Spinal Muscular Atrophy

Spinal muscular atrophy (SMA) is a progressive neuromuscular disorder in which the anterior horn cells of the spinal cord degenerate. Before gene therapy it was the leading genetic cause of infant death, and new disease-modifying therapies have since changed its prognosis dramatically.

Genetics. Inheritance is autosomal recessive, from a mutation or deletion of SMN1 (survival motor neuron 1) on chromosome 5q. Incidence is approximately 1 in 11,000 live births, the international consensus figure, and the carrier frequency roughly 1 in 40-60.

The lesion. The anterior horn cells, the lower motor neurons, degenerate and the skeletal muscles they supply are denervated. Weakness is proximal more than distal, and the lower limbs are typically weaker than the upper.

Why scoliosis develops. Trunk weakness and loss of antigravity control allow a long, flexible neuromuscular curve to progress. Scoliosis is universal in untreated Type I and II, and pelvic obliquity and poor sitting balance are common in non-ambulant children. Curves are often long thoracolumbar sweeps, but configuration varies, so do not define SMA by a single mandatory C-shaped pattern.

Treatment is changing the curve. Disease-modifying therapy is changing the onset and progression of scoliosis, so follow each child's radiographic and functional trajectory. Orthopaedic manifestations may still occur after early treatment, but they are less severe.

View larger

Two genes, one protein. SMN1 is the primary gene producing full-length, functional survival motor neuron (SMN) protein, and homozygous loss or mutation of SMN1 causes SMA. SMN2 is a near-identical paralogue in the same chromosome 5q region. It differs from SMN1 by a single translationally silent C-to-T nucleotide change within exon 7.

The splicing defect. That single base change disrupts an exonic splicing enhancer, so most SMN2 transcripts skip exon 7 and produce a truncated, rapidly degraded protein (SMNΔ7). Only roughly 10 to 15 percent of SMN2 output is full-length functional SMN.

Why copy number is prognostic. SMN2 is present in variable copy number, from 1 to 4 or more. Each extra copy adds a little more full-length SMN and partially compensates for the absent SMN1, so more copies correlate with a milder phenotype. Typical copy numbers by type:

• Type I: 1 to 2 copies
• Type II: 3 copies
• Type III: 3 to 4 copies

Copy number predicts severity at a population level but is not an absolute individual guarantee.

Every therapy raises SMN. All the disease-modifying therapies target the same bottleneck, the supply of functional SMN. Nusinersen, an antisense oligonucleotide, and risdiplam, a small molecule, both promote exon-7 inclusion so that SMN2 makes more full-length protein. Onasemnogene abeparvovec instead delivers a replacement SMN1 transgene in an AAV9 vector. The principles of the gene-replacement platform itself are covered in the gene therapy and tissue engineering topic.

View larger

The four types are defined by age at onset and the best motor milestone reached.

Treatment has rewritten Type I. With gene therapy, survival in Type I has improved dramatically and many children are achieving motor milestones. Early treatment, before symptom onset, leads to the best outcomes.

History. Establish the age of symptom onset, the motor milestones achieved and those since lost, the family history and the respiratory status.

Examination. The findings to look for:

• Hypotonia, the floppy infant of Type I
• Weakness, proximal greater than distal and symmetric
• Reflexes absent or diminished
• Tongue fasciculations, which are characteristic
• The spine, for scoliosis
• The hips, for dysplasia and contractures

Genetic testing. An SMN1 deletion or mutation confirms the diagnosis, and the SMN2 copy number gives the prognosis. EMG and nerve conduction studies show a denervation pattern but are rarely needed now that genetic testing is available.

Pulmonary function. Monitor the forced vital capacity (FVC).

Spine radiographs. Obtain seated or standing whole-spine radiographs according to motor status, curve progression and treatment planning. Assess pelvic obliquity and sagittal alignment, not the Cobb angle alone.

View larger

Hip radiographs. Pelvic radiographs are symptom- and surveillance-driven. When hip displacement is being followed, document three measurements, and follow them together when progressive displacement is being considered for intervention:

• Migration percentage (Reimer), the proportion of each femoral head lateral to Perkin's line. Use reproducible serial measurements to follow displacement rather than describing the hip simply as located or dislocated.
• Acetabular index, measured between Hilgenreiner's line and the acetabular roof. It records acetabular dysplasia separately from femoral-head migration, which helps explain why identical migration percentages may represent different hip morphology.
• Head-shaft angle, which quantifies proximal-femoral valgus.

View larger

View larger

View larger

Key Differentiators for SMA: SMN1 genetic testing is confirmatory; tongue fasciculations are characteristic; CK is normal or only mildly elevated; reflexes are absent or diminished.

Disease-modifying therapy. Three agents are in use:

• Onasemnogene abeparvovec (Zolgensma): gene therapy, a single IV dose in infants
• Nusinersen (Spinraza): intrathecal injections every 4 months
• Risdiplam (Evrysdi): an oral SMN2 splicing modifier

Respiratory and nutritional care. Respiratory failure is the primary cause of death, and respiratory care is the primary focus of treatment: BiPAP, cough assist and suctioning. A gastrostomy is placed if swallowing is impaired.

Scoliosis. Seating systems and braces support posture and function but do not reliably halt structural progression. Surgery is for progressive deformity that compromises sitting balance, pelvic level, pain, care or respiratory mechanics despite non-operative support. Decide from the combined picture of progression, growth remaining, sitting balance, pelvic obliquity, pain, care difficulty and respiratory trajectory rather than a single Cobb-angle or FVC threshold; the Cobb angle informs timing but is not a stand-alone trigger.

Hip displacement. Displacement is common in Type I and II and is usually observed. Many painless displaced hips are managed this way, especially when reduction would not improve function or seating. Monitor symptoms, migration and pelvic obliquity, and escalate for pain, progressive displacement, loss of function or care difficulty through a multidisciplinary, goal-directed plan.

Contractures. Contractures at the hips, knees, ankles and feet affect positioning and function. Preserve comfortable range with positioning, stretching, splinting and therapy, and operate selectively, with releases, when a contracture blocks function, hygiene, seating or orthotic use.

Growth-friendly surgery. In younger children, traditional growing rods, magnetically controlled rods and other growth-friendly strategies trade repeated procedures or implant complications against preservation of thoracic growth. Selection depends on age, curve magnitude and flexibility, chest growth, bone quality, respiratory reserve and family goals. Some children later undergo definitive fusion; others follow different paths as disease-modifying therapy changes growth and motor function.

Posterior spinal fusion. For sitters, the guidance summarised below recommends a long fusion to the pelvis once growth is adequate. Even so, fusion extent and whether to include the pelvis depend on curve flexibility, pelvic obliquity, ambulatory potential, bone quality and fixation strategy. Iliac and S2-alar-iliac fixation are both established options, and pelvic fixation is not mandatory for every child.

View larger

View larger

Scoliosis surgery in SMA carries a high perioperative risk and needs an experienced team. The airway and ventilation, not the spine, are what threaten the patient.

Respiratory reserve. Intercostal weakness produces restrictive lung disease with a low and falling FVC, and a low pre-operative FVC predicts post-operative ventilatory dependence. A low FVC increases risk but should prompt planning rather than act as an isolated exclusion.

Preparation. Treat any chest infection, support nutrition, and train the child with non-invasive ventilation (NIV/BiPAP) and mechanical cough assist before surgery. Long instrumented fusions carry significant blood loss, so plan cross-match, cell salvage and blood management, with neuromonitoring and pressure protection.

Extubation. Plan for elective post-operative NIV and aggressive secretion clearance rather than reactive re-intubation. Protocolised extubation directly onto NIV reduces re-intubation after scoliosis fusion.

Airway and drugs. Bulbar weakness and impaired cough mean a real aspiration risk, so airway protection and secretion management need care. Avoid depolarising neuromuscular blockers (suxamethonium): denervated, atrophic muscle carries up-regulated extrajunctional acetylcholine receptors and risks life-threatening hyperkalaemia. Expect heightened and prolonged sensitivity to non-depolarising relaxants and to opioids and sedatives; titrate carefully, use quantitative neuromuscular monitoring and ensure full reversal.

The child is admitted to intensive care after scoliosis surgery and may require prolonged ventilation. Analgesia is multimodal, and the child sits early in the wheelchair. Bracing is discontinued after fusion, while respiratory and nutritional support continue long term.

What fusion gives. Scoliosis surgery transforms sitting balance and quality of life: independent sitting rose from 22.7% to 77.3% in one severe-deformity series.

What it does not. Respiratory function is preserved rather than gained. Mean FVC remained stable in that series, but all six SMA patients in a separate pulmonary-function series lost vital capacity after fusion (see the evidence below). The honest consent line is that surgery does not reverse the respiratory decline and may reduce vital capacity.

Global epidemiology

• Incidence approximately 1 in 11,000 live births; pan-ethnic, with carrier frequency roughly 1 in 40-60.
• SMN2 copy number is the principal genetic modifier worldwide: 2 copies predict the most severe phenotype, while 3 or more copies predict milder disease.

Major guidelines, side by side

Registries & access

• No single global SMA implant registry; outcome data come from national neuromuscular registries (e.g. TREAT-NMD, SMArtCARE in Europe, Cure SMA in North America) and tertiary-centre series.
• Newborn screening for SMA is now established across much of North America, Europe and parts of Asia-Pacific, but coverage remains uneven globally.

High- vs limited-resource practice variation

• High-resource settings: newborn screening, early disease-modifying therapy, MAGEC/growth-friendly constructs, intrathecal-access-preserving fusion, ICU-supported neuromuscular spine surgery.
• Limited-resource settings: diagnosis often clinical or delayed; disease-modifying drugs frequently unaffordable; management centres on respiratory support, seating, contracture care, and fusion where safe perioperative care exists.

• Spinal surgery in the gene-therapy era: Disease-modifying therapy is reducing curve severity in early-treated children, but it is unclear whether it will eliminate the need for fusion or merely delay it. Long-term natural history of the treated cohort is still being defined.
• Timing of fusion vs growth-friendly constructs: When to convert magnetically controlled growing rods to definitive fusion, and whether early definitive fusion in older children is preferable, remains debated; both strategies show durable correction in current series.
• Preserving intrathecal access: Nusinersen requires repeated intrathecal injection, so "skip" or fenestrated constructs and laminotomy windows are used to keep access; the optimal technique and whether oral risdiplam removes this constraint are unresolved.
• Pulmonary outcome of fusion: Fusion reliably improves sitting and halts deformity, but vital capacity often falls rather than improves after surgery — the respiratory benefit is mainly preservation, not gain.
• Lower FVC threshold for surgery: There is no universally agreed FVC cut-off; very low FVC raises perioperative risk but is not an absolute contraindication when seating and quality of life are at stake.
• Hip management: Whether asymptomatic hip subluxation/dislocation ever warrants reconstruction in non-ambulators is contested; most centres observe because reconstruction has high redislocation rates and limited functional gain.