CADTH Reimbursement Review

Incobotulinumtoxina (Xeomin)

Sponsor: Merz Therapeutics, a business of Merz Pharma Canada Ltd.

Therapeutic area: Chronic sialorrhea associated with neurological disorders

This multi-part report includes:

Clinical Review

Pharmacoeconomic Review

Clinical Review

Executive Summary

An overview of the submission details for the drug under review is provided in Table 1.

Introduction

Sialorrhea, or drooling, can occur when there is excessive saliva production or when saliva pools in the mouth due to poor swallowing and/or neuromuscular dysfunction.² Sialorrhea is associated with several neurologic conditions in adults, including Parkinson disease (PD), atypical parkinsonism (AP), stroke, traumatic brain injury (TBI), amyotrophic lateral sclerosis (ALS), multiple sclerosis (MS), cerebral palsy (CP), and dementias such as Alzheimer disease. Sialorrhea is linked with the severity of the underlying neurologic condition. Chronic troublesome sialorrhea can lead to speech difficulties, facial skin maceration, halitosis, infections, and, potentially, dehydration, choking, aspiration, and pneumonia; together, these have a significant negative impact on patient health-related quality of life (HRQoL) (speaking, eating, social interaction, emotional distress, and social isolation).³

The prevalence and incidence of chronic troublesome sialorrhea in adult patients with neurologic disorders in Canada is unclear, in part because of the lack of a standardized definition for the condition. According to the clinical expert consulted by CADTH for this review, patients with PD represent the largest group of patients treated for sialorrhea in Canadian clinical practice, although the incidence of sialorrhea is higher in patients with conditions that occur more rarely in the Canadian population (e.g., CP, ALS, and TBI). Diagnosis of chronic troublesome sialorrhea in adult patients with neurologic conditions is made by a neurologist or physiatrist based on clinical evaluation. Patients with mild sialorrhea may be treated with chewing gum, hard candy, mouth exercises and/or speech therapy. The need for further treatment arises when symptoms worsen and patients need to carry a cloth to wipe away saliva, experience skin breakdown, or begin to choke on their saliva. Only a subset of these patients will choose to receive pharmacological treatments, including 1% atropine drops or anticholinergics such as amitriptyline (both used off-label). However, according to the clinical expert consulted by CADTH for this review, the therapeutic effects of atropine drops are often temporary, while anticholinergics have systemic side effects and are not well tolerated by many patients.

According to the clinical expert consulted by CADTH for this review, an ideal treatment for sialorrhea would have minimal adverse effects and effectively reduce the frequency and severity of sialorrhea. The treatment goals are to reduce social isolation and prevent or ameliorate maceration of the skin around the mouth, dehydration, speech disturbances, interference with eating, and risk of aspiration. Injection of off-label botulinum neurotoxins (BoNTs) into the salivary glands has been used clinically for many years to reduce sialorrhea in patients with neurologic disorders.^4-6 According to the clinical expert consulted by CADTH for this review, BoNT injections are desirable because they are focal treatments for symptomatic therapy, easy to administer (typically requiring less than 5 minutes), and have limited side effects.

IncobotulinumtoxinA is a purified botulinum neurotoxin type A (BoNT-A) that inhibits acetylcholine production and contraction of the salivary glands. It is the only BoNT and drug of any type approved by Health Canada for the treatment of chronic sialorrhea associated with neurologic disorders in adults. According to the Health Canada–recommended dose, incobotulinumtoxinA is injected at a total dosage of 100 U (30 U per side in the parotid glands and 20 U per side in the submandibular glands) every 16 weeks.⁷

The objective of this review was to perform a systematic review of the beneficial and harmful effects of incobotulinumtoxinA (100 U) administered by intraglandular injection for the treatment of chronic sialorrhea associated with neurologic disorders in adults.

Stakeholder Perspectives

The information in this section is a summary of input provided by the patient groups who responded to CADTH’s call for patient input and from a clinical expert consulted by CADTH for the purpose of this review.

Patient Input

Input for this review was provided by 1 patient group, Parkinson Québec, which is a not-for-profit organization that supports patients with PD in Québec through advocacy, service development, research funding, revenue development, communication, and network management. Parkinson Québec distributed an online survey to traditional users of their services (individuals living with PD and their caregivers). The survey was promoted through the group’s newsletter and social networks between January 19, 2021, and March 1, 2021. Respondents had to be individuals living with PD and sialorrhea or their caregivers, at least 18 years of age, and Québec residents. Among the respondents, 138 individuals living with PD (47%) and 44 caregivers (40%) reported sialorrhea; of these, 116 individuals living with PD and 36 caregivers fully completed the survey.

Respondents were asked how sialorrhea affected their lives. Approximately 1/3 of individuals with PD reported that sialorrhea affected various aspects of their day-to-day lives, including their self-esteem, social discomfort, ability to eat or swallow, and ability to speak or communicate. Approximately 40% to 1-half of caregivers reported that sialorrhea affected their loved ones’ self-esteem, social comfort, personal relationships, ability to speak or communicate, and ability to eat or swallow. The most common methods used by individuals living with PD to manage sialorrhea were tissues or cloths to wipe drool (87%), followed by chewing gum (17%) and muscle exercises (16%). Few individuals living with PD had used medications (5%) or BoNTs (1%) to manage sialorrhea. Respondents were asked to indicate their perceptions regarding the effectiveness of methods currently used to manage sialorrhea. Overall, 61% to 63% of individuals living with PD and 40% to 47% of caregivers were satisfied with the management of their sialorrhea and felt it was being well managed. Approximately 1/3 of individuals with PD and 43% of caregivers agreed that there was a need for new treatments to manage sialorrhea. Respondents were asked to indicate their expectations for new treatments for sialorrhea. Overall, 82% of individuals with PD and 77% of caregivers desired government coverage of treatments, while 65% of individuals with PD and 71% of caregivers desired treatments with rare and mild side effects. Also desired were treatments that reduced the frequency and severity of sialorrhea, oral treatment options, and treatments with longer durations of action.

None of the survey respondents had any previous experience with incobotulinumtoxinA and only 1 respondent had received BoNT injections. No specific treatment outcomes or measures for reduced sialorrhea were identified in the patient input.

Clinician Input

Input from Clinical Experts Consulted by CADTH

One clinical specialist with expertise in the diagnosis and management of chronic troublesome sialorrhea associated with neurologic disorders in adults provided input for this review. The clinical expert stated that there is a significant unmet therapeutic need among adult patients with sialorrhea. Unlike pharmacological or surgical interventions, BoNT injections are easy to administer, have limited side effects, and are helpful for symptomatic therapy. However, they not covered by drug plans and special access must be requested through pharmacare support programs that have limited resources.

IncobotulinumtoxinA does not modify the disease process, but has several advantages compared to other options. It is already part of the current treatment paradigm but cannot be easily accessed by many patients due to funding limitations. Patients best suited for treatment with incobotulinumtoxinA would be those with significant disabling sialorrhea (e.g., those who need to use a cloth to wipe away drool and those for whom the condition is socially isolating). Patients would need to attend injection sessions every 3 to 6 months and have no major swallowing difficulties due to risk of worsening. Patients with sialorrhea that is too mild or patients with swallowing difficulties would be least suitable for treatment. Many neurologic patients have high risks of urinary retention and confusion, and anticholinergics would not be appropriate for many of these patients.

The objective measures used in trials to assess sialorrhea (e.g., radioisotope scanning, collection cups, and counting napkins) are impractical and not used in clinical practice. Response is usually assessed by taking a history. If necessary, a visual analogue scale (VAS), or tools such as the Drooling Severity and Frequency Scale (DSFS), can be used to assess response. A clinically meaningful response would be an improvement in the patient’s HRQoL as described previously. Response can be assessed subjectively at each visit as the drug is an injectable treatment. Treatment should be discontinued when it is not efficacious or when patients develop adverse events (AEs) such as swallowing problems or dental issues. IncobotulinumtoxinA should be administered in a hospital outpatient or community setting. Neurologists or physiatrists would typically be the specialists involved in the care of patients with neurologic conditions and would perform the injections.

Clinician Group Input

No clinician group input was provided for this review.

Drug Program Input

Drug programs identified several key issues related to implementation. The first is whether coverage would be restricted to the specific neurologic conditions assessed in the pivotal phase III trial of incobotulinumtoxinA. The clinical expert consulted by CADTH for this review noted that the study enrolled primarily patients with PD for feasibility reasons but that the results were most likely generalizable to patients with sialorrhea arising from other neurologic conditions who may also benefit from treatment. Second, drug programs asked which criteria would be used to assess the severity of sialorrhea necessitating treatment. The clinical expert noted that eligibility would be based on patient needs and clinician decisions; even patients with moderate but daily issues with drooling may benefit from treatment. Third, drug programs asked whether patients should try off-label systemic medications such as anticholinergics before treatment with incobotulinumtoxinA. The clinical expert stated that these medications are not often used in clinical practice, primarily due to the risks of side effects, but that disease-specific therapy would be routinely optimized in clinical practice before starting treatment with a BoNT. Fourth, drug programs asked whether a combination of incobotulinumtoxinA and anticholinergics would be excluded from coverage. The clinical expert stated that stable concomitant therapies such as anticholinergics have different mechanisms and indicated that there could be a combined benefit. Fifth, drug programs asked whether coverage would be considered for doses other than those studied in the pivotal phase III trial and the Health Canada–approved dose of 100 U. The clinical expert stated that most clinicians would use a dose close to 100 U to avoid side effects. Sixth, drug programs asked whether specific assessment scales such as the DSFS or the Global Impression of Change Scale (GICS) would be used to determine whether treatment should be continued. The clinical expert responded that questions similar to those used in these scales are routinely asked in clinical practice and that treatment decisions would be grounded in assessment of response by both the patient and clinician. Finally, drug programs had questions related to resumption of treatment following discontinuation. The clinical expert stated that treatment could be restarted and used as necessary to manage symptoms; even if treatment was discontinued due to lack of efficacy, sialorrhea may subsequently become more severe or more frequent and patients may benefit from re-treatment at a later stage. The only exceptions would occur in patients who experienced severe side effects of incobotulinumtoxinA treatment such as swallowing impairment; in these patients, treatment might not be resumed if the risk was too high as judged by the clinician.

Clinical Evidence

Pivotal Studies and Protocol-Selected Studies

Description of Studies

One phase III, double-blind, placebo-controlled, multi-centre study (SIAXI^8-10) with an extension period (EP) of dose-blinded active treatment was included. The study enrolled adults aged 18 to 80 years with moderate to severe sialorrhea resulting from neurologic conditions (PD/AP, stroke, or TBI; N = 184). Chronic troublesome sialorrhea was defined as sialorrhea lasting for at least 3 months, with a DSFS sum score of 6 or greater, DSFS scores for both severity and frequency of at least 2, and a modified Radboud oral motor inventory for Parkinson disease (mROMP) Section III “Drooling,” Item A score of 3 or greater at both screening and baseline. The objective of the study was to investigate the efficacy and safety of injection of 2 doses of incobotulinumtoxinA (75 U or 100 U) into the salivary glands, compared with placebo, in reducing the unstimulated salivary flow rate (uSFR) as well as the frequency and severity of chronic troublesome sialorrhea as evaluated by patients, caregivers, and investigators using multiple rating tools (GICS, DSFS, and mROMP), drooling scores, and HRQoL evaluated using a VAS. The study comprised 4 consecutive 16-week treatment cycles. Following each incobotulinumtoxinA injection, patients were assessed over the course of each cycle through in-person visits to study sites and telephone calls. In the main period (MP) of the study (cycle 1), patients were randomized 2:2:1 to receive 75 U of incobotulinumtoxinA, 100 U of incobotulinumtoxinA, or placebo (saline) via 4 bilateral injections in the parotid and submandibular glands. For the EP, which covered cycles 2 to 4, patients who received placebo were re-randomized 1:1 to receive either 75 U or 100 U of incobotulinumtoxinA. All participants were blinded to dose level. The total duration of the study was 64 weeks. Efficacy outcomes for the 75 U incobotulinumtoxinA dose are not presented in this report because these data are not aligned with the Health Canada–approved dose (100 U).

The co-primary efficacy outcomes in SIAXI were the change in uSFR from baseline to week 4 and patient-reported GICS score at week 4 of the MP. The secondary outcomes were change in uSFR from baseline to weeks 8 and 12 and patient-reported GICS score at weeks 1, 2, 8, and 12 of the MP. Exploratory outcomes included DSFS sum and subscores, mROMP speech and drooling scores, and HRQoL assessed using the EuroQol 5-Dimensions 3-Levels questionnaire (EQ-5D-3L) during the MP and the EP.

The mean age of the study population at the MP baseline was 65.2 years (standard deviation [SD] = 11.4 years). Patients were mostly men (70.7%), White (99.5%), and predominantly had sialorrhea secondary to PD (70.7%) or stroke (19.0%). Smaller numbers of patients had AP (8.7%) or TBI (2.7%). The mean duration of sialorrhea was 32.7 months (SD = 34.5 months). Patients had moderate to severe sialorrhea based on DSFS and mROMP scores. Baseline demographic and clinical characteristics (including baseline uSFR, DSFS sum scores, DSFS severity scores, DSFS frequency scores, and mROMP drooling scores) were generally well balanced between study arms in the MP, as well as between the MP and EP. However, 13.9% of placebo-treated patients compared to 24.3% of incobotulinumtoxinA 100 U–treated patients reported receiving prior and concomitant deep brain stimulation (DBS). The clinical expert consulted by CADTH for this review stated that this imbalance was unlikely to affect the internal validity of the study, as patients were kept on the same therapy (medications and/or DBS) before and throughout the study.

Efficacy Results

In the co-primary efficacy analysis, change in uSFR from baseline and patient-reported GICS scores were assessed at week 4 post-injection using a mixed model for repeated measures (MMRM) analysis (Table 2). In exploratory efficacy analyses, DSFS and mROMP scores and HRQoL were also assessed at multiple time points post-injection, including at week 4.

At week 4 of the MP, the least squares mean (LSM) change in uSFR in the incobotulinumtoxinA 100 U arm was −0.13 g/min (standard error [SE] = 0.026; 95% confidence interval [CI], −0.18 to −0.08) compared to −0.04 g/min (SE = 0.033; 95% CI, −0.11 to 0.03) in the placebo arm. The LSM difference in uSFR between the incobotulinumtoxinA 100 U arm and the placebo arm of −0.09 g/min (SE = 0.031; 95% CI, −0.15 to −0.03) was statistically significant in favour of incobotulinumtoxinA 100 U (P = 0.004). In the EP (cycles 2, 3, and 4), similar mean changes in uSFR from study baseline to week 4 were observed for patients treated with incobotulinumtoxinA 100 U, although mean changes with reference to the baseline for each cycle were much smaller in magnitude (−0.03 to −0.06 g/min).

At week 4 of the MP, the LSM patient GICS score in the incobotulinumtoxinA 100 U arm was 1.25 (SE = 0.144; 95% CI, 0.97 to 1.53) compared to 0.67 (SE = 0.186; 95% CI, 0.30 to 1.04) in the placebo arm. The LSM difference in GICS scores between the incobotulinumtoxinA 100 U arm and the placebo arm of 0.58 (SE = 0.183; 95% CI, 0.22 to 0.94) was statistically significantly in favour of incobotulinumtoxinA 100 U (P = 0.002). In the EP (cycles 2, 3, and 4), similar mean GICS scores at week 4 were reported by patients treated with incobotulinumtoxinA 100 U to describe changes in sialorrhea since the previous injection.

At week 4 of the MP, the LSM change in DSFS sum score in the incobotulinumtoxinA 100 U arm was −1.66 (SE = 0.234; 95% CI, −2.12 to −1.20) compared to −0.50 (SE = 0.296; 95% CI, −1.08 to −0.09) in the placebo arm; the LSM difference in DSFS sum scores between the incobotulinumtoxinA 100 U arm and the placebo arm was −1.17 (SE = 0.278; 95% CI, −1.71 to −0.72). In the EP (cycles 2, 3, and 4), similar mean changes in DSFS sum scores for patients treated with incobotulinumtoxinA 100 U were observed with respect to study baseline.

At week 4 of the MP, larger mean decreases were observed in mROMP drooling scores in the incobotulinumtoxinA 100 U arm (−5.66 [SD = 6.16]) compared to the placebo arm (−1.00 [SD = 4.71]) were observed. In the EP (cycles 2, 3, and 4), similar or larger mean changes in mROMP drooling scores for patients treated with incobotulinumtoxinA 100 U were observed with respect to study baseline.

No significant changes in HRQoL measured using the EuroQol 5-Dimensions Visual Analogue Scale (EQ VAS) were observed during the MP or EP for patients treated with incobotulinumtoxinA 100 U or placebo.

Consistent differences of similar magnitudes in efficacy outcomes (uSFR, GICS, DSFS, and mROMP), but not in HRQoL, were observed between incobotulinumtoxinA 100 U and placebo-treated patients at weeks 8 and 12 of the MP. For patients treated with incobotulinumtoxinA 100 U, similar magnitudes of change from study baseline were observed during each of the additional 3 treatment cycles of the EP.

According to the clinical expert consulted by CADTH for this review, the LSM differences in GICS and DSFS scores between the incobotulinumtoxinA 100 U and placebo arms observed during the MP of the study were clinically meaningful.

Harms Results

In the MP of the SIAXI study, AEs and serious adverse events (SAEs) occurred at similar frequencies in the placebo arm (41.7% and 8.3%, respectively) and incobotulinumtoxinA 100 U arm (45.9% and 12.2%, respectively); withdrawal due to adverse events (WDAEs) were extremely rare (0% and 1.2%, respectively) and no deaths occurred. In 48-week follow-up EP, only slightly higher rates of AEs and SAEs were observed in patients treated with incobotulinumtoxinA 100 U (60.7% and 15.7%, respectively). During the EP, WDAEs occurred in 9.0% of patients treated with incobotulinumtoxinA 100 U, more than half of whom (4.5%) discontinued due to a dry mouth. Adverse events of special interest (AESIs) considered by investigators as potentially related to toxin spread occurred in 6.8% of patients in the incobotulinumtoxinA 100 U arm (but no placebo-treated patients) in the MP, as well as 13.5% of incobotulinumtoxinA 100 U–treated patients in the EP. These AESIs were generally manageable in most patients. Dysphagia occurred in 4.5% of incobotulinumtoxinA-treated patients in the EP. Dental-related AEs did not occur more frequently in patients treated with 100 U of incobotulinumtoxinA compared with placebo.

Critical Appraisal

The SIAXI trial was rigorously designed with no major risks of bias. Some areas of potential concern that may affect interpretation of the study results should be noted. The treatment arms were imbalanced in terms of some concomitant medications and therapies, most notably DBS. The clinical expert consulted by CADTH for this review stated that this imbalance was unlikely to affect the internal validity of the study as patients were kept on the same therapy (medications and/or DBS) before and throughout the study. The study used unvalidated outcome measures and no evidence was available to support validity, reliability, and responsiveness to change; placebo effects were observed for all outcomes. For categorical outcomes measured using Likert scales, such as the GICS, the degree to which these constructs were sensitive in delineating true treatment responses from placebo effects was unclear. The study was overpowered for efficacy (based on effect sizes from a prior study of rimabotulinumtoxinB) because of the larger sample size required for safety evaluation, but still detected relatively small mean differences in efficacy outcomes between incobotulinumtoxinA 100 U and placebo. The clinical meaningfulness of differences of these magnitudes was uncertain in part because no evidence was available to suggest a minimal important difference (MID) for any of the outcome measures. Despite these caveats, consistent differences in favour of incobotulinumtoxinA were observed across all study outcomes with similar timing (weeks 4, 8, and 12 post-injection).

The characteristics of patients treated in SIAXI were generally similar to the Canadian context, although there were no study sites in Canada. However, patients were mostly White, male, and from only 2 countries (Germany and Poland), and almost all had sialorrhea secondary to either PD/AP or stroke. In addition, over the complete study (MP plus EP), patients were followed and monitored frequently for 64 weeks, and whether the study’s findings are generalizable to patients with different levels of background care or less-stringent dosing schedules is unclear. None of the efficacy outcomes used in SIAXI are used routinely in clinical practice, and their clinical relevance, importance to patients, and correlation with HRQoL was not clear.

Indirect Comparisons

No indirect evidence was identified for this review. A feasibility assessment conducted by the sponsor also concluded that no data were currently available to inform an indirect treatment comparison between incobotulinumtoxinA and other interventions, including injection of other BoNTs.

Other Relevant Evidence

One additional exploratory single-centre, double-blind, randomized controlled trial (RCT) was summarized to provide additional evidence from patients with other neurologic conditions and comparative evidence for incobotulinumtoxinA and onabotulinumtoxinA. This study was not designed as a direct head-to-head comparison of these 2 BoNTs.

Description of Studies

The study by Restivo et al.¹¹ recruited a consecutive series of patients (aged 18 to 75 years) with PD, stroke, TBI, ALS, or CP (N = 90) with severely disabling sialorrhea. The primary goal of the study was to assess the relationship between efficacy in reducing sialorrhea and the number of glands injected; however, analyses of interest to this review included comparative efficacy assessment of incobotulinumtoxinA versus onabotulinumtoxinA and of BoNT efficacy in patients with different neurologic conditions. Patients were randomized to receive BoNT-A injections (either incobotulinumtoxinA or onabotulinumtoxinA) in different numbers of salivary glands (2, 3, or 4), resulting in a total dose received of 50 U, 75 U, or 100 U. At baseline and 2 weeks post-injection, salivary production was measured by weighing dental rolls placed in the patient’s mouth for 5 minutes. The change in salivary production from baseline was evaluated on a Likert scale (0 = no reduction, 1 = 25% reduction, 2 = 50% reduction, and 3 = 75% reduction in salivary weight).

Efficacy Results

There was a clear pattern in the dose response for both BoNT-A types, with Likert scores increasing with the number of glands injected (P < 0.001), but there was no interaction between BoNT-A type and number of glands injected. The Likert scores of patients treated with the 2 types of BoNT-A appeared to be similar, although the numerical data were not reported (P = 0.12). Subgroup analysis by etiology of sialorrhea in the overall population treated with all doses of a BoNT-A (either incobotulinumtoxinA or onabotulinumtoxinA) suggested a potential difference in treatment effect by neurologic condition (P < 0.001).

Harms Results

Harms were not formally analyzed.

Critical Appraisal

The study by Restivo et al. was described in limited detail and there was significant uncertainty regarding its internal and external validity. Because randomization was by number of glands injected rather than BoNT received, the comparative evidence from this study (incobotulinumtoxinA versus onabotulinumtoxinA) was potentially susceptible to bias and confounding. Furthermore, inability to account for imbalances in the type of BoNT administered to patients with different neurologic conditions (and vice versa) weakened analysis of either factor. Only 8 patients in the study were treated with the Health Canada–approved dose of incobotulinumtoxinA (100 U) and none of these had neurologic conditions that differed from those assessed in the SIAXI study. The study was therefore unable to address the evidence gaps relating to the efficacy of incobotulinumtoxinA in patients with neurologic conditions other than PD/AP and stroke and to comparative efficacy versus other BoNT-A injections for this indication

Conclusions

Evidence from the SIAXI study suggested that injection of incobotulinumtoxinA 100 U into the salivary glands of adult patients with neurologic disorders resulted in reduced salivary production and improvements in patients’ perceptions of frequency and severity of sialorrhea. At 4 weeks post-injection, the mean difference in change from baseline on the uSFR and patient GICS scores was statistically significant in favour of incobotulinumtoxinA versus placebo. Treatment effects in the uSFR and GICS were also observed at weeks 8 and 12 post-injection, and similar results were obtained on the investigator-rated DSFS. The clinical significance of post-treatment changes in sialorrhea between incobotulinumtoxinA- and placebo-treated patients was uncertain because the outcome measures were unvalidated, not used in clinical practice, and subjective (apart from uSFR), and therefore the magnitudes of treatment effects were of unclear relevance to patients. However, the clinical expert consulted by CADTH for this review indicated that questions similar to those asked in the GICS, DSFS, and mROMP drooling scales are part of the clinical assessment, and that the differences in GICS and DSFS between the incobotulinumtoxinA 100 U and placebo arms were clinically meaningful. Despite the uncertain clinical relevance of the magnitude of treatment differences between incobotulinumtoxinA and placebo, and despite the observation of a placebo effect for most outcomes, consistent mean changes with similar timings were observed in favour of incobotulinumtoxinA across all assessed outcomes. Numerical differences in the effects of treatment with incobotulinumtoxinA 100 U versus placebo were observed (via GICS responses) but not statistically significant at weeks 1 and 2 post-injection, clearly manifested at weeks 4, 8, and 12, and then waned by week 16, at which point a subsequent dose was administered. However, this did not translate into improvement for incobotulinumtoxinA-treated patients in terms of HRQoL measured via the EQ VAS. Injection with incobotulinumtoxinA was tolerated in most patients and side effects were generally manageable, with some infrequent but expected notable harms related to toxin spread (e.g., dry mouth and dysphagia). Key evidence gaps included a lack of comparative data on the efficacy of different BoNTs and a lack of evidence from patients with a variety of neurologic conditions

Introduction

Disease Background

Sialorrhea, or drooling, occurs when excess saliva spills over the lip margin. In healthy individuals, approximately 1 L of saliva is continuously produced by 3 pairs of major salivary glands (parotid, sublingual, and submandibular) and swallowed each day. Saliva flow is mediated by acetylcholine binding to muscarinic receptors in the salivary glands.¹² Sialorrhea arises when there is excessive saliva production or when saliva pools in the oral cavity because of poor swallowing; pooling can be anterior (resulting in spilling of saliva from the open mouth), posterior (resulting in spilling into the pharynx with increased risk of aspiration and infection), or both.² Hypersalivation can be caused by some medications or conditions (e.g., gastroesophageal reflux disease), while poor swallowing can be due to anatomic abnormalities (e.g., macroglossia, oral incompetence, or dental malocclusion), neuromuscular dysfunction (e.g., PD, stroke, TBI, or CP), and/or decreased swallowing reflexes (e.g., AP). In adults with neurologic disorders, sialorrhea is linked with the severity of the underlying neurologic condition. Sialorrhea becomes chronic and troublesome when the frequency and/or severity of drooling begins to significantly and consistently disrupt the patient’s life. For example, the patient may require frequent changes of clothes or regular use of a cloth to wipe away saliva. The adverse effects of chronic troublesome sialorrhea include speech difficulties, facial skin maceration, halitosis, infections, and potentially, dehydration, choking, aspiration, and pneumonia; together, these have a significant negative impact on HRQoL (speaking, eating, social interaction, emotional distress, and social isolation).³ Sialorrhea can also be burdensome for caregivers, who may need to regularly monitor loved ones for drooling and risk of aspiration.¹³

The exact prevalence and incidence of sialorrhea and of chronic troublesome sialorrhea in adult patients with neurologic disorders is unclear in Canada and elsewhere, in part because of the lack of a standardized definition for this condition. The prevalence of sialorrhea in patients with PD ranges from 32% to 74%.¹⁴ According to the clinical expert consulted by CADTH for this review, patients with PD represent the largest group treated for sialorrhea in Canadian clinical practice. However, only a relatively small subset of all patients with PD (typically those with more advanced disease) would receive pharmacological treatment for sialorrhea. While patients with PD and sialorrhea may be the most numerous, the clinical expert consulted by CADTH for this review stated that the incidence of sialorrhea is higher in patients with conditions that occur more rarely in the Canadian population, such as CP, ALS, and TBI. Based on 2010 to 2011 estimates of the prevalence of various neurologic conditions (PD, TBI, stroke, ALS, MS, AD, and other dementias) from a Canadian Community Health Survey, a National Population Health Study of Neurologic Disorders, and the Ontario Federation for Cerebral Palsy,^15-18 and multiplying these by estimates of the proportions of patients experiencing severe neurologic disease^19-24 and the proportions experiencing sialorrhea,^14,25-27 the sponsor estimated that more than 20,000 Canadians may be living with chronic sialorrhea associated with a neurologic disorder.

Diagnosis of chronic troublesome sialorrhea in adult patients with neurologic conditions is made by a neurologist or physiatrist based on clinical evaluation.

Standards of Therapy

According to the clinical expert consulted by CADTH for this review, an ideal treatment for sialorrhea would have minimal adverse effects (an issue with anticholinergics) and effectively reduce the frequency and severity of sialorrhea. The treatment goals would be to reduce social isolation and prevent or ameliorate maceration of the skin around the mouth, dehydration, speech disturbances, interference with eating, and risk of aspiration. The most relevant indicator of response to treatment and change in sialorrhea is clinical history and/or self-reporting by patients or caregivers. Drooling scales used in clinical trials (e.g., the DSFS) are typically not used in clinical practice, while saliva collection and measurement are impractical.

According to the clinical expert consulted by CADTH for this review, mild sialorrhea can be treated with chewing gum or hard candy, oral exercises, and/or speech therapy. These methods become ineffective when a certain level of impairment is reached. There is no standardized definition of chronic or troublesome sialorrhea. According to the clinical expert consulted by CADTH for this review, treatment would be considered when patients need to carry a cloth to wipe saliva multiple times a day, when skin breakdown is observed, or when caregivers report that the patient is choking on saliva. Only a subset of patients with neurologic disorders and sialorrhea would choose to receive pharmacological treatments for sialorrhea. When symptoms worsen, patients may be treated with 1% atropine drops or anticholinergics such as amitriptyline (both off-label). However, according to the clinical expert consulted by CADTH for this review, the therapeutic effect of atropine drops is often temporary, while anticholinergics have systemic side effects and are not well tolerated in all patients, particularly those with PD who have significant comorbidities (e.g., urinary retention or confusion). Dopaminergic medications for parkinsonism have a limited impact on sialorrhea, although optimization of anti-parkinsonian medications can be effective in reducing neuromuscular dysfunction and is typically accomplished before additional interventions. Surgery (e.g., gland excision and duct ligation) or irradiation of the salivary glands can be used to control sialorrhea but these interventions are expensive, highly invasive, and, according to the clinical expert consulted by CADTH for this review, are not used in most patients.

Since the first description of sialorrhea treatment with BoNT injections into the salivary glands,²⁸ multiple studies have evaluated the safety and efficacy of BoNT injections into the salivary glands in reducing sialorrhea among adult patients with PD or ALS.^6,29-33 Most studies used doses of 55 U to 450 U of BoNT-A^34-37 and 2,500 U to 4,000 U of botulinum neurotoxin type B^31,32,38 injected into the salivary glands. Preliminary evidence based on non-randomized studies or studies with small sample sizes was suggestive of a clinical benefit in reducing salivary production and drooling severity.³⁹ There is a general consensus among clinicians and specialists that BoNT-A injections are potentially efficacious in controlling drooling in patients with sialorrhea.^4-6 However, due to the potential for worsening of swallowing issues, patients need to be chosen carefully. According to the clinical expert consulted by CADTH for this review, BoNT injections have several desirable characteristics: they are focal treatments for symptomatic therapy, injections are easy to administer (typically requiring less than 5 minutes for 4 injections), and they have limited side effects. The number of salivary glands injected and the dose can be customized to each patient to optimize the therapeutic effect. The clinical expert stated that the treatment effect of BoNT-A injections begins within a few days, peaks within several weeks, then wanes over several months; repeat injections 3 or 4 times a year are therefore required. Injections with BoNTs in the form of a BoNT-A (incobotulinumtoxinA [Xeomin], onabotulinumtoxinA [Botox], abobotulinumtoxinA [Dysport], or prabotulinumtoxinA [Nuceiva]) or botulinum neurotoxin type B (rimabotulinumtoxinB [Myobloc]) are also used off-label to treat sialorrhea in adults with neurologic conditions. RimabotulinumtoxinB is not currently marketed in Canada and prabotulinumtoxinA is indicated for cosmetic use only. According to the clinical expert consulted by CADTH for this review, most commonly onabotulinumtoxinA would be used off-label in Canada at present as abobotulinumtoxinA is less widely available and dosing requires an inconvenient calculation. As this treatment is not covered, it can only be provided by compassionate access through pharmaceutical companies. IncobotulinumtoxinA is the only approved treatment for sialorrhea in adults with chronic troublesome sialorrhea resulting from neurologic conditions.

Drug

IncobotulinumtoxinA is a purified BoNT-A free from complexing proteins that is produced from anaerobic fermentation of the Hall strain of Clostridium botulinum. The drug blocks transmission at the neuromuscular junction by inhibiting the release of acetylcholine from peripheral cholinergic nerve endings (Table 3). The mechanism of action involves: (1) binding to an as-yet uncharacterized site on presynaptic cholinergic axon terminals, (2) uptake within an endocytic vesicle, (3) pore formation and translocation of the light chain into the cytosol, and (4) proteolytic cleavage of SNAP 25, which is a component of the vesicle fusion machinery required for the release of acetylcholine.⁷ Following injection into salivary glands, muscular contraction and salivary production controlled by acetylcholine may be reduced.

IncobotulinumtoxinA is indicated for the treatment of chronic sialorrhea associated with neurologic disorders in adults.⁷ In this indication, the regimen recommended by Health Canada is a total dosage of 100 U (30 U per side in the parotid glands and 20 U per side in the submandibular glands) every 16 weeks. Note that the definition of units is specific to incobotulinumtoxinA and thus dosing is not interchangeable between different BoNTs. The drug received a Notice of Compliance from Health Canada for this indication on November 17, 2020. The sponsor’s reimbursement request is for the Health Canada–approved indication. The timing for repeat treatment should be determined based on the clinical needs of the individual patient, and no sooner than every 16 weeks.

Stakeholder Perspectives

Patient Group Input

This section was prepared by CADTH staff based on input provided by patient groups. The statistical data reported have been reproduced as is according to the submission, without modification.

About the Patient Group and Information Gathered

Input for this review was provided by 1 patient group, Parkinson Québec, which is a not-for-profit organization that supports patients with PD in Québec through advocacy, service development, research funding, revenue development, communication, and network management. Parkinson Québec serves the entire province of Québec, and its services and resources are available free of charge to the 25,000 Quebecers living with PD and their families.

Parkinson Québec distributed an online survey to traditional users of their services (individuals living with PD and their caregivers). The survey was promoted through a newsletter and social networks between January 19 and March 1, 2021. Respondents had to be individuals living with PD and sialorrhea or their caregivers, at least 18 years of age, and Québec residents. A total of 405 individuals responded, 295 of whom were living with PD and 110 of whom were caregivers. Among the respondents, 138 individuals living with PD (47%) and 44 caregivers (40%) reported sialorrhea. Results were reported only for respondents who had completed the entire survey (116 individuals living with PD and 36 caregivers). The characteristics of respondents are shown in Table 4.

Disease Experience

Respondents were asked how sialorrhea affected their lives. Approximately 1/3 of individuals with PD reported that sialorrhea affected various aspects of their day-to-day lives, including their self-esteem, social comfort, ability to eating or swallow, and ability to speak or communicate. Approximately 40% to 50% of caregivers reported that sialorrhea affected their loved ones’ self-esteem, social comfort, personal relationships, ability to speak or communicate, and ability to eat or swallow.

Experiences with Currently Available Treatments

The most common methods used by individuals living with PD to manage sialorrhea were tissues or cloths to wipe away drool (87%), followed by chewing gum (17%) and muscle exercises (16%). Comparatively few individuals living with PD had used medications (5%) or BoNTs (1%) to manage sialorrhea.

Respondents were asked to indicate their perceptions of the effectiveness of methods currently used to manage sialorrhea. Overall, only 61% to 63% of individuals living with PD and 40% to 47% of caregivers were satisfied with the management of their sialorrhea and felt it was being managed well. Approximately 1/3 of individuals with PD and 43% of caregivers agreed that there was a need for new treatments to manage sialorrhea.

Improved Outcomes

Respondents were asked to indicate their expectations for new treatments for sialorrhea. Overall, 82% of individuals with PD and 77% of caregivers desired government coverage of treatments, while 65% of individuals with PD and 71% of caregivers desired treatments with rare and mild side effects. Many respondents also desired treatments that would reduce the frequency and severity of sialorrhea. Oral treatment options and treatments with longer durations of action were also preferred.

Experience with Drug Under Review

No respondents reported having experience with incobotulinumtoxinA.

Additional Information

Parkinson Québec’s input noted that most respondents had mild to moderate PD. The survey methodology may have recruited a biased and younger set of respondents with fewer significant sialorrhea-associated problems. However, studies of more advanced cases of PD show that severe and frequent sialorrhea can significantly affect patients’ HRQoL. New treatments to effectively manage sialorrhea in individuals with PD are therefore needed.

Clinician Input

Input from Clinical Experts Consulted by CADTH

All CADTH review teams include at least 1 clinical specialist with expertise in the diagnosis and management of the condition for which the drug is indicated. Clinical experts are a critical part of the review team and are involved in all phases of the review process (e.g., providing guidance on the development of the review protocol; assisting in the critical appraisal of clinical evidence; interpreting the clinical relevance of the results; and providing guidance on the potential place in therapy). The following input was provided by 1 clinical specialist with expertise in the diagnosis and management of chronic troublesome sialorrhea associated with neurologic disorders in adults.

Unmet Needs

In contrast to pharmacological or surgical interventions, BoNTs are easy to administer and have limited side effects. Because injection of BoNTs is helpful for symptomatic therapy but is not covered by drug plans, special access must be requested through pharmacare support programs, which may not be approved due to limited resources.

Place in Therapy

IncobotulinumtoxinA would not modify the disease process, but has several advantages compared to other options. This treatment is already part of the current treatment paradigm but cannot be accessed by many patients due to funding limitations. Clinicians can sometimes use extra stock on hand or a support program, but supplies are limited. In some patients it would be reasonable to use sublingual atropine before incobotulinumtoxinA; however, anticholinergics are not appropriate for use in many patients, including those at high risk of urinary retention or confusion.

Patient Population

Patients best suited for treatment with incobotulinumtoxinA would be those with significant disabling sialorrhea (e.g., those who need to use a cloth to wipe away drool and those for whom the condition is socially isolating). Patients would need to attend injection sessions every 3 to 6 months and have no major swallowing difficulties due to risk of worsening. These patients would be identified based on clinical diagnosis. Patients with sialorrhea that is too mild or patients with swallowing difficulties would be least suitable for treatment. Patients with sialorrhea resulting from diverse neurologic conditions may benefit from treatment. Patient selection by a neurologist or physiatrist is essential.

Assessing Response to Treatment

The objective measures used in trials to assess sialorrhea (e.g., radioisotope scanning, collection cups, and counting napkins) are impractical and not used in clinical practice. Response is usually assessed by taking a history. If necessary, a VAS or tools such as the DSFS can be used to assess response. A clinically meaningful response would be an improvement in the patient’s HRQoL relating to the issues described previously. Response can be assessed subjectively at each visit as this is an injectable treatment.

Discontinuing Treatment

Treatment should be discontinued when it is not efficacious or when patients develop AEs such as swallowing problems or dental issues.

Prescribing Conditions

IncobotulinumtoxinA should be administered in a hospital outpatient or community setting. Neurologists or physiatrists would typically be the specialists involved in the care of patients with neurologic conditions and would perform the injections.

Additional Considerations

The clinical expert consulted by CADTH for this review noted that pediatric patients with sialorrhea would not be covered in the indication under review but could benefit from treatment. The sponsor emphasized that incobotulinumtoxinA is indicated for treatment of chronic sialorrhea associated with neurologic disorders in adults only.

Clinician Group Input

No clinician group input was received for this review.

Drug Program Input

The drug programs provide input on each drug being reviewed through CADTH’s reimbursement review processes by identifying issues that may affect their ability to implement a recommendation. The implementation questions and corresponding responses from the clinical expert consulted by CADTH are summarized in Table 5.

Clinical Evidence

The clinical evidence included in the review of incobotulinumtoxinA is presented in 2 sections. The first section, the systematic review, includes pivotal studies provided in the sponsor’s submission to CADTH and Health Canada, as well as those studies that were selected according to an a priori protocol. No indirect evidence met the inclusion criteria for this review. The second section includes an additional relevant study that was considered to address an important gap in the evidence included in the systematic review.

Systematic Review (Pivotal and Protocol-Selected Studies)

Objectives

To perform a systematic review of the beneficial and harmful effects of incobotulinumtoxinA (100 U) intraglandular injection for the treatment of chronic sialorrhea associated with neurologic disorders in adults.

Methods

Studies selected for inclusion in the systematic review included pivotal studies provided in the sponsor’s submission to CADTH and Health Canada, as well as those meeting the selection criteria presented in Table 6. Outcomes included in the CADTH review protocol reflect outcomes considered to be important to patients, clinicians, and drug plans.

The literature search for clinical studies was performed by an information specialist using a peer-reviewed search strategy according to the Peer Review of Electronic Search Strategies checklist.⁴⁵ Published literature was identified by searching the following bibliographic databases: MEDLINE All (1946‒) via Ovid and Embase (1974‒) via Ovid. The search strategy comprised both controlled vocabulary, such as the National Library of Medicine’s MeSH (Medical Subject Headings), and keywords. The main search concepts were Xeomin (incobotulinumtoxinA) and sialorrhea. Clinical trials registries searched included the US National Institutes of Health’s clinicaltrials.gov, WHO’s International Clinical Trials Registry Platform (ICTRP) search portal, Health Canada’s Clinical Trials Database, and the European Union Clinical Trials Register. For the main search, no filters were applied to limit the retrieval by study type. A supplemental search was also run using generic BoNT terms; for this search, search filters were applied to limit retrieval to RCTs or controlled clinical trials. Retrieval was not limited by publication date or by language. For the supplemental search, where possible, retrieval was limited to the human population. Conference abstracts were excluded from all search results.

Appendix 1 provides detailed search strategies. The initial search was completed on March 26, 2021. Regular alerts updated the search until the meeting of the CADTH Canadian Drug Expert Committee on July 21, 2021.

Grey literature (literature that is not commercially published) was identified by searching relevant websites from the Grey Matters: A Practical Tool For Searching Health-Related Grey Literature checklist.⁴⁶ Included in this search were the websites of regulatory agencies (US FDA and European Medicines Agency). Google was used to search for additional internet-based materials. See Appendix 1 for more information on the grey literature search strategy.

These searches were supplemented by reviewing bibliographies of key papers and through contacts with appropriate experts. In addition, the sponsor of the drug was contacted for information regarding unpublished studies. Two CADTH clinical reviewers independently selected studies for inclusion in the review based on titles and abstracts, according to the predetermined protocol. Full-text articles of all citations considered potentially relevant by at least 1 reviewer were acquired. Reviewers independently made the final selection of studies to be included in the review, and differences were resolved through discussion.

A focused literature search for network meta-analyses dealing with sialorrhea was run in MEDLINE All (1946–) on March 25, 2021. No limits were applied to the search. No relevant network meta-analyses were identified in the search.

Findings from the Literature

One study was identified from the literature for inclusion in the systematic review (Figure 1). The included study is summarized in Table 7. A list of excluded studies is presented in Appendix 2.

Figure 1: Flow Diagram for Inclusion and Exclusion of Studies

Description of Studies

The SIAXI (Sialorrhea in Adults Xeomin Investigation, N = 184)^8-10 trial was a prospective, double-blind, placebo-controlled, parallel-group, multi-centre, phase III RCT with an EP of dose-blinded active treatment. The study was funded by the sponsor. The objective of the study was to investigate the efficacy and safety of 2 doses of incobotulinumtoxinA (incobotulinumtoxinA, 75 U or 100 U), compared with placebo in reducing salivary flow rate as well as the frequency and severity of chronic troublesome sialorrhea in adults with various neurologic conditions. The study enrolled adults aged 18 to 80 years with chronic troublesome sialorrhea resulting from PD or AP (multiple system atrophy, corticobasal degeneration, or progressive supranuclear palsy), stroke, or TBI. A total of 216 patients were screened at 12 sites in Germany and 21 sites in Poland. Screened patients were given a screening number and members of the subset of randomized patients were given a randomization number through an interactive web response system. Initially, a target was set of at least 20% of the study population for each of the etiology subgroups (PD/AP, stroke, and TBI) but this was dropped to ensure an adequate sample size. Patients were enrolled from April 11, 2014, to August 26, 2015, and the database was closed on January 4, 2017.

Following screening, the study comprised 4 consecutive 16-week treatment cycles (Figure 2). The length of treatment cycles was based on previous observations that BoNT treatment effects occur approximately 2 weeks post-injection then wane after 8 to 12 weeks.^47,48 At baseline, inclusion and exclusion criteria and randomization were re-checked. Following each incobotulinumtoxinA injection, patients were assessed over the course of the 16 ± 2–week cycle through in-person visits to study sites and telephone calls. Randomization was conducted using an unspecified method implemented in RANCODE version 3.6 and stratified by the etiology (neurologic condition) of sialorrhea. In the study’s MP, which comprised the first treatment cycle, patients were randomized 2:2:1 to receive 75 U incobotulinumtoxinA, 100 U incobotulinumtoxinA, or placebo (saline) via 4 bilateral injections in the parotid (100 U incobotulinumtoxinA dose: 30 U per side; 75 U incobotulinumtoxinA dose: 22.5 U per side) and submandibular (100 U incobotulinumtoxinA dose: 20 U per side; 75 U incobotulinumtoxinA dose: 15 U per side) glands. Eligibility for the EP was based on agreement between patient and investigator regarding continued need for treatment, continued absence of clinically relevant dysphagia (an mROMP score for Section II “Swallowing Symptoms,” Item A of no more than 2 and Item C of no more than 3), absence of AEs, absence of infection and/or inflammation at injection sites, a negative pregnancy test, and low risk of suicidality. For cycles 2 to 4 (EP), patients who received placebo were re-randomized 1:1 to receive either 75 U or 100 U of incobotulinumtoxinA. Those receiving active treatment in the MP remained in the same dose group unless dose reduction was required due to AEs. A third, dose-reduced group (56 U, corresponding to a 25% reduction from 75 U) was planned for patients in the 75 U group in the MP who experienced AEs but was not used. All participants were blinded to dose level. The fourth injection occurred 48 ± 6 weeks after the first, making for a total study duration of 64 ± 6 weeks (not including the screening period).

Efficacy outcomes for the incobotulinumtoxinA 75 U arms of both the MP and EP are not presented in this report because these data are not aligned with the Health Canada–approved dose (100 U).

Figure 2: SIAXI Study Design

NT 201 = incobotulinumtoxinA; SCR = screening; t = telephone.

* If dose reduction for patients receiving 75 U in the MP due to AEs became necessary in the EP, then a third dose group was planned (56 U, corresponding to a 25% reduction of 75 U). Dose reduction was to be allowed only once in a single step at the third or fourth injection as an alternative to withdrawal.

Source: SIAXI Clinical Study Report.⁸

Populations

Inclusion and Exclusion Criteria

Key inclusion and exclusion criteria for the SIAXI study are summarized in Table 7. Patients (age 18 to 80 years inclusive) with chronic troublesome sialorrhea resulting from PD/AP, stroke, or TBI (onset ≥ 6 months before screening) were included. All diagnoses of PD were made according to the UK PD Brain Bank criteria.⁴⁹ It was initially planned that 20% of patients be recruited from each etiology subgroup; however, this requirement was dropped for feasibility to achieve the necessary sample size. Chronic troublesome sialorrhea was defined as sialorrhea lasting for 3 months or longer, with a DSFS sum score of at least 6, DSFS scores for both severity and frequency of at least 2, and mROMP Section III “Drooling,” Item A score of 3 or greater at both screening and baseline. Troublesome sialorrhea had to persist after stabilization and/or optimization of medications influencing sialorrhea (e.g., anti-parkinsonian medication). Only patients with no swallowing difficulties were included (an mROMP score for Section II “Swallowing Symptoms,” Item A of no more than 2 and Item C of no more than 3 at screening and baseline). Patients with non-neurologic causes of sialorrhea, generalized neuromuscular junction disorders of muscle activity (e.g., myasthenia gravis), extremely poor dental condition or oral hygiene, history of aspiration pneumonia, previous and/or current infections or tumours of the salivary glands or injection sites, or any concurrent diseases or conditions (hematological, hepatic, renal, gastrointestinal, endocrine, pulmonary, musculoskeletal, or psychiatric) judged by investigators as potential risks were excluded. Patients unable to tolerate keeping cotton rolls in their mouths for 5 minutes (as judged by investigators), patients unable to open their mouths voluntarily, patients at significant risk of suicidality (based on investigator judgment or the electronic Columbia Suicide Severity Rating Scale⁵⁰) were also excluded. Pregnant women were excluded and contraceptive use was required for women of child-bearing potential. Patients with unstable concomitant medications influencing sialorrhea (e.g., anticholinergics for parkinsonism), other concomitant medications known to cause hypersalivation (e.g., clozapine), prior or planned surgery or irradiation to treat sialorrhea, recent treatment with BoNTs (≤ 1 year for sialorrhea and ≤ 14 weeks for other indications), or known hypersensitivities to BoNTs were excluded. Detailed information on inclusion and exclusion criteria related to prior and concomitant medications and therapies is provided in the Interventions section.

Baseline Characteristics

The mean age of the study population in the MP was 65.2 years (SD = 11.4 years) (Table 8). Most patients were men (70.7%) and nearly all were White (99.5%). The most common etiology of sialorrhea was PD (70.7%) followed by stroke (19.0%), AP (8.7%), and TBI (2.7%). Among patients with PD/AP, the mean unified Parkinson disease rating scale (UPDRS)⁵¹ section III “motor examination” score was 31.2 (SD = 15.6). Among patients with stroke or TBI, the mean modified motor assessment scale (MMAS)⁵² score (left/right) was 38.7/39.9 (SD = 10.2/9.9). The mean duration of sialorrhea was 32.7 months (SD = 34.5 months). Baseline mean (and SD) uSFR, DSFS sum scores, DSFS severity scores, DSFS frequency scores, and mROMP drooling scores were well balanced between the placebo and the incobotulinumtoxinA 100 U arms. Baseline demographic and clinical characteristics were generally well balanced between study arms.

Most patients had at least 1 medical history finding; the most common finding, beyond the underlying neurologic condition, was DBS (placebo: 13.9%; incobotulinumtoxinA 100 U: 24.3%). Patients had a variety of concomitant diseases that occurred with similar frequency across arms; however, dysarthria (placebo: 8.3%; incobotulinumtoxinA 100 U: 20.3%), dysphonia (placebo: 2.8%; incobotulinumtoxinA 100 U: 6.8%), and dysphagia (placebo: 5.6%; incobotulinumtoxinA 100 U: 8.1%) were more common in the incobotulinumtoxinA 100 U arm compared with the placebo arm. Compared with the placebo arm, more patients in the incobotulinumtoxinA 100 U arm had a treatment history of DBS (placebo: 13.9%; incobotulinumtoxinA 100 U: 24.3%).

The demographic characteristics of the 173 patients participating in the EP were similar to those in the MP (Appendix 3). Data on severity of sialorrhea (uSFR, DSFS sum scores and subscores, and mROMP drooling scores) at the beginning of the EP are not presented in this table as these values were identical to the week 16 efficacy data for the MP. Baseline demographic and clinical characteristics were again well balanced between arms. Medical history was not evaluated separately for the EP. Concomitant diseases and medications in the EP were similar to those in the MP.

Interventions

In the MP, patients were randomized 2:2:1 to receive 1 of 2 doses of incobotulinumtoxinA (75 U or 100 U) or placebo (physiologic saline) via 4 intraglandular injections administered on a single day by study personnel at the baseline visit. To maintain the double-blind study status, similarly marked vials were provided containing defined amounts of incobotulinumtoxinA that, when resuspended in 4 mL of physiologic saline, yielded solutions of 50 U/mL incobotulinumtoxinA (100 U dose group) or 37.5 U/mL incobotulinumtoxinA (75 U dose group). The placebo vials contained an excipient only (human serum albumin and sucrose). Identical volumes were injected in all patients (0.6 mL bilaterally per parotid gland and 0.4 mL bilaterally per submandibular gland). Injections were guided either by ultrasound or using anatomic landmarks; individual patients consistently received injections using 1 or the other technique. Based on input from the FDA, at least 50% of patients were targeted to receive ultrasound-guided injections, and investigators received special training for this technique. The EP involved 3 additional 16-week treatment cycles with injections administered identically as in the MP, unless dose reduction was required. If patients met the eligibility criteria for the next treatment cycle, the injection could take place at the assessment visit; otherwise, the next injection could be postponed for up to 2 weeks.

Patients treated with BoNTs for sialorrhea within the past 1 year or for other indications within 14 weeks were excluded; treatment with BoNTs (other than protocol therapy) was forbidden during the screening period and during the entire study period. Concomitant use of aminoglycoside antibiotics, curare-like agents, or other medications that could interfere with neuromuscular function was forbidden. Pharmacological treatments for sialorrhea and medications known to cause hypersalivation (e.g., clozapine) were forbidden from 4 weeks before baseline throughout the MP. Concomitant medications influencing sialorrhea (e.g., anticholinergics) were forbidden from 4 weeks before baseline throughout the MP, except for drugs taken at a stable dosage throughout this period. Changes in dosing of medications to treat PD/AP were forbidden from 4 weeks before screening throughout the MP. Salivary gland surgery or irradiation was forbidden at any time before or during the entire study period including screening. Prior (≤ 6 months) functional neurosurgery (e.g., pallidotomy, DBS) or functional neurosurgery planned during the MP was forbidden, as was any other prior (≤ 3 months) or planned (during the MP) surgery, except for minor surgery outside the head and neck region. Anticoagulants were forbidden, but Aspirin and platelet aggregation inhibitors were allowed. Use of illegal and legal drugs (except for alcohol/tobacco) for recreational purposes was forbidden. Excessive use of alcohol and/or tobacco (as judged by the investigator) was forbidden. Smoking was disallowed within 1 hour of assessments. During protocol therapy, a dry mouth could be eased by ice chips, cold water, sugar-free chewing gum, sour sugar-free sweets, sprays, gels, and mouth rinses. Patients were advised to try saliva substitutes and choose their preferred agent, but to avoid use of these agents before and during saliva collection.

Nearly all patients received concomitant medications, including dopaminergic agents for PD (placebo: 77.8%; incobotulinumtoxinA 100 U: 79.7%) during the MP (Table 9), but few received anticholinergic agents (placebo: 0%; incobotulinumtoxinA 100 U: 2.7%). Compared with the placebo arm, more patients in the incobotulinumtoxinA arm received psychoanaleptics (placebo: 19.4%; incobotulinumtoxinA 100 U: 36.5%) and fewer received diuretics (placebo: 27.8%; incobotulinumtoxinA 100 U: 14.9%), calcium channel blockers (placebo: 27.8%; incobotulinumtoxinA 100 U: 14.9%), or cardiac therapy (placebo: 19.4%; incobotulinumtoxinA 100 U: 6.8%). Compared with the placebo arm, more patients in the NT201 arm reported concomitant DBS (placebo: 13.9%; incobotulinumtoxinA 100 U: 24.3%).

Concomitant medications received during the EP were similar to those received during the MP (Appendix 3).

Stopping criteria included withdrawal of informed consent, pregnancy, AEs for which treatment continuation constituted an unacceptably high risk (described in the Outcomes sections), and positive suicidality assessment (a “yes” response to sections 4 and 5 of the electronic Columbia Suicide Severity Rating Scale, any suicidal behaviour, or any non-suicidal self-injurious behaviour). Patients who were withdrawn due to AEs were to be followed up until the end-of-study visit or 28 days after the AE occurred, whichever was later. In addition, dose reductions of incobotulinumtoxinA due to AEs were allowed once and only once at either the third or the fourth injections in the EP; if the AE did not recur, the full dose could be given in the fourth injection for patients receiving reduced doses in the third injection. Dose reduction included planned movement of patients from the 100 U to the 75 U dose group as well as, potentially, movement to a third dose-reduced group (56 U, corresponding to a 25% reduction from 75 U). All dose reductions were based on 25% reductions in the injection volume.

Outcomes

A list of efficacy end points identified in the CADTH review protocol that were assessed in the clinical trials included in this review is provided in Table 10. A detailed discussion and critical appraisal of the outcome measures is provided in Appendix 4. No evidence was available to inform the validity, reliability, responsiveness to change, or MID for any of the outcome measures used in the SIAXI trial.

All primary and secondary outcomes were evaluated in the MP. In the EP, all outcomes were exploratory and were measured as in the MP. In patients with PD/AP and “on-off” motor fluctuations, all efficacy outcomes were assessed in comparable “on” states (typically 60 to 180 minutes after taking anti-parkinsonian medication).

The uSFR was evaluated during onsite assessment visits at screening, baseline, and various times post-injection by direct saliva collection using the swab method. A negative change in the uSFR represents a reduction in salivary flow and therefore an improvement in sialorrhea. Four cotton rolls were placed in the patient’s mouth (1 between the cheek and gums and 1 between the tongue and gums on each side) for 5 minutes, then weighed; after 30 minutes, the procedure was repeated, and the 2 measurements were averaged. In the MP, change in uSFR was evaluated from baseline to all visits. In the EP, change in uSFR was evaluated from baseline to all visits as well as from each injection to the respective assessment visits and the end-of-cycle visit.

Scores for the GICS were evaluated by patients and by caregivers independently on a 7-point Likert scale (−3, very much worse; −2, much worse; −1, minimally worse; 0, no change; 1, minimally improved; 2, much improved; and 3, very much improved) at various times post-injection via self-reported questionnaires or phone interviews. The recall period for GICS was since before the previous injection. If patients were unable to answer, caregiver responses were used in place of the patient’s responses. The number of responders (defined as patient GICS response of at least 1, minimally improved; denominator all patients receiving injections) was calculated at all post-injection assessments. The rationale for defining responders using a threshold of a GICS score of at least 1 was not stated.

Sialorrhea and speech symptoms were evaluated by patients at screening, baseline, and at various times post-injection using the self-administered mROMP inventory containing 8 and 9 questions in the speech and drooling domains, respectively; each domain was evaluated on a 5-point Likert scale yielding total scores from 8 to 40 and 9 to 45, respectively. Changes from study baseline were evaluated.

In addition, sialorrhea severity and frequency were rated on the DSFS by study investigators at screening, baseline, and various times post-injection using 5- and 4-point Likert scales, respectively; the resulting DSFS sum score ranged from 2 to 9. The recall periods for both the mROMP and DSFS were over the previous 1 week. Changes from study baseline were evaluated.

The HRQoL was assessed by patients at baseline and at various times post-injection using the EQ-5D-3L;⁵³ patients descriptively rated each of 5 dimensions (mobility, self-care, usual activities, pain/discomfort, and anxiety/depression) on a 3-point scale (1, no problems; 2, some problems; or 3, extreme problems) and rated their overall health using a VAS ranging from 0 (worst imaginable health state) to 100 (best imaginable health state). Changes from study baseline were evaluated.

Harms outcomes included treatment-emergent AEs, SAEs, WDAEs, and AESIs. All harms outcomes were analyzed for the MP, the full EP, and each cycle in the EP separately. AEs were defined as any untoward medical occurrence and were coded according to Medical Dictionary for Regulatory Activities version 18.1. Incidence was calculated at the system organ class and preferred term (PT) levels. Adverse events in the MP were defined as AEs with onset or worsening after the first injection up to the cycle 2 injection (the first injection of the EP), 176 weeks after the first injection, or the last study visit, whichever was later. Adverse events in the EP were defined as AEs with onset or worsening after the cycle 2 injection (the first injection of the EP) up to 16 weeks after the cycle 4 injection (the last injection of the EP) or the date of last study visit, whichever was later. Serious AEs were defined as an untoward medical occurrence that was life-threatening, requires inpatient hospitalization or prolongation of hospitalization, results in persistent or significant disability and/or incapacity, results in death, is a congenital abnormality, or consists of any other medically important condition as per the International Conference on Harmonisation's E2A guidelines.⁵⁴ Adverse events of special interest (AESI) were defined as in Table 11 to capture all AE PTs potentially reflective of toxin spread. In addition, a dry mouth that was severe, serious, or irreversible was reported as an AESI. At post-injection visits, a focused physical exam was used to assess AESIs for potential toxin spread; if necessary, a full physical exam was administered.

Statistical Analysis

Power Calculation

The planned sample size for co-primary analysis of uSFR and GICS in SIAXI was based on a study of rimabotulinumtoxinB.⁴⁷ For uSFR, the study found mean changes of −0.381 (SD = 0.34) for the rimabotulinumtoxinB 2,500 U group and −0.052 (0.33) for the placebo group 4 weeks post-injection. Assuming a dropout rate of 5% by week 4, a mean change in uSFR from baseline to week 4 of −0.362 (SD = 0.341) was calculated for the incobotulinumtoxinA 100 U group, with a corresponding change of −0.049 (SD = 0.322) for the placebo group. Based on these assumptions and a 2:2:1 randomization ratio, 46 patients in each incobotulinumtoxinA treatment arm and 23 patients in the placebo arm would provide 95% power to show a statistically significant difference between either of the incobotulinumtoxinA groups and placebo (2-sided Satterthwaite t-test, alpha = 0.05). The total sample size required was at least 115 patients.

For the GICS, the study of rimabotulinumtoxinB⁴⁷ found mean changes in the Clinical Global Impression score of 2.0 (SD = 1.04)⁵⁵ for the rimabotulinumtoxinB 2,500 U group and 4.1 (SD = 1.03) for the placebo group 4 weeks post-injection (smaller values reflect greater improvement). Assuming a dropout rate of 5% by week 4 and converting to GICS as (−Clinical Global Impression − change + 4), a mean GICS score at week 4 of 2.100 (SD = 1.103) was calculated for the incobotulinumtoxinA group, with a corresponding change of 1.195 (SD = 1.344) in the placebo arm. Based on these assumptions and a 2:2:1 randomization ratio, 20 patients in each incobotulinumtoxinA treatment arm and 10 patients in the placebo arm would provide 95% power to show a statistically significant difference between either of the incobotulinumtoxinA groups and placebo (2-sided Satterthwaite t-test, alpha = 0.05). The total sample size required was at least 50 patients.

Statistical Models

Two-sided hypotheses were used to test for between-treatment differences. Methods used for statistical analysis of primary and secondary outcomes are summarized in Table 12 and those used for analysis of all outcomes are shown in Appendix 3. For the co-primary outcomes (change in uSFR from baseline to week 4 in the MP and patient GICS score at week 4 in the MP), an MMRM analysis (2-sided, alpha = 0.05) comparing LSMs was used for the confirmatory analysis of differences between the incobotulinumtoxinA treatment and placebo groups. The dependent variables were change from baseline in uSFR or GICS score. The independent variables were treatment group, etiology subgroup, use of ultrasound, country, and gender as fixed factors, visit × treatment as an interaction term, and visit as a repeated factor. To adjust for baseline status, the MMRM analysis of change in uSFR included baseline uSFR as a covariate and the MMRM analysis of GICS scores included baseline DSFS as a covariate. For the co-primary outcomes, a sequence test procedure was used to control type I error when comparing the LSMs from the MMRM analysis for the 2 incobotulinumtoxinA treatment groups versus placebo at week 4. Because of the hierarchical test procedure, no additional adjustment for multiplicity was performed. First, the hypotheses of equivalence versus non-equivalence for change in uSFR and GICS score between the incobotulinumtoxinA 100 U arm and placebo were tested. Subsequently, and only if both previous equivalence hypotheses could be rejected, the hypotheses of equivalence versus non-equivalence for change in uSFR and GICS between the incobotulinumtoxinA 75 U and placebo arms were tested. The confirmatory analyses were performed in the full analysis set (FAS), and the MMRM approach accounted for missing values by assuming they were missing at random. Sensitivity analyses were performed in the same manner using the per-protocol set (PPS). Additional sensitivity analyses were performed for the FAS and PPS using the baseline observation carried forward (BOCF) approach for uSFR and imputing missing GICS entries at week 4 as “no change,” and without replacing missing data (observed cases). For the BOCF and observed cases analyses, analysis of covariance (ANCOVA) models were used without visit × treatment as an interaction term and without visit as a repeated factor. Finally, a non-parametric Wilcoxon rank sum test (FAS and PPS, using BOCF and observed cases analysis) was used as a sensitivity analysis of the change in uSFR to account for potential deviation from normality.

Secondary efficacy outcomes (change in uSFR from baseline to weeks 8 and 12 in the MP and patient GICS score at weeks 1, 2, 8, and 12 in the MP) were analyzed in the same manner as the co-primary outcomes in both the FAS and PPS. These tests were descriptive, not adjusted for multiplicity, and interpreted in an exploratory manner. Comparisons were between the incobotulinumtoxinA 100 U and placebo arms as well as the incobotulinumtoxinA 75 U and placebo arms. There was no hierarchical test procedure or strategy for multiplicity control.

Exploratory efficacy outcomes in the MP were evaluated in the FAS, except for change in uSFR at week 16 and patient GICS score at week 16, which were evaluated in both the FAS and PPS. Change in uSFR at week 16 and patient GICS score at week 16 were analyzed in the same manner as the co-primary outcomes (BOCF and observed cases). Comparison of the number of responders based on patient GICS entries across treatment groups were performed in a manner similar to that used in the primary efficacy analysis, except using logistic regression models with response to treatment as the dependent variable and treatment group, use of ultrasound, etiology subgroup, gender, and country as factors. To account for baseline status, the models included baseline DSFS score as a covariate. Odds ratios and 95% Wald CIs were calculated. Analysis was performed in the FAS without imputation of missing values and imputing missing GICS entries as nonresponders. If logistic regression models were not estimable, descriptive summary statistics were provided. For analysis of GICS responders at week 16, only summary statistics, 95% CIs, and P values from Fisher’s exact tests or chi-square tests were reported, both without imputation of missing values and imputing missing GICS entries as nonresponders. Percentages were based on all patients (BOCF) or patients with data at the respective visit (observed cases). Carer GICS scores at all time points were summarized using descriptive statistics and frequency tables. Changes in DSFS sum scores from baseline to weeks 4, 8, and 12 were analyzed as per the primary efficacy analysis. For changes in DSFS sum scores from baseline to week 16 and changes in DSFS subscores from baseline to weeks 4, 8, 12, and 16, descriptive statistics were provided. For changes in mROMP speech and drooling scores from baseline to all time points, analyses were performed of observed cases with the BOCF applied to single questions; descriptive statistics were provided. Single items from the EQ-5D-3L were analyzed using frequency tables and shift tables for changes from baseline, and the EQ VAS was analyzed using summary statistics for all visits and changes from baseline; analysis was performed without imputation of missing values. None of the exploratory analyses were controlled for multiplicity.

Exploratory efficacy outcomes in the EP were evaluated in the safety evaluation set, extension period (SES-EP) and missing data were not imputed. None of the exploratory analyses of the EP were controlled for multiplicity.

Subgroup analyses were performed as described previously for selected outcomes (uSFR, GICS, and DSFS). All subgroup analyses were exploratory and not controlled for multiplicity. Among others, the SIAXI trial pre-specified a subgroup analysis by etiology of sialorrhea (PD/AP, stroke, or TBI) that was identified as of interest in the CADTH review protocol. This subgroup was well balanced in the treatment arms. A post hoc subgroup analysis of efficacy outcomes was conducted by severity of sialorrhea (uSFR ≤ or > median uSFR at baseline), which was also identified as of interest in the CADTH review protocol but not presented in this report due to its post hoc nature. No subgroup analyses by severity of underlying neurologic condition were conducted.

Analysis Populations

The safety evaluation set, main period (SES-MP) was defined as the set of all patients who received study medication (incobotulinumtoxinA or placebo) during the MP. The FAS was defined as the subset of patients within the SES-MP for whom the primary efficacy outcome was measured (i.e., all patients who were treated and had uSFR measured at baseline; no post-baseline measurement was required for inclusion in the FAS). The PPS was defined as the subset of patients within the FAS without major protocol deviations. If patients were not treated according to the randomization list, analyses of the SES-MP and FAS were by intention to treat.

In the EP, the SES-EP was defined as the set of all patients who received study medication (incobotulinumtoxinA) at least once during the EP. Because medication was assigned for each cycle individually, treatment not according to the randomization list could occur in some but not all cycles. For efficacy outcomes over the entire EP, analysis was by intention to treat even if treatment deviated from the randomization list. For efficacy outcomes for each treatment cycle, analysis was by intention to treat. Patients receiving reduced doses were analyzed as randomized. For harms outcomes, analysis was based on treatment received.

Efficacy analyses in the MP were based on the FAS, and for sensitivity analyses, the PPS. Efficacy analyses in the EP were based on the SES-EP. In the EP, analysis was performed by combined MP/EP treatment group: placebo (MP) + incobotulinumtoxinA 100 U (EP); placebo (MP) + incobotulinumtoxinA 75 U (EP); incobotulinumtoxinA 100 U (MP) + incobotulinumtoxinA 100 U (EP); and incobotulinumtoxinA 75 U (MP) + incobotulinumtoxinA 75 U (EP). Analyses were also performed by treatment group in the EP (incobotulinumtoxinA 100 U or 75 U).

Results

Patient Disposition

A total of 216 patients were screened, of whom 184 were randomized and treated in the MP (incobotulinumtoxinA 100 U: N = 74; incobotulinumtoxinA 75 U: N = 74; placebo: N = 36) (Table 13). Among the 32 screening failures, the most common causes were violation of inclusion and/or exclusion criteria (19; 59.4%), withdrawal of consent (7; 21.9%), inability to participate in study follow-up visits (3; 9.4%), and physician decision or unknown (3; 9.4%). The most common inclusion and/or exclusion criterion violated was extremely poor oral and/or dental condition (9; 28.1%).

The SES-MP and FAS of the MP comprised all 184 patients. The PPS comprised 165 patients (89.7%) (incobotulinumtoxinA 100 U: 85.1%; incobotulinumtoxinA 75 U: 93.2%; placebo: 91.7%). Overall, 11 patients (6.0%) discontinued the study prematurely (incobotulinumtoxinA 100 U: 2.7%; incobotulinumtoxinA 75 U: 6.8%; placebo: 11.1%). The most common reason for discontinuation was withdrawal by patient (4.3%). A higher proportion of patients in the placebo arm discontinued the MP (11.1%) compared with the incobotulinumtoxinA 100 U arm (2.7%).

A total of 173 patients completed the MP and entered the EP (incobotulinumtoxinA 100 U: N = 89; incobotulinumtoxinA 75 U: N = 84) (Table 14). These patients constituted the SES-EP. Overall, 22 patients (12.7%) discontinued the EP (14 patients [15.7%] in the incobotulinumtoxinA 100 U arm and 8 patients [9.5%] in the incobotulinumtoxinA 75 U arm). Among discontinuations during the EP, 10 (5.8%) occurred in cycle 2, 6 (3.5%) occurred during cycle 3, and 6 (3.5%) occurred during cycle 4. The most common reasons given for study discontinuation were AEs (6.9%) and withdrawal by patient (6.9%). A trend toward higher study discontinuation was evident in the incobotulinumtoxinA 100 U arm.

A total of 19 patients (10.3%) experienced major protocol deviations in the MP and were excluded from the PPS (Table 15). A higher number of major protocol deviations (11; 14.9%) occurred in the incobotulinumtoxinA 100 U arm compared with the placebo arm (3; 8.3%). The most common major protocol deviation was concomitant medication influencing sialorrhea (7; 3.8%). Four patients (5.4%) in the incobotulinumtoxinA 100 U arm but no patients in the placebo arm had no chronic troublesome sialorrhea as measured by the DSFS and mROMP drooling scores. All protocol deviations in the EP were classified as minor.

Exposure to Study Treatments

Because injections were administered by investigators at study site visits, adherence was 100% in both the MP and EP. Injection was ultrasound-guided in more than 50% of patients treated in both the MP (Table 16) and the EP (Table 17).

Therapy received was according to the randomization schedule for 100% of patients apart from discontinuations, dose reductions, or erroneous treatment. In the MP, all patients received the study drug or placebo as intended; the mean injection cycle length was 16.56 (SD = 1.19) weeks in the placebo arm, 16.13 (SD = 3.11) weeks in the incobotulinumtoxinA 75 U arm, and 16.13 (SD = 2.20) weeks in the incobotulinumtoxinA 100 U arm. In the EP, greater than 97% of patients received the intended dose of incobotulinumtoxinA; only 7 of 173 patients (4.0%) deviated from the planned randomized treatment, 5 (2.9%) because of dose reductions due to AEs and 2 (1.1%) because of missed cycles or erroneous treatment. The mean lengths of injection cycles 2, 3, and 4 were 16.18 week (SD = 1.95), 16.22 weeks (SD = 1.23), and 15.77 weeks (SD = 1.99), respectively, in the incobotulinumtoxinA 75 U arm, and 16.13 weeks (SD = 1.73), 16.21 weeks (SD = 1.29), and 16.46 weeks (SD = 1.92), respectively, in the incobotulinumtoxinA 100 U arm.

Efficacy

Only those efficacy outcomes and analyses of subgroups identified in the review protocol are reported. Appendix 3 provides detailed efficacy data.

Reduction in Salivary Production

Main Period

As part of the co-primary efficacy analysis, change in the uSFR from baseline to week 4 of the MP was analyzed in each treatment group (Table 18). The LSM change in uSFR in the incobotulinumtoxinA 100 U arm was −0.13 g/min (SE = 0.026; 95% CI, −0.18 to −0.08) compared to −0.04 g/min (SE = 0.033; 95% CI, −0.11 to 0.03) in the placebo arm. There was a statistically significant LSM difference in uSFR in favour of the incobotulinumtoxinA 100 U arm compared with the placebo arm of −0.09 g/min (SE = 0.031; 95% CI, −0.15 to −0.03) (P = 0.004).

Sensitivity analyses of change in the uSFR from baseline to week 4 of the MP in the PPS, and in the FAS and PPS using ANCOVA models and the Wilcoxon rank sum test, produced similar results (Appendix 3). A planned subgroup analysis by the etiology of sialorrhea suggested that the treatment effects of incobotulinumtoxinA 100 U on change in uSFR from baseline to week 4 occurred in both patients with PD/AP and stroke, while the number of patients with TBI was too small to reach any conclusion (Appendix 3). It was unclear from this analysis whether patients with PD/AP and stroke saw similar magnitudes of treatment effects. The subgroup analysis of change in uSFR from baseline to week 4 by baseline severity of sialorrhea was conducted post hoc and therefore was not presented in this report.

Change in the uSFR from baseline to weeks 8 and 12 of the MP was analyzed as a secondary outcome in each treatment group (Table 19). This analysis was not controlled for multiplicity. At week 8, the LSM change in uSFR in the incobotulinumtoxinA 100 U arm was −0.13 g/min (SE = 0.026; 95% CI, −0.19 to −0.08) compared to −0.02 g/min (SE = 0.033; 95% CI, −0.08 to 0.05) in the placebo arm; the LSM difference in uSFR between the incobotulinumtoxinA 100 U arm and the placebo arm was −0.12 g/min (SE = 0.030; 95% CI, −0.18 to −0.06). At week 12, the LSM change in uSFR in the incobotulinumtoxinA 100 U arm was −0.12 g/min (SE = 0.026; 95% CI, −0.17 to −0.07) compared to −0.03 (SE = 0.033; 95% CI, −0.09 to 0.04) in the placebo arm; the LSM difference in uSFR between the incobotulinumtoxinA 100 U arm and the placebo arm was −0.09 g/min (SE = 0.031; 95% CI, −0.15 to −0.03).

Sensitivity analyses of change in the uSFR from baseline to weeks 8 and 12 of the MP in the PPS, and in the FAS and PPS using ANCOVA models and the Wilcoxon rank sum test, showed similar results, with a trend toward lower differences between the incobotulinumtoxinA 100 U and placebo groups at week 12 (data not shown in this report).

The change in the uSFR from baseline to week 16 of the MP was an exploratory outcome (Table 20). This analysis was not controlled for multiplicity. The LSM change in uSFR in the incobotulinumtoxinA 100 U arm was −0.11 g/min (SE = 0.027; 95% CI, −0.17 to −0.06) compared to −0.01 g/min (SE = 0.035; 95% CI, −0.08 to 0.05) in the placebo arm. The LSM difference in uSFR between the incobotulinumtoxinA 100 U arm and the placebo arm was −0.10 g/min (SE = 0.033; 95% CI, −0.17 to −0.04).

Extension Period

The change in uSFR from baseline to each visit in the EP, as well as change from the prior injection, was assessed in an exploratory fashion (Table 21). This analysis was not controlled for multiplicity. Mean decreases (improvements) in uSFR from study baseline and from the baseline for each cycle were observed at each visit in the EP for the incobotulinumtoxinA 100 U treatment arm (Table 35). The magnitude of changes in uSFR from baseline was larger than the magnitude of changes from any of the individual cycles in the EP.

Mean uSFR values over the entire course of the study at all assessment visits are shown graphically in Appendix 3.

Change in Patient and Caregiver Perceived Frequency and Severity of Sialorrhea

Main Period

Patient GICS Score

As part of the co-primary efficacy analysis, patient GICS scores at week 4 of the MP was analyzed in each treatment group (Table 22). The LSM patient GICS score in the incobotulinumtoxinA 100 U arm was 1.25 (SE = 0.144; 95% CI, 0.97 to 1.53) compared to 0.67 (SE = 0.186; 95% CI, 0.30 to 1.04) in the placebo arm. The LSM difference in GICS scores was statistically significantly in favour of the incobotulinumtoxinA 100 U arm compared with the placebo arm of 0.58 (SE = 0.183; 95% CI, 0.22 to 0.94; P = 0.002).

Sensitivity analyses of patient GICS scores at week 4 of the MP in the PPS, and in the FAS and PPS using ANCOVA models, produced similar results (Appendix 3). A planned subgroup analysis by the etiology of sialorrhea suggested that the treatment effects of incobotulinumtoxinA 100 U on both co-primary efficacy outcomes occurred in both patients with PD/AP and stroke, while the number of patients with TBI was too small to reach a conclusion (Appendix 3). It was unclear from this analysis whether patients with PD/AP and stroke saw similar magnitudes of treatment effects. The subgroup analysis of patients GICS at week 4 by baseline severity of sialorrhea was conducted post hoc and therefore is not presented in this report.

The LSM difference in GICS scores at week 4 of the MP between the incobotulinumtoxinA 100 U and the placebo treatment groups in the primary efficacy analysis resulted from decreased numbers of patients treated with incobotulinumtoxinA 100 U responding with “no change in function” (placebo: 50.0%; incobotulinumtoxinA 100 U: 20.3%) and increased numbers of patients selecting “minimally improved function” or “much improved function” (Table 23). At week 4, the category showing the largest difference between treatment groups was a GICS score of 2, “much improved function” (placebo: 11.1%; incobotulinumtoxinA 100 U: 28.4%).

Patient GICS scores at weeks 1, 2, 8, and 12 of the MP were analyzed as a secondary outcome in each treatment group (Table 24). This analysis was not controlled for multiplicity. At week 1, the LSM GICS score in the incobotulinumtoxinA 100 U arm was 0.96 (SE = 0.133; 95% CI, 0.70 to 1.23) compared to 0.67 (SE = 0.170; 95% CI, 0.34 to 1.00) in the placebo arm; the LSM difference in GICS score between the incobotulinumtoxinA 100 U arm and the placebo arm was 0.29 (SE = 0.158; 95% CI, −0.02 to 0.60). At week 2, the LSM GICS score in the incobotulinumtoxinA 100 U arm was 1.11 (SE = 0.139; 95% CI, 0.84 to 1.38) compared to 0.83 (SE = 0.178; 0.47 to 1.18) in the placebo arm; the LSM difference in GICS between the incobotulinumtoxinA 100 U arm and the placebo arm was 0.29 (SE = 0.171; 95% CI, −0.05 to 0.62). At week 8, the LSM GICS score in the incobotulinumtoxinA 100 U arm was 1.30 (SE = 0.148; 95% CI, 1.01 to 1.59) compared to 0.47 (SE = 0.192; 95% CI, 0.09 to 0.84) in the placebo arm; the LSM difference in GICS score between the incobotulinumtoxinA 100 U arm and the placebo arm was 0.84 (SE = 0.192; 95% CI, 0.46 to 1.21). At week 12, the LSM GICS score in the incobotulinumtoxinA 100 U arm was 1.21 (SE = 0.152; 95% CI, 0.91 to 1.51) compared to 0.56 (SE = 0.197; 95% CI, 0.17 to 0.95) in the placebo arm; the LSM difference in GICS scores between the incobotulinumtoxinA 100 U arm and the placebo arm was 0.65 (SE = 0.201; 95% CI, 0.25 to 1.04).

Sensitivity analyses of patient GICS scores at weeks 1, 2, 8, and 12 of the MP in the PPS, and in the FAS and PPS using ANCOVA models and the Wilcoxon rank sum test, showed similar results, with a trend toward smaller differences between the incobotulinumtoxinA 100 U and placebo groups at week 12 (data not shown in this report).

The LSM differences in patient GICS at weeks 8 and 12 between the incobotulinumtoxinA 100 U and the placebo treatment groups in the secondary efficacy analysis resulted from decreased numbers of patients treated with incobotulinumtoxinA 100 U responding with “no change in function” (week 8, placebo: 61.1%; incobotulinumtoxinA 100 U: 18.9%; week 12, placebo: 50.0%; incobotulinumtoxinA 100 U: 20.3%) and increased numbers of patients selecting “minimally improved function” or “much improved function” (Table 25). At weeks 8 and 12, the category showing the largest difference between treatment groups was a GICS score of 2, “much improved function” (week 8, placebo: 8.3%; incobotulinumtoxinA 100 U: 31.1%; week 12, placebo: 11.1%; incobotulinumtoxinA 100 U: 31.1%).

Patient GICS scores at week 16 of the MP were analyzed as an exploratory outcome (Table 26). This analysis was not controlled for multiplicity. The LSM GICS score in the incobotulinumtoxinA 100 U arm was 0.93 (SE = 0.152; 95% CI, 0.63 to 1.23) compared to 0.41 (SE = 0.199; 95% CI, 0.02 to 0.080) in the placebo arm. The LSM difference in GICS score between the incobotulinumtoxinA 100 U arm and the placebo arm was 0.52 (SE = 0.203; 95% CI, 0.12 to 0.92).

GICS Responder Rate

The proportion of responders (GICS score ≥ 1) based on patient GICS responses at all assessment visits and telephone interviews during the MP was analyzed in an exploratory fashion (Table 27). This analysis was not controlled for multiplicity. The proportion of responders ranged from 59.5% (week 1) to 76.4% (week 8) in the incobotulinumtoxinA 100 U group and from 28.6% (week 8) to 48.6% (week 2) in the placebo group. The largest differences in GICS response rates between the incobotulinumtoxinA 100 U arm and placebo arm were observed at weeks 4 (72.6% versus 44.4%), 8 (76.4% versus 28.6%), and 12 (70.8% versus 38.9%).

Carer GICS Score

Carer GICS entries were analyzed in an exploratory fashion at all assessments and telephone interviews during the MP (Table 28). This analysis was not controlled for multiplicity. As for patient GICS entries, mean carer GICS scores were higher at all time points in the incobotulinumtoxinA 100 U arm compared to the placebo arm, notably at week 4 (1.14 [SD = 0.98] versus 0.48 [SD = 0.83]), week 8 (1.20 [SD = 0.87] versus 0.32 [SD = 0.86]), and week 12 (1.05 [SD = 1.02] versus 0.30 [SD = 0.87]).

mROMP Speech and Drooling Scores

Main Period

Changes in mROMP speech scores from baseline were analyzed in exploratory fashion at weeks 4, 8, 12, and 16 in the MP (Table 29). This analysis was not controlled for multiplicity. Similarly modest improvements in mROMP speech symptoms from baseline were observed in both the incobotulinumtoxinA 100 U group and the placebo group at all time points.

Changes in mROMP drooling scores from baseline were analyzed in an exploratory fashion for each treatment group at weeks 4, 8, 12, and 16 in the MP (Table 30). This analysis was not controlled for multiplicity. The greatest differences in the mean decreases in mROMP drooling scores in the incobotulinumtoxinA 100 U arm compared to the placebo arm were observed at week 8 (−6.58 [SD = 5.90] versus −1.26 [SD = 4.91]) and week 12 (−6.40 [SD = 5.20] versus −1.77 [SD = 4.54]).

Extension Period

Patient GICS Score

Patient GICS scores at each visit in the EP were assessed in an exploratory fashion (Table 31). This analysis was not controlled for multiplicity. Positive GICS values (improvements) were observed for the incobotulinumtoxinA 100 U arm at all visits in the EP. The magnitudes of mean patient GICS responses for the EP was similar to that observed during the MP. Patient GICS tended to peak at weeks 4 and 8 of each cycle (note that GICS referred to changes compared with the previous injection).

GICS Responder Rate

The number of responders based on patient GICS entries (score ≥ 1) at each visit in the EP was assessed in an exploratory fashion (Table 32). This analysis was not controlled for multiplicity. Response rates in the incobotulinumtoxinA 100 U arm ranged from 58.4% to 80.9% throughout the EP and tended to peak at weeks 4 and 8 of each cycle (GICS referred to changes compared with the previous injection).

Carer GICS Score

Carer GICS scores at each visit in the EP were assessed in an exploratory fashion (Table 33). This analysis was not controlled for multiplicity. Positive carer GICS values (improvements) were observed for the incobotulinumtoxinA 100 U arm for all visits in the EP. The magnitudes of carer patient GICS responses for the EP were similar to those observed during the MP. Carer GICS scores tended to peak at weeks 4 and 8 of each cycle (GICS referred to changes compared with the previous injection).

mROMP Speech and Drooling Scores

Changes in mROMP speech scores from study baseline to all assessment visits in the EP were analyzed in an exploratory fashion (Table 34). This analysis was not controlled for multiplicity. The mROMP speech scores for the incobotulinumtoxinA 100 U treatment group remained generally stable over the course of the entire EP.

Changes in mROMP drooling scores from study baseline to all assessment visits in the EP were analyzed in an exploratory fashion (Table 35). This analysis was not controlled for multiplicity. Mean decreases (improvements) in mROMP drooling scores were observed at the week 4 visits for the incobotulinumtoxinA 100 U treatment arm at each injection cycle compared with baseline. Magnitudes of changes in mROMP drooling scores were similar or greater to those observed during the MP.

Change in Clinician-Perceived Frequency and Severity of Sialorrhea

Main Period

Changes from baseline in DSFS sum scores at weeks 4, 8, 12, and 16 were exploratory end points in the MP (Table 36). This analysis was not controlled for multiplicity. At week 4, the LSM change in DSFS sum score in the incobotulinumtoxinA 100 U arm was −1.66 (SE = 0.234; 95% CI, −2.12 to −1.20) compared to −0.50 (SE = 0.296; 95% CI, −1.08 to 0.09) in the placebo arm; the LSM difference in DSFS sum score between the incobotulinumtoxinA 100 U arm and the placebo arm was −1.17 (SE = 0.278; 95% CI, −1.71 to −0.72). At week 8, the LSM change in DSFS sum score in the incobotulinumtoxinA 100 U arm was −1.97 (SE = 0.239; 95% CI, −2.44 to −1.49) compared to −0.68 (SE = 0.305; 95% CI, −1.28 to −0.08) in the placebo arm; the LSM difference in DSFS sum score between the incobotulinumtoxinA 100 U arm and the placebo arm was −1.29 (SE = 0.291; 95% CI, −1.86 to −0.71). At week 12, the LSM change in DSFS sum score in the incobotulinumtoxinA 100 U arm was −1.62 (SE = 0.237; 95% CI, −2.09 to −1.16) compared to −1.00 (SE = 0.301; 95% CI, −1.59 to −0.40) in the placebo arm; the LSM difference in DSFS sum score between the incobotulinumtoxinA 100 U arm and the placebo arm was −0.63 (SE = 0.286; 95% CI, −1.19 to −0.06). At week 16, the LSM change in DSFS sum score in the incobotulinumtoxinA 100 U arm was −1.18 (SE = 0.232; 95% CI, −1.64 to −0.73) compared to −0.75 (SE = 0.294; 95% CI, −1.33 to −0.17) in the placebo arm; the LSM difference in DSFS sum score between the incobotulinumtoxinA 100 U arm and the placebo arm was −0.43 (SE = 0.275; 95% CI, −0.98 to 0.11).

Changes from baseline in DSFS severity and frequency scores were exploratory end points in the MP (Table 37). This analysis was not controlled for multiplicity. Mean decreases in DSFS severity scores were of greater magnitude in the incobotulinumtoxinA 100 U arm compared with the placebo arm, particularly at week 4 (−1.06 [SD = 1.03] versus −0.37 [SD = 0.77]) and week 8 (−0.92 [SD = 1.00] versus −0.64 [SD = 0.87]). Mean decreases in DSFS frequency scores were of greater magnitude in the incobotulinumtoxinA 100 U arm compared with the placebo arm, particularly at week 4 (−0.71 [SD = 0.70] versus −0.22 [SD = 0.64]) and week 8 (−0.83 [SD = 0.73] versus −0.22 [SD = 0.64]).

Extension Period

Changes from baseline to each visit in the EP for DSFS sum scores as well as severity and frequency scores were assessed in an exploratory fashion (Table 38). This analysis was not controlled for multiplicity. Sum scores and subscores for the DSFS and decreased (improved), from study baseline to the week 4 visit for the incobotulinumtoxinA 100 U treatment arm at each injection cycle. Magnitudes of changes in DSFS sum scores were similar or greater to those observed during the MP.

Health-Related Quality of Life

Main Period

Changes in EQ-5D-3L single items and EQ VAS from baseline were analyzed as exploratory end points in the MP (Table 39). This analysis was not controlled for multiplicity. At baseline the mean EQ VAS score was 59.31 (SD = 18.00) in the placebo arm and 58.62 (SD = 17.08) in the incobotulinumtoxinA 100 U arm. No clear changes in the EQ VAS from baseline were observed in any treatment group, nor were differences between treatment groups evident. No clinically relevant shifts in EQ-5D-3L single items were observed (not shown in this report).

Extension Period

Changes in EQ-5D-3L single items and EQ VAS from baseline to the week 4 assessment for each cycle in the EP were analyzed in an exploratory fashion (Table 40). This analysis was not controlled for multiplicity. No major changes in the EQ VAS from baseline were observed for the incobotulinumtoxinA 100 U treatment arm, nor were clinically relevant shifts in EQ-5D-3L single items observed (not shown in this report).

Harms

Only those harms identified in the review protocol are reported below. Table 41 and Table 42 provide detailed harms data for the EP.

Adverse Events

In the MP, AEs occurred in 34 patients (45.9%) treated with incobotulinumtoxinA 100 U and 15 patients (41.7%) treated with placebo. Apart from AESIs and notable harms (see below), the most common AEs reported at the PT level in the incobotulinumtoxinA 100 U group were diarrhea (3 patients; 4.1%) and hypertension (3 patients; 4.1%). No AEs in the placebo group occurred in more than 1 patient.

In the EP, AEs occurred in 54 patients (60.7%) treated with incobotulinumtoxinA 100 U. Apart from AESIs and notable harms, the most common AEs reported at the PT level in the incobotulinumtoxinA 100 U group were nasopharyngitis (6 patients; 6.7%), fall (5 patients; 5.6%), PD (3 patients; 3.4%), bronchitis (3 patients; 3.4%), amylase increased (3 patients; 3.4%), and hypertension (2 patients; 2.2%).

Serious Adverse Events

Overall, SAEs occurred in 9 patients (12.2%) treated with incobotulinumtoxinA 100 U and in 3 patients (8.3%) treated with placebo. At the PT level, all SAEs occurred in single patients.

In the EP, SAEs occurred in 14 patients (15.7%) treated with incobotulinumtoxinA 100 U. At the PT level, the only SAEs occurring in more than 1 patient were PD (2 patients; 2.2%) and urethral stenosis (2 patients; 2.2%).

Withdrawals Due to Adverse Events

In the MP, AEs leading to study discontinuation occurred in 1 patient (1.4%) treated with incobotulinumtoxinA 100 U (gastrointestinal obstruction) and in no patients treated with placebo.

In the EP, AEs leading to study discontinuation occurred in 8 (9.0%) patients treated with incobotulinumtoxinA 100 U. The most AE causing discontinuation was a dry mouth (4 patients, 4.5%).

Mortality

No patients died during the MP. In the EP, 2 patients (2.2%) treated with incobotulinumtoxinA 100 U died. Both deaths occurred during the fourth treatment cycle. The causes of death were dopamine dysregulation syndrome and pulmonary embolism.

AESIs and Notable Harms

In the MP, AESIs occurred in 5 patients (6.8%) in the incobotulinumtoxinA 100 U arm and in no patients in the placebo arm. Dry mouth (3 patients; 4.1%), dysarthria (1 patient; 1.4%), and dysphonia (2 patients; 2.7%) occurred in the incobotulinumtoxinA 100 U group but not in the placebo group. Dysphagia and speech disorder did not occur in any patients. Dental-related AEs occurred in 4 patients (5.4%) in the incobotulinumtoxinA 100 U arm and 3 patients in the placebo group (8.3%; dental caries, noninfective gingivitis, and tooth fracture). Swallowing scores for the mROMP were stable from study baseline to all assessment visits in the MP for both groups (placebo and incobotulinumtoxinA 100 U).

In the EP, AESIs occurred in 12 patients (13.5%) treated with incobotulinumtoxinA 100 U. Dry mouth (10 patients; 11.2%), dysphagia (4 patients; 4.5%), and pneumonia aspiration (1 patient; 1.1%) occurred in patients treated with incobotulinumtoxinA 100 U. No patients treated with incobotulinumtoxinA 100 U experienced speech disorder, dysphonia, dysarthria, or dyspnea. Dental-related AEs occurred in 10 patients treated with incobotulinumtoxinA 100 U (tooth extraction, 4 patients, 4.5%; dental caries, dental implantation, noninfective gingivitis, gingivitis, tooth loss, and tooth repair, 1 patient each, 1.1%). Swallowing scores for the mROMP were stable from study baseline to all assessment visits in the EP.

Critical Appraisal

Internal Validity

SIAXI was a double-blind, placebo-controlled, multi-centre, phase III RCT with an EP of dose-blinded active treatment (N = 184). The study was rigorously designed and the risk of bias was generally low. Randomization appeared adequate in balancing the baseline characteristics of treatment arms (e.g., age, gender, neurologic condition, concomitant diseases and medications, MMAS/UPDRS, sialorrhea severity evaluated based on uSFR, DSFS, and mROMP), although the specific algorithm used was not disclosed, and allocation was concealed using computer systems. According to the clinical expert consulted by CADTH for this review, there would be no noticeable effect of incobotulinumtoxinA injection that could lead to unblinding of patients or investigators, apart from reduction in sialorrhea due to an effect of therapy. No interim analyses or unblinding of any individuals took place until after completion of the MP. Relatively few patients (placebo: 11.1%; incobotulinumtoxinA 100 U: 2.7%) discontinued the MP of the study in which co-primary and secondary outcomes were evaluated, and there were few missing data for analysis of the co-primary outcomes (uSFR change from baseline and patient GICS score at week 4). Despite the higher discontinuation rate in the placebo arm, no patients cited lack of efficacy as the reason for discontinuation. According to the clinical expert consulted by CADTH for this review, discontinuation rates of these magnitudes are to be expected in this population and would not be likely to affect the results of the study.

Randomization was generally successful in balancing treatment arms for baseline demographic and clinical characteristics. The only notable exceptions were in the use of DBS (placebo: 13.9%; incobotulinumtoxinA 100 U: 24.3%), psychoanaleptics (placebo: 19.4%; incobotulinumtoxinA 100 U: 36.5%), diuretics (placebo: 27.8%; incobotulinumtoxinA 100 U: 14.9%), calcium channel blockers (placebo: 27.8%; incobotulinumtoxinA 100 U: 14.9%), and cardiac therapy (placebo: 19.4%; incobotulinumtoxinA 100 U: 6.8%). None of these imbalances, including DBS, were judged by the clinical expert consulted by CADTH for this review as likely to affect the internal validity of the study, as patients were kept on the same therapy (medications and/or DBS) before and throughout the study. However, several unmeasured variables, which, if imbalanced, could confound the relationship between treatment administered and salivary production measured using the uSFR. For example, rates of smoking and use of saliva substitutes were not presented, nor was the volume of water consumed by patients in each group before uSFR measurements. Testing conditions for the uSFR were standardized to the extent feasible, but the measurement of small changes in salivary production could be susceptible to various forms of confounding and bias.

The statistical approach used for analysis of co-primary efficacy outcomes in SIAXI was generally appropriate. Multiplicity control was only considered for the co-primary outcome using a fixed sequence test procedure (incobotulinumtoxinA 100 U versus placebo for both uSFR and GICS at week 4, followed by incobotulinumtoxinA 75 U versus placebo for both outcomes). This hierarchical test procedure was considered adequate by Health Canada reviewers and biostatisticians.⁵⁶ Data handling and imputation procedures were satisfactory, with MMRM analysis and ANCOVA (BOCF and observed cases) used for both primary and secondary analyses, as well as several exploratory analyses, in both the FAS and PPS. Although ANCOVA with BOCF were originally specified in the statistical analysis plan and later changed to an MMRM, this was done on advice from the FDA, given that an MMRM is more appropriate for outcomes with repeated measurements over time.⁵⁶

However, interpretation of the co-primary efficacy outcomes was complicated by several factors and statistical issues that should be considered. First, the sample size required in the SIAXI study for safety evaluation (N = 184) was probably significantly overpowered to detect treatment effects of the magnitudes reported in a previous study of rimabotulinumtoxinB.⁴⁷ The study was powered to have a 95% probability of detecting a difference in mean change in uSFR at week 4 between the incobotulinumtoxinA and placebo arms of 0.246 g/min, and a difference in mean GICS scores at week 4 of 0.955. The actual mean changes determined in the study were 0.09 g/min (uSFR) and 0.58 (GICS). Given that the sample-size calculation was based on a study of rimabotulinumtoxinB, a potential explanation is that the treatment effect of incobotulinumtoxinA was substantially smaller in the SIAXI study compared with the rimabotulinumtoxinB study. The power of the SIAXI study to detect treatment effects of the magnitudes actually reported was not stated but may have been lower than 95%. Second, the rationale for inclusion of baseline DSFS (investigator-assigned), rather than uSFR or mROMP, as a covariate in MMRM models of GICS was not explained; whether another construct may have been more appropriate to model baseline sialorrhea severity was unclear. Third, the GICS was scored on a 7-point Likert scale with only 3 levels denoting a positive change in sialorrhea (minimally improved, much improved, and very much improved) and then modelled as a continuous variable in MMRM. A similar strategy was applied to the DSFS in exploratory analyses. Few patients in any treatment group selected a GICS score of 3 (very much improved), and many placebo-treated patients selected a score of 1 (minimally improved). These patient-reported scales therefore may not have been responsive to relatively modest but clinically significant differences in function, and comparisons of LSM differences in GICS scores may not have been sensitive to such changes. Along similar lines, the definition used in SIAXI of responder (a GICS score ≥ 1) was not explained; using this definition, 30% to 90% of placebo-treated patients responded to therapy across the MP. Analysis of GICS categories suggested that the LSM difference between incobotulinumtoxinA 100 U and placebo was driven by a relatively small proportion of incobotulinumtoxinA -treated patients with large GICS values (≥ 2). According to the clinical expert consulted by CADTH for this review, moderate to marked improvement in sialorrhea (a GICS score of 2 or 3), rather than minimal improvement (a score of 1), would constitute a clinically meaningful response to treatment. Sensitivity analyses of categorical GICS data other than the proportion of responders (e.g., using logistic regression or Fisher’s exact test) were not reported. Finally, there was no strategy for multiplicity control for secondary or exploratory outcomes (following hierarchical testing of the co-primary outcomes), all of which were treated in an exploratory fashion.

Subgroup analyses by the etiology of sialorrhea were specified a priori but the study was not specifically powered to evaluate differences in treatment effect by neurologic condition. The number of patients with stroke and TBI in particular was relatively small. There was therefore uncertainty in the extent to which treatment effects were similar or different across patients with PD/AP and stroke.

As described in Appendix 4, no evidence was available to support the validity, reliability, responsiveness to change, or MID for any of the outcome measures used in the SIAXI trial. Many of these outcome measures (e.g., GICS, DSFS, mROMP, EQ VAS) capture subjective impressions of change from patients, caregivers, or clinicians. According to the clinical expert consulted by CADTH for this review, none of the outcome measures used in the study are used routinely in clinical practice, although questions similar to those asked on the GICS, DSFS, and mROMP drooling scales are part of clinical assessments. Similarly, none of these outcomes were specifically identified by patient groups as important to them. A clear placebo effect was observed for all efficacy outcomes analyzed; for patient- or investigator-rated scales such as GICS and DSFS, this may have been partially due to limited categories of choice, such that any marginal improvement gained via the placebo effect would be scored as positive. According to the clinical expert consulted by CADTH for this review, placebo effects would be expected in this population but would wane over time. However, even uSFR (an objective measure) showed consistently negative changes from baseline to all time points in the MP, suggesting a placebo effect mediated via increased salivation in placebo-treated patients, a subconscious bias in weighing saliva on the part of investigators, or a true treatment effect of saline injection into the salivary glands. Despite these caveats, the clinical expert consulted by CADTH for this review interpreted the results (particularly changes from baseline in GICS and DSFS) as clinically significant, in part based on the consistency of effect across 4 different measures (uSFR, GICS, DSFS, and mROMP). Consistency of effect was also seen across time points in the MP, with the greatest difference in treatment effect between incobotulinumtoxinA 100 U and placebo occurring at weeks 4, 8, and 12, then waning. Given the limited dynamic ranges of the scales used in the SIAXI trial (e.g., 7 points on the GICS and 9 points on the DSFS sum score), even modest differences could be clinically meaningful, according to the clinical expert consulted for this review, provided these differences resulted in a meaningful change in the patient’s lifestyle (e.g., by reducing social isolation or allowing the patient to remain in the workforce).

External Validity

The SIAXI trial was conducted at 33 sites (12 in Germany and 21 in Poland). According to the clinical expert consulted by CADTH for this review, the ages of patients in the study (mean age = 65.2 years; SD = 11.4 years) were representative of the population seen in Canadian practice. However, most patients were men (70.7%) and nearly all were White (99.5%). Baseline disease parameters and sialorrhea severity (e.g., UPDRS/MMAS, mROMP, and DSFS) were reflective of patients typically treated with incobotulinumtoxinA, according to the clinical expert consulted by CADTH for this review. Acknowledging the restrictions and limitations of the study population, the clinical expert did not think this would impede generalization of the study results to patients in other countries and regions or patients of different genders or races; based on the mechanism of action, similar treatment effects would be expected in all adult patients with neurologic conditions. The clinical expert expected that the treatment effect would be similar irrespective of underlying neurologic condition, such that the findings of the SIAXI study (whose population consisted of 79.4% PD/AP patients and 19.0% patients with stroke) would most likely be generalizable to patients with other neurologic conditions, resulting in sialorrhea such as ALS, dementia, and CP. However, the etiology of sialorrhea was moderately associated with patient GICS scores at week 4 in MMRM models (P = 0.034), and some subgroup analyses in the SIAXI trial were unable to conclusively demonstrate homogeneity of treatment effect in patients with PD/AP versus stroke. Whether the results of the SIAXI study can be generalized to patients with other types of neurologic conditions is therefore uncertain.

Adults with many neurologic conditions, particularly PD and stroke, are expected to deteriorate over time, and the clinical expert consulted by CADTH for this review stated that the treatment effect of incobotulinumtoxinA may no longer be sufficient in such patients, especially if swallowing dysfunction becomes worse. The findings of the SIAXI study therefore may not be generalizable to older patients, those with more difficulty swallowing, and patients with more severe neuromuscular dysfunction (e.g., based on UPDRS/MMAS scores). Despite an exclusion criterion for prior (within 6 months) or planned functional neurosurgery including DBS, 13.9% of patients in the placebo arm and 24.3% of patients in the incobotulinumtoxinA 100 U arm reported previous and concomitant DBS during the study. It is therefore unclear whether the results are generalizable to patient populations with higher or lower rates of DBS.

According to the clinical expert consulted by CADTH for this review, the inclusion/exclusion criteria for SIAXI would be expected to recruit a population similar to that treated with incobotulinumtoxinA in clinical practice in Canada. To be included in the SIAXI study, patients had to have DSFS sum scores of 6 or greater, DSFS scores for both severity and frequency of at least 2, and mROMP Section III “Drooling,” Item A scores of 3 or greater at both screening and baseline. As the most common cause of screening failure was poor oral and/or dental condition, the findings may also not be generalizable to these patients. The SIAXI study population was BoNT-naive (within 1 year for sialorrhea), and study results may not be generalizable to patients recently or persistently treated with BoNTs who might develop resistance and/or neutralizing antibodies.

The treatment regimen (100 U) used in the SIAXI study was aligned with the Health Canada–approved dose and, according to the clinical expert consulted by CADTH for this review, with the doses used in Canadian clinical practice. Few patients received comedications influencing sialorrhea (e.g., anticholinergics), although these are also routinely available in Canadian clinical practice. The dosing of 100 U every 16 weeks over the 4 cycles in the EP means that the safety and efficacy results may not be generalizable to patients treated with other, potentially more sporadic, dosing schedules. This is supported by the relatively smaller changes in uSFR in the EP comparing week 4 to cycle baselines compared with study baseline, or with the MP examining BoNT-naive patients.

Although SIAXI study investigators received specialized training in administration of ultrasound-guided injections, the clinical experiment consulted by CADTH for this review stated that both ultrasound-guided and anatomic landmark–guided injections can be administered successfully and are used in clinical practice. Training for use of ultrasound to guide injections, or lack thereof, would therefore be unlikely to affect the generalizability of the study findings.

Over the course of the SIAXI study (64 weeks, approximately 15 months), patients made 15 visits for assessment and consultation and were contacted in 7 telephone interviews. This may have led to a different standard of background care than the average adult patient with a neurologic condition would receive in the Canadian context. Whether and how this affects generalizability of the study findings to patients receiving less-frequent care (e.g., optimization of anti-parkinsonism medications) is unclear.

The efficacy outcomes used in the SIAXI trial (uSFR, GICS, DSFS, mROMP, and EQ-5D-3L) are not used routinely in clinical practice according to the clinical expert consulted by CADTH for this review. However, questions such as those included on the GICS and DSFS are routinely asked as part of clinical evaluations. According to the clinical expert consulted by CADTH for this review, the duration of follow-up (1 cycle, 16 weeks for the MP; 4 cycles in total, 64 weeks) was sufficient for evaluation of efficacy and safety outcomes in this population.

Indirect Evidence

No indirect evidence was identified for this review. A feasibility assessment conducted by the sponsor also concluded that no data were available to inform an indirect treatment comparison between incobotulinumtoxinA and other interventions, including injection of other BoNTs.

Other Relevant Evidence

This section includes additional relevant studies included in the sponsor’s submission to CADTH that were considered to address important gaps in the evidence included in the systematic review.

Studies in Other Populations

One additional RCT of BoNT-A injections (incobotulinumtoxinA and onabotulinumtoxinA) for treatment of sialorrhea in adults with neurologic conditions was summarized to provide additional evidence in patients with other neurologic conditions (ALS and CP) as well comparative evidence for these 2 BoNTs. According to the clinical expert consulted by CADTH for this review, onabotulinumtoxinA is regularly used off-label for the treatment of sialorrhea in adults with neurologic disorders. This study was not a direct head-to-head comparison of these 2 BoNTs.

Methods

Restivo et al.¹¹ conducted an exploratory double-blind RCT to evaluate the efficacy of BoNT-A (incobotulinumtoxinA or onabotulinumtoxinA) injections into the salivary glands for treatment of adults with chronic sialorrhea resulting from neurologic disorders (PD, stroke, TBI, ALS, and CP; N = 90). The study was unrelated to the pivotal SIAXI study and was not funded by the sponsor.

The primary goal of the study was to assess the relationship between efficacy in reducing sialorrhea and number of glands injected. A consecutive series of patients were randomized to receive ultrasound-guided injections with BoNT-A in either 4, 3, or 2 salivary glands (groups A, B, and C, respectively; N = 30 per group). The primary analysis was of the number of glands injected with either of the BoNT-A types, and not the specific type. The basis for selection of incobotulinumtoxinA or onabotulinumtoxinA for injections was not stated and this analysis was post hoc. Salivary production was measured at baseline and 2 weeks post-injection by an investigator blind to treatment-group assignments (patients were also blind to group assignments). Dental rolls were weighed, placed in the patient’s mouth for 5 minutes, then weighed again. The difference in weights was calculated. The procedure was repeated 15 minutes later, and the 2 measurements were averaged. Finally, differences in salivary weight from baseline to 2 weeks post-injection were converted to a 4-point Likert scale (0: no reduction, 1: 25% reduction, 2: 50% reduction, and 3: 75% reduction in salivary weight). The specific break points for this scale were not stated.

Populations

Patients aged 18 to 75 years inclusive with PD, stroke, TBI, ALS, or CP and severely disabling sialorrhea lasting for at least 6 months were included. The diagnostic criteria for neurologic disorders and “severely disabling sialorrhea” were not stated. Patients had to have wet dental-roll weights that were at least 10-fold greater than dry roll weights to be included. Patients with other neurologic conditions or patients who could not provide informed consent due to cognitive impairment were excluded.

Among the included patients (N = 90), 59 (65.6%) were men and 31 (34.4%) were women. The mean age was 53.4 (SD = 17.6) years. Neurologic conditions included PD (30; 33.3%) stroke (21; 23.3%), CP (20; 22.2%), ALS (11; 12.2%), and TBI (8; 8.9%). Gender and age were generally well balanced across treatment arms (group A, 66.7% male, age range = 17 to 73 years; group B, 66.0% male, age range = 18 to 72 years; group C, 66.7% male, age range = 21 to 73 years). However, the etiology of sialorrhea (underlying neurologic condition) was not well balanced: the proportions of patients with PD, stroke, CP, ALS, or TBI were 53.3%, 23.3%, 13.3%, 0%, and 10.0% in group A; 36.7%, 20.0%, 26.7%, 13.3%, and 3.3% in group B; and 10.0%, 26.7%, 23.3%, 23.3%, and 16.7% in group C, respectively.

Interventions

Patients were injected percutaneously with 25 U (0.5 mL) of BoNT-A (onabotulinumtoxinA or incobotulinumtoxinA) per salivary gland (parotid and/or submandibular). Injections were ultrasound-guided using a 27G 3-quarter-inch needle. Injections were performed only once. The total dose of BoNT-A received was therefore 100 U in patients receiving injections in 4 glands, 75 U in patients receiving injections in 3 glands, and 50 U in patients receiving injections in 2 glands. In patients receiving BoNT injections in 2 or 3 glands, the remaining glands were injected with an equivalent volume of saline solution. Information on concomitant medications and therapies was not provided.

Outcomes

The singular efficacy outcome assessed in the study by Restivo et al. was the change in salivary production from baseline to 2 weeks post-injection. Change in salivary production was converted to a 4-point Likert scale as described in the Methods section, and differences between numbers of glands injected and other parameters were analyzed by 1- and 2-way analysis of variance (ANOVA). Safety outcomes were not formally analyzed.

Statistical Analysis

Differences in efficacy (based on the Likert scale) in patients receiving injections into different numbers of glands were assessed using 2-way ANOVA. Differences in efficacy based on type of toxin injected (incobotulinumtoxinA versus onabotulinumtoxinA) were assessed using 2-way ANOVA. Subgroup analyses within patients receiving injections into 3 glands (2 parotid + 1 submandibular or 1 parotid + 2 submandibular) and patients receiving injections into 2 glands (2 parotid, 2 submandibular, or 1 parotid + 1 submandibular) were conducted using 1-way ANOVA. Subgroup analysis by neurologic condition was conducted using 1-way ANOVA. For ANOVA with 3 levels, tests were adjusted for multiple comparisons using the Bonferroni correction. Analysis was in the FAS and there were no missing data.

Patient Disposition

A consecutive series of 107 patients were evaluated for inclusion in the study, 90 of whom were randomized. No information was provided on the 17 screening failures (15.9%). All 90 patients (100.0%) received study treatment were evaluated at 2 weeks post-injection and completed the study.

Exposure to Study Treatments

All patients received the planned injections. A total of 57 patients (63.3%) were injected with onabotulinumtoxinA and 33 patients (33.7%) were injected with incobotulinumtoxinA. Among the overall study population, only 8 patients (8.9%) received the Health Canada–approved dose (100 U) of incobotulinumtoxinA (6 patients with PD, 1 patient with stroke, and 1 patient with TBI).

In group A, 8 patients (26.7%) received incobotulinumtoxinA and 22 patients (73.3%) received onabotulinumtoxinA; in group B, 15 (50.0%) patients received incobotulinumtoxinA and 15 (50.0%) patients received onabotulinumtoxinA; and in group C, 10 patients (33.3%) received incobotulinumtoxinA and 20 (67.7%) patients received onabotulinumtoxinA. Within each study group, the proportions of patients receiving incobotulinumtoxinA versus onabotulinumtoxinA were not balanced by neurologic condition.

Efficacy

Overall, 82 (91.2%) patients responded to treatment (Likert score ≥ 1) and the mean dental-roll weight was 0.25 g (SD = 0.1 g) at 2 weeks post-injection compared to 0.8 g (SD = 0.08 g) at baseline.

Differences in the treatment effects of incobotulinumtoxinA and onabotulinumtoxinA were explored using ANOVA (Figure 3). There was a clear pattern of dose response for both BoNT-A types, with Likert scores increasing along with increasing number of glands injected (P < 0.001, 2-way ANOVA, for both BoNT-A types), but no interaction between BoNT-A type and number of glands injected (P = 0.069, 2-way ANOVA with Bonferroni correction).

No difference between the incobotulinumtoxinA-treated and onabotulinumtoxinA groups was observed based on a 1-way ANOVA (P = 0.12). However, no numerical data were available in the report to substantiate this conclusion.

Subgroup analysis by etiology of sialorrhea in the overall population treated with all doses of BoNT-A (either incobotulinumtoxinA or onabotulinumtoxinA) suggested a potential difference in treatment effect by neurologic condition (P < 0.001, 1-way ANOVA). Based on the Bonferroni multiple comparison, this difference was driven by paired differences between ALS patients and CP patients (P = 0.0033), PD patients (P < 0.001) and patients with stroke (P = 0.0022), with ALS patients (N = 8) showing generally poorer responses.

Figure 3: Likert Scores for Change in Salivary Production Following Injection of BoNT-A (Restivo et al.)

Incobotulinum = incobotulinumtoxinA (Xeomin); ns = not significant; Onabotulinum = onabotulinumtoxinA (Botox); *** = P < 0.001.

Note: Scores represent Likert transformation of response scores: 0 = no response, 1 = poor response, 2 = intermediate response, and 3 = good response. Error bars represent standard errors. Large bars (bottom) represent results for the entire study population; narrow bars (top) represent subgroup analysis for injections into 4, 3 and 2 glands.

Source: Restivo et al. (2018). Copyright 2019. This work is licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0) licence. Full text available here: https://www.mdpi.com/2072-6651/10/2/55.¹¹

Harms

Safety was not formally assessed in the study by Restivo et al. One patient (1.1%) experienced dysphagia 7 days after BoNT-A injection and 2 patients (2.2%) developed hematomas at injection sites. The specific treatments (incobotulinumtoxinA versus onabotulinumtoxinA), dose, and number of glands injected that resulted in these AEs were not stated.

Critical Appraisal

The exploratory RCT by Restivo et al. was described in limited detail, preventing CADTH from conducting a thorough critical appraisal. The evidence and any conclusions derived therefrom are therefore associated with substantial uncertainty.

Internal Validity

Limited information on the study design or statistical analysis was provided in the published study. Randomization groups were assigned by number of salivary glands injected, and the basis for treatment of some patients with incobotulinumtoxinA and others with onabotulinumtoxinA was not explained. The comparative evidence from this study for different BoNT-A types is therefore potentially susceptible to various forms of confounding and bias. Bias could not be evaluated; the randomization strategy and allocation concealment, the methods used for maintaining the DB, and reasons for screening failures were not stated. Importantly, the neurologic condition/etiology of sialorrhea was not well balanced across treatment arms. The comparative analyses of the BoNT-A drugs (incobotulinumtoxinA versus onabotulinumtoxinA) and neurologic condition (both using ANOVA) did not take into account the confounding effects of imbalances in neurologic condition (for the former) and which BoNT-A was administered (for the latter). The rationale and break points for converting salivary weight measurements into a 4-point Likert scale were not explained, and the numerical data for this outcome (mean, SD, or SE) for the comparisons between BoNT-A drug and neurologic outcomes were not provided.

External Validity

This was a study conducted at a single Italian hospital. Of the 33 patients treated with incobotulinumtoxinA, only 8 were treated with the Health Canada–approved dose of 100 U (6 patients with PD, 1 with stroke, and 1 with TBI). Limited information was provided on inclusion and/or exclusion criteria. No information was provided on patient background, including ethnicity, neurologic disease diagnosis and severity, severity of sialorrhea, concomitant therapies, and background treatments. The clinical relevance of the outcome measure (salivary production converted into a Likert scale using unclear breakpoints) was uncertain. As no control group (placebo) was assessed, the true degree of treatment effect of either BoNT-A could not be evaluated. The follow-up time (2 weeks) was relatively short and potentially insufficient to evaluate efficacy. Safety data were not collected.

In summary, the study by Restivo et al. was unable to address the evidence gaps relating to the efficacy of incobotulinumtoxinA in patients with neurologic conditions other than PD/AP and stroke and to comparative efficacy versus other BoNT-A injections for this indication. This was primarily due to the small number of patients receiving the Health Canada–approved dose of incobotulinumtoxinA (100 U), all of whom had PD, stroke, or TBI (the same groups covered included in the SIAXI study), and the design of the study, which, because it involved a non-randomized comparison of incobotulinumtoxinA with onabotulinumtoxinA, had the potential for confounding bias and limited confidence in this comparison.

Discussion

Summary of Available Evidence

One phase III, double-blind, placebo-controlled RCT (SIAXI,^8-10 N = 184) was included in the CADTH systematic review. The study enrolled primarily patients with PD/AP (approximately 80%) and stroke (approximately 20%) who had moderate to severe sialorrhea (DSFS sum score ≥ 6, DSFS subscores ≥ 2, and mROMP drooling scores ≥ 3). Following injection of incobotulinumtoxinA (100 U) or placebo, patients in the MP were followed for 16 weeks. All patients in the study were subsequently treated with incobotulinumtoxinA for 3 additional cycles (the dose-blinded active treatment EP) and the same outcomes were examined. The co-primary efficacy outcomes were the mean difference in change from baseline in uSFR at week 4 and mean difference in patient-reported GICS scores at week 4 among incobotulinumtoxinA-treated patients compared with placebo-treated patients. The secondary outcomes were change in uSFR from baseline to weeks 8 and 12 and patient-reported GICS scores at weeks 1, 2, 8, and 12 of the MP. The EQ-5D-3L was used to assess HRQoL as an exploratory outcome.

The baseline characteristics of the SIAXI population were generally representative of Canadian adults requiring treatment for chronic sialorrhea. However, participants were mostly male and White, and almost all patients had sialorrhea secondary to either PD/AP or stroke. Baseline and disease characteristics were generally well balanced at baseline, there were few discontinuations, and the relatively minor degree of missing data (similar between treatment arms) was not expected to affect the results.

In addition to the SIAXI trial, 1 additional exploratory single-centre double-blind RCT was summarized to provide additional evidence from patients with other neurologic conditions and to provide comparative evidence for treatment with incobotulinumtoxinA and onabotulinumtoxinA. The study by Restivo et al.¹¹ recruited patients with PD, stroke, TBI, ALS, and CP (N = 90) with sialorrhea. Patients were randomized to receive BoNT-A injections in 2, 3, or 4 salivary glands. Two weeks post-injection, change in salivary production was evaluated on a Likert scale. Only 33 patients were treated with incobotulinumtoxinA and only 8 patients were treated with the Health Canada–approved dose of 100 U (6 patients with PD, 1 patient with stroke, and 1 patient with TBI).

Interpretation of Results

Efficacy

According to the sponsor, there are no reports in the literature to support the validity of, or an MID for, any of the outcome measures used in the SIAXI trial (uSFR, GICS, DSFS, or mROMP). In addition, no studies were identified by CADTH to support the validity, reliability, responsiveness to change, or MID for any of the outcome measures. This is because of a lack of standard definitions and diagnostic criteria for sialorrhea, leading to wide variability in assessment tools across studies and in clinical practice.⁵⁷ However, an international consensus statement on assessment of BoNTs for treatment of pediatric and adult drooling recommended 2 measures (salivary flow measurements such as uSFR and the DSFS), both of which have been used in several clinical studies.⁶ In an interview conducted by the sponsor, some Canadian neurologists suggested that the GICS aligns with outcomes that are meaningful to patients, and 1 neurologist suggested that a reduction of 1 on the DSFS (range = 2 to 9) would be an important difference for patients and clinicians.¹

In the co-primary efficacy analysis, mean difference in uSFR change from baseline to week 4 showed a statistically significant difference in favour of incobotulinumtoxinA versus placebo using a variety of analytic approaches. Also as part of the co-primary efficacy analysis, the mean difference in patient-reported GICS scores at week 4 showed statistically significant differences in favour of incobotulinumtoxinA versus placebo using a variety of analytic approaches. Similar results and similar uncertainties were documented for secondary outcomes in the MP (uSFR at weeks 8 and 12 and GICS scores at weeks 1, 2, 8, and 12), exploratory outcomes in the MP (DSFS and mROMP), and all outcomes in the EP; none of these analyses were controlled for multiplicity. However, the clinical relevance of these differences was uncertain due to the unclear validity and MID for these outcomes. The uncertain clinical meaningfulness in changes in uSFR and patient- or investigator-perceived sialorrhea may be connected with overpowering of the SIAXI study for efficacy outcomes, allowing it to detect much smaller treatment effects than those targeted based on past investigations of BoNT injections. Despite the uncertain clinical relevance of the magnitude of treatment differences between incobotulinumtoxinA and placebo, and despite observation of a placebo effect for most outcomes, consistent mean changes, with similar timings, were observed in favour of incobotulinumtoxinA across all outcomes assessed: the effect of treatment was numerically observed (but not statistically significant) at weeks 1 and 2 post-injection, manifested at weeks 4, 8, and 12, and then waned by week 16. However, this did not translate into improvement for incobotulinumtoxinA-treated patients in terms of HRQoL measured via the EQ VAS.

The change in uSFR from cycle baseline to week 4 of each cycle was not consistent across the study. In the EP, the magnitudes of LSM differences from cycle baseline to week 4 for incobotulinumtoxinA 100 U–treated patients were much smaller than those of LSM differences from study baseline to week 4 of any of the cycles. This could imply a long-term treatment effect, with uSFR not returning to baseline even at 16 weeks post-injection. Notably, GICS scores in the EP reflected perceived improvement in sialorrhea at each cycle, and this was evaluated with reference to the previous injection, not study baseline.

Subgroup analyses in the SIAXI study did not provide substantive evidence for the presence or absence of potentially important differences in treatment effect by neurologic condition and by baseline severity of sialorrhea. A potentially relevant subgroup analysis by uSFR at baseline (≤ or > median uSFR) was conducted post hoc and therefore was not presented in this report. Neither the pre-specified subgroup analysis by neurologic condition nor the post hoc subgroup analysis by baseline uSFR provided useful evidence regarding the potentially important issue of variation in treatment effect by neurologic condition or sialorrhea severity.

According to the clinical expert consulted by CADTH for this review, the effect of sialorrhea on patient HRQoL does not correlate with neurologic disease severity. For example, in patients with moderate PD, sialorrhea may be the most troublesome issue and affect their lives in terms of social interactions. By contrast, in patients with advanced PD, sialorrhea may cause choking or skin breakdown. This suggests there may be impacts of therapy at different stages. Both identifying patients for treatment and evaluating response will often be highly dependent on circumstances. For example, a patient achieving a mild improvement in their sialorrhea who is still in the workforce might find this treatment effect sufficient to continue working and go about their daily life. It therefore may not be possible to identify a specific MID for sialorrhea that is generalizable for patients with a single neurologic condition, or even for an individual patient whose needs may change as their disease evolves. For this and other reasons, salivary production, although an objective measure, was considered by the clinical expert consulted by CADTH for this review to be less clinically relevant compared with patient- and investigator-rated scales of sialorrhea improvement.

The clinical expert stated that treatment effects following incobotulinumtoxinA injection should be marked and that clinicians are looking for at least moderate, and not minimal, improvement. Along these lines, the mean difference in GICS scores at week 4 between the incobotulinumtoxinA 100 U and placebo groups was driven by a larger proportion of patients selecting a “much improved” score of 2, (incobotulinumtoxinA 100 U: 33.8% versus placebo: 11.1%). However, the majority of patients in both the placebo and incobotulinumtoxinA 100 U groups reported a GICS score of 1 (minimal improvement) or less. A weakness of the analytic approach involving mean change and mean difference in the SIAXI study was that the mean does not describe individual patients, and potentially important variation in the proportion and degree of treatment effect was not analyzable. Many of the characteristics of an ideal treatment identified by the clinical expert consulted by CADTH for this review (minimal side effects and address social inhibition, maceration of skin, dehydration, speech disturbances, interference with eating, and risk of aspiration) were not directly addressed by the outcomes evaluated in the SIAXI study.

A limitation of the SIAXI study was unclear generalizability to patients with neurologic conditions other than PD/AP and stroke and to patients with different treatment schedules and less-regular medical care. However, the clinical expert consulted by CADTH for this review stated that, based on the mechanism of action of incobotulinumtoxinA, it would be reasonable to expect that treatment would reduce sialorrhea irrespective of neurologic condition. The exploratory double-blind RCT conducted by Restivo et al.¹¹ was unable to extend the evidence regarding the efficacy of incobotulinumtoxinA to patients with other neurologic disorders. The CADTH review of this study was based on the limited results presented in the published study and therefore a thorough critical appraisal of internal and external validity was not possible. Only 8 patients in the study were treated with the Health Canada–approved dose of incobotulinumtoxinA (100 U) and all of these patients had sialorrhea secondary to PD or stroke, the same conditions studied in SIAXI.

Although multiple studies have evaluated injection of different BoNT preparations (in isolation) to control sialorrhea, and a variety of BoNTs are used off-label for this purpose in clinical practice, the SIAXI study provided no comparative evidence for incobotulinumtoxinA versus other BoNTs. The study by Restivo et al.¹¹ suggested that treatment effects occurred following injection of both incobotulinumtoxinA and onabotulinumtoxinA in the salivary glands, although this issue was not the primary focus of the study. In addition, Restivo et al. was not a head-to-head comparison; the comparison of BoNT-A type was susceptible to bias and confounding, based on a partially described Likert scale of unclear clinical relevance, and not reported numerically. Whether treatment effects of the 2 BoNTs were similar therefore could not be evaluated.

Harms

The toxicity profile of incobotulinumtoxinA in the SIAXI trial was as expected, based on previous experience with salivary injection of BoNTs, including incobotulinumtoxinA, and was consistent with the product monograph. In the MP, both AEs and SAEs occurred at similar frequencies in the placebo arm (41.7% and 8.3%, respectively) and incobotulinumtoxinA 100 U arm (45.9% and 12.2%, respectively); WDAEs were extremely rare (0% and 1.2%, in the placebo and incobotulinumtoxinA 100 U groups, respectively) and no deaths occurred. In the EP, which consisted of a 48-week follow-up period, only slightly higher rates of AEs and SAEs were observed in incobotulinumtoxinA 100 U–treated patients (60.7% and 15.7%, respectively) compared to the rates of these events observed during the 16-week period of the MP. During the EP, WDAEs occurred in 9.0% of patients treated with incobotulinumtoxinA 100 U, half of whom (4.5%) discontinued due to dry mouth. Harms related to toxin spread (dry mouth, dysarthria, dysphonia, dysphagia, and pneumonia aspiration) occurred in 6.8% of patients in the incobotulinumtoxinA 100 U arm, no placebo-treated patients in the MP, and 13.5% of incobotulinumtoxinA 100 U–treated patients in the EP. These AESIs were generally manageable in most patients. Dysphagia occurred in 4.5% of incobotulinumtoxinA-treated patients in the EP, although there was no evidence of decreased mROMP swallowing scores associated with incobotulinumtoxinA treatment. Dental-related AEs did not occur more frequently in patients treated with incobotulinumtoxinA 100 U compared with those treated with placebo.

Other Considerations

Patient group input for this review indicated that no patients had experience with incobotulinumtoxinA and did not identify specific outcomes important to patients. The characteristics desired by patients were limited side effects, government coverage of treatments, treatments effective in reducing severity and frequency of drooling, and self-administered treatments. There appears to be a research gap in the assessment of drooling severity and frequency, and in connecting measurement of drooling directly with HRQoL.

The pivotal SIAXI trial did not include study sites in Canada. There are currently no reimbursed treatments for sialorrhea in adults with neurologic disorders in Canada. Although injection of several BoNTs is used off-label for this purpose, only extremely limited comparative evidence was identified to address their relative efficacy. Other evidence gaps included lack of data in patients with neurologic conditions other than PD/AP and stroke as well as lack of data regarding potential differences in treatment effects based on the severity of sialorrhea. Another research gap lies in defining the characteristics of patients who will or will not manifest treatment responses to injections with BoNTs such as incobotulinumtoxinA.

The clinical expert consulted by CADTH for this review stated that, although the SIAXI trial assessed incobotulinumtoxinA injections in adults with PD and stroke, other groups of patients would also benefit from treatment and would not be covered by the indication under review. The expectations and backgrounds of adult patients with PD may differ substantially from those of patients with CP and other childhood neurologic disorders that involve profuse sialorrhea. On a per-capita basis, these patients have higher rates of sialorrhea and it is distressing for them.

Conclusions

Evidence from the SIAXI study suggested that injection of incobotulinumtoxinA (100 U) into the salivary glands of adult patients with neurologic disorders resulted in reduced salivary production and improvements in patients’ perceptions of the frequency and severity of sialorrhea. At 4 weeks post-injection, the mean difference in change in score from baseline in the uSFR and GICS was statistically significantly in favour of incobotulinumtoxinA versus placebo. Treatment effects as measured by the uSFR and GICS were also observed at weeks 8 and 12 post-injection, and similar results were obtained on the investigator-rated DSFS. The clinical significance of post-treatment changes in sialorrhea between incobotulinumtoxinA- and placebo-treated patients was uncertain because the outcome measures were unvalidated, not used in clinical practice, and subjective apart from uSFR, and therefore the magnitudes of treatment effects were of unclear relevance to patients. However, the clinical expert consulted by CADTH for this review indicated that questions similar to those asked in the GICS, DSFS, and mROMP drooling scales are part of clinical assessments, and that the differences in GICS and DSFS between the incobotulinumtoxinA 100 U and placebo arms were clinically meaningful. Despite the uncertain clinical relevance of the magnitude of treatment differences between incobotulinumtoxinA and placebo, and despite evidence for a placebo effect in most outcomes, consistent mean changes with similar timing in favour of incobotulinumtoxinA were observed across all outcomes assessed: differences in the effects of treatment with incobotulinumtoxinA 100 U versus placebo were numerically observed (via GICS responses) but not statistically significant at weeks 1 and 2 post-injection, clearly manifested at weeks 4, 8, and 12, and then waned by week 16, at which point a subsequent dose was administered. However, this did not translate into improvement for incobotulinumtoxinA-treated patients in terms of HRQoL measured via the EQ VAS. Injection with incobotulinumtoxinA was tolerated in most patients and side effects were generally manageable, with some infrequent but expected notable harms related to toxin spread (e.g., dry mouth and dysphagia). Key evidence gaps included a lack of comparative data on the efficacy of different BoNTs and lack of evidence in patients with a variety of neurologic conditions.

References

1.Drug Reimbursement Review sponsor submission: Xeomin (incobotulinumtoxin A, clostridium botulinum neurotoxin type A (150 kD), free from complexing proteins): powder for solution for injection, 50 and 100 units per vial [internal sponsor's package]. Burlington (ON): Merz Pharma Canada Ltd.; 2021 Mar 5.

2.Lakraj AA, Moghimi N, Jabbari B. Sialorrhea: anatomy, pathophysiology and treatment with emphasis on the role of botulinum toxins. Toxins (Basel). 2013;5(5):1010-1031. PubMed

3.Leibner J, Ramjit A, Sedig L, et al. The impact of and the factors associated with drooling in Parkinson's disease. Parkinsonism Relat Disord. 2010;16(7):475-477. PubMed

4.Miller RG, Jackson CE, Kasarskis EJ, et al. Practice parameter update: the care of the patient with amyotrophic lateral sclerosis: multidisciplinary care, symptom management, and cognitive/behavioral impairment (an evidence-based review): report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology. 2009;73(15):1227-1233. PubMed

5.Naumann M, So Y, Argoff CE, et al. Assessment: Botulinum neurotoxin in the treatment of autonomic disorders and pain (an evidence-based review): Report of the Therapeutics and Technology Assessment Subcommittee of the American Academy of Neurology. Neurology. 2008;70(19 PART 1):1707-1714. PubMed

6.Reddihough D, Erasmus CE, Johnson H, McKellar GM, Jongerius PH, Cerebral Palsy Institute. Botulinum toxin assessment, intervention and aftercare for paediatric and adult drooling: international consensus statement. Eur J Neurol. 2010;17 Suppl 2:109-121. PubMed

7.Xeomin (incobotulinumtoxin A, clostridium botulinum neurotoxin type A (150 kD), free from complexing proteins): powder for solution for injection, 50 and 100 units per vial [product monograph]. Burlington (ON): Merz Pharma Canada Ltd.; 2020 Nov 3: https://www.merzcanada.com/files/XEOMIN%20Therapeutic%20Product%20Monograph%20Nov%2017%202020.PDF. Accessed 2021 Apr 29.

8.Clinical Study Report: MRZ60201_3090_1. Prospective, randomized, double-blind, placebo-controlled, parallel-group multicenter study, with an extension period of dose-blinded active treatment, to investigate the efficacy and safety of two dose levels of NT 201 in treating chronic troublesome sialorrhea in various neurological conditions [internal sponsor's report]. Frankfurt (DE): Merz Pharmaceuticals GmbH; 2017 May 16.

9.Jost WH, Friedman A, Michel O, et al. SIAXI: Placebo-controlled, randomized, double-blind study of incobotulinumtoxinA for sialorrhea. Neurology. 2019;92(17):e1982-e1991. PubMed

10.Jost WH, Friedman A, Michel O, et al. Long-term incobotulinumtoxinA treatment for chronic sialorrhea: Efficacy and safety over 64 weeks. Parkinsonism Relat Disord. 2020;70:23-30. PubMed

11.Restivo DA, Panebianco M, Casabona A, et al. Botulinum toxin A for sialorrhoea associated with neurological disorders: evaluation of the relationship between effect of treatment and the number of glands treated. Toxins (Basel). 2018;10(2):27. PubMed

12.Prommer E. Anticholinergics in palliative medicine: an update. Am J Hosp Palliat Care. 2013;30(5):490-498. PubMed

13.Hockstein NG, Samadi DS, Gendron K, Handler SD. Sialorrhea: a management challenge. Am Fam Physician. 2004;69(11):2628-2634. PubMed

14.Kalf JG, de Swart BJ, Borm GF, Bloem BR, Munneke M. Prevalence and definition of drooling in Parkinson's disease: a systematic review. J Neurol. 2009;256(9):1391-1396. PubMed

15.Wong SL, Gilmour H, Ramage-Morin PL. Parkinson's disease: Prevalence, diagnosis and impact. Health Rep. 2014;25(11):10-14. PubMed

16.Public Health Agency of Canada. Canadian Best Practices Portal. Neurological Conditions. 2016; https://cbpp-pcpe.phac-aspc.gc.ca/chronic-diseases/neurological-conditions/. Accessed 2021 Apr 23.

17.A guide to cerebral palsy. Toronto (ON): Ontario Federation for Cerebral Palsy; 2011: https://www.ofcp.ca/pdf/Web-Guide-To-CP.pdf. Accessed 2021 Apr 23.

18.Mapping connections: an understanding of neurological conditions in Canada. The National Population Health Study of Neurological Conditions. Toronto (ON): Neurological Health Charities Canada; 2014: https://central.bac-lac.gc.ca/.item?id=HP35-45-2014-eng&op=pdf&app=Library. Accessed 2021 Apr 23.

19.GBD 2016 Neurology Collaborators. Global, regional, and national burden of neurological disorders, 1990-2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol. 2019;18(5):459-480. PubMed

20.GBD 2016 Motor Neuron Disease Collaborators. Global, regional, and national burden of motor neuron diseases 1990-2016: a systematic analysis for the Global Burden of Disease Study 2016. Lancet Neurol. 2018;17(12):1083-1097. PubMed

21.Himmelmann K, Beckung E, Hagberg G, Uvebrant P. Gross and fine motor function and accompanying impairments in cerebral palsy. Dev Med Child Neurol. 2006;48(6):417-423. PubMed

22.Krueger H, Koot J, Hall RE, O'Callaghan C, Bayley M, Corbett D. Prevalence of individuals experiencing the effects of stroke in Canada: trends and projections. Stroke. 2015;46(8):2226-2231. PubMed

23.Nguyen R, Fiest KM, McChesney J, et al. The international incidence of traumatic brain injury: a systematic review and meta-analysis. Can J Neurol Sci. 2016;43(6):774-785. PubMed

24.Yuan J, Maserejian N, Liu Y, et al. Severity distribution of Alzheimer's disease dementia and mild cognitive impairment in the Framingham Heart Study. J Alzheimers Dis. 2021;79(2):807-817. PubMed

25.Glader L, Delsing C, Hughes A, et al. Sialorrhea in cerebral palsy. Milwaukee (WI): American Academy for Cerebral Palsy and Development Medicine; 2017: https://www.aacpdm.org/publications/care-pathways/sialorrhea. Accessed 2021 Apr 23.

26.Moller E, Karlsborg M, Bardow A, Lykkeaa J, Nissen FH, Bakke M. Treatment of severe drooling with botulinum toxin in amyotrophic lateral sclerosis and Parkinson's disease: efficacy and possible mechanisms. Acta Odontol Scand. 2011;69(3):151-157. PubMed

27.Morgante F, Bavikatte G, Anwar F, Mohamed B. The burden of sialorrhoea in chronic neurological conditions: current treatment options and the role of incobotulinumtoxinA (Xeomin R). Ther Adv Neurol Disord. 2019;12:1756286419888601. PubMed

28.Jost WH. Treatment of drooling in Parkinson's disease with botulinum toxin. Mov Disord. 1999;14(6):1057. PubMed

29.Costa J, Rocha ML, Ferreira J, Evangelista T, Coelho M, de Carvalho M. Botulinum toxin type-B improves sialorrhea and quality of life in bulbaronset amyotrophic lateral sclerosis. J Neurol. 2008;255(4):545-550. PubMed

30.Intiso D, Basciani M. Botulinum toxin use in neuro-rehabilitation to treat obstetrical plexus palsy and sialorrhea following neurological diseases: A review. NeuroRehabilitation. 2012;31(2):117-129. PubMed

31.Jackson CE, Gronseth G, Rosenfeld J, et al. Randomized double-blind study of botulinum toxin type B for sialorrhea in ALS patients. Muscle Nerve. 2009;39(2):137-143. PubMed

32.Lagalla G, Millevolte M, Capecci M, Provinciali L, Ceravolo MG. Long-lasting benefits of botulinum toxin type B in Parkinson's disease-related drooling. J Neurol. 2009;256(4):563-567. PubMed

33.Lipp A, Trottenberg T, Schink T, Kupsch A, Arnold G. A randomized trial of botulinum toxin A for treatment of drooling. Neurology. 2003;61(9):1279-1281. PubMed

34.Lagalla G, Millevolte M, Capecci M, Provinciali L, Ceravolo MG. Botulinum toxin type A for drooling in Parkinson's disease: a double-blind, randomized, placebo-controlled study. Mov Disord. 2006;21(5):704-707. PubMed

35.Mancini F, Zangaglia R, Cristina S, et al. Double-blind, placebo-controlled study to evaluate the efficacy and safety of botulinum toxin type A in the treatment of drooling in parkinsonism. Mov Disord. 2003;18(6):685-688. PubMed

36.Nobrega AC, Rodrigues B, Torres AC, Enzo A, Melo A. Does botulinum toxin decrease frequency and severity of sialorrhea in Parkinson's disease? J Neurol Sci. 2007;253(1-2):85-87. PubMed

37.Porta M, Gamba M, Bertacchi G, Vaj P. Treatment of sialorrhoea with ultrasound guided botulinum toxin type A injection in patients with neurological disorders. J Neurol Neurosurg Psychiatry. 2001;70(4):538-540. PubMed

38.Ondo WG, Hunter C, Moore W. A double-blind placebo-controlled trial of botulinum toxin B for sialorrhea in Parkinson's disease. Neurology. 2004;62(1):37-40. PubMed

39.Vashishta R, Nguyen SA, White DR, Gillespie MB. Botulinum toxin for the treatment of sialorrhea: a meta-analysis. Otolaryngol Head Neck Surg. 2013;148(2):191-196. PubMed

40.Botox (onabotulinumtoxinA for injection Ph. Eur., Clostridium botulinum type A neurotoxin complex (900kD)): sterile vacuum-dried concentrate powder for solution for injection, 50, 100 and 200 Allergan units per vial [product monograph]. Markham (ON): Allergan, Inc.; 2014: https://allergan-web-cdn-prod.azureedge.net/allergancanadaspecialty/allergancanadaspecialty/media/actavis-canada-specialty/en/products/pms/botox-pm-2014-07-07_e.pdf. Accessed 2021 Apr 29.

41.Dysport therapeutic (abobotulinumtoxinA for injection Ph. Eur.): sterile lyophilized powder for solution for injection, 300 and 500 units per vial [product monograph]. Mississauga (ON): Ipsen Biopharmaceuticals Canada Inc.; 2017: https://www.cadth.ca/sites/default/files/cdr/monograph/dysport-therapeutic-product-monograph.pdf. Accessed 2021 Apr 29.

42.Nuceiva (prabotulinumtoxinA for injection): sterile vacuum-dried powder for solution for injection, 100 units per vial [product monograph]. Santa Barbara (CA): Evolus, Inc.; 2018: https://pdf.hres.ca/dpd_pm/00046932.PDF. Accessed 2021 Apr 29.

43.Myobloc (botulinum toxin type B): 5000 units/mL, sterile solution for intramuscular injection [product monograph]. San Diego (CA): Solstice NeuroSciences, LLC.; 2004: https://pdf.hres.ca/dpd_pm/00000633.PDF. Accessed 2021 Apr 29.

44.Elavil (amitriptyline hydrochloride tablets USP): 10, 25, 50 and 75 mg tablets [product monograph]. Vaughan (ON): AA PHARMA INC.; 2018: https://pdf.hres.ca/dpd_pm/00043995.PDF. Accessed 2021 Apr 29.

45.McGowan J, Sampson M, Salzwedel DM, Cogo E, Foerster V, Lefebvre C. PRESS Peer Review of Electronic Search Strategies: 2015 guideline statement. J Clin Epidemiol. 2016;75:40-46. PubMed

46.Grey matters: a practical tool for searching health-related grey literature. Ottawa (ON): CADTH; 2019: https://cadth.ca/grey-matters-practical-tool-searching-health-related-grey-literature. Accessed 2021 Mar 22.

47.Chinnapongse R, Gullo K, Nemeth P, Zhang Y, Griggs L. Safety and efficacy of botulinum toxin type B for treatment of sialorrhea in Parkinson's disease: a prospective double-blind trial. Mov Disord. 2012;27(2):219-226. PubMed

48.Solstice Neurosciences. NCT00515437: A multi-center study of MYOBLOC for the treatment of sialorrhea in Parkinson's disease patients. ClinicalTrials.gov. Bethesda (MD): U.S. National Library of Medicine; 2019: http://clinicaltrials.gov/ct2/show/NCT00515437. Accessed 2021 May 12.

49.Hughes AJ, Daniel SE, Kilford L, Lees AJ. Accuracy of clinical diagnosis of idiopathic Parkinson's disease: a clinico-pathological study of 100 cases. J Neurol Neurosurg Psychiatry. 1992;55(3):181-184. PubMed

50.Posner K, Brown GK, Stanley B, et al. The Columbia-Suicide Severity Rating Scale: initial validity and internal consistency findings from three multisite studies with adolescents and adults. Am J Psychiatry. 2011;168(12):1266-1277. PubMed

51.Goetz CG, Tilley BC, Shaftman SR, et al. Movement Disorder Society-sponsored revision of the Unified Parkinson's Disease Rating Scale (MDS-UPDRS): scale presentation and clinimetric testing results. Mov Disord. 2008;23(15):2129-2170. PubMed

52.Loewen SC, Anderson BA. Reliability of the Modified Motor Assessment Scale and the Barthel Index. Phys Ther. 1988;68(7):1077-1081. PubMed

53.Janssen MF, Pickard AS, Golicki D, et al. Measurement properties of the EQ-5D-5L compared to the EQ-5D-3L across eight patient groups: a multi-country study. Qual Life Res. 2013;22(7):1717-1727. PubMed

54.International Conference on Harmonisation of Technical Requirements for Registration of Pharmaceuticals for Human Use. ICH Harmonised Tripartite Guidelines - Clinical safety data management: definitions and standards for expedited reporting. Geneva (CH): International Council for Harmonisation of Technical Requirements for Pharmaceuticals for Human Use (ICH); 1994: https://database.ich.org/sites/default/files/E2A_Guideline.pdf. Accessed 2021 May 12.

55.Busner J, Targum SD. The clinical global impressions scale: applying a research tool in clinical practice. Psychiatry (Edgmont). 2007;4(7):28-37. PubMed

56.Health Canada reviewer's report: Xeomin (incobotulinumtoxinA) [internal sponsor's report]. Ottawa (ON): Therapeutics Products Directorate, Health Canada; 2020.

57.Hill F, Miller N, Walsh RA, Mockler D, McDowell R, Walsh M. Botulinum toxin for drooling in Parkinson's disease. Cochrane Database Syst Rev. 2016;10:CD012408.

58.Narayanaswami P, Geisbush T, Tarulli A, et al. Drooling in Parkinson's disease: A randomized controlled trial of incobotulinum toxin A and meta-analysis of Botulinum toxins. Parkinsonism Relat Disord. 2016;30:73-77. PubMed

59.Kalf JG, Borm GF, de Swart BJ, Bloem BR, Zwarts MJ, Munneke M. Reproducibility and validity of patient-rated assessment of speech, swallowing, and saliva control in Parkinson's disease. Arch Phys Med Rehabil. 2011;92(7):1152-1158. PubMed

60.Brooks R. EuroQol: the current state of play. Health Policy. 1996;37(1):53-72. PubMed

61.EuroQoL Group. EuroQol--a new facility for the measurement of health-related quality of life. Health Policy. 1990;16(3):199-208. PubMed

Appendix 1: Literature Search Strategy

Note that this appendix has not been copy-edited.

Clinical Literature Search

Overview

Interface: Ovid

Databases:

• MEDLINE All (1946–present)

• Embase (1974–present)

• Note: Subject headings and search fields have been customized for each database. Duplicates between databases were removed in Ovid.

Date of Search: March 26, 2021

Alerts: Bi-weekly search updates until project completion

Study Types: No filters were applied to limit the main search retrieval by study type. For the search for generic terms: randomized controlled trials; controlled clinical trials.

Limits:

• No publication date limits

• For the search for generic terms: Humans

• No language limits

• Conference abstracts: excluded

Table 43: Syntax Guide

Multi-Database Strategy

Search Strategy

1. (xeomin* or xeomeen* or incobotulin* or inco A or incoA or Ainco or A inco or inco BoNT* or incoBoNT* or NT 201 or NT201 or bocouture*).ti,ab,ot,kf,hw,nm,rn.

2. Sialorrhea/ or Salivation/ or Saliva/ or exp Salivary Glands/

3. (sialorrh* or drool* or dribbl* or saliva* or hypersaliva* or hypersialorrh* or ((oral or oropharyngeal) adj1 (secret* or hypersecret*)) or polysialia or ptyalism or hyperptyalism or ptyalorrh* or sialism*).ti,ab,kf.

4. 2 or 3

5. 1 and 4

6. 5 use medall

7. (xeomin* or xeomeen* or incobotulin* or inco A or incoA or Ainco or A inco or inco BoNT* or incoBoNT* or NT 201 or NT201 or bocouture*).ti,ab,kw,dq.

8. exp hypersalivation/ or salivation/ or salivation disorder/ or saliva/ or exp salivary gland/

9. (sialorrh* or drool* or dribbl* or saliva* or hypersaliva* or hypersialorrh* or ((oral or oropharyngeal) adj1 (secret* or hypersecret*)) or polysialia or ptyalism or hyperptyalism or ptyalorrh* or sialism*).ti,ab,kw,dq.

10. 8 or 9

11. 7 and 10

12. 11 not conference abstract.pt.

13. 12 use oemezd

14. 6 or 13

15. remove duplicates from 14

Supplemental Generic Terms Search: Multi-Database Strategy

Search Strategy

1. Botulinum Toxins/ or Botulinum Toxins, Type A/

2. (Abobotulin* or AGN 191622 or AGN191622 or Allergan* or Azzalure* or Botox* or Botulax* or CNT 52120 or CNT52120 or daxibotulin* or DWP450 or DWP 450 or Dysport* or Evabotulin* or evosyal* or GSK 1358820 or GSK1358820 or HSDB 7796 or HSDB7796 or jeuveau* or Lanxoz* or letibotulin* or Mediotoxin* or Medytox* or MT 10109 or MT10109 or Nabota* or nabotulin* or Neuronox* or nuceiva* or oculinum* or onabotulin* or onaclostox* or prabotulin* or prosigne* or purtox* or QM1114 or QM 1114 or relabotulin* or reloxin* or revance* or RTT 150 or RTT150 or vistabel* or vistabex*).ti,ab,ot,kf,hw,nm,rn.

3. (Botulinum* or botulinium* or (botulin* adj2 (toxin* or neurotoxin*)) or BoNT-A or BoNTA or BTXA or BTX-A or toxine botulinique A or E211KPY694 or EINECS 297-253-4).ti,ab,ot,kf,hw,nm,rn.

4. 1 or 2 or 3

5. Sialorrhea/ or Salivation/ or Saliva/ or exp Salivary Glands/

6. (sialorrh* or drool* or dribbl* or saliva* or hypersaliva* or hypersialorrh* or ((oral or oropharyngeal) adj1 (secret* or hypersecret*)) or polysialia or ptyalism or hyperptyalism or ptyalorrh* or sialism*).ti,ab,kf.

7. 5 or 6

8. 4 and 7

9. 8 use medall

10. *botulinum toxin/ or *botulinum toxin a/

11. (Abobotulin* or AGN 191622 or AGN191622 or Allergan* or Azzalure* or Botox* or Botulax* or CNT 52120 or CNT52120 or daxibotulin* or DWP450 or DWP 450 or Dysport* or Evabotulin* or evosyal* or GSK 1358820 or GSK1358820 or HSDB 7796 or HSDB7796 or jeuveau* or Lanxoz* or letibotulin* or Mediotoxin* or Medytox* or MT 10109 or MT10109 or Nabota* or nabotulin* or Neuronox* or nuceiva* or oculinum* or onabotulin* or onaclostox* or prabotulin* or prosigne* or purtox* or QM1114 or QM 1114 or relabotulin* or reloxin* or revance* or RTT 150 or RTT150 or vistabel* or vistabex*).ti,ab,kw,dq.

12. (Botulinum* or botulinium* or (botulin* adj2 (toxin* or neurotoxin*)) or BoNT-A or BoNTA or BTXA or BTX-A or toxine botulinique A or EINECS 297-253-4).ti,ab,kw,dq.

13. 10 or 11 or 12

14. exp hypersalivation/ or salivation/ or salivation disorder/ or saliva/ or exp salivary gland/

15. (sialorrh* or drool* or dribbl* or saliva* or hypersaliva* or hypersialorrh* or ((oral or oropharyngeal) adj1 (secret* or hypersecret*)) or polysialia or ptyalism or hyperptyalism or ptyalorrh* or sialism*).ti,ab,kw,dq.

16. 14 or 15

17. 13 and 16

18. 17 not conference abstract.pt.

19. 18 use oemezd

20. 9 or 19

21. (Randomized Controlled Trial or Controlled Clinical Trial or Pragmatic Clinical Trial or Equivalence Trial or Clinical Trial, Phase III).pt.

22. Randomized Controlled Trial/

23. exp Randomized Controlled Trials as Topic/

24. 24.."Randomized Controlled Trial (topic)"/

25. Controlled Clinical Trial/

26. exp Controlled Clinical Trials as Topic/

27. "Controlled Clinical Trial (topic)"/

28. Randomization/

29. Random Allocation/

30. Double-Blind Method/

31. Double Blind Procedure/

32. Double-Blind Studies/

33. Single-Blind Method/

34. Single Blind Procedure/

35. Single-Blind Studies/

36. Placebos/

37. Placebo/

38. Control Groups/

39. Control Group/

40. (random* or sham or placebo*).ti,ab,hw,kf,kw.

41. ((singl* or doubl*) adj (blind* or dumm* or mask*)).ti,ab,hw,kf,kw.

42. ((tripl* or trebl*) adj (blind* or dumm* or mask*)).ti,ab,hw,kf,kw.

43. (control* adj3 (study or studies or trial* or group*)).ti,ab,kf,kw.

44. (Nonrandom* or non random* or non-random* or quasi-random* or quasirandom*).ti,ab,hw,kf,kw.

45. allocated.ti,ab,hw.

46. ((open label or open-label) adj5 (study or studies or trial*)).ti,ab,hw,kf,kw.

47. ((equivalence or superiority or non-inferiority or noninferiority) adj3 (study or studies or trial*)).ti,ab,hw,kf,kw.

48. (pragmatic study or pragmatic studies).ti,ab,hw,kf,kw.

49. ((pragmatic or practical) adj3 trial*).ti,ab,hw,kf,kw.

50. ((quasiexperimental or quasi-experimental) adj3 (study or studies or trial*)).ti,ab,hw,kf,kw.

51. (phase adj3 (III or "3") adj3 (study or studies or trial*)).ti,hw,kf,kw.

52. or/21-51

53. 20 and 52

54. exp animals/

55. exp animal experimentation/ or exp animal experiment/

56. exp models animal/

57. nonhuman/

58. exp vertebrate/ or exp vertebrates/

59. or/54-58

60. exp humans/

61. exp human experimentation/ or exp human experiment/

62. or/60-61

63. 59 not 62

64. 53 not 63

65. remove duplicates from 64