A pragmatic approach to Botulinum Toxin safety

In April of 2009, the FDA announced the “black box warning”, and a Risk Evaluation and Mitigation Strategy highlighting the risk of use and spread of the Botulinum Toxin (BoNT) outward from the area of injection. The serious side effects highlighted by the “black box warning” for all neuromodulators such as Botox (abobotulinumtoxin A), Dysport (onabotulinumtoxin A), Xeomin (incobotulinumtoxin A), and Myobloc (rimabotulinumtoxin B) include ‘problems with swallowing, speaking, or breathing, due to weakening of associated muscles, can be severe and may result in death’ [1]. The use of drugs with “black box warnings” rightfully promote hesitation and concern among adult patients, but certainly these concerns are much more amplified for parents when deciding on the use for their children. Even before the “Black box warning,” the injection process itself and the use of a ‘toxin,’ for most parents and families heightens the anxiety for potential complications and risks in BoNT use for treatment of hypertonia.

As a clinician, risks and benefits of BoNT use should be thoroughly reviewed with the patients and parents who present for their consultations, as well as the options for the procedure and injection. Communication and rapport become essential during these encounters, particularly in relaying the associated risks and serious side effects in accessible terms. In addition, providing ample time and opportunity for parents to ask questions during the consultation and injection visits is necessary to review alternative treatment paradigms, and discuss the short- and long-term objectives of BoNT.

Over the past decade, clinicians have debated treatment paradigms to reduce risk to our patients. We find it essential to be open and honest about the use of BoNT in order to establish shared decision making on the most appropriate management of spasticity and hypertonia, and importantly to inform parents and patients how, as clinicians, we will mitigate adverse events and risks. A successful aspect of our practice has been to provide literature including current scientific research to family and patients on the use of BoNT. Usually this approach helps parents to understand our decision-making process. While infrequent, the media has sensationalized sentinel events, defined by the Joint Commission as a “patient safety event that results in death, permanent harm, or severe temporary harm” [2], that occur in children with cerebral palsy (CP) associated with BoNT injections which raises additional anxiety. The essential issue for physicians is how to assess the true risks of adverse events to each of our patients and to minimize the occurrence.

Some studies report adverse events as high as 56%, but these include events such as bleeding and bruising, which are expected to routinely occur with any form of injection, and certainly should not be a significant deterrent from treating a typical patient. Based on the review of the literature, significant systemic effects are estimated to occur 1–3.6% of the time [3, 4, 5, 6]. Severe sentinel events from BoNT have caused significant morbidity such as dysphagia, respiratory issues, generalized muscle weakness, systemic botulism or even death. O’Flaherty et al. noted that sentinel events occurred 2.3% of the time out of 596 injections [3]. Of those events, 13 out of 14 occurred in children that were Gross Motor Function Classification System GMFCS levels 4 or 5 [3]. A similar study published in 2018 showed a comparable rate of 2.4%; however, the published rate was 3.4% as they added “later nausea and vomiting” as a systemic adverse event. Swinney et al. also noticed an increased prevalence of sentinel events in patients of GMFCS 4 or 5 [7]. Of note, O’Flaherty et al. did a comparison of health problems (upper respiratory infections, lower respiratory infections, seizures and dysphagia) in children with CP the month prior and the month after BoNT injections. He noted that in the month prior to BoNT these children had an equal or greater number of events compared to the month after BoNT injections [3]. For patients with higher GMFCS levels 4 and 5, we thoroughly educate parents and patients regarding the signs and symptoms of local and systemic reactions. The relationship between BoNT and adverse events is complex, due to the concurrent comorbidities and medical conditions of the children with GMFCS levels of 4 and 5. Pre-existing medical comorbidities such as respiratory and orofacial weakness may contribute to a higher risk of an unintended event and may correlate to adverse outcomes with higher GMFCS classifications [4, 8]. It is our responsibility to be vigilant and aware of the adverse events and not to discount their occurrences as insignificant. Since the population of children that are under our care are vulnerable, the physician must diligently establish practice parameters and determine pragmatic ways to reduce these events.

Traditionally, in avoiding adverse outcomes, clinicians’ attention has been focused on BoNT: total number of units used, units per kilogram body weight (U/kg), units per muscle, size of muscle, injection site, extent of toxin spread within the muscle, and the clinical response to different dosages [9]. Over the years, clinicians have become more comfortable using BoNT and have begun to escalate the doses and substantially increase the amount of BoNT that is injected per patient. Prior to 2005, physicians were using 12–20 U/Kg of abobotulinumtoxin A. After a 2006 European consensus, some dosing protocols increased their ceiling to 25 U/Kg, with a maximum of 500 Units of abobotulinumtoxin A in the pediatric population [9]. However, Allergan has recently changed its pediatric Botox dosing to 10 U/Kg or 340 Units maximum [10]. A survey of 148 pediatric physiatrists in the United States published in 2018 revealed that 54% exceeded the previous United States pediatric standard of 400 Unit maximum dose and 86.5% utilized more than 10 U/kg in their patients [11]. Does this change in dosing parameters increase patient safety or limit the clinician’s ability to control hypertonia?

Although there appears to be a link between escalating BoNT dosages and adverse events among the adult population [12], there are complex factors beyond dosing that must be considered to ensure patient safety when all patient populations are taken as a whole. There is evidence in the literature supporting higher dose ceilings in both the pediatric and adult populations [12, 13, 14, 15, 16, 17, 18], but appropriate care needs to be taken in how the medication is administered to mitigate the risk of adverse events. In shaping our practice over the years, the patient’s clinical presentation and goals of injection, along with clinical experience, have contributed to our management plans to a greater degree than simply utilizing the calculated maximal dosage for my patients. Through a combination of clinical examination, observation, and assessment of function, a risk analysis is formulated and optimization of BoNT injection dosages can be accomplished while staying under the maximal dose. Specifically, our group has been able to achieve short and long-term objectives with regards to improving range of motion, hygiene management, pain reduction, and ambulation goals among our patients. In our opinion, the physician’s responsibility lies in their diligence to optimize dosage applications with close adherence to the available guidelines in order to reduce unwanted adverse events.

In addition to minimizing adverse events related to underlying medical comorbidities and dosage related complications, the setting in which BoNT injections are performed should be considered to further ensure patient safety. An example in our practice is that the vast majority of our injections occur in the office setting. Only 4% of patients who received BoNT injections in an office setting report significant post-injection pain [19]. Exceptions are made for patients who are concurrently receiving phenol, having other sedated procedures simultaneously, and those that express feelings of emotional distress or have behavioral issues that would make office injections unsafe. Intranasal midazolam is offered to patients who have expressed anxiety in order to avoid unnecessary anesthetics. Pre-existing medical comorbidities such as respiratory and orofacial weakness may contribute to a higher risk of having an adverse event [4, 8], and medications such as anesthetics and benzodiazepines that are used for sedation during procedures may increase these risks [8, 20]. However, intranasal midazolam has a rapid onset as well as a quicker recovery time at a lower dosage in comparison to oral midazolam, and may provide an alternative to more invasive interventions [21]. While it may be customary for some practices to sedate a significant percentage of their patients for injections, the families should be made aware of risks and alternatives to sedation [11]. In the literature above, the majority of studies use a form of sedations, anesthesia, oral midazolam, and/or nitrous oxide [3, 4, 6, 7]. As a result, clinicians should carefully consider the necessity of anesthetics during administration of BoNT, especially in patients with medical comorbidities, and provide an office-based option if possible to further minimize procedure associated adverse events.

In our practice, there is a focus on improving the localization and technique of injecting BoNT to improve accuracy and reduce spread. BoNT works by inhibiting release of acetylcholine at the motor end plate (MEP). The closer the BoNT is injected to a MEP, the stronger the paralytic effect. This requires that the physician be knowledgeable of the orientation of the MEPs. In studies comparing injections near the MEPs to injections distal from the MEP, not only was the effect of the medication improved when injected closer to the MEP, but less volume and/or lower doses were able to be utilized [22, 23, 24, 25]. However, in 2018 only 17.1% of pediatric physiatrists targeted motor points when performing BoNT injections [11]. Since the introduction of ultrasound (US) to our clinical practice, we use a combination of published MEP location data and the EUROMUSCULUS/USPMR spasticity approach when localizing muscles for injection [26, 27, 28]. Electrical stimulation is our secondary localization technique of choice to confirm MEP location, especially in muscles that have atypical architecture.

Equally important to localization is the awareness of the quality of muscle that is being treated, which can be assessed with localization techniques such as US. With the introduction and expansion of US in medicine, evaluation of muscle “quality” may help with improvement of BoNT injections. BoNT is most effective in non-contracted and nonfibrotic muscle, as the presence of increased collagen and sarcopenia content inherently reduces the biological efficacy of BoNT [29]. It is likely that patients who have had significant hypertonia for a prolonged period of time and patients with higher GMFCS classifications may have more fibrotic and contracted muscles, superimposed upon intrinsic collagen and extracellular matrix differences [30]. Children with CP also have an increase in muscle lipid content [31, 32]. A meta-analysis of children under the age of 2 years who received BoNT for musculoskeletal problems such as clubfoot and brachial plexus injuries revealed that there were no BoNT related adverse events [33]. A review of the dosing in these patients showed that they received proportionally higher dosing than what is typically injected into children with spasticity. It is possible that no adverse events occurred because these children have “normal” muscle in comparison to children who have had chronic muscle changes from a prolonged history of spasticity. Booth et al. found a correlation between severity of CP and degree of collagen accumulation and fibrosis within muscles [34]. As a result, patients with CP may not be as receptive to BoNT and have an increase in risk of adverse events from medication diffusing from the site of injection. However, “normal” muscle does not necessarily preclude adverse events, and while rare, the media have highlighted sentinel events associated with cosmetic BoNT injections such as neck weakness, dysphagia, hoarseness, dry eye syndrome, strabismus, diplopia, gangrene, fasciitis, superficial temporal artery pseudoaneurysm, and death [35]. Yet a literature search of the occurrence of adverse events such as an instance of facial anaphylaxis following the use of BoNT in the cosmetic treatment of masseter hypertrophy indicated that they continue to be far and in between [36].

Ultrasound is being studied to define local pattern texture anisotropy, and may be a potential tool for assessing muscle quality differences among Duchenne’s Muscular Dystrophy (DMD), Inclusion Body Myositis (IBM), and Normal Healthy Controls [37]. Shear wave elastography has also been studied to assess muscle stiffness in patients with CP [38, 39]. If there are additional concerns regarding the quality of the muscle, a secondary localizing technique, such as Electromyography or electrical stimulation, could be used in order to assess the quality of muscle and to confirm the MEP location. Drawing from our institution’s clinical experience, if we have encountered the estimated MEP location to be fibrotic, we identify an alternative location based on muscle quality and response to stimulation. In our practice, we often comment on the quality of the muscle in the post procedure note. This has been helpful in evaluating the results of the procedure and also in determining if there are other merits, such as pain relief, for continuing the procedures. Furthermore, discussing muscle morphology and expected transformation of the muscle with the patient’s family has helped guide expectations and open discussions towards other treatment paradigms.

The fundamental goal of spasticity reduction in a physiatrist practice is to promote function, improve care and comfort, and reduce complications in our patients. Research has demonstrated that after use of BoNT there have been significant improvement in Gross Motor Function Measure (GMFM-88) among both non-ambulatory and ambulatory children with CP [40]. Adolescents who underwent repeated BoNT injections demonstrated improvement or maintenance of GMFCS level [41]. BoNT is a safe and effective treatment for patients with hypertonicity when careful consideration is taken in assessing the patient’s goals, history of injections, options, quality of technique and awareness of risk factors associated with the medication. In addition to being cognizant of body weight dosing, the physician needs to consider other factors that may influence outcomes, such as need for anaesthesia, medical comorbidities, targeting techniques and muscle morphology. While our comprehension of the factors that can lead to systemic adverse events is limited, it is the responsibility of the physician to evaluate the data and make proactive decisions that decrease risk while maximizing the effectiveness of BoNT. This highlights the need for more efficient use of BoNT, improved assessment tools, and strategies in the management of hypertonia.