Article Text

Download PDFPDF

Bilateral caudal zona incerta nucleus stimulation for essential tremor: outcome and quality of life
  1. Puneet Plaha,
  2. Shazia Javed,
  3. David Agombar,
  4. Genevive O' Farrell,
  5. Sadaquate Khan,
  6. Alan Whone,
  7. Steven Gill
  1. Institute of Neurosciences, Frenchay Hospital, Bristol, UK
  1. Correspondence to Professor Steven Gill, Department of Neurosurgery, Frenchay Hospital, Bristol BS16 1LE, UK; steven.gill{at}


Background Over the past few years, bilateral stimulation of the caudal or motor part of the zona incerta nucleus (cZI) has been performed by the authors in patients with essential tremor (ET). Outcomes including quality of life data in 15 patients with a follow-up period of up to 84 months (mean 31.7±28.6 months) are presented.

Methods 15 consecutive ET patients underwent MRI guided bilateral cZI deep brain stimulation implantation. Patients were assessed by applying the Fahn–Tolosa–Marin Tremor Rating Scale and the Short Form Health Survey-36 (SF-36) to assess quality of life.

Results The total tremor score improved by 73.8% (p<0.0001). The part A score (items 1–9) improved by 86.6% (p<0.0001). Postural tremor improved by 88.2% (p<0.0001) and action tremor by 82.2% (p<0.0001). The part B score, which evaluates the functional activities of the upper limbs, improved by 60.1% (p<0.0001). Part C score, which evaluates the activities of daily living, improved by 80.0% (p<0.0001). The SF-36 physical component score improved by 23.7% (p<0.0001) and the mental component score by 22.4% (p<0.0001). There was one wound infection and three patients developed stimulation related transient dysarthria. None developed any disequilibrium or tolerance to stimulation.

Conclusion Bilateral cZI stimulation is safe and effective in suppressing the postural and action component of ET. It is associated with a low incidence of stimulation related complications and patients do not develop tolerance to stimulation with maintained clinical benefit over a follow-up period of up to 7 years.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


Essential tremor (ET) is a common movement disorder that occurs in 300–415 persons per 100 000 population.1 The incidence increases with the age of the person and the disorder affects both men and women equally.

ET is a kinetic tremor usually with a postural and action component although an intention component can occur in advanced ET. ET can affect either the proximal or distal musculature. Although the terminology is sometimes used interchangeably, kinetic tremor is subclassified into tremor that occurs when maintaining limb position or posture (postural tremor), during active limb movement (action tremor) or when a limb is approaching a target (intention tremor).

Severe ET can be functionally disabling and drug treatment is effective in only 50% of patients.2 The medically refractory patients are offered either lesional or deep brain stimulation (DBS) surgery.

Typically, the ventralis intermedius (Vim) nucleus of the thalamus is the target of choice by implantation of DBS electrodes. However, proximal tremor and the action component of distal tremor respond poorly to DBS3 with only a third of patients showing any significant improvement.4 5 Perhaps the key limiting factor in effective suppression of tremor by bilateral DBS of the Vim nucleus is the high incidence (30–50%) of dysarthria and dysequilibrium.4 6–9 Therefore, in order to avoid the above complications most centres lesion or implant unilaterally, contralateral to the tremor dominant side. Another notable side effect of prolonged Vim stimulation is tolerance and lack of therapeutic effect, necessitating the device to be switched off for a variable duration of time.4 9 10

The subthalamic region comprises the subthalamic nucleus (STN), caudal zona incerta (cZI) and white matter tracts, including the cerebellothalamic prelemniscal radiation and the pallidothalamic H1 and H2 field of Forel. This area has been targeted since the 1960s to suppress ET by making large lesions involving the nuclei and white matter fibre tracts.11–13

Over the past few years a couple of centres have implanted unilateral DBS electrodes into this region to suppress tremor,13–15 with some specifically targeting the prelemniscal radiation.16 We implanted bilateral DBS leads into this area with good results.16

More recently, we targeted the cZI nucleus and implanted bilateral DBS leads with effective tremor control at the 12 month follow-up.17 In this paper, we present our outcomes, including quality of life data, on bilateral stimulation of the cZI nucleus in 15 patients with a follow-up period of up to 84 months (mean 31.7±28.6 months). Twelve month outcome for six of these patients (patient Nos 1, 2, 3, 6, 7 and 13) has been presented previously.17

Material and methods

Patient population

Fifteen consecutive patients (seven women, eight men) with medically refractory ET diagnosed according to the Consensus Statement of the Movement Disorder Society on Tremor were selected for surgery. Mean age of the patients was 65.4±7.9 years. Mean duration of the disease was 21.5±13.5 years (for details see table 1). Three patients (Nos 2, 3 and 13) have received stimulation for up to 84 months and one patient (No 6) for 48 months, with mean follow-up of the group 31.7±28.6 months.

Table 1

Patient demographics and clinical history

The Frenchay Hospital Local Ethics Committee gave approval to perform the stereotactic procedures under general anaesthesia using implantable guide tubes to deliver the DBS leads. After obtaining informed consent, all patients underwent bilateral DBS of the cZI nucleus.

Clinical evaluation

Non-blinded assessments were performed by applying the Fahn–Tolosa–Marin Tremor Rating Scale.18 Preoperative assessments were performed with patients off all antitremor medications for 12 h overnight. Postoperatively, they were assessed 12 h after stopping antitremor medications and switching off the stimulation (off medication/off stimulation) and then after switching on the DBS (off medication/on stimulation). A specialist movement disorder nurse performed all evaluations and was responsible for programming the stimulator under the supervision of the senior author and neurologist.

Outcome measures

The primary outcome measure was the percentage change between each of the baseline subscores of the Tremor Rating Scale (part A, part B and part C) when compared with the off medication/on stimulation score at follow-up. The Short Form Health Survey-36 (SF-36) was used to assess the improvements seen in the patient's quality of life following surgery. Surgery and stimulation related complications were recorded.


Surgery was performed under general anaesthesia using high resolution long acquisition MRI scans and plastic guide tubes (Neuroguide, Renishaw Plc, UK) for intraoperative verification of accurate DBS lead implantation. This MRI directed technique of performing functional surgery for both Parkinson's disease and non-parkinsonian tremor has been described by us in detail previously.16 17 19 20 The surgical planning with three-dimensional image construction of the STN, red nucleus and the planned target was done using the Neuroinspire surgical planning software (Renishaw Plc, Wotton-under-Edge, UK). The target area was confined to the cZI, which is located posteromedial to the posterodorsal STN (figure 1).

Figure 1

Left: Long acquisition axial T2 weighted preoperative MRI scan showing the subthalamic nucleus (STN) and the red nucleus. Right: Intraoperative inverted T2 weighted axial scan with the STN and red nucleus outlined. The red arrows point to the guide tube stylettes in the caudal zona incerta nucleus. After the intraoperative scan, the stylettes are removed and replaced with the deep brain stimulation leads.

A transfrontal trajectory to the target, 45o to the AC–PC plane, was planned and the anatomical position of each contact on the DBS lead was defined such that the second contact from its distal end (contacts 1 or 5) was placed at the target site (figure 2). Bilateral DBS leads (model 3389; Medtronic Inc, Minneapolis, USA) were implanted and connected to a DBS pulse generator (Kinetra; Medtronic Inc), which was implanted in the infraclavicular region.

Figure 2

Long acquisition MRI image acquired in all three planes and reconstructed using the Neuroinspire surgical planning software. Image shows the subthalamic and red nucleus in three-dimensions with the planned transfrontal trajectory of the deep brain stimulation lead in the caudal zona incerta nucleus.

Postoperative management

The Kinetra generator was switched on immediately after surgery. The patients were programmed by the movement disorder nurse who was guided by both the neurologist and the neurosurgeon. Antitremor medications were reduced as deemed appropriate by the neurologist. All patients subsequently stopped their antitremor medications except one patient (No 3) who remained on propranolol for hypertension.

Anatomical location of active contacts

The active contact programmed was noted and its anatomical location was identified by superimposing the intraoperative MRI scan onto the preoperative MRI plan scan on which we had defined the anatomical location of each DBS lead contact. If there was a displacement of the stylette from the planned target on the intraoperative MRI scan, the anatomical location of each of the contacts on the DBS lead could still be determined.

Statistical analysis

The primary efficacy was analysed using the paired Wilcoxon signed rank and sign test. The test of significance was applied to the scores of the tremor rating score and the SF-36 questionnaire.


Part A scores

The total tremor score improved by 73.8% (baseline mean score of 63.9±16.2 to 16.7±9.3, p<0.0001). The part A score (items 1–9) improved by 86.6% (baseline mean score of 21.5±9.8 to 2.9±2.9, p<0.0001) (figure 3). All patients had severe tremor in both upper limbs (mean postural score 10.2±5.6, mean action tremor score 9±2.8). Patient Nos 8, 9 and 12 also had severe rest tremor. Postural tremor improved by 88.2% (mean baseline score 10.2±5.6 to a score of 1.2±1.3, p<0.0001). Action tremor improved by 82.2% (mean baseline action tremor score of 9±2.8 to a score of 1.6±1.9, p<0.0001). The combined posture and action component tremor score improved by 86.1% (from a mean baseline score of 9.6±4.4 to a score of 1.4±1.7, p<0.0001).

Figure 3

Improvement in the total tremor score and subscores at baseline off medication (Baseline-off) and postoperative off medication/on stimulation (Postop-off/On).

Axial (face, head, voice and truncal) tremor

Face tremor improved by 90.4% (baseline mean score of 0.7±1.4 to 0.07±0.3, p=0.07). Head tremor improved by 93.2% (baseline mean score of 1.9±2.5 to 0.13±0.4, p=0.01). Voice tremor improved by 33.3% (baseline mean score of 0.6±1.1 to 0.4±0.7, p=0.2711). Trunk tremor improved by 100% (baseline mean score of 0.6±0.9 to 0.0, p=0.0230).

Part B scores

The part B score (items 10–14), which evaluates the functional activities of the upper limbs, improved by 60.1% (baseline mean score of 24.7±5.9 to 9.9±3.9, p<0.0001) (figure 3). Handwriting improved by 81.9% (mean baseline score of 2.9±0.8 to a score of 0.5±0.8, p<0.0001). The ability to draw spirals improved by 52.9% (mean baseline score of 12.1±3.3 to a score of 5.7±1.9, p<0.0001). The ability to pour water improved by 71.7% (mean baseline score of 5.4±1.6 to a score of 1.5±1.8, p<0.0001) (figure 3).

We also calculated the improvement by summing the postural and action tremor scores of the upper limb with motor score related to functional activities of the upper limb (writing, drawing and pouring). This category improved by 73.5% (mean baseline score of 7.9±4.6 to a score of 2.1±2.5, p<0.0001).

Part C scores

Part C score (items 15–21), which evaluates the activities of daily living, improved by 80.0% (mean baseline score 2.5±1.2 to 0.5±0.8, p<0.0001) (figure 3, table 2).

Table 2

Part C subscores outcome

Quality of life SF-36

Improvement in the patient's quality of life was assessed by applying the multipurpose SF-36.

The SF-36 physical component score improved by 23.7% (mean baseline score of 42.4±9.3 to a score of 55.6±6.3, p<0.0001) and the mental component score by 22.4% (mean baseline score of 46.6±7.9 to a score of 60.1±6.2, p<0.0001).

Subgroup outcome analysis on four patients with more than 4 years of follow-up

The total tremor score in these four patients improved by 72.6% (patient Nos 2, 3, 6 and 13) (baseline mean score of 58.5±13.5 to 16±4.8, p=0.0045).

The part A score improved by 81.5% (baseline mean score of 17.8±4.4 to 3.3±2.1, p=0.007). The part B score improved by 60% (baseline mean score of 22±3.7 to 8.8±2.2, p=0.0044). Part C score improved by 82.4% (mean baseline score 18.8±6.2 to 3.3±2.1, p<0.0230).

Pulse generator parameters

Following effective tremor control with DBS, all patients were weaned off their antitremor medications. The mean pulse generator parameters for the group are shown in table 3. Tremor in eight patients was controlled with monopolar stimulation while the rest required bipolar stimulation. In all 15 patients the chosen effective stimulation contacts were as planned preoperatively. There was no significant difference in the pulse width and frequency settings evaluated at 6 weeks post surgery and those evaluated at last follow-up. There was however a slight statistically significant increase in the amplitude of the current with time.

Table 3

Parameters of pulse generator at 6 week and at the last follow-up

Pulse generator parameters for four patients with a follow >4 years are shown in table 4.

Table 4

Pulse generator parameters in four patients with more than 4 years of follow-up

All patients received continuous stimulation and none developed tolerance to stimulation with follow-up to 84 months.


One patient developed a wound infection on the background of insulin dependent diabetes (patient No 14); the system was explanted and replaced after 8 weeks following antibiotic treatment. Three patients (patient Nos 6, 10 and 14) developed bilateral stimulation related dysarthria and a hypophonic speech. The dysarthria in all three was mild, such that their speech rate was decreased with suboptimal articulation; however, it was completely intelligible. This effect was directly proportional to an increase in voltage and was completely reversible on switching the stimulator off. The dysarthria was not always noticeable by the patient but was by an observer. In patient Nos 6 and 14, the dysathria was intermittent and tended to exhibit itself towards the end of the day when the patient was fatigued; in patient No 10 however, it was present at all times. None of the patients developed any disequilibrium.


The functional outcome and quality of life data on all 15 patients shows that the therapeutic effect of DBS of the cZI in patients with ET is maintained over a follow-up period of up to 84 months.

Surgical method and localisation of active contacts

The intraoperative axial and the preoperative three-dimensional planning pictures show that we have targeted the caudal or motor part of the ZI nucleus. Our described method relies on visualisation of the brain targets on high resolution MR images acquired in all three planes under strict stereotactic conditions.16 17 19 20 The technique also provides a means of confirming the accuracy of targeting using intraoperative MRI imaging to identify the relationship of an indwelling stylette to the visualised anatomical target.17 19 20 It also helps in identifying the precise anatomical location of each contact on the DBS electrode.

The cZI target lies posteromedial to the posterodorsal STN. This target is posterior to our previously defined target for the suppression of ET, which was medial to the posterior dorsal one-third of the STN and involved the cerebellothalamic fibre tracts in the prelemniscal radiation.16 As published recently, bilateral stimulation in this region in our series of patients with tremor dominant Parkinson's disease resulted in a 40% incidence of hypophonic speech and disequilibrium.19 We therefore moved our target more posterior to the cZI.

In our recent publication,17 the active contact in all six ET patients was within the cZI nucleus (figure 2 in Plaha and colleagues17). In all 15 patients in this paper the active contacts (contact 1 or 5) were also within the cZI nucleus. Contacts 0 and 4 on the DBS leads with our transfrontal trajectory were at the junction of the cZI and the medial lemniscus while the higher contacts (2, 3 and 6, 7) were dorsal to the posterior one-third of the STN.

Patient outcome, complications and tolerance

The total tremor score improved by 73.8%; the part A score by 86.6%, part B by 60.1%, the functional motor score by 73.5% and the part C by 80%. These results show that sustained clinical benefit is obtained with bilateral stimulation and is mirrored in the patient's subjective assessment of outcome (activities of daily living (part C) and SF-36 score). This effect is maintained over a follow-up period of up to 7 years.

Although our patient numbers are small, we believe the results are worthy of a comparison with Vim nucleus stimulation for ET which improves distal action tremor by 50–65% following unilateral or bilateral stimulation.6 21 22 The drawback of bilateral Vim nucleus stimulation is the high incidence of stimulation related dysarthria and dysequilibrium, reported as ranging from 30% to 50%4 6–9 and more recently an even higher incidence in a multicentre study (dysarthria 75% and disequilibrium 56%).22

Vim DBS is not effective in suppressing proximal tremor and the action component of distal tremor,3 with only one-third of patients showing any significant improvement.4 5 To control this component of ET, both unilateral14 15 and bilateral stimulation16 17 of the posterior subthalamic region, including the cZI, has been performed with good results. Some centres that implant DBS leads into the Vim thalamus on retrospective analysis of the anatomical location of the DBS contacts found the most effective contact to control was in fact within the subthalamic region and not the thalamus.23 24

In our series, axial head and neck and face tremor was well controlled. Bilateral thalamic DBS is effective in controlling/suppressing axial head and neck tremor8 21 25 although as discussed above, this is achieved at the cost of stimulation related side effects.

Benabid et al first reported tolerance to Vim stimulation for ET. Tolerance to stimulation may occur after weeks or months and a regular increase in stimulation intensity is necessary to maintain control. Even at a maximally tolerable intensity, tremor may still breakthrough. To prevent this, patients are advised to turn off their stimulators at night, take stimulation holidays10 or only switch on the DBS generator when performing a task (on demand).26 Benabid et al found that tolerance occurs in up to 18.5% of cases by 3–6 months4 9 and more recent long term data have shown that the efficacy of Vim stimulation decreases over time.27 None of our patients developed any tolerance to continuous cZI stimulation with long term follow-up to 84 months. The stimulation voltage remained low and along with the pulse width did not change significantly over time. We speculate that the tolerance seen with Vim stimulation is probably secondary to continuous stimulation current causing long term depression of neuronal synapses between thalamocortical and GABAergic interneurons thereby weakening neuronal transmission to the motor cortex. A second hypothesis could be that the absolute refractory period of thalamocortical neurons during the firing of an action potential is longer than cZI neurons.

Caudal zona incerta and ET

The ZI, an embryological derivative of the ventral thalamus, is a distinct heterogenous nucleus that lies at the base of the dorsal thalamus and is an extension of the reticular thalamic nucleus.28 Its rostral component extends over the dorsal and medial surface of the STN whereas its caudal or motor component lies posteromedial to the STN.29 30

The cZI is bounded by the medial lemniscus posteriorly, the ascending cerebellothalamic fibres in the prelemniscal radiation anteriorly, and lying anteroventral to the cZI is fasciculus Q or area Q. Fasciculus Q lies dorsomedial to the substantia nigra and medial to the ventral half of the STN.31

As discussed in detail in our previous publication17 and summarised here very briefly, the ZI provides a unique GABAergic link between the basal ganglia output nuclei and the cerebello-thalamo-cortical loop. This places it in a key position to transmit synchronised oscillations into these loops (figure 4).

Figure 4

ZI provides a unique GABAergic link between the basal ganglia output nuclei, the cerebello-thalamo-cortical loop and the brainstem nuclei, and synchronises the oscillatory firing of these subcortical nuclei. The GABAergic ZI output is shown as stippled blue to signify its synchronising function. MRF, medial reticular formation; PFC, prefrontal cortex; VA ventralis anterior nucleus of thalamus; VL, ventrolateral nucleus of thalamus; ZI, zona incerta.

In ET synchronised oscillations at 4–12 Hz from excessive electronic coupling between dendrites of the inferior olivary neurons (IO) via GABA mediated gap junctions32 are transmitted to the cerebellar cortex and then distributed along the ascending cerebellothalamic pathway and the descending brainstem medial reticular formation to manifest as tremor. The interpositus nucleus of the cerebellum appears to modulate proximal limb muscles33–35 during movement correction via its descending afferents36 to the medial reticular formation. Distal muscle control is via its ascending efferent fibres to the motor cortex that relay both in the ventrolateral nucleus and the magnocellular red nucleus.36 As the rubrospinal tract which originates from the magnocellular red nucleus is rudimentary in humans, distal muscle control is primarily via the motor cortex.37

ZI has key efferents to the cerebello-thalamo-cortical loop, ventrolateral nucleus of the thalamus, olivary neurons and the medial reticular formation and therefore is in a key position to override tremor oscillations within these loops following DBS.


Bilateral cZI stimulation is safe and effective in suppressing medically refractory ET. It is associated with a low rate of stimulation related complications and patients do not develop tolerance to stimulation, with maintained clinical benefit over a follow-up period of up to 7 years. A randomised controlled trial is required to determine the efficacy of bilateral cZI stimulation versus Vim nucleus stimulation.


The authors thank Dr Peter Heywood for his assistance in the management of patients.



  • Funding PP was supported by a grant from the Medical Research Council UK (G9900797).

  • Competing interests SG and SJ are consultants with Renishaw Plc, UK.

  • Ethics approval This study was conducted with the approval of the Frenchay Hospital Ethics Committee, Bristol.

  • Provenance and peer review Not commissioned; externally peer reviewed.