Action tremor of the arms can be an invalidizing symptom of diseases such as Essential Tremor, Dystonic Tremor, Parkinson's disease and Multiple Sclerosis. In this study we compare the efficacy and safety of two different brain targets for deep brain stimulation (DBS) that both are known to reduce action tremor of the arms. These two targets are called the VIM nucleus of the thalamus (VIM) and the posterior subthalamic area (PSA), which includes the Zona Incerta. Both targets can be reached by one lead (wirh four electrode contact).

Patients that are found eligible for DBS because of severe action tremor of the arms are invited to participate. After randomization, half of them are stimulated first in the VIM for 3 months and then in the PSA for 3 months, and the other half first in the PSA and then VIM for 3 months each. Tremor severity is scored on a clinical quantitative scale at baseline and at the end of each of these two 3-month periods, and eventual side-effects are registered. The best target is then selected and after another 6 months scoring is repeated. We intend to provide robust data about whether one of the two targets is superior to the other both regarding ability to reduce tremor efficiently and to avoid or minimize side-effects, or if there is no significant difference between the two targets. We also carefully check the exact position of the active electrode contact in the brain and compare this with efficacy and safety evaluations. Long-term follow-up is planned after 3, 5, 7 and 10 years.

Tremor is defined as rhythmic and oscillating involuntary movements of a body part. Action tremor occurs when a body part is activated, and is clinically most frequently seen in Essential (idiopathic) Tremor (ET), with an overall prevalence of 0.9 %, and of 4.6 % in people over 65 years. But it can also be the most disabling symptom in many other neurological diseases such as dystonia, multiple sclerosis (MS), advanced Parkinson's disease (PD), and some spino-cerebellar ataxias/hereditary disorders with dysfunction of cerebellar connections. The cerebello-thalamo-cortical pathway is probably involved in all tremors. The exact mechanisms that disrupt these pathways and produce tremor remain, however, largely unknown. The cerebello-thalamic tract originates from the cerebellar nuclei (dentate, emboliform and globose), ascends and crosses in the pons and enters the red nucleus (RN) in the mesencephalon. Most fibers project further to the ventral thalamic nuclei - oralis (VO) and intermedius (VIM). The posterior subthalamic area (PSA) is the region inferior to VO and VIM, dorsal and posterior to the subthalmic nucleus (STN) , and lateral to RN. Its principal components are the nucleus called Zona Incerta (ZI) and the fiber bundle called the prelemniscal radiation (RaPrl). Whereas STN and RN are visible on standard T2-weighted MRI-scans, PSA is only indirectly defined by these surrounding structures. Classification and differential diagnosis of action tremor disorders can be difficult, since with rare exceptions there are no confirmative diagnostic tests or biomarkers for the primary tremor forms. Differential diagnosis of primary action tremor forms such as ET, Dystonic tremor (DT) and Cerebellar tremor (CbT), therefore relies mainly on clinical examination and the exclusion of underlying diseases causing secondary action tremors (such as PD and MS). Action tremor is divided into postural (when a position is maintained), or kinetic (when a movement is made). Other neurological findings (dystonia, cerebellar, pyramidal, parkinsonian, neuropathic signs), and systemic signs must be assessed. Supplementary diagnostic workup includes MRI in all cases to show/exclude tremor-causing lesions such as infarcts, MS plaques or other inflammatory diseases, cerebellar atrophy, or atrophy or pathological signal changes along the cerebello-thalamo-cortical circuits, including the basal ganglia. DAT-scan (131I-FP-CIT SPECT) of the DopAmine Transporter (DAT) shows reduced binding in the striatum in PD and other degenerative parkinsonian disorders, but not in ET, DT or CbT. It is performed in all patients who have rest tremor or other parkinsonian signs. Necessary laboratory tests of blood (+/- CSF and urine) are also performed. Medical treatment available for primary (and most secondary) action tremor disorders are only symptomatic. The two most commonly used drugs for ET, the non-selective beta-blocker propranolol and the anticonvulsant primidone both received a level A recommendation of efficacy by the American Academy of Neurology in 2005. This was based on mostly small and short-term studies (3-6 weeks), showing an average tremor reduction of about 50-60% in about 50% of patients. No controlled long-term studies have been done, but open-label studies indicate development of tolerance and reduced effect in 10-15%. Level B evidence of a probable effect has been given to other beta-blockers (atenolol, sotalol) and the antiepileptics alprazolam, topiramate and gabapentin. There have been no systematic studies for other tremors like dystonic or cerebellar tremors. Thus, for these patients no medications with established efficacy exist, except intramuscular botulinum toxin injections for dystonic head tremor.

Deep brain stimulation (DBS) of the Nucleus Ventralis Intermedius of the thalamus (VIM) is still considered the target of choice for patients with disabling action tremor at most centers. Open-label, mostly short-term studies of VIM-DBS in ET have reported up to 80-90% reduction of action tremor, but long-term clinical experience and studies show that gradual loss of tremor control over months and years occurs in many patients. Masked raters were not used, except for one study. The problem termed "development of tolerance" to stimulation, is emphasized in recent reviews and clearly limits the functional long-term benefit of VIM-DBS. VIM-DBS also tends to be less effective on proximal tremor, which is often the most severe symptom. Because VIM cannot be clearly outlined on MRI or mapped using microelectrode recordings, targeting it employs the use of standard stereotaxic coordinates.The definition of these varies between centers. They are defined in relation to the inter-commissural line (ICL). Final placement of the electrode is decided after peroperative clinical testing of effect on tremor and side effects in the awake patient. Bilateral VIM-stimulation is associated with a relatively high rate of side effects. In a meta-analysis of 430 patients, 19 % had paresthesias, 9 % dysarthria, 7 % headache and 6 % unsteadiness or gait difficulties. Therefore, many centers offer only unilateral stimulation to most patients even if they have bilateral tremor. The limitations of VIM-stimulation have led to a renewed interest in exploring the PSA as an alternative. Open label case series from several centers indicate better results from PSA-DBS than VIM-DBS, particularly in reducing kinetic tremor, including proximal arm tremor and intention tremor. The targeted area within the PSA varies among centers. Plaha et al reported results from blinded 1-year follow-up of bilateral stimulation of the caudal ZI in 5 patients with PD tremor, and 13 patients with action tremors from a range of diseases (ET, MS, CbT, DT, Holmes tremor). Average improvement of postural/action tremor was 88% and of rest tremor 95%. In the MS-patients, both severe proximal arm tremor, truncal ataxia and severe head-and-neck tremor improved. There was no significant decline of clinical efficacy or significant change in stimulation parameters during serial follow-ups, but some side-effects were seen in a few patients due to misplaced electrodes or edema. Blomstedt et al reported results of PSA-DBS in 21 patients with ET (most unilateral), achieving on average 95% tremor reduction and 87% functional improvement in the contralateral arm. Transient, mild expressive dysphasia was observed in eight patients. In summary, the most frequent side-effects reported in PSA-DBS series are paresthesias, dysarthria, dysequilibrium and blurred vision, and thus similar to those observed with VIM-DBS, but are consistently reported to be mild and transient in almost all patients. So far no controlled comparison of the efficacy and safety between stimulation of the two targets has been reported.With respect to the ICL, the targets for PSA used in the above studies were 10-14 mm lateral, 6-7.5 mm posterior to the midpoint of the ICL, and 2-4 mm inferior. Few authors have studied different targets in the same patients. Herzog et al compared stimulation with the upper electrode contacts the thalamic level, and with the lower contacts 1.5mm below the ICL in 21 patients with ET or MS-tremor. The results favoured the contacts below the ICL.

To overcome the limitations of today´s standard method of treating severe action tremors with VIM-DBS, we are performing a double-blind, randomized, controlled and comparative study of VIM-DBS versus PSA-DBS in patients with disabling action tremor. We will implant one quadripolar electrode (each contact 1.5 mm, interspace 0.5 mm) so that it covers both targets on one side, unilaterally or bilaterally as indicated clinically. The efficacy of stimulating each target will be studied in individual patients and between patient groups in a cross-over design, in two consecutive randomized 3-month periods, by the use of validated tremor scores. Further continuous stimulation of the most effective target in each patient will be evaluated after another 6 months. This will provide unbiased efficacy-data about the two targets, both within each patient, and between patient groups. Stimulation efficacy will also be evaluated in relation to the exact target localization, defined by merging post-op helical CT scan (which shows the four electrode contacts), with pre-op MRI, including Diffusion Tensor Imaging/fiber tracking. Such high-resolution correlation data has to date only been published in a few action tremor patients world-wide, and will add important new data to aid DBS targeting in the future. Since little is known about whether cognitive functions are altered by the disease or the treatment, we will include neuropsychological testing before and after effective DBS treatment.

Methods: i) Clinical evaluation and scoring. Before the patients are invited to participate, they have been thoroughly evaluated for correct diagnosis, including a complete clinical neurological examination, MRI of the head/brain (+/- spine/spinal cord), laboratory tests, etc. The inclusion examination ensures that inclusion criteria are fulfilled and that none of the exclusion criteria are met. The study examinations include clinical examination and scoring performed at baseline -end of 1. Random.per./3 mo. - end of 2. Random.per./6 mo.- 1 year

Performed at all study visits:

• The Fahn-Tolosa-Marin tremor scale contains 21 items scored 0-4, with 0 indicating no tremor/disability and 4 severe tremor (constant, high amplitude)/inability to perform a task due to tremor. Item 1-9 score tremor for face, tongue, voice, head, trunk and the four extremeties. For face only rest tremor is scored, for tongue, head and trunk rest and postural tremor, and for the extremities action/intention tremor is also scored. Items 10-13 test handwriting, and spiral drawing with right and left hand, and line drawing between defined points. Item 14 tests pouring of water from one plastic cup to the other, and Item 15-21 describe functional disability of speaking, feeding (not liquids), bringing liquids to mouth, hygiene, dressing, writing, working.
• Quality of life in Essential Tremor- scale (QUEST). A 30 item questionnaire developed specifically for tremor patients, evaluating quality of life, cognitive, and emotional aspects of the tremor.
• VAS of Global Disease Burden (patient evaluated). Visual analogue 0-10 centimeter continuous scale, on which the patient scores at which severity they assess the global tremor disease burden.
• The Hospital Anxiety and Depression Scale. A well-established and validated self-report questionnaire which contains 14 items expressing emotions, alternating between 7 emotions of anxiety and 7 of depression. Scores for each item ranging from 0-3. Cut-off sum-scores exist for clinically significant symptoms of anxiety and depression, respectively.

In addition,complete neurological exam. and neuropsychological testing is performed at baseline and at 1 year of DBS.

ii) Operation. Preoperative axial MRI sequences (T2-weighted fast-spin-echo, diffusion-weighted spin-echo echo-planar imaging, and 3D inversion prepared T1-weighted gradient-echo) will be obtained on a 3 Tesla MRI-scanner the day before the operation. In local anaesthesia a CRW™ stereotactic frame, (Radionics, MA, USA) is mounted, placed parallell to the AC-PC line, before performing a stereotactic 3D CT scan. The MRI and CT scans are merged using the iPlan™ (version 3.0 or later) computer-aided neuronavigation system (BrainLAB, München, Germany), which is also used to plan the targets based on predefined stereotactic coordinates for VIM and PSA, refined according to relationship to the MRI-visible STN and RN. Trajectory angle will be chosen to cover VIM and to enter PSA slightly posterior and medial to STN, at the level of maximal diameter of RN. To ensure safety, trajectories are always planned to avoid vessels, sulci and ventricles. Microelectrode recordings (MER) will not be performed.

Clinical test stimulation will be performed in the awake patient, at four 2 mm steps, from 8 mm above the most caudal target in the PSA (Caudal Zona Incerta) using the permanent quadripolar electrode, and the placement accepted if good tremor suppression is confirmed and no unacceptable side effects observed. Electrode position is checked using perioperative radiography, and the electrodes fixed to the scull. After removal of the stereotactic frame, the electrodes are connected to a pulse generator implanted in the subclavicular or abdominal region.

iii) Randomization periods/stimulation treatment. Half the patients are randomized to receive VIM-stimulation from 0-3 months after surgery and then switched to receive PSA-stimulation for the next 3 months, and the other half to receive PSA-stimulation the first three months and VIM-stimulation the next 3 months. During these two randomization periods the pulse width will be held constant at 60 microseconds and frequency 130 Hz, monopolar mode (bipolar mode only if monopolar leads to intolerable side effects). Current should not exceed 3 mA during the randomization periods. After scoring at the end of the last 3-month randomization period, the electrode contact yielding the best tremor reduction with no or only minor/possible side effects is chosen for further treatment. During follow-up to 1-yr post-op there may be 2 additional visits for optimal adjustment of stimulation parameters. Randomization is performed through a web-based program provided by Faculty of Medicine, unit for Applied Clinical Research at the Norwegian University of Science and Technology, Trondheim.

iv) Neuropsychological examinations. A standard set of well-known and validated neuropsychological tests will be performed pre-op and 1-yr post-op.

Inclusion Criteria:

• Min 5-yr duration chronic action tremor of upper limb(s), w/ or w/out tremor of head/neck, trunk, lower limbs, who have insufficient relief from adequate trials of recommended medications. The arm tremor must be so severe that it interferes with the patient's work performance or activities of daily living, such as drinking/eating, dressing/hygiene, and/or writing. Clinical diagnosis (according to criteria defined by the Consensus Statement of the Movement Disorders Society) include Essential Tremor, Dystonic Tremor, Cerebellar Tremor- idiopathic or secondary (to e.g.MS or SCA), or severe PD tremor (who are not eligible for DBS of the subthalamic nucleus).

Exclusion Criteria:

• Brain MRI showing so marked general atrophy, or supra-tentorial white matter changes, that the safety of the procedure is affected. Co-morbidity of dementia or other severe neuropsychiatric disorder (major depression or anxiety disorder, active/recent psychosis, drug or other substance abuse). Increased risk of bleeding, cancer at advanced/not stabilized stage, other severe co-morbidity (e.g. With chronic pain, short life expectancy,). Other surgical contraindications.

Both the doctor who assesses the primary outcome measures at 3 and 6 months and the participant are unaware of which of the two targets is stimulated in the two randomization periods.