Spinal presentations in children with type 1 spinal muscular atrophy on nusinersen treatment across the SMA-REACH UK network: a retrospective national observational study

Abstract

Background

Prior to the introduction of disease-modifying treatments (DMTs), children with type 1 spinal muscular atrophy (SMA) typically did not survive beyond the age of 2 years; management was mainly palliative. Novel therapies have made this a treatable condition, resulting in increased life expectancy and more time spent upright. Survival and improved function mean spinal asymmetry is a new complication with limited data on its prevalence and severity and no current guidelines on management and treatment. This study aimed to evaluate the spinal presentation and management of type 1 SMA children on nusinersen across the SMA-REACH UK network.

Methods

Spinal presentation and management of 80 children (age range 4 months–14 years, median 4 years 2 months) with type 1 SMA on nusinersen across the SMA-REACH UK network were reviewed through retrospective data analysis.

Results

There were 60 type 1 children who developed a spinal asymmetry, of which 40 had kyphosis and 50 used a supportive thoraco-lumbar-sacral orthosis (TLSO). TLSOs were predominantly a one-piece jacket with abdominal hole, advised to be worn when upright during the day. Reduced neck range of movement was found in 33, 1 of these had plagiocephaly and 5 had torticollis. Of those with reduced neck range of movement, 26 (79%) had spinal asymmetry. Spinal surgery was performed in 7.

Conclusions

Our study confirms high prevalence of spinal asymmetry in this cohort, requiring long-term management planning. It provides information on presentation and treatment options, facilitating development of guidelines for these new complications observed in children surviving longer with DMTs.

STRENGTHS AND LIMITATIONS OF THIS STUDY.

• There is good representation across the UK and collection involved a holistic view of patients with a range of outcomes.

• This is not longitudinal data and provides a snapshot only. There are no robust objective measurements (eg, X-ray) due to lack of availability.

Background

Spinal muscular atrophy (SMA) is a rare autosomal recessive genetic disorder with an estimated incidence rate of 1 in 11 000 live births.¹ SMA is caused by mutations in the survival motor neuron (SMN) 1 gene resulting in a deficiency of the SMN protein. This causes degeneration of the spinal motor neurons leading to progressive generalised muscle weakness.²,⁵ A homologous SMN2 gene is present in all SMA patients but physiological processing, namely alternative splicing, results in the exclusion of exon 7 from the mature mRNA, leading to the inability to produce sufficient SMN protein. The copy number of the SMN2 gene can vary in patients and its number is grossly correlated with the SMA severity. Based on the age at symptom onset and maximum motor ability achieved, SMA is classified into three main paediatric subtypes. Type 1 SMA is the most severe form and accounts for 50–60% of incident cases.⁶ Affected children are unable to sit and muscle weakness can cause further secondary complications including spinal asymmetry, joint contractures and difficulties with swallowing and breathing.⁴ The traditional SMA classification divides type 1 SMA in further subtypes, a, b and c. Type 1a is the most severe with symptoms presenting at birth to 2 weeks. In type 1b, the generalised weakness and hypotonia occur by 3 months with poor head control, and in type 1c, from 3 to 6 months with better head control.⁷ Children with type 2 SMA are milder, presenting with symptoms before 18 months, and children are able to sit but not walk. Type 3 SMA children are able to walk and symptoms are detected after 18 months old.

Prior to the recent introduction of novel disease-modifying treatments (DMTs), children with type 1 SMA typically experienced progressive muscle weakness; management was mainly palliative, and historically, death usually occurred by the age of 2 years (with a median age of 13.5 months).^{8 9} Recent DMTs have made type 1 SMA a treatable condition requiring a proactive approach and a change in management to meet the evolving clinical complications.

Nusinersen is an antisense oligonucleotide designed to increase the amount of functional SMN protein by altering splicing of SMN2 pre-mRNA. Nusinersen has met the primary endpoints of multiple randomised placebo-controlled clinical trials in type 1 SMA including Endear and SHINE (NCT02594124). It has received Food and Drug Association and European Medicines Agency approval and is available in the UK for patients with type 1 SMA as well as milder variants via a managed access agreement.⁹,¹² Nusinersen has been proven to be safe for patients with SMA.^{8 11 13 14}

Since the introduction of nusinersen, the natural history of children with type 1 SMA has significantly changed. Treated children survive longer, achieve previously unattainable milestones and maintain these acquisitions, in striking contrast to the natural history of type 1 SMA prior to DMTs.¹⁵,¹⁸ A number of these treated children can now attend nursery and school. As a result of their improved strength, these children are spending more time in an upright position (supported by wheelchairs/seating), with some achieving head control and independent sitting ability. With increased life expectancy and improved function, spinal asymmetry including scoliosis and kyphosis has been noticed as a complication which was not managed before as the children did not survive. Scoliosis is defined as a curvature of the spine in the frontal/coronal plane with a Cobb angle of >10°. Spinal asymmetry can be a clinical observation of this which does not require a radiographical finding. Kyphosis is an excessive curvature of the thoracic region in the sagittal plane.

Neuromuscular scoliosis (spinal asymmetry) is a common issue in SMA. The incidence of scoliosis varies between 60 and 90% in children with type 1 and 2 SMA^{7 19} and is approaching 100% in children with the most severe SMA forms.²⁰,²² These figures were reported prior to the introduction of DMTs so they are not representative of the current evolving clinical condition of children with type 1 SMA. It is not yet fully known if improved motor function will result in increased spinal asymmetry. Scoliosis is a common feature and almost invariably requires surgery in the natural history of type 2 SMA,⁶ and it is not known if the scoliosis trajectory of treated type 1 SMA will be similar to or different from type 2 SMA.

Spinal surgery has not previously been routinely considered in type 1 SMA other than in milder cases, based on clinical experience due to the limited survival of these children and their severe respiratory compromise. However, this is changing following the introduction of DMTs.

Currently, there is limited consensus among physiotherapists, spinal surgeons and neurologists on how to treat these children with regard to maintaining or achieving symmetrical and functional posture. The presentation and management are changing due to the DMTs and this is reflected in an increasingly complex cohort. As nusinersen is a relatively new treatment in the UK, there are limited data in this area. This study aimed to gather information on the prevalence and severity of spinal asymmetries, and the current management strategies for spinal posture and asymmetries in children with type 1 SMA receiving nusinersen.

Methods

Identifying patient cohort

SMA patients with genetic confirmation of SMN1 gene mutations and clinical diagnosis of type 1 SMA were included in the study. This was registered as a service evaluation under Great Ormond Street Hospital (Audit Number: 2840) and covered by the national ethics REC (SMA-REACH UK ethics REC IRAS 122521). This was a retrospective national observational study.

The neuromuscular centres which are part of the SMA-REACH UK network were contacted in July 2020, when nusinersen was the only DMT available in the UK. SMA-REACH UK is an SMA REsearch And Clinical Hub consisting of 17 paediatric centres across the UK. It is a network which collects data on patients diagnosed with all types of SMA, with the aim to establish a national agreement on clinical and physiotherapy assessments and standards of care. Each patient diagnosed with SMA is reviewed in the clinic and offered to voluntarily consent to having their clinical and physiotherapy data anonymously stored in the SMA-REACH UK database. Copies of their signed parents’ consent form and age adequate patient assent form are provided to the families. Physiotherapists at each centre were requested to provide retrospective data from the network focusing on spinal asymmetry presentation and management. Data were collected in August 2020. All children with a diagnosis of type 1 SMA of any age on nusinersen were included. Children involved in research trials were not included.

Patient and public involvement

Patients or the public were not involved in the design, conduct, reporting, or dissemination plans of our research.

Patient information collection

Patient data included sex, age (years and months), type 1 SMA subtype (type 1a/1b/1c), SMN2 copies (where available), age at first dose of nusinersen, sitting ability, seating type and respiratory support. Predefined questions gathered information regarding spinal posture and management of the type 1 SMA population including presence of spinal asymmetry (any observed scoliosis), kyphosis, reduced neck range of movement or torticollis and, when available, X-ray results. Spinal assessment is part of routine physiotherapy review. X-rays are not routinely completed rather when there is a clinical indication of asymmetry forming. Details on spinal management were requested from each centre including thoraco-lumbar-sacral orthosis (TLSO)/spinal jacket use.

Data were collected anonymously with no patient-identifiable information.

Data analysis and statistics

All analyses were performed using IBM SPSS Statistics V.25 software (IBM Corporation).

Characteristics of the sample are presented as median or frequency (percentage) unless otherwise stated.

Results

Of the 17 neuromuscular centres, 13 provided data on 80/101 (79%) type 1 SMA patients in the UK. There was representation across the UK including England, Wales, Scotland and Northern Ireland.

All 80 patients had type 1 SMA, and all were on nusinersen (table 1).

Table 1. Cohort characteristics (n=80).

Gender (M:F)	38 (47.5%):42 (52.5%)
Age	Median: 4 years 2 months. Range: 4 months to 14 years
Type 1 SMA, subtype b	31 (39%)
Type 1 SMA, subtype c	15 (19%)
Type 1 SMA, subtype unknown	34 (42%)
SMN2 copy number 2	40 (50%)
SMN2 copy number 3	15 (19%)
SMN2 copy number unknown	25 (31%)
Age started on nusinersen	Median: 12 months. Range: 1 month to 12 years 4 months
Duration of nusinersen treatment to reported spinal assessment	Median: 3 yearsRange: 0 months to 4 years 8 months

FfemaleMmaleSMAspinal muscular atrophySMN2survival motor neuron 2

Spinal presentation

Out of the 80 children, 33 had kyphosis and spinal asymmetry, 26 had spinal asymmetry but no kyphosis and 6 had kyphosis only with no spinal asymmetry. There were 14 who had no asymmetry or kyphosis on clinical observation. A child had spinal asymmetry, but kyphosis status was unrecorded (table 2).

Table 2. Spinal presentation, thoraco-lumbar-sacral orthosis (TLSO) use and sitting ability.

Spinal presentation, TLSO use and sitting ability	Sitting ability	Age range years (y) months (m)/median	SMA type	SMN2 copies	Age nusinersen initiated years (y) months (m)/median
Spinal asymmetry onlyn=26 (32.5%)TLSO use=15 (57.69%)	8 independent sitting	2 y 7 m–7 y 8 m/4 y 8 m	1b: 0/1c: 2	2 SMN2: 1/3 SMN2: 1	7 m–5 y/1 y 9 m
	4 sitting with support	7 m–10 y 7 m/4 y	1b: 4/1c: 0	2 SMN2: 4/3 SMN2: 0	2 m–6 y/4 y
	11 not sitting	7 m–14 y/4 y 8 m	1b: 4/1c: 1	2 SMN2: 6/3 SMN2: 0	2 m–12 y 4 m/1 y 4 m
	3 unknown	2 y 8 m–4 y/3 y 8 m	1b: 0/1c: 0	2 SMN2: 1/3 SMN2: 1	3 m–2 y 10 m/13 m
Spinal asymmetry and kyphosis n=33 (41.25%)TLSO use n=27 (81.81%)	13 independent sitting	2 y 2 m–8 y 9 m/4 y 1 m	1b: 5/1c: 4	2 SMN2: 6/3 SMN2: 2	1 m–5 y 6 m/6 m
	13 sitting with support	1 y 3 m–12 y 8 m/4 y 8 m	1b: 5/1c: 4	2 SMN2: 7/3 SMN2: 4	3 m–10 y 6 m/1 y 9 m
	6 not sitting	2 y 6 m–7 y 9 m/5 y	1b: 6/1c: 0	2 SMN2: 5/3 SMN2: 2	2 m–4 y 6 m/1 y 9 m
	1 unknown	2 y 9 m	1b: 0/1c: 0	2 SMN2: 1/3 SMN2: 0	9 m
Spinal asymmetry and unknown kyphosis n=1 (1.25%)TLSO use=1 (100%)	1 unknown	2 y 8 m	1b: 0/1c: 0	2 SMN2: 0/3 SMN2: 1	3 m
Kyphosis onlyn=6 (7.5%)TLSO use n=3 (50%)	2 independent sitting	4 y 7 m and 8 y 6 m	1b: 1/1c: 1	2 SMN2: 1/3 SMN2: 0	1 y 6 m and 5 y 4 m
	3 sitting with support	4 m–5 y 7 m/3 y 4 m	1b: 2/1c: 1	2 SMN2: 2/3 SMN2: 1	6 m–5 y/8 m
	1 not sitting	3 y 8 m	1b: 1/1c: 0	2 SMN2: 1/3 SMN2: 0	7 m
No asymmetry or kyphosisn=14 (17.5%)TLSO use n=4 (28.57%)	4 independent sitting	1 y–5 y/1 y 6 m	1b: 0/1c: 2	2 SMN2: 1/3 SMN2: 3	3 m–1 y 7 m/1 y 3 m
	2 sitting with support	1 y 3 m and 2 y 5 m	1b: 2/1c: 0	2 SMN2: 2/3 SMN2: 0	2 m and 7 m
	8 not sitting	5 m–7 y 2 m/4 y 4 m	1b: 1/1c: 0	2 SMN2: 2/3 SMN2: 0	1 m–4 y/1 y 4 m

The 20 children with no spinal asymmetry (including kyphosis only) ranged from 4 months to 8 years 6 months old with a median age of 3 years 8 months. There was an equal split of males and females with 60% type 1b and 40% type 1c. There were no type 1a’s. Further spinal presentations with cohort characteristics are shown in table 2.

There were 33 children (41%) with reduced neck range of movement including 1 with plagiocephaly and 5 with torticollis. Only one patient had plagiocephaly without reduced neck range of movement. A total of 26/33 (79%) of those with reduced neck range of movement had spinal asymmetry. A total of 35 children (44%) had a full range of movement and 12 (15%) were not documented.

Orthoses and Lycra

From the centres, six answered qualitative questions on TLSOs. All centres had children using TLSOs. For this patient population, a one-piece, rigid (front opening) spinal jacket with an abdominal hole was generally provided, and all were through the National Health Service in the UK. Semirigid braces provide less postural control so were only recommended if a rigid brace was not tolerated. If the family struggled to put on a one-piece brace occasionally, a two-piece TLSO was used. Lycra orthotics were used in one case. Hours recommended for TLSO use were varied between centres. They were advised when spending long periods in an upright position and when children were fatigued and leaning laterally due to poor postural control. Use during sleep or when in a reclined position was generally not advocated.

A total of 50/80 children used a supportive TLSO (table 2). TLSO median start age was 2 years 1 month, with the earliest start being at 8 months and the latest 5 years 5 months. From the 66 children with spinal asymmetry and/or kyphosis, TLSOs were used in 46 (table 2). There were 30 children who did not use a TLSO despite 11 of them having spinal asymmetry. Of the children not wearing a TLSO, 10 were reported to not have any spinal asymmetries or kyphosis which was the case in 5 that were under 18 months. These children had an age range of 5–18 months (median 1 year). There were three who had two SMN2 copies and one had three. Another two children were reported to be 3 years 8 months and 4 years with one recorded as having three SMN2 copies and both sitting independently with no spinal asymmetries. When no spinal asymmetries were reported in the older children, they were not able to sit due to their severe respiratory complications and high non-invasive ventilation use. These three children were aged 5–7 years 2 months with a median age of 5 years 6 months and one recorded as type 1b with two SMN2 copies. There were four that used a TLSO prophylactically with no spinal asymmetries. TLSO use was reportedly well tolerated in most children with low compliance only mentioned in three cases and other reasons such as respiratory needs preventing use or not needed following spinal surgery were highlighted as additional causes for not all children wearing TLSOs. Some highlighted initial difficulties but improved tolerance with time or spinal asymmetry progression.

Sitting ability

In our study, 27/80 children sat independently (ie, not wearing a TLSO with no hand support/propping), 22 sat with support and 26 were unable to sit. Sitting ability data were not recorded for five children. Prevalence of spinal asymmetry increased with achievement of sitting with most children with no spinal asymmetry being non-sitters (either too young to sit, or older weaker children who did not gain the ability to sit). Only four children who sat independently had no spinal asymmetries (age range 1–5 years with a median age of 1 year 6 months). Kyphosis was most prevalent in children who sat with support, followed by those who sat independently (figure 1).

Figure 1. Spinal descriptors in different sitting abilities.

X-rays and spinal surgery

There were 40/80 children who had been monitored with X-rays (table 3). Fixed fusion spinal surgery was performed in seven, an additional four were considered by the multidisciplinary team and the decision was to not proceed due to factors such as poor bone density or severe respiratory involvement. There were two out of seven who had already undergone spinal surgery prior to starting nusinersen treatment. The median age for spinal surgery was 8 years 4 months (range 4 years 8 months–12 years). All children continued on nusinersen.

Table 3. Summary of X-ray and spinal surgery.

	Yes	No	Not known
X-ray performed	40 (50%)	28 (35%)	12 (15%)
Spinal surgery performed	7 (8.75%)	67 (83.75%)	6 (7.5%)

Discussion

Spinal presentation

Due to advances in DMTs, prolonged survival and improved motor function have been observed in type 1 SMA patients who previously many would have died by age 2 years. However, some children who were initially of milder phenotype did survive resulting in later initiation of DMTs. These children, while alive, now have a more severe presentation and less response to the DMT. Kyphosis has been highlighted as a common factor in this group of treated type 1 SMA supported by previous literature.²³ These children may achieve sitting but are not actively standing. Therefore, they are potentially resting in a kyphotic posture while sitting to counteract their hamstrings pulling their pelvis back into a posterior tilt. TLSOs do not control a kyphosis well. Due to the young presenting age of these patients, alternative bracing needs to be considered. This may be difficult due to their size and potential respiratory involvement and requires discussions with orthotists and clinicians.

There is an association between spinal asymmetry and reduced neck range of movement, and it is therefore important that this is identified and monitored as part of spinal management.

Orthoses and Lycra

Descriptors regarding spinal management for sitters and non-sitters can be found in the international standards of care for SMA.²³ Thoracic bracing for postural stability is the standard management for all children with SMA with the aim to prevent spinal asymmetry and to promote motor development.^{6 23} Modifications, including abdominal holes, to allow for respiratory involvement should be standard. For children with the ability to sit, it is recommended that rigid braces should be applied for more than 60 min, five times a week, to be effective.²³ It is advised that children with more than a 20° Cobb angle use bracing; however, there is no consensus on positioning or length of wearing the brace, as well as no specific recommendation on prophylactic use of spinal braces in SMA. Consensus statements have been published by Vitale et al but do not differentiate between the different types of SMA.²⁴ In this consensus statement, 82% said there was a role in orthotic bracing for management.

Limited data are currently available on how effective TLSOs will be to manage spinal asymmetry for this group of type 1 SMA children as they grow. A longitudinal study is needed to investigate this.

Older studies report that bracing is not well tolerated and limited to untreated type 1 SMA.^{25 26} This does not appear to be the case in most of the population reviewed. In our study, the reason for not using a TLSO was mostly related to not having any spinal asymmetries, having already undergone spinal surgery or spending limited time sitting due to respiratory compromise. Poor compliance was only mentioned as the primary concern in three cases in our study. There were two children, aged 12 and 18 months respectively, who sat independently and did not yet have a TLSO; both had been sitting independently for 2 months.

Lycra is rarely used and there is currently limited evidence to support its use for scoliosis in SMA. Evidence is based on single case studies and conditions such as idiopathic scoliosis, spinal tumours and cerebral palsy that have different causes for their spinal asymmetry. As children with SMA present with muscle weakness, Lycra is unlikely to provide enough support to counteract the effect of gravity and the professional opinion of our SMA-REACH UK network is that they should not be recommended as the first-line intervention.

Spinal surgery

Spinal surgery is an option and needs to be considered given the prolonged life expectancy and increased asymmetry. Surgery is not possible in all children due to factors including respiratory complications and poor bone health and requires careful consideration and thorough investigations for timely intervention.²⁷ All patients in our cohort who received fixed fusion spinal surgery were treated with nusinersen after the age of 2 years and had a chronic presentation. The use of MAGnetic expansion control (MAGEC) growing rods has been reported in the severely affected SMA population.^{28 29} There was one child in our cohort with spinal asymmetry, which was detected early initially at the age of 4 months and was reported to need scoliosis surgery. This was on hold due to their young age, COVID-19 and the lack of MAGEC growing rods in the UK. In the UK, MAGEC rods were on hold due to safety concerns and traditional growing rods were used instead. The suspension has recently been lifted but they are not yet back in use in the UK.³⁰

Vitale et al published a consensus statement saying spinal instrumentation should be implemented with above 50° Cobb angle and growth friendly instrumentation should be used as needed in children from 4 to 8 years old.²⁴ Not all patients develop the same and spinal rigidity presents differently; thus, each individual needs to be considered independently for optimal intervention including orthotics and orthopaedic surgery.⁶ COVID-19 has had an impact on spinal surgery as many potential surgeries were put on hold during the height of the pandemic and hospitals continue to have long waiting lists.

Halo traction is not routinely used in patients with SMA and has limited research other than reported cases clinically. Studies have limited numbers and do not differentiate between conditions or are in adult populations.³¹,³⁴

Limitations and recommendations for future research

A limitation of our study is that it provides a snapshot of the UK population in 2020 and more children with spinal asymmetries have presented and been managed since then. Physiologically, it is possible that gravity and the acquisition of sitting ability could be having an impact on spinal asymmetry, posture and alignment due to axial weakness, and longitudinal data would need to be collected to confirm if this is the case.

Our study may have some selection bias; while a good representation from across the UK was gained data for 21 children (20.8%) were not accounted for. Unfortunately, not all sections were completed for all children with some data missing which may have affected overall results.

Routine spinal X-rays were not performed in many children with type 1 SMA, especially those under 4 years of age, for clinically mild and flexible spinal asymmetry. In these cases, surgery would not have been possible, either because of young age or respiratory complications. This made quantifying the severity of spinal asymmetry difficult. Further research is needed to include regular monitoring of Cobb angle.

This study did not look at the effect of therapy, nor at the speed and progression of curvature, and would be beneficial to look at longitudinally. It would also be beneficial to look at the incidence of spinal asymmetry before and after the introduction of DMTs and symptom onset compared with time of observation of spinal asymmetry. The current frequency of routine imaging may make this challenging. Increased life expectancy beyond 2 years warrants a review of the standard monitoring protocol, with the potential for increased frequency of X-rays seen in the future. This needs to be reviewed based on clinical need to avoid unnecessary radiation in children, for example, with no spinal asymmetry. While important in the early stages, this does not change management, as this is conservative including TLSOs until surgery is indicated.

Additional work with orthotists would be beneficial to investigate orthotic management especially for kyphosis management, and with the younger and therefore smaller patients who are more difficult to brace. While good compliance was reported, further investigation is needed looking at how many hours TLSOs are worn for and if this impacts presentation longitudinally.

With few children undergoing scoliosis surgery, the impacts of surgery on areas such as respiratory and function (including upper limb) are not yet known and require further investigation.

Children with type 1 SMA are changing rapidly with new challenges and presentations. This study focused on nusinersen as at the time of data collection this was the only treatment available in the UK as part of the Expanded Access Programme. This is a large and non-homogeneous group including young and longer surviving older type 1 SMA children. At that time, we had not established a standardised prospective protocol to collect the information regarding the effect of therapeutic intervention on spine complications. More analysis into different groups would be beneficial in the future. Other DMTs have more recently become available including the use of gene therapy (Zolgensma) and SMN2 splice modifying small molecules (Evrisdy). The impact of these on spines is not yet known, therefore needing further investigation.

There are little data and limited studies in this area and gaps in knowledge have been identified^{6 35} with more specific consensus needed for management.

Conclusions

Our study confirms a high prevalence of spinal asymmetry in treated type 1 SMA needing long-term management. It provides information on the presentation and treatment options for this patient cohort including a new option of spinal surgery previously not considered. With an expected increased type 1 SMA population, early and ongoing assessment of these children is essential and consideration for the new presenting management challenges needs to be made. Our study will facilitate the development of guidelines to optimise and standardise care and management of this group of patients.

Data availability statement

All data relevant to the study are included in the article or uploaded as supplementary information.