Pain Reduction in Cervical Dystonia Following Treatment with IncobotulinumtoxinA: A Pooled Analysis

This analysis pooled pain severity data from four phase 3 and 4 studies of incobotulinumtoxinA (incoBoNT-A) for the treatment of cervical dystonia (CD) in adults. CD-related pain severity was assessed at baseline, each injection visit, and 4 weeks after each injection of incoBoNT-A using the Toronto Western Spasmodic Torticollis Rating Scale pain severity subscale or a pain visual analog scale. Both were analyzed using a score range of 0–10 and pain was categorized as mild, moderate, or severe. Data for 678 patients with pain at baseline were assessed and sensitivity analyses evaluated pain responses in the subgroup not taking concomitant pain medication (n = 384 at baseline). At Week 4 after the first injection, there was a mean change of −1.25 (standard deviation 2.04) points from baseline pain severity (p < 0.0001), with 48.1% showing ≥ 30% pain reduction from baseline, 34.4% showing ≥50% pain reduction from baseline, and 10.3% becoming pain free. Pain responses were sustained over five injection cycles with a trend to incremental improvements with each successive cycle. Pain responses in the subgroup not taking concomitant pain medication demonstrated the lack of confounding effects of pain medications. These results confirmed the pain relief benefits of long-term treatment with incoBoNT-A.


Introduction
Cervical dystonia (CD), a chronic neurological disorder characterized by involuntary contractions of cervical muscles, causes abnormal resting positions and movements of the head, neck, and/or shoulders. The term isolated CD (iCD) is used when other movement disorders, except tremors, are not involved [1]. CD is the most common form of adult-onset focal dystonia, with a pooled prevalence estimate of 9.95 cases per 100,000 in a recent meta-analysis of epidemiological studies [2]. The patient numbers at each injection visit (IV) and control visit (CV; 4 weeks postinjection) for 5 injection cycles for the total population and by individual study are shown in Figure S1. As the data for CV6 and subsequent injection cycles were provided by only one study, we did not include the results for injection cycle 6 in our presentation of the pooled data. The number of patients with pain assessment in each injection cycle is given in Table S1.

Change in Pain Severity from Baseline to Week 4
At baseline, the mean (standard deviation [SD]) pain severity score was 4.26 (2.32) for the total population (n = 678) and 1.73 (1.06), 4.91 (0.80), and 7.40 (0.83) for the mild, moderate, and severe pain groups, respectively (Table 1). Patients treated with incoBoNT-A showed a significant reduction in the mean pain severity score at CV1 (Figure 1). There was a mean (SD) change of −1.25 (2.04) points from baseline pain severity (p < 0.0001) in the total population with pain data at CV1 (n = 669) ( Figure 1a At baseline, the mean (standard deviation [SD]) pain severity score was 4.26 (2.32) for the total population (n = 678) and 1.73 (1.06), 4.91 (0.80), and 7.40 (0.83) for the mild, moderate, and severe pain groups, respectively (Table 1). Patients treated with incoBoNT-A showed a significant reduction in the mean pain severity score at CV1 (Figure 1). There was a mean (SD) change of −1.25 (2.04) points from baseline pain severity (p < 0.0001) in the total population with pain data at CV1 (n = 669) ( Figure 1a) and this change was consistent across all four individual studies (Figure 1b).
Toxins 2023, 15, x FOR PEER REVIEW 4 of 17 Figure 1. Mean change from baseline in pain severity at control visit 1 (CV1) 4 weeks after the first injection of incobotulinumtoxinA for: (a) the total population with pain assessment at CV1 (n = 669) and by baseline pain severity (mild, moderate, severe); (b) by individual study [19,20,27,28]; and (c) for patients with pain assessment at CV1 not taking concomitant pain medication (n = 379). Pain severity score ranges from 0 to 10. Bars represent standard deviation. p-values are from one-sample t-tests.
Change in pain severity at Week 4 (CV1) following a single injection of incoBoNT-A by baseline pain severity category is presented in Figure 2a. The figure shows a shift to a lower level of pain severity at Week 4 for a good proportion of the patients in each of the baseline pain categories. Of the 244 patients with mild pain at baseline, 43 (17.6%) had no pain at CV1. Of the 285 patients with moderate pain at baseline, 21 (7.4%) had no pain and 113 (39.6%) had mild pain at CV1. Of the 140 patients with severe pain at baseline, 5 (3.6%) had no pain, 31 (22.1%) had mild pain, and 65 (46.4%) had moderate pain at CV1. Notably, 10.3% of all patients with pain at baseline (mild, moderate, or severe) reported no pain at Week 4. Additionally, the proportion of patients with mild pain increased from 36.5% at baseline to 48.0% at Week 4, while the proportion with moderate pain decreased from 42.6% at baseline to 33.3% at Week 4, and the proportion with severe pain decreased from 20.9% at baseline to 8.4% at Week 4 ( Figure S2a). Mean change from baseline in pain severity at control visit 1 (CV1) 4 weeks after the first injection of incobotulinumtoxinA for: (a) the total population with pain assessment at CV1 (n = 669) and by baseline pain severity (mild, moderate, severe); (b) by individual study [19,20,27,28]; and (c) for patients with pain assessment at CV1 not taking concomitant pain medication (n = 379). Pain severity score ranges from 0 to 10. Bars represent standard deviation. p-values are from one-sample t-tests.
Change in pain severity at Week 4 (CV1) following a single injection of incoBoNT-A by baseline pain severity category is presented in Figure 2a. The figure shows a shift to a lower level of pain severity at Week 4 for a good proportion of the patients in each of the baseline pain categories. Of the 244 patients with mild pain at baseline, 43 (17.6%) had no pain at CV1. Of the 285 patients with moderate pain at baseline, 21 (7.4%) had no pain and 113 (39.6%) had mild pain at CV1. Of the 140 patients with severe pain at baseline, 5 (3.6%) had no pain, 31 (22.1%) had mild pain, and 65 (46.4%) had moderate pain at CV1. Notably, 10.3% of all patients with pain at baseline (mild, moderate, or severe) reported no pain at Week 4. Additionally, the proportion of patients with mild pain increased from 36.5% at baseline to 48.0% at Week 4, while the proportion with moderate pain decreased from 42.6% at baseline to 33.3% at Week 4, and the proportion with severe pain decreased from 20.9% at baseline to 8.4% at Week 4 ( Figure S2a).

Pain Severity, Response Rates, and Complete Pain Relief Following Multiple Injection Cycles
Changes in pain severity by category of pain severity were maintained over multip injection cycles ( Figure S3). Of the 669 patients with pain data at CV1, 322 (48.1%) patien had ≥30% pain reduction from baseline and 230 (34.4%) patients had ≥50% pain reducti from baseline; these patients were classified as responders.
Pain response rates (≥30% or ≥50% reduction in pain score from baseline) were su tained following multiple injections of incoBoNT-A (Figure 3a). The results also showed th pain had not returned to the baseline level by the time of their next injection, demonstrati a cumulative effect throughout the cycles. In total, 29.9% at IV2 to 36.1% at IV5 were clas fied as responders (≥30% pain reduction from baseline) and 15.3% at IV2 to 23.5% at I were classified as responders (≥50% pain reduction from baseline) (Figure 3a).

Pain Severity, Response Rates, and Complete Pain Relief Following Multiple Injection Cycles
Changes in pain severity by category of pain severity were maintained over multiple injection cycles ( Figure S3). Of the 669 patients with pain data at CV1, 322 (48.1%) patients had ≥30% pain reduction from baseline and 230 (34.4%) patients had ≥50% pain reduction from baseline; these patients were classified as responders.
Pain response rates (≥30% or ≥50% reduction in pain score from baseline) were sustained following multiple injections of incoBoNT-A (Figure 3a). The results also showed that pain had not returned to the baseline level by the time of their next injection, demonstrating a cumulative effect throughout the cycles. In total, 29.9% at IV2 to 36.1% at IV5 were classified as responders (≥30% pain reduction from baseline) and 15.3% at IV2 to 23.5% at IV5 were classified as responders (≥50% pain reduction from baseline) (Figure 3a).

Pain Severity, Response Rates, and Complete Pain Relief Following Multiple Injection Cycles
Changes in pain severity by category of pain severity were maintained over multiple injection cycles ( Figure S3). Of the 669 patients with pain data at CV1, 322 (48.1%) patients had ≥30% pain reduction from baseline and 230 (34.4%) patients had ≥50% pain reduction from baseline; these patients were classified as responders.
Pain response rates (≥30% or ≥50% reduction in pain score from baseline) were sustained following multiple injections of incoBoNT-A ( Figure 3a). The results also showed that pain had not returned to the baseline level by the time of their next injection, demonstrating a cumulative effect throughout the cycles. In total, 29.9% at IV2 to 36.1% at IV5 were classified as responders (≥30% pain reduction from baseline) and 15.3% at IV2 to 23.5% at IV5 were classified as responders (≥50% pain reduction from baseline) (Figure 3a).  The proportion of patients with complete pain relief at the control visit increase slightly over the first 5 injection cycles from 10.3% at CV1 to 16.8% CV5 (Figure 4a). Not bly, a small proportion of patients were pain free at the time of the second and subsequen injection visits. The proportion of patients with complete pain relief at the control visit increased slightly over the first 5 injection cycles from 10.3% at CV1 to 16.8% CV5 (Figure 4a). Notably, a small proportion of patients were pain free at the time of the second and subsequent injection visits.

Pain Severity Responses in Patients Not Taking Concomitant Pain Medication
Of the 678 patients with pain at baseline, 435 (64.2%) were not taking concomitant pain medication at IV1. The proportions of patients taking any pain medication were similar at IV1 and CV1 (Table S2), and 9% of patients received benzodiazepines or muscle relaxants at IV1 and CV1. There was no trend towards increased use of analgesic medication during the study. The pain severity responses in the subgroups of incoBoNT-Atreated patients with pain at baseline and patients not taking concomitant pain medication were consistent with the findings presented above for patients with pain at baseline regardless of concomitant pain medication. These results indicated that concomitant pain medication did not influence the pain response rates. At IV1, the mean (SD) pain severity score was 3.83 (2.41), with 43.2% reporting mild pain, 39.3% reporting moderate pain, and 17.5% reporting severe pain in the subgroup not taking concomitant pain medication (n = 384). The mean (SD) change in pain severity score from baseline at CV1 following a single injection of incoBoNT-A for the subgroup without concomitant pain medication and pain data at CV1 (n = 379) was −1.29 (1.96; p < 0.0001) (Figure 1c). The change in pain severity at CV1 by baseline pain severity category ( Figure 2b) showed a shift towards less severe pain. Of the 379 patients with pain at IV1 (mild, moderate, or severe) and pain data at CV1, 50 (13.2%) reported no pain at CV1 ( Figure S2b). Additionally, the number of patients (%) with mild pain increased from 164 (43.3%) at IV1 to 205 (54.1%) at CV1, while there were decreases in the numbers (%) of patients with moderate pain from 148 (39.1%) to 99 (26.1%) and with severe pain from 67 (17.7%) to 25 (6.6%) ( Figure S2b).
The response rates at CV1 among patients not taking concomitant pain medication were 54.4% (≥30% pain reduction from baseline) and 41.4% (≥50% pain reduction from

Pain Severity Responses in Patients Not Taking Concomitant Pain Medication
Of the 678 patients with pain at baseline, 435 (64.2%) were not taking concomitant pain medication at IV1. The proportions of patients taking any pain medication were similar at IV1 and CV1 (Table S2), and 9% of patients received benzodiazepines or muscle relaxants at IV1 and CV1. There was no trend towards increased use of analgesic medication during the study. The pain severity responses in the subgroups of incoBoNT-A-treated patients with pain at baseline and patients not taking concomitant pain medication were consistent with the findings presented above for patients with pain at baseline regardless of concomitant pain medication. These results indicated that concomitant pain medication did not influence the pain response rates. At IV1, the mean (SD) pain severity score was 3.83 (2.41), with 43.2% reporting mild pain, 39.3% reporting moderate pain, and 17.5% reporting severe pain in the subgroup not taking concomitant pain medication (n = 384). The mean (SD) change in pain severity score from baseline at CV1 following a single injection of incoBoNT-A for the subgroup without concomitant pain medication and pain data at CV1 (n = 379) was −1.29 (1.96; p < 0.0001) (Figure 1c). The change in pain severity at CV1 by baseline pain severity category ( Figure 2b) showed a shift towards less severe pain. Of the 379 patients with pain at IV1 (mild, moderate, or severe) and pain data at CV1, 50 (13.2%) reported no pain at CV1 ( Figure S2b). Additionally, the number of patients (%) with mild pain increased from 164 (43.3%) at IV1 to 205 (54.1%) at CV1, while there were decreases in the numbers (%) of patients with moderate pain from 148 (39.1%) to 99 (26.1%) and with severe pain from 67 (17.7%) to 25 (6.6%) ( Figure S2b).
The response rates at CV1 among patients not taking concomitant pain medication were 54.4% (≥30% pain reduction from baseline) and 41.4% (≥50% pain reduction from baseline), and the response rates remained high over subsequent injection cycles (Figure 3b). The proportion of patients not taking concomitant pain medication with complete pain relief increased from 13.2% at CV1 to 22.1% at CV5 (Figure 4b), showing a cumulative effect over time. The proportion of patients who remained pain-free at the time of their next injection of incoBoNT-A ranged from 5.5% to 9.7%.

Discussion
In this pooled analysis, we investigated the clinical course of pain severity in a large cohort of patients with CD from a wide geographic area (Europe and USA) treated with incoBoNT-A. We observed significant pain reduction during the first injection cycle, which was sustained throughout several treatment cycles. These reductions in pain severity were also seen in patients who had already been successfully treated with BoNTs before study enrolment (82% of cases in this analysis). Pain reduction/relief was also consistently observed in the subgroup of patients not taking concomitant pain medication, confirming the benefits of long-term treatment with incoBoNT-A. Effective pain control in CD is important to prevent the development of chronic pain [29].
As expected, a high proportion of patients had pain at baseline (64%), which was moderate or severe in 64% of these patients. Our total population of incoBoNT-A-treated patients with pain at baseline comprised mostly middle-aged women (68% women, mean age 53.6 years), consistent with the prevalence data that women are approximately two times more likely than men to have CD and the mean age of CD onset is 42 years [30].
Our results showed a reduction in pain severity after treatment with incoBoNT-A. At Week 4 after the first injection, we observed a mean reduction of 1.25 points in the pain severity score from baseline (score range 0-10), and a large proportion of patients met the response criteria: 48.1% with ≥30% pain reduction from baseline level and 34.4% with ≥50% pain reduction from baseline level. The pain reduction shown by this analysis can be considered clinically relevant, as it is consistent with the following criteria. The CD-PROBE study suggested that a 2-point change in the TWSTRS pain score (score range 0-20) could be considered a minimum clinically important change [31], and the Initiative on Methods, Measurement, and Pain Assessment in Clinical Trials (IMMPACT) recommendations on chronic pain [32] state that reductions in pain intensity of ≥30% and ≥50% appear to reflect at least moderate or substantial clinically important improvements.
The observed reduction in pain severity was sustained over repeated injections of incoBoNT-A and there was a trend over the five treatment cycles towards incremental pain reduction/relief. These results showed that even if pain was still present at the control visit (4 weeks post-injection), it was less severe than at the injection visit of that treatment cycle. Moreover, the proportion of patients with mild pain increased, while the proportion of patients with moderate or severe pain decreased relative to that at the injection visit. Importantly, we found that 10.3% of patients with pain at baseline were pain-free at Week 4 after the first incoBoNT-A injection, and the proportion of patients who experienced complete pain relief increased over each subsequent injection cycle to 16.8% at the 5th injection cycle. As only 18% of the patients in our analyses were BoNT naïve, previous treatment with other BoNTs may have attenuated the analgesic effect of incoBoNT-A.
Previous studies of BoNT-A treatment in CD have shown persistent reductions in pain during long-term treatment with aboBoNT-A [33,34], onaBoNT-A [5,35], and incoBoNT-A [21,27,28] when pain was scored using the TWSTRS and/or a pain numeric rating scale. A ≥30% improvement from baseline in pain score was also noted in a large cohort of patients with moderate-to-severe pain treated with onaBoNT-A in the CD-PROBE observational study [24] and in a meta-analysis of observational studies of aboBoNT-A in patients with CD [26], supporting the use of ≥30% reduction in pain as a clinically relevant measure.
Pain severity was assessed using established and validated pain scales (TWSTRS severity subscale and pain visual analog scale [VAS]) that are widely used in clinical trials and have been shown to correlate well with each other [5]. The TWSTRS covers a range of CD features, including disease severity, functional ability, and pain [36]. IncoBoNT-A has been shown to effectively control CD symptoms after each injection, as reflected by improvements in the TWSTRS total score and TWSTRS subscale scores for motor severity, disability, and pain [19][20][21]27,28]. However, the TWSTRS does have several drawbacks, including its complexity for use in clinical practice and it does not consider the location of pain [36], which may lead to an underestimation of pain severity.
The TWSTRS pain subscale not only measures pain severity but also the duration of pain and the contribution of pain to disability, but these were inconsistently assessed across the studies included in the pooled analysis. Therefore, these factors were not analyzed. We also did not examine changes in the TWSTRS disability score in this pooled analysis. However, it is well known that pain contributes to disability in CD [11] and that BoNT treatment, including incoBoNT-A, reduces both pain and disability [20,21,27,28,37]. A recent case-control study found that patients with CD-related pain had significantly higher TWSTRS severity and disability scores than those without pain and that pain reduction following BoNT injection lasted longer than muscle relaxation in 85.3% of patients with pain, although muscle relaxation preceded pain improvement for 53% of these patients [8].
As we examined changes in pain severity using TWSTRS severity subscale or pain VAS scores, comparing our data with those of other studies is difficult. However, despite different patient populations and methods of assessing pain, clinical studies with other BoNT-A formulations in CD have reported changes in pain severity that are consistent with the mean change of −1.25 on a scale of 0-10 reported for incoBoNT-A in our analysis ( Figure 1a). For example, in a pooled analysis of data for aboBoNT-A in CD, an adjusted mean change of −3.2 ± 0.2 points at Week 4 was reported for aboBoNT-A using the TWSTRS pain score (scale range 0-20) [9], with a separate study reporting a mean change of −3.7 points from baseline at Week 4 for aboBoNT-A using the TWSTRS pain subscale score (scale range 0-20) [33].
Many CD patients report having re-emergent motor symptoms before the next BoNT-A injection. In the Carenity survey, 88% of patients reported symptom re-emergence at approximately 10.5 weeks from injection [38]. The benefits of BoNT-A start to wear off after about 8 weeks, creating a cyclical response known as the "yo-yo" effect [39] that may contribute to patient dissatisfaction, reduced QoL, and discontinuation of therapy [40]. The "yo-yo" effect varies between individuals and by the BoNT-A preparation used [39]. A novel finding from our analyses was that CD-associated pain demonstrated this "yo-yo" effect across BoNT-A injection cycles. Re-emergent pain before the next injection is likely to have an impact on patients. In the Carenity patient survey, pain was one of the first symptoms to re-emerge as the effects of BoNT-A wore off, and this waning effect had an impact on the patient's ability to work, their daily activities, and QoL [38]. Our results showed that although around 5% of patients remained pain free at the time of their next incoBoNT-A injection (Figure 4a), pain levels increased between CV and the next IV, as reflected in the lower proportion of patients with ≥30% or ≥50% pain reduction at the next IV (Figure 3a). Indeed, the increasing proportion of patients classified as responders at successive injection visits implied progressive improvements in pain with repeated incoBoNT-A treatments. It would be interesting to further explore these findings of a cyclical pattern of pain reduction and recurrence between incoBoNT-A injections as CD therapy aims to reduce pain or keep patients pain-free.
Pain is also an important determinant of QoL in patients with CD and can be assessed using various scales. The CDQ-24, a 24-item disease-specific health-related QoL measure, includes a pain subscale that has demonstrated sensitivity to change following BoNT treatment and good correlation with other pain rating scales [41,42]. The CDQ-24 pain scale also measures frequency of pain, which is not measured by the TWSTRS, and is useful for assessing the impact of pain on QoL. Further clinical studies on pain reduction in CD with BoNT therapy should include both scales.
Pain is strongly associated with emotional well-being and depression [43] and is a relevant contributor to the severe negative impact of CD on health and disability as it limits patients' activities and participation, as measured using the International Classification of Functioning, Disability, and Health [44]. Higher pain scores (pain VAS and TWSTRS pain score) have been associated with worse patient satisfaction with BoNT-A treatment [45], but patient satisfaction is a different measure from pain rating and we did not assess patient satisfaction in the current pooled analysis. Although we focused on pain severity, we can speculate that the reduction in pain severity seen with incoBoNT-A injections is likely to be associated with greater patient satisfaction with treatment, less disability, and improved QoL.
It has been reported that approximately two-thirds of patients with CD use oral/systemic analgesics to manage their CD-related pain [46,47]. In our pooled analysis, only 36% of patients with pain at baseline reported using concomitant pain medication at the first injection visit. This may be because most participants (82%) had already been successfully treated with BoNTs before study enrolment, which would contribute to an underestimation of the pain-relieving effect of incoBoNT-A.
Our results in the subgroup of CD patients with pain not taking pain medication (about two-thirds of the total population) confirmed that the pain reduction/relief we observed was a result of incoBoNT-A treatment and not due to analgesic medication.
The limitations of our analysis include potential bias due to different patient numbers and varied numbers of injection cycles in the different studies used in the pooled analysis. The effect of incoBoNT-A on pain may have been underestimated because most (82%) patients had been previously treated with other BoNTs. Additionally, pain was assessed as a secondary outcome measure in all four studies, and pain assessment at the control visit (4 weeks after injection) may not have been the peak time of pain relief, potentially underestimating the effect of incoBoNT-A on reducing pain. The scales used to assess pain in the studies included in the pooled analysis did not give a full profile of pain, so we focused on pain severity. One study in the pooled analysis used a pain VAS and, even though the pain VAS scores were transformed to a 0-10 scale, the pain VAS may not have provided an identical measure of pain severity as the TWSTRS severity subscale. Nevertheless, a pain VAS is recommended by IMMPACT for use in pain studies to measure pain intensity and the effectiveness of treatments [32]. Patients with no pain at baseline were excluded from our pooled analysis, but it is possible that some of these patients developed pain during the study.

Conclusions
This pooled analysis showed that patients with CD-related pain experienced clinically relevant and sustained reductions in pain during repeated incoBoNT-A injections, with a trend toward greater pain relief over successive treatment cycles. The level of pain after incoBoNT-A injection was reduced in severity for many patients, leading to increases in the proportions of patients with mild or no pain. After repeated injections, 16.8% of patients became pain-free at 4 weeks post-injection. The pain reduction/relieving effects of incoBoNT-A were also observed in the large subgroup of patients not taking concomitant pain medication. Our results provide further evidence to support the long-term use of incoBoNT-A for pain relief in patients with CD. Active questioning about CD-related pain in clinical practice may be important for the timing of incoBoNT-A re-injection and to improve pain management, thereby avoiding cycles of pain relief and recurrence.

Studies Included in the Pooled Analysis
Analysis was based on the pooled results from four prospective, multicenter, phase 3 and 4 studies conducted for incoBoNT-A treatment of CD in adults. These studies included a phase 3, double-blind, randomized, active-control (onaBoNT-A), non-inferiority study conducted in the European Union [19]; a phase 3, randomized, double-blind, placebocontrolled study of fixed doses in the USA (NCT00407030) [20,21]; a phase 4 study in Germany with flexible injection intervals and dosing (NCT00541905) [27]; and a phase 4 randomized, non-inferiority study in the USA comparing two injection schedules (shorter interval of 8 ± 2 weeks vs. longer interval of 14 ± 2 weeks) (NCT01486264) [28]. The main details of the four studies included in the analysis are summarized in Table 2. Two studies were randomized, double-blind, multicenter, phase 3 studies and two studies were openlabel phase 4 studies. One study was placebo-controlled [20] and one study had an active control group [19]. Two studies enrolled patients who had prior treatment with botulinum toxin [19,26], while the other two studies included both previously treated and BoNT-A naïve patients. All pre-treated subjects had their last BoNT injection at least 10 weeks prior to study entry and their first incoBoNT-A injection. The number of injection cycles of incoBoNT-A varied from 1 [19] to 11 [28]. Two studies had a main period after a single injection of up to 20 weeks [20] or 24 weeks [27], followed by a long extension period. Both the total dose of incoBoNT-A and the interval between injection sessions were fixed in some studies and flexible in others.
All four studies used the TWSTRS [48,49]. The TWSTRS provides a total score (range 0-85) and three subscores for CD severity (range 0 = mild to 35 = severe), disability due to CD (range 0-30), and CD-related pain (range 0-20). The TWSTRS pain score is based on the patient's subjective assessment and has three items: severity of pain, duration of pain, and degree of disability due to pain. In the pain severity score, the usual, worst, and least levels of pain during the last week are each rated on an 11-point scale (range 0-10). The pain severity score is calculated as (2 x usual + best + worst)/4 and has a maximum score of 10 points. Duration of pain and disability due to pain are each scored on 6-point scales (range 0-5). The total pain score is calculated as the sum of the three scores, with lower values indicating less pain.
The TWSTRS severity subscale was rated by patients in three studies [20,27,28] and used in the present analysis. Patients in the study by Benecke et al. [19] only completed the TWSTRS disability subscale, but they used a 100 mm pain VAS ranging from 0 (no pain) to 100 (worst possible pain) to rate their present sensation of pain. Randomized, double-blind, active-controlled, parallel-group, multicenter, non-inferiority study in the EU to investigate the safety and efficacy of incoBoNT-A compared to onaBoNT-A in patients with CD.
n = 463 Adults with CD (predominantly of the rotational form; i.e., spasmodic torticollis) who had shown stable responses in at least two previous onaBoNT-A injection sessions prior to study entry. Last injection session at least 10 weeks prior to randomization. Baseline TWSTRS severity score ≥10.
Change in CD severity using the TWSTRS severity score from baseline to control visit on day 28 ± 7 after injection of study medication.

Analyses
This pooled analysis evaluated data only for incoBoNT-A-treated patients with pain at baseline. We did not compare incoBoNT-A results with placebo or active control as these data were derived from one study each. Data for different incoBoNT-A dosages used across the studies were pooled, as the change in TWSTRS pain scores did not differ between groups in the studies using different doses or dosing intervals of incoBoNT-A [20,28].
Data were analyzed at the IV and at the CV 4 weeks (±3 or 7 days) after each injection of incoBoNT-A.
This pooled analysis focused on pain severity data, including TWSTRS severity scores (range 0-10) and pain VAS scores (range 0-100), which were transformed to a scale of 0-10 by dividing by 10. Pain severity was categorized as no pain (score = 0), mild pain (score >0 to <3.5), moderate pain (score 3.5 to <6.5), or severe pain (score 6.5-10). The proportion of responders was calculated with response defined as ≥30% or ≥50% reduction from baseline in the pain severity score, reflecting at least moderate or substantial clinically important improvements, respectively [32]. The percentage of patients with complete pain relief (pain score = 0) was also evaluated at each IV and CV.
Data are presented using descriptive statistics as mean (SD) or n (%). Change in pain severity from baseline to CV1 (Week 4 in the first injection cycle) was assessed using one-sample t-tests.
Since concomitant pain relief medications (analgesics and non-steroidal anti-inflammatory drugs) were allowed in all four studies, with no restrictions on dosing and frequency of intake, the percentage of patients using concomitant pain medication at each IV and CV was determined. As a sensitivity analysis, we examined pain severity responses in the subgroup of patients not taking concomitant pain-relieving medication.
All analyses were performed using Statistical Analysis Software (SAS) version 9.4.

Supplementary Materials:
The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/toxins15050333/s1, Figure S1: Number of incobotulinumtoxinAtreated patients with pain severity assessed during injection cycles 1-5 for the total population (numbers given outside each bar) and by study; Figure S2: Percentage of patients in each pain severity category at baseline and control visit 1 (CV1) for: (a) total population (n = 669) and (b) subgroup of patients not taking concomitant pain medication (n = 379); Figure S3: Category of pain severity at each visit following multiple injections of incobotulinumtoxinA. The percentages in each bar are based on the number of patients with a pain severity assessment at that visit (n value). CV, control visit IV, injection visit; Table S1: Number of incobotulinumtoxinA-treated patients with pain severity assessed at the injection visit (IV) and control visit (CV) of each injection cycle for the total population and by individual study; Table S2: Use of concomitant pain medication in the first injection cycle by patients who had pain at baseline and data available on pain medication use.