Biology Atlas

Map / Drugs

drug · intermediate · complete

Zolgensma (onasemnogene abeparvovec)

A one-time intravenous gene therapy that delivers a working SMN gene inside an AAV9 shell to infants with spinal muscular atrophy.

Also called: onasemnogene abeparvovec, AVXS-101

This describes published medicine for readers who want to understand how it works and what happened. It is not medical advice and says nothing about what any person should take.

What it is and what happened

Zolgensma is a one-time infusion of a harmless-shell virus carrying a working copy of the SMN gene, for babies with spinal muscular atrophy. In the pivotal trial 19 of 21 infants were alive without permanent ventilation, against roughly one in four surviving past 14 months in natural history. It is approved, but the label warns of liver failure. 1

Reported, not yet replicated

The decisive evidence more

Phase 1 (abstract read): 15 infants, 3 at a low dose and 12 at a high dose, all alive and event-free at 20 months against 8% in a historical cohort; of the 12 on the high dose, 11 sat unassisted and 2 walked.

Label Study 1: 21 infants, mean age 3.9 months; 19 alive without permanent ventilation at data cutoff, 13 of the 19 had reached 14 months, and 10 of 21 (47.6%) sat for 30 seconds. The label puts natural history at about 25% surviving past 14 months with none sitting. The low-dose cohort, about a third of the dose, had no one sitting; high dose 9 of 12.

EMA (news item of 27 March 2020 recommending conditional authorisation): 22 patients, 20 (91%) alive without permanent ventilation at 14 months and 14 (64%) sitting before 18 months. Phase 3 STR1VE-EU: 14 of 32 (44%) sat independently for at least 10 seconds by 18 months versus 0 of 23 natural-history patients; 31 of 32 (97%) survived free of permanent ventilation at 14 months versus 6 of 23 (26%). 2134

Reported, not yet replicated

Why it worked more

Four things line up. The disease is the loss of one gene, so supplying one gene copy is the right repair. The target, motor neurons, no longer divides, and AAV genomes are lost mostly when cells divide, so the effect lasts. The AAV9 capsid reaches the nervous system after an intravenous dose. And treatment is early, before most motor neurons have died. Added to this, the label wraps the dose in an antibody screen, steroid cover and liver tests.

The label itself does not explain why durability follows; the episome-and-division account is the standard explanation from the AAV review, not a statement of the label. 15

Reported, not yet replicated

Dose, units and the restated phase 1 doses deep

The label dose is 1.1 x 10^14 vg/kg. The phase 1 paper gave 6.7 x 10^13 (low) and 2.0 x 10^14 (high) vg/kg as dosed at the time. FDA called the early titre assay inaccurate and imprecise; re-measured 44 months later, the doses were restated as 3.7 x 10^13 and 1.1 x 10^14 vg/kg, and FDA could not determine the doses actually given (as our own earlier report cites the FDA review; not re-opened here). Doses of different products are also not comparable by number: vector genomes of AAV and particles of adenovirus are different units, and a number per kg says little across vectors. 26

Reported, not yet replicated

Persistence and immune memory deep

There is no off switch after dosing. After infusion, anti-AAV9 titres rose to at least 1:102,400 in every patient in Study 2, and the label says re-dosing has not been evaluated. Random integration has been reported and tumours have been reported after approval; the label states no causal link has been established. Whether this is coincidence or a class effect is open. 1

Reported, not yet replicated

Where the evidence is thin deep

The pivotal comparison is single-arm against a natural-history cohort, so differences in age at treatment, SMN2 copy number and supportive care can mislead; the checks are the dose-response between low and high dose and the replication in STR1VE-EU, which also used natural history. Long-term safety is stated by the authors of STR1VE-EU as needing further study. 14

Reported, not yet replicated

Modality

Gene therapy

Targets

Pathways

Indications

Identifiers

Nonproprietary name

onasemnogene abeparvovec

Vector

self-complementary AAV9

Mechanism

A self-complementary AAV9 capsid carries SMN cDNA under a cytomegalovirus-enhancer / chicken beta-actin hybrid promoter, a strong near-ubiquitous switch. Self-complementary means the DNA folds into a double strand, so expression does not wait for a slow conversion step. The label states that the DNA persists largely in episomal form, a loose circle in the nucleus outside the chromosomes, with random integration reported. 1

Reported, not yet replicated

Delivery route

Must survive

The dose is one 60-minute intravenous infusion of 1.1 x 10^14 vector genomes per kg. The vector must avoid neutralising anti-AAV9 antibodies (the trials required baseline titres of 1:50 or lower and the label says to test), cross from blood into the nervous system, and enter motor neurons. Because the infusion is systemic, much of the dose also reaches the liver. 1

Reported, not yet replicated

Side effects

Acute liver injury and liver failure (boxed warning)

The label carries a boxed warning for acute serious liver injury or acute liver failure; liver monitoring is required (weekly for the first month, then every other week, per the label). 1

Reported, not yet replicated

Fever

Pyrexia in 22 of 33 infants (67%) in STR1VE-EU. 4

Reported, not yet replicated

Upper respiratory infection

11 of 33 infants (33%) in STR1VE-EU. 4

Reported, not yet replicated

Raised alanine aminotransferase

9 of 33 infants (27%) in STR1VE-EU. In phase 1, liver enzymes rose in 4 of 15 infants and were attenuated by prednisolone. 42

Reported, not yet replicated

Serious adverse events judged related to the drug

6 of 33 (18%) in STR1VE-EU. 4

Reported, not yet replicated

Thrombotic microangiopathy, thrombocytopenia, troponin rise, tumour reports

Listed in the current label. Tumours have been reported after approval; the label says a causal link is not established. 1

Reported, not yet replicated

Verdict

Outcome

worked

Why

Treated infants survived and gained motor milestones far beyond the natural history of untreated type 1 disease, in a single-arm trial against historical comparators, with dose-response and a second trial in another region pointing the same way. It works because the disease is a single-gene loss, the target cells are non-dividing motor neurons that keep episomes, the capsid reaches them, treatment comes before most neurons are lost, and steroid cover and liver monitoring blunt the main toxicity. The same label carries a boxed liver warning, so the benefit is real and the toxicity wall is still there. 1243

Reported, not yet replicated

Teach this

In one sentence

Zolgensma works because a single missing gene was replaced in non-dividing nerve cells, by a shell that reaches them, early enough to matter, with a safety net around the dose. 1

Reported, not yet replicated

The five-minute version

A virus shell (AAV9) is emptied and filled with the SMN gene plus a strong switch. One infusion sends it through the blood to motor neurons, where the gene sits as a loose DNA circle and keeps making the protein for as long as the cell lives and does not divide. Infants given it in trials survived and sat when untreated infants mostly did not. The cost: the liver sees a large dose too, so the label has a boxed warning, and the dose cannot be taken back or easily repeated. 14

Reported, not yet replicated

What a listener will ask

Why not just redose if it fades?

The immune system makes antibodies to the shell; after treatment titres were very high and the label says re-dosing is not evaluated.

Is it a cure?

It is a lasting replacement of one gene product, not an edit of the chromosome. Durability beyond the follow-up so far is not established.

Why did a similar idea kill someone years earlier?

A different virus, a different dose-immune relationship and weaker process discipline; see the adenovirus OTC trial record.

The common wrong idea

Gene therapy is a one-time fix that edits your DNA.

This one adds a working copy that mostly stays outside the chromosomes, so it fades in dividing tissue, cannot be switched off, and was tested against an immune system that reacts to the shell.

The one citation to hand over: 1

Still open here

Not yet written: Assays used, Off targets.

Linked from

Sources

  1. US Food and Drug Administration. ZOLGENSMA (onasemnogene abeparvovec-xioi) prescribing information, revision 6/2026. FDA (2026). regulatory, not peer reviewed; opened 2026-11-17 all records citing this
  2. Mendell JR et al.. Single-Dose Gene-Replacement Therapy for Spinal Muscular Atrophy. New England Journal of Medicine (2017). paper; opened 2026-11-17 all records citing this
  3. European Medicines Agency. EMA news item: new gene therapy to treat spinal muscular atrophy (CHMP recommendation of conditional authorisation). EMA (2020). regulatory, not peer reviewed; opened 2026-11-17 all records citing this
  4. Mercuri E, Muntoni F, Baranello G, et al.. Onasemnogene abeparvovec gene therapy for symptomatic infantile-onset spinal muscular atrophy type 1 (STR1VE-EU): an open-label, single-arm, multicentre, phase 3 trial. The Lancet Neurology (2021). paper; opened 2026-11-17 all records citing this
  5. Wang D, Tai PWL, Gao G. Adeno-associated virus vector as a platform for gene therapy delivery. Nature Reviews Drug Discovery (2019). review; opened 2026-11-17 all records citing this
  6. US Food and Drug Administration. FDA Summary Basis for Regulatory Action: Zolgensma (2019). FDA (2019). regulatory, not peer reviewed; opened 2026-11-17 all records citing this
  7. US Food and Drug Administration. FDA product page: Zolgensma. FDA (2019). regulatory, not peer reviewed; opened 2026-11-17 all records citing this
Where it sits on the map
History of this record (6)

This record has been changed 6 times, on the dates below. The note kept against each change is written for the people who maintain the atlas, in their own shorthand, so it is not shown here; it is in the record itself.

Year
2019
Regulatory status
approved
Status
complete
Last updated
2026-11-19
Record
content/drugs/zolgensma.json