SMA Disease-Modifying Therapies: Core Approaches Compared

Why SMA Disease-Modifying Therapies Matter

Spinal muscular atrophy is an autosomal recessive motor neuron disease caused by biallelic loss of the SMN1 gene. The nearly identical SMN2 gene produces mostly a truncated, unstable protein, so SMN2 copy number blunts but does not prevent the deficit.

Untreated type 1 disease is the leading genetic cause of infant mortality. Most affected infants never sit unsupported, and the majority do not survive beyond two years without ventilatory support.

Within a few years, three mechanisms to raise SMN protein moved from concept to approval, transforming a uniformly fatal infantile disease into a treatable one. A fourth, muscle-directed strategy is now in late-stage testing, reframing the question from whether SMA can be treated to how these approaches compare.

The Options Compared

Four agents anchor the current landscape. Three raise SMN protein by distinct routes; the fourth leaves SMN untouched and instead targets muscle. The table summarises each pivotal trial with its verified design, sample size, and headline result.

Survival and Ventilation — How They Differ

In infantile-onset disease, survival free of permanent ventilation is the outcome that matters most, because it separates a fatal course from a survivable one.

In ENDEAR, nusinersen led to a 47% lower risk of death or permanent ventilation than sham control, with a hazard ratio of 0.53 (P=0.005). The trial was stopped early once its motor-milestone endpoint was met.

Gene replacement produced comparably strong survival signals. In STR1VE-US, 91% of treated infants were alive and ventilation-free at 14 months, against roughly 26% in a historical natural-history cohort.

Risdiplam showed a similar direction in FIREFISH Part 2, where 85% of infants were alive without permanent ventilation at 12 months versus a natural-history criterion of 42%. Because the gene therapy and risdiplam infantile trials were single-arm against historical controls, their survival figures cannot be placed on the same statistical footing as the randomised nusinersen comparison.

Motor Function — How They Differ

Motor endpoints diverge by age group and by the scale each trial used, which complicates direct comparison. Infantile trials track milestone attainment; later-onset trials track change on continuous motor scales.

Among infants, gene therapy produced the highest proportion reaching independent sitting in its pivotal cohort — 64% (14 of 22) sat for at least 30 seconds — while risdiplam’s FIREFISH Part 2 recorded 29% sitting unsupported for at least 5 seconds. The differing sitting definitions and single-arm designs mean these percentages describe each therapy against natural history, not against each other.

In later-onset disease, both randomised trials cleared their thresholds by modest margins. CHERISH showed more than half of nusinersen-treated children gaining at least 3 HFMSE points versus about a quarter of controls. SUNFISH produced a statistically significant but small MFM32 treatment difference of 1.55 points.

Apitegromab addresses a different gap. Because SMN-targeted therapies preserve motor neurons but leave residual muscle weakness, SAPPHIRE tested whether adding a muscle-directed antibody improves function on top of background SMN treatment. It produced an HFMSE mean difference of 1.8 points over placebo, with early separation by week 8. This provides evidence for a combination strategy rather than a replacement for SMN therapy.

Subgroups That Change the Choice

The single most consistent finding across every agent is that earlier treatment predicts better outcomes. In ENDEAR, infants treated within 12 weeks of symptom onset had markedly higher milestone-response rates than those treated later.

Presymptomatic treatment performs best of all. In gene-therapy and risdiplam presymptomatic cohorts, many children achieved age-appropriate milestones, including sitting and standing, that are essentially never reached in untreated type 1 disease. This underpins the case for newborn screening, which shortens the window from diagnosis to dosing.

SMN2 copy number also shapes expectations: more copies generally predict a milder baseline phenotype and greater room for gain. Age, weight, and prior anti-AAV9 antibody status further constrain which route is feasible — gene therapy is restricted to younger, lighter, seronegative children, whereas the oral and intrathecal options span a wider age range.

What the Evidence Does Not Show

Several clinically important questions remain unresolved by the pivotal trials.

No randomised trial directly compares nusinersen, gene therapy, and risdiplam, so claims that one is superior rest on cross-trial inference rather than direct evidence. Long-term durability of gene replacement beyond the current follow-up, and whether a single infusion suffices for life, are not yet fully answered.

The value of switching or combining SMN-targeted agents is not established from controlled data. For apitegromab, evidence is limited to a defined non-ambulant type 2/3 population on background therapy; benefit in ambulant patients or as measured over longer horizons is not yet demonstrated. Its regulatory status also remains unsettled.

Practical Implications

The evidence supports, rather than dictates, several practice patterns. Because earlier treatment consistently predicts better outcomes, prompt genetic diagnosis and access to therapy appear central to results.

For infantile-onset disease identified early, all three SMN-raising agents have evidence of benefit, and the choice often turns on delivery burden, age and weight eligibility, and anti-AAV9 serostatus rather than on a demonstrated efficacy gap. For later-onset disease, the two oral and intrathecal options have randomised support for slowing or reversing motor decline.

Muscle-targeted therapy is best understood, on current evidence, as a potential add-on to SMN treatment for residual weakness rather than a standalone option — and, importantly, as investigational. Drug dosages and eligibility should be verified against current prescribing information.

Article Information and References