Essential tremor (ET), a common movement disorder affecting 4.6% of individuals ≥65 years of age [1
], is characterized by uncontrollable trembling of the upper limbs [2
]. Other areas, including the head, face, jaw, and vocal cords, can also be affected [2
]. ET can significantly impair activities of daily living (ADL) [3
Treatment is based on severity and impact on the quality of life [5
]. Propranolol and primidone, the most commonly used therapies, reduce tremor amplitude by ~50% [2
]. However, approximately 30–70% of patients fail to respond, and side effects are relatively common [2
Two placebo-controlled studies evaluated botulinum neurotoxin type A (BoNT-A; onabotulinumtoxinA) in patients with hand ET [7
]. The first study randomized 133 patients to onabotulinumtoxinA 50 U, 100 U or placebo injected into wrist muscles (flexor carpi radialis and ulnaris, extensor carpi radialis and ulnaris) for 16 weeks [7
]. The authors reported a significant reduction in postural tremor with onabotulinumtoxinA versus placebo at 6, 12 and 16 weeks and in kinetic tremor at 6 weeks using a tremor severity rating scale; however, grip strength significantly reduced with onabotulinumtoxinA versus placebo, hand weakness was reported as an adverse event by half of the patients overall, and only minimal changes in motor function and functional disability occurred. The second study randomized 25 patients to onabotulinumtoxinA 50 U or placebo into the same wrist muscles as the first study for 16 weeks [8
]. The authors reported a significant reduction in postural tremor by Week 4, which was maintained for the duration of the study; similar to the first study, finger weakness was reported by up to half of patients receiving onabotulinumtoxinA versus none receiving placebo, and functional rating scales did not significantly improve with onabotulinumtoxinA. Due to concerns of increased risk of hand or finger muscle weakness [7
], BoNT-A has not been widely adopted for ET. However, there is growing evidence that a more customized selection of muscles and doses reduces muscle weakness and improves activities of daily living and quality of life of patients with ET [9
Characterization of movement is important for selecting BoNT-A dose and the muscles to be injected, and is difficult to define in the upper limb due to complex joint biomechanics [13
]. Kinematic analysis to record the dynamics of movement is well established for analysis of upper-limb movement [14
] and has demonstrated benefits in improving the outcome of BoNT-A treatment for upper-limb Parkinson’s disease tremor [13
]. Two open-label single-site studies of incobotulinumtoxinA were conducted in patients with ET [12
]. The first study enrolled 31 patients to receive three incobotulinumtoxinA injections over 30 weeks and captured upper limb kinematics using goniometers and a torsiometer at the forearm, wrist, elbow and shoulder joints; the computerized tremor analysis was used to aid and plan a customized dosing algorithm [12
]. Tremor amplitude was significantly reduced as early as 6 weeks and maintained throughout the treatment course, with only mild but not bothersome hand weakness and a reduction in grip strength that did not substantially impact hand function [12
]. The second study enrolled 24 patients to receive up to six incobotulinumtoxinA treatments for up to 96 weeks [17
]. Motion sensor devices on the forearm, wrist, elbow and shoulder captured tremor severity, and were used to select the most appropriate muscles for injection. A significant decline in tremor severity was reported as early as 6 weeks and maintained throughout the study course; perceived muscle weakness was not long-lasting, and the vast majority of patients experienced a functional benefit from treatment [17
]. The present study investigated whether this promising approach was reproducible under placebo-controlled conditions in a multicenter setting, with the objective of assessing the efficacy and safety of a single, kinematic planned intramuscular injection of incobotulinumtoxinA, compared with placebo, in moderate-to-marked ET of the upper limb.
In this randomized, double-blind, placebo-controlled clinical trial in patients with moderate-to-marked ET of the upper limb, all efficacy variables consistently favored the incobotulinumtoxinA group after a single treatment cycle, until at least Week 12. Kinematic planned incobotulinumtoxinA treatment statistically significantly and persistently decreased tremor severity for up to 24 weeks post-injection, with greater improvement in physician-rated GICS versus placebo 4 weeks post-injection. IncobotulinumtoxinA-treated patients experienced a slight and transient reduction in maximal grip strength versus placebo, but no significant effect on self-perceived muscle weakness. IncobotulinumtoxinA conferred a statistically significant improvement in FTM part B score for motor function at Weeks 4 and 8 versus placebo, suggesting that the observed transient reduction in maximal grip strength did not inhibit the patients’ ability to perform coordinated motor tasks. Notably, although greater improvements in FTM tremor score were observed with incobotulinumtoxinA versus placebo at Weeks 4 and 8, these were not statistically significant at Week 4. The contrast between FTM tremor and Part B motor performance scores may have been due to the tremor score being visually assessed and subjective, which causes crude variabilities, while the motor performance score is a more reliable measure of improvement. Results from this multicenter study highlight the reproducibility of this technique and are consistent with those of a recent open-label study, in which kinematic planned incobotulinumtoxinA injections for ET resulted in a reduction in tremor amplitude and significant improvement in arm function, despite mild, yet non-bothersome, weakness in the treated muscles [17
The effective treatment of ET remains an unmet clinical need [5
]. Although propranolol and primidone have shown efficacy in ET, many patients fail to respond, and side effects are common [6
]. Consistent with this, the inclusion criteria for the current study required that patients had a moderate-to-marked tremor and any patients taking concomitant anti-tremor medication received a stable dose, from at least 4 weeks prior to the screening visit until the study end. The high proportion of patients (93.3%) on concomitant beta-blocking agents and anti-epileptics while still meeting the inclusion criteria of moderate-to-marked tremor, indicates that the majority had an inadequate therapeutic effect with these therapies prior to the study.
Based on recent evidence, it has been postulated that the symptoms of ET may be part of degenerative brain pathophysiology characterized by structural changes to the cerebellum, particularly the Purkinje cell population [21
]. Side effects of currently available oral therapies may result in a reduction of cerebral performance, including confusion and cognitive difficulties [2
]. Therefore, a localized treatment, such as BoNT, may provide an alternative to a systemic oral approach, or too invasive, complication-prone treatment such as deep brain stimulation.
Two previous studies followed a standard non-personalized protocol for the injection of wrist muscles with onabotulinumtoxinA. One study injected onabotulinumtoxinA 15 U or 30 U into each of the flexor carpi radialis and ulnaris and 10 U or 20 U into each of the extensor carpi radialis and ulnaris. Significant improvements from baseline in postural and kinetic tremor at 6 weeks were reported, with significant reduction in grip strength [7
]. The second study injected onabotulinumtoxinA 15 U into each flexor carpi radialis and flexor carpi ulnaris and 10 U into each extensor carpi radialis and extensor carpi ulnaris. The authors reported improvements in postural tremor and treatment response ratings, but no significant improvement in functional score with BoNT-A versus placebo [8
]. In both studies, the extent of the improvement was limited by the development of functional impairment caused by disproportionate wrist and hand weakness [7
]. In one study, depending on the administered BoNT-A dosage, up to 70% of patients experienced weakness in the fingers and/or wrist, as measured by maximum grip strength and MRC clinical ratings, which impacted ADL in some patients [7
In light of this evidence, the American Academy of Neurology recommended that the benefits of BoNT-A should be considered in the context of the “common adverse effect of muscle weakness” and that BoNT should be considered as a treatment option for ET of the hand in those patients who fail to respond to oral therapy [22
]. Results of a more recent single injection placebo-controlled study suggest that personalized injections with a broader spectrum of injected muscles improves hand weakness [9
]. However, selection of the tremor-causing muscles remains challenging. While Mittal et al. used electromyography to identify rhythmic burst potentials, the kinematic analyses for each plane of motion used in this study appears to provide a standardized and reproducible approach with additional help to aid the injector in the customization of dose selection for individual muscles [12
]. Furthermore, the customized protocol used in this study, combined with the significantly reduced MRC MMT scores at Week 4, indicate that this protocol may overcome the concerns seen in previous trials using a standard protocol.
In the current study, customized dosing patterns on kinematic tremor analysis was well tolerated: only two patients (10.5%) in the incobotulinumtoxinA group experienced muscular weakness localized to the fingers of the injected limb. No patient discontinued the study due to muscle weakness.
Furthermore, although the effect of muscle weakness on ADL was not assessed in this study, mean MRC MMT scores in both the incobotulinumtoxinA and placebo groups ranged between 24.1 and 25.0 at all post-injection visits. As a maximum score of 25 on the MRC MMT scale indicates normal muscle strength, these scores demonstrate overall good muscle strength, which one may assume had little or no impact on ADL. The results suggest that customized incobotulinumtoxinA injections can be a promising treatment for ET with a lower rate of side effects than shown in previous fixed-dose studies.
This study demonstrated the efficacy and tolerability of incobotulinumtoxinA for the treatment of ET and the feasibility of using kinematic analysis at multiple sites. While the study was exploratory in nature, lacked a primary endpoint and was not powered to demonstrate the superiority of incobotulinumtoxinA versus placebo, statistically significant improvements in efficacy were shown with incobotulinumtoxinA; these findings would need to be confirmed in larger studies. A further limitation was the single treatment cycle. Measurements over additional treatment cycles would allow better estimation of long-term safety and may demonstrate enhanced treatment efficacy, as shown previously [17
]. Using such an approach in a recent open-label pilot study, a continuous trend of further improvements in FTM scales was noted up to 96 weeks when follow-up treatments were further optimized with kinematic based outcome measures [18
5. Materials and Methods
5.1. Study Design and Patients
This trial was a prospective, randomized, double-blind, placebo-controlled, parallel-group, multicenter, exploratory study in patients with upper-limb ET (NCT02207946). Patients were enrolled from December 2014 to April 2016 by six investigators in Canada and the United States (US) (at Medicine Professional Corporation, Toronto, ON, Canada; Movement Disorders Clinic, Toronto Western Hospital, Toronto, ON, Canada; Movement Disorder Clinic, Winnipeg, MB, Canada; David King, Inc., Halifax, NS, Canada; Mount Sinai Medical Center, New York, NY, USA).
Eligible patients were adults with a first onset of ET ≥6 months before screening and with stability of symptoms over 4 weeks; whose diagnosis of “definite ET” was otherwise in accordance with the Tremor Investigation Group criteria applicable at the time [23
] (bilateral postural tremor with/without kinetic tremor, involving hands and forearms, that was visible and persistent); moderate-to-marked upper-limb postural and/or kinetic tremor at wrist level (FTM Part C, items 17–23) in the limb to be treated, with a rating of ≥2 in at least two categories; visible tremor at wrist level in at least one of the four positions/tasks used in kinematic assessment; tremor deemed by the investigator to require treatment with a total dose of 30–200 U incobotulinumtoxinA for up to 3 joints (wrist joint mandatory; shoulder and elbow joints optional); and receiving concomitant anti-tremor medication, if any, at a stable dose (≥4 weeks before screening and until the study end). Principal exclusion criteria are listed in the supplementary materials
Patients were randomized 2:1 to receive, in a single treatment cycle, unilateral, intramuscular injections of incobotulinumtoxinA (total dose 30–200 U), or an equivalent volume of placebo, into muscles of the wrist (mandatory) and, optionally, into the elbow and/or shoulder, of the limb that (in the patient’s opinion) had the greatest impact on their ADL (the doses received are shown in Table 2
). Injections were guided using ultrasound, electromyography, and/or electrical nerve stimulation as determined by the investigator. Injections were conducted by experienced investigators familiar with the Pictorial Atlas of Botulinum Toxin Injection: Dosage, Localization, Application by Jost [24
The treatment planning for muscles to be treated was based on kinematic tremor analysis using the tremor kinematic analytics investigational device, and followed kinematic tremor procedures as previously described [17
] (this technology is pending regulatory approval, and is available to clinical investigators for research purposes from the corresponding author). The customized treatment pattern for total dose and dosing regimen was based on kinematic tremor analysis; however, this clinical decision algorithm could be modified at the investigator’s discretion. Patients were monitored regularly over 24 weeks, including telephone contact 1 week post-injection and clinic visits every 4 weeks until the end-of-study visit at Week 24.
5.3. Efficacy Assessments
No primary efficacy variables were defined in this exploratory study. Secondary efficacy variables included change from baseline to Week 4 in: maximum angular tremor amplitude of the wrist of the injected limb; maximum log-transformed accelerometric tremor amplitude at hand level of the injected limb; FTM tremor score in the injected limb [19
]; FTM motor performance score [19
]; and physicians’ and patients’ GICS score at Week 4. Changes from baseline to later time points from Weeks 8–24 were assessed every 4 weeks as further efficacy variables.
5.4. Kinematic Tremor Analysis
The maximum log-transformed accelerometric tremor amplitude at the hand and the RMS angular tremor amplitude at the shoulder, elbow, forearm, and wrist of the injected limb were measured by kinematic tremor analysis at the screening and baseline visits, and every 4 weeks from Week 4 to Week 24 post-injection. The analysis was conducted as described previously [16
5.4.1. FTM Tremor Rating Scale
Tremor severity and functional effects were assessed by the investigator using the FTM tremor rating scale [19
] at screening and baseline visits, and Weeks 4, 8, 12, and 24 post-injection. The change from baseline in FTM tremor score for the injected limb was assessed using FTM Part A items 5 or 6 (right or left upper limb, respectively) for three functions (at rest, with posture holding, with action and intention). The scale ranged from 0 (normal) to 4 (severe, amplitude > 4 cm) for each item and from 0 (normal) to 12 (severe) for the sum score of all three functions.
The change from baseline in FTM motor performance score was calculated from the sum of FTM Part B items 11 (handwriting for the dominant hand, irrespective of treatment) and 12–15 (drawing a large/small spiral, drawing a line and pouring for the treated limb only). The scale ranged from 0 (normal) to 4 (severe) for each item and from 0 (normal) to 20 (severe) for the sum of all five motor tasks.
5.4.2. Global Impression of Change
The physicians’ and patients’ GIC were assessed using the GICS, a 7-point Likert scale [25
] from −3 (very much worse) to +3 (very much improved) every 4 weeks from Week 4 to Week 24 post-injection.
5.5. Safety Assessments
5.5.1. Adverse Events
Adverse events were monitored, and predefined adverse events of special interest thought to be a possible indication of toxin spread, based on Medical Dictionary for Regulatory Activities-preferred terms (version 19.0), were monitored by targeted questioning. TEAEs, including those of special interest, were analyzed. These were defined as adverse events with onset or worsening at/after date and time of the first administration of incobotulinumtoxinA.
5.5.2. Other Safety Assessments
Maximum grip strength (kg) was measured for both hands using a hand-held dynamometer. Details of additional safety assessments (MRC MMT and self-perceived weakness) are presented in the supplementary materials
5.6. Statistical Analysis
No formal sample size calculation was performed, and all analyses were exploratory. Two-sided 95% CIs (based on the t-distribution) were calculated for the change from baseline and for treatment differences for efficacy variables. The treatment groups were compared using sensitivity ANCOVA, mixed model repeated measures (MMRM), and the Wilcoxon rank-sum test, using observed cases. Descriptive p-values from ANCOVA, MMRM, and the Wilcoxon rank-sum test were provided as appropriate.
Efficacy analyses were based on the full analysis set (all treated patients with a post-baseline value for any efficacy variable) and on the per-protocol population (patients in the full analysis set, for whom no major protocol violations were recorded) for the sensitivity analysis. Safety data were based on the safety evaluation set (all patients who received study medication) and summarized descriptively. Statistical analyses were performed using SAS® version 9.2 (SAS Institute, Cary, NC, USA).
5.7. Standard Protocol Approvals, Registrations, and Patient Consents
The study was conducted in accordance with the International Conference on Harmonization Guidelines, applicable regulations, and the principles of the Declaration of Helsinki. The study protocol was reviewed by an independent ethics committee or institutional review board at each study site. The study was approved by the: Western Institutional Review Board, Puyallup, WA, Canada; University Health Network Research Ethics Board, Toronto, Canada; University of Manitoba Bannatyne Campus Research Ethics Board, Manitoba, Canada; and BRANT IRB Accreditation Consulting Services, Lake Success, New York, NY, USA (further details in Table S5
). All patients provided written informed consent.