Assessing Bulbar Function in Spinal Muscular Atrophy Using Patient-Reported Outcomes

INTRODUCTION

Spinal muscular atrophy (SMA) is an autosomal recessive motor neuron disease characterized by progressive muscle weakness caused by degeneration of lower motor neurons in the spinal cord and brainstem. The disease affects an estimated 10–16 per 100,000 infants [1]. SMA has a broad phenotypic spectrum of disease severity with a clinical classification based on age of onset and maximal function achieved. Patients with the intermediate form, type 2, are moderately affected as they achieve the ability to sit but never walk. Patients with the mildest form, type 3, are less affected as they achieve the ability to walk independently but may lose this ability over time with varying degrees of disability [2].

Individuals with more severe forms of SMA have historically experienced bulbar dysfunction that includes problems in communication and deglutition resulting from weakness in the oropharyngeal musculature controlled by the cranial nerves in the pons and medulla [3–8]. Abnormal bulbar function is a well-known complication of SMA with common physical complaints and prevalence previously reported [6, 9–13] with a negative impact on activities of daily living, health, well-being and social participation [3–5, 7, 8].

Previous questionnaire-based studies examining specifications of these deficits using patient-reported outcome (PRO) measures reported the three most common deficits in deglutition in individuals with SMA type 2 and 3 (age 3 to 45 years) to be choking (31%), difficulty conveying food to the mouth (20%), and difficulty chewing (20%), with patients with lower levels of motor function having 7.6 greater odds of reporting dysphagia [10]. Another commonly reported impairment using a PRO in those age 1 to 75 years is the perception of fatigue during mealtimes that prohibits an individual’s ability to safely consume full oral nutrition [6, 14]. Though fewer investigations have reported on specifications of communication deficits, those that have examined this domain, using questionnaires and structured interviews in individuals age 31 to 67 years, have indicated high reports of voice and speech problems that, similar to swallowing, were more common in patients with type 1 and 2 (45%) than in those with type 3 (17%), although no significant correlation was found between these symptoms and disease duration [13].

The approval of disease-modifying therapies for SMA was based on clinical trials of infants, children, and young adults [15–23] with no formal studies conducted in the chronic adult SMA population. Real-world evidence from adults with SMA has suggested improvement on treatment in several motor function and respiratory measures distinct from the progressive decline typically seen in untreated chronic SMA [24–30]. However, there is limited objective data regarding the impacts of disease-modifying therapies on communication and deglutition due, in part, to the scarcity of tools specific to assessing bulbar function in individuals with SMA [4, 31].

The use of PROs has been strongly supported by regulatory authorities both in US and in Europe to provide a better understanding of disease burden and impact of treatment on patients [4]. A Canadian adult SMA consensus-derived clinical outcome measure toolkit recognized gaps and poor availability of assessments to detect changes in multiple aspects of life including bulbar function (communication and swallowing), particularly in non-sitter and sitter populations [32], as this was previously identified as a priority for stabilization for persons with SMA type 2 and 3 [33]. Monitoring perceived changes in communication and deglutition is important in severely affected adult patients, in whom motor function testing of limb musculature is typically less sensitive or applicable due to the advanced disease stage and floor effects of available measures [34].

The primary objective of this study is to systematically assess perceived bulbar function (voice and deglutition) among treated and untreated non-ambulatory patients with SMA using two currently available PROs, the Voice Handicap Index (VHI) and Eating Assessment Tool-10 (EAT-10), and determine their relationships with clinical characteristics. Understanding perceived bulbar function was deemed clinically feasible as these PROs do not require the use of instrumental assessments or need for specialized equipment or training and could be administered during a routine clinic visit. We also sought to examine if perceived voice and deglutition deficits are associated with one another. We hypothesize that more severe, lower functioning individuals with SMA will experience greater perceived voice and deglutition deficits, and that these deficits will be weakly related to age.

MATERIALS AND METHODS

For this cross-sectional study, data were acquired prospectively as part of an IRB-approved multicenter natural history study of SMA at two of the Pediatric Neuromuscular Clinical Research (PNCR) Network sites (Stanford University School of Medicine and Boston Children’s Hospital / Harvard Medical School). All participants spoke English, and informed consent was obtained prior to enrollment. This study is in accord with the Helsinki Declaration of 1975.

RESULTS

We recruited 47 non-ambulatory participants (mean (SD) age 29.8 (13.7) years, range 10.3 –73.2) with a genetically confirmed diagnosis of SMA who completed both bulbar function PROs at their clinic visit. Twenty-five participants (53.2%) were recruited from Stanford University School of Medicine and 22 participants (46.8%) from Boston Children’s Hospital / Harvard Medical School. Demographic and clinical characteristics are described in Table 1. A majority of the 47 participants were SMA type 2 (70.2%), non-sitters (78.7%), with 3 SMN2 copies (77.5%), using any respiratory support (66.0%) and receiving disease-modifying therapy at the time of PRO completion (66.0%). In all cases the disease-modifying therapy that had been received was nusinersen.

Motor function testing (Table 2) revealed weaker phenotypes with all participants scoring≤22 out of 66 on the HFMSE and nearly half (42.9%) scoring zero, exhibiting a floor effect. Mean RULM score was 9.6 out of 37 with only 19.1% scoring zero, exhibiting a floor effect. Of the 40 participants assessed with RULM entry item A, 16 (40.0%) were unable to lift hands against gravity (entry item A score 0 or 1). No participants could raise their arms above their head with or without compensations (entry item A score 5 or 6). The CHOP-ATEND was able to capture subtler distal upper and lower extremity movements with a mean score of 24.0 out of 64 and no participants demonstrating a floor effect. Of the 22 participants assessed with the CHOP-ATEND, 11 exhibited full head control with trunk supported as described by item 12 [49]. Not all motor function testing outcomes were appropriate for administration based on SMA severity level and therefore limited data are available for each motor outcome (Table 2).

Patient-reported issues of bulbar function

Ninety-four percent of SMA participants reported perceived voice deficits, noted primarily in 8 of the 30 VHI items (Fig. 1). The problems most frequently reported include items related to sound production and articulation: ‘people have difficulty understanding me in a noisy room’ (87.2%); ‘my voice makes it difficult for people to hear me’ (74.5%); and ‘people ask me to repeat myself when speaking face-to-face’ (72.3%). Percentages of item scoring responses for all individual items of the VHI are provided (Supplementary Fig. 1 and Supplementary Table 1). Seventy-two percent of participant VHI total scores fell in the mild severity level, 21% in the moderate severity level, and 6% in the severe severity level (Table 2).

Fig. 1

Items in which greater than 50% of participants reported any problems VHI, Voice Handicap Index; F, Functional Domain; P, Physical Domain; EAT-10, Eating Assessment Tool-10.

Eighty-five percent of SMA participants reported perceived swallowing deficits, noted primarily in 3 of the 10 EAT-10 items (Fig. 1). The problems most frequently reported include: ‘swallowing pills takes extra effort’ (68.1%); ‘when I swallow food sticks in my throat’ (66.0%); and ‘swallowing solids takes extra effort’ (61.7%). Percentages of item scoring responses for all individual items of the EAT-10 are provided (Supplementary Fig. 2 and Supplementary Table 2). Sixty-eight percent of participants scored greater than or equal to 3 out of 40 indicating a problem with swallowingefficiency/safety.

Fig. 2

Association between VHI and EAT-10 total scores (r = 0.66, p < 0.0001). VHI, Voice Handicap Index; EAT-10, Eating Assessment Tool-10.

Although the EK2 is not a PRO of bulbar function, there are a few items assessed that are applicable to voice/speech and swallowing problems. Sixty-four percent of participants reported they are unable to cough effectively (item 8), 40.0% do not exhibit powerful speech as they were unable to sing or speak loudly (item 9), 64.0% require head support when driving or sitting still in wheelchair (item 12), 56.0% modify their foods in any way in order to eat them (item 14), 72.0% are unable to consume a whole meal in the same time as others (item 15), and 64.0% experience problems with swallowing, including choking (item 16).

Bulbar function associations

The VHI and EAT-10 total scores were moderately associated (r_s = 0.66, p < 0.0001) (Fig. 2). Significant differences in VHI total scores were found (Fig. 3) between those who scored 0–1 (median = 19) and those who scored 2–4 (median = 7) on RULM entry item A (W = 407, p = 0.03). No other clinical characteristics were associated with the perceived voice issues.

Fig. 3

Distributions of VHI total score in those with RULM entry item A 0-1 and 2-4.VHI, Voice Handicap Index; RULM, Revised Upper Limb Module.

The EAT-10 total scores were weakly associated with age (r_s = 0.32, p = 0.03) and symptom duration (r_s = 0.35, p = 0.02). Significant differences in EAT-10 total scores (Fig. 4) were found between those with SMA type 2 (median = 9) and type 3 (median = 1) (W = 136, p = 0.002), those with current functional status of non-sitter (median = 8) and sitter (median = 1) (W = 121, p = 0.002), those with RULM entry item A score 0–1 (median = 11) and score 2–4 (median = 2) (W = 456, p = 0.0004), and those using non-invasive ventilation (median = 8) and not on any respiratory support (median = 1) (W = 338, p = 0.04).

Fig. 4

Distributions of EAT-10 total scores by: a) SMA type, b) current functional status, c) RULM entry item A, and d) use of non-invasive ventilation. EAT-10, Eating Assessment Tool-10; RULM, Revised Upper Limb Module.

DISCUSSION

In this investigation we evaluated voice and deglutition using two PROs of bulbar function among non-ambulatory patients with SMA. Our results demonstrate: 1) weaker individuals with SMA experience high rates of perceived deficits in specific components of communication and deglutition and their relationship to one another; and 2) perceived deglutition is associated with other metrics of disease characteristics such as SMA type, current upper extremity function, and use of respiratory support.

These results are consistent with those of previous investigations that have reported high rates of bulbar impairment among patients with SMA that are correlated with disease severity [6, 10]. Specifically, work by Chen et al. (2012) reported 77% of individuals with SMA type 2 reported dysphagia, while only 23% of those with SMA type 3 reported such deficits. Investigators in this study found the best predictors of bulbar integrity, however, to not be in the broad SMA classifications but in the more refined measures of motor function [10]. It is notable that the primary perceived dysphagia symptoms identified within these previous investigations as well as our current findings were not in the individual’s ability to ingest liquids, but in their ability to ingest solids. This is in contrast to deficits in swallowing liquids that are frequently reported in other patient populations (e.g., stroke), and may reflect differences in underlying impairments. For example, it is plausible that in contrast to impairments in airway closure, patients with SMA more frequently suffer from deficits in bolus propulsion and mandibular movement required for effective mastication. Future investigations examining the correlation between these PROs including the EAT-10 and instrumental assessments are required to better understand how perceived deficits relate to true physiology.

We found age to be weakly correlated with perceived swallowing ability, suggesting that swallowing and mastication complaints become more common with increasing age for individuals with SMA [6, 9, 10, 12]. This is consistent with the known progression of muscle weakness, atrophy, and fatty infiltration of SMA muscles that mediate mouth opening (lateral pterygoid) and mastication [9, 52]. Postmortem examination in SMA has shown degeneration of motor neurons in the brainstem as the trigeminal motor nucleus is affected in patients with severe SMA, but less than lower cranial nerve nuclei [53], perhaps due to differences in motor unit size [52]. Alternatively, specific bulbar muscle groups (or specific motor neuron pools) may differ in vulnerability to SMN deficiency [52]. Progressive weakness of neck and bulbar muscles can impact temporomandibular joint morphology, head balance, and function leading to contracture and reduced mouth opening. These limitations have been shown to occur at a very young age (7 years); therefore, it is important to investigate clinically whether earlier therapeutic intervention might be more efficacious in maintaining function [54].

Interestingly, our participants reported low mean perceived swallowing deficits despite the severity of our SMA cohort as 34.0% had a gastrostomy tube. Clinically, older individuals with SMA often report only supplemental use of their gastrostomy tubes for their meals and primarily eat by mouth. The frequency and volume of gastrostomy intake would be helpful information when interpreting these results, but this information was not collected in our cohort. Also, of clinical significance are the reported deficits in communication that impeded the ability to be understood. Despite the fact that the majority (57.4%) of our participants required respiratory support, the reports of mild voice difficulties in the older, chronic SMA cohort queried in this investigation are consistent with the limited previous investigations examining this area of function; in one such investigation, 27% of individuals with SMA reported ‘a weak voice’ [6]. The ability to effectively communicate, i.e., able to be understood by an unknown listener in a variety of environments, requires integrity within the respiratory system to enable sufficient breath support for phonation, as well as the neuromuscular and skeletal system to enable rapid and precise movement of the articulators including the lips, tongue, and jaw. While the evaluation of all of these separate components of communication was beyond the scope of this investigation, it is likely that the aforementioned cranial nerve degradation in advanced SMA influences all of these sub-systems. Determining the sub-systems involved in these functional communication deficits is certainly a critical next step in establishing future optimal rehabilitation approaches.

Utilization of more refined communication assessments is also important in fully understanding the true physiologic impairment. It is notable that unlike PROs examining motor control, an individual’s perceived communication effectiveness is largely based on their ability to effectively transmit their intended message to their listener. Research in other populations who have profound communication deficits indicates it is common for these individuals to limit communication to close family members and friends most familiar with the communicator’s unique communication limitations, or to use these communication partners to translate their message to those communication partners less familiar with their deficits. In both cases, it is common for the communicator to not perceive a profound communication impairment as their communicative intent gets transmitted effectively [55], further reiterating the need to combine these functional outcomes with refined clinical assessments.

The evaluation of bulbar function in SMA is becoming an increasingly greater area of interest especially given the need to detect clinically meaningful response to changes in treatment with the growing availability of FDA approved therapies [31, 56]. Previous studies have included PROs or semi-structured interviews to describe bulbar dysfunction, although many focused-on infants and children with more severe SMA type 1 [4, 57–59]. A recent German study examined patient-reported treatment expectations for adult SMA individuals receiving nusinersen [34]. Although adults had high expectation of treatment effectiveness regarding increases in muscle strength and disease stabilization, only 4% expected an improvement in bulbar function in the type 2 population, although perhaps it was not a prominent impairment in their cohort. Encouragingly, 13% of individuals reported experiencing improvement in ventilatory and bulbar function, both mainly in SMA type 2, and no individuals reported deterioration of bulbar or respiratory function over a follow-up period of ten months [34]. This highlights the need to identify dedicated bulbar assessments, both physiologic and PRO, in SMA [4, 31] as the use of PROs is tremendously beneficial in elucidating the functional impacts of the disease burden on individuals [4]; however, the reliance on perception likely limits their sensitivity to some extent [6, 9, 13].

Our study was limited given the novelty and application of using these standardized PROs (often used for clinical dysarthria and dysphagia in other disease groups) to assess bulbar function in SMA. Neither PRO used in this study was shown to clearly assess all aspects and sub-systems involved in the perception of voice and deglutition deficits, and our results need to be considered with caution as limited data were available for pediatric SMA, and for those participants under age 18 years (n = 9), the Pediatric Eating Assessment Tool (PEDI-EAT-10) and Pediatric Voice Handicap Index (pVHI) for caregivers would have been more appropriate and informative. There are currently no longitudinal data available on either the VHI or EAT-10 in SMA, nor any investigations of validity, reliability, or minimal detectable changes, which will be necessary to help assess normal patient variability and determine clinically meaningful changes with treatment. The perceived impaired vocal communication issues our patients reported might reflect a complex combination of dysphonia, dysarthria, impaired stamina, and ventilatory insufficiency making it challenging to evaluate these impairments with a patient-reported assessment of voice. Further research is needed to identify relevant items to develop a SMA-specific PRO that yields a reliable, valid, responsive, and clinically relevant measure of bulbar function.

Perceived voice and swallowing issues in SMA are common yet may be underreported in the chronic SMA cohort. Our findings highlight the need for objective and quantitative measures of these aspects of this neurodegenerative disorder. Future studies of a larger sample may provide more insight on the ability of these PROs to measure a range of bulbar function, and a supplementary review of other currently available PROs may enable the development of a better disease-specific measure of bulbar function, which may be more sensitive to the effects of disease-modifying treatments for SMA. Multidisciplinary efforts are currently underway including collaboration of speech language pathologists, physical/occupational therapists, neurologists, and dentists with expertise in SMA to identify and refine assessment tools for relevance and evaluation of bulbar function in SMA [60]. The ideal tool should be standardized and aim to determine the presence, severity, and progression of dysarthria and dysphagia, and help to identify rehabilitation and management interventions, as well as quantify treatment effects[24, 61].