Tremor and disability outcomes following deep brain stimulation for multiple sclerosis-associated tremor

Abstract

Background:

Tremor in multiple sclerosis (MS) is often refractory to medications. Deep brain stimulation (DBS) may help, but possible side effects and disease exacerbation limit its use.

Objectives:

Evaluate outcomes of DBS for MS-associated tremor.

Methods:

Retrospective review of 18 patients with MS-associated tremor who underwent DBS, analyzing Expanded Disability Status Scale (EDSS) and Fahn–Tolosa–Marin (FTM) tremor scores.

Results:

Median age at surgery was 49.5 years (range, 28–75), with postoperative follow-up 2.5 years (range, 0–22). No relapses (0/18) or new lesions (0/8) occurred postoperatively. All had ventral intermediate nucleus leads; other targets were not used in programming. EDSS remained stable (median 6 to 6.25; p = 0.2). FTM scores improved (median 14.5 to 10, p = 0.008). Patients with EDSS 3.5–6 had variable outcomes (p = 0.02).

Conclusions:

DBS is a potential treatment for severe MS tremor and did not provoke relapse. Targeting VIM alone may optimize tremor control and maintain magnetic resonance imaging (MRI) access.

Introduction

Tremor is common in MS, occurring in about one quarter of patients,^1,2 and correlating with greater disability and impaired quality of life.³ MS-related tremor can reduce independence and mobility. MS-associated tremor is most commonly cerebellar type, with postural and kinetic components,⁴ though Holmes tremors may be encountered.^5,6 Many classes of pharmacologic therapies have been tried for MS-related tremor^7,8 with inconsistent success and dose limiting side effects. Surgical interventions for patients with refractory tremor have been investigated. Lesional options such as thalamotomy have had little long-term success with many patients having recurrence of tremor as early as 3 months,^9,10 and increased risk of secondary neurologic impairment due to surgical side effects.¹¹ Deep brain stimulation (DBS) remains an option for patients with severe and medication refractory tremor. A recent meta-analysis demonstrated approximately 75% of patients with MS benefited from DBS for MS-associated tremor, however methodological variability across studies limits conclusions regarding optimal patient selection and surgical targets.¹² Given the lack of standardized selection criteria and the need for long-term outcome data, we evaluated the efficacy of DBS for MS-associated tremor, and analyzed factors associated with optimal outcomes.

Methods

Patient identification and consent

The Mayo Clinic Institutional Review Board approved this study, and patients or their representatives provided written informed consent for use of their deidentified medical information. We identified 18 patients from January 1, 1998, to August 31, 2023, with a diagnosis of MS and tremor. We retrospectively evaluated charts for inclusion criteria of (1) Diagnosis of MS according to 2017 McDonald criteria¹³ and (2) DBS for tremor at Mayo Clinic Rochester. Some patients were described in prior reports.^6,9

DBS patient selection

Prior to DBS all patients were evaluated by a movement disorders DBS trained neurologist to confirm an exam consistent with disabling tremor, rather than only ataxia. In addition, upper limb surface EMG and accelerometry recordings were performed in our clinical movement disorders laboratory with arms held in posture and during finger-to-nose testing. The presence of rhythmic tremor was confirmed by analyzing regularity of burst amplitudes and burst-to-burst intervals and confirming a sharp peak on the power spectrum. To exclude functional tremor, the presence of distractibility and/or entrainment was determined by monitoring for amplitude reduction or frequency shift of the postural tremor on the limb with maximum tremor while the patient performed a paced tapping task with the contralateral hand at multiple frequencies. Patients were also seen by a functional neurosurgeon and neuropsychologist. Risks and benefits of DBS surgery were discussed at a multidisciplinary consensus conference.

Chart review

Records of patients were reviewed by an MS (S.A.B) and movement disorders (B.T.K.) neurologist. Tremor severity prior to surgery was self-reported using the Bain tremor severity scale (BTS; maximum 100).¹⁴

Outcome metrics

Outcomes evaluated included the retrospective¹⁵ Expanded Disability Status Scale (EDSS; maximum score 10)¹⁶ for overall disability and modified Fahn–Tolosa–Marin¹⁷ scale (FTM; excluding writing, drawing, pouring; maximum 76) for tremor severity measured by trained nurses. MS disease activity was defined by new T2 hyperintense or enhancing lesion on magnetic resonance imaging (MRI) of the brain or spine, or clinical relapse determined by treating neurologist.

DBS lead visualization

DBS lead placement was reconstructed using Lead-DBS.¹⁸ Preoperative MRI and postoperative CT imaging were co-registered using Advanced Normalization Tools (ANTs)¹⁹ and Statistical Parametric Mapping (SPM12),²⁰ with brain shift correction. DBS contacts were automatically prereconstructed using the PaCER method²¹ and manually refined as needed. Final reconstructions were normalized to MNI152 space,²² and segmented using the DISTAL atlas. Group visualizations were performed using the Lead group²³ toolbox.

Statistics

All statistical analyses and visualizations were performed using MATLAB (MathWorks, MA, USA). Nonparametric tests were applied where possible. Bonferroni correction was used with all post hoc tests. Statistical tests were two-tailed, with alpha of .05. Ranges are reported throughout the text as minimum to maximum values.

Results

Baseline clinical features

In total, 18 patients were identified who underwent DBS for MS-related tremor. Median age of onset for MS was 32 (range 21–65) and tremor was 41 (range 20–76). The cohort included patients with relapsing remitting (two), secondary progressive (eight), and primary progressive MS (eight). Twelve of the patients were on MS disease-modifying therapy (DMT) at any point; six were on DMT at the time of surgery (Table 1). Tremor was the initial symptom of MS in 3 (17%) and involved the dominant hand in 16 (89%). Median preoperative BTS for essential tremor (ET) was 63.5 (range 44–77). Patients tried a median of 2 (0–6) medications for tremor prior to surgery. Patients underwent surgery a median of 7 years (0–54) after tremor onset, at age of 49.5 (28–75). The cohort had overall follow-up from initial visit of 8 years (0–35), and 2.5 (0–22) years follow-up after surgery.

Table 1.

Demographics and clinical features.

Characteristic	Data (n = 18)
Demographics
Female sex	11 (61%)
Race	17 White, 1 African American
Median age MS onset (range)	32 years (21–65)
Median age tremor onset (range)	41 years (20–76)
Median age DBS surgery (range)	49.5 years (28–75)
Median tremor duration prior to DBS (years)	7 (0–54)
Median follow-up (range)	8 years (0–35)
Median follow-up after DBS (range)	2.5 years (0–22)
MS Clinical Characteristics
MS disease course	RRMS 2 (11%)
	SPMS 8 (44%)
	PPMS 8 (44%)
DMTs used during disease course	Any DMT used 12 (67%)
DMT at time of surgery	6 (33%)
	Ocrelizumab (2)
	Mitoxantrone (1)
	Teriflunomide (1)
	Dimethyl fumarate (1)
	Glatiramer acetate (1)
Tremor Clinical Characteristics
Tremor as initial symptom of MS	3 (17%)
Family history of ET	4 (22%)
Preoperative movement lab completed	16 (89%)
Tremor in dominant hand	16 (89%)
Tremor components	Rest 5 (28%)
	Postural 17 (94%)
	Action 18 (100%)
Median number of tremor medications trialed prior to DBS(range)	2 (0–6)*
MS-Related Outcomes
Clinical relapse after DBS	0
New MRI lesion after DBS	0/8
Median Preop EDSS (range)	6 (3–6.5)
Median EDSS at last follow-up (range)	6.25 (1–8)
EDSS worsened	8 (44%)
Subjective gait decline	11 (61%)
Subjective balance decline	11 (61%)
Subjective cognitive decline	1 (6%)
Subjective speech decline	8 (44%)
Immediate surgery-related worsening	1 (6%)

Key: DBS, deep brain stimulation; DMT, disease-modifying therapy; EDSS, Expanded Disability Status Scale, ET, essential tremor; MS, multiple sclerosis; PPMS, primary progressive multiple sclerosis; RRMS, relapsing remitting multiple sclerosis; SPMS, secondary progressive multiple sclerosis.

^*Treatments for tremor included: Propranolol (10), primidone (9), gabapentin (6), carbidopa levodopa (4), topiramate (3), carbamazepine (2), benzodiazepine (2), levetiracetam (1), metoprolol (1), baclofen (1).

DBS surgical targets

DBS leads were placed in ventral intermediate nucleus (VIM) of the thalamus in all patients (bilateral = 12, unilateral = 6). Additional leads were placed in the ventralis oralis anterior/posterior (VOA/VOP, n = 2) and in the VOA/VOP/subthalamic nucleus (STN, n = 1) regions (Figure 1(a) and (b)). These additional leads were not used in final stimulation parameters as they did not provide clinical benefit.

Figure 1. Deep brain stimulation (DBS) outcomes in Multiple Sclerosis (MS)-related tremor.

(a) Lead-DBS Reconstruction of final DBS lead placement for patients receiving bilateral ventral intermediate nucleus (VIM) leads. *One bilateral and all four unilateral VIM DBS patients did not have postoperative imaging for reconstruction. (b) Reconstruction of final DBS lead placement for patients with multiple target constructs, including bilateral VIM + ventralis oralis anterior/posterior (VOA/VOP; blue), right VIM + VOA/VOP/subthalamic nucleus (STN; purple), and left VIM + VOA/VOP (green). Only the VIM leads were ultimately selected for optimal programming settings with clinical benefit in all 3. (c) There was no significant change in expanded disability status scales (EDSS) between preoperative EDSS (median 6) and at last follow-up after DBS (median 6.25, p = 0.2). (d) DBS therapy significantly improved Fahn–Tolosa–Marin (FTM) tremor rating scores at last follow-up (median 14.5 to 10, p = 0.008). (e) The effect of DBS therapy significantly differed across disability subgroups (p = 0.02; EDSS < 3 median FTMs: pre = 6, first post = 5, latest = 3; 3 < EDSS < 6.5 median FTMs: pre = 13, first post = 10, latest = 11.5; EDSS > 6.5 median FTMs: pre = 18, first post = 16, latest = 12).

MS-related outcomes

There were no clinical relapses (0/18) or new lesions (0/8) at last follow-up (MRI follow-up median 44 months, range 6–97). MRIs were not able to be obtained in 3 patients due to dual lead placement in one hemisphere preventing MRI compatibility, which ultimately led to difficulty with MS management in these cases. EDSS did not change from pre (median 6, range 3–6.5) to postoperative (median 6.25, range 1–8; Wilcoxon signed rank; p = 0.2; Figure 1(c)). One patient had an early surgical complication with acute worsening of speech and gait related to lead placement. No patients developed infection, hemorrhage, or upper extremity weakness (Mayo strength grading system).²⁴

Tremor outcomes

Subjectively, 15 patients reported improved independence in activities of daily living such as eating and drinking, 1 with improved walking and returned work, while 2 reported no improvement. There was a significant reduction in FTM scores (n = 16) between the preoperative baseline (median 14.5 (range 4–24)) and at last follow-up (median 10 (range 1–16), Freedman repeated measure analysis of variance (ANOVA), p = 0.008, Table 2, Figure 1(d)).

Table 2.

Tremor outcomes.

Variable	Preoperative	First postoperative	Latest postoperative
Median months from surgery	4 (0–21)	3 (0–6)	24 (4–211)
Median Total Fahn–Tolosa–Marin Scale (range)*	13 (4–22)	10 (3–19)	11 (2–16)
Median Fahn-Tolosa Marin Scalc of most impaired limb (range)*	7 (2–12)	5 (2–12)	5 (1–9)

^*Serial Fahn–Tolosa–Marin Scale data available on 16 patients.

A two-way repeated measures ANOVA was used to examine the effects of baseline EDSS and timepoint of FTM score. There was an interaction between baseline EDSS and treatment timepoint (p = 0.02), suggesting that the changes in FTM scores over time with stimulation varied by disability subgroups. The patients with EDSS ⩽ 3 (mild) or ⩾ 6.5 (severe) had sustained tremor improvement, while the moderate group had variability as seen in Figure 1(e). There were no correlations between change in FTM score and other baseline characteristics including age, duration of MS or tremor, or preoperative BTS score (Spearman, p > 0.05).

Discussion

In this cohort of patients with MS with severe baseline tremor, most patients had a moderate sustained reduction in tremor after DBS surgery. The response to DBS was most sustained in those with mild (⩽3) or severe (⩾6.5) baseline EDSS. DBS surgery was not associated with new clinical or radiographic inflammatory disease activity in this group of patients with predominantly progressive MS.

Prior work outlining DBS outcomes for MS-related tremor has similarly shown that VIM targeted DBS provides a modest benefit in tremor severity^9,12,25–28 and quality of life.²⁶ One patient in our series had a surgical complication related to lead placement and none had infection, similar to prior studies of patients who underwent DBS for MS^12,29 or other indications.^30–32

The question of whether DBS surgery may provoke inflammatory relapsing MS disease activity arises in clinical practice, but there are no prior studies to our knowledge that detail this clearly with MRI outcomes. In our cohort of patients who mostly had progressive MS, there was no evidence of new MRI lesions nor clinical relapse following surgery.

Baseline disability may serve as a predictor of tremor outcome in pwMS, which to our knowledge is a novel observation. We found that patients with moderate gait impairment (EDSS 3.5–6) but without gait aid use had a more variable response to DBS. We hypothesize that this group had reduced reserve to tolerate gait ataxia, a recognized side effect of thalamic DBS.^33–38 In contrast, patients whose primary disability was from tremor without preexisting gait dysfunction (EDSS ⩽ 3) tolerated higher stimulation settings and generally achieved better tremor control. Finally, patients with severe disability who required a mobility device preoperatively (⩾ 6.5) were often able to compensate for ataxia using their assistive devices.

In addition to VIM leads, 3 of the patients in this cohort had additional leads placed to complement VIM leads for symptom control, while potentially mitigating risk of tolerance or cerebellar side effects.³⁸ However, these additional leads did not add additional benefit. There has been one small, randomized trial evaluating VIM and VOP stimulation in MS.²⁹ DBS led to sustained tremor improvement in 8 of 11 at 6 months, without a clear difference between VIM, VOP, and dual lead stimulation. Notably, according to Food and Drug Administration (FDA) labeling, patients with a single electrode per hemisphere may undergo MRI. As demonstrated in this cohort, most patients needed an MRI for clinical care after DBS surgery, however dual lead placement in 1 hemisphere prevented this in 3. Therefore, with the current evidence that additional leads lack clear clinical benefit, and that these patients were not able to obtain MRIs for MS care, we now favor single VIM lead per hemisphere for these patients as we cannot reliably predict which patients would not need MRIs in the future.

Multidisciplinary decision-making involving the movement and MS neurologists and experienced neurosurgeons should take place to evaluate potential candidates with severe refractory MS-related tremor. Nuances in this population include possible infection risk in patients on DMT, the need for postoperative MRIs to monitor MS activity, and the potential for worsening ataxia or cognition. Further work is needed to better understand the potential impact on progressive disease activity in this population. With the advent of high efficacy MS DMTs and new DBS technology, continued work is needed in this field.

This study has limitations due to its single center retrospective cohort design. Arm strength was assessed retrospectively using the Mayo strength grading system, a qualitative clinical scale; future studies should incorporate standardized quantitative measures of upper extremity function and strength to better delineate the relationship between tremor and weakness. In addition, studies should prioritize patient reported quality-of-life measures, applying standardized functional assessment for tremor (such as complete Fahn’s Tremor Rating Scale or quantitative electromyography), and assessing gait and cognition to better refine patient selection. Finally, larger datasets will be needed to evaluate treatment response across different tremor subtypes.

In conclusion, DBS targeting the VIM offered a reduction in tremor severity and improved quality of life for most people with MS, without associated relapsing MS activity. Outcomes varied by baseline disability, emphasizing the need for individualized treatment strategies. These findings reinforce the role of DBS as a therapeutic option for refractory tremor in MS.

Data Availability Statement

The data presented in this study are available on request from the corresponding author. The data are not publicly available.