J Neurosurg 84:203–214, 1996

Chronic electrical stimulation of the ventralis intermedius

nucleus of the thalamus as a treatment of movement
disorders
ALIM LOUIS BENABID, M.D., PH.D., PIERRE POLLAK, M.D., DONGMING GAO, M.D.,
DOMINIQUE HOFFMANN, M.D., PATRICIA LIMOUSIN, M.D., EMMANUEL GAY, M.D.,
ISABELLE PAYEN, M.D., AND ABDHELHAMID BENAZZOUZ, PH.D.
Department of Clinical and Biological Neurosciences, INSERM Preclinical Neurobiology, Joseph
Fourier University of Grenoble, Hôpital A. Michallon, Grenoble, France; and Department of
Physiology, Jinzhou Medical College, Jinzhou, Liaoning, Peoples Republic of China

 Tremor was suppressed by test stimulation of the thalamic ventralis intermedius (VIM) nucleus at high frequency
(130 Hz) during stereotaxy in nonanesthetized patients suffering from Parkinson’s disease or essential tremor.
Ventralis intermedius stimulation has since been used by the authors over the last 8 years as a treatment in 117 patients
with movement disorders (80 cases of Parkinson’s disease, 20 cases of essential tremor, and 17 cases of various dys-
kinesias and dystonias including four multiple sclerosis). Chronic electrodes were stereotactically implanted in the
VIM and connected to a programmable stimulator. Results depend on the indication. In Parkinson’s disease patients,
tremor, but not bradykinesia and rigidity, was selectively suppressed for as long as 8 years. Administration of L-Dopa
was decreased by more than 30% in 40 Parkinson’s disease patients. In essential tremor patients, results were satis-
factory but deteriorated with time in 18.5% of cases, mainly for patients who presented an action component of their
tremor. In other types of dyskinesias (except multiple sclerosis), results were much less favorable. Fifty-nine patients
underwent bilateral implantation and 14 other patients received implantation contralateral to a previous thalamotomy.
Thirty-seven patients (31.6%) experienced minor side effects, which were always well tolerated and immediately
reversible. Three secondary scalp infections led to temporary removal of the implanted material. There was no per-
manent morbidity. This tremor suppression effect could be due to the inhibition or jamming of a retroactive loop.
Chronic VIM stimulation, which is reversible, adaptable, and well tolerated even by patients undergoing bilateral
surgery (74 of 117 patients) and by elderly patients, should replace thalamotomy in the regular surgical treatment of
parkinsonian and essential tremors.

KEY WORDS • Parkinson’s disease • essential tremor • thalamus •

ventralis intermedius nucleus • deep brain stimulation • stereotaxy

modern surgical treatment of tremor was intro-
T
HE
duced 30 years ago when an unexpected intraoper-
ative complication, which occurred during pyrami-
dotomy,59 led Cooper and colleagues22–24 to the discovery
that destruction of portions of the thalamus suppressed
tremor in a patient with Parkinson’s disease. The acciden-
tal lesion of the anterior choroid artery induced an infarc-
tion of the anterior and lateral parts of the thalamus, and at
the end of the surgical procedure, the patient was perma-
nently relieved of her tremor, although the pyramidotomy
had not been performed. According to the previous work
of Hassler,37 this accidental observation was exploited and
extensive work was done to determine the best functional
target. Although multiple lesions were advocated to
relieve both bradykinesia and tremor, which led to per-
forming pallidotomy for bradykinesia and thalamotomy
for tremor during the same procedure, it appeared quite
rapidly that the most effective target was in the thalamus
and that the most sensitive symptom was tremor. Over
the years, this target was progressively restricted to the
posterior part of the motor area of the lateral thalamus and
finally to the ventralis intermedius (VIM) nucleus, where
smaller lesions were achieved.4,36,38–40,42,51,62,68,69,71,77,89,90,
93,94
Although a spectacular and total suppression of trem-
or could be obtained in a high percentage of patients
undergoing surgery, recurrences were nonetheless ob-
served in approximately 4% to 20%42 of these patients
after several weeks or years. When the lesion size was
increased to prevent this recurrence, morbidity such as
motor deficit, dystonia, speech disturbance, and sensory
loss were observed with a reported frequency of approxi-
mately 25% transitory deficits and 2% to 9% permanent
deficits.87,89 In addition, bilateral procedures, although
reported to be feasible,57 were often associated with severe
neuropsychological deficits56 and were thus rarely per-
formed as a routine procedure.

J. Neurosurg. / Volume 84 / February, 1996 203

 A. L. Benabid, et al.

The advent of L-Dopa27 therapy totally changed the TABLE 1

therapeutic landscape of the disease, and for decades the Description of patient population
surgical treatment of Parkinson’s disease was practical-
ly withdrawn from the therapeutic arsenal. In the early No. of No. of No. of
No. of Bilat Contralat Bilat
1970s, the follow-up examination of L-Dopa–treated pa- Disease Patients Implantation Thalamotomy Surgery
tients began to reveal some drawbacks of the substitutive
treatment such as intolerance, disappearance of efficacy, Parkinson’s disease 80 38 8 46
or even added complications such as abnormal involun- essential tremor 20 13 2 15
tary movements, due not only to the evolution of the dis- dyskinesia 17 9 4 13
total 117 60 14 74
ease but also to the therapy itself. A recurrent need for sur-
gical treatment was born and stereotactic procedures for
thalamotomies were once again performed.9,30,32,33,85
To avoid surgical complications, accurate localization ease and essential tremors and even more when a bilater-
of the most effective target must be achieved during sur- al procedure is necessary.
gery, using electrophysiological methods: deep brain re-
cordings3,4,35,82,91 aimed at the recognition of specific dis-
charge patterns in the surrounding structures such as the Clinical Material and Methods
ventroposterolateral (VPL) sensory thalamus in front of Patient Selection
which is situated the VIM target or inside the VIM target
itself.68,69,73,77 When electrical recording is not available, From January 1987 to July 1994, 117 patients under-
electrical stimulation can be used to detect areas sur- went implantation with uni- or bilateral electrodes (177
rounding the VIM nucleus that must be avoided, such operated sides). Tremor amplitude at rest, during posture
as the sensory VPL nucleus or the pyramidal tract. Dur- handling, or during action and intention maneuvers, was
ing these procedures we observed that electrical stimula- considered severe in all patients and scored as four of four
tion was able to inhibit completely and immediately both on a five-point scale31 with an assessment of overall sever-
parkinsonian rest tremor and postural tremor. However, ity of four of four by both the patient and the examiner.
this effect, which was immediately reversible when the This protocol received the approval of the Grenoble
stimulation was discontinued, was only obtained at high University Hospital ethical committee. All patients under-
frequency (100 Hz and more). Similar observations have went a hospitalization period first during which they were
been previously reported: microelectrode recordings of evaluated.
neuronal activity in the human thalamus revealed the
presence of burst discharges synchronous with tremor in Clinical Examination
VIM.3,5,38,39,46,63,72,93 These authors observed that stimula- Clinical examination was undertaken to confirm the
tion at high frequency through the same microelectrode diagnosis of tremor and its etiology. Quantification of the
suppressed parkinsonian rest tremor immediately and that patients’ symptoms was achieved by clinical measure-
electrocoagulation within that area stopped tremor perma- ments of frequency, amplitude, and type of the tremor and
nently. by videotaping. The tremor was also quantified by ac-
We, as others,42,93 used this effect for several years as an celerometry. Neuropsychological evaluation was system-
intraoperative test for electrode placement. The implanta- atically performed, investigating global cognitive func-
tion of a chronic deep brain stimulation (DBS) electrode tions as well as frontal, attentional functions and motor
was initially proposed as an experimental therapeutic pro- neglect. In all patients, this evaluation protocol was per-
cedure in a patient already thalamectomized on one side formed again between the 3rd and 6th months after
who persistently requested a contralateral procedure and surgery.
who gave his informed consent.13 The result was excel-
lent, and a pilot study of long-term high-frequency VIM Pharmacotherapeutic Test
stimulation in the treatment of disabling tremor was initi-
ated for patients in whom a second operation contralater- All patients submitted to several pharmacological trials
al to a previous thalamotomy was under consideration and at the maximum tolerable dosages. Patients with Par-
for those in whom bilateral surgery was advocated.13–15 In kinson’s received dopaminergic agonists plus peripheral
view of the very encouraging results, we extended the decarboxylase inhibitors (L-Dopa up to 1100 mg/day;
method to other types of dyskinesias, including dystonias bromocriptine up to 30 mg/day; apomorphine up to 50
of various origins and multiple sclerosis tremor, to replace mg/kg per subcutaneous injection) and anticholinergic
conventional destructive thalamotomy.16,77 We report here drugs (trihexylphenydyl up to 6 mg/day). Patients with
our experience with 117 consecutive patients implanted essential tremor received propranolol up to 320 mg/day
with uni- or bilateral electrodes in VIM (177 operated and primidone up to 750 mg/day. Patients who responded
sides), with a follow-up period of more than 7 years (90 to medical therapy were not recommended for surgical
months) for the first patient. The absence of permanent intervention. The details of the population of patients are
complication, the minor side effects when any occurred reported in Table 1.
and their immediate reversibility, the possibility of bilat- Parkinson’s Disease, Essential Tremor, or Other
eral implantation in the same session, and the persistence Dyskinesias
of tremor relief are strong arguments that support chronic
VIM stimulation as the method of choice when a surgical There were 100 patients (with 151 implanted thalami)
procedure is indicated for the treatment of Parkinson’s dis- who exhibited either tremor due to Parkinson’s disease

204 J. Neurosurg. / Volume 84 / February, 1996

Ventralis intermedius stimulation in movement disorders
FIG. 1. A: Schematic drawings over x-ray films for determina-
tion of the ventralis intermedius (VIM) target. The third ventricle
dimensions did not significantly differ between Parkinson’s dis-
ease, essential tremor, and dyskinesia groups. The anterior and pos-
terior commissures (AC–PC) = 26.25 mm ⫾ 2.02 mm. Thalamus
height = 17.38 mm ⫾ 1.76 mm. Third ventricle width at the level
of electrode implantation = 6.69 mm ⫾ 2.51 mm. B: Graphs
showing general representation of the actual position of the mono-
polar VIM electrodes in the first 73 patients (108 thalami) and the
active contacts of the tetrapolar VIM electrodes in the last 44
patients (69 thalami).
(80 cases, 118 sides) or essential tremor (20 cases, 33
sides). Seventeen patients with 26 implanted thalami
(four multiple sclerosis, five dystonias, one writer’s
cramp, seven posttraumatic or posthemorrhagic midbrain
tremors) were operated on according to the same proce-
dure. One patient with dystonia muscularis deformans was
admitted to our service in critical condition, and the pro-
cedure was agreed upon with the patient’s parents and
medical team as a humanitarian therapeutic trial. This was
later justified by the favorable outcome.
Surgical Procedures
After induction of brief, reversible general anesthesia
(diprivan administered intravenously) spontaneously
breathing patients were placed in a Talairach stereotac-
tic frame (at the focus of a long distance (3.5 m) biorthog-
onal x-ray system). Anteroposterior and lateral x-ray films
were obtained at an average magnification coefficient of
1.05. Contrast ventriculography was performed by free-
hand tapping of the frontal horn of the ventricle through a
twist drill (90 mm from nasion, 25 mm from midline, 65
mm long Cushing cannula). Serial images were obtained
during and after a 6.5-ml bolus injection of contrast medi-
um (Iopamiron) with the patient supine and then prone.
J. Neurosurg. / Volume 84 / February, 1996
This allowed a very precise delineation of the midline of
the third ventricle and of the anterior commissure (AC)
and posterior commissure (PC).
The target of implantation has been VIM, considered to
be the target for thalamotomy most effective in relieving
tremor.36,62,66,69,71,73,76,77,93,94 The coordinates of VIM have
been calculated from the lateral x-ray view according to
a proportional geometric scheme based on the AC–PC
line34,50,88 (Fig. 1A). The posterior and anterior limits of
VIM on the AC–PC line are situated at the 2/12 and 3/12
of the AC-PC line length, ahead of the PC. The laterality
of the theoretical target was equal to 11.5 mm from the lat-
eral wall of the third ventricle.93 On the lateral view, the
electrode projection was superimposed on the main axis
of the VIM nucleus. After assessment in the initial cases,
electrodes began to be placed more on the anterior than on
the posterior border of the VIM, to diminish or even avoid
the spreading of current to the primary somatosensory tha-
lamic relay VPL, which induces contralateral paresthe-
sias. On the anteroposterior view, the trajectory of the
electrode was parallel to the midsagittal plane in 109
patients (162 electrodes). In eight patients (15 electrodes),
an oblique trajectory was chosen (6˚–10˚ from the mid-
sagittal plane) to be aligned on the main axis of the VIM.
In the first 22 patients (28 sides), an electrode with an
outer diameter of 2.3 mm (Radionics, Burlington, MA)
was inserted into the VIM target through a 2.3-mm diam-
eter burr hole and was used to stimulate structures during
the electrode placement procedure. In the last 95 patients
(149 sides), it was replaced by a semi-microelectrode69 or
by a bipolar concentric stimulating-recording electrode
with an outer diameter of 0.62 mm (model 17-300-1;
Frederic Haer and Co., Brunswick, ME). Spontaneous as
well as evoked multiunit activities were recorded at vari-
ous sites along the trajectory down to the AC–PC line.
The neuronal activity was recorded using conventional
preamplifiers (model DAM-5A; World Precision Instru-
ments, Hertfordshire, England), AC–DC amplifiers (Neu-
rolog NL106; Digitimer Research Instruments, Hertford-
shire, England), filters (Neurolog NL125; Digitimer
Research Instruments), and spike triggers (Neurolog
NL201; Digitimer Research Instruments), and then
processed through a MacLab 4 WPI system and a Mac-
Intosh II cx computer. Stimulation was also performed at
various sites along the trajectory, using the same semimi-
croelectrode and a constant-current stimulator with an iso-
lation unit (models Accupulser A310 and A365R; World
Precision Instruments). The exact position of each record-
ing and/or stimulating site was checked on x-ray film and
imported into the final operating scheme. The effect of
stimulation on tremor was quantified using an accelerom-
eter attached to the patient’s finger. Usually, several sub-
sequent electrode tracks were performed parallel to the
midsagittal plane and aimed at detecting the anterior and
posterior limits of the VIM; these were also used to check
the laterality. The expected effects of 130-Hz stimulation
on tremor (tremor suppression with the lowest (0.2–2 mA)
electrical intensity) was the major criterion in choosing
the final placement of the chronic electrode. When they
could be recorded, bursting cells synchronous to tremor
were also considered to be a highly significant feature of
the VIM. Finally, the body area in which paresthesias
were induced by stimulation of VPL behind VIM could
205

FIG. 2. Graph depicting the evolution of the current threshold
along the electrode track, exploring the thalamus from anterior
commissure (AC) to posterior commissure (PC), in 11 patients.
also help in defining the best laterality according to the
known somatotopic organization, with lower-limb senso-
ry neurons lying more laterally than hand and mouth neu-
rons. Because of individual variations, the final target
could therefore be significantly different from the theoret-
ical target. When localization was considered satisfactory,
the recording-stimulating electrode was removed and re-
placed by a chronic DBS electrode (Medtronic, Inc.,
Minneapolis, MN). The first 108 electrodes were mono-
polar (model SP5535; Medtronic, Inc.) with an insulated
tip 1.2 mm in diameter and 4 mm long. The last 64 elec-
trodes were tetrapolar (model 3387; Medtronic, Inc.) with
four contacts 1.2 mm in diameter and 1.5 mm long each
separated by 1.5 mm. The actual placement of electrodes
in the 117 patients is reported on Fig. 1B. In both cases,
this electrode was secured to the skull by a knot anchored
in the bone and sealed by a drop of methylmethacrylate
dental cement. This electrode was connected to a lead
externalized through the skin at the level of the parietal
area. All incisions were closed by nonresorbant sutures.
The patient was allowed to recover for 24 hours and
then underwent test stimulation for approximately 1 week.
When the test was considered satisfactory, a programma-
ble stimulator (Itrel I in the first 23 patients and Itrel II in
the last 94 patients; Medtronic, Inc.) was implanted in the
subclavicular region on the same side as the electrode
while the patient received general anesthetic. In patients
undergoing bilateral operations, two stimulators allowed
separate and independent stimulation of each VIM nucle-
us. All surgical procedures were performed without sys-
temic antibiotic administration. Skin incisions were local-
ly irrigated with rifampin during surgical procedure and
prior to being closed.
Stimulation Parameters
During the stereotactic procedure, test stimulation was
done with a 60-␮sec pulse width frequency at 130 Hz and
a current intensity that varied from 0.1 to 10 ␮A. The
parameters of the implanted stimulator were set at 60-
␮sec pulse width, 130 Hz, and 0.5 to 8 volts. They were
adjusted according to the needs of each patient during the
follow-up period. If needed, the frequency range of the
stimulator (Itrel II) could be extended to 185 Hz.
Evaluation of Benefits and Follow Up
After implantation of the stimulators, patients were kept
206
A. L. Benabid, et al.
FIG. 3. Electrophysiological recording showing effects of ven-
tralis intermedius stimulation on the parkinsonian tremor in a rep-
resentative patient and polarity dependence of stimulation-induced
tremor suppression. The electrode penetration was anteroposterior-
ly into the right thalamus. The position of the electrode was ⫺3
mm to the anterior and posterior commissure line and the stimula-
tion intensity was 0.48 mA at 130 Hz. The tremor was recorded at
the left index with an accelerometer. The accelerometric recording
of tremor (upper trace) shows that negative polarity stimulation
(lower trace) is more effective than positive polarity.
at the hospital for 1 week to evaluate the effects of stimu-
lation on the tremor. Video recordings as well as com-
puterized tomography, magnetic resonance imaging, so-
matosensory evoked potentials, and in four cases, positron
emission tomography evaluation of the cerebral blood
flow28 were obtained. The clinical assessment of VIM
stimulation effects was based on a five-point scale: 4 =
complete disappearance of tremor in all circumstances;
3 = reappearance of a slight tremor on rare occasions (for
example, under stress, mental calculation, or motor acti-
vation); 2 = moderate benefit; 1 = slight but definite ben-
efit without any real improvement in daily living; and 0 =
no benefit.
Results
Intraoperative Effects of VIM Stimulation
During insertion of the stimulating–recording test elec-
trode (Fig. 2), stimulation at 130 Hz, 60-␮sec pulse width
was able to suppress the tremor with current intensities
that decreased as long as the electrode approached the
optimum site of stimulation, which usually corresponded
to the VIM area, as determined by the scheme of Guiot
and colleagues.34–36 The lowest value of this threshold was
used to determine the site of maximum effectiveness into
which the DBS electrode was finally implanted. Con-
tinuing progression of the electrode beyond this point usu-
ally necessitated increasing current values to suppress
tremor, whereas permanent paresthesias actually induced
by stimulation of VPL were progressively and more in-
tensely induced. In the VIM area, stimulation with cur-
rents as low as 0.2 mA induced immediate suppression of
the tremor, in a current intensity–dependent manner. At
sufficient stimulation intensity, the tremor disappeared at
the onset of the current, with no more than 1 or 2 seconds
of delay, and recurred almost as immediately when the
current was turned off. When a posteffect was observed, it
did not last more than 10 to 20 seconds. The effect was
polarity sensitive, the stimulating electrode being nega-
J. Neurosurg. / Volume 84 / February, 1996
Ventralis intermedius stimulation in movement disorders
FIG. 4. Electrophysiological recordings in ventralis intermedius
(VIM). A: Multiunit recording showing that the neural noise
amplitude is high in the caudate nucleus (⫺23 mm from the ante-
rior and posterior commissure line), then decreases (from ⫺15
to ⫺8 mm), and increases again when the electrode enters the
VIM. B and C: Single-unit recordings (upper trace) of bursting
cells recorded in the VIM (B) synchronous to tremor (accelerome-
try, lower trace). These cells are replaced by an electrical silence
when the electrode leaves the VIM to enter the internal capsule (C).
tive. When the polarity was reversed, the effect disap-
peared or at least was strongly decreased (Fig. 3). Brief
paresthesias, which disappeared within 10 seconds, were
often described by the patient at the onset of stimulation.
During progression of the electrode toward the final tar-
get, a decrease in spontaneous tremor intensity was often
observed, together with increased stimulation posteffects
and difficulty in reinducing the tremor. This was consid-
ered a minor equivalent to the thalamotomy-like effect,
which could be classically observed when entering the
thalamotomy target with a larger electrode. In the VPL
area, stimulation that was less effective in reducing trem-
or as the electrode was more distant from VIM somato-
topically induced permanent paresthesias in various areas
of the body.
Extracellular Recording of Neuronal Activity
Neuronal activity was recorded along the track. With
a semimacroelectrode, only multiunit activity was record-
ed: the amplitude of the spikes, described as “neural
noise,”70,71 varied along the track and provided informa-
tion about the limits of the different nuclei. This neural
noise was high in the VIM and strongly diminished when
the electrode entered the internal capsule (Fig. 4A). With
the microelectrode, single units were recorded in the VIM
and VPL. In the VIM, cells fired by bursts consisting of
five to 10 large spikes, and some cells had a spontaneous
bursting activity independent of any peripheral stimula-
tion or muscular activity. Some other cells were definitely
synchronous with the tremor (Fig. 4B and C) and also
responded to passive movements of the limbs. Similarly,
evoked activities were easily recorded in the VPL in
J. Neurosurg. / Volume 84 / February, 1996
FIG. 5. Electrophysiological recordings in the ventroposterolat-
eral sensory thalamus. Neural activity evoked by skin taps and light
pressure was easily recorded.
response to superficial stimulation of the skin (touch,
pressure) (Fig. 5).
Postoperative Effects of VIM Stimulation
Effects of VIM Stimulation on Tremor and Other Ex-
trapyramidal Symptoms. Tremor was essentially the only
symptom significantly influenced by VIM stimulation
(Table 2). Rigidity was only slightly affected because the
cogwheel syndrome was reduced when tremor was sup-
pressed. There was almost no change in bradykinesia or
in any other symptom of Parkinson’s disease. Unilateral
pain, which accompanied severe tremor and rigidity in
many cases, also was greatly reduced.
During the test period, the external extension of the
DBS electrode was used to study the current threshold as
a function of frequency, and this showed, as Fig. 6 demon-
strates, that the optimum frequency was in the 100 Hz to
1000 Hz range with a minimum at approximately 250 Hz.
In all patients, clinical modifications in the tremor were
quite obvious: at the onset of stimulation, tremor disap-
peared within 2 seconds and remained absent provided
that the intensity of the stimulation was sufficient. Dis-
continuation of stimulation always resulted in an im-
mediate return of the tremor.
In patients with Parkinson’s disease, drug therapy was
often reduced after VIM stimulation: dopamine agonists
were maintained in all but two patients to treat bradykine-
sia or to enhance the effects of VIM stimulation. Thirty-
nine (48.7%) of 80 patients with Parkinson’s disease had
their dopamine dosage decreased by 20% at 3-month fol-
low-up review, but only 12 (15%) at the last follow up
continued at the decreased dosage, due to progression of
the disease. Because stimulation effects were adjustable,
results are given according to each patient’s choice, based
on an individual compromise between benefit and adverse
effects.
During the initial postoperative period (3 months),
scores of 3 or 4 were obtained in the upper limbs in 102
(91.9%) of the 111 operated sides (Table 2), whereas 7.2%
of the sides had scores of 2 or 1, and 0.9% had a score of
0. At the last follow-up examination, these scores were
88%, 10.8%, and 0.9%, respectively. Global scores for the
four limbs were slightly lower, rated 4 or 3 in 86% at 3
months and then 85% at the last follow up. Resting or pos-
tural tremor seemed to be better controlled than action
207

TABLE 2
Effects of ventralis intermedius stimulation according to
disease type*
No. of Implanted Sides (%)
First Follow Up Last Follow Up
Score at UL† (3 mos) (at least 6 mos)
Parkinson’s disease (111 electrodes)
0 1 (0.9) 1 (0.9)
1 3 (2.7) 5 (4.5)
2 5 (4.5) 7 (6.3)
3 31 (27.9) 55 (49.5)
4 71 (64.0) 43 (38.7)
global score (86) (83)
essential tremor (36 electrodes)
0 0 (0.0) 6 (16.7)
1 3 (8.3) 2 (5.6)
2 6 (16.7) 6 (16.7)
3 13 (36.1) 18 (50.0)
4 14 (38.9) 4 (11.1)
global score (69) (59)
* Abbreviations: UL = upper limb; LL = lower limb.
† The results are related to the number of implanted sides, because, in
bilaterally implanted patients, they may be different from one side to the
other. The global scores represent the evaluation of patients who had
scores of 3 or 4 at both UL and LL (see Fig. 8), and this explains why the
percentage of these scores is lower than the sum of scores 3 and 4 at the
UL alone.
tremor. In addition, distal limb tremor was more easily
suppressed than proximal or axial tremor.
In patients with essential tremor, during the initial post-
operative period, scores of 3 or 4 were obtained in the
upper limbs in 27 (75%) of the 36 operated sides (Table
2), whereas 25% of the sides had scores of 2 or 1, and
none had a score of 0. At the last follow up, these scores
were obtained in 61.1%, 22.2%, and 16.7% of operated
sides, respectively. Global scores for the four limbs were
identical, as essential tremor involves mainly the upper
limbs, rated 4 or 3 in 75% of operated sides at 3 months
and then 61% at the last follow up.
Other kinds of dyskinesias (such as writer’s cramp,
posttraumatic dyskinesia, multiple sclerosis tremor, dys-
tonias) were inconsistently, less significantly, or not im-
proved. If improvement was achieved, it lasted only a few
months. This is comparable to the results of thalamoto-
my.9,10 However, whereas the effects of VIM stimulation
did not achieve a satisfactory result with regard to the clin-
ical scoring scale used for tremor, there was significant
improvement observed in the quality of daily living. In the
two patients with familiar dystonia the parameters that
proved to be the most effective were significantly differ-
ent from those used in Parkinson’s disease tremor and
essential tremor, namely pulse width, which was 450
msec. The establishment of the effect was delayed, and the
parameters had to be changed during rather long periods
of observation to check the effects. When the stimulators
were shut off or when the batteries were depleted, the
patients’ clinical status quickly worsened and the stimula-
tors had to be changed immediately, after which it took
several days for patients to recover to their previous clin-
ical status. The effect of VIM stimulation in these two
patients consisted mainly of a decrease in limb rigidity
208
A. L. Benabid, et al.
FIG. 6. Graph depicting frequency–intensity relationship of the
ventralis intermedius stimulation threshold for tremor suppression
effect.
and in saliva inhalation, an increased facility in swallow-
ing, and reduction in hip, knee, and elbow flexed posi-
tions, all leading to a global improvement in general sta-
tus. Two of the four multiple sclerosis patients had a good
or fair benefit.
Side Effects. Side effects were always reversible, mild,
and accepted, as long as VIM stimulation intensity could
be set at a level at which these side effects became tolera-
ble for the patient and tremor relief was still good or excel-
lent. Paresthesias (9%) were usually induced at intensities
of stimulation higher than those that suppressed the trem-
or and lasted a few minutes or even seconds after the on-
set of VIM stimulation. However, they were much less
intense when electrode placement was more anterior,
close to the anterior border. These paresthesias were never
painful except in one patient whose electrode was placed
2 mm below the AC–PC line and who experienced burn-
ing sensations in the perioral and ophthalmic area. These
disappeared when the electrode was withdrawn 2 mm.
Higher amplitudes than those needed to suppress tremor
could result in the induction of a slight cerebellar dys-
metria interpreted as a spreading of current to cerebellar
pathways. In 9% of patients, a dystonia of the foot was
observed after approximately 12 months of stimulation.
This was reversible when stimulation was discontinued
but resumed immediately at onset and therefore required
the patients to use slightly lower parameters to avoid this
dystonia. Dysarthria was observed in 23 patients (19.6%;
18 Parkinson’s disease, five essential tremor): only five
(4.2%) had unilateral stimulation, 14 (12%) had bilateral
stimulation, and four (3.4%) had a previous contralateral
thalamotomy. When compared to the numbers of patients
in each group, dysarthria was observed in 14 (27.5%) of
the 51 bilaterally stimulated patients but in four (40%) of
the 10 patients who underwent thalamotomy on one side
and stimulation on the other side. This shows that thala-
motomy has more side effects than VIM stimulation. Dis-
equilibrium was observed in 10 patients (six Parkinson’s
disease, four essential tremor), of whom six were bilater-
ally and four unilaterally stimulated, among whom three
had previous contralateral thalamotomy. Contralateral
dystonia was observed in six patients (four Parkinson’s
disease, two essential tremor), of whom four were bilater-
J. Neurosurg. / Volume 84 / February, 1996
Ventralis intermedius stimulation in movement disorders
FIG. 7. Graphs showing the evolution of the intensity of ven-
tralis intermedius stimulation required to suppress totally the trem-
or with time (A), and the evolution of electrode impedance mea-
sured on Itrel II stimulators (B).
ally stimulated and two unilaterally. Bilateral stimulation
induced perioral paresthesias in one patient with Par-
kinson’s disease and hypersalivation in one patient with
essential tremor. Bilateral stimulation did not induce the
neuropsychological deficits usually reported following
bilateral thalamotomies. No patient spontaneously com-
plained of symptoms of frontal disturbance, but neuropsy-
chological tests showed a slight, lateralized decrease in
performance fluency, especially affecting verbal perfor-
mance, when the left VIM was stimulated, and spatial per-
formance, when the right VIM was stimulated. Detailed
results will be reported elsewhere.
Follow-Up Data on VIM Stimulation Parameters. In 22
(20.5%) of 107 patients and 23 (15%) of 153 electrodes,
introduction of the test electrodes during the procedure
induced total suppression of the tremor when the electrode
reached the target. Assessment of the efficacy of external
stimulation had to be delayed in many patients because of
temporary suppression or reduction of the tremor induced
by the microthalamotomy-like effect of electrode implan-
tation. The tremor reappeared after 1 to 10 days, in many
cases at a lower amplitude than before surgery. This
“target impact effect” (or thalamotomy-like effect) usual-
ly disappeared within 1 week and the tremor resumed.
Nevertheless, in all cases the intensity of VIM stimulation
necessary to alleviate tremor had to be increased during
the first 3 weeks from an average initial value of 1 V ⫾
0.5 V to approximately 4 V ⫾ 0.5 V. It reached a plateau
after approximately 2 months at a mean value of 3.06 V
(Fig. 7A). Using the Itrel II stimulators, it was possible to
measure the electrical impedance of the brain structures
stimulated by the implanted electrode. This parameter
increased significantly during the first days, from a mean
of 794 ohm (range 499–1238 ohm) on Day 14 and then
more slowly during the following weeks to 1057 ohm
J. Neurosurg. / Volume 84 / February, 1996
FIG. 8. Bar graphs showing clinical results for Parkinson’s dis-
ease, essential tremor, and dyskinesia patients, expressed in per-
centage versus the clinical status, rated at 3 months after surgery
(initial) and at the last follow up (final).
(range 828–1483 ohm) on Day 106 (Fig. 7B). This in-
crease in impedance could account, at least in part, for the
increase in threshold current intensity observed during
clinical follow up of treated patients. The average current
charge density with the first electrodes (DBS SP 5535;
Medtronic) (60-msec pulse width, 130 Hz) was 2.89
mA/16.2 mm2 = 179 A/m2.
The effect of VIM stimulation (Fig. 8) remained stable
more often in patients with Parkinson’s disease (96% of
the 102 stimulated sides) than in patients with essential
tremor (81% of the 27 stimulated sides), although essen-
tial tremor is considered to be the best indication for thal-
amotomy:36 complete suppression of the tremor was
achieved with rather low stimulation voltages, which were
stable after the initial postoperative increase. Dis-
continuation of VIM stimulation led to a recurrence of the
tremor almost identical to that observed during the preop-
erative test period.
However, in 36% of cases (two of 20 with essential
tremor, 33 of 78 with Parkinson’s disease), discontinua-
tion of the VIM stimulation resulted in a rebound effect,
consisting of an increased intensity in the tremor as com-
pared to the preoperative intensity. Patients who previous-
ly had no tremor while sleeping then required VIM stim-
ulation 24 hours a day, because their tremor without
stimulation had increased to a level that did not allow
them to fall asleep. Sixty-four percent of patients (18 of 20
with essential tremor, 45 of 78 with Parkinson’s disease)
were able to stop their stimulator at night.
In four (44%) of the nine patients who had an initial
score at 4 or 3 and were then rated at 2 or 1 at their last
follow up, improvement in the activities of daily life was
209

still significant. This was mainly the case for Parkinson
patients with rest tremor, and this loss of efficacy appeared
at 17.2 ⫾ 9.6 months.
In five (55%) of these nine patients, this decrease in
score was no longer associated with a significant func-
tional improvement. These were primarily patients with
an action component to their tremor, and this loss of effi-
cacy appeared more quickly, at 7.2 ⫾ 1.5 months postop-
eratively.
Mortality and Morbidity. There was no operative mortal-
ity. One patient died suddenly on the 11th postoperative
day from pulmonary embolism due to previously existing
cardiovascular insufficiency, although he had recovered
from stereotactic surgery and was able to walk into the
neurosurgery department. Five others died from various
nonneurological diseases at 3, 6, 7, 10, and 23 months. In
six patients, a microhematoma was induced by electrode
insertion: three were asymptomatic, discovered only on
routine postoperative computerized tomography scanning,
two were responsible for transient motor neglect, and one
occurred in a patient with multiple sclerosis and induced
an acute deficit on the 8th day, also reversible in 3 months.
Five patients suffered from skin problems: three had a late
scalp infection, due to skin necrosis in front of the cable
connectors in two female patients with thin scalps. In
these three cases, electrode and connector removal and
further replacement were needed to heal the wounds. Two
patients had a granuloma along the connector extension
track, and one patient had transient fluid collection in
the subclavicular pocket of the stimulator. In this series
no epileptic seizures were induced by thalamic kindling.
There was no complication of ventriculography. This
procedure is safe when strict landmarks are observed, and
reported complications are often due to use of cannulas
longer than 65 mm.21,55 Comparable low morbidity has
been already observed during VPL chronic stimulation for
pain.1,43
Discussion
The intraoperative suppressive effect of thalamic stim-
ulation on parkinsonian tremor has long since been re-
ported in the same site in which thalamotomy would also
suppress tremor.6,8,42,63,76,80,92,93 It has also been reported
that stimulation increased or triggered the tremor.66,69
Dependence on the effect of frequency was observed,64
but it was not clearly established that high frequency was
a critical parameter, although in some reports6 this stimu-
lation was performed at 200 Hz. Surprisingly, attempts to
apply this observation as a permanent treatment were rare,
reported as poorly effective and short lived92 or involving
the centrum medianum and the intralaminar nuclei,11,12,64
zona incerta,19 the sensory VPL nucleus,25,58,64,84 the pul-
vinar and dentate nucleus,61 but not the VIM nucleus.
Efficacy of VIM Stimulation on Tremor Arrest
As a general rule, the effectiveness of VIM stimulation
reproduces that classically obtained for thalamotomy: Par-
kinson’s disease tremor and essential tremor are the best
indications, and complete arrest can be expected when the
patients are carefully selected and the surgical stereotactic
procedure is correctly performed. The results are immedi-
210
A. L. Benabid, et al.
ate, spectacular, and totally reversible without significant
posteffect and without permanent side effect. However,
this effect is selective, as the different effects of VIM stim-
ulation on Parkinson’s disease tremor, essential tremor,
and other dyskinesias suggest that the corresponding un-
derlying mechanisms are also different, at least with re-
gard to the role of the VIM.
Reduced Morbidity of Unilateral and Bilateral VIM
Stimulation
Because the procedure is safe and devoid of tissue
destruction after electrode insertion, it minimizes the risk
of bleeding or progressive edema. The procedure was
always well tolerated, even by the oldest patient who was
81 years old. Bilateral implantation during one surgical
session (in 60 of 117 patients; 51%) or complementary
implantation on the contralateral side of a previous con-
tralateral thalamotomy (in 14 of 117 patients; 11.9%) did
not induce any of the neuropsychological deficits fre-
quently reported in cases of bilateral thalamotomy. This,
by itself, is a unique advantage of VIM stimulation and
provides a surgical solution for patients in whom bilateral
thalamotomy would be indicated. Because the side effects
are immediately reversible, the patient has the opportuni-
ty to choose between the benefit of VIM stimulation,
which suppresses the tremor, and the eventual side effect,
or simply to lower the intensity of stimulation to the level
at which the side effects are reduced and the benefit still
significant.
Tolerance Phenomenon in VIM Stimulation
In this series, stimulation amplitude had to be increased
in all patients. The initial increase was more rapid during
the first 3 weeks and was related to tissue changes around
the electrode. The need for the late increase in stimulation
amplitude was because of the progression of Parkinson’s
disease or the development of tolerance. The first hypoth-
esis could explain the late recurrences of tremor several
months after thalamotomy, although this does not really
happen when placement of the lesion has been correct.29,36
The second hypothesis assumes that chronic stimulation
of the VIM nucleus would become less effective with
time, therefore demanding an ever increasing amplitude of
stimulation. The observed increase in electrode imped-
ance can account only for the early and not the late toler-
ance to VIM stimulation. Tolerance was not due to deple-
tion of the power batteries: after changing the generator
parameters were not significantly different. The Itrel I
stimulators have already been replaced in 15 of 19 patients
after 34 ⫾ 15 months (range 17–63 months). The life time
of the four Itrel I stimulators not yet replaced ranges from
56 to 80 months. Four Itrel II stimulators have already
been replaced, with a life time of 41.8 ⫾ 7.7 months
(range 31–49 months). The life time of the 77 Itrel II stim-
ulators not yet changed ranges from 1.9 to 56 months.
Tolerance could also be due to a decreased biological
response (habituation) of the neuronal network. This
hypothesis is supported by the fact that after turning off
VIM stimulation, the tremor recurred with a temporary
rebound of its amplitude. The threshold intensity for alle-
viation of tremor had to be regularly increased up to a
final level that could no longer be increased due to the
J. Neurosurg. / Volume 84 / February, 1996
Ventralis intermedius stimulation in movement disorders
induction of paresthesias. This led to a loss of functional
benefit, mainly in patients with intense initial tremor who
needed high intensities 24 hours a day. This tolerance was
reversible after a stimulation holiday when VIM stimula-
tion arrest was acceptable by the patient. A similar phe-
nomenon has been observed during central gray matter
stimulation for pain.43
Mechanism of Action
The intimate mechanisms underlying the effect of VIM
stimulation are unknown as, too, are the mechanisms of
tremor65,74,75 and remain to be studied. It is surprising that
both destruction and stimulation of the same structure
achieve the same effect. Actually, VIM stimulation sup-
presses tremor only when the frequency of stimulation is
at least 100 Hz (Fig. 6). The similarity of this curve with
that commonly known as the intensity–frequency rela-
tionship recorded in frog muscle nerve fibers26 is striking.
This suggests, by analogy, that VIM stimulation involves
mostly passing fibers rather than cell bodies of VIM
neurons.
This paradoxical effect of stimulation when using low-
frequency versus high-frequency currents has been also
reported during stimulation of the intralaminar and mid-
line thalamic nuclei for relief of epilepsy in human
patients98 and has long been known and documented in
experimental animals.2,41,44,47,60 The precise structure that
is stimulated in our series is more than probably the VIM,
although no anatomical control is available. During thala-
motomy, it has been already observed that stimulation in
the VIM can suppress tremor and that coagulation at that
site will achieve the same effect.6,11,92,93 The anatomical
landmarks from the third ventricular commissures are
accurate when a proportional graphic system is used to
construct the position of the target.88,97 The pattern of
semimicroelectrode recording cannot fully characterize
VIM. Detection of the best target using the changes in
neural noise69,71,73,76,77 can be equivocal. Observation of
rhythmic patterns in the VIM3–6,68,69,71,73,76,93 is not always
easy to achieve. Recognition of the specific somatosenso-
ry areas of the thumb and labial commissure in VPL
immediately behind the VIM target seems to be a reliable
landmark5,29,36 and is a strong argument for the identifica-
tion of the effective target as the VIM, and not the VPL.
In our last 72 cases in which a preliminary electrode track
crossed the VIM from front to back, the stimulation
threshold required to suppress tremor was minimal in the
area corresponding to the geometrically defined VIM
from the AC–PC landmarks of Guiot, et al.,34 (Fig. 2) and
significantly reincreased when the electrode entered the
VPL, where the threshold of induced paresthesias was in
turn minimal. Therefore, our target is definitely situated
immediately in front of the VPL and is more than likely
the VIM.
Role of VIM Nucleus in the Arrest of Tremor During
Thalamic Stimulation
Knowledge of the afferent and efferent connections of
the VIM and of the biochemical basis of Parkinson’s dis-
ease does not provide an explanation for this mechanism.
The VIM probably receives vestibular afferents37 as well
as proprioceptive inputs from limbs76–78 and projects onto
J. Neurosurg. / Volume 84 / February, 1996
the motor cortical area.49 However, there are three main
populations of cells that are situated successively and en-
countered by the electrode along a track passing through
the reticular thalamic nucleus, the ventrooral anterior and
ventroposterior parts of the ventrolateral nucleus, and then
the VIM.78 In the reticular thalamic nucleus, cells respond
to verbal command; in the ventrooral–ventroposterior
parts of the ventrolateral nucleus and the VIM, the popu-
lations of cells responding to voluntary and passive move-
ments are mixed, and the rhythmic activities recorded at
these levels either precede or follow the motor activity
recorded by electromyographic (EMG) monitoring.4–6,
54,76,78
It can be assumed that the VIM is a proprioceptive
relay,7,68 receiving postural inputs from the peripheral
joints and muscles. It is also suggested that the VIM can
trigger tremor. Rhythmic activities, recorded in VIM cells
of patients17,46,73,78,93 preceding EMG discharges and disap-
pearing just before tremor arrest, are not suppressed in
monkeys with tremor by section of the dorsal spinal
roots67 or after injection of curare.48,52 Therefore, the VIM
would act as part of a feed-back loop,81 impinging on the
main corticospinal motor pathway and aiming at modula-
tion of the transfer function of the sensorimotor system.
The gain in this loop can be regulated by other afferents,
among which the dopaminergic nigrostriatal system pre-
sumably plays a key role.79 Deletion of this dopaminergic
control would detune the VIM-containing loop and there-
fore lead to a noncritically dampened loop responsible for
the oscillatory behavior that is constitutive of the tremor.
Interruption of this ill-regulated feed-back loop (by
destruction or by stimulation-induced inhibition of the
VIM) would suppress this abnormally oscillatory behav-
ior (and then suppress the tremor) but would also suppress
a normally needed tuning system, thus explaining why
voluntary movement after thalamotomy is never as pre-
cise and skilled as it is in normal persons. On the basis of
the studies of motor activities under the effect of DBS in
humans, it has been proposed that resonance properties of
the motor control circuit are basic features of the motor
system, which are normally dampened by a suppression
mechanism. When this mechanism weakens, external
stimuli or internal impulses may elicit oscillations and
then tremor. This model does not postulate a thalamic
rhythmic center to explain the tremorogenic process.95
Jamming of the VIM, in this hypothesis, would also result
in the same tremor suppression. However, this cannot be a
unique system because we did not have such good results
in other types of dyskinesias (17 patients with 26 thalam-
ic-implanted sides). It may be possible that the VIM is not
an adequate target for these dyskinesias and has to be
replaced by a more appropriate structure, which has yet to
be discovered. For instance, stimulation of a target, which
is 8 mm below and 2.5 mm more medial than ours, has
been reported to be effective on multiple sclerosis inten-
tion tremor,19 compared with the lack of satisfactory and
long-lasting results in our series for this type of dyskine-
sia. However, tremor in multiple sclerosis may be of vari-
ous origins—cerebellar or extrapyramidal—depending on
the location of the responsible sclerotic lesion. Therefore,
VIM stimulation may suppress the extrapyramidal com-
ponent of the tremor and leave the other component unaf-
fected, in particular the cerebellar tremor. Ventralis inter-
211

medius stimulation has to be compared to the VIM thala-
motomy, which is commonly reported to be effective
regardless of the type of dyskinesia, although clinical
description of patients as well as precisely evaluated post-
operative scores are not often available in the neurosurgi-
cal literature.86
Assuming that the data obtained in the monkey45,49,96 on
afferent and efferent connections of the thalamus could be
extrapolated to humans,49 one might consider that the
VIM has strong reciprocal connections from the cerebel-
lum,53 whereas the VPL receives the somatosensory lem-
niscal projections and the ventroposterior part of the ven-
trolateral nucleus receives mainly pallidal inputs. When
the electrode is correctly placed in the medial part of the
VIM or even better in its anterior half, immediately
behind the posterior limit of the ventroposterior part of the
ventrolateral nucleus, the tremor-suppressing effect is
consistently obtained at low thresholds and is free from
any side effects. In these ideal cases, when the intensity of
stimulation is increased purposely, for as long as the
patient feels it is acceptable, the appearance of cerebellar
symptoms such as a dysmetria during the finger-to-nose
maneuver and hypotonia can be observed.
Conclusions
In our experience VIM stimulation is a remarkable ther-
apeutic agent against the tremor of Parkinson’s disease as
well as essential tremor, although it is much less effective
in other types of dyskinesias. This method should there-
fore be applied after careful clinical analysis of the pa-
tient’s case. Ventralis intermedius stimulation is effective
at high frequency. Good results are strongly correlated
with accurate placement of electrodes. Additional electro-
physiological methods such as microstimulation and re-
cording must be used to locate the target precisely and to
achieve the best placement. Criteria such as the intensity
of neural noise, the close vicinity of the thumb–labial
commissure sensory representation in the VPL, as well as
recording of cells bursting synchronously with tremor, are
valuable tools. The types of dyskinesias that do not re-
spond to VIM stimulation depend on other targets that
remain to be discovered. The lack of permanent side
effects and, moreover, the possibility of performing bilat-
eral implantations in one session without any permanent
neuropsychological side effect are the most important
arguments in support of this method. This method can be
proposed systematically as a first alternative, even if thal-
amotomy later becomes necessary at the same site, which
never occurred in our experience. The cost of the proce-
dure has to be taken into consideration. Finally, the mech-
anism of action is still unknown, and if the hypothesis of
a high frequency–induced inhibition or jamming accounts
for the observed effects, a precise explanation is still need-
ed, which might also help in avoiding the phenomenon of
tolerance. Ventralis intermedius stimulation is increasing-
ly used as a treatment for tremor,83 of L-Dopa induced
dyskinesias,18 of persistent hemiballism (T Tsubokawa, et
al., unpublished data), and might have other effects than
tremor suppression.20 Moreover, this procedure does not
exclude the patient from consideration for another type of
treatment such as fetal transplant in the future. It therefore
212
A. L. Benabid, et al.
appears that VIM stimulation can be proposed as a stan-
dard neurosurgical approach to tremor, with the same indi-
cations as those imposed for thalamotomy. Finally, VIM
stimulation provides an experimental opportunity, be-
cause of its reversibility and adjustability, to study and
understand the mechanisms of tremor in Parkinson’s dis-
ease as well as of normal motor control in humans.
Acknowledgments
The authors thank Dr. Elizabeth Boogusch for useful help with
the English text and Mrs. A. Abbadie for typing the manuscript.
References
1. Adams JE, Hosobuchi Y, Fields HL: Stimulation of internal
capsule for relief of chronic pain. J Neurosurg 41:740–744,
1974
2. Akimoto H, Yamaguchi N, Pkabe K, et al: On sleep induced
through electrical stimulation on dog thalamus. Folia Psychiatr
Neurol Jpn 10:117–146, 1956
3. Albe-Fessard D: Interactions entre recherches fondamentale et
clinique. Deux exemples tirés d’une expérience personnelle.
Can J Neurol Sci 15:324–332, 1988
4. Albe-Fessard D: Physiopathologie du Parkinson, in: Le Point
sur la Maladie de Parkinson. Bruxelles: Merck Sharp &
Dohme, 1973, pp 1–20
5. Albe-Fessard D, Arfel G, Guiot G: Activités électriques carac-
téristiques de quelques structures cérébrales chez l’homme.
Ann Chir 17:1185–1214, 1963
6. Albe-Fessard D, Arfel G, Guiot G, et al: Dérivations d’activités
spontanées et évoquées dans les structures cérébrales profondes
de l’homme. Rev Neurol 106:89–105, 1962
7. Albe-Fessard D, Arfel G, Guiot G, et al: [Identification and pre-
cise delimitation of certain subcortical structures in man by
electrophysiology. Its importance in stereotaxic surgery of dys-
kinesia.] C R Acad Sci Paris 253:2412–2414, 1961 (Fr)
8. Alberts WW, Feinstein B, Levin G, et al: Electrical stimulation
of therapeutic targets in waking dyskinetic patients. Electro-
encephalogr Clin Neurophysiol 20:559–566, 1966
9. Andrew J: Surgical treatment of tremor, in Findley LJ, Capildeo
R (eds): Movement Disorders: Tremor. London: MacMillan,
1984, pp 339–351
10. Andrew J, Fowler CJ, Harrison MJG: Stereotaxic thalamotomy
in 55 cases of dystonia. Brain 106:981–1000, 1983
11. Andy OJ: Parafascicular-center median nuclei stimulation for
intractable pain and dyskinesia (painful-dyskinesia). App Neu-
rophysiol 43:133–144, 1980
12. Andy OJ: Thalamic stimulation for control of movement disor-
ders. Appl Neurophysiol 46:107–111, 1983
13. Benabid AL, Pollak P, Gervason C, et al: Long-term suppres-
sion of tremor by chronic stimulation of the ventral intermedi-
ate thalamic nucleus. Lancet 337:403–406, 1991
14. Benabid AL, Pollak P, Hommel M, et al: Traitement du trem-
blement parkinsonien par stimulation chronique du noyau ven-
tral intermédaire du thalamus. Rev Neurol 145:320–323, 1989
15. Benabid AL, Pollak P, Louveau A, et al: Chronic VIM-thalam-
ic stimulation in movement disorders, in Crossman AR, Sam-
brook MA (eds): Neural Mechanisms in Disorders of Move-
ment. Current Problems in Neurology, Vol 9. London: John
Libbey, 1988, pp 413–415
16. Benabid AL, Pollak P, Louveau A, et al: Combined (thalamot-
omy and stimulation) stereotactic surgery of the VIM thalamic
nucleus for bilateral Parkinson disease. Appl Neurophysiol
50:344–346, 1987
17. Bertrand G, Jasper H, Wong A: Microelectrode study of the
human thalamus: functionnal organization in the ventro-basal
complex. Confin Neurol 29:81–86, 1967
J. Neurosurg. / Volume 84 / February, 1996
Ventralis intermedius stimulation in movement disorders
18. Blond S, Caparros-Lefebvre D, Parker F, et al: Control of trem-
or and involuntary movement disorders by chronic stereotactic
stimulation of the ventral intermediate thalamic nucleus. J
Neurosurg 77:62–68, 1992
19. Brice J, McLellan L: Suppression of intention tremor by con-
tingent deep-brain stimulation. Lancet 1:1221–1222, 1980
20. Burleigh AL, Horak FB, Burchiel KJ, et al: Effects of thalamic
stimulation on tremor, balance, and step initiation: a single sub-
ject study. Mov Disord 8:519–524, 1993
21. Cheshire WP, Ehle AL: Hemiparkinsonism as a complication of
an Ommaya reservoir. Case report. J Neurosurg 73:774–776,
1990
22. Cooper IS: Ligation of anterior choroidal artery for involuntary
movements—parkinsonism. Psychiatric Quart 27:317–319,
1953
23. Cooper IS: The Neurosurgical Alleviation of Parkinsonism.
Springfield, Ill: Charles C Thomas, 1956
24. Cooper IS, Bravo GJ: Anterior choroidal occlusion, chemo-pal-
lidectomy and chemo-thalamectomy in Parkinsonism, in Fields
WS (ed): Pathogenesis and Treatment of Parkinsonism.
Springfield, Ill: CC Thomas, 1958, p 325
25. Cooper IS, Upton ARM, Amin I: Chronic cerebellar stimulation
(CCS) and deep brain stimulation (DBS) in involuntary move-
ment disorders. Appl Neurophysiol 45:209–217, 1982
26. Coppée G: La pararésonance dans l’excitation par les courants
alternatifs sinusoïdaux. Arch Int Physiol Biochim Biophys
40:1–58, 1934
27. Cotzias GC, Van Woert MH, Schiffer LM: Aromatic amino
acids and modification of Parkinsonism. N Engl J Med 276:
374–379, 1967
28. Deiber MP, Pollak P, Passingham R, et al: Thalamic stimulation
and suppression of parkinsonian tremor. Evidence of a cerebel-
lar deactivation using positron emission tomography. Brain
116:267–279, 1993
29. Derôme P: Le Noyau Ventral Postérieur de l’Homme. Thesis.
Paris, 1965
30. Derôme PJ, Jedynak CP, Visot A, et al: Traitement des mouve-
ments anormaux par lésions thalamiques. Rev Neurol 142:
391–397, 1986
31. Fahn S, Tolosa E, Marin C: Clinical rating scale for tremor, in
Jankovic J, Tolosa E (eds): Parkinson’s Disease and Move-
ment Disorders. Munich: Urban and Schwarzenberg, 1988, pp
225–234
32. Fox MW, Ahlskog JE, Kelly PJ: Stereotactic ventrolateralis
thalamotomy for medically refractory tremor in post-levodopa
era Parkinson’s disease patients. J Neurosurg 75:723–730,
1991
33. Goldman MS, Ahlskog JE, Kelly PJ: The symptomatic and
functional outcome of stereotactic thalamotomy for medically
intractable essential tremor. J Neurosurg 76:924–928, 1992
34. Guiot G, Arfel G, Derôme P: La chirurgie stéréotaxique des
tremblements de repos et d’attitude. Gaz Med 75:4029–4056,
1968
35. Guiot G, Derome P, Arfel G, et al: Electrophysiological record-
ings in stereotaxic thalamotomy for parkinsonism. Prog Neu-
rol Surg 5:189–221, 1973
36. Guiot G, Derôme P, Trigo JC: Le tremblement d’attitude. In-
dication la meilleure de la chirurgie stéreotaxique. Presse Med
75:2513–2518, 1967
37. Hassler R: Forels Haubenfaszikel als Vestibuläre Empfindungs-
bahn mit Bemerkungen über einige andere sekundäre Bahnen
des Vestibularis und Trigeminus. Arch Psychiatry 180:23–53,
1948
38. Hassler R: The influence of stimulations and coagulations in the
human thalamus on the tremor at rest and its physiopathologic
mechanism. Excerpta Medica Foundation 2:637–642, 1955
39. Hassler R: The pathological and pathophysiological basis of
tremor and parkinsonism. Excerpta Medica Foundation 1:
29–40, 1955 [Unverified]
J. Neurosurg. / Volume 84 / February, 1996
40. Hassler R, Riechert T, Mundinger F, et al: Physiological obser-
vations on stereotaxic operations in extrapyramidal motor dis-
turbances. Brain 83:337–350, 1960
41. Hess WR: The diencephalic sleep center, in Adrian ED (ed):
Brain Mechanisms and Consciousness. Springfield, Ill: CC
Thomas, 1954, pp 117–136
42. Hirai T, Miyazaki M, Nakajima H, et al: The correlation be-
tween tremor characteristics and the predicted volume of effec-
tive lesions in stereotaxic nucleus ventralis intermedius thala-
motomy. Brain 106:1001–1018, 1983
43. Hosobuchi Y: Subcortical electrical stimulation for control of
intractable pain in humans. Report of 122 cases (1970–1984). J
Neurosurg 64:543–553, 1986
44. Hunter J, Jasper HH: Effects of thalamic stimulation in un-
anaesthetised animals. The arrest reaction and petit mal-like
seizures, activation patterns and generalized convulsions. Elec-
troencephalogr Clin Neurophysiol 1:305–324, 1949
45. Ilinsky IA, Kultas-Ilinsky K: Sagittal cytoarchitectonic maps of
the Macaca mulatta thalamus with a revised nomenclature of
the motor-related nuclei validated by observations on their con-
nectivity. J Comp Neurol 262:331–364, 1987
46. Jasper HH: Recording from microelectrode in stereotactic sur-
gery for Parkinson’s disease. J Neurosurg 24:219–221, 1966
47. Jasper HH: Unspecific thalamocortical relations, in American
Physiological Society (eds): Handbook of Physiology. Sect
1. Neurophysiology. Washington DC: J Field, 1960, pp
1307–1321
48. Joffroy AJ, Lamarre Y: Rhythmic unit firing in the precentral
cortex in relation to postural tremor in a deafferented limb.
Brain Res 27:386–389, 1971
49. Jones EG: The Thalamus. New York: Plenum Press, 1985, p
935
50. Kelly PJ, Derôme P, Guiot G: Thalamic spatial variability and
the surgical results of lesions placed with neurophysiologic
control. Surg Neurol 9:307–315, 1978
51. Laitinen LV: Brain targets in surgery for Parkinson’s disease.
Results of a survey of neurosurgeons. J Neurosurg 62:
349–351, 1985
52. Lamarre Y, Mercier LA: Neurophysiological studies of harma-
line-induced tremor in the cat. J Physiol Pharmacol 49:
1049–1058, 1971
53. Larson SJ, Sances A Jr, Wetzel N: Cerebello-pallido-thalamic
connections in man. Appl Neurophysiol 45:549–562, 1982
54. Lenz FA, Tasker RR, Kwan HC, et al: Single unit analysis of
the human ventral thalamic nuclear group: correlation of tha-
lamic “tremor cells” with the 3-6 Hz component of the parkin-
sonian tremor. J Neurosci 8:754–764, 1988
55. Marks PV, Wild AM, Gleave JRW: Long-term abolition of
parkinsonian tremor following attempted ventriculography. Br
J Neurosurg 5:505–508, 1991
56. Matsumoto K, Asano T, Baba T, et al: Long-term follow-up
results of bilateral thalamotomy for parkinsonism. Appl Neu-
rophysiol 39:257–260, 1976/77
57. Matsumoto K, Shichijo F, Fukami T: Long-term follow-up
review of cases of Parkinson’s disease after unilateral or bilat-
eral thalamotomy. J Neurosurg 60:1033–1044, 1984
58. Mazars G, Merienne L, Cioloca C: Control of dyskinesias due
to sensory deafferentation by means of thalamic stimulation.
Acta Neurochir Suppl 30:239–243, 1980
59. Meyers R: Surgical interruption of the pallidofugal fibers. Its
effect on the syndrome of paralysis agitans and technical con-
siderations in its applications. NY State J Med 42:317–325,
1942
60. Monnier M, Kalberer M, Krupp P: Functional antagonism
between diffuse reticular and intralaminary recruiting projec-
tions in the medial thalamus. Exp Neurol 2:271–289, 1960
61. Mundinger F, Neumüller H: Programmed stimulation for con-
trol of chronic pain and motor diseases. Appl Neurophysiol
45:102–111, 1982
213

62. Nagaseki Y, Shibazaki T, Hirai T, et al: Long term follow-
up results of selective VIM-thalamotomy. J Neurosurg 65:
296–302, 1986
63. Narabayashi H: Stereotaxic Vim thalamotomy for treatment of
tremor. Eur Neurol 29 (Suppl 1):29–32, 1989
64. Nashold BS, Slaughter DG: Some observations on tremors, in
Gillingham FJ (ed): Third Symposium on Parkinson’s Dis-
ease, 1968. Livingstone Ltd, 1969, pp 241–246 [Unverified]
65. Ohye C: Rôle des noyaux thalamiques dans l’hypertonie et le
tremblement de la maladie de Parkinson. Rev Neurol 142:
362–367, 1986
66. Ohye C, Bouchard R, Larochelle L, et al: Effect of dorsal rhi-
zotomy on postural tremor in the monkey. Exp Brain Res
10:140–150, 1970
67. Ohye C, Fukamashi A, Narabayashi H: Spontaneous and
evoked activity of sensory neurons and their organization in the
human thalamus. Z Neurol 203:219–234, 1972
68. Ohye C, Hirai T, Miyazaki M, et al: Vim thalamotomy for the
treatment of various kinds of tremor. Appl Neurophysiol 45:
275–280, 1982
69. Ohye C, Imai S, Nakajima H, et al: Experimental study of spon-
taneous postural tremor induced by a more successful tremor-
producing procedure in the monkey. Adv Neurol 24:83–91,
1979
70. Ohye C, Maeda T, Narabayashi H: Physiologically defined Vim
nucleus. Its special reference to control of tremor. Appl Neu-
rophysiol 39:285–295, 1977
71. Ohye C, Nakamura R, Fukamachi A, et al: Recording and stim-
ulation of the ventralis intermedius nucleus of the human thal-
amus. Confin Neurol 37:258, 1975
72. Ohye C, Narabayashi H: Physiological study of presumed ven-
tralis intermedius neurons in the human thalamus. J Neurosurg
50:290–297, 1979
73. Ohye C, Saito Y, Fukamashi A, et al: An analysis of sponta-
neous rhythmic and non-rhythmic burst discharges in the hu-
man thalamus. J Neurol Sci 22:245–259, 1974
74. Ohye C, Shibazaki T, Hirai T, et al: Further physiological obser-
vations on the ventralis intermedius neurons in the human thal-
amus. J Neurophysiol 61:488–500, 1989
75. Ohye C, Shibazaki T, Hirai T, et al: A special role of the par-
vocellular red nucleus in lesion-induced spontaneous tremor in
monkeys. Behav Brain Res 28:241–243, 1988
76. Ohye C, Shibazaki T, Nagaseki Y, et al: Possible descending
pathways mediating spontaneous tremor in monkeys. Adv
Neurol 40:181–188, 1984
77. Pollak P, Benabid AL, Gervason CL, et al: Long-term effects of
chronic stimulation of the ventral intermediate thalamic nucle-
us in different types of tremor. Adv Neurol 60:408–413, 1993
78. Raeva S: Localization in human thalamus of units triggered
during “verbal commands,” voluntary movements and tremor.
Electroencephalogr Clin Neurophysiol 63:160–173, 1986
79. Sanderson P, Mavoungou R, Albe-Fessard D: Changes in sub-
stantia nigra pars reticulata activity following lesions of the
substantia nigra pars compacta. Neurosci Lett 67:25–30, 1986
80. Schaltenbrandt G, Spuler H, Wahren W, et al: Electroanatomy
of the thalamic ventro-oral nucleus based on stereotaxic stimu-
lation in man. Z Neurol 199:259–276, 1971
81. Schnider SM, Kwong RH, Kwan HC, et al: Detection of feed-
back in the central nervous system of parkinsonian patients.
IEEE WA 11:291–294, 1986
82. Sem-Jacobsen CW: Depth-electrographic observations related
to Parkinson’s disease. Recording and electrical stimulation in
the area around the third ventricle. J Neurosurg 24:388–402,
1966
83. Siegfried J: Chronic electrical stimulation of the VL-VPL com-
plex and of the pallidum in the treatment of extrapyramidal and
214
A. L. Benabid, et al.
cerebellar disorders: personal experience since 1985, in XIIth
Meeting of the World Society for Stereotactic and Functional
Neurosurgery. Abstract 125, Ixtapa, Mexico, 1993 [Unverified]
84. Siegfried J: Effets de la stimulation du noyau sensitif du thala-
mus sur les dyskinésies et la spasticité. Rev Neurol 142:
380–383, 1986
85. Siegfried J: Neurosurgical treatment of Parkinson’s disease.
Present indications and value, in Rinne UK, Klingler M, Stamm
G (eds): Parkinson’s Disease. Current Progress, Problems
and Management. Amsterdam: Elsevier, 1979, pp 369–376
86. Speelman JD: Parkinson’s Disease and Stereotaxic Neu-
rosurgery. Thesis. Amsterdam: Rodopi, 1991, p 218
87. Stellar S, Cooper IS: Mortality and morbidity in cryothalamec-
tomy for parkinsonism. A statistical study of 2868 consecutive
operations. J Neurosurg 28:459–467, 1968
88. Talairach J, David M, Tournoux P, et al: Atlas d’Anatomie
Stéréotaxique des Noyaux Gris Centraux. Paris: Masson,
1957, p 294 [Unverified]
89. Talairach J, de Ajuriaguerra J, David M: A propos des coagula-
tions thérapeutiques sous-corticales: étude topographique du
système ventriculaire en fonction des noyaux gris centraux.
Presse Med 58:697–701, 1950
90. Talairach J, Hecaen H, David M, et al: Recherches sur la coag-
ulation thérapeutique des structures sous-corticales chez
l’homme. Rev Neurol 81:4–24, 1949
91. Taren J, Guiot G, Derome P, et al: Hazards of stereotaxic thala-
mectomy. Added safety factor in corroborating X-ray target
localization with neurophysiological methods. J Neurosurg
29:173–182, 1968
92. Tasker R: Effets sensitifs et moteurs de la stimulation thala-
mique chez l’homme Applications cliniques. Rev Neurol 142:
316–326, 1986
93. Tasker RR, Organ LW, Hawrylyshyn P: Investigation of the
surgical target for alleviation of involuntary movement disor-
ders. Appl Neurophysiol 45:261–274, 1982
94. Tasker RR, Siquiera J, Hawrylyshyn P, et al: What happened to
Vim thalamotomy for Parkinson’s disease? Appl Neurophys-
iol 46:68–83, 1983
95. Toth S, Solyom A, Vajda J, et al: The rhythmic properties of the
motor system. Stereotact Funct Neurosurg 53:95–104, 1989
96. Van Buren JM, Borke RC, Modesti LM: Sensory and nonsen-
sory portions of the nucleus “ventralis posterior” thalami of the
chimpanzee and man. J Neurosurg 45:37–48, 1976
97. Velasco F, Velasco M, Ogarrio C, et al: Electrical stimulation
of the centromedian thalamic nucleus in the treatment of con-
vulsive seizures : a preliminary report. Epilepsia 28:421–430,
1987
Manuscript received March 3, 1995.
Accepted in final form July 26, 1995.
This study was supported by INSERM, CNAMTS, Grenoble
University, Région Rhone-Alpes, France Parkinson, Fondation pour
la Recherche Medicale and Joseph Fourier University. Dr. Gao
received a fellowship from Jinzhou University and later Grenoble
University.
This work was presented in part at the 60th annual meeting of the
American Association of Neurological Surgeons in San Francisco,
California, April 11–16, 1992.
Address reprint requests to: Alim Louis Benabid, M.D., Depart-
ment of Clinical and Biological Neurosciences, INSERM Pre-
clinical Neurobiology U-318, Joseph Fourier University of Gren-
oble, Hôpital A. Michallon, Pavilion B, BP 217, 38043 Grenoble,
France.
J. Neurosurg. / Volume 84 / February, 1996