New Treatments14 September 2026

Three ways to treat spinal muscular atrophy

In less than four years, a disease with no treatment gained three, each using a different technology. SMA has become a case study in how modality shapes access.

By NextGenHealth Editorial Team2 min read

A clinician listening to a baby’s chest with a stethoscope — Photo: CDC / Unsplash

In brief

Nusinersen, onasemnogene abeparvovec and risdiplam compared: three SMA treatments, three technologies, and why starting before symptoms matters most.

Nusinersen, approved in 2016, is an antisense drug given into the spinal fluid.
Onasemnogene abeparvovec, approved in 2019, is a one-time gene therapy.
Risdiplam, approved in 2020, is an oral small molecule taken daily.

Spinal muscular atrophy (SMA) is caused by loss of the SMN1 gene, which makes a protein that motor neurons need to survive. In its most severe form, babies lose the ability to move, swallow and breathe, and historically most did not survive early childhood.

Humans have a backup gene, SMN2, that makes a small amount of working protein. All three approved SMA treatments exploit that biology in different ways.

Nusinersen: an antisense drug

Approved by the FDA in December 2016, nusinersen is an antisense oligonucleotide that changes how SMN2 is spliced so that it produces more full-length protein. It does not cross into the brain and spinal cord from the blood, so it is injected into the spinal fluid through a lumbar puncture, with loading doses followed by maintenance doses every four months.

Onasemnogene abeparvovec: a gene therapy

Approved in 2019 for children under two, onasemnogene abeparvovec uses an AAV9 vector to deliver a working copy of SMN1 in a single intravenous infusion. Its launch price made it one of the most expensive medicines ever sold, and it put outcomes-based and instalment payment models on the agenda.

Risdiplam: an oral small molecule

Approved in 2020, risdiplam also modifies SMN2 splicing, but it is a small molecule taken by mouth every day. It reaches tissues throughout the body and avoids repeated lumbar punctures. Its label was later extended to newborns.

Time is everything

Across all three, results are best when treatment starts before symptoms appear, because lost motor neurons do not come back. That evidence drove the addition of SMA to newborn screening programmes in the US and a growing number of other countries.

Clinicians now face questions that did not exist a decade ago: which therapy to use first, whether combining them helps, and how to judge long-term benefit in children who would not previously have survived to be studied.

Spinal muscular atrophy
Antisense oligonucleotides
Gene therapy
Newborn screening

More in New Treatments. Pipelines, approvals and first-in-child therapies, from small molecules to gene editing.

Treatments

Gene therapy’s pipeline is overwhelmingly paediatric

Many of the conditions that one-time genetic treatments target first appear in infancy. That puts children at the centre of the field’s scientific, commercial and ethical debates.
Treatments

Bespoke medicines: genetic therapies designed for a single child

Doctors have begun designing genetic medicines for one patient at a time. The science is moving faster than the regulatory and funding models built around it.
Treatments

CAR-T began with children: the paediatric origins of cell therapy

The first CAR-T therapy approved anywhere was for children and young adults with leukaemia. That history still shapes how the field thinks about paediatric patients.
Treatments

CFTR modulators and the push to treat cystic fibrosis earlier

Drugs that correct the underlying protein defect have changed cystic fibrosis. The next question is how early in life they should start.
Treatments

Gene editing reaches the clinic: what CRISPR therapy means for younger patients

The first approved CRISPR-based medicine treats sickle cell disease from age 12. Extending it to younger children raises questions about conditioning, fertility and timing.
All New Treatments 6 articles

Safety

Why long-term follow-up matters more when the patient is a child

A medicine given to a child can affect growth, development and fertility years after the last dose. Trials end long before those effects would show.
Policy

Why paying for paediatric medicines is different

Health technology assessment was built around adult trials and adult lives. Children’s medicines strain almost every assumption it relies on.
Conditions

Duchenne muscular dystrophy and the accelerated approval debate

No paediatric disease has tested the FDA’s accelerated approval pathway like Duchenne. The arguments over dystrophin, function and risk are far from settled.
Policy

Why children still wait years for medicines adults already have

Two decades of incentives and mandates on both sides of the Atlantic have transformed paediatric drug development. They have not closed the gap.
Medicines

The formulation problem: making adult drugs work for small bodies

A tablet that suits a 70kg adult is often useless for a toddler. Getting the dose, the taste and the ingredients right is a development programme in its own right.
Evidence

Extrapolation is changing how paediatric trials are designed

Regulators increasingly accept that adult efficacy data can carry over to children when the disease and the drug’s effect are similar enough. The argument is about how much.