Abstract

Introduction
For this year’s Needlepoints we asked leaders in the care of children with cerebral palsy (CP) from around the globe how they would response to the same clinical scenario. We created a clinical presentation, and we asked each author to formulate their care plan without knowing what every other group was considering.
The clinical presentation is as follows:
3-year-old female with spastic diparetic CP secondary to prematurity-related periventricular leukomalacia (PVL). She has been working with a physical therapist (PT) for the past 2 years. She presents to your clinic as she has plateaued with her function. She has a scissoring, jump gait presentation. The PT is requesting botulinum toxin (BoNT) injections and emailed you ahead of time. When would you incorporate further interventions (e.g., oral medications, injectable medications, etc.)? What additional treatments would you offer and why (e.g., serial casting, phenol, ethyl alcohol, other modalities)? What are the goals of your treatment?
Pertinent lower extremity exam: She was noted to have increased muscle tone with spasticity on the Hypertonia Assessment Tool without noted dystonia in her legs. Modified Ashworth Scale (MAS) of 1 + in her hip adductors, knee flexors and hip flexors and MAS 2 in her ankle plantar flexors. Range of Motion (ROM) Tardieu hip abduction R1 30 R2 60, hamstrings R1 -60 R2-20 ankle knee extended R1 -10 R2 0, strength 4/5 except ankle dorsiflexor 2/5, Selective Control Assessment of the Lower Extremity –left-side score of 7 and right-side score of 8.
Gait pattern: Mild scissoring jump gait presentation, bilateral in-toeing, mild bilateral Trendelenburg gait appreciated Please set treatment goal(s) based on the therapist’s request, parents’ concern, and your findings. Please discuss your treatment plan and reasoning, including: Chemoneurolysis plan to achieve above goals –selection of muscles, dosing, injection agent(s) (BoNT and/or phenol, guided injections, sedation etc.). Pharmacological intervention Casting/bracing Other. When would you repeat the hip x-rays? Please write post-injection recommendations if you recommended chemoneurolysis therapy (therapy modalities, ankle brace, hip brace, castings, oral medications, sports activities, etc.). What kind of outcome measurement tools would you use to evaluate the efficacy of your treatment?
Four approaches are presented, from comprehensive discussions to focused and personalized plans. The process from all four emphasizes the way Pediatric Rehabilitation Medicine organizes and conceptualizes outcomes and treatment for a common clinical issue. We do not endorse any one approach but offer you the opportunity and encourage you to compare and contrast your management with the experts!
Australian authors: Drs. Gauden and Scheinberg
Big picture considerations
Clinical care and research in CP is an exciting and ever-evolving field involving an array of medical subspecialities and professionals not limited to the healthcare domain. Understandably, therefore, clinical approaches and opinions are varied. The complex interplay of the patient’s health condition, goals and personal and environmental factors, as described by the International Classification of Functioning, Disability and Health (ICF), outlines the dynamic, multidimensional and unique nature of patient care [1]. The role of the rehabilitation physician is therefore an essential conduit between professionals and the patient and their family/caregivers, providing diagnostic education, outlining clinical expectations and identifying short and long-term challenges.
Research suggests that clinical care approaches which incorporate the ICF or six F-words (function, family, fitness, fun, friends and future) as described by Rosenbaum, provide a more holistic care model and are more likely to promote discussions with the patient, their family and caregivers on care challenges and individual goals [2]. However, clinical goals may differ between the clinician and patient and/or family. Patient and family goals primarily center around participation, whereas the goals of the clinician tend to focus on impairment, activity and prevention.
Intervention decision making
Clinical reasoning on available interventions should be assessed on a case-by-case basis taking into account factors that may impact the final decision. Limiting factors that may alter the decision-making process include restrictions in clinical resources, reduced access, and availability of optimal interventions and limitations in clinical expertise. For example, what would be the appropriate intervention in Low and/or Middle-Income Countries where intramuscular BoNT is not an option or in a center that offers selective dorsal rhizotomy (SDR) but not intrathecal baclofen?
Open parental discussions on the clinical decision-making process should include exploration of treatment indications, associated intervention risks and benefits, availability, optimal timing and short- and long-term outcomes. A multidisciplinary care approach is essential for clinical assessment, planning and monitoring and understanding the patient’s likely clinical trajectory. Most clinical interventions in CP can also be considered throughout the child’s life with the support of early and collaborative subspeciality team involvement.
In the clinical case presented, medical and surgical treatment options could include oral and injected pharmacological treatments (oral baclofen, intramuscular BoNT type A [BoNT-A] injections), orthopedic intervention and SDR. Additional neurosurgical interventions for spasticity may also include intrathecal baclofen delivered via a programmable pump, which may be considered in patients with severe, generalized spasticity.
Choosing the intervention
The child described in the clinical case is reported to have good lower limb strength with weakness noted at the level of the pelvis and ankle. She walks with a jump gait pattern. Other pertinent impairments noted in the clinical case presented include a large dynamic range in hip abduction, early contractures in the ankles with a small dynamic range, evidence of knee and hamstring straining and lowered selective motor control distally.
The clinical goals described include improving ambulation, base of support and endurance to participate in school activities. We would support these goals, aiming to promote her developmental progression including learning to run, climb, and explore, and to support emerging playground and tricycle skills. Further defining ‘school activities’ may help contextualize the goals for the child and family.
The primary clinical concern in this case is that of an early mild ankle contracture. It is unclear whether this will accelerate and contribute to contracture at the knees and development of patella alta and crouch gait. Additionally, this patient does not have the plantarflexion knee extension coupling at the level of the knee suggesting that there is a role in this early stage for intervention at the level of the knees with intramuscular BoNT-A in the hamstrings and/or gastrocnemii.
Proposed intervention: Intramuscular BoNT-A injections
Despite recent debate on the safety and efficacy of intramuscular BoNT-A injections in children with CP, BoNT-A remains an accepted treatment modality within this population [3, 4]. Cases should be assessed on an individual basis with a multidisciplinary review prior to the procedure. Informed consent from parents or caregivers should always be obtained.
Following clinical assessment, technical procedural aspects including muscle selection, dosage, dilution and sedation techniques should be documented.
We propose here two clinical approaches that could be considered. The first approach involves intramuscular BoNT-A injections targeting both the hamstrings and gastrocnemii. The alternative approach involves targeting only the hamstring muscles and monitoring the outcome at follow-up appointments. Our clinical decision would be to inject both the hamstring and gastrocnemii but remain cautious with the dose delivered to gastrocnemii.
The BoNT-A preparation principally used in our clinical department is onabotulinumtoxin A (Botox®). In this clinical case we would opt to deliver a conservative total dosage of 12 units/kg. This would include 4 units/kg to the hamstrings bilaterally and 2 units/kg to the gastrocnemii bilaterally. The dilution used is 100 units/2 ml. Clinical practice across Australia with regards to dosing and dilution has been shown to be largely consistent, with dilution of 1–2.2 ml normal saline for 100 units Botox® and dose range of 12–20 units/kg (maximum dosage 400 units) for pediatric patients [4]. Botox® is registered by the Australian Therapeutic Goods Administration for use in the treatment of focal spasticity in children with CP and subsidized by the Australian government through the Pharmaceutical Benefits Scheme.
Procedural preparation in our center follows a streamlined process involving the proceduralist, nursing staff, procedural assistant and child life therapists who provide procedural support and distraction [5]. The process begins with a clinical review with patient and family to discuss the child’s medical history, procedural risks and benefits, procedural process and follow-up plans. Informed consent is obtained from the parent or caregiver, and an information handout is provided to the family. The muscle injecting sites are marked and local anesthetic cream is applied to the skin. Traditionally, muscle localization was guided by anatomical landmarks, palpation and electrical stimulation. Current practices now include ultrasound guidance which is the preferred practice in our center. This allows the proceduralist to have direct and continuous visualization of the target, surrounding structures and the procedural needle [6].
Successful procedural sedation requires clear communication, planning and clinical support. Procedural management guidelines include pharmacological and non-pharmacological strategies. Pharmacological options frequently used include inhaled nitrous oxide (NO) and oral medications, primarily midazolam. Other non-pharmacological and psychological strategies commonly performed include distraction, relaxation or deep breathing techniques, music, guided imagery and virtual reality technology. General anesthetic is available to those children not able to tolerate conscious sedation, or when other interventions such as phenolization of obturator nerves are required.
Post BoNT-A recommendations in this clinical scenario would include follow-up with the medical and physiotherapy teams for consideration of serial casting if the patient has up to 10 degrees of plantarflexion range with the knee extended. Orthotics prescription will depend on the end ROM achieved following serial casting. If the child is able to achieve 5 degrees of ankle dorsiflexion, a hinged ankle foot orthosis (AFO) could be considered over a solid AFO. A clear physical therapy plan focusing on goals to improve participation and play would also be beneficial. A physical therapy program unique to this clinical case would incorporate a stretching program including long sitting, climbing activities, high kneeling, sit-to-stand activities and calf raises out of orthotics.
Decisions on future BoNT-A injections will be dependent on clinical response, assessment findings and patient/family goals. Repeat injections could be considered after at least six months from the initial procedure. Additionally, there may be clinical benefit in using BoNT-A during the most rapid period of motor development. This is supported by the understanding that the natural history of spasticity in children with CP peaks at 4–5 years and then slowly decreases [7].
Measurement tools –feasibility vs. reliability vs. utility
Clinical assessment tools used in this patient cohort must fulfill the outcome measure characteristics of utility, feasibility, and reliability. However, it may not be realistic for a detailed assessment to be completed in some settings.
In terms of feasible assessment tools, we would recommend that the patient be assessed using the MAS and/or the Modified Tardieu Scale for muscle tone, the Medical Research Council Scale for muscle strength and the Selective Control Assessment for the lower extremity. A video of the patient’s gait would also be beneficial for further discussions.
The Gross Motor Function Measure (GMFM) could also be completed in full or specific domains that align with the patient’s goals such as standing or walking. Additional formal assessments may include a 3D gait analysis, 6-minute walk test, Edinburgh Visual Gait Score or single leg stance test. In centers that do not have access to 3D gait analysis, these additional assessments are easy to perform, reliable and valid.
To assess patient and parental outcomes we would perform the Goal Attainment Scale and Canadian Occupational Performance Measure (COPM). Finally, from the perspective of the clinician, the Physician Rating Scale could be performed.
Hip surveillance
The child described in this case had a recent pelvic x-ray performed at 2.5 years for surveillance of hip displacement. The findings showed a femoral migration index of 15% on left and 10% on right. Our center follows the national Australian Hip Surveillance Guidelines for Children with Cerebral Palsy recently updated in 2020 [8–10]. The recommendation for this child would be to repeat a pelvic x-ray at 5 years of age or earlier if there were any concerns noted on clinical assessment during medical reviews.
This case highlights the importance of collaborative multidisciplinary care in the care and support of young patients and their families across their lifetime.
German authors: Drs. Schroeder, Heinen, and Berweck
Summary & reasoning:
Etiology: we read the report of an independently ambulant female young child with bilateral, possibly left dominant, spastic CP (BSCP) secondary to prematurity with a dynamic jump gait pattern, reduced range of ambulation in her community and no history of musculoskeletal pain. Etiology of BSCP would most likely be due to perinatal ischemia during premature birth or due to hypoxia during intrauterine fetal stress, but this is not further specified. Assuming this etiology, cranial MRI examination would most likely show periventricular white matter lesions and could support the etiology of spastic movement disorder in the specific patient. Radiological hip surveillance shows signs of (secondary?) dysplasia not yet “at risk,” which would complete the typical phenotype of a child with BSCP Gross Motor Function Classification System (GMFCS) Level II frequently presenting in CP clinics across the globe [11]. Interpretation of phenomenology: clinical examination points to velocity-dependent tone increase in both lower limbs; no assessment of the upper limb is provided, but typically, there should not be “much spasticity”. During gait, internal rotation of the lower limbs and mild scissoring is stated, which typically can be explained predominantly by weakness of the gluteal muscles, femoral anteversion as a consequence of muscle tone imbalance, and reduced weigh- bearing and reduced verticalization during the previous 2 years. Boney hip internal and external rotation assessment is not further addressed in the clinical assessment; however hip x-ray shows mild subluxation not being rated “at risk” referring to the German “hip traffic light” (green light). For the distal lower limb, we can assume that jump gait pattern implies retrograde heel contact during stance phase probably with significant instability in the lower ankle joint (valgus or varus deformity), which is combined with increased knee (and probably hip) flexion as a sum of slightly reduced strength (4/5). Hyper-resistance and muscle stiffness, which could also be caused by the non-neurological secondary components of impaired muscle function in children with CP will also add to the clinical picture [12]. Toe walking during terminal stance phase would be a result of increased plantar flexion activity of the calf muscles. This could be (i) due to primary spasticity in the gastrocnemius muscle, (ii) secondary due to mentioned non-neurological components of impaired muscle function and (iii) tertiary voluntary activation of the plantar flexor muscles. This could overcome weakness in the contralateral ankle dorsiflexion muscles (strength 2/5) with the functional goal of clearing swing phase and avoiding stumbling. The complex mechanisms of adaptation on the musculoskeletal level in children with CP have been discussed for almost 30 years [13] and are being further elaborated in recent literature [14–16]. Pelvic ventral tilt is not mentioned in the description of the patient but can frequently be seen in children with jump gait pattern, especially at later ages [17]. Pelvic tilt can be due to primary weakness of abdominal wall muscles, gluteal muscles and hamstring muscles, secondary due to spasticity of the hip flexor muscles (Tardieu maneuver for hip flexor muscles is not mentioned in the clinical examination), or tertiary due to compensatory vertical alignment of knee flexion and toe walking to maintain balance. Typically, children present a combination of all the above-mentioned issues. Rotation of the forefoot could also be attributed to all levels of the lower limb starting from hip flexion and femoral anteversion, knee flexion and over activity of bi-articular muscles (such as the medial hamstrings and the gracilis muscle). Supination or pronation can aggravate the distal lower limb due to hypertonicity in the muscles moving the forefoot (e.g., the tibialis posterior muscle and/or the peroneus longus muscle, flexor hallucis longus muscle, and m. abductor hallucis). This typically would be combined with a reduction of strength in the antagonistic muscles (e.g., tibialis anterior muscles, extensor muscles of the toes and peroneus brevis muscle, as is stated in strength assessment with 2/5). Additional secondary bony deformities or connective tissue adaptations are rare in the mentioned age group of children with BSCP but belong to the non-neurological factors impairing muscle function. Orthotics: The patients wears an AFO; however it is not stated whether the patient tolerates them in daily activities. Using a static AFO aims to correct equinus deformity, resulting in a normal heel strike and a more normal sequence of ankle [12] probably with regular shoes? In our clinical experience, orthotic management can be very challenging in the mentioned age group as children have difficulties in changing their positioning during play on the ground, getting up and walking inside the room and then playing on the floor again. Some dynamic AFOs try to compensate for this; however they need to be light, correct/stabilize mid-foot deformity and easy to apply. No further information is presented with regard to additional aids or splints, e.g., night splinting for contracture prevention. Participation: Reading the presented information for ICF-Children and Youth version (ICF-CY) [18] domains “structure and function” we could summarize: (i) Significant velocity-dependent increase (spastic) in muscle tone in both lower limbs, (ii) significant weakness in the muscles of both lower limbs, and (iii) bilateral mild musculoskeletal contractures in proximal and more prominent in both distal lower limb joints. From the “activity and participation” domain, reduced endurance in the community setting can be stated. Stumbling and falling does not seem to be an issue during activities of daily living. There is no additional information about personal and/or environmental factors contributing to the actual situation of the child and the family. Trying to read the information with a focus on the six F-Words [2, 19], the following concerns can be addressed: (i) functioning and some limitations affecting (ii) family, (iii) fitness, and (iv) fun. From the developmental perspective, concern with (v) friends is not clear, but it will become more relevant when the child will be engaged in pre-school and peer group activities. Therefore, (vi) future is the major issue in this child at the moment, which is addressed by the therapist and the parents in the presented case.
Treatment plan & reasoning
Clinical decision-making and therapy would be based on: Establishing continuous, outpatient, multidisciplinary care in a social pediatric center (SPZ) embedded in the German health care system [20]. As of 2021, there are more than 160 SPZs in Germany and their number continues to grow. These institutions are unique with respect to the combination of social and pediatric services allowing biopsychosocial care for children with medical complexity and their families (or caretakers). The principal goal is to foster (neuro) development in the community using the guiding framework of the ICF-CY [18]. Although there is still a long way until knowledge transfer on implementing the framework into clinical practice is finished, first attempts focus on establishing this as standard of care in the German health care system [21]. Applying valid assessment tools in order to specify patient-centered goals. With a focus on ambulation, for example, the Gait Outcomes Assessment List (GOAL) questionnaire offers semi-structured communication [22, 23] and could be enhanced by the Both Hands Assessment test for measuring bimanual performance, including quantification of possible asymmetry of hand use [24]. Following this, the family therapeutic goals could be transferred into a Goal Attainment Scale, which helps to rate the responsiveness to interventions that will be mentioned later. Helpful in clinical routine, the COPM can be used and adapted to the above-elaborated goals [25]. Expanding the clinical examination with a focus on hip flexor muscles to disentangle the dynamic contribution of M. iliopsoas, M. tensor fasciae latae, and M. rectus femoris using the Tardieu maneuver and/or the presence of stiff knee from M. quadriceps spasticity (e.g., positive Duncan Ely sign). In prone position, bony hip internal rotation could be assessed by rotating the flexed lower limbs to look for anteversion/internal rotation at the hip joint level beyond muscle activity imbalance. In case of significant bony anteversion of the femur, internal rotation will be difficult to be improved using tone modulation (e.g., with BoNT injections). Only continuous strengthening of the external rotating muscles at the hip level (mainly gluteus maximus and medius muscles) would stimulate musculoskeletal development towards hip maturation. Documenting a structured 2D video gait analysis in sagittal plane with a focus on terminal stance phase: plantarflexion, knee (hyper-)extension activity, contralateral foot clearance during swing, and reduced knee extension during terminal swing phase. Video gait documentation in the frontal plane should focus on scissoring and rotation of the legs during activity, e.g., using patella orientation during stance and swing phase versus forefoot positioning. Gait scales such as the Observational or the Edinburgh Visual Gait scale may help to structure the video assessment [26, 27] but require extra time for classification. 3D gait analysis is not available on a regular basis in Germany, as access to is limited, and the child is very young to tolerate the assessment and the added value compared to 2D gait observation at that age may even be questioned. However, it would become more meaningful in the course of development in this ambulatory child. At this age, using practicable classification tools may be more quickly applied in non-academic settings, as published by Graham et al. in 2021 [14]. The combination of [3] and [4] may help to improve communication about gait and musculoskeletal pathology (MSP) at present and future development with the family. Applying the referenced classification system of MSP for this patient would state a “stage 1” with beginning of “stage 2” as the ankle joint shows significantly reduced dorsiflexion at 0° and a dynamic pes equinus with a catch at –10° plantarflexion representing “not much” dynamic tone increase during passive clinical examination. Discussing splinting. Knowing the controversies about pes equinus contracture prevention in the available literature, we would supply the patient with custom-made dynamic night splints with circular stabilization of the lower ankle joint, the mid- and forefoot (to correct varus or valgus foot deformity) in order to stretch the triceps surae during night sleep. We know that this procedure is often limited in acceptance, but if the child is not impaired during night sleep, this may postpone the development of musculoskeletal deformity. In addition to static AFOs, we would supply the child with custom-made insoles to support the lower ankle joint and forefoot deformity within a regular shoe. This allows the child to better actively move her ankle during ambulation in her community and therewith strengthens the calf- and foot-lifting muscles, as these are usually weak and need strengthening without continuous immobilization. Managing contracture. Therapeutic casting of the ankle and foot could be performed for 7–10 days in order to improve ROM of 0° in both ankle joints towards 5–10° dorsiflexion. However, this should not be performed to a much greater extent in order to avoid focal muscle inactivity and secondary development of muscle atrophy. However, the indication for this remains weak and individually decided, as evidence for this procedure remains controversial [28]. Injecting with BoNT. In clinical examination, velocity dependent increase in muscle tone is clearly pointing towards responsiveness to injections with BoNT. The following muscles “could be considered” from the mentioned clinical examination: hamstrings and calf muscles on both sides. However, there are points that need to be discussed: The time course of BoNT injection with temporarily limited effect. Short-term functional improvements (in reducing spasticity) could be beneficial for the patient if they transfer into activity and participation levels. The temporary effect could also be beneficial as neurodevelopment is still dynamic in a child with BSCP at the age of 3 years and ideally allows more ambulation in a more physiological ambulation pattern, which could positively affect neuroplasticity, and therewith, motor learning/development at early ages. However, as to our knowledge, this hypothesis has never been scientifically proven for children with CP –probably also due to the limitation of our assessment methodologies. On the opposite side, there is growing evidence that BoNT injections induce longer lasting neurogenic atrophy as there is apparent clinical effect [29] and repeated injections may support development of premature sarcopenia [30] in muscles of children with CP. The controversy of growth and muscle development versus premature degradation in spastic muscle remains a matter of debate in literature [30]. However, these debates should not only focus on BoNT but should be addressed to any motor intervention for children with CP. Therefore, standardized (semi-structured) assessments are mandatory to clarify the benefit of short-term benefits for the individual patient and may be helpful to estimate the possibility of continuous elimination of spasticity using neurosurgical SDR (see below). Communicating this to the parents/caretakers before the injection is necessary for any child being treated with BoNT. Muscle selection would depend on more detailed clinical information, e.g., examination of M. psoas (positive Tardieu maneuver, patella alta), M. tensor fasciae latae (if it contributes to hip flexion and internal rotation activity), M. gracilis, M. semimembranosus et -tendinosus. If spasticity is present –as could be expected –we would focus with our injection on the upper part of both lower limbs. We would inject the gastrocnemius muscle with a low dose (if at all) and possibly the tibialis posterior muscle, if internal rotation of the foot (in spite of pelvic tilt and internal rotation at the hip) is relevant for the patient’s gait pattern. Dosage. Safety, tolerability, and efficacy have been shown in a broad range of randomized controlled trials (RCTs) for different preparations [31–33], but off-label use still represents a major issue when treating children with BoNT. However this is not a limitation for treatment in Germany. Dosing guidelines as published more than a decade ago remain valid and can be used [34–36]. However, while in the past we were aiming for the highest possible therapeutic dosage in order to have a maximum effect for the longest duration [34, 37], recent RCTs have shown significant effectiveness in the low dosage treatment groups [32, 33]. Considering the new evidence about muscle development and BoNT, our clinical practice therefore is now focusing on “how low can you go” with dosing in order to modulate spasticity on multiple musculoskeletal levels. This generally means more muscles injected with lower dose per muscle with more volume/dilution using ultrasound-guided injection with a patient receiving effective analog sedation during injection; this advocates the best standard of care for our patients. Discussing SDR if the patient reports relevant benefits from BoNT injection at follow-up. A multidisciplinary team including social- and neuro-pediatricians, neuro-orthopedics, neurorehabilitation specialists and neurosurgeons should discuss timing for surgery, pre- and aftercare. If the decision is pro surgery, standardized protocols should be applied when patients enter an SDR program. Very specific protocols have been set up in countries outside of Germany with well-structured health care systems [38] which will allow us to enhance our knowledge about which patient will benefit the most at what age from this irreversible procedure. In any case, a minimally-invasive SDR procedure should be applied to avoid the development of scoliosis [39, 40]. Intense, post-surgery, task-specific interventions focusing on strength and gait endurance need to be communicated to the parents prior to surgery. Prior to and post-surgery, clinical assessments for our clinical care setting include 3D gait analysis, GMFM-66, GOAL questionnaire, and the Goal Attainment Scale. Intensive, task-specific rehabilitation blocks including robotic-enhanced therapies [41–43] or serious gaming technologies using inpatient, outpatient and/or home-based protocols should be based on the capacity of the family. The interventions should be accompanied by the above-mentioned assessments at 6 and 24 months after surgery and from there on as necessary. Functional therapies. In any of the above-mentioned aspects, neurodevelopmental support should be maintained by a community-based physiotherapist. Intermittent intensive therapy blocks with task-specific training sessions for a limited period of time (either on an outpatient basis or inpatient during rehabilitation) are the most relevant aspect in achieving the therapeutic goals of the parents. However, the interventions should be participation-oriented and consider the individual resources of the family within the daily activities as these are significantly affected by personal and environmental factors. In Germany and for this specific case, fostering motor development with the goal to enhance motor capacity in a three-year old child with BSCP GMFCS II remains the main focus for any intervention before school and contact with peers and the community become the primary focus of life from five to six years of age onward (“F-words”). Muscle lengthening procedures, which might be necessary as the child grows older. Therefore, regular interdisciplinary consultations are recommended.
Conclusion
A simple question raises a complex answer. Luckily, the toolbox for physicians caring for children with complex medical needs (such as children with CP) continues to grow and evidence will continue to improve beneficial decision-making. In the presented case, the patient seems to be “doing quite well,” so why should we struggle too much with time and resource-demanding neurodevelopmental support? Not using the broad spectrum of therapeutic options in a three-year-old child with BSCP GMFCS Level II seems unethical. Evidence-based recommendations on what to do at what age have emerged [44, 45] and show that, when carefully selected, they are worth the effort. However, any intervention we propose to the child and its family needs to be critically appraised with respect to the specific impact on musculoskeletal development, improvement of meaningful activities, as well as the individual and environmental aspects of each child and family. Therefore, the simple question stated in the beginning may be thoroughly answered differently for each patient considering development (age) and capacity of the family and the health care system the patient is living in. In this context the injection of BoNT in ambulant children with CP has changed towards a carefully chosen tool within a toolbox of many different options and surely is not a long-term option.
Korean author: Dr. Hong
Treatment goal
✓ Improve the gait pattern by reducing spasticity with botulinum toxin (BoNT) injection in the lower extremities and intensive physical therapy or exercise. by increasing the ROM of the ankle with BoNT injection and serial casting. After chemoneurolysis, gait patterns (scissoring, in-toeing, jump knee gait) could be improved with reduced muscle tone. In addition, the base of support could be increased with relieved calf muscle tightness, which could also improve balance.
Treatment plan and reasons
A. Chemoneurolysis plan to achieve above goals
✓ Where and which materials to use
The child shows mild scissoring and jump knee gait, which suggests spasticity of the hip adductors, hamstrings, and hip flexors in addition to calf muscles. Target muscles for focal spasticity control with BoNT might be bilateral hip adductor, iliopsoas, medial hamstring, and gastrocnemius muscles. As the difference of R1 and R2 (R2-R1) are 30 to 40 degrees in hip abduction and knee flexion, we could recognize the dynamic component of stiffness in these parts. Therefore, hamstring and hip adductors may be effectively treated with chemoneurolysis. Final muscles and dose for each muscle might be medial hamstring 30/30 IU, gastrocnemius (both medial and lateral heads) 35/25 IU, and gracilis 15/15 IU with a dilution of 100 U/2 ml (onabotulinumtoxinA). As phenol or alcohol is not available in Korea nowadays, there’s no option but BoNT. The child’s physical examination revealed a weaker right ankle dorsiflexor compared to the left side. In case of weakened motor power, it is more difficult to overcome the calf spasticity, so a higher dose was administered to the right side of the calf, unlike other areas. Gracilis muscle is attached to the tibia, and it works as a knee flexor, hip adductor, and tibial medial rotator on the femur. The girl showed jump knee gait and in-toeing, so gracilis muscle was chosen rather than adductor longus. The patient is only 3 years old and, if it is the first BoNT injection, its efficacy might be better than in older children or children who have had multiple previous injections [46, 47]. Therefore, it would be better to adjust the dose about 10 U/kg in total without overdoing it, taking into consideration the potential for weakness and general side effects such as iatrogenic botulism. In Korea, BoNT injections in the lower extremities in children with spastic CP are covered by national health insurance. And the Health Insurance Review and Assessment Service reviews the legitimacy. Although the policy continues to change, the current (January 2022) standard stipulates that the dose is less than 300 IU at intervals of 3 months or more from 2 years old or older. Although 300 IU is the maximum dose, it is administered in relation to bodyweight for small children.
✓ How to do a chemoneurolysis
As one mL of onabotulinumtoxinA spreads about 3 cm from injection site when 100 units of onabotulinumtoxinA is diluted in 1 mL of preservative-free normal saline [48], distributing the doses to two or more points of the innervation zone in one muscle might be helpful [49–51]. In the case of the forearm, anesthesia is sometimes recommended to maximize the desired effect and minimize unwanted weakness by increasing the accuracy of small muscles. However, with the legs, whether anesthesia is administered or not is determined by the child and their caregivers. Usually, children between the ages of 2 to 5 are terrified of procedures and are very resistant. Therefore, in this age group, anesthesia is done with the help of an anesthesiologist if the parents want. It is easy to use electrical stimulation and ultrasound guidance with anesthesia. There are several modalities for analgesia and sedation during the procedure [52]. Eutectic Mixture of Local Anesthetics (EMLA®) cream had been used in the past. However, as only skin anesthesia is available, it does not help much relieve pain, and children’s fear persists, so it no longer commonly used. Electrical stimulation and ultrasound are prepared for localization of each muscle in addition to anatomic guidance, sedated or not. However, without sedation, two methods may or may not be used depending on the degree of cooperation of the child. It is safe and most helpful to finish the procedure as soon as possible if the child is very irritable and has severe fear. Injection under electrical stimulation and ultrasound guidance is preferred because it is more accurate and reassuring to the operator than without guidance. In addition, when the effect of injection does not meet expectations, the possibility of injection into undesired muscles could be excluded. As a result, the operator can adjust the dose or selection of muscles when performing the next injection.
B. Pharmacological intervention
Many children participating in social life, such as in kindergarten or schools, do not like drugs that cause drowsiness. Furthermore, spasticity with MAS 1+ 2 in large muscles without other complications is usually not too much of a problem after BoNT injection. Therefore, pharmacological therapy in this child is not primarily considered, and muscle tone will be checked after injection.
C. Casting/bracing
The child’s ankle dorsiflexion ROM is 0 degrees, which could be regarded as decreased ROM and early/moderate contracture. In a recent systematic review, serial casting was applied to effectively reduce contractures in the short term when a contracture has begun to develop [45]. Combining BoNT injection and casting is effective, and there is sufficient evidence to make a recommendation [45, 53]. For the first casting, a splint is applied for 3 or 4 days to check tolerance and skin problems. Although some evidence suggests that changing the casts at 3-day intervals is more effective than a week interval [45], it is not easy to come to the hospital often, so, in many cases, it is done after the first 3–4 days and then at intervals of 1 week for a 3-week period [54]. With improved ROM after casting, wearing an AFO might be more comfortable with relieving spasticity. Also, the application of AFOs and improvements in weight-bearing could prolong the effects of chemodenervation [55]. However, the effect of AFOs on various gait parameters is controversial, and restriction of motion may lead to inhibited muscle use [56]. Therefore, the duration of wear of orthoses might be adjusted while the BoNT is effective if there is no concern about a breakdown of the midfoot [56]. Also, additional braces such as hip abduction braces or twister cables might not be recommended as their effectiveness has not yet been sufficiently proven [45]. It is questionable whether the benefits are large compared to the cost and effort required to wear the orthosis.
D. Others
Various physical activities and conventional physical therapy is recommended for children with spastic CP after BoNT injection [57]. These can strengthen the weakened muscles and help to improve the pattern of movement. BoNT indeed relieves stiffness; however, it is not easy to achieve the desired goal of “improvement of gait pattern” with BoNT alone [58, 59]. After reducing muscle tone, the ability to control isolated muscles can be improved, and combining physical therapy or comprehensive rehabilitation at this time is helpful in making a synergistic effect [60]. Regular physical therapy is also helpful in maintaining the efficacy of BoNT and ROM [61, 62] in children with spasticity. Several reports show that children with CP do not achieve healthy physical activity levels compared to typically developing children [63, 64]. However, we all know that physical activity and exercise are essential for health care benefits including cardiorespiratory function, gait stability, quality of life, and happiness in people with CP [65, 66]. Therefore, any sports or exercise, including swimming, martial art, ballet, dancing, climbing, etc., that children want to do is recommended if it is not dangerous.
Hip x-rays
Given GMFCS II and good migration percentage at age 2.5 years old, hip x-ray could be delayed until the age of five if she maintains her ambulatory function [10, 68]. However, regular follow-up every 6 months or at least every year for clinical examination and checking her gait pattern and ambulatory function is needed, because some GMFCS II children may change to a lower or higher GMFCS level as they grow [69–71]. Furthermore, as the patient is young, there is a higher chance of changing her ambulatory function [71]. In addition, Korea has relatively good access to medical care in terms of distance and cost. Therefore, regular follow-up is possible only with the will of the patient and caregiver. And x-ray could be done if it is necessary at any time.
Post-injection recommendations
As described above, chemoneurolysis only is not sufficient to attain a gait improvement goal. The girl is showing a ‘jump knee’ gait pattern, and the doses of BoNT may not be sufficient to relieve the spasticity fully. Follow-up after the injection is needed, but AFOs will probably be necessary. Also, there might be some weakness and balance problems, so physical therapy or other forms of physical training are usually beneficial. Physical therapy is covered by health insurance in Korea, and many children with CP are receiving physical therapy, but supply has not kept pace with demand. However, physical therapy is usually more static than active and involves less physical activity than aerobic exercise or children’s leisure activities. Moreover, it is not fun. Children with GMFCS I and II could participate in more active leisure activities. However, resources and other limitations are always problems for leisure activities. Taekwondo, a type of martial arts, is popular among typically-developing children in Korea, and nearly every town has a Taekwondo studio. If a child wants to do Taekwondo, I encourage the parent to try it although there is limited direct evidence. Usually, Taekwondo programs include sitting upright, stretching, and various isolated movements of arm and leg motions. Even better is that children can socialize with their friends while learning at the studio and feel a sense of accomplishment as they progress. Recently, good practice recommendations for interventions to improve physical function for children and youth with CP were published [44]. These also mentioned that interventions should target the child’s chosen goals and be enjoyable and motivating, and many will agree with this.
Since the child is young and the degree of ankle contracture is not severe, orthopedic surgery is unlikely to be discussed immediately. Repeat injections will be considered after efficacy of injection is assessed. The repeat BoNT injection will be considered in 4–6 months to prevent muscle atrophy [55, 72]. SDR may be recommended sooner if the BoNT injection is ineffective or spasticity becomes more severe because SDR plus intensive physiotherapy could improve gross motor function and quality of life in children with spastic diplegic CP [73]. Furthermore, there was more long-term gait improvement in children with GMFCS I and II compared to III after SDR [74]. However, exact eligibility criteria for SDR are not fully defined, and consensus guidelines should be developed [73, 75].
Outcome measurement tools
Spasticity is usually checked by the modified Tardieu scale and MAS. Also, hip, knee, and ankle ROM measurements should be done in patients with lower limb spasticity to confirm the fixed contracture. In addition, a check-up at two positions for ankles, knee flexed and extended, both spasticity and ROM is necessary to identify whether the gastrocnemius muscle is the causative muscle or the soleus muscle. Unfortunately, there is no gait lab in the author’s hospital, so it is hard to analyze the gait pattern objectively. However, observational analysis of the child’s gait with and without AFO at the clinic is done to see how relieved tone affects the functional gait. Sometimes, recording their gait to measure knee ROM with a smartphone application could complement observation more accurately.
I’m a physiatrist in Korea working at a university hospital, and I meet many children with developmental delays, including CP, in my clinic. It has been a little over 10 years now that I have seen mainly children.
United States author: Dr. Gormley
Preface
This child presents at 3 years old with spasticity that may be interfering with her functional development. Typically, I like to address spasticity anytime it may be interfering with development, even if the child is still making some developmental gains. If addressing the spasticity in a child could potentially foster additional functional gains or minimize the consequences of spasticity, then I would consider treating them at any age even if the child is non-ambulatory. Some clinicians will not treat spasticity if the child is still making gains, but I usually like to maximize functional improvement, particularly if I think their spasticity is slowing their development. Ideally, I would be following this child along with her therapists and other providers from infancy since CP is considered a risk.
This child is a little too young for a rhizotomy at our center but the potential for a future rhizotomy would influence my decision to treat her spasticity at this age. At our center, the ideal candidate for a rhizotomy is an ambulatory child with spastic diplegic CP from prematurity and PVL who is between the ages of 4–8 years old, has good cognition and behavior, and has good social support. I usually advocate for a more conservative surgical technique for patients who are GMFCS II since excessive tone reduction can cause weakness. When this child got older, I typically would recommend a rhizotomy from L1 to S1 (not S2 since this may overly weaken plantarflexion long-term), cutting only about 25% of rootlets and dissecting at least a total of 100 rootlets. The goal for this child would be to improve her gait pattern, stability when walking, and walking speed and endurance. A. Chemoneurolysis to reduce this child’s lower extremity spasticity would be appropriate. I would consider BoNT and phenol. The most problematic spastic muscles are the plantarflexors, hamstrings, hip adductors, and hip flexors. The dynamic hypertonia and how it adversely impacts their gait and function may be different when watching this child walk than on her passive physical exam. Although this child does not have a significant amount of proximal lower extremity hypertonia on passive exam, there appears to be a significant amount of proximal dynamic hypertonia when she walks which may warrant chemoneurolysis. Since this child has only “mild scissoring” while walking, I would not initially inject phenol into the obturator nerves. I would consider phenol if after the initial BoNT injections into the hip adductors scissoring persisted or if I was restricted in the amount of BoNT I could inject in the other areas if some toxin was used in the hip adductors.
For this child’s first round of injections, I would inject only BoNT and not phenol. I would address her plantarflexors, medial hamstrings, and hip adductors. I probably would not inject her hip flexors unless significant subluxation of her hips was noted on x-ray, which would be unusual in this scenario [76].
Muscle injection plan: Gastrocnemius/soleus: onabotulinumtoxinA or incobotulinumtoxinA 50 U/side, or abobotulinumtoxinA 100 U/side. Medical hamstrings: onabotulinumtoxinA or incobotulinumtoxinA 25 U/side, or abobotulinumtoxinA 50 U/side. Hip adductors: onabotulinumtoxinA or incobotulinumtoxinA 25 U/side, or abobotulinumtoxinA 50 U/side. For localization, I would use a combination of electrostimulation (e-stim) and electromyography (EMG) feedback and possibly ultrasound. I usually use a machine that can toggle between e-stim and EMG. I will toggle depending on the patient’s response during the injections. I use ultrasound localization routinely for the salivary glands and occasionally for the hip flexors. But some of my partners will routinely use ultrasound guidance for all muscles. At my institution, which localization technique to use is more of a personal preference, but everyone uses some form of guidance. For sedation, we most commonly use NO but also midazolam or general anesthesia depending on the clinic setting and the patient’s perceived tolerance. Both NO and midazolam sedation are considered conscious sedations. For NO sedation, a respiratory therapist administers the NO, and the patient is continuously monitored with oximetry and by a separate nurse. For midazolam sedation, oximetry and nurse monitoring are continually present. Most combination injections with phenol and BoNT are done under general anesthesia, but patients who only receive BoNT injections but cannot tolerate conscious sedation will often use general anesthesia. After the child’s initial injections, if she had only minimal or short-lived improvement in her scissoring, then I would consider low-dose phenol blocks to the obturator nerves and distributing the BoNT to the other areas injected. I would consider adding BoNT to the hip flexors with subsequent injections if the child continued to walk with increased hip flexion or an anterior pelvic tilt or if hip subluxation of more than 20% was seen on x-ray. I would also consider hip flexor injections if her Thomas test was > 10 degrees.
B. Another pharmacological intervention to consider is oral baclofen. But oral baclofen usually only has modest functional benefits if any in this setting. Before starting baclofen in a child younger than 3 years old, consider the gastrointestinal side effects sometimes seen with baclofen in younger children. If this child was started on oral baclofen I would start at a low dose and titrate slowly up to 1 mg/kg/day divided TID, then reassess her. If she tolerated the baclofen and had some clinical benefit, then I would consider titrating up to a higher dose (about 2 mg/kg/day). If there was no significant clinical benefit, then I would taper the child off baclofen. I would not typically start diazepam, tizanidine, or dantrolene in this child because of the limited functional benefit and risk of sedation or muscle weakness.
Most spasticity management is a combination of treatments. I usually treat with a combination of interventions and not one intervention in isolation. If a combination of injections, oral medications, therapy, bracing, and other measures are more effective collectively than if used individually then I will use them in combination. I will not necessarily start with the most conservative intervention and get more aggressive as the treatment progresses. If early in treatment I feel that a more aggressive measure would have the best outcome, then I will initiate that treatment early. I will usually separate the initiation of various treatments by several weeks so their individual effect can be better measured.
C. This child may benefit from serial casts; I would probably cast her with bilateral short-leg walking casts for 2–4 weeks with weekly cast changes until her ankles can be easily dorsiflexed passively with her knee fully extended 10 degrees past neutral. In this child’s scenario, usually only 2 weeks of casting would be needed. For convenience, I usually start casting the same day as the injections, but occasionally I will wait a week after the injections to let the BoNT have time to reduce their spasticity and improve the tolerance of the cast in the first few days. Some clinicians would not cast this patient since BoNT injections and daily stretching exercises will improve her ROM, but casting can improve her ROM more quickly and significantly, and the walking casts can more quickly facilitate a plantigrade walking pattern.
The brace I would most likely recommend for this child during the day is a hinged AFO. If a patient can walk on their toes across the room, then they have enough plantarflexion power to take advantage of the ankle movement in hinged AFOs. Solid ankle or posterior leaf-spring AFOs may help increase the knee extension moment in stance but make it more difficult to go up and down steps, get up off the floor, and make lateral movements. I would check the patient’s ankle strength while walking after the injections before formally deciding on the type of AFOs. If a patient walks with a plantigrade foot position, has poor push-off and a crouch, I would consider a solid or stiff posterior leaf-spring AFO to improve knee extension in stance. If a patient has enough plantarflexion strength to walk on their toes across a room, then I want to allow them to use this dynamic strength while walking and use a hinged AFO with a plantarflexion stop at 90 degrees to minimize excessive plantarflexion while walking. Supramalleolar orthoses probably wouldn’t provide the needed stability and restriction of plantarflexion. I would get a hip x-ray when initially seeing the patient. Subsequent x-rays intervals would depend on the degree of hip subluxation and the American Academy for Cerebral Palsy and Developmental Medicine hip surveillance protocol [67]. I would continue with aggressive physical therapy after the injections at 2–3 times per week for about 1 month and then 1–2 times per week. She should have a home exercise program and walk regularly during the day.
I would consider knee immobilizers or AFOs at night if significant hamstring or plantarflexor contractures persisted. The current contractures are not severe and may not warrant night splinting after injections, but if contractures did persist, I would consider bilateral or alternating splints at night. If a child has popliteal angles > 45 degrees or plantarflexion contractures greater than neutral, then I will commonly use night splints. Night splints can help improve ROM [77, 78]. The typical outcome measures I use are ROM, MAS, Tardieu, and predetermined gait and functional goals. These functional goals are not necessarily formulized within a standard outcome measure, but the Goal Attainment Scale or GMFM can be used.
Conflict of interest
No conflicts of interest or funding to disclose.
