Abstract
Objective: To determine the safety and efficacy of botulinum toxin type B (BoNT/B) in patients with cervical dystonia (CD).
Background: BoNT/B is a form of chemodenervation therapy for the treatment of patients with CD.
Methods: The authors performed a 16-week, randomized, multicenter, double-blind, placebo-controlled trial of BoNT/B in patients with CD who continue to respond to botulinum toxin type A. Placebo, or 5,000 U or 10,000 U of BoNT/B was administered in two to four muscles involved clinically in CD. The Toronto Western Spasmodic Torticollis Rating Scale (TWSTRS)-Total score at week 4 was the primary efficacy measure. Clinical assessments and adverse events were recorded for treatment day 1 and at weeks 2, 4, 8, 12, and 16.
Results: A total of 109 patients were enrolled randomly across all three treatment groups. The mean improvement in the TWSTRS-Total scores in each group at week 4 was 4.3 (placebo), 9.3 (5,000 U), and 11.7 (10,000 U). For the prospectively defined primary contrast (10,000 U versus placebo), highly significant differences were noted for the primary (TWSTRS-Total, baseline to week 4, p = 0.0004) and supportive secondary (Patient Global Assessment, baseline to week 4, p = 0.0001) outcome measures. Improvement in pain, disability, and severity of CD occurred for patients who were treated with BoNT/B when compared with placebo-treated patients. Overall, improvements associated with BoNT/B treatment were greatest for patients who received the 10,000-U dose. The duration of treatment effect for BoNT/B was 12 to 16 weeks for both doses.
Conclusion: Botulinum toxin type B (NeuroBloc) is safe and efficacious at 5,000 U and 10,000 U for the management of patients with cervical dystonia.
Intramuscular injection of botulinum toxin type A (BoNT/A) into affected muscles has revolutionized therapy for cervical dystonia (CD) and is generally accepted as the primary therapy for CD.1-3 The results of both controlled and open-label clinical trials of BoNT/A for the treatment of CD substantiate its use in clinical practice.4,5
Of the six other botulinum toxin serotypes found in nature, types B,6,7 C,8 and F,9,10 have also been studied in the treatment of CD. BoNT/B is an antigenically distinct serotype synthesized by Clostridium botulinum,11,12 and antibodies to type A and B BoNT do not cross-react in in vivo studies.13,14 Type B toxin inhibits the release of acetylcholine (ACh) at the neuromuscular junction by a different mechanism than type A toxin.15
At least 4% and perhaps more then 7% of treated patients may develop neutralizing antibodies to BoNT/A,16and some patients who develop neutralizing BoNT/A antibodies17,18 have benefited from injections of immunologically distinct preparations, such as BoNT/F19,20 and BoNT/B. In the first reported, double-blind, placebo-controlled study of BoNT/B (formulated as NeuroBloc), it was found safe and effective in the treatment of 122 patients with CD.7 To assess further the safety and efficacy of this formulation of BoNT/B in type A–responsive patients, we performed a randomized, multicenter, double-blind, placebo-controlled trial.
Methods.
Inclusion/exclusion criteria.
Patients were eligible for enrollment if they had CD for at least 1 year that involved two or more of the following muscles: levator scapulae, scalene complex, semispinalis capitis, splenius capitis, sternocleidomastoid, or trapezius; and if their CD continued to respond to BoNT/A treatment. Patients had to have met the following criteria at baseline: Toronto Western Spasmodic Torticollis Rating Scale (TWSTRS)-Total score of ≥20 with at least a TWSTRS-Severity score of ≥10, a TWSTRS-Disability score of ≥3, and a TWSTRS-Pain score of ≥1 (discussed later). Patients enrolled in this study were at least 18 years of age, weighed at least 46 kg, and had baseline physical and neurologic examinations and laboratory tests that were acceptable clinically.
Patients were excluded if they had received a BoNT injection in the last 4 months for their CD, had previously participated in a BoNT/B trial, had neck contractures or cervical spine disease so that a substantial decrease in passive range of motion of the neck was present, had pure retrocollis or anterocollis, used drugs irregularly that could interfere with the evaluation of the safety and efficacy measures (e.g., narcotics, muscle relaxants, or benzodiazepines), had an acute or chronic medical condition or known drug hypersensitivity to the study drug formulation, had a history of myotomy or denervation surgery of the neck and/or shoulder region, received tetanus toxoid in the last 4 months, or used another investigational drug or device within 30 days of entry into this study. Also excluded were patients with a history of clinically persistent neurologic or neuromuscular disorder; with a cardiovascular, renal, hepatic, gastrointestinal, dermatologic, major psychiatric, or hematologic illness; or women of child-bearing potential who were pregnant or breast-feeding.
Study design.
Efficacy assessments.
The TWSTRS-Total score (range, 0 to 87) was the primary efficacy measure.21,22 It is composed of three subscales: Severity (range, 0 to 35), Disability (range, 0 to 32), and Pain (range, 0 to 20). The TWSTRS-Total score was determined at each visit and was used to assess efficacy. Three visual analog scale assessments were also used to assess efficacy: Patient Global Assessment of Change, Principal Investigator (PI) Global Assessment of Change, and Patient Analog Pain Assessment.
Safety assessments.
Adverse event data were collected at each visit during the study. An adverse event was defined as any undesirable event that occurred to a participant during the course of the study or within a reasonable time after study termination, regardless of whether the event was considered to be related to the study drug. The severity and relationship to the study drug were determined by the appropriate investigator for each adverse event. Blood and urine samples, collected at screening, at week 4, and again during the final visit, were assayed for hematology, blood chemistry, and urinalysis variables. Vital signs were determined and recorded at each visit.
Treatment and study activities.
Eligible patients were assigned randomly to one of three treatment groups: BoNT/B, 10,000 U; BoNT/B, 5,000 U; or placebo. Master randomization tables were generated by an independent organization (Pharmaceutical Research Associates, Charlottesville, VA), and the investigators, patients, and the sponsor were blinded to drug assignment until after the database was locked and analyzed.
After obtaining informed consent, the following were obtained/performed: medical history, physical and neurologic examination (including vital signs), laboratory tests, pregnancy test, and serum sample for BoNT/B antibody determination. Efficacy assessments were performed at the treatment visit (day 1) and at weeks 2, 4, 8, 12, and 16 (termination). Vital signs, adverse events, and concomitant medications were recorded at all visits. Additionally, a telephone call was made to the patient at week 1 to assess safety.
To ensure the maintenance of the blind, two investigators—the PI and the administrative investigator (AI)—were designated at each site. The PI completed all screening, day 1 activities, and subsequently completed only the TWSTRS and PI Global Assessments of Change. After study-drug injection, no other information about the patient was provided to or discussed with the PI by either the patient or the site staff. The AI or a designee (but not the PI) conducted all other activities for weeks 1, 2, 4, 8, 12, and the termination visits. At each visit, the patient global and pain assessments were completed independently by the patient. Patients were discontinued or withdrawn from the study at the patient’s request, or at the discretion of the investigator or sponsor if deemed appropriate for any reason. Patients who withdrew prematurely from the study completed the termination procedures.
Test material formulation and dose selection.
The study drug was provided by Athena Neurosciences, Inc. (San Francisco, CA), in 3.5-mL vials containing either 5,000 U of NeuroBloc (botulinum toxin type B) or placebo (same solution without toxin) in a 1.0-mL sterile solution, buffered to a pH of 5.5, and refrigerated. The study drug was assigned randomly to each patient. A total of 2.0 mL of the study drug was injected into two to four involved CD muscles selected by the PI at the day 1 visit with or without the use of electromyography. Based on the PI’s judgment, the proportionate volume per muscle was divided and injected into one to five sites. Each patient received only one treatment.
Statistical analyses.
The intent-to-treat dataset was used for all analyses and consisted of data for every efficacy variable at every visit for all patients who were randomized to treatment. If a patient had missing data for an efficacy variable at any visit, the value of that variable from the previous visit was carried forward. The level of significance for the main effects (e.g., center and treatment) was set at α = 0.05, and the level of significance for interactions was set at α = 0.10. Analyses were performed by using SAS, version 6.12 (SAS Institute, Inc., Cary, NC).23,24
Analyses of the differences among the three treatment groups for gender and race were performed using Fisher’s exact test. Analyses of the differences for age, height, and weight were performed by using analysis of variance (ANOVA). The primary efficacy analysis compared the TWSTRS-Total scores at week 4 among the treatment groups by using analysis of covariance (ANCOVA), and the primary contrast was a test with a single degree of freedom that compared the 10,000-U and the placebo treatment groups. The dependent variable was the score at week 4, and the independent variables were baseline score, center, and treatment group. The study was powered to assess the primary contrast (10,000-U versus placebo groups at week 4 using the TWSTRS-Total scores) based on the results of a previously reported NeuroBloc study.7 ANOVA of the Patient Global Assessment of Change at week 4 for the primary contrast was defined as the supportive secondary efficacy analysis. Other secondary efficacy analyses were an ANOVA of the PI GlobalAssessment of Change at week 4 and an ANCOVA of the TWSTRS-Total scores at weeks 8 and 12. Tertiary efficacy analyses consisted of ANCOVAs of the Patient Analog Pain Assessments at week 4, TWSTRS subscale scores at week 4, and TWSTRS-Total scores at week 16. Baseline scores are those that were recorded on the day of treatment (day 1).
Effects in the preliminary models for the ANCOVA were baseline value, center, treatment, and the center-by-treatment and baseline-by-treatment interactions. Effects in the preliminary models for the ANOVA were center, treatment, and the center-by-treatment interaction. If the interaction was significant, exploratory descriptive analyses were performed. Otherwise, the interaction was removed from the model, and a reduced model was used.
An exploratory Kaplan–Meier survival analysis25 was performed to assess the duration of treatment effect. The event of interest was the time to return to the baseline value as measured by the TWSTRS-Total score. This time was defined as the number of days until the TWSTRS-Total score equals or exceeds the baseline value, and was based on all assessments at or greater than week 4. The median time until return to baseline was then estimated for each treatment group. Using the log-rank statistic,26 the three distributions were compared. Using the Cox27 proportional hazards regression model with two indicator covariates, two contrasts with one degree of freedom each were tested (NeuroBloc versus placebo and 10,000 U versus 5,000 U).
Results.
Patient disposition and population.
A total of 109 patients was enrolled: 36, placebo; 36, 5,000 U; and 37, 10,000 U. A diagram of the study design is presented in figure 1. Four patients discontinued the study prematurely: two patients (one each in the placebo and the 5,000-U groups) because of lack of study drug effect, one placebo patient at the patient’s request (new job), and one 10,000-U patient died following coronary artery bypass surgery (discussed later). Demographic and baseline data are summarized in table 1 for age, gender, race, height, and weight by treatment group. There were no significant differences among the groups for any demographic or baseline characteristics.
Figure 1. The screening and randomization of this clinical trial of Neuro-Bloc in the treatment of A–responsive cervical dystonia patients. *Patient died after triple-vessel coronary bypass surgery. TWSTRS = Toronto Western Spasmodic Torticollis Rating Scale.
Demographic and baseline characteristics
Efficacy results.
TWSTRS-total scores.
Table 2 summarizes TWSTRS-Total scores by visit and treatment group, and figure 2 summarizes the pairwise comparisons, p values, and 95% CIs for selected efficacy variables for the 10,000-U versus the placebo groups (the primary contrast) for the intent-to-treat dataset.
Summary of Toronto Western Spasmodic Torticollis Rating Scale–Total scores
Figure 2. Pairwise comparisons, 95% CIs, and levels of significance for selected efficacy outcome variables for botulinum toxin type B treatment in type A–responsive cervical dystonia patients. A, primary efficacy; B, secondary efficacy variables; C, tertiary efficacy variables. TWSTRS = Toronto Western Spasmodic Torticollis Rating Scale; PI = principal investigator.
Primary efficacy outcome variable.
For the TWSTRS-Total scores, the mean improvement at week 4 was 4.3 (placebo), 9.3 (5,000 U), and 11.7 (10,000 U). For the prospectively defined primary contrast for the TWSTRS-Total scores at week 4, significant differences occurred between the placebo and the 10,000-U groups (estimated difference, 7.2; 95% CI, 3.3, 11.1;, p = 0.0004; see figures 2 and 3⇓), as well as between the placebo and the 5,000-U groups (estimated difference, 5.0; 95% CI, 1.2, 8.9; p = 0.0115; see figure 3).
Figure 3. Botulinum toxin type B therapy in type A–responsive cervical dystonia patients: primary and supportive secondary efficacy outcome variables. (A) TWSTRS—Total change from baseline to week 4. (B) Patient global visual analog scale assessment at week 4. *p = 0.001; **p = 0.0001. TWSTRS = Toronto Western Spasmodic Torticollis Rating Scale.
Supportive and other secondary efficacy outcome variables.
For the prospectively defined supportive secondary efficacy outcome measure, mean values of the Patient Global Assessment of Change at week 4 were 43.6 (placebo), 60.6 (5,000 U), and 64.6 (10,000 U; table 3). For the primary contrast, the Patient Global Assessment of Change at week 4 was significant (estimated difference, 21.2; 95% CI, 11.3, 31.1; p = 0.0001; see figures 2 and 3⇑), as well as between the placebo and the 5,000-U groups (estimated difference, 17.0; 95% CI, 7.0, 26.9; p = 0.0010). For the other secondary efficacy variables the following were noted (see table 3): Mean values for the PI Global Assessment of Change at week 4 were 52.0 (placebo), 65.3 (5,000 U), and 64.2 (10,000 U), with significant differences noted for the 10,000-U versus the placebo groups (estimated difference, 11.8; 95% CI, 3.9, 19.7; p = 0.0038; see figure 2), as well as for placebo versus 5,000 U (estimated difference, 13.4; 95% CI, 5.5, 21.3; p = 0.0011). Mean improvements in TWSTRS-Total scores from baseline to week 8 (table 2) were 2.3 (placebo), 7.0 (5,000 U), and 8.4 (10,000 U), with significant differences noted between the placebo and the 10,000-U groups (estimated difference, 6.1; 95% CI, 2.1, 10.1; p = 0.0032; see figure 2) and between the placebo and the 5,000-U groups (estimated difference, 4.8; 95% CI, 0.8, 8.8; p = 0.0191). Mean improvements in TWSTRS-Total scores from baseline to week 12 were 1.4 (placebo), 3.5 (5,000 U), and 4.1 (10,000 U), with no significant differences noted for the 10,000-U (see figure 2) or the 5,000-U comparisons.
Summary of visual analog scales
Tertiary efficacy outcome variables.
For the Patient Analog Pain Assessment at week 4 (see table 3), the differences were significant between placebo versus 10,000 U (estimated difference, 21.8; 95% CI, 10.5, 33.0; p = 0.0002; see figure 2), as well as for the placebo versus the 5,000-U comparison (estimated difference, 19.2; 95% CI, 8.0, 30.4; p = 0.0010).
For the TWSTRS subscales at baseline and week 4 (table 4), significance was reached for the primary contrast (placebo versus 10,000 U) for both the TWSTRS-Severity and -Pain subscale scores at week 4 (Severity, p = 0.0016; Pain, p = 0.0008; see figure 2). However, for the placebo versus the 5,000-U comparison, significance was not reached on the TWSTRS-Severity subscale but was reached for the TWSTRS-Pain at week 4 (p = 0.0023). For the TWSTRS-Total comparisons at week 16, significance was not reached for either treatment group versus placebo.
Summary of mean TWSTRS subscale baseline and week 4 assessments
Duration of treatment effect.
Figure 4 displays the Kaplan–Meier survival curves by treatment group. The event of interest is the estimated median time until return to baseline: 63 days for placebo, 114 days for the 5,000-U group, and 111 days for the 10,000-U group.
Figure 4. Kaplan–Meier survival curve analysis of the number of days to return to baseline in botulinum toxin type A–responsive patients after treatment with botulinum toxin type B or placebo. Solid line = placebo; large asterisks = 5,000 U of botulinum toxin type B toxin; circles = 10,000 U of botulinum toxin type B toxin.
Based on the log-rank statistic, the three distributions differ significantly (p = 0.01). Using the Cox proportional hazard regression model with two indicator covariates, the one degree of freedom contrast comparing the two BoNT/B groups to placebo is also significant (p = 0.008). No statistical difference was noted between the two active groups.
Adverse events.
Adverse events were reported by similar percentages of patients in each treatment group, although adverse events that were considered to be related to the study drug occurred more frequently in the 10,000-U group than in the 5,000-U or placebo groups. Thirty patients in the placebo group (83%), 32 in the 5,000-U group (89%), and 30 in the 10,000-U group (81%) reported at least one adverse event. Overall, the majority of adverse events was considered to be mild in intensity.
The numbers and percentages of patients who reported those adverse events that occurred in more than 5% of patients are summarized by adverse event and treatment group in table 5. Thirteen of 109 patients (1 in the placebo, 4 in the 5,000-U, and 8 in the 10,000-U groups) reported dysphagia. Of the 13 reported cases of dysphagia, 0 were severe (as determined by the AI), 3 were moderate (1 in the placebo and 2 in the 10,000-U groups), and 10 were mild (4 in the 5,000-U and 6 in the 10,000-U groups). Of the three moderate cases of dysphagia reported, the duration of the dysphagia was as follows: 12 days (10,000 U), 38 days (placebo), and 62 days (10,000 U). For the 10 mild cases of dysphagia reported, the duration was as follows: 1 minute; 2, 4, 6, 11, 16, 21, 22, and 29 days; and ongoing. The latter patient who reported ongoing dysphagia had a preexisting history of mild baseline dysphagia before study entry.
Number (%) of patients who reported each of the adverse events that occurred in >5% of study patients*
Seven serious adverse events were reported that were distributed evenly among the treatment groups; none were deemed to be study-drug related. Reports of serious adverse advents were as follows: placebo group, bladder cancer and atrial flutter; 5,000-U group, coronary occlusion and bladder stenosis; and 10,000-U group, myocardial infarction (patient died), pathologic fracture, and colitis. The one death occurred in a 67-year-old man with a past medical history that included coronary artery disease (with residual shortness of breath and intermittent angina), cerebrovascular accident, hypertension, and hypercholesterolemia who had an acute myocardial infarction 56 days after receiving the study drug (10,000 U). He ultimately died on study day 75 from complications of his triple-vessel coronary artery bypass surgery that was performed on study day 67.
Other safety assessments.
No clinically significant treatment-emergent laboratory abnormalities or vital sign changes occurred.
Discussion.
The general mechanism of action of all BoNTs in blocking neuromuscular function is believed to be a three-step process: extracellular binding of the toxin to the presynaptic site of the neuromuscular junction, internalization and release of the toxin into the cytosol of the nerve terminals, and ultimate inhibition of ACh release from the nerve terminals. BoNT inhibits the presynaptic release of ACh, causing a chemical denervation that results in muscle weakness. Different mechanisms of action for the inhibition of ACh release have been proposed recently for BoNT/A versus BoNT/B.15,28 These neurotoxins are reported to act as zinc-dependent endopeptidases, each of which appears to have a different intraneuronal target protein. The target proteins identified are believed to be part of the intracellular docking proteins responsible for the release of ACh from synaptic vesicles into the neuromuscular junction. BoNT/A cleaves (and thus presumably inactivates) synapse-associated protein-25,28 whereas BoNT/B cleaves vesicle-associated membrane protein (also known as synaptobrevin).15 Therefore, even though the clinical effects of these two neurotoxins are generally similar (e.g., muscle paralysis), their target intraneuronal proteins and their mechanisms of action appear to be quite different.
In this study we have shown that treatment with BoNT/B for CD was consistently better than placebo, evidenced by significant differences for the following prospectively defined efficacy variables: TWSTRS-Total at week 4, Patient Global Assessment at week 4, PI Global Assessment at week 4, and TWSTRS-Total at week 8. Additional support for efficacy was demonstrated by the significant differences that occurred between the 10,000-U and the placebo groups at week 4 for the following tertiary efficacy variables: Patient Analog Pain Assessment (p = 0.0002), TWSTRS-Severity subscale (p = 0.0016), and TWSTRS-Pain subscale (p = 0.0008). Although both dosages were effective, in general, the overall benefits for the 10,000-U group exceeded those of the 5,000-U group.
To estimate the duration of treatment effect of BoNT/B we performed an exploratory analysis using the Kaplan–Meier survival method. Significant differences were noted between the estimated median survival of BoNT/B versus placebo, suggesting an overall duration of treatment effect of 12 to 16 weeks in our patients.
The most commonly reported, clinically relevant adverse events were dry mouth and dysphagia. For the most part, the incidence of these adverse events increased with increasing dosage but were self-limited and well tolerated.
Our current study of BoNT/B (NeuroBloc) found it safe and efficacious at 5,000 U and 10,000 U for the management of BoNT/A–responsive patients with CD. Our safety and efficacy data are consistent with previously reported studies.6,7,29-31 Additionally, we performed an exploratory analysis suggesting a 12 to 16-week duration of treatment effect for BoNT/B. This more robust analysis supports the conclusions of a previous study.7
Our data are strong and compelling evidence for the safety and efficacy of BoNT/B in the treatment of patients with CD. Those who received 10,000 U appeared to have a greater benefit. BoNT/B in this liquid preparation (NeuroBloc) is refrigerated, does not require reconstitution, and offers a new therapeutic option for treating physicians and their patients with CD.
Acknowledgments
Supported by a grant from Athena Neurosciences, Inc., a wholly owned subsidiary of Elan Corporation, plc.
Acknowledgment
The authors acknowledge Drs. Joanne M. Wojcieszek, Cheryl H. Waters, Tanya Baxter, Paul Tuite, Eric J. Pappert, Eric S. Molho, Brad Davis, Peter Lewitt, Alberto Martinez, Keema Sharma, Lisa M. Shulman, and Fabio Danisi for participation as the AIs in this study.
Disclosure. A.W-.H., J.D.W., and M.K. are full-time employees of Athena Neurosciences Inc., and receive salary and other compensation therefor. They have no significant interest in Athena Neurosciences Inc. or its parent, Elan Corporation, plc. They certify that they have no affiliation with any other organization or entity with a direct financial interest in the subject matter or material discussed in this article.
Footnotes
-
See also page 1431
- Received November 24, 1998.
- Accepted June 3, 1999.
References
Disputes & Debates: Rapid online correspondence
NOTE: All authors' disclosures must be entered and current in our database before comments can be posted. Enter and update disclosures at http://submit.neurology.org. Exception: replies to comments concerning an article you originally authored do not require updated disclosures.
- Stay timely. Submit only on articles published within the last 8 weeks.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- 200 words maximum.
- 5 references maximum. Reference 1 must be the article on which you are commenting.
- 5 authors maximum. Exception: replies can include all original authors of the article.
- Submitted comments are subject to editing and editor review prior to posting.