Introduction

Tremor, defined as an involuntary, rhythmic, oscillatory movement of a body part [1] is a key semeiological feature of Parkinson disease (PD) and essential tremor (ET) [2, 3], as well as the cardinal symptom of less common but equally disabling movement disorders, namely Holmes tremor (HT), orthostatic tremor (OT), and fragile X-associated tremor/ataxia syndrome (FXTAS) [1, 4]. In addition, tremor represents a secondary yet disabling feature of immune-mediated neurological disorders such as multiple sclerosis (MS) and demyelinating peripheral neuropathies [1].

Tremulous disorders are classified basing on two axes, consisting of (a) the clinical and semeiological characteristics of tremor and (b) the underlying etiology [1]. HT is a rest, postural, and intention low-frequency (< 5 Hz) tremor [1] usually associated with an ischemic or demyelinating lesion of the cerebello-thalamo-cortical or dentate-rubro-olivary pathways, with superimposed nigrostriatal dysfunction [1, 5, 6]. OT is a 13–18 Hz tremor affecting the weight-bearing limbs and resulting in a sensation of unsteadiness and imbalance when standing [1, 4, 7]. Although the exact location of its generator remains unclear, available evidence suggests the involvement of the cerebellum and pons [7, 8]. MS-associated tremor is a heterogeneous syndromic entity that includes at least three different subtypes of tremor [1, 4]: a 4–5 Hz cerebellar tremor predominantly involving the upper extremities; a 3–5 Hz ataxic tremor of the head (“head titubation”) and the trunk; and a 7–12 Hz mild-amplitude postural hand tremor [9]. Neuropathy-associated tremor is a posture and intention 3–6 Hz tremor predominantly involving the upper limbs [1, 10], and possibly associated with an altered peripheral sensory input, which might prevent the cerebellum ability to correct limb position and velocity during voluntary movements [11, 12]. Finally, FXTAS-associated tremor is a posture and intention tremor associated with ataxia, parkinsonism, cognitive decline, peripheral neuropathy, autonomic dysfunction, and psychiatric disorders [13, 14], and caused by a trinucleotide repeat expansion in the premutation range (CGG block lengths 55–200) in the FMR1 gene.

While pharmacological therapies, mostly beta-blockers, benzodiazepines, and anticonvulsants represent the first line of treatment for common tremor disorders [15], uncommon tremors may be refractory to oral medications and, in selected cases, require advanced surgical treatments such as deep brain stimulation (DBS) [16, 17]. The risk/benefit profile of DBS in uncommon tremor disorders, however, remains to be clarified.

In this review, we sought to analyze the literature related to the safety and efficacy of DBS in uncommon tremor disorders, discussing clinical indications, anatomical targets, and programming strategies.

Materials and methods

We reviewed the literature for studies reporting the outcome of DBS in patients with HT, OT, MS-associated tremor, neuropathy-associated tremor, or FXTAS. Relevant articles were identified through electronic search of PubMed.gov and ClinicalTrials.gov using the following keywords: “tremor”, “Holmes tremor”, “orthostatic tremor”, “multiple sclerosis”, “multiple sclerosis tremor”, “neuropathy”, “neuropathic tremor”, “fragile X-associated tremor/ataxia syndrome”, and “fragile X.” No language restrictions were applied.

We selected studies involving humans. There were no restrictions applied to gender, age, disease duration, or disease severity. The data extracted included the following: (a) DBS targets including, but not limited to, ventral intermediate (Vim) nucleus, ventro-lateral (VL), ventralis oralis anterior (VOA) and ventralis oralis posterior (VOP) nuclei, posterior subthalamic area (PSA), subthalamic nucleus (STN), globus pallidus pars interna (GPi), and zona incerta (ZI); (b) magnitude of response, according to the available presentation of clinical data; (c) follow-up duration; (d) adverse events (AEs); and (e) stimulation settings.

Results

A total of 263 cases of uncommon tremor disorders treated with DBS were identified through a systematic review of PubMed.gov and ClinicalTrials.gov. Of these, 44 had Holmes tremor (HT); 18 had orthostatic tremor (OT); 177 had multiple sclerosis (MS)-associated tremor; 14 had neuropathy-associated tremor; and 10 had fragile X-associated tremor/ataxia syndrome (FXTAS).

Holmes tremor

We identified 24 studies (17 single case reports and 7 case series), which totaled 44 patients treated with DBS for HT (Table 1).

Table 1 Summary of literature data on Holmes tremor cases treated with deep brain stimulation
Full size table

There were 31 patients treated with Vim-DBS: 13 received unilateral Vim-DBS, 5 bilateral Vim-DBS, and 13 Vim-DBS plus an additional target (GPi n = 5; STN n = 5; VOA/VOP border n = 3). Tremor improved in all cases, with a reduction ≥ 80% in 10/31 patients (32.3%). The extent of improvement was not specified in 12/31 patients (38.7%). The DBS electrode was removed in 1 patient treated with unilateral Vim-DBS due to lack of efficacy. Time to follow-up varied from 3 to 52 months. AEs were reported in 4/31 patients (12.9%), consisting in 1 infection of the implantable pulse generator (IPG), 1 case of facial paresthesia, 1 case of transient dysarthria, and 1 case of ataxia, paresthesia and dystonic movements. Stimulation frequency ranged from 100 to 180 Hz, intensity from 1.7 to 4.8 V, and pulse width from 60 to 210 μs (Table 1) [18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37].

GPi-DBS was used in 14 patients with HT: 7 received unilateral GPi-DBS, 2 bilateral GPi-DBS, and 5 GPi-DBS plus Vim-DBS. Tremor improved in all cases. A reduction ≥ 80% occurred in at least one component of tremor in 11/14 patients (78.6%). Follow-up spanned from 6 to 52 months. All but one study reported no occurrence of AEs, and 1 study did not report AEs. Stimulation frequency ranged from 130 to 185 Hz, intensity from 2.0 to 6.0 V, and pulse width from 90 to 210 μs (Table 1) [26, 30, 35,36,37].

STN-DBS was used in 6 patients with HT: 5 received unilateral STN-DBS plus Vim-DBS, and 1 STN-DBS plus Vo-DBS. Tremor improved in all cases, with a reduction ≥ 80% in 1/6 patients (16.7%). The follow-up duration ranged from 24 to 72 months. All but one study reported no occurrence of AEs, and one study did not report AEs. Stimulation frequency ranged from 135 to 145 Hz, intensity was 2.0 V, and pulse width varied from 90 to 210 μs (Table 1) [21, 34, 38].

Three cases were treated with ZI-DBS (1 bilateral ZI, 1 unilateral VOA/ZI, 1 unilateral lenticular fasciculus/ZI). Tremor improved in all cases, with a reduction ≥ 80% in 1/3 patients (33.4%). The follow-up duration ranged from 12 to 48 months. There was a lead infection 4 years after surgery, requiring removal of the system, and one case of transient dysphagia. Stimulation frequency ranged from 145 to 185 Hz, intensity from 1.8 to 3.0 V, and pulse width from 60 to 120 μs (Table 1) [39,40,41].

Orthostatic tremor

We identified 9 studies (6 single case reports and 3 case series), reporting a total of 18 patients treated with DBS for OT (Table 2).

Table 2 Summary of literature data on orthostatic tremor cases treated with deep brain stimulation
Full size table

Vim-DBS was used in 18 patients: 17 received bilateral Vim-DBS, and 1 unilateral Vim-DBS. Tremor amplitude reduction or increased latency to symptom onset after standing was reported in 15/18 patients (83.3%). One patient treated with bilateral Vim-DBS reported no improvement, and the patient receiving unilateral Vim-DBS reported only a temporary improvement, which lasted less than 6 months. Most of patients experienced a mild/moderate waning of tremor improvement over time. The follow-up duration ranged from 6 to 102 months. A total of 14 AEs was reported in 13/18 patients (72.2%), consisting in 1 skin infection over IPG, 2 surgical revisions due to cervical pain and lead dislocation, 1 generalized tonic–clonic seizure 3 days after surgery, 2 transient paresthesia, 2 gait ataxia, 1 unilateral foot dystonia, 1 case of dizziness, and 4 cases of speech difficulties. Stimulation frequency varied from 130 to 185 Hz, intensity from 1.5 to 4.0 V, and pulse width from 60 to 90 μs (Table 2) [42,43,44,45,46,47,48,49,50].

Multiple sclerosis-associated tremor

We identified 26 studies (3 single case reports, 22 case series, and 1 randomized clinical trial), for a total of 177 patients treated with DBS for MS-associated tremor (Table 3).

Table 3 Summary of literature data on multiple sclerosis-associated tremor cases treated with deep brain stimulation
Full size table

Vim-DBS was used in 146 patients: 93 received unilateral Vim-DBS, 26 bilateral Vim-DBS, 25 Vim-DBS plus an additional target (STN n = 11; VO n = 13; pre-lemniscal radiations n = 1), and 2 unspecified Vim laterality. Symptoms improved to some extent in 126/146 patients (83.6%). A transitory response was observed in 20/146 patients (13.7%). The follow-up duration ranged from 3 to 127 months. Several patients experienced AEs, encompassing MS exacerbation, seizures, intracerebral hematoma, gait/balance disturbance, asthenia and transient lower limb paresis, ataxia, dysarthria, paresthesias, and infections. Stimulation frequency ranged from 130 to 185 Hz, intensity from 0.5 to 8.0 V, and pulse width from 60 to 210 μs (Table 3) [25, 51,52,53,54,55,56,57,58,59,60,61,62,63,64,65,66,67,68,69,70].

ZI-DBS was used in 25 patients with MS-associated tremor: 4 received bilateral ZI-DBS, and 21 bilateral ZI-DBS plus VOP-DBS. Improvement of all the tremor components was reported for most of treated patients. Some cases gradually deteriorated across a follow-up of 12–62 months. AEs were reported in 8/25 patients (32%), consisting in 1 infection with scalp erosion leading to DBS removal, 2 other wound infections, 2 peri-operative seizure, 1 transient hemiparesis, 1 mild dysarthria, and 1 case of post-operative prolonged lethargy and reduced mobility. Stimulation frequency ranged from 40 to 105 Hz, intensity from 1.9 to 3.8 V, and pulse width from 90 to 330 μs (Table 3) [39, 71, 72].

VL thalamic DBS was used in six patients with MS-associated tremor: four received bilateral VL-DBS and 2 VL-DBS plus STN-DBS. Tremor improved in all cases, but follow-up was limited to fewer than 12 months. One study did not report AEs, and 1 study reported stimulation-induced paresthesia, dysarthria, and gait ataxia in a cohort of mixed patients, without specifying whether AEs occurred in MS patients. Stimulation frequency ranged from 130 to 145 Hz, intensity from 2.0 to 3.6 V, and pulse width was 60 μs in all cases (Table 3) [73, 74].

Neuropathy-associated tremor

We identified 9 studies (7 single case reports and 2 case series), reporting a total of 14 patients treated with DBS for tremor associated with neuropathy (Table 4).

Table 4 Summary of literature data on neuropathy-associated tremors treated with deep brain stimulation
Full size table

Vim-DBS was used in 13 patients: 9 received bilateral Vim-DBS, and 4 unilateral Vim-DBS. Tremor improved in all cases. There was a complete resolution of tremor in 1/14 patient (7.1%), and a “marked” or “dramatic” tremor improvement in 2/14 patients (14.3%). Follow-up ranged from 6 months to 9 years. AEs were reported in 4/13 patients (30.8%), with 1 case of mild transient paresthesia, 1 case of dysarthria and worsening of gait ataxia, 1 case of mild dysarthria, and 1 skin erosion over IPG. Two studies reported no occurrence of AEs, and two studies did not report AEs. Stimulation frequency varied from 130 to 185 Hz, intensity from 0.6 to 4.5 V, and pulse width from 60 to 210 μs (Table 4) [12, 75,76,77,78,79,80,81].

One patient received unilateral PSA-DBS with a significant improvement of symptoms over a follow-up of 12 months. Intraoperative occurrence of visual phenomena, speech alterations, and paresthesia were reported. Stimulation frequency was 145 Hz, intensity 3.0 V, and pulse width 60 μs (Table 4) [82].

Fragile X-associated tremor/ataxia syndrome

We identified eight studies (seven case reports and one case series), reporting a total of ten patients treated with DBS for FXTAS (Table 5).

Table 5 Summary of literature data on Fragile X-associated tremor/ataxia syndrome treated with deep brain stimulation
Full size table

Vim-DBS was used in eight patients: four received bilateral Vim-DBS, two unilateral Vim-DBS, one bilateral Vim-DBS plus PSA-DBS, and one bilateral Vim-DBS plus bilateral STN-DBS. Tremor improved in all cases, with a reduction ≥ 50% in 5/8 patients (62.5%). In four patients, a concomitant improvement of ataxia was reported. The follow-up duration ranged from 6 to 60 months. AEs were assessed in all studies and reported in 5/8 patients (62.5%), consisting in three cases of ataxia and speech worsening (in one case with associated cognitive decline), and two cases of slight worsening of cognition after surgery. Two studies reported no occurrence of AEs. Stimulation frequency varied from 125 to 185 Hz, intensity from 1.5 to 5.4 V, and pulse width from 60 to 150 μs (Table 5) [83,84,85,86,87,88].

Vo/ZI-DBS was used in one patient with unquantified improvement of tremor and associated ataxia. Follow-up was 30 months, during which no AEs were reported. Stimulation settings were as follows: frequency = 176 Hz; intensity = 3.0 V (right) and 3.25 V (left); and pulse width = 90 μs (Table 5) [89].

PSA-DBS was used in one patient, resulting in a 58% improvement in tremor severity with no AEs occurrence. The follow-up duration was 8 months. Stimulation frequency was 160 Hz, intensity 2.8 V, and pulse width 120 μs (Table 5) [90].

Discussion

We reviewed and summarized studies that reported the outcomes of DBS in uncommon tremor disorders, namely HT, OT, MS-associated tremor, neuropathy-associated tremor, and FXTAS. Apart from one randomized clinical trial that tested the efficacy and safety of DBS in MS [70], the majority of studies consisted of single case reports or small case series. While this relevant limitation should be recognized, suggesting the possibility that a “file drawer effect” might have influenced our results, data reported in the literature seem to suggest that DBS may be clinically useful in uncommon tremor disorders.

The majority of cases received Vim-DBS, but other surgical targets were also tested, including GPi, ZI, PSA, VOA or VOP nuclei, STN, and VL thalamus (Fig. 1). A few patients with HT and MS-associated tremor received dual-target DBS, mostly Vim-DBS in association with GPi-DBS or STN-DBS. Available data do not permit one to draw final conclusions on the best target in each subtype of tremor. Nonetheless, promising results were observed with PSA-DBS and with dual-target DBS. In fact, the most conventional Vim-DBS target was sometimes confounded by worsening pre-existent ataxia in FXTAS, MS-, and neuropathy-associated tremor. Similar complications have been reported in patients with OT, where the accuracy of Vim-DBS targeting is complicated by the somatotopic arrangement of thalamic fibers. In fact, the thalamic region corresponding to the lower limbs is located laterally and is in close proximity to the internal capsule. Such specific anatomical organization accounts for the programming’s narrow therapeutic window that is frequently observed in uncommon tremor disorders. Motor and sensory stimulation-induced side effects consist of speech impairment, sensory issues, motor pulling, and possibly gait ataxia. New directional DBS systems capable of greater anatomical resolution may potentially address these difficulties, leading to an improvement in the risk/benefit profile of DBS treatment in uncommon tremor disorders.

Fig. 1
figure 1

Main DBS targets for uncommon tremor disorders. Holmes tremor = Vim; Voa; Vop; STN; Gpi. Orthostatic Tremor = Vim; Zi. MS-associated tremor = Vim; Voa; Vop; Zi; Raprl. FXTS = Vim; Vop; Zi; PSA. Neuropathy-associated tremor = Vim; PSA. DBS deep brain stimulation, FXTAS fragile X-associated tremor/ataxia syndrome, MS multiple sclerosis, GPe globus pallidus pars externa, GPi globus pallidus pars interna, IC internal capsule, MS multiple sclerosis, PUT putamen, PSA posterior subthalamic area, Raprl pre-lemniscal radiation, SN substantia nigra, STN subthalamic nucleus, TH thalamus, Voa ventralis oralis anterior, Vop ventralis oralis, ZI zona incerta

Full size image

Overall, our main findings can be summarized as follows:

Holmes tremor Vim-DBS, GPi-DBS, and STN-DBS appear to represent effective targets in HT, with a potential to modulate both cerebellar-thalamic and nigrostriatal outflows. However, given the multiple pathways involved in its pathogenesis, a combination of two targets might also be considered. In particular, positive results were observed in patients treated with Vim-DBS and GPi-DBS or with Vim-DBS and STN-DBS. The vast majority of studies reported a programming strategy based on single monopolar stimulation, with high frequency (160–180 Hz in most cases) and a pulse width of 60–90 μs (≥ 100 μs in few cases). Bipolar electrode configurations were occasionally used, mostly due to stimulation-induced side effects. Vim-DBS can be considered as a viable therapeutic option for HT patients that have failed multiple oral treatments, including levodopa, dopamine agonists, anticholinergics, and anticonvulsants [15, 91].

Orthostatic tremor Bilateral Vim-DBS demonstrated long-lasting efficacy in reducing tremor in most patients; a few cases treated with unilateral Vim-DBS reported transient and unsatisfactory results. The vast majority of studies reported a programming strategy based on single monopolar stimulation, with a frequency of 160–180 Hz (less frequently 130 Hz) and a pulse width of 60 or 90 μs. More complex programming strategies, such as bipolar configurations and interleaving, were occasionally used. Although partial efficacy in reducing OT has been demonstrated by benzodiazepines, followed by beta-blockers and anticonvulsants [7], the severity of OT symptoms usually progresses over time. This, therefore, lessens the efficacy of oral medications. Results from an international registry suggest that bilateral Vim-DBS might significantly improve the severity of symptoms in patients with incomplete response to oral medications.

Multiple sclerosis-associated tremor Dual-lead thalamic DBS (one targeting the VIM border and one targeting the ventralis oralis anterior-ventralis oralis posterior border) has the highest level of clinical trial support [70], providing a better control likely by stimulating the major pathways involved in MS-associated tremor, and cerebello-thalamo-cortical and pallidal pathways. Almost all studies reported the use of single monopolar stimulation, with a frequency ranging from 130 to 180 Hz and variable pulse width, ranging from 60 to 210 μs, but typically higher than other tremors. Interestingly, a very short frequency stimulation (40 Hz) was reported as efficacious in controlling the tremor of four patients. The application of DBS in MS-associated tremor is debated. Approximately 46% of MS patients have some form of tremor, severe enough to impair daily living activities in 5.8% of cases [9]. Beta-blockers and botulinum toxin injection [92, 93] should represent the first line of treatment. DBS can be considered in cases with particularly disabling symptoms, refractory to oral medications.

Neuropathy-associated tremor Vim-DBS appears to provide substantial improvement and improvement was reported in the only case treated with STN-DBS. The typical programming strategy was based on single monopolar stimulation, with a frequency of 130 or 180 Hz and a pulse width of 60 or 90 μs (in 3 cases > 100 μs). While the employment of DBS in neuropathy-associated tremor requires further confirmation by larger clinical trials, the significant impact played by tremor on functional activities in patients with immune-mediated neuropathies suggest that DBS might be a feasible therapeutic option for patients with neuropathy-associated tremor failing propranolol, primidone, and benzodiazepines [94].

Fragile X-associated tremor/ataxia syndrome Vim-DBS seems to be effective in improving tremor and sometimes also ataxia; however, a high rate of cases showed a worsening of ataxia, speech, and cognition after surgery, raising the question of the balance between the possible benefit and the potential harm deriving from DBS surgery in these patients. Most studies reported a stimulation frequency of 160–185 Hz and a pulse width of 90 μs or higher. Complex programming strategies such as bipolar or double monopolar stimulation were occasionally used, mostly due to stimulation-induced side effects. Further controlled clinical trials would help to clarify the extent of benefit achievable with Vim-DBS, as compared to oral medications, in patients with FXTAS.

Conclusions

Our review summarized the limited evidence available as well as some of the challenges associated with the use of DBS in uncommon tremor disorders. First, the inherent rarity of these conditions is, indeed, a limitation, which lends itself to certain publications being underpowered, and thus with an elevated risk of Type II errors. Second, the frequent association of uncommon tremor disorders with other movement disorders, such as ataxia, dystonia, and chorea, enhances the litany of confounders endemic to the analyses, which make disentanglement of cause-and-effect relationships challenging. Specifically, such clinical heterogeneity hampers the ability to determine the nature, and magnitude, of certain interventions and what outcomes they portend with great reliability. Third, the lack of validated scales for the objective measurements of disability and DBS-associated clinical benefits limits the generalizability of the results.

Given these limitations, particular efforts ought to be capturing the entirety of the patient’s clinical picture, including identification of co-existing movement disorders. Other variables to consider may include, but are not limited to, relevant genotyping, especially in syndromes that have been not well characterized, selection of DBS target(s), and ascertainment of optimal stimulation settings. Recent and future advances in DBS technology, such as directional stimulation, lead with more contacts, and adaptive stimulation stands to improve outcomes when using DBS for uncommon tremor disorders. An international registry to gather data from DBS-treated patients with uncommon tremor disorders is warranted to clarify selection criteria, long-term results, and optimal surgical targets.