The Swedish Essential Tremor (SWEET) Trial

PURPOSE AND AIMS

• Aim 1 - To evaluate the effect of deep brain stimulation (DBS) vs best medical treatment in essential tremor (ET) in a randomized, single-blinded controlled trial.
• Aim 2 - To compare the effect of DBS in the established target in the ventrolateral (VL)-thalamus/ nucleus ventralis intermedius thalami (Vim) and in a new target in the posterior subthalamic area (PSA)/caudal Zona incerta (cZi).
• Aim 3 - To map the target area in the VL-thalamus and PSA concerning effects and side effects of stimulation in order to identify and delineate the optimal target.
• Aim 4 - To evaluate the long-term effects of DBS for ET in a longitudinal non-randomized evaluation.

SURVEY OF THE FIELD Deep Brain Stimulation (DBS) In DBS, thin quadripolar electrode leads (with four 1.5 mm long and 1.27 mm thick contacts, separated by 1.5 mm space), connected to a neuropacemaker are implanted with stereotactic technique into the central parts of the brain where the neuronal activity is modulated with electrical current. DBS has revolutionized the treatment of Parkinson´s disease and other movement disorders, and more than 150.000 patients have so far been operated. Currently, new indications and targets are emerging.

However, many of the brain targets subjected for DBS are not well defined, and one of the main problem with DBS is likely to be the high number of suboptimal placed electrodes with lack of effect, unacceptable side effects and costly revisions. Further, even though this is an invasive and highly resource demanding therapy, most clinical indications for DBS (asides from Parkinson´s disease and dystonia) are not evidence based. They are considered as "established treatments" and often provided as clinical treatment and not within the context of trials or scientific studies.

DBS for Essential Tremor ET is the most common adult movement disorder with a prevalence of about 5% in the population above 65 years of age. Of patients who seek medical care up to 50% do not respond adequately to drug therapy.

The thalamic nucleus ventralis intermedius (Vim) is the projection structure of the cerebello-thalamic fibers mediating tremor. Hence, Vim has been an "established" target for surgical treatment of ET during the thalamotomy era, and when DBS was introduced and replaced thalamotomy, the Vim was logically the target for DBS for tremor. However, data from recent years indicate that it might not be stimulation of the Vim itself, but rather, the stimulation was affecting the pathologic tremor oscillations mediated by the cerebello-thalamic-cortical projections, located in the PSA and the caudal zona incerta (cZi). Thus, a number of studies have demonstrated that electrode contacts that reach the PSA have a better effect on tremor than those located more dorsally-rostrally in the Vim. This subthalamic area, which is situated immediately ventral-caudal to the Vim, is particularly dense with cerebello-thalamic axons, dispersing into the Vim. Stimulation is thus likely to involve more axons in the PSA than in the Vim, since these axons in the PSA constitute so to speak a bottle neck of the axonal traffic mediating tremor. It is also plausible that stimulation of axons rather than stimulation of nuclei actually affects by antidromic and orthodromic propagation more neurons and therefore also alters tremor oscillations more efficiently. A number of studies directly targeting the cZi have also recently presented results comparing favorably with the results published for Vim DBS.

Level of evidence regarding DBS in ET:

Neither the Vim nor the cZi are evidence based targets, and there are no evidence-based DBS procedures for ET (level IV evidence only), even though this is the second most common indication for DBS. However, the Vim is by tradition the established target for ET and the procedure is recommended in the Swedish guidelines for treatment of tremor.

The results of Vim DBS ET have been demonstrated in a number of uncontrolled, mostly minor studies. A problem is that many of these studies are likely to contain electrodes placed blindly within the PSA36 and that the actual location of the active electrode contacts of the DBS lead is seldom taken into consideration. Further, many reports lack a preoperative baseline and the results are often reported in a heterogeneous manner, making comparison between different studies difficult. Regarding cZi DBS, excluding our own studies, the results of cZi DBS in 33 patients have been presented in five small and non-randomized studies. Concerning the relative merits of the two targets, we have previously presented a tremor reduction of 86 % following cZi, compared to only 60% in the Vim4, and demonstrated that the electrode contacts with the best effect in what was called Vim were often placed in the PSA. There is further a scarcity of long-term data regarding both targets.

Identifying the brain target in the individual patient In a recent study of 28 000 procedures, electrode revisions constituted up to one third of all DBS procedures. It is likely that misplaced electrodes due to difficulties in identifying the target is one of the major causes explaining a poor result in DBS for movement disorders.

There are a number of different ways in which the target can be identified. Even for brain targets that can be visually identified on MRI, some centers rely mostly on statistical coordinates of the targets based on anatomical atlases, but this may result in poorly placed electrodes, due to significant inter-individual differences in the anatomy. However, concerning the Vim, identification of that target in relation to ventricular landmarks is a necessity since the Vim cannot be visualized as such in MRI.

Currently, 60 % of the centers are relying on micro-electrode recording (MER) in order to identify the target in the individual patient. This is an invasive neurophysiological method necessitating considerable resources. It is also associated with an increased risk of hemorrhages, since it involves introducing several sharp cannulas and electrodes into the target area. The third method is visual anatomical targeting, whereby the target is identified on the patients´ MRI, either directly or in relation to visible very closely related landmarks (Figure 2). We have over the years made significant contributions in the development of this last method, both concerning the pallidum, STN and cZi. Unfortunately this method demands a high level of dedication and relies to a high degree on personal experience. Scientifically based guidelines delineating the target areas and identifying the optimal target do not readily exist.

Common for all different methods is that it is normally required that the patient is awake during surgery in order to allow for intra-operative testing of effect/side effects, even if some today perform the procedure in general anesthesia concerning some of the targets. If the brain targets could be identified with a high degree of certainty, and the electrode position within that target can be verified during surgery, then having the patients awake during surgery would no longer be necessary.

Identifying and delineating the optimal targets The literature regarding optimal target point/volume for the different targets is surprisingly meagre. The most common target, the subthalamic nucleus (STN), is also the most studied regarding this issue, but the effect has been evaluated in relation to the location of the electrode in no more than 260 patients in a total of 13 different studies. Interestingly, most of these studies have found that the best effect is not achieved in the target structure itself, but in another adjacent structure, the more rostral part of the Zona incerta overlying the STN. The same is true regarding the few studies of thalamic surgery for tremor, where most have demonstrated that the best effect is achieved from contacts outside the thalamic target, in the PSA. Concerning other targets/indications the literature is even more limited.

The existing studies are further hampered by the heterogeneity of the methods. Often it is not the actual location of the electrode that is reported, but the intended one. Even when the actual location is reported it is most often presented according to distant landmarks, even if the targeting was done visually, which is a significant problem, taking into consideration the inter-individual variability. Further, even though the stimulation parameters used can differ widely and hence the electrical current affect different neighboring structures, this is virtually never taken into account.

DBS for ET - Conclusions and aim DBS for ET is thus not an evidence based therapy and the same is true regarding both target areas used for ET. The relative merits of the different targets used have not been clearly demonstrated and the individual targets are not well delineated. There is thus a need to address these issues.

Overview - What are we going to study? The effect of DBS for ET will be evaluated in a multicentre study recruiting 100 patients randomized to immediate surgery (group A) or best medical treatment only (as decided by the movement disorder neurologist, with few exceptions likely to be identical to their current medication) with delayed surgery at 6 months if still needed /indicated (group B). Primary outcome is tremor reduction as measured with Essential tremor rating scale11 (ETRS), with focus on items 5/6 (hand tremor) \& 11-14 (hand function), at baseline before surgery and at 6 months. Secondary outcome includes the essential tremor rating assessment scale (TETRAS), quality of life measured with QUEST and electrical energy consumption18. After 6 months group B is operated -as indicated-, and group A and B are thereafter evaluated as one single group for long-term effects and analyses of electrode location. The electrodes will in all patients be inserted with a trajectory developed for this study allowing stimulation both in the established and traditionally used target Vim as well as in the new target cZi. Each contact will be evaluated individually concerning location, effects and side effects, in relation to stimulation parameters. This is done in order to, a) compare the two targets, but more importantly, b) to create a map of the whole area with identification/delineation of the optimal target point/volume. The time necessary to gather 100 patients is estimated to 2.5 - 3 years.

Study design:

• Step 1 is designed as a randomized controlled trial in order to compare the effect of DBS (Group A) and medical treatment (Group B) on ET (Aim 1).
• In step 2 - 4 the patients in group A are operated and group A and B joined into one single group.
• Step 2 - all electrode contacts are individually analysed regarding chronic and acute stimulation effects in relation to their location in either the Vim or the cZi in order to decide which target is more effective (Aim 2).
• Step 3 - Effects and side effects, with consideration to field of stimulation, are analyzed in relation to neighboring structures in order to map the whole area and delineate the optimal target points (Aim 3).
• Step 4 - The long-term effect is evaluated in a non-randomized longitudinal study at 1, 3, 5, 10 years (Aim 4).

Sample size and power:

In step 1, with a confidence interval of 95%, and an expected improvement of 85 % in the surgical group and none (or at least < 25%) in the medical group, regarding contralateral items 5/6 \& 11-14, a power of 100 % is achieved. In step 2, with an expected improvement of 85% in one target and 60% in the other, a power of 98% is achieved. The high number of patients is, however, necessary for step 3, mapping of the area, in order to give a good cover of the periphery (Length of contact carrying electrode surface 7.5 mm, expected mean non-random intended deviation from target point (as defined in the study) 1 mm, expected mean random deviation from intended target point 1.5 mm). Based on the previous experience, less than 5% of patients is expected to be lost during the follow up, which will have an insignificant effect on the power.