Diagnosis-led treatment

Antisense oligonucleotides (ASO)

Splice-switching oligonucleotides, exon skipping, gapmers, TANGO, n-of-1 ASOs

The platform that has delivered more licensed medicines in child neurology than any other gene-directed approach — and the only one flexible enough to be built for a single child.

Where it stands, condition by condition

The same therapy can be well supported for one problem and completely untested for another. This is the single most common place families are misled.

What it is

An antisense oligonucleotide is a short synthetic strand of chemically modified DNA or RNA, typically 15 to 25 letters long, designed to stick to one specific RNA message inside the cell.

It is not gene therapy. No gene is added, removed or rewritten. The DNA is left exactly as it was; what changes is how one message is read, and only for as long as the drug is present.

Because the active ingredient is a sequence, the manufacturing process is the same whatever disease is targeted. That is the single most important fact about this platform: once a variant is known, designing the molecule is comparatively fast, which is why ASOs reached the clinic first and why they are the only route realistically open to a child whose variant is unique to them.

Almost all neurological ASOs are given by lumbar puncture into the spinal fluid, because these molecules cannot cross from the blood into the brain. Dosing is repeated — every four months for several of them — and that is both the main burden and the main safety advantage.

How it is meant to work

Four distinct jobs, often confused with one another:

Splice switching — adding protein. The strand covers a site the cell uses when cutting and joining the RNA message, forcing it to include an exon it would otherwise discard. Nusinersen works this way in SMA; so does the TANGO approach in Dravet, where the healthy SCN1A copy is made more productive.

Exon skipping — restoring a reading frame. The opposite manoeuvre: a damaged exon is deliberately left out so the rest of the message can still be read. The protein is shorter but functional. This is the Duchenne strategy, and it is mutation-specific — a drug that skips exon 51 does nothing for a child who needs exon 53 skipped.

Knockdown — removing protein. A gapmer recruits an enzyme that destroys the target message. This is the strategy where the problem is too much of something, or a toxic product: zilganersen lowering GFAP in Alexander disease, tofersen lowering mutant SOD1.

Which one a child needs depends on the variant, not the diagnosis. In a gain-of-function variant you want less; in a loss-of-function variant you want more. Giving the wrong direction would be expected to make things worse, which is why the functional interpretation on a genetic report matters as much as the gene name.

What has actually been tested

• Spinal muscular atrophy is the proof. Nusinersen changed the natural history of a condition that previously had none, and together with newborn screening it reshaped what the diagnosis means. Everything that followed on this page followed from that.
• Alexander disease is the most recent and in some ways the most striking: a condition with no treatment at all until September 2026, now with a licensed medicine that met a walking-speed endpoint in a controlled trial.
• The Duchenne exon-skippers are the cautionary case. They were licensed on a laboratory marker — how much dystrophin protein appeared — rather than on how children walked. Years later the size of the real-world benefit is still argued over. A regulatory approval is not the same as a settled clinical question.
• The pattern across all of them: benefit depends on starting before damage accumulates. These drugs protect neurons that are still alive. None restores what has already been lost.

What we still do not know

• How long treatment must continue. Nobody has stopped these drugs in a stable child to find out, and for most there is no reason to think the effect persists after dosing ends.
• Whether starting later still helps, and how much. Trials recruit the children most likely to show a measurable change.
• How to get a molecule designed for one child through a system built for medicines given to thousands — the central problem the n-of-1 field is working on.
• Long-term effects of repeated intrathecal dosing through childhood, since the first treated children are only now reaching adolescence.

Risks and unwanted effects

• The lumbar puncture, repeated for years. Post-puncture headache, back pain, and in young children the need for sedation each time. For a child with scoliosis or spinal rods the procedure can become genuinely difficult, and some need imaging guidance or an implanted port.
• Raised protein in the spinal fluid is common across this class and is usually without clinical consequence, but it needs monitoring.
• Kidney and platelet effects are watched for with some agents; the monitoring schedule is part of the prescription, not an optional extra.
• The reversibility is a genuine safety advantage. If something goes wrong, dosing stops and the effect washes out. Gene replacement does not offer that.
• The practical risk for families is different: these are specialist-centre treatments, and access depends on licensing and funding in each country rather than on whether the child would benefit.

Questions to ask before you agree

1. Does my child's variant suit this approach — and in which direction, more protein or less?
2. Is the specific drug licensed for my child's gene and age, or is this a trial?
3. How often is the lumbar puncture, will sedation be needed, and who performs it?
4. What is measured before and after, and at what point would we conclude it is not working?
5. If there is no programme for this gene, is there a registry or a natural-history study we should join now, so we are ready if one opens?

More in this section