The effect of botulinum toxin A (Botox) injections used to treat limb spasticity on speech patterns in children with dysarthria and cerebral palsy: A report of two cases

1. Introduction

A variety of treatments are used to improve muscle tone and movement patterns in children with cerebral palsy. Two recently introduced pharmacologic treatments, intrathecal baclofen and Botulinum toxin A (Btx-A) injections, improve gross and fine motor control and may also yield improvements in other aspects of function. Intrathecal baclofen treatment has been shown to improve quality of life and ease of care, speech and oral motor activities, nutrition and appetite, respiratory support, and saliva control [1, 2, 3, 4]. Improved muscle tone and gross and fine motor movements have been reported in patients receiving Btx-A injections to reduce limb spasticity. However, the best outcomes for reducing spasticity and improving motor function may occur when Btx-A is used in conjunction with other treatment measures, particularly physiotherapy. Among the specific treatments used in addition to Btx-A were strength training and the neurodevelopmental treatment approach. In addition to improvement in limb tonicity and movement, changes in postural control have been noted [5, 6, 7, 8, 9, 10]. However, there are no published reports of improvement in oral motor function or speech production associated with treatment of spasticity or dystonia with Btx-A. There are reports of changes in respiratory support for speech production with changes in sitting posture and trunk control [11, 12, 13]. Anecdotal clinical observations of improved speech production in children with cerebral palsy being treated with Btx-A injections in the limbs by two of the authors (MW and JM) led to this investigation. This report presents results of longitudinal speech evaluation of two children with cerebral palsy given injections of Btx-A for treatment of lower limb spasticity. Speech evaluations were accomplished at baseline (date of injections) and 4- and 10-weeks post-injection. The Marshfield Clinic Institutional Review Board approved this case series, and parents of the subjects provided written informed consent to the research and to publication of the results.

2. Case report

Subject 1 was a 6-year-old female born prematurely at 28 weeks gestation with a primary diagnosis of spastic quadriparesis. Function was rated at a Gross Motor Function Classification System (GMFCS) level of IV and a Communication Function Classification System (CFCS) level of II. Previous assessments indicated normal hearing acuity and language development and a primary communication disorder of dysarthria. Subject 1’s speech was characterized by significantly reduced respiratory support resulting in reduced loudness; variable voice quality (strained-strangled to breathy); adequate resonant voice quality; articulatory imprecision; and multiple articulation errors. Intelligibility was reduced if the context was not known or if the environment was noisy.

Subject 2 was an 11-year-old male born prematurely at 33 weeks gestation with a primary diagnosis of spastic triplegia. Prior assessment indicated GMFCS and CFCS levels of IV, normal hearing acuity, and a primary communication disorder of dysarthria. Subject 2’s speech was characterized by reduced respiratory support resulting in variable loudness (normal to inadequate); variable voice quality (normal to strained-strangled to breathy); adequate resonant voice quality; articulatory imprecision; and multiple articulation errors. Intelligibility was reduced if the context was not known and if voice volume was reduced.

Informed consent was obtained from the parents of each child prior to Btx-A injections and speech evaluation per Marshfield Clinic’s Institutional Review Board requirements. Subjects served as their own controls.

Both children were receiving speech and language therapy services through their respective school programs. Speech therapy services were limited to language goals during the month preceding the subjects’ initial visit to provide a speech therapy “washout” period in an effort to reduce confounding effects of ongoing treatment. Both children had received prior Btx-A injections for spasticity. Most recent Btx-A injections were administered 12 and 7 months prior to the injections for this investigation for Subjects 1 and 2, respectively. Subject 1 had received six previous administrations of Btx-A, and Subject 2 had received 14 previous administrations of Btx-A. Both subjects experienced improved motor function with no adverse reactions.

The initial visit for this investigation included acquisition of a baseline speech sample by the speech pathologist (MW) and physical evaluation, including passive range of motion in the lower extremities, by the pediatric physiatrist (JM). Initial Ashworth scores were obtained at visits 6 and 4 weeks prior to the injection date for Subjects 1 and 2, respectively. Following speech and physical evaluations, Btx-A injections were administered by the pediatric physiatrist (JM) according to international standard of care guidelines [14]. Subject 1 received a total of 200 units Btx-A (9 units/kg) divided amongst singular injections of 33 1/3 units each in the right and left adductor longus, right and left gracilis, and right and left hamstrings. Subject 2 received a total of 400 units Btx-A (7.87 units/kg), which included four injections of 33 1/3 units each in the right hamstring, five injections of 33 1/3 units each in the left hamstring, and two injections of 25 units each in the right and left gracilis muscles. Large muscles were identified by palpation. The needle was aspirated prior to injection to avoid intravascular injection. Neither subject received Btx-A injections in the upper extremities, neck, or salivary glands. Follow-up sessions were scheduled at 4 and 10 weeks post-injection for repeat speech sample recording and physiatrist evaluation.

The speech-language pathologist (MW) and aspeech scientist (RK) assessed speech samples. Each sample was recorded in a sound treated booth. Subjects were practiced in the repetition of sentences at least once. Stimuli for the sentence repetition task were presented in random order via audio recording. Recordings were made on a MacBook computer with a high quality unidirectional microphone placed 15 cm from the speaker’s mouth. Samples were transferred to a master CD and scored twice by both listeners for reliability purposes. Speech samples included the following components:

Sustained vowels: Subjects were instructed to take in all the air they could and to sustain each vowel as long as possible with encouraging gestures from the investigator during production. Three repetitions of each vowel were recorded (/i/, /a/, /u/, and /ae/). Maximum loudness: Subjects were instructed to take in all the air they could and produce the vowel “ah” (/a/) as loud as they could for three trials. Sentence Repetition: Each subject produced five repetitions of each of the following sentences:

Automatic (overlearned/rote) speech: Subjectswere asked to count from 1 to 20 at a comfortable rate. Spontaneous speech (novel response to a stimulus) sample: Each subject was asked to describe his/her home.

Listeners rated severity of dysarthria based on the automatic speech sample. Intelligibility was rated on the listener’s orthographic transcription of the first 30 words of the spontaneous speech sample. The listener was allowed to listen to the connected speech sample twice while transcribing. The listener also rated five perceptual dimensions (loudness control, rate, voice quality, articulation, and phrasing) on an interval scale from 1 (performance within normal limits) to 7 (extremely severe involvement) [15, 16]. Listeners were allowed to hear connected speech samples as many times as necessary for scoring the five dimensions. Finally, a parent questionnaire was administered to a single parent at each time point to assess parental perception of their child’s speech quality and functionality (Table 1). Fifteen questions required a response based on a numerical scale. Fourteen item responses were based on a 5-point scale, and one item response was based on a 3-point scale.

Increased PROM and improved Ashworth scores compared to baseline measures indicated that theBtx-A was effective in reducing limb spasticity (Table 2). No change was observed in maximum phonation times for either subject (Table 3) or for subject 1 with respect to maximum loudness (Table 4). Subject 2 showed similar scores for maximum loudness in each session, but had increasing stability of performance over the three sessions (Tables 3 and 4). Average syllables per breath for repeated sentences showed improvement (Table 4). For the perceptual rating tasks, inter- and intra-listener reliability was high. Mean values were calculated for severity of dysarthria and intelligibility. The five perceptual dimensions for each subject are displayed in Table 5. Differences in scores were small, usually no more than 1 scale value. Changes in intelligibility reflected no more than a three word difference.

The percentage of consonants correct in repeated sentences varied over the three sessions. For Subject 1, the percentage of consonants correct improved from initial to final session for 6/12 sentence groups judged, omission errors were reduced for 5/12 groups, and distortion errors were reduced for 1/12 groups. For Subject 2, the percentage of consonants correct was improved for 3/12 sentence groups, omission errors were reduced for 2/12 groups, substitution errors were reduced for 1/12 groups, and distortion errors were reduced for 5/12 groups.

Responses to the parent questionnaire over time were compared via Wilcoxon’s matched-pairs signed-ranks test. This non-parametric test was applied to parent responses for the 18 item parent questionnaire responses at baseline and at four and ten weeks post-injections though questions 16–18 were excluded from this analysis since the responses were not quantitative (Table 1). Of the 15 remaining items in the questionnaire, 14 item responses were based on a 5-point scale, and one item response was based on a 3-point scale. Parental responses were compared at four and ten weeks post-injections to the baseline response. As parental responses to the questionnaires were linked to each child, all time points were considered matched data for analysis. The 15 items used for baseline, 4 weeks, and 10 weeks were averaged from patient 1 and patient 2 for a total sample size of 15 before the Wilcoxon signed-rank test was performed. Scores at 4 and 10 weeks post-injection demonstrated statistically significant improvement over pretreatment baseline measures ( 0.05). For Subject 1, gains were seen in intelligibility for unknown material and willingness to speak with friends in a one-on-one or group setting. For Subject 2, gains were seen in intelligibility for known material and in a noisy background and willingness to communicate with friends one-on-one or in a group setting.

Though the use of Btx-A injection to treat spasticity in children with cerebral palsy is thought to result in functional improvements, effects on oral motor function and speech production have not been assessed. In the two children described here, speech evaluation was performed before Btx-A injection and at 4 and 10 weeks of follow-up. Physiatrist evaluation indicated that the Btx-A was in effect when speech samples were captured. As neither subject received speech therapy services during the 10-week follow-up period, changes in speech production are not thought to be a direct result of speech treatment.

These preliminary findings indicate improvement in aspects of speech in two children with dysarthria. Specifically, improvements were noted in production of consonants, loudness control, and syllables produced per breath, all of which may reflect improved respiratory support for speech production. Interestingly, the measurement of syllables per breath showed most improvement at 4 weeks post-injections. Although this was reduced at 10 weeks post-injection for both subjects, the values still exceeded baseline. The mechanism for improved respiratory support may relate to improved postural control resulting from reduced tone and improved function in the extremities. Parents perceived improved intelligibility and willingness to speak. In written comments, the parent of Subject 1 described their child’s improved speech production in week 10 by saying “…seems to have gotten louder and can hold sentences longer.” Subject 2’s parent commented in week 4 that the child was “not slurring words so much, seems to have for [sic] stamina for a conversation” and also reported “less drooling” at week 10.

The tasks chosen to analyze the effects of Btx-A injections on speech patterns for two patients reflected performance in the areas of respiratory support for speech production, vocal quality, and oromotor function for consonant sound production. For this small sample, tasks that showed most changes were for maximum loudness, transcription of sentences for assessment of consonant production, and length of breath groups. Perceptual assessment for severity of dysarthria, five dimensions of speech, and voice production and intelligibility ratings showed small differences but not distinctive changes. Perhaps for this type of intervention observed over a relatively short period of time, perceptual impressionistic measures may not be as sensitive to changes in performance as more quantitative and objective measures. Future case reports and larger trials may include additional objective measures of respiratory function for speech, such as the ability to develop and maintain intraoral air pressure or the ability to prolong vowels and fricatives without interruptions or marked variations in sound pressure level. More detailed acoustic measures suited to single word or multi-word productions could include segment durations, formant-frequencies for target vowels, and variation in vocal fundamental frequency and intensity. Because no single form of intelligibility assessment is suited to various purposes and different levels of speech/language abilities, it is recommended that intelligibility assessment include two or more measures, such as a single-word test to evaluate features of monosyllabic production, a rating scale completed by the clinician or other trained individual, or a parent questionnaire (e.g., the Intelligibility in Context Scale [17] or a syllable-based intelligibility index [18]). The latter three assessments are appropriate to rating conversational speech or connected speech tasks, such as sentence repetition. The parent questionnaire, which reflected quality and intelligibility of speech production outside the testing situation and willingness to participate, was useful and related to changes seen in objective measures; however, we recognize our initial sample size was small and our findings are based on two patients. Further work is needed to analyze the qualitative and quantitative changes in speech production post-Btx A injection in pediatric patients with cerebral palsy.

It may be that as with gains in gross and fine motor skills resulting from Btx-A injections, gains in speech production could be enhanced and maintained with an intense program of speech therapy conducted over a set period of time following the injections such as LSVT LOUD, which is directed toward increased respiratory-phonatory effort [19]. Additional work with a larger sample size is warranted to further assess the nature of changes seen following Btx-A injections and type, intensity, and duration of speech treatment to achieve maximal results. Changes in saliva control and swallowing were not analyzed for these patients, but anecdotal evidence from one patient’s parents indicates that saliva control could also be improved with Btx-A injections. These variables could be incorporated into future case studies and clinical trials. Future researchers could also investigate the effects of Btx-A injections to other sites, such as upper extremities and parotid glands, for effects on speech production and oromotor function.

Even though many candidates were identified and approached for inclusion in this case series, the number of patients included in the final analysis was smaller than projected. Parents were willing to travel to the medical center for the Btx-A injections, but they cited cost of travel and job and family obligations as deterrents. Future case studies and clinical trials could include follow-up data collection in centers where patients can be accessed without additional travel (e.g., a team consisting of a physical therapist and speech-language pathologist traveling to school settings) and may also include untrained listeners to reduce bias that may have been present with using only experienced listeners. In conclusion, treatment of limb spasticity with Btx-A resulted in improved speech production for both children discussed in this report, but further research with more patients is warranted to replicate and extend these observations.