Abstract
PURPOSE:
The aim of this study was to determine the effectiveness of botulinum toxin type A (BoNT-A) injections in infants with congenital muscular torticollis (CMT) who were refractory to conservative management.
METHODS:
This was a retrospective study in which all subjects included were seen between 2004 and 2013 and were deemed appropriate for BoNT-A injections. A total of 291 patients were reviewed for inclusion in the study, and 134 patients met the inclusion criteria. Each child was injected with 15–30 units of BoNT-A into each of the following muscles: ipsilateral sternocleidomastoid, upper trapezius, and scalene muscles. The key outcome and variable measurements analyzed included age at time of diagnosis, age at time of initiation of physical therapy, age at time of injection, total number of injection series utilized, muscles injected, and degrees of active and passive cervical rotation and lateral flexion pre- and post-injection. A successful outcome was documented if a child could achieve 45° of active lateral flexion and 80° of active cervical rotation post-injection. Secondary variables including sex, age at time of injection, number of injection series utilized, surgery required, adverse effects of botulinum toxin, presence of plagiocephaly, side of torticollis, orthosis used, presence of hip dysplasia, skeletal anomalies, complications during pregnancy or birth, and any other pertinent information regarding the delivery were also measured.
RESULTS:
Based on this criteria, 82 children (61%) had successful outcomes. However, only four of the 134 patients required surgical correction.
CONCLUSION:
BoNT-A may be an effective and safe method for treatment in refractory cases of congenital muscular torticollis.
Introduction
Congenital muscular torticollis (CMT) is a postural deformity of the neck detected at birth or shortly after birth, largely resulting from unilateral shortening and fibrosis of the sternocleidomastoid (SCM) muscle [1]. CMT has a slight male predominance and is associated with firstborn children [1–3], involvement of the right more than the left SCM, facial asymmetry, and less frequently other musculoskeletal anomalies including brachial plexus injury, hip dysplasia, metatarsus adductus, and talipes equinovarus [1, 2]. Other diagnoses that can present with asymmetrical head posturing include Klippel-Feil, neurologic disorders, ocular disorders, paroxysmal torticollis that alternates sides, spinal abnormalities, and SCM neoplastic masses such as rhabdomyosarcoma [1].
Thus, a thorough differential diagnosis and diagnostic evaluation are important to rule out these other associated anomalies. Diagnosis of CMT is based on clinical examination. Studies support that early intervention leads to improved outcomes, prevents secondary sequalae, and decreases episodes of care, so early referral to physical therapy is important [1]. Most cases of CMT (studies range from 50–95%) resolve within a few months, either spontaneously or with therapy [3–5]. CMT can be classified into three types: postural, muscular, and SCM mass [1–5]. Postural CMT is the infant’s postural preference, whereas muscular CMT presents with decreased passive range of motion and SCM tightness [1]. SCM mass CMT consists of fibrotic thickening of the SCM with resultant decreased passive range of motion [1]. Typically, infants who are identified early with postural CMT and who get therapy have the shortest duration of treatment time, while those who are identified later (3–6 months of age) with a SCM mass, require longer and possibly more invasive treatments, such as botulinum toxin injections or surgery for correction [1]. Also of note, as children age and become more resistant to therapy, stretching can become more difficult to perform effectively.
Treatment of torticollis involves early implementation of physical therapy [1]. Physical therapy consists of active and passive cervical range of motion exercises and a home exercise program consisting of proper positioning, carrying, and handling techniques that can be taught to the caregivers [1]. Improvements in cervical range of motion can be monitored through measurements of active and passive cervical range of motion through use of visual inspection, photographs, and a goniometer [1]. Another tool that can be used to assess torticollis in infants is the Muscle Function Scale [1]. It assesses the function of the lateral flexors of the neck in infants through active range of motion using a scale of zero to six. Cervical orthoses have been used as well for the treatment of CMT. Historically, surgical intervention has been required to release the fibrotic tissues when conservative measures have failed [1, 4]. However, surgery is invasive and may not be necessary. Children who do not respond to physical therapy should be considered candidates for botulinum toxin injections to help improve cervical range of motion [4, 7]. Botulinum toxin can enhance the effectiveness of stretching on the side of the muscle shortening and allow strengthening of overstretched and weakened muscles on the opposite side of the neck [4, 5]. The purpose of this study was to determine the effectiveness of botulinum toxin type A (BoNT-A) injections in infants with CMT refractory to physical therapy, as well as look at different secondary variables to see if they would affect success rates. To date, there have been no studies looking into how these secondary variables affect outcomes, nor setting guidelines for the injections regarding muscle selection and dosing.
Methods
This was a retrospective study in which the protocol was reviewed and approved by the Institutional Review Board before initiation. This project received a waiver of permission/assent.
Participants
All subjects included in the study were seen in the Torticollis Management Clinic between January 1, 2004, and December 31, 2013, and were deemed appropriate for BoNT-A injections. A medical record review of ICD-9 code 723.5 (torticollis, unspecified) and CPT code 64613 (chemodenervation of muscle[s]; neck muscle[s]) yielded 291 potential patients; 157 were not included due to the wrong diagnosis or insufficient documentation. A total of 134 patients with CMT were included in this analysis. All were classified as having muscular torticollis. Those with postural torticollis were not included as management with botulinum toxin would not be warranted. No patients had a documented SCM mass.
Inclusion criteria for this study were diagnosis of CMT and at least two months of physical therapy (including active and passive range of motion exercises, as well as a home exercise program) with either no improvement in initial measurements or having hit a plateau regarding improvements in active and passive cervical range of motion. Inclusion criteria also included infants aged six to 24 months at time of first injection, who received 15 to 30 units of botulinum toxin into each muscle injected and had recorded pre- and post-treatment measurements for active or passive cervical rotation and flexion. Exclusion criteria included children who fell outside the age range of six to 24 months, had postural torticollis, warranted further work up for the torticollis, had less than two months of physical therapy, received more than 30 units of botulinum toxin into each muscle injected, and did not have follow up within the hospital system after the injections.
Each child was injected with 15 to 30 units of BoNT-A following standard clinical procedure into each of the following muscles: ipsilateral SCM, upper trapezius, and scalene muscles. These muscles were selected based on the understood anatomy and mechanics involved in head rotation, lateral tilt, and associated shoulder elevation. The dose used was based on clinical judgement, within the range of 15–30 units per site. No sedation was used. All infants were positioned by the assisting nurse. The muscles were palpated, and electrical stimulation guidance was used for location of muscles. Injections were followed by continued outpatient physical therapy in all patients. SPSS was used for all the statistical analyses in this study. Chi-squared tests were used to determine if there was a relationship between the different categorical variables and successful outcomes. Fisher’s exact test was used when a chi-squared test was not appropriate. In all statistical analyses, a P value of 0.05 or less was considered statistically significant.
Participants involved were infants aged six to 24 months who were diagnosed with refractory or unresolved CMT. This study defined refractory as CMT in patients who underwent two months of physical therapy and hit a plateau in improvement of passive or active cervical rotation and/or flexion as documented by the treating therapist. A plateau was defined as when the child did not reach normal range of motion despite improvements with therapy. In this study, normal cervical range of motion was 45° of active lateral flexion and 80° of active cervical rotation. Participants were considered for botulinum injections if they did not have one or both ranges of motions intact pre-injection. Physical therapy both pre- and post-injections consisted of active and passive range of motion exercises and a home exercise program. Active range of motion exercises are activities the infant actively does to improve range of motion, whereas passive range of motion exercises are stretches performed on the infant by either a therapist or parents. A home exercise program was given to parents to continue the exercises and stretches at home while not in therapy. All participants in the study had to be re-evaluated by the physician or physical therapist within the hospital system following the procedure with documented measurements for active cervical rotation and flexion. The physician and therapist each had over 10 years’ experience treating CMT prior to the start of the study. A thorough neurological exam was performed, and cervical x-rays were ordered to rule out any skeletal anomalies such as a C1-C2 subluxation. Any patient with an abnormal exam, thus warranting further imaging or consultation (such as ophthalmology), was excluded from the study.
The frequency and duration of therapy were at the discretion of the treating therapist for both pre- and post-injections. A total of 127 of the patients were injected in all three of the muscles with four receiving injections in the SCM only. One received upper trapezius only, and two received upper trapezius and scalenes. Muscle selection was based on patient presentation. If the patient demonstrated deficits in lateral flexion as well as rotation, then the ipsilateral SCM, upper trapezius, and scalene muscles were all injected. If there was an isolated deficit in rotation, then only the SCM was injected. Likewise, if there was an isolated deficit in lateral flexion, then only the upper trapezius and scalenes were injected. Injections were performed by the same physician using Dantec Clavis electrical stimulation for guidance.
The physical therapist present in the Torticollis Management Clinic measured both active and passive cervical rotation and lateral flexion. This was done by moving the infant’s neck both actively (looking at an object or toy) and passively to their end points of bi-directional horizontal rotation as well as lateral flexion. A second person took anterior view photographs in upright resting head position and endpoint range of motion in lateral flexion. A goniometer was then used on the photographs to get the measurements. Angles were measured in the anterior views between points on the sternal notch, the chin, and the mid forehead. For cervical rotation, measurements were made with a goniometer placed under the chin, measuring rotation away from neutral. Since the Muscle Functional Scale was used inconsistently, it was not utilized in the data. After the injections, the patient was to follow up with their treating physical therapist. Their home exercise program would be tailored by the treating physical therapist based on the progression shown.
Follow-up after injections occurred between two months and two years post-injection. The primary outcome measures were the degrees measured of active lateral flexion and active cervical rotation post-injection, as well as whether surgery was required. A successful outcome was documented if a child could achieve 45° of active lateral flexion and 80° of active cervical rotation post-injection. Although passive lateral flexion and cervical rotation were documented as part of the examination in most patients, it was not used as a primary outcome, as passive range of motion alone does not equate to functional improvement. For that reason, only active range of motion was used. These outcome measures were determined by the clinic. Normal range of motion was considered 45° of active lateral flexion and 80° of active cervical rotation. Secondary variables that were documented included sex, age at time of injection, number of injection series utilized, presence of plagiocephaly, side of torticollis, orthosis used, presence of hip dysplasia, skeletal anomalies, complications during pregnancy or birth, and any other pertinent information regarding the delivery such as prematurity, breech positioning, Cesarean section, and multiple gestation. Presence or absence of SCM tumor and scoliosis and presence of initial head tilt were also documented. However, none of the children had a documented SCM tumor or scoliosis, so these factors were not included in the statistical analysis. Also of note, all the children initially had a documented head tilt, so this variable was likewise not included as a factor affecting outcomes. There were no adverse effects documented following the injections; hence, this was not included in the statisticalanalysis.
Results
A total of 134 children (61% male and 39% female) received BoNT-A injections for treatment of refractory CMT between 2004 and 2013 and met the inclusion criteria. All the infants included in this study were classified as having muscular torticollis, meaning they had clinical thickening or tightness of the SCM muscle on the affected side without a palpable or visible tumor. This was a retrospective study, and no infants included had a documented SCM tumor during the Torticollis Clinic evaluation or prior documentation from the parents or treating therapist. The characteristics documented are listed in Table 1. Age at time of injection was between six and 12 months of age in 90 patients (67%), between 13 months and 18 months in 38 patients (28.5%), and between 19 months and 24 months in six patients (4.5%). There was no statistically significant difference in time to injection between successful and unsuccessful outcomes with a p value of 0.8557. The pre-injection and post-injection measures are listed in Table 2.
Characteristics of 134 Patients with Congenital Muscular Torticollis
Characteristics of 134 Patients with Congenital Muscular Torticollis
SD: standard deviation.
Measurements Pre- and Post-injections
Eighty-two children (61%) had successful outcomes, obtaining 45° of active lateral flexion and 80° of active cervical rotation post-injection (Table 3). When considering only lateral flexion, 83 (62%) achieved success, compared to 112 (84%) successful outcomes for cervical rotation. When lowering the outcomes for lateral flexion to 35°, only one child (63%) could be added to the success column, and another 15 children (74%) could be added when lowering it to 30°. In reviewing the secondary variables of the children, there was no significant determinant that predicted a successful outcome (Table 4). None of the secondary variables documented had any bearing on whether a treatment was successful or unsuccessful.
Successful Treatments vs Unsuccessful Treatments
Characteristics and Predictability of Successful Treatment
Successful outcome = 45° of active lateral flexion and 80° of active cervical rotation post-injection. Unsuccessful outcome = patient did not achieve 45° of active lateral flexion and 80° of active cervical rotation post-injection. *Fisher’s Exact Test.
Four children out of the 134 (3%) with refractory CMT who received botulinum toxin injections went on to have surgical correction. All four proceeded with surgery after one round of injections. In all cases, the parents opted to seek surgical treatment over trying botulinum toxin injections again. One of the patients who proceeded to surgery was born prematurely due to oligohydramnios and was also diagnosed with a brachial plexus injury at birth and had meningitis at two months of age. The injection was performed at 12 months of age. Another patient was born with multiple congenital heart defects requiring surgical repair. The other two had no distinguishing characteristics documented. There were no reported adverse reactions to the botulinum toxin injections in any of the patients.
Botulinum toxin can potentially augment the effect of stretching by temporarily weakening the affected muscles that cause torticollis [7]. Children who do not respond to physical therapy should be considered candidates for BoNT-A injections to help improve range of motion [4, 7]. Previous studies have demonstrated similar results in much smaller cohorts; however, this study evaluated the role of a specific dosing protocol and explored secondary factors and their effect on success rates.
Oleszek et al. performed a retrospective study in 2005 on 27 children who received botulinum toxin injections for CMT [7]. The SCM muscle (20–50 units) and/or the upper trapezius muscles (25–35 units) were injected [7]. Three children received repeat injections [7]. Twenty out of the 27 (74%) had documented improvement in neck rotation or active head tilt, measured by either quantitative or qualitative data [7]. One child ultimately underwent surgery (having received a botulinum toxin injection at age 17 months) [7]. The study was relevant to the patient population in this study based on age and muscles injected; however, there was not a standard for measuring range of motion, and it lacked measurements for nine of the patients.
Joyce et al. retrospectively evaluated 15 patients in 2005 who received botulinum toxin for refractory muscular torticollis (average age of 7.6 months) and had persistent loss of at least 10 to 15 degrees of passive lateral flexion of the neck despite a course of physiotherapy and stretching exercises [8]. Therapy was initiated between ages one and six months, and the length of therapy was between one and 14 months before initiation of botulinum toxin [8]. The patients underwent general anesthesia and were injected in 3-4 sites along the SCM with 25–50 units of botulinum toxin [8]. Full range of motion was considered to be 90° of cervical rotation and 45° of lateral flexion [8]. Fourteen of the 15 patients had significant improvements in neck range of motion subsequent to the botulinum injections, and only one child underwent surgery six months later [8]. The outcome measure was slightly different from the current study, which considered post-injection achievement to be 80° of active cervical rotation and 45° of active lateral flexion due to what was considered to be functional range of motion. This study was also interested in examining the two motions, cervical rotation and lateral flexion, separately.
In 2019, Limpaphayom et al. evaluated the combined use of physical therapy and botulinum toxin in the treatment of resistant CMT [9]. They prospectively looked at 39 children aged 13.5 months to 27.5 months who got 1–4 rounds of BoNT-A injections into the SCM muscle only [9], similar to Joyce et al. The patients were injected with 2-3 units/kg of BoNT-A over 2-3 sites along the SCM [9]. They looked at the improvement in range of motion, which can vary greatly depending on the initial range of motion documented. So, while the study demonstrated a significant improvement in range of motion, achievements may not be considered full range of motion; also, only the SCM was injected. In contrast, the current study set strict criteria for a successful outcome and injected multiple muscles as most of the patients had a laterocollis on examination.
Qui et al. reported a meta-analysis study in 2020 that evaluated the effectiveness of the botulinum toxin in CMT [10]. The study reviewed 10 articles with a total of 411 patients. It found the overall effective rate of botulinum toxin for CMT to be 84%, with a conversion rate to surgery at 9%. The adverse reaction rate was 1% [10]. Transient dysphagia and reaction site erythema were the most common adverse reactions reported [10]. The study concluded that botulinum toxin injections for the treatment of CMT are safe and effective [10]. Despite the 1% adverse reaction rate reported in this meta-analysis, the current study had no documented adverse reactions.
This study looked at age at time of injection and found that there was no significant difference between the age groups (6–12 months, 13–18 months, and 18–24 months) regarding successful outcomes. Although it is ideal to initiate therapy as soon as possible, this study would suggest that a child with refractory CMT can still get satisfactory results with botulinum toxin injections even if therapy is delayed, regardless of any of the other associated factors predicting outcome following injections. Despite the strict criteria for a successful outcome in this study, most of the parents reported that the treatment resulted in significant improvement in their child’s appearance and range of motion, as documented in the child’s follow-up note. Only four of 134 participants went on to have surgical correction due to lack of acceptable clinical improvement in cervical flexion and/or rotation. Additionally, the child only followed up with the physician if there were ongoing concerns. Most of the children were discharged back to their therapists after the injection. There may have been several more successful outcomes not captured in this study if the therapists were not within the clinic’s hospital system.
Study limitations
The main study limitation was that it was retrospective, with some vital statistics and temporal relationships that can be challenging to measure. There was also a risk of information bias. When dealing with retrospective studies, outcomes rely heavily on accurate recordkeeping. Several patients were excluded because they did not follow up in the clinic’s hospital system or due to lack of documentation. As stated above, follow-up with the physician was only scheduled if there were ongoing issues. So, there may have been several successful outcomes that were missed due to lack of follow-up or due to receiving therapy at an outside institution. The physician and therapist may have measured the child differently post-injection than the initial pre-injection measurement, which brings in to question inter-rater reliability for post-injection measurements. All pre-injections measurements were performed by the same physical therapist using the same measuring technique. The Muscle Function Scale is a validated measurement tool used for reliability in CMT [1]. Although the initial physical therapist doing the pre-injection measurements did use the Muscle Function Scale in their documentation, none of the post-injection measurements included it, so it was not able to be used. For that reason, active cervical rotation and lateral flexion were used as the primary measurements. Duration of treatment pre- and post-injection were recorded but varied among the patients. Pre-injection duration of therapy ranged between two months and almost two years, with post-injection therapy ranging from one month to one year. Another limitation was the lack of comparison group. This study did not have a comparison group because the patients who were referred to the clinic were sent for evaluation for botulinum toxin injections for their refractory torticollis. While some patients’ CMT did not warrant injections, there were not enough of them to compare to the injection group for meaningful comparison. Also, there was no control over the frequency and duration of therapy as they were at the discretion of the therapist. Going forward, a prospective randomized trial may help alleviate some of these limitations.
Conclusion
As an adjunct to physical therapy, botulinum toxin may be considered a safe and effective method for treatment in refractory, or unresolved, cases of CMT. It is a quick and simple procedure that is a suitable alternative to surgical correction. Although, it has been used in the treatment of torticollis for over 15 years now, this is the first study that provides dosing and muscle selection, as well as evaluates associated variables noted with torticollis. However, the results are encouraging and suggest that the next step would be to perform a prospective comparative randomized study evaluating different dosing ranges.
Footnotes
Acknowledgments
Thank you to Julie Musick, the main therapist in the Torticollis Clinic at Children’s Mercy Hospital, for helping to conceptualize and design the study. There was no funding for this study.
Conflict of interest
The authors indicate that they have no potential conflicts of interest to disclose.