Biology Atlas

Map / Drugs

drug · intermediate · complete

Nusinersen (Spinraza)

Nusinersen is an antisense oligonucleotide injected into the spinal fluid that changes how the SMN2 gene is spliced so that children with spinal muscular atrophy make more full-length SMN protein.

Also called: Spinraza

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.

In brief

Nusinersen is a short chemically armoured strand that sticks to one RNA in spinal muscular atrophy and changes how it is spliced, so a missing protein is made. It is injected into the spinal fluid and lasts for months. In infants it beat a sham procedure: 51 percent of treated infants met the motor-milestone endpoint against 0 percent of controls. 12

Reported, not yet replicated

What the decisive trial showed more

ENDEAR was randomised, double-blind and sham-controlled in infantile-onset SMA. The interim analysis stopped the trial early: motor-milestone response 21 of 51 (41 percent) versus 0 of 27, P below 0.001. The final numbers were 37 of 73 (51 percent) versus 0 of 37, with a hazard ratio of 0.53 for death or permanent ventilation and 0.37 for death. Earlier treatment did better. Only the abstract was read.

The FDA label for 2016 gives the same endpoint as 21 of 52 (40 percent) versus 0 of 30, because the regulator used its own eligible set; the two counts were not reconciled in the sources opened. 12

Reported, not yet replicated

How it works: switching a splice, not cutting an RNA more

The cell makes RNA from SMN1 and SMN2. SMN1 is lost in the disease. SMN2 is nearly identical but mostly skips exon 7, so the protein is short-changed. Nusinersen binds a sequence in the intron after exon 7 of SMN2 pre-mRNA (label section 11). The trial protocol describes an intronic silencer there (ISS-N1) and up to about 90 percent skipping of exon 7. Covering the silencer lets exon 7 be included, and the NEJM abstract states that nusinersen promotes full-length SMN protein.

Nothing is destroyed. Every position of the strand is modified, so the cutting enzyme RNase H cannot engage (this is the standard design logic, described in the review). It is a steric blocker. 2314

Reported, not yet replicated

Why it worked, and what it costs more

It worked in a disease with a known cause and a drug that reaches the compartment where the neurons are. The long half-life in the spinal fluid (135 to 177 days) is why the label spaces doses months apart; this describes the 2016 label and is not guidance for anyone.

The costs are class effects of the backbone. In the label, 6 of 56 treated patients (11 percent) dropped below the normal platelet range versus 0 of 28 sham, and raised urine protein occurred in 17 of 51 (33 percent) versus 5 of 25 (20 percent). Both are monitored. 2

Reported, not yet replicated

Chemistry and pharmacokinetics in detail deep

The label describes an 18-mer with phosphorothioate linkages and 2'-O-methoxyethyl sugars; the review lists nusinersen as a fully modified steric blocker, unlike the gapmers that keep a DNA centre. Mean terminal half-life is 135 to 177 days in CSF and 63 to 87 days in plasma. It is metabolised by exonucleases, not CYP450 enzymes. Labelled schedule in 2016: four loading doses, then one every four months; later label revisions changed regimens and were not opened, so any current schedule is unverified here.

The tiling screen that found the optimal oligos (31 screened, two strongly raised exon 7 inclusion in mouse liver and kidney) is a sequence-to-activity dataset of the same shape as a QSAR table. That the intron-7 silencer of the screen is the site nusinersen covers is the accepted reading, but the label and abstracts opened do not state it in those words. 2453

Reported, not yet replicated

Where the story is thinner than it looks deep

The pivotal evidence read here is the ENDEAR abstract and the 2016 label; the full paper, the regulatory review documents and later trials were not opened. Mouse and tissue-distribution data (the autopsy series is three people) say where the drug goes, not how much of it is needed. The route matters: plasma levels say little about brain and cord levels. Why exon 7 skipping leads to an unstable protein was not in the sources opened. 215

Reported, not yet replicated

Modality

Antisense oligonucleotide (ASO)

Targets

Pathways

Indications

Identifiers

Endear trial

NCT02193074

Fda label first approval

2016

Mechanism

The SMN1 gene is lost or defective in spinal muscular atrophy. A near-copy gene, SMN2, mostly leaves out exon 7 when its RNA is spliced, up to about 90 percent of the time according to the trial protocol, so little full-length SMN protein is made. Nusinersen is an 18-letter oligonucleotide, fully modified with phosphorothioate links and 2'-O-methoxyethyl (MOE) sugars, that binds a sequence in the intron just after exon 7 of SMN2 pre-mRNA. It does not cut the RNA. By covering a splicing silencer it lets exon 7 be included, so more full-length SMN protein is made. It turns a gene up, the opposite of a cutting (gapmer) antisense drug. 2341

Reported, not yet replicated

Delivery route

Must survive

Nucleases (the phosphorothioate and MOE chemistry is the protection). The blood-brain barrier is bypassed rather than crossed: the drug is injected by lumbar puncture into the cerebrospinal fluid. In the label pharmacokinetics, plasma trough levels are low compared with CSF, and drug distributes through the central nervous system and also to muscle, liver and kidney (autopsy, three people). It is removed by exonuclease metabolism, not by CYP450 enzymes. 24

Reported, not yet replicated

Assays used

Antisense tiling screen of the target region

Early work screened 31 antisense oligonucleotides along the SMN2 region and found two optimal ones that strongly raised human SMN2 exon 7 inclusion in transgenic mouse liver and kidney. This is a sequence-to-activity map: the molecule is a string and the readout is a splicing outcome. Only the abstract was read. 5

Reported, not yet replicated

Sham-controlled randomised trial with motor-milestone endpoint

A drug given by lumbar puncture is blinded with a sham procedure (a needle prick without injection). The decisive endpoint was the proportion of infants who met motor-milestone criteria. 1

Reported, not yet replicated

Off targets

Off-target RNA binding

No off-target RNA is named in the label sections or abstracts read for this record. The class toxicities that were observed are listed under side effects. Whether sequence-matched off-target RNAs matter for this drug was not established from the sources opened. 2

Unverified: from our own report, source not opened

Side effects

Low platelets and coagulation abnormalities

In the 2016 label, 6 of 56 treated patients (11 percent) fell below the lower limit of normal for platelets versus 0 of 28 sham-control patients; none fell below 50,000 per microlitre. The label warns about coagulation abnormalities and thrombocytopenia and asks for testing. 2

Reported, not yet replicated

Renal toxicity (urine protein)

In the 2016 label, raised urine protein occurred in 17 of 51 treated patients (33 percent) versus 5 of 25 sham-control patients (20 percent); the label asks for urine protein testing. 2

Reported, not yet replicated

Adverse events overall in ENDEAR

The paper abstract states that adverse events were similar between the treated and sham groups. 1

Reported, not yet replicated

Verdict

Outcome

worked

Why

In the randomised sham-controlled ENDEAR trial in infantile-onset SMA, the interim analysis stopped the trial early at a motor-milestone response of 21 of 51 (41 percent) versus 0 of 27 (0 percent), P below 0.001. The final analysis gave 37 of 73 (51 percent) versus 0 of 37 (0 percent); the hazard ratio for death or permanent ventilation was 0.53 (P=0.005) and for death 0.37 (P=0.004), and infants with shorter disease duration did better. It worked because the target was a known, causal RNA event; the drug was injected beside the sick motor neurons; the half-life in that compartment is long; and untreated disease is bleak, so a flat 0 percent control made the effect easy to see. The 2016 label counts its interim set slightly differently (21 of 52, 40 percent versus 0 of 30); the sources do not reconcile the counts. 12

Reported, not yet replicated

Teach this

In one sentence

Nusinersen is an RNA-level drug that covers a splicing silencer on SMN2 so the cell makes the full-length protein it was missing. 21

Reported, not yet replicated

The five-minute version

Spinal muscular atrophy is caused by loss of the SMN1 gene. A backup gene, SMN2, mostly skips exon 7 so it makes little working protein. Nusinersen is a chemically armoured strand that pairs with a site in SMN2 pre-mRNA and covers a splicing silencer, so exon 7 is kept and full-length SMN rises. It cannot cross the blood-brain barrier, so it is injected into the spinal fluid, where its half-life is months. In a sham-controlled trial in infants, 37 of 73 treated infants met the motor-milestone endpoint against 0 of 37 controls. The costs are class effects: platelet and kidney monitoring. 213

Reported, not yet replicated

What a listener will ask

Why inject it into the spine?

It is a large charged polymer that does not cross the blood-brain barrier well, so it is placed in the spinal fluid beside the motor neurons (label pharmacokinetics: low plasma, high CSF). 21

Reported, not yet replicated

Does it cut the RNA?

No. It is fully modified so the RNA-cutting enzyme RNase H cannot work on it; it only blocks a splicing silencer. 21

Reported, not yet replicated

Why was the control a sham?

A lumbar-puncture drug can only be blinded by a sham procedure; with untreated disease so severe, a 0 percent responder control made the effect clear. 21

Reported, not yet replicated

The common wrong idea

Antisense drugs work by destroying a bad RNA.

Some do, but nusinersen does the opposite: it leaves the RNA intact and changes how it is spliced so more of a good protein is made.

The one citation to hand over: 1

Still open here

Linked from

Sources

  1. Finkel, R. S., et al.. Nusinersen versus sham control in infantile-onset spinal muscular atrophy (ENDEAR). New England Journal of Medicine (2017). paper; opened 2026-11-17 all records citing this
  2. U.S. Food and Drug Administration (label text supplied by the sponsor). SPINRAZA (nusinersen) injection, for intrathecal use: US prescribing information (initial US approval 2016, revised 12/2016). FDA Drugs@FDA label archive (2016). regulatory, not peer reviewed; opened 2026-11-17 all records citing this
  3. Sponsor-written trial protocol. Clinical trial protocol NCT02594124 (nusinersen extension study), full document. ClinicalTrials.gov large documents (2015). regulatory, not peer reviewed; opened 2026-11-17 all records citing this
  4. Egli, M., Manoharan, M.. Chemistry, structure and function of approved oligonucleotide therapeutics. Nucleic Acids Research (2023). review; opened 2026-11-17 all records citing this
  5. Hua, Y., et al.. Antisense masking of an hnRNP A1/A2 intronic splicing silencer corrects SMN2 splicing in transgenic mice. American Journal of Human Genetics (2008). paper; opened 2026-11-17 all records citing this
Where it sits on the map
History of this record (1)

This record has been changed 1 time, 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
2016
Regulatory status
approved
Status
complete
Last updated
2026-11-17
Record
content/drugs/nusinersen.json