Brain, cognition, and language development in spinal muscular atrophy type 1: a scoping review
Abstract
Aim
To summarize the current knowledge on brain involvement in spinal muscular atrophy (SMA) type 1, focusing on brain pathology, cognition, and speech/language development.
Method
A scoping review was performed using the methodology of the Joanna Briggs Institute. Five databases and references from relevant articles were searched up to December 2019. Articles were screened on the basis of titles and abstracts. Full-text papers published in peer-reviewed journals in English were selected.
Results
Nineteen articles met eligibility criteria. Eight case series/reports on brain pathology showed abnormalities in few SMA type 0/1 cases, supported by findings in three post-mortem examinations in mice. Four studies (three case–control, one cross-sectional) on cognition reported contradictory results, with impaired cognitive performances in recent, small groups with SMA type 1. Four studies (three cross-sectional, one observational) on speech/language showed that untreated SMA type 1 patients rarely achieve functional and intelligible speech, with data limited to parent reports/non-formal evaluations.
Interpretation
Brain involvement is an under-investigated aspect of SMA type 1, requiring further exploration in longitudinal studies. A deeper knowledge of brain involvement would improve the interpretation of clinical phenotypes and the personalization of rehabilitation programmes supporting patients' autonomies and quality of life. Additionally, it may help to define further outcome measures testing the efficacy of current and new developing drugs on this domain.
What this paper adds
- Brain involvement is under-investigated in spinal muscular atrophy (SMA) type 1.
- Neuropathological data suggest progressive brain involvement in severe forms of SMA.
- Impaired cognitive performances are reported in small groups with SMA type 1.
- Data on language in those with SMA type 1 are limited to parent reports and non-formal assessments.
What this paper adds
- Brain involvement is under-investigated in spinal muscular atrophy (SMA) type 1.
- Neuropathological data suggest progressive brain involvement in severe forms of SMA.
- Impaired cognitive performances are reported in small groups with SMA type 1.
- Data on language in those with SMA type 1 are limited to parent reports and non-formal assessments.
Abbreviations
-
- JBI
-
- Joanna Briggs Institute
-
- SMA
-
- Spinal muscular atrophy
-
- SMN
-
- Survival motor neuron
Spinal muscular atrophy (SMA) encompasses a group of neuromuscular disorders characterized by degeneration of alpha motor neurons in the spinal cord with progressive muscle atrophy, weakness, and paralysis.1 The most common form of SMA is due to a defect in the survival motor neuron 1 (SMN1) gene located on chromosome 5q11.2–q13.3, resulting in insufficient SMN protein levels.2 However, humans have at least one copy of the highly homologous SMN2 gene producing a low amount of functional full-length SMN protein, which is sufficient to allow survival in the absence of full-length SMN from the SMN1 gene.2, 3 The incidence is 1 in 7000 to 10 000 live births and the carrier frequency is approximately 1 in 50.4 The disease presents a wide range of phenotypes that are classified into five clinical groups (types 0–4) depending on age at onset and maximum motor milestone achieved, with types 0 and 1 being the most severe. SMA type 0 is an extremely severe prenatal/congenital form, with reduced fetal movements, congenital contractures, and early respiratory failure. SMA type 1, also called Werdnig–Hoffman disease, presents shortly after birth and before 6 months of age with inability to achieve independent sitting and limited life expectancy (high mortality rate by 2y). This form is further classified into three subgroups according to the age at symptom onset: within the first 2 weeks of life (type 1a), by 3 months of age (type 1b), and between 3 and 6 months of age (type 1c). Overall, SMA type 1 accounts for 60% of all patients with SMA and is the most common genetic cause of death in infants. Published literature on SMA type 1 has mainly focused on survival and respiratory, bulbar, and motor function,5-7 while less attention has been paid to other features of the disease, including brain involvement. However, clinical practice shows that some patients with SMA type 1 may show cognitive impairment, and most children untreated for SMA type 1 never achieve functional verbal skills.8
Over the past few years, the natural history of the disease – in particular of the type 1 form – has radically changed thanks to the availability of new pharmacological treatments. Nusinersen (Spinraza), the first SMN-modulating treatment targeting the RNA splicing of the SMN2 gene, resulted in prolonged survival and improved motor function in clinical trials9-12 and was approved by the US Food and Drug Administration and the European Medicine Agency in 2016 and 2017, respectively. Onasemnogene abeparvovec-xioi (Zolgensma), the first gene replacement therapy for this disease, has also shown positive results in clinical trials in patients with SMA type 1,13-15 and was approved by the US Food and Drug Administration in 2019 and more recently by the European Medicine Agency. Other drugs are at a very advanced stage of clinical development, including the orally administered SMN2 splicing modifier risdiplam,16 recently approved by the US Food and Drug Administration, and represent promising additional pharmacological options for SMA.
With the increasing number of long-term survivors of SMA type 1 worldwide, it has become obvious that treated children show new phenotypes, presenting changes not only in respiratory, bulbar, and motor function, but also in other areas of functioning, including cognition and speech and language development. A better understanding of the characteristics and extent of brain involvement in SMA type 1 is crucial for a deeper comprehension of the clinical features of the disease and for personalized patient management.
In this scoping review, we first aim to explore and summarize the current knowledge of brain involvement in SMA type 1, analysing the domains of brain pathology, cognition, and speech/language development through an extensive search of the published literature. Second, by providing up-to-date information on structural and functional brain involvement in SMA type 1, we aim to pave the way for future research focusing on these domains. We believe that understanding the nature and degree of brain involvement in this disease may help to better characterize the newly emerging phenotypes of treated patients. This, in turn, would shed light on aspects that are still unclear, including the impact of new pharmacological treatments on cognitive and speech/language functions and the role of recovered motor abilities on brain development. These aspects may become increasingly important in the future when choosing between different drugs and planning personalized rehabilitation programmes, particularly in children with SMA type 1, who may present with a developmental disorder in addition to the neuromuscular disease.
Method
The Joanna Briggs Institute (JBI) methodology for scoping reviews described in the online JBI reviewer's manual17 was used to conduct the review. Results are presented following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR) checklist.18
No a priori protocol was registered. Further information on the process can be obtained from the corresponding author on request.
Inclusion criteria
The inclusion criteria used to select the articles for the review were based on the population, concept, and context elements reported below.
Population
We included articles addressed to the 5q11.2–q13.3 SMA, excluding all other forms of SMA. Afterwards, only papers on 5q SMA type 1 were analysed, according to the aim of the review. The only exception was the brain pathology domain, where we also discussed the findings from other 5q SMA subtypes (0 and 2). The reason was that the biological mechanisms underlying brain pathology could be considered similar in all forms of 5q SMA, although with different degrees of severity.19-21
Concept
We selected articles analysing the following concepts: ‘brain pathology’, for which we included pathological and imaging studies on both humans and animal models; and ‘cognition’ and ‘speech/language development’, for which we included studies testing cognitive functions as well as receptive and expressive communication skills with any tests for the paediatric population (both validated and non-validated in large paediatric cohorts).
Context
No cultural, geographical, ethnicity, or sex-specific limits were considered for our review, the reason being the equal presentation of the disease in these categories.
Search strategy
The review covered data available up to December 2019. Selected keywords were combined to create search strategies, adjusted for each screened database. Articles were searched in the following databases: PubMed/MEDLINE (1950–2019), Scopus (1996–2019), Embase (1980–2019), PsycINFO (1806–2019), and Web of Science (1990–2019). Search terms included ‘spinal muscular atrophy’, ‘brain’, ‘magnetic resonance imaging’, ‘central nervous system’, ‘cognition’, ‘intellectual disability’, ‘speech’, ‘language’, ‘communication’, ‘augmentative alternative communication’, ‘attention’, ‘executive functions’, ‘working memory’, and ‘neuropsychology’. Table S1 (online supporting information) shows the search process used to retrieve the final articles discussed in the review. Search strategies and search terms are reported for PubMed/MEDLINE, Scopus, Embase, and PsycINFO. References from relevant articles were searched for inclusion of additional papers not previously identified through the systematic search.
Study screening and selection
Articles were initially screened on the basis of titles and abstracts according to the population, concept, and context elements previously described (data on central nervous system [CNS] morphology and functions in SMA type 1). Duplicates were removed. Only full-text papers published in peer-reviewed journals and in English were selected. The articles were examined by two authors (RM and CB), and eligibility for inclusion was performed independently; in case of discordant opinion between the reviewers, the eligibility of the article was discussed until consensus was reached.
Level of evidence and qualitative analysis of eligible articles
The strength of evidence for each article was assessed according to the levels of evidence developed by the JBI.22
In addition, the methodological quality of papers focusing on cognitive and speech/language development was assessed following amended criteria from Cross and Hare23 reported by Pearson et al.24 The articles were rated from 0 to 2 in six areas: control group; sample size; recruitment; syndrome diagnosis; methodology; and appropriate statistics/comparisons. Table S2 (online supporting information) shows the scoring criteria used to assess the methodological quality of eligible articles. A total score was obtained for each article, with papers scoring in the upper tertile of possible scores (9+) deemed to be of reasonable methodological quality. The qualitative assessment was performed by two authors (RM and CB) independently; in case of discordant opinion between the reviewers, the scoring was discussed until consensus was reached.
Extraction and presentation of results
All data relevant to inform the scoping review objectives and questions were extracted and are summarized in Tables 1 to 4 (Tables 3 and 4 also report the quality assessment of articles focusing on cognitive and speech/language development). Results were grouped according to the domains explored: brain neuropathology, brain neuroimaging, cognition, and speech/language development.
| Reference | Study design | Sample (size; SMA type; age range) | Assessment | Results | Level of evidence22 |
|---|---|---|---|---|---|
| Towfighi et al.26 | Case series | n=4; SMA type 1; 9d–8.5mo | Autopsy examination | Neuronal ballooning, chromatolysis, degeneration and neuronophagia in ventral thalamic nuclei, primary spinal sensory neurons | 4.b |
| Devriendt et al.25 | Case report | n=1; SMA type 0; 25d | Autopsy examination | Severe neurodegenerative changes including ballooned neurons and neuronophagia in brainstem nuclei, thalami, cerebellum, dorsal root ganglia | 4.b |
| Araki et al.28 | Case series | n=2; SMA type 2; 5y and 37y | Autopsy examination | Reduction in the number of the Betz cells in the precentral gyrus and reduction of the large myelinated fibres in the spinal–pyramidal tract | 4.b |
| Harding et al.27 | Case series | n=5; SMA type 1; 18d–10y | Autopsy examination | Neuronal degeneration with ballooning and chromatolysis in cerebral cortex, hippocampus, basal ganglia, thalami, brainstem, pigmented nuclei, cerebellum, dorsal root ganglia | 4.b |
| Wishart et al.29 | Bench research (mouse model) | Smn−/−; SMN2 mice | Post-mortem examination | Decreased cell density, reduced cell proliferation and impaired neurogenesis in primary motor cortex, hippocampus | 5.c |
| Liu et al.30 | Bench research (mouse model) | Smn−/−; SMN2 mouse embryos | Post-mortem examination | Dramatic increase in cell death in the developing telencephalon at both the dorsal and ventral sides around the lateral ventricle | 5.c |
| d’Errico et al.31 | Bench research (mouse model) | Smn−/−;SMN2+/+;SMNΔ7+/+ mice | Post-mortem examination | Selective decrease in the number of large layer V pyramidal neurons in the motor cortex | 5.c |
| Reference | Study design | Sample (size; SMA type; age range) | Assessment | Results | Level of evidence22 |
|---|---|---|---|---|---|
| Yohannan et al.32 | Case series | n=8; SMA type 1; 21d–15mo | Baseline brain computed tomography | Generalized cerebral cortical atrophy, low attenuated non-enhancing areas in the white matter of both frontal lobes | 4.b |
| Ito et al.33 | Case report | n=1; SMA type 1; 6y | Baseline brain MRI | High signal intensity lesions in the anterolateral portions of the bilateral thalami | 4.b |
| Mendonça et al.34 | Case series | n=3; SMA type 0; first assessment: by 2mo, follow-up assessment: at 11mo, 1y, and 3y respectively | Longitudinal brain MRI |
First assessment: supratentorial atrophy of subcortical predominance, tapered corpus callosum and widening of sulcus and ventricles Follow-up assessment: severe reduction of the white matter (3/3), severe hippocampal atrophy (2/3); symmetrical signal abnormalities in the putamen and thalamus—lateral and pulvinar (2/3), with additional atrophy of caudate (1/3); marked ventricular dilatation (1/3) |
4.b |
| Maeda et al.35 | Case report | n=1; SMA type 0; first assessment: neonatal, second assessment: 7mo, third assessment: 2y | Longitudinal brain MRI |
First assessment: no findings suggestive of hypoxic–ischemic encephalopathy Second assessment: atrophy of the cerebral cortex, subcortical white matter, thalamus and basal ganglia Third assessment: additional reduction of volumes of cervical cord, brainstem and cerebellum |
4.b |
- MRI, magnetic resonance imaging.
| Reference | Study design | Sample (size; SMA type; age range) | Assessment | Results | Level of evidence22 | Quality score23, 24 |
|---|---|---|---|---|---|---|
| Billard et al.36 | Case–control study |
n=11; SMA type 1 and 2; 8y 3mo–13y 6mo n=21; DMD; 8y 6mo–13y 6mo n=42; typically developing comparison group; 6y 3mo–14y 0mo |
WISC-R subtests; Batterie d’Evaluation du Langage, Test de vocabulaire actif e passif, North Syntax screening test; reading and processing tests: le pipe et le rat, single-word lists reading | Verbal IQ, verbal memory, language and reading deficits in patients with DMD compared with SMA and comparison group | 4.b | 6/12 |
| Von Gontard et al.37 | Case–control study |
n=96; SMA (18 type 1); 6y 0mo–18y 11mo n=45; non-affected siblings; age-matched n=59; typically developing comparison group; age-matched |
Raven coloured and standard progressive matrices, subtests of the Kaufman assessment battery for children and of the WISC | Similar results in SMA and control groups, no differences according to SMA type; environmentally mediated aspects of intelligence higher in adolescents with SMA | 4.b | 9/12 |
| Chung et al.38 | Cross-sectional study |
n=83; SMA (22 type 1); age not specified No comparison group |
Interview-based questionnaire for caregivers: Functional Independence Measure for Children – Chinese version | Overall performance of children with SMA below normal, with the worst performances in patients having type 1 SMA | 4.b | 6/12 |
| Polido et al.39 | Case–control study |
n=12; SMA type 1; 6y 0mo±2y 4mo n=12; typically developing comparison group; age-matched |
Four pair-matching tasks (assessed with an eye-tracking device): task 1, matching objects, animals, and fruits; task 2, matching figures and colours; task 3, matching letters; task 4, matching numbers | Poorer performances in patients with SMA than in comparison group in terms of correct answer rate and time of performance in all tasks | 4.b | 9/12 |
- DMD, Duchenne muscular dystrophy; WISC-R, Wechsler Intelligence Scale for Children – Revised.
| Reference | Study design | Sample (size; SMA type; age range) | Assessment | Results | Level of evidence22 | Quality score23, 24 |
|---|---|---|---|---|---|---|
| Hoshi et al.40 | Cross-sectional study | n=36; SMA type 1; 11mo–15y | Postal questionnaire for parents/carers on language development milestones classified into three main items: communication skills using devices; communication skills with no devices; communication methods used at the time of the assessment | The ability to communicate with no devices (eye fixation, vocalization, signs) is delayed in SMA type 1 compared with typically developing children, but it is acquired earlier than the ability to communicate using electronic or non-electronic devices | 4.b | 4/12 |
| Pane et al.8 | Cross-sectional study | n=122; SMA type 1; 3mo–266mo | Clinical observation of the ‘ability to produce at least short sentences that could be understood by the examiner and not just by the carers’ (reported as yes/no) | 34/122 (28%) had acquired comprehensible speech | 4.b | 7/12 |
| Ball et al.41 | Cross-sectional study | n=32; SMA type 1; 6mo–30y | Researcher-developed online survey for parents/carers on interaction characteristics associated with the following: communication using natural speech; communication methods used; issues with speech-generating devices |
Speech deficits related to clarity, independence, intelligibility; methods used: speech (n=19), gesture (n=22), speech-generating device (n=12), no-tech picture or symbol board/book (n=5), all methods (n=3) Several obstacles to the acquisition and implementation of speech-generating devices |
4.b | 4/12 |
| Al-Zaidy et al.14 | Observational study |
n=12; SMA type 1; 0.9mo–7.9mo at baseline No comparison group |
Clinical observation of the ‘ability to speak’ at baseline and 2y after (reported as yes/no) | 11/12 (92%) were able to speak by the end of the 2y follow-up period | 3.e | 4/12 |
Results
Search results
A total of 19 articles were included in the review after study screening and selection: 11 focusing on brain pathology, four on cognition, and four on speech/language development, as shown in Figure S1 (online supporting information). All but one of these papers included patients untreated for SMA type 1.
Brain pathology
We identified 11 different articles on brain involvement in SMA: seven on neuropathological examinations (three out of seven performed in mouse models), and four on neuroimaging findings in patients with SMA.
Neuropathological studies: patients
Specific neuropathological abnormalities have been reported in several CNS areas other than lower motor neurons from autoptic examinations performed in patients with both the type 025 and the type 126, 27 forms of the disease. Articles published before 1980 were already summarized by Towfighi et al.26 and they were not counted individually as part of the results of this review. The involved CNS structures included brainstem nuclei, pigmented nuclei, thalami, basal ganglia, frontal and temporal cortices, hippocampi, and cerebellum. In addition, few papers documented different degrees of involvement according to disease severity. Brain samples from patients with different subtypes of SMA type 1 showed milder findings in patients with the less severe forms of SMA type 1 (1b and 1c) compared with patients with the most severe form of the disease (type 1a).27 These neuropathological findings were considered primarily related to the underlying condition by the authors. In addition, neuropathological examinations of patients with a clinically and genetically confirmed diagnosis of SMA type 2 showed no neuronal changes in the areas previously described as affected in patients with SMA type 0/1 (e.g. brainstem nuclei, pigmented nuclei, thalami, basal ganglia, hippocampi, cerebellum).28 Changes in areas such as the precentral gyrus and the large myelinated fibres in the spinal–pyramidal tract were reported instead. The overall level of evidence about the presence of neuropathological abnormalities in humans was 4.b according to the JBI criteria.22
Neuropathological studies: mouse models
Changes in brain morphology were reported in a severely affected SMA mouse model.29 Size reduction was observed in areas normally associated with higher SMN protein levels in the healthy postnatal brain – especially the hippocampus – with more modest morphological reductions in the primary motor cortex. According to the authors, these data show for the first time that high levels of SMN protein are required for typical brain development in vivo and, as a result, reduced expression of SMN protein causes atypical brain development, particularly affecting regions such as the hippocampus. Neuropathological abnormalities in the developing telencephalon and in the motor cortex were also reported in two other SMA mouse model studies.30, 31 The overall level of evidence about the presence of neuropathological abnormalities in mouse models was 5.c according to the JBI criteria.22 Details of each article are reported in Table 1.
Neuroimaging studies
Alterations of several CNS areas other than alpha motor neurons in the spinal cord have been demonstrated both by brain computed tomography (CT)32 and by brain magnetic resonance imaging (MRI)33-35 studies. The largest case series reported results of CT scans performed in eight children with a clinical diagnosis of SMA type 1.32 Images showed generalized cerebral cortical atrophy in all but one patient who presented mild abnormalities in the white matter of both frontal lobes (the authors could not exclude chronic hypoxic-ischemic brain injury as a cause of the described abnormalities). All subsequent neuroimaging case series/reports presented brain MRIs performed in patients with both SMA type 133 and type 034, 35 with no documented history of hypoxic-ischemic events. A first case report of a patient with SMA type 1 showed thalamic abnormalities in the anterolateral portions.33 Very recently, marked progressive CNS alterations have been documented in a longitudinal neuroimaging case series in three patients with SMA type 0.34 Predominant brain MRI findings at the first scan were supratentorial atrophy of subcortical predominance, tapered corpus callosum, and widening of sulcus and ventricles; imaging follow-up showed a marked progression of the supratentorial brain atrophy in all patients, characterized by severe reduction of white matter (three out of three), severe hippocampal atrophy (two out of three) and relative sparing of the cerebellum. Symmetrical signal abnormalities were detected in the putamen and thalamus (lateral and pulvinar) in two patients, with additional atrophy of caudate in the third. Marked ventricular dilatation was also detected in the third patient. Finally, a progressive atrophy of the cerebral cortex, subcortical white matter, thalami, basal ganglia, brainstem, and cerebellum accompanied by epilepsy was reported at follow-up MRI scans in a very severely affected patient with SMA type 0.35 The overall level of evidence about the presence of neuroimaging abnormalities was 4.b according to the JBI criteria.22 Details of each article are reported in Table 2.
Cognition
According to the review inclusion criteria, only four papers on cognitive performances in type 1 SMA patients were found, and the available data seem contradictory.
The comparison of a group of patients with SMA (types 1 and 2, age 8–13y) with those having Duchenne muscular dystrophy and typically developing children in verbal and non-verbal intelligence (Wechsler Intelligence Scale for Children – Revised), language (Batterie d’Evaluation du Langage; Test de vocabulaire actif e passif; North Syntax screening test), and reading (le pipe et le rat; single-word list reading) showed lower performances in the Duchenne muscular dystrophy group (verbal IQ, language, and reading) than in patients with SMA.36 Overall, the SMA group was described as having similar performances to typically developing individuals, but a direct statistical comparison was not made between these two groups in the study. Another study of cognitive skills in 96 children with SMA (18 with type 1, age 6–18y), assessed through the Raven Coloured and Standard Progressive Matrices, the Kaufman Assessment Battery for Children, and the Wechsler Intelligence Scale for Children, showed very similar results between the SMA and comparison groups (non-affected siblings and typically developing peers), with no differences between the types of SMA.37 It should be noted that patients classified in SMA type 1 in both of these studies were probably not representative of this form of disease, as only some of them were able to sit and the age was higher than expected by natural history. Furthermore, the adolescent patients described in the second study seemed to perform better than a typically developing comparison group on the verbal IQ subscale of Weschler tests and Raven standard progressive matrix, suggesting that the environmentally mediated aspects of intelligence were higher in patients with SMA than in the comparison groups. The authors speculated that this was a compensatory effect for the restrictions resulting from physical disability.
On the other hand, the description of the functional status of 83 patients with SMA (22 type 1) through the WeeFIM (an interview-based questionnaire for caregivers) showed that the overall performance of children with SMA was below normal, with the worst performance in patients with SMA type 1, who required help or assistance in most of the cognitive functioning domains, particularly in expression, social interaction, and problem solving.38 Furthermore, even the most recent study on cognition39 confirmed a poorer performance in patients with SMA than in a comparison group. It was specifically addressed to patients with SMA type 1 (age 3–9y), compared with a sex- and age-matched typically developing comparison group. Cognitive assessment was completed with pair-matching tasks, consisting of verbal requests to match objects, figures and colours, letters (upper and lower cases), numbers (Arabic and words), by using an eye tracker device. Patients with SMA performed worse than the comparison group in correct answer rate and time performance in all tasks.
The overall level of evidence about cognitive impairment was 4.b according to the JBI criteria.22 Details of each article are reported in Table 3.
Speech and language development
Four articles were found on speech and language development in patients with SMA type 1.
Two cross-sectional studies provided information on patients’ communication abilities on the basis of researcher-developed questionnaires completed by parents/carers (a postal questionnaire40 and an online survey via e-mail invitation41 respectively). Both surveys focused on questions related to the communication methods used by the child, including electronic and non-electronic communication devices. In the first study, the authors observed that the five communication methods most frequently used were signs (e.g. eye movements indicating ‘yes’ or ‘no’, 50%), eye fixation (47%), electronic communication devices (47%), vocalizing (30%), and non-electronic communication devices (22%). No communication methods were used by 19.4%. The acquisition of the ability to communicate without using devices, such as through eye fixation, vocalization, or use of signs was delayed compared with typically developing peers. In this population, children showed the ability to communicate using electronic devices from the age of 3 years and non-electronic devices from the age of 4 years. This is the first study providing insight on the communication skills of children with SMA type 1 according to age and use of devices. In the second study, variability in the communication methods used was reported, including speech (59%), gesture (69%), speech-generating device (38%), and no-tech picture or symbol board/book (16%). Three parents reported that all four methods were used by their children (9%). In patients able to speak, speech deficits were reported and were related to clarity, independence, and intelligibility. Among children with no functional natural speech, 22% used gesture only. Results also illustrated parental perceptions of greater receptive than expressive language abilities. The benefit of implementing speech-generating devices in terms of communication effectiveness and quality of life was highlighted, although several obstacles to their acquisition and implementation were reported.
Two further studies provided information on patients' communication abilities on the basis of clinicians' observations. In one study,8 the authors reported functional abilities and other clinical findings – including speech – in a cohort of 122 untreated children with SMA type 1. In their cross-sectional analysis only 28% had acquired comprehensible speech, defined as the ability to produce at least short sentences that could be understood by the examiner and not just by the carers. These were mainly patients with SMA type 1c who were at the milder end of the spectrum. The last study14 was the only one providing information about patients treated for SMA type 1, in particular on health outcomes – including the acquisition of the ability to speak – in 12 children with SMA type 1 treated with gene replacement therapy. It showed an increase to 92% of patients able to speak by the end of the 2-year follow-up period.
The overall level of evidence about speech/language impairment was 4.b in all but one study (level 3.e) according to the JBI criteria.22 Details of each article are reported in Table 4.
Discussion
This review suggests that a comprehensive understanding of brain involvement in SMA type 1 is still lacking. One of the reasons for this limited knowledge is the difficulty of assessing cerebral functions and other cognitive abilities in severely affected patients. Severe muscle weakness as well as respiratory and bulbar dysfunction significantly limit the interactions with the environment. Augmentative and alternative communication and the use of eye-tracking devices have been recommended for patients with SMA, but there are some limitations in the use of a non-physiological way of communication to study cognitive and verbal skills compared with typically developing peers.
Furthermore, although the expression of the SMN protein throughout the CNS has been known since the late 1990s,19, 20 neuropathological studies in SMA type 1 are sparse. The study of brain involvement through autopsies of very young infants was probably slowed down by ethical considerations, given the apparent secondary interest of the topic in such a devastating neuromuscular disorder.42, 43 Neuroimaging studies in this population are very limited as well.
Available data show that brain structures can be primarily affected in the severe forms of SMA. This was originally documented before the identification of the SMN1 gene, on the basis of neuropathological findings at post-mortem examinations.26 Subsequent studies in patients with a genetically confirmed diagnosis of 5q SMA clearly demonstrated brain involvement,25, 27, 28, 33-35 especially in the most severe forms (types 1a or 0). Several structures, including thalami, basal ganglia, temporal and frontal cortices, hippocampi, and cerebellum have been reported to be variably affected. Interestingly, a marked progression of initial brain abnormalities has been documented by follow-up brain scanning.34, 35 This is something to be considered and further investigated, especially now that treated patients are surviving longer. Experimental evidence from mouse models – although very limited29-31 – further supports a possible primary brain involvement in patients with SMA type 1. Of note, neuropathological alterations in other nervous system areas such as the primary spinal sensory neurons of dorsal root ganglia were reported as well, both in human25-27 and mouse model44, 45 studies. Although the study of the peripheral nervous system is beyond the scope of this review, it should be noted that sensory inputs may play a role in the overall brain functioning of these patients.
The clinical correlates of these findings in terms of cognitive and other neuropsychological functions including speech and language are still largely unclear, and no study has investigated the correlation between neuropathology/neuroimaging reports and cognitive and neuropsychological functioning. As already mentioned, this was mainly due to the clinical severity of the disease and the high mortality rate by 2 years of age of patients untreated for SMA type 1, who were often not able to provide verbal or gestural responses during the assessments. Most of the studies focusing on the intellectual abilities in SMA were initially performed in less severely affected patients (SMA types 2 and 3), who could be more easily assessed using the available validated tests and assessments. These studies showed that patients with SMA types 2 and 3 obtained normal to higher than normal IQ and speech/language abilities scores compared with their peers.46 Average vocabulary scores and above-average early grammar scores,46, 47 as well as significantly higher levels of lexical and semantic development48 and rich spatial language abilities, were reported, despite children with SMA type 2, by definition, never experiencing locomotion abilities.49
More recent studies38, 39 specifically investigating cognitive abilities in patients with SMA type 1, also using adapted assessments,39 showed that they have poorer performances than their peers, particularly in the attention and executive function domains. Speech and language development is also affected, with published data showing that functional and intelligible speech is rarely achieved in children untreated for SMA type 1.12, 40, 41 Considering that the limited interaction with the environment owing to poor expressive communication skills has been shown to further impact on cognitive development in several neurodevelopmental disorders,50 alternative and augmentative ways to communicate are recommended from early in life.
The advances in multidisciplinary care and the recent clinical implementation of SMN-modulating treatments have contributed to prolonging event-free survival and improving motor function in treated patients. However, it is still unclear whether other areas of functioning, including cognitive development and the achievement of effective speech and language abilities, which can significantly impact on independence and quality of life, may equally benefit from treatments. In fact, there are no published studies directly addressing cognition and language in children treated for SMA type 1. The report by Al-Zaidy et al.14 is the only study providing information on the effect of a treatment (gene replacement therapy) on one of the domains considered in our review (speech/language), simply reporting that the vast majority of patients were ‘able to speak’ 2 years after receiving the treatment. Additionally, the role of bulbar function in the development of articulate speech abilities is also unclear. Whether children with SMA type 1 treated with the new SMN-modulating treatments will recover bulbar function and improve speech, having the possibility to experience and develop their pre-verbal and verbal social and communication skills as typically developing children do, or whether residual structural and/or functional brain pathology will affect achievement of mature language abilities, will need to be clarified.
This scoping review provides very preliminary information about this, and the included articles present some limitations. One of the main limitations is the methodological quality of papers focusing on cognitive and speech/language development, with most articles presenting scores below the threshold for a reasonable methodological quality.8, 14, 36, 38, 40, 41 This is due to the lack of comparison groups, the very limited sample sizes, and the lack of appropriate statistical data analyses. Other weaknesses are the lack of validated and/or standardized measures, with the necessary use of information reported by parents/carers or by clinicians, and the difficulties of assessing severely disabled children with alternative methods (e.g. eye tracker) and of measuring these results against the comparison groups.
Longitudinal multicentric studies with standardized assessments of cognitive and speech/language abilities are required in children with SMA type 1. The use of adapted methodology to perform these assessments should be considered. Moreover, longitudinal neuroimaging studies performed alongside the clinical assessments would provide additional valuable information on the new emerging clinical phenotypes in this population, especially in patients at the most severe end of the spectrum.
Data collected from these studies would also provide a significant contribution to the growing evidence that SMA may be a multi-system disorder.51 The SMN protein is ubiquitously expressed throughout the body and is involved in several physiological mechanisms, including the assembly of spliceosomal small nuclear ribonucleoproteins,52 critical to RNA splicing of multiple genes. One of the most intriguing features in the pathogenesis of SMA is the selective vulnerability of motor neurons. However, several studies have shown that various cell types other than motor neurons can be affected in SMA,53 particularly in the most severe early-onset forms. There may be differential thresholds between different cell types, with spinal motor neurons being the most sensitive to a reduction in SMN protein expression.19 Several gene modifiers have been demonstrated to affect disease severity in SMA, with the number of copies of the SMN2 gene playing a major role. Patients with only one copy of the SMN2 gene and the lowest levels of SMN protein compatible with survival have shown the highest degree of brain involvement, as well as of other organ dysfunctions. Additionally, SMN protein levels have been shown to decrease during development not only in the spinal cord but also in the brain, with a substantial decline between fetal and postnatal stages, reaching very low levels after 3 months of age.54 This emphasizes the crucial role of the SMN protein during the early stages of brain development, and further supports the need for longitudinal standardized assessments of cerebral structures and functions in children with SMA type 1.
Conclusion
In conclusion, brain pathology, cognition, and speech/language development are under-investigated aspects of SMA type 1. In the literature there is some limited evidence of potentially progressive brain involvement in the severe forms of the disease. Impaired cognitive performances are reported in recent small SMA type 1 cohorts, while data on speech/language development in SMA type 1 are limited to parent-reported information or non-formal evaluations. Future longitudinal studies focusing on standardized assessments of cognitive and speech/language development in SMA type 1 are required, as well as longitudinal neuroimaging evaluations performed alongside. Data obtained will contribute to a better knowledge on new emerging phenotypes in treated patients with SMA type 1, and will guide more accurate, personalized rehabilitation programmes, thus supporting patients’ emerging abilities, autonomies, and quality of life. The information gathered will also help to define additional outcome measures to test the efficacy of current and new developing drugs for SMA.
Acknowledgements
No funding was received. RM is an investigator in SMA trials sponsored by AveXis, Roche, and Novartis and has received consultancy fees from Roche, AveXis, Biogen. CB has no competing interest to declare. MS has received compensation for participation at symposia and scientific advisory boards from Roche, Biogen, and Avexis and is an investigator for trials Sponsored by AveXis, Roche, and Biogen. GB has been an investigator in SMA trials sponsored by AveXis, Roche, and Novartis and has received compensation for participation at symposia and scientific advisory boards from Roche and AveXis.
Open Research
Data availability statement
Data sharing is not applicable to this article as no new data were created or analyzed in this study.
