Ionis Pharmaceuticals and Otsuka Pharmaceutical reported positive topline results on 22 September 2026 from the Phase 3 FUSION study. The trial tested ulefnersen in amyotrophic lateral sclerosis caused by mutations in the fused in sarcoma gene. Ionis describes the drug as the first potential disease-modifying FUS-ALS treatment. John Kraus, Chief Medical Officer at Otsuka, calls FUSION the first FUS-ALS trial to meet its primary endpoint.
The announcement carries one statistic. Ionis reported a statistically significant improvement on the primary endpoint at p=0.0005 in 73 participants. No effect size, confidence interval, median or event count accompanies it. On the same day Otsuka opened a global early access programme. For a trade readership the open question is not whether the trial worked. It is what a regulator or a payer can do with a rank statistic and no published numbers.
What the FUSION study reported
FUSION is a global multicentre, randomised, double-blind, placebo-controlled Phase 1-3 trial. Part 1 ran a 72-week double-blind period. Part 2 is an open-label extension in which all participants receive ulefnersen. The primary analysis covers the Part 1 cohort only.
The releases describe the primary endpoint in two ways. The first is a joint rank analysis of three measures. Those are time to death or permanent ventilation and time to rescue. The third is change in the revised ALS Functional Rating Scale to Day 505. The second gives the endpoint as functional impairment and survival at 72 weeks. That wording gives the third component as ventilation assistance-free survival. LSDN publishes both constructions rather than choosing between them.
Secondary endpoints reaching statistical significance included change from baseline in serum neurofilament light chain. They also included time to the earliest of death, permanent ventilation, rescue, or withdrawal due to disease progression. Neither release says whether any other secondary endpoint reached significance. Results for the other listed measures were not given. They include slow vital capacity, handheld dynamometry, quality of life and cerebrospinal fluid markers.
A registry mirror of the record for NCT04768972 gives further design detail. It lists actual enrolment of 89 participants and primary completion on 17 June 2026. It records roughly 25 sites across 16 countries, including the United Kingdom. Eligibility opened at age 10 with a confirmed FUS variant. Dosing is an intrathecal bolus every 12 weeks after a loading schedule.
What the announcement does not quantify
LSDN applies the same test to positive readouts as to negative ones. On this release the missing detail is extensive. Neither company published the functional rating scale change in either arm. Neither published a hazard ratio or confidence interval for survival. Neither gave the number of deaths, ventilation events or rescue events. Neither published the size of the neurofilament reduction. Neither gave a p value for any secondary endpoint. Neither disclosed baseline characteristics or the dose.
The registry figure of 89 enrolled participants also sits above the 73 in the primary analysis population. Neither release explains the difference, and LSDN draws no inference from it. Both companies say detailed results will go to a medical congress and to a peer-reviewed journal. LSDN named the FUSION readout as a near-term catalyst in its review of the 2026 neurodegeneration pipeline.
How this differs from the SOD1 precedent
Tofersen is the nearest precedent. The ALS Association lists it as the only genetically targeted therapy among the FDA-approved ALS medicines. Its United States label states that approval rested on reduction in plasma neurofilament light chain. It adds that continued approval “may be contingent upon verification of clinical benefit in confirmatory trial(s)”. That is a biomarker-based accelerated approval. Biogen reported in October 2021 that the Phase 3 VALOR study missed its primary endpoint. That endpoint was change in the functional rating scale at 28 weeks.
FUSION is different in kind. Its primary endpoint is a composite of function and survival, not a biomarker. Holly Kordasiewicz is Executive Vice President and Chief Development Officer at Ionis. She called ulefnersen the first investigational medicine to show a statistically significant Phase 3 benefit. The claim is specific to a prespecified joint-rank analysis. LSDN reports that as a company claim and takes no position on it.
The distinction matters to anyone modelling a FUS-ALS treatment. A rank-based composite establishes that the treated group did better. It does not by itself produce the absolute difference a health technology assessment body needs. Otsuka says it will review the results with the US Food and Drug Administration. It will also continue discussions with other health authorities on potential expedited submission pathways. No filing date has been given.
The safety record so far
Both releases describe a favourable profile, with most adverse events mild or moderate in severity. No rates, serious adverse event counts or discontinuation figures were published. LSDN treats that as an absence of disclosure, not as a finding.
Two other datasets carry more detail. The investigator-initiated jacifusen case series, published in The Lancet on 22 May 2025 by Neil A. Shneider and colleagues, treated 12 patients aged 16 to 45. The authors described the drug as generally well tolerated. Reported events included back pain, headache, nausea and post-lumbar puncture headache. Two deaths occurred and were reported as unrelated to treatment.
The second dataset belongs to a different molecule. The tofersen US prescribing information carries three warnings. They cover myelitis and radiculitis, papilloedema and elevated intracranial pressure, and aseptic meningitis. Its most common adverse reactions include pain at 42 per cent and fatigue at 17 per cent. That is a separate product and a separate programme, and the two were never designed for comparison. It is still the nearest licensed reference point for a long-term intrathecal FUS-ALS treatment regimen.
Otsuka opens an early access route
The second release of the day sets out an access route. Otsuka announced a global early access programme for ulefnersen on 22 September 2026. It is a company-run FUS-ALS treatment route outside the trial. It follows earlier investigator-led expanded access, under which the jacifusen case series was run. Eligibility requires a confirmed genetic diagnosis of FUS-ALS. Applicants must also be unable to join an ongoing clinical trial.
Requests must come from the treating physician through the Otsuka early access portal. The company says it does not accept requests direct from patients or carers. It also says timing from application to treatment will vary. The stated variables are regulatory approvals, ethics clearance, centre readiness and patient-specific requirements.
That wording matters in England. A global programme is a company offer, not a national supply route. Each country applies its own import, ethics and funding arrangements. NICE records that tofersen is offered through a company-sponsored early access programme at a number of NHS Trusts. LSDN found no MHRA or NHS England statement on ulefnersen supply as at 22 September 2026. LSDN draws no conclusion from that.
What a FUS-ALS treatment would face in England
Three gates sit between a positive trial and an English prescription. The first is diagnosis, because no FUS-ALS treatment can reach a patient whose variant is unknown. NHS England’s National Genomic Test Directory covers this group through panel R460, amyotrophic lateral sclerosis, delivered by whole genome sequencing. The NHS Genomics Education Programme states that all patients with confirmed motor neurone disease are eligible. It names FUS alongside SOD1 and TARDBP among the genes investigated.
The clinical guideline has not kept pace. NICE guideline NG42 on motor neurone disease dates from 2016. On LSDN’s reading of the recommendations, it contains no recommendation on genetic testing, genetic counselling or referral to clinical genetics. Recommendation 1.2.7 lists “Will my children get MND?” among the topics to discuss. The guideline anticipates the question and offers no pathway. That is an LSDN observation about the guideline, not a NICE position on any product.
The second gate is health technology assessment, and the tofersen file offers the nearest benchmark. The MHRA authorised tofersen on 22 July 2025 through the International Recognition Procedure. NICE is still evaluating it. Project GID-HST10062 is listed as in progress on the Highly Specialised Technologies standard route. The topic was selected on 18 November 2024. NICE records that, following discussions with the company, the evaluation was due to start in late March 2026. The first committee meeting is scheduled for 22 October 2026 and publication is expected on 17 March 2027. That is roughly 20 months from UK authorisation to expected guidance, part of it set by company timing.
The third gate is scale. The MND Association reports more than 5,000 adults living with the condition in the UK. It gives six new diagnoses a day and a lifetime risk of about one in 300. Otsuka puts FUS-ALS at an estimated 0.6 per cent of all ALS cases. The International Alliance of ALS/MND Associations gives a range of 0.3 to 0.9 per cent. Applying those shares to the UK prevalence figure implies a candidate population in the low tens. LSDN offers that only as arithmetic on two published figures, since the sources count different populations.
Why the age profile complicates eligibility
Otsuka states that FUS-ALS accounts for an estimated 43 to 52 per cent of juvenile and paediatric ALS cases. It describes earlier onset and rapid progression. In early-onset and juvenile cases it reports respiratory failure and death often within one to two years of symptom onset. FUSION enrolled from age 10. The European Medicines Agency agreed a paediatric investigation plan on 15 May 2023. That decision included a deferral and a waiver.
England’s genomic testing route for this group sits in the adult-onset neurodegenerative pathway. LSDN found no published NHS policy on how a juvenile presentation reaches the same panel. LSDN draws no conclusion about current practice. Any paediatric FUS-ALS treatment decision would turn on that question.
Commercial terms and what comes next
Otsuka licensed ulefnersen from Ionis under an agreement announced on 22 November 2024. The upfront payment was US$10 million, with regulatory and sales milestones and royalties to follow. Otsuka said it would apply for regulatory approvals and sell the product worldwide. The milestone amounts have not been disclosed.
Ionis is running a separate neurology launch of its own. The FDA approved its antisense medicine Zanvastro for Alexander disease on 3 September 2026. Ulefnersen itself holds fast track and orphan drug designation from the FDA. It holds orphan designation from the European Medicines Agency and from Swissmedic. Ionis also sponsored a retrospective natural history chart review in FUS-ALS with UK sites. The Health Research Authority record, reference 23/YH/0283, gives a favourable opinion on 17 January 2024. It reported documented mortality in 63 per cent of participants, averaging 32 months from symptom onset.
The near-term tests for a licensed FUS-ALS treatment are specific. They are the congress presentation and the peer-reviewed paper, the regulatory route Otsuka selects, and whether a UK filing follows. Until the effect sizes are published, the case for the first FUS-ALS treatment rests on company assertion.
