateur implante programmable. Les effets de I’amplitude, de la durCe et de la frequence d e stimulation ont Cte CtudiCs. L’enregistrement en rkference non cephalique du potentiel tvoque par la stimulation du nerf vague a mis en evidence quelques proprietts inhabituelles: une composante negative survient au niveau du scalp avec une latence de 12 ms, une amplitude tres importante (iusqu’h 60 pV) et une distribution tres Ctendue sur le scalp. Les Ctudes de distribution de champs ont indique que ce potential Ctait d’origine myogtnique, et qu’il est genere au niveau de la region des electrodes de stimulation, places au niveau du cou. Une paralysie musculaire induite chimiquement confirme cette observation. Un enregistrement bi-polaire sur le scalp a mis en evidence plusieurs potentiels de faible amplitude distincts sur I t plan topographique, survenant en moins de 30 ms. Aucun effet, aigu ou chronique, n’a pu t t r e mis en Cvidence sur les potentiels 6voquCs par inversion de pattern, sur les potentiels CvoquCs auditifs du tronc cCrebral, sur les potentiels CvoquCs auditifs k 40 Hz ou sur les potentiels CvoquCs cognitifs. (P. Genton, Marseille)

ZUSAMMENFASSUNG Aus tierexperimentellen Ergebnissen geht hervor, dalj die elektrische Vagusstimulation nicht nur das Elektroenzephalogramm beeinflussen, sondern auch zu evozierter Aktivitat in spezifischen Hirngegenden fuhren kann. Wir berichten uber die Wirkung einer elektrischen Vagusstimulation bei 9 Patienten mit therapierefraktaren Anfallen als Teil einer klinischen Untersuchung zur Epilepsiebehandlung mit chronischer Vagusstimulation. Die Stimulation des linken N. vagus im Nacken wurde mit einem programmierbaren implantierten Stimulator vorgenommen. Wirkungen der Reizamplitude, Dauer und Reizfrequenz wurden untersucht. Die nichtcephale Referenzableitung des Vagus- evozierten Potentials zeigte einige ungewohnliche Eigenschaften: Eine negative Komponente trat mit einer Latenz von 12 ms, sehr hoher Amplitude (60 uV) und einer weiten Verteilung uber der Kopfoberflache auf. Die Untersuchungen zur Feldverteilung zeigten, daR das Potential myogenen Ursprungs war und in der Gegend der stimulierenden Elektroden im Nacken erzeugt wurden. Eine chemisch induzierte Muskelparalyse bestatigten diese Beobachtung. Eine bipolare Skalpableitung zeigte verschiedene niedrigamplitudige, topographisch unterschiedliche Potentiale innerhalb von 30 ms. Weder ein akuter, noch chronischer Effekt wurde bei den Musterumkehr-Potentialen, den auditorischen Hirnstammpotentialen, den auditorischen 40 Hz-Potentialen oder den kognitiven Potentialen gefunden. (C. K. Benninger, Heidelberg)