Epilepsia, 33(6): 1005-1012, 1992

Raven Press, Ltd., New York

0 International League Against Epilepsy

Inhibition of Experimental Seizures in Canines by

Repetitive Vagal Stimulation
Jacob Zabara
Department of Physiology, Temple University School of Medicine, Philadelphia, Pennsylvania, U.S.A.

Summary: Repetitive electrical stimulation of the canine

cervical vagus nerve interrupts or abolishes motor seizures induced by strychnine and tremors induced by pentylenetetrazol (PTZ). Tremors were defined as rhythmic
alternating contractions of opposing muscle groups, exerting much less force than seizure contractions. Seizures
were induced by injection boluses of strychnine or PTZ at
1- to 4-min intervals until sustained muscle activity was
observed electromyographically (EMG). Vagal stimulation terminated seizures in 0.5-5 s. There were prolonged
periods with no spontaneous EMG activity after stimulation. The period of protection was approximately four

times the stimulation period. The antiseizure actions of

vagal stimulation were not altered by transection of the
vagus distal to the stimulating electrode. Optimal stimulus parameters were estimated: strength, -20 V (electrode resistance 1-5 0);frequency 20-30 Hz; duration,
-0.2 ms. These data suggest that the antiseizure effects
derive from stimulation of small-diameter afferent unmyelinated fibers in the vagus nerve. These results may form
the basis of a new therapeutic approach to epilepsy. Key
Words: Electrical stimulation-Epilepsy-Pentylenetetrazol-Seizures-Strychnine-Dogs-Vagus
nerve.

Inhibition of motor activity by activation of

visceral vagal afferents was first reported by
Schweitzer and Wright (1937) and was later confirmed by Paintal(1973). Identification of vagal projections to the cerebellum (Dell and Olson, 1951;
Sobusiak et a!., 1971; Hennemann and Rubia,
1978), cortex (Bailey and Bremer, 1938; Siegfried,
1961; Aubert and Egros, 1963; O'Brien et al., 1971),
brainstem (Grastyan et al., 1952; Kimehiko-Too
and Dussardier, 1963; Padel and Dell, 1965), thalamus (Juhasz et al., 1985), and hippocampus (Serkov
and Bratus, 1970) followed. Stimulation of vagal afferent fibers was also shown to cause profound
changes in the EEG (Zanchetti et al., 1952; Garnier
and Aubert, 1964; Chase and Nakamura, 1968;
Varbanova, 1972) and that low-voltage vagal stimulation significantly reduced EEG spiking of a cortical epileptic focus caused by topical application of
strychnine (Stoica and Tudor, 1967, 1968). Moreover, y-aminobutyric acid (GABA), considered a
primary inhibitory neurotransmitter, was observed
in brain areas of vagal innervation (D'Amelio et al.,
1987). These observations suggest that stimulation

of vagal afferents might prevent, reduce, or terminate spontaneous seizures.

Preliminary results from clinical trials (Penry and
Dean, 1990; Wilder, 1990) of vagal stimulation from
implanted electrodes, given in accordance with the
Neurocybernetic Prosthesis (NCP"; Cyberonics,
Webster, TX, U.S.A.), provide strong indications
that vagal stimulation is effective in reducing complex partial seizures in patients who are refractory
to drug therapy. The experimental basis for this
therapy was a series of studies of the effects of cervical vagal stimulation on motor seizures or tremors
induced in dogs. The details of this work, heretofore published only in abstract form (Zabara,
1985a,b, 1987),are reported. The aim of these studies was to determine the ranges of stimulus parameters (amplitude, repetition, frequency, and duration) that produce antiseizure activity and, if possible, to estimate the optimum stimulus parameters.
METHODS
Twenty dogs, 10 male and 10 female, of mixed
breed weighing between 10 and 22 kg were anesthetized with a-chloralose (Chloralose, Sigma Chemi-

Received March 1989; revision accepted June 1992.

Address correspondence and reprint requests to Dr. J. Zabara
at Department of Physiology, Temple University School of Medicine, 3223 N. Broad St., Philadelphia, PA 19140, U.S.A.

* Zabara J . Neurocybernetic Prosthesis. U.S. Patent

4,702,254-1987; U.S. Patent 4,867,164-1989.
1005

1006

J . ZABARA

cal, St. Louis, MO, U.S.A.) 100 mg/kg intravenously (i.v.). An indwelling catheter was then
placed in the right femoral vein for administration of
strychnine or pentylenetetrazol (PTZ; Knoll Pharmaceutical, Whippany, NJ, U.S.A.) to induce seizure or tremor, respectively, and for supplemental
anesthesia as dictated by the needs of the dog
throughout the experimental trials. For this study,
tremors were defined as rhythmic alternating contractions of opposing muscle groups, exerting much
less force than seizure contractions. The dogs were
cared for in compliance with the Principles of Laboratory Animal Care (National Society for Medical
Research, U.S.A.) and the Guide for the Care and
Use of Laboratory Animals (National Institutes of
Health, Bethesda, MD, U.S.A.).
Electrode application
The cervical vagus nerve was exposed high in the
neck between the branching points of the superior
pharyngeal and recurrent nerves. Care was taken to
eliminate excessive connective tissue and to avoid
drying of the nerve. Electrical stimulation was applied to the nerve through either a cuff (Stein et al.,
1977; Hoffer et al., 1981) or hook electrodes. The
nerve cuff, consisting of stainless-steel braided-wire
electrodes embedded in silicone rubber or Teflon,
maintains interelectrode separation and contact of
the electrodes with the cervical vagus nerve so that
the geometry of current is relatively constant for
repeated stimulation. The cylindrical cuff was positioned parallel to the vagus nerve with the opening
facing ventrally, and the nerve was slipped into the
cuff through the opening. Two wire hook electrodes
slipped under the nerve and pulled away from surrounding tissue were used for biphasic stimulation.
Monitoring
Pressure in the right femoral artery was recorded
with a P23AC transducer (Grass Instrument,
Quincy, MA, U.S.A.). The electrocardiogram
(ECG) was recorded through pin electrodes inserted under the skin. Seizure activity was recorded
by electromyography (EMG) with electrode pairs
inserted in both gastrocnemius muscles. Respiration was recorded through a pressure cuff wrapped
around the dogs abdomen or thorax and connected
to a PT5A transducer (Grass). All variables were
charted on a polygraph (Grass).
A functional monitor based on a respiratory response to cervical vagal stimulation, charted on the
same polygraph, was used as an initial control procedure to assure effective activation of inhibitory
neurons in the appropriate nerve bundle. Stimulation of the cervical vagus produces a respiratory
pattern resembling hyperventilation (Rice and Joy,
Epilepsia, Vol. 33, No. 6,1992

1947), a mild increase in the frequency of respiration. During the initial phase of the experiment after
anesthesia was induced and before strychnine or
PTZ was injected, stimulation parameters were adjusted in each dog to produce this respiratory response.
Stimulation parameters
The electrical stimulus was delivered from the
pulse generator through lightweight, flexible,
braided stainless-steel wire. To test the integrity of
the electrodes, the impedance of each contact was
measured at 30 Hz with a Grass impedance meter
(Grass). Impedance values were steady and in the
range of 1-5 R. Square pulses were applied at 15-100
V, 20-150 Hz; brief stimulus pulses (0.2-2 ms) were
used to minimize electrode polarization. From Ohms
law, currents were in the range of 3-100 mA, most
likely in the range of 5-35 mA. Ranges for the threshold of effect in stopping seizures were then determined by steadily increasing the voltage until an increase in respiratory response to cervical vagal stimulation (described above) was obtained and verified.
Stimulation trials
In all trials, stimulation was started 1-3 min after
initial signs of seizure were observed on the EMG
recording. After each stimulation trial, seizure activity was allowed to return spontaneously or, if it
did not recur within 30 min, was again induced.
Strychnine seizures were produced by injecting
0.2-0.5 ml 1% solution of strychnine in distilled water through a catheter in the femoral vein at 1- to
4-min intervals until initial seizure activity was observed on the EMG recording; e.g., the convulsion
shown in Fig. 1 resulted from 0.5-ml injections of
strychnine at 0, 3.5, 7.4, 9.8, 11.8, 14.8, and 16.7
min. Individual convulsive twitches began after the
last injection and climaxed in a high-frequency continuously occurring seizure after 14.2 min. Repetitive vagal stimulation was then started. Latency
from initiation of stimulation to cessation of the seizure episode was recorded for each trial. In one dog
that received strychnine, a ramp-down was used to
terminate stimulation: one of the stimulus parameters (amplitude, frequency, or duration) was gradually reduced (four trials) to determine whether an
off effect, a myoclonic jerk occurring at termination of stimulation, could be diminished. In another dog, a ramp-up of one of the stimulus parameters was performed (four trials) to determine if latency in seizure suppression could be reduced. In
two dogs that received strychnine, three trials were
performed; the vagus nerve was then transected
distal to the stimulating electrodes to determine if
this would abolish seizure suppression.

VAGAL INHIBITION OF SEIZURES

1007

1A
RESPIRATION
TIME
EMG 1

EMG 2
EKG

1 MIN

STIMULATION
ON

STIMULATION

OFF

1B
RESPIRATION

TIME
EMGl--

EMG 2

FIG. 1. Termination of strychnine-induced seizure activity by

repetitive stimulation of the vagus nerve. A: Stimulation (60
Hz, 0.6 ms, 40 V) continued for a
total duration of 4 min. Time intervals are in seconds. EMG 1,
electromyogram from the left
gastrocnemius; EMG 2, from
the right gastrocnemius muscle. Heart beat (electrocardiogram) and respiration ceased
temporarily at onset of stimulation. 8 : Continuation of record
shown in A. A 10.25-min interval
occurred between the end of
the recording shown in A and
the start of the recording shown
in B. Seizure activity was absent
for a total duration of 20 min;
i.e., during 4-min stimulation
and for 16 min after stimulation.
Seizure activity resumed near
the end of the record.

1 MIN

PTZ was administered at relatively small doses to

induce tremors, but not seizures, in two dogs. A
0.1% solution of PTZ in distilled water was prepared: 0.2-0.5 ml solution was injected through a
catheter in the femoral vein at 1 - to 4-min increments until tremor was observed on the EMG recording. Brief, repetitive stimulus trains were applied for -30 s each (three trials in each dog). In
three dogs that had spontaneous tremors apparently
due to chloralose induction, the effects of vagal
stimulation were tested (one trial in each dog) before strychnine was administered.
Because of potential injury to the dogs from the
movement generated by seizure activity, each dog
was loosely restrained in a supine position on a dogboard resting on top of a long rectangular table. The
momentum of the seizure was dissipated by movement of the dog-board along the top of the table.
RESULTS
Strychnine produced sustained tonic-clonic seizures lasting a 2 0 min. PTZ in the relatively low
doses administered produced tremors in the two
dogs so tested. In three dogs, spontaneous tremors
occurred, apparently due to chloralose induction.
The hyperventilatory response to vagal stimulation
(described above) was quite consistent and was used
as a reliable indication that the stimulating current
was activating the fiber groups in the cervical vagus
nerves that cause inhibition of motor seizures.

Effect of vagal stimulation on strychnine-induced

seizure activity
The intent of this study was to define the ranges
of stimulus parameters that produce antiseizure activity and to estimate optimum values of the parameters. An ongoing seizure was clearly identified in
the polygraph recordings based on the massive excursions of all monitors during seizure activity. Repetitive stimulation was started 1-3 min after the
initial signs of a seizure were detected. Seizures
could be either interrupted or terminated by repetitive stimulation of the cervical vagus (Figs. 1-3).
Transection of the vagus nerve distal to the electrodes had no observable effect on seizure suppression.
The recordings in Fig. 1 show that the time between the start of vagal stimulation and cessation of
seizure activity was short: <1 s. Although strychnine usually produces several periods of seizure activity at intervals of several minutes, vagal stimulation often prevents the return for a considerable
time after the end of stimulation (Fig. 1A and B).
The longest observed latency from start of stimulation to termination of seizure was 14 s. In a few
cases, a myoclonic jerk occurred at termination of
stimulation.
Although clonic movements did not always terminate immediately after initiation of vagal stimulation (Figs. 2 and 3), their rate decreased quickly
to zero (4 s in Fig. 2, and 14 s in Fig. 3). Characteristically, a continuous seizure lasts 6 2 0 min.
Epilepsia, Vol. 33, No. 6 , 1992

1008

J. ZABARA

FIG. 2. A strychnine-induced seizure

was observed as massive excursions
in all channels. At stimulation onset
(150 Hz, 0.8 ms, 90 V), seizure frequency decreased, and the seizure terminated in 4 s. A slow but steady decrease in respiration and heart rate
(electrocardiogram) was apparent after initiation of stirnulation. Beyond
the end of this record, stimulation
continued for a total duration of 6.33
min. The seizure continued to be suppressed during this period and for 17
m i n more after s t i m u l a t i o n was
stopped (not shown). The total period
of seizure absence was 23.33 min; after this period, seizure immediately returned in full strength and frequency.

Apparently, therefore, vagal stimulation did not

merely coincide with the spontaneous end of the
seizure.
During seizures, the limb excursions appeared to
remain at a constant level. Thus, either the limb
excursions in the seizure appeared to occur in full
intensity or they did not occur at all, as in a threshold phenomenon.
After suppression, the seizure did not recur immediately after stimulus termination, but rather after a variable period of time that was directly related to stimulation duration (Fig. 1). In several
cases, a delay of 4 min to seizure recurrence occurred for each 1 min of stimulation; e.g., after a
4-min stimulation period there was a 16-min hiatus
before a seizure occurred (Fig. 1). Typically, application of stimulus trains produced long-term control
of seizures and, after several sequential stimulation
periods, seizures did not recur. In such cases, after
a 30-min interval, another injection of strychnine
was given to continue the experimental trials.
Heart and respiratory function (Figs. 1-4) were
not impaired during seizures or vagal stimulation.

I
STIMULATION

ON

Epilepsia, Vol. 33, No. 6, 1992

.5MIN

STIMULATION

.5 MIN

Rarely, there was a brief hiatus in the heart beat at

onset of stimulation (Fig. lA), but the heart rate
quickly stabilized.
Threshold for antiseizure effect of vagal stimulation
A wide range of stimulus voltages, durations, and
frequencies were used in an attempt to define the
optimal stimulus parameters for seizure suppression (Table 1). There were two main categories:
stimulation that abolished seizures, and stimulations that reduced the frequency of clonic jerks
250%. The data show that the antiseizure effects of
vagal stimulation were maximum at stimulation frequencies 230 Hz, and stimulus voltages a20 V are
independent of duration of stimuli over the range
used (0.2-2 ms). In addition, some evidence shows
that stimulation frequencies 280 Hz are less effective than 60 Hz.
Off effect
At cessation of stimulation, an off effect consisting of one or several myoclonic jerks occurred in
some trials (Fig. 1A). This effect could be prevented

FIG. 3. A strychnine-induced seizure

was observed as high-frequency excursions in all channels. Seizure frequency was greatly reduced immediately as stimulation (100 Hz, 0.8 ms, 85
V) started and abruptly terminates
within 14 s of stimulus onset. This delay in seizure termination can be reduced to less than a second by a
ramp-up in stimulus parameters. Stimulation continued beyond the end of
this recording for a total duration of
4.67 min. After s t i m u l a t i o n was
stopped, the seizure did not return for
a 21-min period (not shown). The time
scale in this recording is the same as
that in Fig. 2.

VAGAL INHIBITION OF SEIZURES

FIG. 4. A pentylenetetrazolinduced, preseizure tremor is
evident in the electromyogram
(EMG) tracings. Stimulation (70
Hz, 0.6 ms, 70 V) began at the
first black rectangle on the time
trace and ended at the second
black rectangle The tremors
stopped at onset of stimulation
and were absent for -85 s, well
past cessation of stimulation.
Thereafter, tremors returned in
both limb muscles almost simultaneously and gradually increased in amplitude.

RESPIRATION^^^^^^^^^^^^^
TIME-

r P

EMG
- 1

T---"-

___

--

I
_
.
-

EMG 2-

1009

STIMULATION

OFF

by a ramp-down in stimulation parameters, e.g., by

reducing either the stimulus voltage or pulse width
to zero in 10 s rather than terminating the stimulus
abruptly. The brief, myoclonic jerk at cessation of
stimulation shown in Fig. 1A is the same size as a
single oscillation of the dog's limbs during a seizure
episode. The seizure consisted of a high-frequency
repetition of virtually identical jerks. These discrete
movements are not distinguishable in the recording.
The myoclonic jerk at the end of stimulation appeared to be a breakthrough of the seizure, an off
effect, indicating that the full-blown seizure would
have recurred at this time if it were not for a slowly
developing and decaying inhibitory effect of the
stimulation (Fig. 1B).
Effect of vagal stimulation on PTZ-induced
subthreshold seizures (tremors)
Tremor induced with PTZ also was controlled by
vagal stimulation in the two dogs so tested (Fig. 4).
The tremors were observed on EMG recordings
from the left and right gastrocnemius muscles. With
initiation of stimulation, the tremors stopped and
did not reappear for -55 s after stimulation was
terminated. When the tremors returned, they usually did so with slowly increasing magnitude. Seizures, in contrast, recurred in their full magnitude.
The level of voltage necessary to produce the inhibitory effect on tremors was usually less than that
required to control seizures. Figure 4 shows that the
prestimulation EMG 1 was greater than poststimulation; the reverse was true for EMG 2.
Three other dogs had spontaneous tremors, presumably as a result of the chloralose anesthesia
(Fig. 5). Repetitive, brief stimulations were performed (Fig. 5 ) , but there was no discernible effect
until the tremors stopped abruptly soon after the
end of the last burst of stimuli. The tremors may
have been terminated by a persistent inhibitory effect that was incremented by each burst of stimulation until it was strong enough to suppress the tremors. The data are not sufficient, however, to elimi-

-~

ON

STIMULATION

_.

1 MIN

nate the possibility that the tremors disappeared

spontaneously at this time by coincidence.
DISCUSSION

Our results provide direct evidence that repetitive electrical stimulation of the vagus nerve in the
neck can interrupt or terminate strychnine-induced
seizures and PTZ-induced tremors in dogs. This inhibitory effect is not affected by transection of the
vagus nerve distal to the stimulating site. Bilateral
vagal stimulation produces no measurably greater
effect than does unilateral stimulation, and right or
left vagal stimulation is equally effective in controlling motor seizures (J. Zabara, unpublished observations, 1990). There appears to be a common site
or mechanism which either vagal nerve bundle can
activate equally to prevent seizures and which can
be maximally activated by input from either vagal
nerve. In previous studies (Chase and Nakamura,
1967; Chase et al., 1968), vagal stimulation-induced
EEG changes appeared to be equivalent over both
cortexes. These observations indicate that cervical
vagal impulses develop bilateral activity in the brain.
Another consistent finding was that the inhibitory
effects of repetitive vagal stimulation persist for a
considerable time after termination of stimulation.
A rough rule of thumb is that seizures are suppressed for a period four times as long as the duration of the stimulation; e.g., in the seizure shown in
Fig. 1, stimulation duration was 4 min and seizure
activity did not reappear for 16 min more. It is also
true that seizures are terminated within several seconds after start of stimulation. These two findings
indicate the presence of at least two components of
the inhibitory process: a rapidly rising and decaying
component (time scale of seconds) and a slowly rising and decaying component (time scale of many
minutes).
The results of this study of canine seizures have
been corroborated in experiments in other animals.
Chronic focal seizures induced in monkeys with
Epilepsia, Vol. 33, No. 6 , 1992

J . ZABARA

1010

TABLE 1. Effect of vagal stimulation on seizures in

relation to stimulation frequency, voltage, and pulse
width: 100% (category A ) or 50% (category B)
seizure decrease"
Seizure
decrease"

(%I
Category A
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
100
Category B
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50
50

Stimulation
frequency
(HzI

Pulse
width
(ms)

150
100
100
100
100
80
80
80
80
80
70
60
60
60
60
60
60
60
60
60
60
40
40
30

90
85
60
40
10
70
70
30
20
15
80
60
60
40
50
50
50
50
95
95
100
100
30
80

0.8
0.8
1.0
0.3
0.2
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.6
0.8
0.6
0.6
0.6
0.6
0.2
0.2
0.3
0.3
0.6

100
100
I00
100
100
80
80
80
80
80
80
80
60
60
60
60
60
60
60
60
60
20

80
80
100
20
80
60
50
60
15
15
15
15
10
10
10
10
10
15
10
10
10
20

1.0
1.0
0.5
0.2
0.2
0.5
0.5
0.6
0.2
0.6
0.6
0.6
2.0
2.0
2.0
2.0
2.0
0.6
0.6
0.6
0.2
0.2

__

Experimental trial results: The aim of these studies was to

find the ranges of stimulus parameters that produce antiseizure
activity. In the 100% reduction category (A), the seizure was
completely stopped during the stimulation period and usually for
a period of time after the end of stimulation. In the 50% reduction
category (B), the average frequency of clonic jerks was reduced
>50% during the stimulation period. The parameters in these
studies were 4 to 200 Hz, 1-100 V, and 0.2-2 ms, thus demonstrating a relatively wide range of parameter effectiveness. The
trials were presented as a series of decreasing stimulation frequencies, with associated voltages and pulse widths.
a

Epilepsia, Vol. 33, No. 6 , 1992

alumina gel were then tested for the effects of repetitive vagal stimulation on seizure frequency. In
two monkeys, seizures were abolished; the interseizure intervals became invariable in the remaining
two monkeys (Lockard and Congdon, 1986; Lockard et al., 1990). Anticonvulsant effects of cervical
vagal stimulation were also observed in seizures induced in rats with PTZ, 3-mercaptoproprionic acid,
and maximal electroshock (Woodbury and Woodbury, 1990, 1991).
Fiber types involved and estimation of optimum
stimulus parameters
Apparently two different fiber groups exist, with
opposite effects on the EEG (Garnier and Aubert,
1964; Chase and Nakamura, 1968). Stoica and Tudor (1967, 1968) observed significantly reduced
EEG spiking of a cortical epileptic focus caused by
strychnine with low-voltage stimulation of the cervical vagus, and obtained increased spiking with
high-voltage stimulation. They concluded that this
dual effect may result from activation of different
categories of fibers.
Chase et al. (1967) investigated the refractory periods of the nerve groups in the cervical vagus of
cats. Increases in stimulus voltage led to excitation
of fiber groups with successively higher thresholds
and slower conduction velocities. When the voltage
and duration of the stimulus were adjusted to excite
fibers submaximally in a specific fiber group, an
increase in the frequency of stimulation >20/s resulted in a reduction in spike amplitude. The data in
Table 1 indicate the types of nerve fibers that produce the antiseizure effect and a preliminary estimate of the optimum stimulus parameters.
Stimulus amplitude
Antiseizure effects appear to plateau at or below
20 V. Electrode impedances ranged from 1 to 5 R at
30 Hz. Taking the middle value, the current flow at
20 V was -7 mA (range 4-20 mA). This is a high
value, indicating that the inhibitory fibers have
small diameters and thus high thresholds. This is
consistent with the findings of Woodbury and
Woodbury (1990), who concluded that stimulation
of small unmyelinated (C) fibers produced the antiseizure effect. For comparison, the values used in
studies in humans are usually 0.5-2.0 mA (Hammond et al., 1990; Penry and Dean, 1990).
Stimulus frequency
Maximum antiseizure effects were obtained at
>20-Hz frequencies, with some indication that
>60-Hz frequencies reduce the effect. This finding
again suggests that the inhibitory influence is car-

VAGAL INHIBITION OF SEIZURES

EMG- 1

--__rc

EMG 2

1011

FIG. 5. Effects of vagal stimulation

on spontaneous leg tremors. These
are prominent in electromyogram
(EMG) 2 tracings. Stimulation periods
are indicated by black rectangles on
the time trace. At the end of several
brief periods of stimulation (60 Hz, 1 .O
ms, 15 V), the tremors terminated and
did not return.

.5 MIN

ried on small-diameter fibers that cannot sustain impulse frequencies 260 Hz. Probably only unmyelinated fibers fit this category.
Stimulus duration
Stimulus durations between 0.2 and 2 ms had no
detectable effect on the strength of the antiseizure
effect. This result is useful because it is important to
minimize average current flow through the electrodes. At any frequency, the average current increases directly with duration; hence, the shortest
pulse that does not decrease the antiseizure effect
should be used. Thus, the optimum stimulus duration is -0.2 ms.
The data indicate that small unmyelinated nerve
fibers must be stimulated to produce an antiseizure
effect. The optimum stimulus parameters estimated
from the data shown in Table 1 are: stimulus frequency 20-30 Hz, stimulus strength 10-20 V, and
stimulus duration -0.2 ms.
Long-term suppression of seizures has been observed in patients implanted with the NCP for cervical vagal stimulation (Hammond et al., 1990;
Penry and Dean, 1990; Wilder et al., 1991). Their
results are consistent with the results of the present
study.
Acknowledgment:I thank Drs. J. Walter Woodbury and
Dixon Woodbury for reading the manuscript and making
valuable suggestions for revision.

REFERENCES
Aubert ML, Egros J. Projections du nerfvague sur le neocortex
du chat. J Physiol 1963;2:109.
Bailey P, Bremer F. A sensory cortical representation of the
vagus nerve. J Neurophysiol 1938;1:405-12.
Chase MH, Nakamura Y. Cortical and subcortical EEG patterns
of response to afferent abdominal vagal stimulation: neurographic correlates. Physiol Behav 1968;3:605-10.
Chase MH, Nakamura Y , Clemente CD. Afferent vagal stimulation: neurographic correlates of induced EEG synchronization and desynchronization. Brain Res 1967;5:23649.
DAmelio FE, Mehler WR, Gibbs MA, et al. Immunocytochemical localization of glutamic acid decarboxylase (GAD) and
glutaminesynthetase (GS) in the area postrema of the cat.
Light and electron microscopy. Brain Res 1987;410:232-44.
Dell P, Olson R. Projections secondaires mesencephaliques, di-

encephaliques et amygdaliennes des afferences viscerales vagales. C R Soc Biol (Paris) 1951;145:1088-91.
Garnier L , Aubert M. Modifications delelectroencephalogramme du chat consecutives a la stimulation du nerf vague.
C R Soc Biol (Paris) 1964;158:2405-8.
Grastyan E, Hasznos T, Lissak L. Activation of the brain stem
activating system by vegetative afferents. A c t a Physiol
Scand Sci Hung 1952;3:102-22.
Hammond EJ,Ramsay ER, Uthman BM, Reid SA, Wilder BJ.
Vagus nerve stimulation in humans: neurophysiological studies and electrophysiological monitoring. Epilepsia 1990;31
(SUPPI2):S51-9.
Hennemann HE, Rubia FJ. Vagal representation in the cerebellum of the cat. Pflugers Arch 1978;375:119-23.
Hoffer JA, Loeb GE, Pratt CA. Single unit conduction velocities
from averaged nerve cuff electrode records in freely moving
cats. J Neurosci Methods 1981;4:211-25.
Juhasz G , Detari L, Kukorelli T. Effects of hypnogenic vagal
stimulation on thalamic neuronal activity in cats. Bruin Res
Bull 1985;15:437-41.
Kimehiko-Too T, Dussardier M. Convergence sur les cellules de
la formation reticulaire bulbaire dafferences vagales et dafI
Physiol 1963;2: 179.
ferences des membres. .
Lockard JS, Congdon WC. Effects of vagal stimulation on seizure rate in monkey model. Epilepsia 1986;27:626.
Lockard JS, Congdon WC, DuCharme LL. Feasibility and
safety of vagal stimulation in monkey model. Epilepsia
1990;31(suppl 2):S20-7.
OBrien JH, Pimpaneau A, Albe-Fessard D. Evoked cortical responses to vagal, laryngeal and facial afferents in monkeys
under chloralose anaesthesia. Electroencephalogr Clin Neurophysiol 1971 ;3 1:7-20.
Padel Y, Dell P. Effets bulbaires et reticulaires des stimulationsendormantes du tronc vago-aortique. J Physiol (Paris) 1965;
57:269-70.
Paintal AS. Vagal sensory receptors and their reflex effects.
Physiol Rev 1973;53:159-227.
Penry JK, Dean JC. Prevention of intractable partial seizures by
intermittent vagal stimulation in humans: preliminary results.
Epilepsiu 1990;3l(suppl 2): S40-4.
Rice HV, Joy MS. Modifications of respiratory movements by
vagal stimulation. A m J Physiol 1947;149:24-42.
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 1937;88:459-75.
Serkov FN, Bratus NV. Electrical responses of the hippocampus
to stimulation of the vagus nerve. In: Rusinov VS, ed. Electrophysiology of the central nervous system. New York: Plenum, 1970;391402.
Siegfried J. Topographie des projections corticales du nerf vagus
chez le chat. Helv Physiol Pharmacol Acta 1961;19:269-78.
Sobusiak T,Zimny R, Matlosz Z. Primary glossopharyngeal and
vagal afferent projection into the cerebellum in the dog. J
Hirnforsch 1971;13:117-34.
Stein RB, Nichols TR, Jhamandas J, et al. Stable long-term recordings from cat peripheral nerves. Brain Res 1977;128:2138.
Epilepsia, Vol. 33, No. 6 , 1992

J . ZABARA

1012

Stoica I, Tudor I. Effects of vagus afferents on strychninic focus

of coronal gyrus. Rev Roum Neurol 1967;4:287-95.
Stoica I, Tudor I. Vagal trunk stimulation influences on epileptic
spiking focus activity. Rev Roum Neurol 1968;5:203-10.
Varbanova A. The role of the rhythmic vagal impulses in the
appearance of paroxysmal EEG activity (in English). Izv lnst
Fiziol (Sofiya) 1972;14:63-83.
Wilder BJ, ed. Vagus nerve stimulation for the control of epilepsy. Epilepsia 1990;31(suppl 2):S140.
Wilder BJ, Uthman BM, Hammond EJ. Vagal stimulation for
control of complex partial seizures in medically refractory
epileptic patients. PACE 1991;14: 108-15.
Woodbury DM, Woodbury JW. Effects of vagal stimulation and
experimentally induced seizures in rats. Epilepsia 1990;31
(SUPPI2):S7-20.
Woodbury JW, Woodbury DM. Vagal stimulation reduces the
severity of maximal electroshock seizures in intact rats: use
of a cuff electrode for stimulating and recording. PACE 1991;
14:94-107.
Zabara J. Control of hypersynchronous discharge in epilepsy.
Electroencephalogr Clin Neurophysiol 1985a;61:162.
Zabara J. Time course of seizure control to brief, repetitive stirnuli. Epilepsia 1985b;26:5 18.
Zabara J. Controlling seizures by changing GABA receptor sensitivity. Epilepsia 1987;28:604.
Zanchetti A, Wang SC, Moruzzi G. The effect of vagal stimulation on the EEG pattern of the cat. Electroencephalogr Clin
Neurophysiol 1952;4:357-61.

&SUME
La stimulation Clectrique rCpCtCe du nerf vague cervical chez
le chien interrompt ou abolit les crises motrices induites par la
strychnine et les tremblements induits par pentylenetCtrazol
(PTZ). Ces tremblements sont dCfinis comme des contractions
alternantes rythmiques de groupes musculaires opposCs, exerEant une force bien infkrieure B celle des contractions critiques.
Les crises ont CtC induites par des bolus dinjections de strychnine ou de PTZ a intervalles de 1 a 4 minutes, jusquh obtention
sur IEMG dune activitC musculaire soutenue. La stimulation

Epilepsia, Vol. 33, No. 6 , 1992

vagale a interrompu les crises en 0.5 a 5.0 secondes. De longues

pCriodes sans activite EMG spontanee ont CtC observees apres
stimulation. La ptriode de protection ttait denviron 4 fois la
periode de stimulation. Lactivite anti-crise de la stimulation vagale na pas CtC modifiCe par la transection du vague distale a
ltlectrode de stimulation. Les parametres optimaux de stimulation ont CtC estimes comme suit : intensit6 environ 20 volts (resistance de IClectrode de 1 a 5 Kohm); frtquence 20 2 30 Hz;
durCe environ 0.2 rns. Ces donnCes suggbrent que les effets anticrises proviennent de la stimulation des fibres amytliniques affkrentes de petit diametre du nerf vague. Ces rCsultats peuvent
constituer une base pour une nouvelle approche thkrapeutique
de lepilepsie.

(P. Genton, Marseille)

ZUSAMMENFASSUNG
Die repetitive Stimulation des zervikalen N. vagus bei Nagern
unterbricht oder bringt motorische durch Strychnin erzeugte Anfalle sowie durch Pentylentetrazol (PTZ) erzeugten Tremor zum
Verschwinden. Tremor wurde definiert als rhythmische alternierende Kontraktion von antagonisierenden Muskelgruppen mit
weniger Kraftentfaltung als bei Antfallen. Die Anfalle wurden
durch Bolusinjektion von Strychnin oder PTZ in 1 4 Minuten
Intervallen bis zu anhaltender Muskelaktivitat im EMG erzeugt.
Die Vagusstimulation beendete die Anfalle in 0.5 bis 5.0 Sek. Es
gab verlangerte Perioden ohne spontanes EMG nach der Stimulation. Die Protektionsdauer betrug ungefahr das Vierfache der
Stimulationsperiode. Die antikonvulsive Wirkung der Valgusstimulation wurde durch Transektion des Vagus distal zur Stimuluselektrode nicht verandert. Als optimaler Stimulusparameter
wurde geschatzt: Reizstarke 20 Volt (Elektrodenwiderstand 1-5
kOhm); Frequenz 20 bis 30 Hz; Reizdauer 0,2 ms. Die Daten
zeigen, daR die antikonvulsive Wirkung von der Stimulation
kleinkalibriger afferenter, unmyelinierter Vagusfasern herruhrt.
Die Ergebnisse konnen vielleicht die Grundlage eines neuen
therapeutischen Ansatzes bei Epilepsie sein.
(C. K. Benninger, Heidelberg)