The Necessity of a Locally Active Antidote in the Clinical Practice of Botulinum Neurotoxin Therapy: Short Communication

Recently, it was demonstrated that copper complexes and 3,4-diaminopyridine can effectively reduce the activity of the botulinum neurotoxin light chain. The aim of the present study was to indicate that treatment with an antidote may have a major influence, not only on the extremely rare disease of botulism, but also on the much more frequently occurring side effects experienced during BoNT therapy. This was a retrospective chart review of patients who were regularly treated with BoNT for various indications. The percentage of patients with clinical signs of overdosing was determined. In patients with facial dystonia, double vision and ptosis occurred as side effects. In patients with cervical dystonia, neck weakness and dysphagia were observed as the most frequent side effects. In oromandibular and oropharyngeal dystonia, abnormal tongue movements and dysphagia occurred frequently. In writer’s cramp and mild post-stroke hand spasticity, severe paresis of the injected and non-injected finger muscles was observed. Additionally, in the BoNT treatment of pain syndromes (such as tension headaches or migraines), neck weakness may occur. Across all indications for clinical BoNT applications, clinical signs of BoNT overdosing may occur in up to 5% of the BoNT-treated patients. Therefore, the development of an antidote for BoNT overdoses would be very much appreciated and would have a major influence on the management of BoNT therapy.


Introduction
Botulinum neurotoxins (BoNT) are among the most potent biological toxins [1], they may cause foodborne botulism [2,3], and they continue to be a bioterrorism threat [4,5]. Therefore, effective antiBoNT compounds are needed to reduce the risk of life-threatening BoNT overdoses. Since the pioneering use of BoNT/A, by A. Scott, to reduce muscular dysbalance in strabism [6], the field of clinical applications for BoNTs is continuously growing [7]. Moreover, in all these indications, the risk of a BoNT overdose exists. In most of these cases, a BoNT overdose is not life-threatening but may lead to a considerable impairment over many weeks before the effect of the overdose fully declines [7]. Therefore, the development of a locally acting BoNT antidote would be very much appreciated.
For the development of an effective antidote, knowledge about the action of BoNTs is essential. The action of BoNTs results from the following: the binding of the heavy chain to a cell membrane; endocytosis; the resolution of the disulfide bond between the heavy and the light chain; the translocation of the light chain (LC) into the cytosol of the cell; and the cleavage of the LC as a zinc metalloprotease of the components (such as SNAP25) of the soluble NSF attachment protein receptor (SNARE) complex, which is responsible for docking vesicles at the presynaptic membrane [8][9][10]. The duration of the BoNT action depends on the survival time of the LC in the cytosol.

Materials and Methods
This retrospective, monocentric study was performed according to the declaration of Helsinki and the guidelines for good clinical practice (GCP). The local ethics committee of the University of Düsseldorf (Germany) allows the publication of retrospective data, as long as names are pseudonymized.
For recruitment, all the charts of the patients in the private BoNT ambulance of the first author (HH), between 1 January and 1 April 2021, who had received at least 4 BoNT/A injections every 3 months, were screened for documented adverse events (ADEV). During our clinical routine BoNT treatment, only those ADEVs that were spontaneously reported by the patients were documented. We distinguished between general adverse events (G-ADEV) and BoNT-associated adverse events (BoNT-ADEV). For a BoNT-ADEV, a clear temporal and spatial relationship to a previous injection must exist. A temporal relationship can be suspected if the ADEV occurs within 3 weeks after an injection. A spatial relationship can be suspected if the anatomical structure involved is in the neighborhood of an injected structure (such as a tension headache on one side after an injection in the neck muscles on the other side).
As a new aspect, BoNT-ADEVs were highlighted that probably would have been treatable by a locally acting BoNT antidote (ADOT-BoNT-ADEV), if such a substance was available. In the following, a complete list of percentages of BoNT-ADEVs per indication in our cohort (Table 1) and a description of ADOT-BoNT-ADEVs are presented (Table 2). Demographical as well as treatment-related data were also extracted from the charts (see Table 1). The total dose per session and doses per site and the preparation used were documented. All three BoNT/A preparations licensed in Europe were used. For the sake of comparison, doses were transformed into unified dose units, following European consensus recommendations [22]. Doses of incobotulinum A (incoBoNT/A; Xeomin ® , Merz Pharmaceuticals, Germany) and onabotulinum A (onaBoNT/A; Botox ® , Allergan, Erwan, CA, USA) remained unchanged; abobotulinum A doses (aboBoNT/A; Dysport ® , IpsenPharma, Paris, France) were divided by 3.
There was only one patient with cervical dystonia (CD) who developed secondary treatment failure (STF) and was switched to incoBoNT/A [23], without any effect. They were then treated with 12500 U rimabotulinum B (rimaBoNT/B; Soliste, Neuro/MyoBloc ® , USA). This patient was not included in Tables 1 and 2.  ADOT Patient 5, with epsilon-glycan-positive dystonia, who was treated because of severe head jerks to the right side, experienced weakness of shoulder elevation after injection of the right splenius capitis and levator scapulae muscle with 200 U Xeomin ® .

HFS 2
Patient 1, with blepharospasm and additional hemifacial spasm on the left side. After injection of 60 U Dysport ® into the right orbicularis oculi muscle, 10 U Dysport ® per edge of the right eye lid, and 7.5 U Dysport ® into the edges of the left eye lid, a moderate ptosis on the left side developed, despite dose reduction on the left side ( Figure 2).

Demographical and Treatment-Related Data
The charts of 198 adult patients who received BoNT/A injections between 1 January and 1 April 2021, by HH were screened. One patient, who was treated with rimaBoNT/B, was excluded from further analysis.
The mean total doses per session were slightly higher than recommended in the license descriptions. This was due to the fact that, in comparison to other cohorts of patients, our patients suffered from a high complexity of clinical presentations. More than 60% of these patients were referred from other BoNT centers because of difficulties during their treatment, the dissatisfaction of the patients or the treating physicians, or suspected secondary treatment failure.

ADEVs and BoNT-ADEVs
Eighteen patients (=9.14%) reported ADEVs. Three of these ADEVs were not classified as BoNT-mediated. All three of these ADEVs occurred in patients with CD. One female patient, who already suffered from segmental dystonia (blepharospasm, oromandibular, and cervical dystonia), reported difficulties in swallowing and breathing during walking after receiving the last injection of 750 U aboBoNT/A. Since these difficulties persisted until the end of the injection cycle, we suspect that this was due to the generalization

Demographical and Treatment-Related Data
The charts of 198 adult patients who received BoNT/A injections between 1 January and 1 April 2021, by HH were screened. One patient, who was treated with rimaBoNT/B, was excluded from further analysis.
The mean total doses per session were slightly higher than recommended in the license descriptions. This was due to the fact that, in comparison to other cohorts of patients, our patients suffered from a high complexity of clinical presentations. More than 60% of these patients were referred from other BoNT centers because of difficulties during their treatment, the dissatisfaction of the patients or the treating physicians, or suspected secondary treatment failure.

ADEVs and BoNT-ADEVs
Eighteen patients (=9.14%) reported ADEVs. Three of these ADEVs were not classified as BoNT-mediated. All three of these ADEVs occurred in patients with CD. One female patient, who already suffered from segmental dystonia (blepharospasm, oromandibular, and cervical dystonia), reported difficulties in swallowing and breathing during walking after receiving the last injection of 750 U aboBoNT/A. Since these difficulties persisted until the end of the injection cycle, we suspect that this was due to the generalization of dystonia, rather than to a systemic side effect. The other two male patients suffered from a complex cervical dystonia and reported difficulty swallowing after receiving the last injection. In both cases, these difficulties persisted for more than 6 weeks and probably indicated a transition from CD to segmental dystonia. In 15 patients (=7.6%), the ADEVs were BoNT-mediated and they are described in detail in Table 2.
The percentage of BoNT-ADEVs varied considerably across the indication subgroups. Because of the small sample size, the BoNT-ADEV percentage was the highest in the HYPER subgroup. The second highest percentage occurred in the OMD/OPD and MEIGE subgroup. No BoNT-ADEVs were reported in the PAIN and GEN DYS subgroup.
The most frequent side effect was neck pain and neck muscle weakness in patients with CD. An example is presented in Figure 1.
The second most frequent side effect was ptosis in the treatment of hemifacial spasms and blepharospasms. An unusual case, with a blepharospasm and an additional hemifacial spasm on the left side is presented in Figure 2.
Third, weakness of the middle finger extensor was demonstrated as an example that shows that a special muscle group may be much more sensitive to BoNT than another muscle group nearby (Figure 3).
Eleven of the fifteen BoNT-ADEVs were classified as ADOT-BoNT-ADEVs. This means that we would have tried to reduce the severity of the BoNT side effects by injecting a locally acting antidote in 5.55% of the patients and in 73% of the BoNT-ADEVs.

Safety and Efficacy of BoNT/A Injections
Treatment with repeated injections of BoNT/A is a generally safe and effective therapy for a variety of indications. Because of its safety and efficacy profile, BoNT/A has received level A recommendations from the American Academy of Neurology (AAN) and the Royal Society of Physicians for a variety of indications [24,25]. The frequency of adverse events depends on the indication (see Table 1 and compare [7,26,27]). In a double-blind study, comparing the ona-and incoBoNT/A treatment of CD [25], 25% to 30% of the patients in both the medication groups experienced ADEVs [26,28]. In a large open-label study that analyzed the safety and efficacy of the first BoNT/A injection in 516 BoNT-naïve CD cases, up to 43% of patients reported ADEVs [29]. Furthermore, in placebo-controlled trials of CD, up to 53% of the placebo-treated patients reported ADEVs [30]. This suggests that a large percentage of reported ADEVs in controlled and open trials were not BoNT-associated.
In a recent review of BoNT/A and BoNT/B treatment of CD, only 21% of the reported ADEVs were BoNT-related [30]. Taking into account this result, the frequency of BoNT-ADEVs appears to be rather low and varies between 5% and 9% in double-blind or openlabel studies. Therefore, the overall frequency of BoNT-ADEVs of 7.6% and the frequency of BoNT-ADEVs of 6.7% in the CD subgroup, corresponds with the numbers reported in the literature.

Two Types of BoNT-ADEVs
In general, the frequency of the side effects declines with repeated applications and with the experience of the treating physician. In the present cohort, most patients were referred from other centers because of complications, a lack of efficacy, or the induction of antibodies and side effects. This is the reason why the mean total doses for different indications were rather high.
In clinical practice, BoNT-related side effects can hardly be avoided. Especially in difficult-to-treat cases or in suspected secondary treatment failure (as in many cases from our cohort), side effects occur quite frequently. This results from the treatment of structures (muscles, glands) close to other BoNT-sensitive structures or the use of high doses with the risk of a spread to neighboring structures.
In the present paper, we have distinguished the following two types of BoNT-associated side effects: (i) side effects that result from the overdosing of a special muscle or gland, and (ii) side effects resulting from the spread of BoNT into neighboring structures. In the first case, an antidote treatment (if available) may be considered; in the second case, it is less likely that an improvement can be achieved by antagonizing the side effect with an antidote.
Both types of side effects occur after the uptake and onset of the action of BoNT. A critical question remains, regarding what delay-dependent amount of improvement can be achieved when the antidote is administered post-BoNT injection. This question can only be answered when an antidote is available and specific studies are designed in humans. In cell-based experiments, on the application of copper complexes, the cleavage of SNAP25 could be inhibited in a time-dependent manner, within a short time window of approximately 8 h [20]. When BoNT/A is injected into humans, the uptake of BoNT/A also takes some time. Therefore, it may very well be that, in humans, the time window will be even larger than 8 h.

Combined Use of BoNT Injections and Antidote Application
Much effort is being invested in the development of mutants of BoNT/A, with a shorter duration of action [31]. The use of such mutants is only reasonable in a few applications. If applied repeatedly, the risk of antibody formation is high, as is known from the treatment with BoNT F [32]. Another strategy could be to combine the application of a BoNT preparation with a normal duration and to antagonize the effect of LC activity, if necessary.
The attempt to influence the duration of the action of BoNT/A by the additional application of BoNT/E, which also cleaves SNAP25 but in a different place than BoNT/A, failed to reduce BoNT/A-induced paralysis [33]. However, continuous infusion with 3,4-DP showed a symptomatic reversal of BoNT-induced paralysis and antidotal efficacy after a lethal systemic injection with BoNT/A [18]. Therefore, the systemic application of 3,4-DP may have a beneficial influence on the second type of BoNT/A-induced side effects.
Recent experiments on the reduction in LC activity by means of heavy metal complexes, especially of copper and mercury, also appear to be promising. In cell cultures, as well as in BoNT-treated mice, a reduction in LC activity was demonstrated by the dependence on the delay after the administration of copper complexes, following the BoNT injection [20]. For humans, mercury is too toxic. Copper also has a high cell toxicity, as is known from WD [34,35]. In WD, the amount of free copper is elevated in the blood, but in clinical practice, this elevation is not high enough to reduce the BoNT/A activity significantly [21]. Furthermore, whether copper complexes can be administered in high enough doses to reduce LC activity locally, without toxic effects has still to be shown.

Conclusions
For the first time, in this study the frequency of BoNT-related side effects was estimated, which can probably be reduced by a locally acting LC antidote (ADOT-BoNT-ADEVs). Such ADEVs seem to occur quite frequently (in approximately 5% of BoNT-treated patients). This underlines the need for the development of such BoNT antidotes in clinical practice, in addition to the need that results from the necessity to treat botulism or a bioterrorist attack.
Therefore, animal experiments are recommended, in which BoNT/A, 3,4-DP, copper complexes, or other substances are locally applicated in different doses and with different delays, to prepare the application of these substances as an antidote in human subjects to reduce the risk of BoNT/A overdosage in clinical practice.

Limitations of the Study
The estimation of the frequency of the ADOT-BoNT-ADEVs was based on a small sample size (n = 197) and on a cohort of patients with a high disease complexity. Nevertheless, we believe that the estimation is realistic since the frequency of BoNT-ADEVs in the present cohort of BoNT-treated patients matches the frequencies of BoNT-ADEVs reported in the literature. Currently, the assumption that a substance can be developed that can effectively reduce the BoNT-LC activity locally is highly speculative.