Highlights
-
A multicenter retrospective observational study exploring the need for respiratory and feeding support post-nusinersen treatment in Canada.
-
Continued need for respiratory and feeding support in most children with SMA type 1 when nusinersen is initiated after symptom onset.
-
These children require close routine monitoring for respiratory and feeding function.
Introduction
Spinal muscular atrophy (SMA) is an autosomal recessive neuromuscular disorder caused by a homozygous deletion or compound heterozygous alterations (mutation and deletion) in the survival motor neuron 1 gene (SMN1).Reference Rodrigues, Owen, Talbot, Ignatius, Dubowitz and Davies1,Reference Wirth2 Characterized by degeneration of the anterior horn cells, persons with SMA experience progressive muscle atrophy and weakness, leading to supportive care needs for feeding and respiratory function.Reference Arnold, Kassar and Kissel3
In recent years, three disease-modifying therapies have been approved by Health Canada: nusinersen (Spinraza®), risdiplam (Evrysdi®) and onasemnogene abeparvovec (Zolgensma®). Nusinersen, the first-approved therapy, is an antisense oligonucleotide that alters pre-mRNA splicing of the SMN2 gene, a gene paralogous to SMN1, which promotes the inclusion of exon 7 in the final SMN2 product and increases the expression of full-length and functional SMN protein.Reference Passini, Bu, Richards, Kinnecom, Sardi and Stanek4 The efficacy of nusinersen on motor function and survival in infantile-onset SMA with two copies of SMN2 was demonstrated in a phase III, randomized, double-blind, sham-controlled clinical trial (NCT 01839656) with initial treatment before 7 months of age.Reference Finkel, Mercuri, Darras, Connolly, Kuntz and Kirschner5 Nusinersen became available through a compassionate access program before Health Canada approval in June 2017,6 eventually obtaining positive reimbursement recommendation by Institut National d’Excellence en Santé et Services Sociaux (INESSS) in December 2018 and by Canada’s Drug and Health Technology Agency (CADTH) in February 2019.7 Although efficacy was shown in less than 7-month-old infants with SMA type 1 who were not dependent on invasive ventilation, clinical access to the drug was made available to older patients across Canada with more advanced disease.
Nusinersen has been generally well-tolerated and shown to improve functional motor outcomes in several studies, with the highest effect when initiated early.Reference Finkel, Mercuri, Darras, Connolly, Kuntz and Kirschner5,Reference Pechmann, Behrens, Dornbrack, Tassoni, Stein and Vogt8–Reference Ergenekon, Yilmaz Yegit, Cenk, Gokdemir, Erdem Eralp and Ozturk12 However, observations of its effectiveness on respiratory and bulbar outcomes have been variable, with some studies showing no significant change, whereas others report a decrease in need for ventilation or tube feeding.Reference Pechmann, Behrens, Dornbrack, Tassoni, Stein and Vogt8,Reference Aragon-Gawinska, Seferian, Daron, Gargaun, Vuillerot and Cances13–Reference Hepkaya, Kilinc Sakalli, Ulkersoy, Baskan, Arslan and Meral16
The aim of this study is to assess the evolution of respiratory and feeding support in symptomatic children with SMA type 1 by 24 months after the initiation of treatment with nusinersen and explore if this is modified by age at initiation of treatment.
Methods
Study design
We conducted a multicenter, retrospective, observational study on children with SMA type 1 treated with nusinersen.
Data source
The Canadian Neuromuscular Disease Registry (CNDR) is a pan-neuromuscular disease registry that prospectively collects data on SMA individuals from 30 clinics (15 pediatric and 15 adult) across Canada to gain real-world evidence on disease progression, natural history and therapy outcomes in children and adults.Reference Hodgkinson, Oskoui, Lounsberry, M’Dahoma, Butler and Campbell17 Clinical data and outcome assessments performed during routine clinic visits were retrieved from medical record reviews and entered into a secure web portal (REDcap) maintained at the University of Calgary. Data on all children with SMA type 1 between 2017 and 2023 were extracted from the CNDR SMA module on July 13, 2023.
Participants
The inclusion criteria for this study were (i) genetically confirmed diagnosis of SMA (homozygous deletion or compound heterozygous alterations of exon 7 of SMN1, (ii) symptom onset in the first 6 months of life, (iii) no ability to sit independently (as per definition of SMA type 1) prior to treatment initiation and (iv) received at least one dose of intrathecal nusinersen injections. We extracted data at baseline and 24 months from the first nusinersen dose date, accepting data captured within 3 months of each visit. If a child discontinued nusinersen treatment, the data for the follow-up visits were still included. The cohort was divided into two groups based on their age at nusinersen treatment initiation: ≤2 years and >2 years.
Variables
The CNDR variable for respiratory support status includes the following categories: none (no ventilation), noninvasive part-time (night only), noninvasive part-time (day and night), noninvasive full-time (≥16 hours per day) or invasive ventilation. When analyzing the data, misclassification was observed for the noninvasive part-time (day and night) and noninvasive full-time (≥16 hours per day) options. These two categories were collapsed together as noninvasive full-time (≥16 hours per day). Hospitalization data due to respiratory complications in the 12 months preceding the baseline and 24-month visits were also extracted.
Feeding support was determined by the insertion and use of a feeding tube. The variables available for feeding support include whether a feeding tube was placed, the date of tube placement, the type of tube and whether it is used for exclusive or supplementary feeding. Sex- and age-specific body mass index (BMI) z-scores were calculated using the 2006 World Health Organization growth charts for children under 24 months of age and the Centers for Disease Control and Prevention growth charts for children aged 24 months and older.Reference Grummer-Strawn and Krebs18 BMI z-scores below −2 SD were classified as underweight and above 2 SD as overweight.
Outcomes
Our primary outcomes were (i) time to death or full-time (≥16 hours per day) ventilatory support (noninvasive or invasive) and (ii) time to initiation of feeding support. For death, the date of death was used, and for respiratory support, the age at initiation of full-time ventilation was used. The date of placement of the feeding tube was used for initiation of feeding support. Our secondary outcome was the change in the respiratory and feeding support 24 months after the first nusinersen treatment.
Statistical methods
Descriptive analysis for categorical data was performed by calculating the absolute frequencies and percentages. Continuous data were analyzed as median with range. Kaplan–Meier survival analyses were done to compare the time to event for the two treatment initiation groups. Log-rank test was performed to compare the survival probabilities between the two groups.
Secondary analysis included Fisher’s exact tests for categorical variables and Mann–Whitney U test for continuous variables for comparison of the 24-month follow-up data of respiratory and feeding support between the two groups. P < 0.05 was considered statistically significant. All analyses were performed in Stata V.13.1.19
Results
At the time of data extraction (July 13, 2023), there were 64 children with SMA type 1 enrolled in the CNDR, of which 38 met inclusion criteria. Six were excluded as baseline data were missing, leaving 32 children included in this study (Figure 1). Baseline characteristics for the full cohort and the two subgroups, nusinersen treatment initiated ≤2 years and >2 years of age, are presented in Table 1. Twenty-four children (75%) had two copies of SMN2, and eight (25%) had three copies of SMN2. The median (range) age at symptom onset for the full cohort was 2 months (0–6), and the median age at nusinersen treatment initiation was 6 months (0.8–183.8). At the time of nusinersen initiation, 50% (n = 16) of the total cohort were already receiving some level of respiratory support with a median age of initiation of 11 months (1.2–39), and 56% (n = 18) had a feeding tube, with a median age of insertion of 9 months (2.2–131). Both groups had similar demographics and clinical characteristics except for age at initiation of nusinersen (as per study design), age of symptom onset (p = 0.03) and children with three copies of SMN2 (p = 0.002).
Study flow chart.
Baseline descriptive characteristics of SMA type 1 children
SMN2 = Survival motor neuron 2; NIV = noninvasive ventilation; BMI = body mass index; CHOP-INTEND = Children’s Hospital of Philadelphia Infant Test of Neuromuscular Disorders; *Mann–Whitney U test. P < 0.05, ** Fisher’s exact test. P < 0.05.
Of the 32 children, 23 had 24-month follow-up data since nusinersen treatment initiation was available. Two children were deceased, and two children had not yet completed 24 months of nusinersen treatment. Five children did not have data on the 24-month visit available in the registry. The overall median (range) duration of nusinersen treatment was 5.8 months (0.9–51.3). Children who discontinued nusinersen treatment were included in the study. Eleven children had discontinued nusinersen treatment before completing 24 months of nusinersen treatment. The median age for discontinuing treatment was 13.5 months (5–111). Ten children received gene therapy (Zolgensma), and one child stopped meeting criteria for treatment with nusinersen due to tracheostomy placement and was treated subsequently with risdiplam.
Primary outcomes
At baseline, of the children treated ≤2 years of age (n = 20), four (20%) were receiving respiratory support, of which two (10%) required full-time ventilation. At 24 months after first nusinersen treatment, eight (81.9%) children required respiratory support, with four (36.4%) requiring full-time ventilation. In the group of children treated after 2 years of age (n = 12), all 12 children were on respiratory support, with 7 (58.3%) needing it full-time at baseline. At 24 months follow-up, six (50%) needed full-time ventilation. Overall, the median age (range) of death or full-time ventilation was 8.8 monthsReference Passini, Bu, Richards, Kinnecom, Sardi and Stanek4–Reference Sansone, Pirola, Albamonte, Pane, Lizio and D’Amico24 (Figure 2). At 24 months from nusinersen initiation, there was no difference in the probability of event-free survival between the two (≤2 years vs >2 years) treatment groups, with rates of 61% versus 45% (p = 0.44). Two children had died at the ages of 6.9 and 12.1 months in the ≤2-year treatment group while none in the >2-year group.
Time to death or full-time respiratory support.
Eight (40%) children had feeding tube support (two nasogastric tube [NG] and six gastrostomy [G] tube) at baseline in the ≤2-year treatment group. At 24-month follow-up, eight (72.7%) children required a G-tube. In the older treatment group, 10 (83.3%) children had feeding tube support (2 NG-tube, 7 G-tube and 1 jejunostomy tube). At 24-month follow-up, no additional children required a feeding tube. The median age (range) for insertion of feeding tube was 9.4 months (1.7–130.6) for the overall cohort (Figure 3). There was no difference in the probability of remaining event-free beyond 24 months from the first nusinersen treatment (p = 0.63), with rates of 30% and 36% in the ≤2-year and >2-year treatment groups, respectively.
Time to initiation of feeding support.
Secondary outcomes
At 24 months from treatment initiation, there was no significant difference in the need for respiratory support between the two treatment groups, with the majority in both groups needing ventilation(p = 0.58 for part-time support and p = 0.41 for full-time support; Table 2). In the early treatment group, 81.9% of children required noninvasive ventilation compared to 100% in the late treatment group. There was also no difference in hospitalization rates due to respiratory problems at 24 months between groups, with rates of 27.3% and 8.3%, respectively.
Comparison of the ventilatory support status at 24-month follow-up
NIV = noninvasive ventilation.
Feeding support
Of the eight (40%) children from ≤2-year treatment group (n = 20) that had a feeding tube, six patients required it for exclusive feeding, and two patients required it for supplemental feeding. By 24 months, no further children had required feeding tube insertion. However, one patient who required supplemental feeding progressed to exclusive feeding.
In >2-year treatment group (n = 12), of the 10 (83.3%) children who had a feeding tube at initiation, 2 required it for supplemental and 8 for exclusive feeding support. There was no shift from supplemental to exclusive tube feeding after treatment initiation. At 24 months from nusinersen initiation, there was no significant difference in need for feeding support between the two groups, with the majority needing some feeding support in both the early and late treatment groups, 72.7% and 83.3%, respectively (Table 3). There was no difference in the BMI z-scores between the two treatment groups at baseline and at 24-month follow-up.
Comparison of the feeding support status at 24-month follow-up
BMI = body mass index.
Discussion
Our study is a multicenter retrospective observational study exploring the respiratory and feeding support status post-nusinersen treatment in Canada. SMA type 1 is the most severe form of SMA, with disease onset before the age of 6 months and a very high probability of morbidity and mortality before 2 years of age without respiratory support or disease-modifying treatment.Reference Finkel, McDermott, Kaufmann, Darras, Chung and Sproule20–Reference Zerres and Rudnik-Schöneborn22 Introduction of respiratory and feeding support has increased the probability of 2-year survival from 30.8% to 73.9% in children with type 1 SMA.Reference Oskoui, Levy, Garland, Gray, O’hagen and De Vivo23 The decision of when to initiate feeding and respiratory support can be center-dependent, with some sites initiating these earlier to support growth and chest wall development.
Our study reaffirms observations in other international cohorts showing a significant decrease in mortality but continued need for feeding and respiratory support in this population after treatment initiation with nusinersen.Reference Ergenekon, Yilmaz Yegit, Cenk, Gokdemir, Erdem Eralp and Ozturk12,Reference Lavie, Diamant, Cahal, Sadot, Be’er and Fattal-Valevski14–Reference Hepkaya, Kilinc Sakalli, Ulkersoy, Baskan, Arslan and Meral16,Reference Sansone, Pirola, Albamonte, Pane, Lizio and D’Amico24,Reference Heitschmidt, Pichlmaier, Eckerland, Steindor, Olivier and Fuge25 Few studies, however, have multicenter post-treatment data at follow-up periods as long as 2 years or more.Reference Pechmann, Behrens, Dornbrack, Tassoni, Stein and Vogt8,Reference Lavie, Diamant, Cahal, Sadot, Be’er and Fattal-Valevski14,Reference Chen, Widger, Teng, Fitzgerald, D’Silva and Farrar15 The ANCHOVY studyReference Cances, Vlodavets, Comi, Masson, Mazurkiewicz-Beldzinska and Saito26 was a global, multicentered study and followed 60 SMA type 1 children who were never on any disease-modifying treatment. They used similar definitions for respiratory and feeding support outcomes. In the ANCHOVY natural history cohort, the probability of event-free survival (not dying or requiring permanent ventilation and insertion of feeding tube) at 24 months of age was 7% and 8%, respectively, whereas in our study, it was 60% and 29%, respectively.
Our study confirmed that the majority of our patients required respiratory support over time, in keeping with previous studies.Reference Lavie, Diamant, Cahal, Sadot, Be’er and Fattal-Valevski14,Reference Sansone, Pirola, Albamonte, Pane, Lizio and D’Amico24 One patient from each treatment group had reduced need for noninvasive ventilation from full-time to part-time after nusinersen initiation, whereas two remained on full-time ventilation. This is in contrast with a study by Hepkaya et al.Reference Hepkaya, Kilinc Sakalli, Ulkersoy, Baskan, Arslan and Meral16 where two SMA type 1 children with tracheostomy transitioned to part-time noninvasive respiratory support by approximately 6 months post-nusinersen treatment initiation. They had two SMN2 copies, and treatment was initiated at the mean age of 15 months. Our results support the observations reported in other studies that patients treated post-symptom onset with nusinersen continue to need respiratory support.
While most studies have looked at the clinical respiratory status to determine the effect of nusinersen treatment, Chacko et al. have analyzed pulmonary function tests and reported there was a decrease in the decline of lung function within the first year of treatment in SMA types 2 and 3.Reference Chacko, Sly, Ware, Begum, Deegan and Thomas11 Although a difference in the functional motor outcomes due to treatment initiated before or after 2 years has been reported,Reference Audic, de la Banda, Bernoux, Ramirez-Garcia, Durigneux and Barnerias27,Reference Pane, Coratti, Sansone, Messina, Catteruccia and Bruno28 this improvement does not necessarily translate into stabilization of lung function or risks associated with decreased lung function such need for respiratory support and acute respiratory exacerbations requiring hospitalization. In our study, we did not observe any significant differences in the respiratory support needed or hospitalization for respiratory problems at 24 months post-treatment initiation between the two treatment groups in our study, suggesting that nusinersen might have a limited impact on respiratory function when initiated after symptoms onset.
The need for tube feeding support after initiation of nusinersen treatment in post-symptomatic patients remains whether treated before or after 2 years of age, in agreement with previous studies.Reference Weststrate, Stimpson, Thomas, Scoto, Johnson and Stewart10,Reference Sansone, Pirola, Albamonte, Pane, Lizio and D’Amico24,Reference van der Heul, Cuppen, Wadman, Asselman, Schoenmakers and van de Woude29 Possible reasons for this include early brainstem motor neuron involvement in this population and potentially reduced rostral exposure to nusinersen from lumbar intrathecal delivery. In infants with two copies of SMN2 treated with nusinersen pre-symptomatically between 7 days and 56 days of age have also been shown to develop bulbar symptoms and subsequent need for feeding support within the first year of life.Reference van der Heul, Cuppen, Wadman, Asselman, Schoenmakers and van de Woude29 In our cohort, the median age for tube feeding was 6.5 and 14.5 months in the early and late treatment groups, respectively. This difference is likely attributed to baseline differences between the two groups, with more patients with three copies of SMN2 in the later treatment group, which also had a later age of symptom onset.Reference Chen, Widger, Teng, Fitzgerald, D’Silva and Farrar15,Reference Finkel, McDermott, Kaufmann, Darras, Chung and Sproule20
Our data are from a national registry and provide insight into the real-world data of treatment and management of rare diseases such as SMA. However, it is limited to its retrospective nature and to the data available in the registry. Missing data accounted for 15% of the cohort that had to be excluded. If the reasons for incomplete data are not random, this could potentially introduce bias. The CNDR is based on data collected for clinical purposes. This approach has the benefit of improving patient compliance and minimizing workload and cost on participating sites. However, this approach has its limitations for research purposes. Respiratory and feeding support variables defined may not capture all clinically meaningful differences, for example, ventilatory pressures used or swallowing studies. Because participation in the registry is voluntary, it may introduce selection bias toward certain demographics. Because recruitment is clinic based, there is also potential for survival bias. The study sample size, although reasonable for a rare disease, may also reduce the power to detect meaningful differences.
Conclusion
Children with SMA type 1 treated with nusinersen post-symptom onset continue to need both respiratory and feeding support. Families should be counseled on disease progression and treatment expectations, and these children should be closely and routinely monitored for respiratory and feeding function and need for additional supportive intervention.
Author contributions
MO, TB, GG and VH conceptualized the study design. GG and TB drafted the initial manuscript. TB and MO contributed to the analysis of data. KS, JKM, HK, MCC, MC, EL, KJ, CC, LM, HG, HM, AM, CO, MT and MO contributed to data acquisition. All authors critically reviewed the manuscript and approved the final version for submission.
Funding support
The SMA registry in the CNDR is supported financially by Biogen. They are not involved in the project development, data collection or data analysis.
Competing interests
Tarannum Behlim has nothing to disclose. Craig Campbell has received consultancy fees (Biogen, Roche) and travel support (Biogen, Roche). Maria Castro-Codesal has nothing to disclose. Megan Crone has participated in an advisory board (Biogen) and received travel support (Biogen). Guillaume Gauvreau has nothing to disclose. Hernan Gonorazky has received consultancy fees (Novartis, Roche and Biogen) and speaker fees. (Novartis). Victoria Hodgkinson has received honoraria for speaker fees (Biogen, Roche) and participation in an advisory board (Novartis). Kristina Joyal has received fees for participation in an advisory board (Biogen, Novartis, Roche). Hanna Kolski reports funding from the Canadian Institutes of Health Research. Edward Leung has nothing to disclose. Jean K. Mah has received research grants from the Alberta Children’s Hospital and has been a site principal investigator on clinical trials (Roche, Biogen). Laura McAdam has nothing to disclose. Anna McCormick has nothing to disclose. Hugh McMillan has been a site principal investigator on clinical trials (Novartis, Biogen and Roche). He has received honoraria for consultancy (Novartis and Roche). Pamela Ng has nothing to disclose. Colleen O’Connell has received research grants from Roche, the Canadian Neuromuscular Disease Registry and Biogen. She has received advisory fees (Roche, Biogen) and lecture fees (Roche, Biogen) and was an unpaid volunteer on the Muscular Dystrophy Canada Scientific and Medical Advisory Board. Maryam Oskoui has been a site principal investigator on clinical trials (Novartis, Biogen and Roche). Kathryn Selby has been a site principal investigator on clinical trials (Biogen). She has received consultancy fees (Biogen, Novartis) for participation in advisory boards, speaker fees (Roche, Novartis) and travel support (Novartis). Monique Taillon has nothing to disclose.
Ethical standards
The CNDR is led by the University of Calgary, and ethics approval was obtained from the Conjoint Health Research Ethics Board at the University of Calgary. All clinical sites participating in the registry obtained local research ethics board approvals for data collection. Eligible patients are consented to participate in the registry and make available their information for research purposes.
Target article
Respiratory and Bulbar Support in Spinal Muscular Atrophy Type I Treated with Nusinersen
Related commentaries (1)
Reviewer Comment on Gauvreau et al. “Respiratory and Bulbar Support in Spinal Muscular Atrophy Type I Treated with Nusinersen”