Botulinum Toxin in the Treatment of Cervical Dystonia: Evidence-Based Review

Cervical dystonia is the most common form of dystonia encountered in a movement disorders clinic. Botulinum toxin has been a long-established first line therapy. Several studies, including nearly two dozen randomized clinical trials, have shown that botulinum toxin is safe and effective in reducing the clinical severity of cervical dystonia. Longitudinal data have demonstrated decades of sustained benefit and safety. Although there is a potential for the development of botulinum toxin immunoresistance, this is quite rare, and partly determined by frequency of administration, cumulative dosage, and properties of the injected product. When immunoresistance does occur, switching to an alternative type of botulinum toxin (e.g., from type A to type B) usually restores the efficacy. In this evidence-based review we highlight the results of published double blind, placebo-controlled studies. We also briefly discuss injection techniques and some unmet needs, such as the development of practical assays to detect immunoresistance and longer-acting formulations of botulinum toxin.


INTRODUCTION
Botulinum toxin (BoNT) is a neurotoxin derived from the anaerobic bacterium, Clostridium botulinum (1)(2)(3).It is one of the most potent toxins found in nature, causing clinical botulism, a condition that leads to widespread paralysis and ultimately diaphragmatic respiratory failure (1,4,5).The toxin acts at the presynaptic nerve terminal by cleaving soluble N-ethylmaleimide-sensitive factor activating receptor (SNARE) proteins involved in docking of acetylcholine vesicles with the presynaptic membrane, thus preventing the release of acetylcholine leading to loss of muscle contraction, manifested by focal weakness when injected locally or generalized weakness, including respiratory, paralysis, when ingested in contaminated food (1,(6)(7)(8).There is also growing evidence that BoNT has a central effect, as well, by impacting antero-and retrograde transportation (9)(10)(11)(12).
Capitalizing on the ability of BoNT to produce weakness by modulating the release of acetylcholine, this most potent biologic toxin known to man has emerged as one of the most multipurposed treatments for a large variety of neurologic and non-neurologic disorders (1,7).In fact, Justinius Kerner, a German physician who initially described the effects of BoNT in 1817 suggested that it may have a therapeutic value in a variety of conditions such as St. Vitus dance, hypersalivation, and hyperhidrosis (1,8).BoNT did not receive approval for human use by the United States Food and Drug Administration (FDA) until 1989, nearly a decade after the first publication of a double-blind, placebo-controlled trial establishing safety and efficacy of BoNT in cranio-cervical dystonia (13,14).Since that time, BoNT has been approved for over a dozen indications including blepharospasm, hemifacial spasm, spasticity, detrusor overactivity, chronic migraines, hyperhidrosis, sialorrhea, and a variety of cosmetic indications (1,15).In this review, we will focus on the use of BoNT in cervical dystonia (CD).

METHODS OF LITERATURE SEARCH
A comprehensive literature review of PubMed database was conducted with search criteria including the following key words: botulinum toxin treatment, cervical dystonia, randomized controlled trials (RCTs).Filters were set to RCTs between the years of 1980-2022.We excluded articles not published in the English language and those which were based on non-human subjects.The query resulted in 66 articles.Those were further reviewed to include both safety and efficacy data.Any ongoing trials or those that were beyond the scope of BoNT in treatment of CD were excluded.

Botulinum Toxin Products
Before reviewing the findings from the published studies of treatment with BoNT in CD, we wish to briefly describe the pharmacology and properties of the different formulations of BoNT as this is critical to the interpretation of the published results.Structurally, BoNT is composed 100 kD heavy chain and 50 kD light chain, linked by a disulfide bond (16).The heavy chain binds the complex to the presynaptic membrane and the light chain contains the active SNARE protein that cleaves either the intracellular synaptosomal-associated protein 25 (SNAP25) in the case of BoNT type A and E, or vesicle-associated membrane protein (VAMP), in the case of BoNT types B, D, and F (3).Some formulations of BoNT also contain non-toxic accessory proteins (NAPs) that include hemagglutinin and non-hemagglutinin, designed to prevent degradation (5,(17)(18)(19).The different forms of BoNT vary in the amount and types of NAPs and unique excipients (8,16,18).
Though most clinical trials have utilized BoNT type A and B, there are eight immunologically distinct forms of BoNT (A-H).There are four different preparations of BoNT that have been approved by the FDA: BoNTA onabotulinumtoxinA (Botox ® ), abobotulinumtoxinA (Dysport ® ), incobotulinumtoxinA (Xeomin ® ), and BoNTB rimabotulinumtoxinB (Myobloc ® ) (1,6,8).Other, novel, formulations are pending approval or in clinical trial, including a promising formulation of BoNTA called daxibotulinumtoxinA.In a phase 2 trial, which enrolled approximately 37 patients, daxibotulinumtoxinA was found to be safe and effective in the treatment of CD (20,21).Results of the follow-up phase 3 clinical trial, with 301 patients, ASPEN-1, are pending full publication but have been presented at international meetings (22).
Generally, the effects of BoNT become evident within 1 week after injection, last 3-4 months, and then the benefits gradually wane.It has been proposed that the relatively short-lived effects are due to axonal sprouting at the presynaptic nerve terminal after injection with return of neuromuscular junction function that correlates with re-establishment of normal (baseline) muscle strength (1,23,24).With newer derivatives, such as daxibotulinumtoxin A, the effects have been found to last an average of 24 weeks, thus potentially reducing the number of injections needed per year to maintain optimal and sustained benefits.This may address a major concern for many patients-the wearing off of benefits before the next injection (25).

Immunogenicity
One of the potential risks associated with long-term BoNT use is the emergence of immunoresistance associated with the development of neutralizing antibodies (NABs) (1,5,21,26).Several factors account for the immunogenicity of BoNT, including the unique properties of the formulation (some BoNT preparation have more accessory and stabilizing proteins that are inherently immunogenic), individual dose load and short inter-injection intervals (1,5,17,(26)(27)(28)(29).The overall risk of immunogenicity varies among available products, ranging from 0%, as reported with incobotulinumtoxinA, to roughly 42% for rimabotulinumtoxinB (1,5).BoNT resistance also varies by indication (5).The frequency of NABs is relatively low, but highest levels of NABs have been reported in patients treated for CD (30), followed by blepharospasm, hemifacial spasm and other conditions that generally require relatively low doses of BoNT, supporting the notion that immunogenicity is linked to total amount of toxin injected (5).AbobotulinumtoxinA (26) and especially, rimobotulinumtoxinB, having the highest risk of immunogenicity (18).
The possibility of immunoresistance should be considered when patients have at least two to three consecutive treatments without at least a 25% improvement.One of the biggest unmet needs in BoNT therapeutics is the lack of reliable assays to measure NABs.Currently available assays include the mouse protection assay (MPA) and mouse hemidiaphragm assay (MHDA) but these are difficult to perform and require sacrificing animals (1,5).The unilateral brow injection is a good surrogate test for immunogenicity.It involves an injection of 20 U of BoNTA or 1,000 U of BoNTB in the medial eyebrow with assessment of ipsilateral procerus and corrugator muscle function at 1-2 weeks (1,5).Other clinical tests for immunoresistance exist, but their discussion is beyond the scope of this article.

Cervical Dystonia
CD (also referred to as spasmodic torticollis) is a form of focal dystonia manifested by abnormal postures or movements produced by involuntary contractions of muscles in the neck, often associated with tremor (31)(32)(33)(34) and pain (1,6).In nearly a third of the cases of CD, regions outside of the neck are also involved with dystonia (35).Dystonia has an estimated prevalence of about 40 per 100,000 persons (36) and CD about 10 per 100,000, with an average age at onset of 40 years (1,(36)(37)(38)(39)(40).In isolated (primary) CD, about 12% of cases have a family history of dystonia, tremor or both (41,42).There are many secondary causes of CD including neck injury (43).Although most patients have persistent, chronic symptoms, up to 20% experience spontaneous, but typically transient, remission, especially in younger individuals (1,44).Patients with CD assume different combinations of head postures such as torticollis, laterocollis, retrocollis, and anterocollis, also categorized according to the "Col-Cap concept" (21,45).Patients with predominant anterocollis or retrocollis are often excluded from clinical trials of BoNT because this form of CD is more challenging to treat (21).In these patients using electromyography (EMG) or ultrasound to guide the injection may improve muscle targeting and outcomes (1).An interesting phenomenon in CD, as in other forms of dystonia, is the alleviating maneuver, also referred to as "sensory trick" or "geste antogniste" (31,46,47).It is used in up to 90% of patients to reduce their dystonic posture and movement (46) which often complicates selection of muscles for injection with BoNT (46,48).Alleviating maneuvers may not only "correct" abnormal postures but also can reduce dystonic head tremor (31,49,50).For example, arm raising has been found to effectively reduce head tremor in CD (31).Besides awareness of alleviating maneuvers, the clinician must be able to differentiate between contractions of the primary (agonist) versus compensatory (antagonistic) muscles in order to select the most appropriate muscles for targeting with BoNT (46,47,51).
The assessment of BoNT efficacy largely relies on clinical rating scales, such as the Toronto Western Spasmodic Torticollis Rating Scale (TWSTRS) and Tsui score (52)(53)(54).Of the two, TWSTRS is the most frequently used and validated scale (1,54).However, the TWSTRS scale is not without limitations, particularly when used in clinically mild CD (31,46).Another limitation of TWSTRS is that it does not assess head tremor (55).Several studies have determined that the minimal clinically important change for improvement is 8-12-point reduction in total TWSTRS score (56).

RESULTS
Several RCTs and long-term observational studies (57) have demonstrated the efficacy of BoNTA in the treatment of CD (1,8).In this review we will focus on RCTs (Table 1) (58-82), although we acknowledge that placebo-controlled trials have important limitations, such as short duration of treatment, placebo effect, inflexible designs in which the therapy is not individualized, and other problems (83)(84)(85).Furthermore, open-label and observational longitudinal studies have provided important information besides efficacy, such as dosing and long-term safety (86)(87)(88)(89)(90).
While most studies involve BoNTA, trials assessing the safety and efficacy of BoNTB have found similar results (1,70,73,75,78,90).There are few head-to-head trials comparing various formulations of BoNT.One study, involving 40 patients with CD, found no difference in efficacy between incobotulinumtoxinA and onabotulinumtoxinA at a 1:1 unit dose ratio (91).Another RCT, involving 103 patients, showed no significant difference between abobotulinumtoxinA and onabotulinumtoxinA at a dose ratio of 2.5:1 units (77).Two studies compared BoNTA with BoNTB at a dose ratios of 1 U BoNTA to 40 U of BoNTB (69) and at 1:66.7 (79), respectively, showed similar clinical efficacy and safety, however the frequency of dry mouth and injection pain was higher with BoNTB and the clinical effect has been found to last longer with BoNTA treatment in the study using a 1: 40 ratio (69).
A Cochrane review conducted in 2016 that included three RCTs found no difference between the two types of BoNT but BoNTB had a higher frequency of dry mouth (92).Several other Cochrane reviews have demonstrated the efficacy of both, BoNTA and BoNTB, in the treatment of CD (93,94), as well as the superiority of BoNTA to traditional oral therapies, such as trihexyphenidyl (95).Several investigators have reported higher risk of immunogenicity with BoNTB compared to BoNTA (96).Furthermore, treatment failure is more frequent with BoNTB compared to other formulations, likely due to increased immunogenicity, though the mechanism remains unclear (5,18,21).
Dysphagia is the most frequently cited adverse effect of BoNT, reported in 5%-42% of patients, followed by muscle weakness in 3%-4%, injection pain 1%-9%, generalized weakness 0.3%, speech difficulties 0.3%, head drop 0.3%, rigidity 0.3%-3%, weight loss 0.3%, xerostomia (or dry mouth) 56%-71% (21).One side effect that it relatively common but rarely reported is flu-like symptoms occurring during the first few days after the injection in about 20% of patients (106).The mechanism of this transient adverse effect is not well understood but one study showed correlation with the interleukin inducible protein 10 (106).
Although BoNT has clearly improved the quality of life of patients with CD, up to 30% of patients discontinue treatment (108).The main reasons for discontinuation of therapy include change in provider, comorbid disease limiting treatment, advanced age, spontaneous improvement, adverse side effects and treatment failure (102,108).The latter accounted for up to a 39% dropout rate from one study.Other studies have reported up to 26.1% drop out rate for primary, and 13.4% drop out rate for secondary nonresponders (21).In other studies, common causes of discontinuation of therapy include a lack of response, short duration of effect, as well as inconvenience and cost of repetitive treatments, usually up to 4 times a year (109,110).In a retrospective study of 568 patients treated with abobotulinumtoxinA for an average of 13 years, about 1.9% of patients were found to have partial secondary treatment failure per year, amounting to 14.5% of patients at 9 years (111).Factors contributing to secondary nonresponse include prior surgery to the area, prior side effects from BoNT, high mean doses of BoNT, and immunoresistance (5,112).One of the major limitations of BoNT is its relatively short duration (about 2-4 months) of benefit.In a survey of 209 patients, 88% reported re-emergence of symptoms in between treatments at an average of 10.5 weeks from injection (25).Generally, patients prefer treatments with longer injection intervals, which happens to be the best prognostic factor in overall satisfaction in symptoms control (25,40).For these and other reasons, efforts at producing more effective and long-lasting formulations have been a priority in experimental therapeutics with BoNT in CD (20,21).
Though most studies use primary endpoints of improvement in clinical rating scales, few evaluate for patient satisfaction or quality of life measures as surrogates for treatment efficacy.This has been highlighted by others, creating a need for improved rating systems that consider the complexity of achieving desirable results with minimal side effects (108).
Clinicians must always strive to optimize the response to BoNT by recognizing the full phenomenology of the patient's CD and by making certain that they target the most relevant (agonist) muscles and avoid injecting compensatory (antagonist) muscles.Although EMG, ultrasound, kinematic guidance and other techniques have been found to improve accuracy of BoNT injections and possibly reduce the risk of BoNT-related side effects (45,(113)(114)(115)(116)(117), these techniques add cost, time, and discomfort and, therefore, should be used in selected cases rather than applied routinely to all BoNT injections (8,38,116).
Adjunctive treatment has an important role in overall patient satisfaction.Several studies have shown that anticholinergic drugs, baclofen, benzodiazepines and other muscles relaxants provide ancillary benefits in patients with CD treated with BoNT (42).Treating comorbid depression, anxiety, pain, fatigue and addressing stigmatization can all be therapeutic in the treatment of CD (108,118).In one RCT studying the effects of physical therapy on CD, there was an improvement in TWSTRS score severity and pain by 31% in patients who received physical therapy in addition to BoNT injections versus 28% in those who only received BoNT (119).Though this study shows only marginal improvement, other studies have shown more robust benefits of physical therapy, particularly focusing on improving range of motion and preventing contractures (120).Another study found that the addition of physical therapy to BoNT injections extended the duration of improvement by 19.7 days with reduced pain and disability with activities of daily living (121).
Despite efforts to optimize efficacy by selecting appropriate targets, mitigating side effects, and reducing immunogenicity, some patients with CD remain refractory to BoNT.In these cases, advanced therapies such as deep brain stimulation (DBS) may provide benefit.A randomized, sham-controlled study conducted in Europe, which included 62 patients (32 assigned to neurostimulation vs. 30 assigned to sham stimulation), found that pallidal neurostimulation for 3 months is more effective than sham stimulation in reducing symptoms of CD (p = 0.0024) (122).A prospective pilot study found bilateral subthalamic nucleus stimulation improves dystonia, with a statically significant reduction in TWSTRS score, suggesting an alternative to pallidal DBS in the treatment of primary dystonia (123).In a smaller study of two patients who were previously implanted with pallidal targets, and subsequently received subthalamic nucleus DBS, one showed a 74% improvement and the other 84.3% improvement in TWSTRS scores (124).Finally, a metanalysis of RCTs evaluating DBS for dystonia found low-quality evidence to support the use of pallidal stimulation for the treatment of cervical, segmental or generalized moderate-to-severe cases of dystonia with improved functionality and reduced symptom severity (125).
Although BoNT is generally effective and safe for the treatment of CD, other adjunct therapies and physical therapy may be needed to optimize the response (Figure 1).

CONCLUSION
BoNT has been shown, through numerous open-label and controlled trials, to be safe and effective in the treatment of CD.These benefits, which clearly translate into improved quality of life, are usually sustained indefinitely.There are currently four formulations of BoNT (three types A and one type B) approved in the United States, but other formulations are available elsewhere and more are in development.The most important factor in favorable outcome following BoNT injection is the identification and appropriate selection of dose in the target muscle.It is likely that in the near future novel formulation of BoNT will be developed with improved properties such as longer duration of action, less diffusibility and immunogenicity, and lower cost which will lead to wider accessibility.

AUTHOR CONTRIBUTIONS
NH was responsible for data collection and manuscript writing.JJ was responsible for manuscript writing and editing.

CONFLICT OF INTEREST
JJ has received research or training grants from AbbVie Inc., CHDI Foundation, Dystonia Coalition, Emalex Biosciences, Inc., Medtronic Neuromodulation, Michael J Fox Foundation for Parkinson Research, National Institutes of Health, Parkinson's Foundation, Revance Therapeutics, Inc., and Teva Pharmaceutical Industries Ltd. JJ has served as a consultant for AbbVie Inc., Aeon BioPharma, Neurocrine; Revance Therapeutics, and Teva Pharmaceutical Industries Ltd.
The remaining author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

TABLE 1 |
Randomized controlled trials of botulinum toxin in cervical dystonia.
n = 111, 55 patients received BoNTA and 56 received BoNTB TWSTRS score improved by −11 in BoNTA and −8.8 in BoNTB groups.Severe SEs similarly reported in both groups.Dry mouth more common in I (Continued on following page)

TABLE 1 |
(Continued)Randomized controlled trials of botulinum toxin in cervical dystonia.