The New England Journal of Medicine -- February 17, 2000 -- Vol. 342, No. 7

New Treatment Options for Tremors

The revival and refinement of stereotactic neurosurgery represent the most
important therapeutic advance in the treatment of movement disorders in the
past 30 years. The majority of patients who undergo surgery have Parkinson's
disease, but this approach is also being used for patients with movement
disorders such as dystonia and tremor, which is sometimes so severe that it
cannot be adequately treated with medication. The study by Schuurman et al.
(1) in this issue of the Journal offers new information on neurosurgical
treatment in such patients.

Tremor is the most common movement disorder. The majority of affected
patients have Parkinson's disease (prevalence, 110 per 100,000) or essential
tremor (prevalence, 306 to 417 per 100,000). The remaining types of tremor,
including various idiopathic forms and that caused by multiple sclerosis or
brain trauma, are less frequent but often lead to severe disability.

Tremor is not a single entity. It is classified according to the underlying
cause, such as Parkinson's disease, or to the presence of syndromes that
have various causes but similar symptoms at presentation, such as cerebellar
tremor. (2) Of the more than 10 types of tremor, the most incapacitating are
essential tremor, tremor due to Parkinson's disease, cerebellar tremor, and
tremor due to neuropathy. The effect on the patient depends on the clinical
manifestations. The more goal-directed movements are distorted by tremor,
the more severe the resulting difficulty in performing daily activities.
Rest tremor can be intolerable because it can severely restrict social
interactions and it is physically exhausting. In patients with Parkinson's
disease, tremor occurs mainly at rest. In patients with essential tremor,
tremor often occurs during goal-directed movements, and in patients with
cerebellar tremor, this is always the case. Tremor in patients with multiple
sclerosis is usually cerebellar in nature but is accompanied by other
cerebellar, sensory, and corticospinal impairments. The study by Schuurman
et al. includes patients with tremor due to Parkinson's disease, essential
tremor, or multiple sclerosis.

The first-line treatment for tremor is oral medication. (3) The average
patient with tremor due to Parkinson's disease can be treated with dopamine
agonists, levodopa, anticholinergic agents, or budipine, and if all other
types of medication have failed, clozapine is often effective. This kind of
treatment generally results in a symptomatic improvement of more than 50
percent. Some patients, however, remain disabled by their tremor. Patients
with essential tremor are treated with beta-blockers (mainly a nonselective
blocker such as propranolol or a (beta)2-selective blocker), primidone, or
both, and 40 to 70 percent of these patients have symptomatic improvement of
50 to 70 percent. (4) There is no generally accepted medical treatment for
cerebellar tremor. Treatment with clonazepam is sometimes successful, as is
treatment with levodopa and anticholinergic agents or clozapine when a
clinically significant rest tremor is present. Unfortunately, for the most
severely affected patients, the degree of improvement afforded by
pharmacotherapy is insufficient. Therefore, new types of treatment are
necessary, and the most promising is neurosurgery. The minimal criteria for
a patient to be considered a candidate for neurosurgery are a lack of
response to pharmacotherapy, severe disability resulting from the tremor,
and the absence of contraindications to neurosurgery.

The use of neurosurgery dates back to the 1950s. In the 1960s the preferred
neuroanatomical target was the nucleus ventralis intermedius thalami, in
which a lesion was created with thermocoagulation. The creation of a lesion
as small as 40 to 60 mm3 can suppress the tremor. The critical causative
role of the nucleus ventralis intermedius thalami is not yet fully
understood, but it is assumed that tremor is generated within loops between
deep cerebral or cerebellar nuclei and cortical areas and that most of these
loops pass through the ventrolateral thalamic nuclei. Initially, most of the
patients who underwent surgery had Parkinson's disease, and the treatment of
nonparkinsonian tremors was a secondary indication at many centers.

The quality of the assessments of early efforts can hardly be compared with
that of the present day, (5) but long-term studies suggest that the
improvement in tremor was clinically significant. Speech problems and
neuropsychological side effects were frequent, especially in patients who
underwent bilateral operations. (6) Therefore, bilateral thalamotomy was
rarely performed. These efforts ended abruptly with the introduction of
levodopa in the early 1970s. Levodopa helped patients with Parkinson's
disease, but not those with essential tremor or cerebellar tremor. Only a
few neurosurgical centers continued to operate on small numbers of patients,
mainly those with nonparkinsonian tremors.

The renaissance of neurosurgery in the past 10 years has occurred for
several reasons. First, in 1987 Benabid and colleagues described the use of
deep-brain stimulation (7) at the same location that was earlier targeted
for thermocoagulation. These researchers implanted small electrodes into the
nucleus ventralis intermedius thalami and connected them to a subcutaneously
implanted stimulator. Stimulation at frequencies of more than 100 Hz is
assumed to work by inhibiting the function of the stimulated area. This
treatment promised similar benefits but fewer side effects than
thermocoagulation. The electrical settings could be adapted for each
patient, and even bilateral operations became relatively safe.

Second, the reevaluation of pallidotomy as a treatment for Parkinson's
disease confirmed that this approach improved akinesia and decreased
levodopa-induced dyskinesia. (8) Third, the advent of new imaging and
microelectrode recording techniques increased the safety and precision of
the operations. Fourth, pathophysiologic studies in animal models of
Parkinson's disease showed that the thalamus is overinhibited and both the
internal pallidum and the subthalamic nucleus are overactive. Lesions of the
subthalamic nucleus decreased the parkinsonian symptoms in monkeys, (9)
prompting neurosurgeons to target this area for stimulation by electrodes.
Meanwhile, the subthalamic nucleus, internal pallidum, and the nucleus
ventralis intermedius thalami all became the subjects of ongoing studies of
thermocoagulation and stimulation therapy. Thus, neurosurgery for
Parkinson's disease suddenly became logical, and the reasons for its success
became apparent. Since the pathophysiology of nonparkinsonian tremors is
still unknown, the use of such surgery for nonparkinsonian tremors is purely
empirical, with the thalamus the only target.

The effect of deep-brain stimulation on tremor in patients with Parkinson's
disease or essential tremor has been well documented in two controlled,
prospective studies. (10,11) In smaller studies, patients with tremor due to
multiple sclerosis have also had a favorable response to deep-brain
stimulation. The study by Schuurman et al., however, is the first
randomized, prospective trial that compares the effects of thalamotomy on
tremor with those of stimulation, and it has solved two key questions. The
degree of improvement in tremor was similar after deep-brain stimulation and
thalamotomy, but overall function was better after stimulation. Schuurman et
al. have convincingly demonstrated that perioperative morbidity is lower
with deep-brain stimulation than with thalamotomy, thereby confirming
uncontrolled data from centers with experience in the use of both
techniques. (12)

In the study by Schuurman et al., both approaches improved tremor in
patients with multiple sclerosis but did not result in functional
improvement. This result is not unexpected, since these patients also have
other complications of their disease. Furthermore, this result confirms
those of earlier studies that reported functional improvement in only about
a third of patients with cerebellar tremor who were treated by the creation
of lesions in the nucleus ventralis intermedius thalami. (13) A rough
estimate of the percentage of patients with long-term functional improvement
after surgery on the nucleus ventralis intermedius thalami is 85 percent for
patients with Parkinson's disease, 50 percent for those with essential
tremor, and 30 percent for those with multiple sclerosis, (14) although the
percentages in whom the severity of tremor is reduced are much higher.

Do the results of Schuurman et al. signal the end of thalamotomy? They offer
a strong argument for the use of deep-brain stimulation as the treatment of
choice for eligible patients. But patients so treated must be followed
closely at specialized centers, and the costs of therapy are much higher
than those of thalamotomy. Studies of the cost effectiveness of thalamic
stimulation are needed. The results of Schuurman et al. cannot be
uncritically extrapolated to approaches involving other targets, such as the
internal pallidum and the subthalamic nucleus.

Progress in the field of stereotactic surgery is very rapid. While Schuurman
et al. were conducting their study, the main target of stimulation therapy
in patients with Parkinson's disease shifted to the subthalamic nucleus,
because stimulation of this nucleus improves not only tremor but also
akinesia by about 70 percent. (15) There is no evidence to date that
stimulation of this nucleus will also improve nonparkinsonian tremors, but
other targets may be discovered for them, too. The factor limiting
therapeutic progress in patients with nonparkinsonian tremors is the lack of
insight into the underlying mechanisms of tremor. We need to understand the
abnormally functioning motor loops within the central nervous system if we
are to find better ways to inhibit them. It is unlikely that the same
mechanisms underlie all nonparkinsonian tremors, and thus thermocoagulation
or stimulation of different nuclei may be necessary. Further basic research
and clinical work are necessary to find better ways to help patients with
disabling tremor.

Gunther Deuschl, M.D.
Christian Albrechts Universitat
24105 Kiel, Germany

Copyright © 2000 by the Massachusetts Medical Society. All rights reserved.