Assessment of Cerebellar Ataxia Associated With Ventral Intermediate Median Nucleus Deep Brain Stimulation in Essential Tremor: a Multimodal Neurophysiologic Study

Essential tremor (ET) is a chronic, progressive neurological disease characterized by a symmetrical 4-12 Hz kinetic tremor constantly affecting hands and arms. The tremor can also involve the neck, the lower limb, the voice, the mouth, the tongue or other body regions. ET is among the most frequent movement disorders, with an estimated prevalence of 0,9% of the global population. While the tremor can be mildly disabling at the beginning of the disease, most patients experience a worsening of their tremor over the time. Eventually, disease progression can lead to severe disability, impacting patients' working abilities and in the most severe cases, activity of daily living. Though ET being a frequent debilitating condition, there are still few therapeutic options for the most disabled patients. Several surgical procedures have been evaluated in ET including Deep brain Stimulation (DBS) of Thalamic Ventral intermediate nucleus (Vim) or within the posterior subthalamic area, surgical thalamotomy, Gamma knife or, more recently, MRI guided focused ultrasound thalamotomy. While Gamma knife and focused ultrasound thalamotomy may be safer due to the absence of craniotomy, Vim-DBS remains currently the reference treatment and the most used procedure for medically refractory ET.

Besides the early side effects of Vim-DBS such as haemorrhagic, ischemic and infectious complications occurring immediately after the electrodes implantation procedure, late side-effects such as cerebellar symptoms including gait or upper limb cerebellar ataxia, paraesthesia, dysarthria, dysphagia, and cognitive difficulties are quite frequent. These long-term consequences of the stimulation may dramatically impact the quality of life of patients suffering from ET years after the surgery and mitigate the benefits of this intervention. Considering gait ataxia, it has been reported that this complication may affect up to 37% of ET patients following Vim-DBS surgery and seem to occur more frequently in patient who had pre-operative gait impairment, greater disease severity, older age and when DBS was performed bilaterally. Upper limb ataxia is also common and characterized by the occurrence or the paradoxical worsening of an intention tremor together with asynergy, dyschronometria and dysdiadochokinesia. It has been shown that this latter can be induced by suprathreshold stimulation.

One of the major assets of Vim-DBS over surgical thalamotomy remain its ability to steadily adapt the current delivered by the electrodes by changing the stimulation parameters. Using traditional Implantable Pulse Generator (IPG), it is possible to modify the stimulation mode (monopolar or bipolar stimulation), amplitude, frequency and pulse width as well as to modify the contacts of the electrodes delivering the current. However, despite DBS being a treatment largely available worldwide as well as being part of the routine care for medically refractory tremor, only few studies have assessed the effects of varying DBS settings on cerebellar symptoms presented by patients suffering from ET and stimulated within the Vim. It has been reported that tremor responds best to amplitude, which is usually increased until a complete resolution of the tremor is obtained. Increasing the stimulation pulse width produces a similar clinical effect and may help to alleviate tremor. Interestingly, it has recently been shown that decreasing the pulse width may improve the gait disorders, widen the therapeutic window and diminishes DBS induced side effects. This clinical finding is supported by the recent dissection of the neuronal network underlying Vim DBS clinical response which suggest that tremor suppression could be due to dentato-rubro-thalamic tract stimulation while gait and limb ataxia could be promoted by the stimulation of the cerebello-rubro-spinal and the rubro-olivo-cerebellar tract meaning the possibility of correcting these side effects by accurately optimizing the DBS parameters.

Among the clinical effect of the different types of parameters, the stimulation frequency has certainly been the less extensively studied in ET. This is at first sight quite surprising given that decreasing the frequency of stimulation of the Subthalamic nucleus (STN) is a valuable therapeutic option in Parkinson's disease especially for patient presenting with dopa-refractory symptoms such as freezing of gait or speech impairment. Lower frequency stimulation has also shown to be beneficial for some patients suffering from dystonia, and it has been suggested that this sustained clinical improvement was obtained by avoiding the stimulation of the structure responsible of side effects.

In comparison, a very low number of studies has focused on the consequences of lower frequency stimulation on tremor. Noteworthy is the fact that most of them have pointed that the tremor suppression was better achieved for a frequency around 100 Hz, with little evidence for a small additional therapeutic effect for frequencies above 130 Hz. Conversely, frequency of 50 Hz or lower seems to not suppress tremor and could even worsen it or induce myoclonic jerks. Based on these clinical findings, it is currently admitted initiating the stimulation with a frequency of 130 Hz and increasing it up to 180 Hz if the therapeutic effect on tremor remains unsatisfying despite concomitant high amplitude and large pulse width. A recent small case series has interestingly shown that some patients presenting with balance disorder years after Vim DBS surgery were having their trouble improved when the stimulation frequency was decreased to 130 Hz. Additionally, there are some evidence that, similarly to short pulse width stimulation, decreasing the frequency up to 100 Hz may help to optimize postural or intention tremor and DBS side effects control by widening the therapeutic window. Taken together, these findings suggest that, similarly to shorter pulse width, a lower frequency of stimulation could be an interesting therapeutic approach for ET patients suffering from balance disorders or upper limb ataxia promoted by Vim-DBS. However, since all the above-mentioned studies were unblinded and non-randomized, the level of evidence supporting such a strategy remain weak. Additionally, none of these works has aimed at evaluating the effect of decreasing DBS frequency on gait and balance, neither by using clinical testing, nor by using quantitative measurement such as 3D motion analysis or computerized posture assessment.