FDA Approves Ultragenyx's Fayuvi, the First Treatment for Sanfilippo Syndrome Type A; Company Sets $3.95 Million Price
The agency granted standard approval on September 17, 2026, for a one-time gene therapy in children with MPS IIIA, based on a single-arm trial compared against untreated patients.
A Drug That Costs $3.95 Million, and a Disease That Doesn't Wait
For a child with Sanfilippo syndrome type A, the brain starts losing ground almost as soon as symptoms show up. Speech goes first, then thinking, then movement. Most children with the disease, also called MPS IIIA, die in their teens[2]. On September 17, 2026, the FDA approved the first treatment ever cleared for it in the United States: a one-time gene therapy called Fayuvi[1].
The same day, its maker, Ultragenyx Pharmaceutical, announced what it will charge: $3.95 million per patient, at list price[3][4]. Both facts are true at once, and neither cancels the other out. A disease with no clock-stopping treatment now has one. That treatment costs more than most people's houses.
Ultragenyx's stock jumped 10% to 13% on the news[9][7]. Somewhere between those two numbers, a genuine argument is happening about how much certainty $3.95 million actually buys.
How a Virus Delivers a Missing Instruction
Fayuvi is given once, through an IV. It uses a modified virus, harmless on its own, to carry a working copy of a gene called SGSH into a patient's cells[1]. That gene is the whole point.
Children with MPS IIIA are missing the instructions to make an enzyme called sulfamidase. Without it, a sugar-like molecule called heparan sulfate builds up inside brain cells and slowly damages them[2]. The therapy is meant to hand cells the missing instruction so they can make the enzyme themselves and start clearing that buildup.
The FDA had reviewed this exact drug before and said no. In July 2025, the agency sent Ultragenyx a rejection letter, but not because the drug failed to work. The issue was chemistry, manufacturing and quality-control problems at the plant[8]. Once those were fixed, the agency's second look ended in approval, along with Orphan Drug, Fast Track and Breakthrough Therapy status, all of which mark a drug as addressing a serious, poorly served condition[1].
The Comparison Nobody Loves but Everybody Uses
Here's the harder question, and it's the one running underneath nearly every version of this story: how do you prove a drug works when giving half the sick children a fake treatment would mean watching them decline on purpose?
Ultragenyx didn't run a standard trial with a treated group and a placebo group. It ran a single-arm trial, meaning every enrolled child got the drug, and then compared their results to a "natural history" record — data on how other children with MPS IIIA declined without any treatment[2]. Treated children scored 23.5 points higher on a cognitive test called the Bayley-III, measured between ages 2 and 5, with a p-value below 0.0001, a statistical measure indicating the result is very unlikely to be chance[2].
The FDA's own reasoning is that MPS IIIA follows a predictable enough downhill path to make that kind of comparison legitimate, and that withholding treatment from a dying child to run a cleaner trial raises its own ethical problem[2]. Notably, the agency didn't just wave the drug through on a lab marker. Ultragenyx had originally planned to seek a faster, conditional approval based on heparan sulfate levels in spinal fluid — a stand-in measurement. Instead, the FDA required the full, standard approval track, built on the cognitive test result[6][2].
Critics of that standard, including voices in medicine and health policy, don't argue the drug should be blocked. Their concern is narrower: kids who get into a trial early, because their families found a specialist fast, may already differ from the historical comparison group in ways that have nothing to do with the drug[6]. They also note the headline number comes from a small group — an earlier presentation to neurologists showed a similar 23.2-point effect in just 17 children treated youngest[6]. Their proposed fix isn't rejection. It's tighter, mandatory follow-up after approval, so the real-world benefit gets pinned down rather than assumed.
That debate had an unusually direct voice inside the FDA until recently. Vinay Prasad, who ran the division reviewing gene therapies and had previously criticized the agency's approval of a Duchenne muscular dystrophy gene therapy on similar grounds, left at the end of April 2026. Commissioner Marty Makary resigned on May 12[3]. The standard used to approve Fayuvi is, in a real sense, still being worked out in real time.
Who Actually Pays for a One-Time, Once-in-a-Lifetime Bill
The price tag traces back to simple arithmetic. Ultragenyx estimates roughly 3,000 to 5,000 patients worldwide live in places where the drug would realistically reach them[2]. Developing and manufacturing a gene therapy costs roughly the same no matter how many patients exist to buy it. With so few patients to spread that cost across, the price per child has to be very high to recoup it. The $3.95 million figure isn't a markup on top of the science — it's a direct result of the small number of people who will ever need the drug[3].
But U.S. health insurance is written and budgeted year by year, while Fayuvi is a single infusion. Whichever insurer or state Medicaid program happens to be covering a child in the year they're infused absorbs the full bill, even though the benefits, if the drug works as intended, play out over decades[3]. That mismatch, not simple stinginess, is why one-shot gene therapies tend to move slowly through insurance approval.
Payers — private insurers, employers, and state Medicaid programs — don't dispute that the disease is devastating. Their argument is that they're being asked to pay a fixed, certain price for a benefit whose size carries real uncertainty, given the single-arm trial design. Tools like outcomes-based contracts, where money comes back if a child doesn't respond, or payments spread over several years, are the kind of arrangements they tend to prefer.
Families and advocates see the same mismatch from the other side, and time is the variable they can't get back. The trial data suggest the benefit is largest in children treated youngest, before more brain cells are lost[6]. For them, a slow prior-authorization process isn't a paperwork delay. It's a countdown running against the exact window when the drug works best. Approval, in their view, is the beginning of the fight, not the end of it.
The Same News, Told as Three Different Stories
Coverage of the approval split along fairly predictable lines. STAT News, read largely by health-policy professionals, put the $3.95 million price high in its story alongside the approval itself, and used the phrase "childhood Alzheimer's" to describe the disease — a framing that pairs medical milestone with cost question in the same breath[3].
Financial outlets told a narrower story. Stocktwits led with Ultragenyx's stock jump and did include the $3.95 million figure, but only as a standalone fact, disconnected from any discussion of insurance or who actually pays it, ahead of trader commentary about further stock upside[9]. The French financial site Boursorama went further, leading its headline with the price itself rather than the approval — a framing more natural in a country where a national health system, not a private insurer, would face the bill[4].
Notably absent from day-one coverage: any independent reporting from major U.S. right-leaning outlets. Fox News, the Wall Street Journal's news desk, the New York Post and National Review don't appear to have covered the approval on their own in the first 24 hours, so there's no conservative editorial frame to weigh against the others yet.
Reuters' wire report, which got redistributed to local outlets, stuck to the approval and the stock move in compressed form, with the price, the label's warnings and the trial design mostly falling away in the shorter republished versions[7]. The FDA's own release, by contrast, is the least colorful and most useful record: mechanism, indication, and designations, with no attempt to characterize how large the benefit actually is[1].
What Happens Between Now and a Child's First Infusion
Fayuvi's label carries warnings for liver injury, low platelets, a blood-clotting complication called thrombotic microangiopathy, and infusion reactions[5]. None of that is in dispute. What's still unsettled is how well the trial's 23.5-point result will hold up once more children, diagnosed at different ages and treated in the real world rather than a study, actually receive the drug.
Ultragenyx says commercial launch is roughly 30 to 60 days out[7]. Before any child gets an infusion, someone has to diagnose the disease early, get an insurer or Medicaid program to approve a $3.95 million claim, and do both fast enough to matter for a disease that doesn't pause for paperwork. The FDA has finished its part. What's left is, in the most literal sense, a coverage question.
Summary
On September 17, 2026, the Food and Drug Administration approved Fayuvi (rebisufligene etisparvovec-hopf), the first treatment cleared in the United States for Sanfilippo syndrome type A[1]. The disease is also called MPS IIIA. It is an inherited disorder that steadily destroys a child's brain. Children lose speech, thinking and movement, and most die in their teens[2]. Until this week, doctors could only manage symptoms[1]. The maker, Ultragenyx Pharmaceutical, said it will charge $3.95 million per patient at list price[3][4].
Fayuvi is given once, through an IV. It uses a modified, harmless virus to carry a working copy of the SGSH gene into the body's cells[1]. That gene tells cells to make an enzyme called sulfamidase. Children with MPS IIIA cannot make enough of it, so a sugar-like molecule called heparan sulfate piles up in brain cells and damages them[2]. The therapy is meant to restart that missing step.
The FDA granted standard approval, not the conditional kind. That matters. Ultragenyx had earlier told investors it planned to seek accelerated approval using the level of heparan sulfate in spinal fluid as a stand-in measure[6]. Instead, the agency accepted a cognitive test result. Treated children scored 23.5 points higher on the Bayley-III cognitive raw score, between ages 24 and 60 months, than untreated children in a natural-history record[2].
The genuine dispute is about that comparison. There was no randomized control group. Treated children were measured against records of other children who were never treated. Supporters — including the FDA itself in approving the drug — say the disease's downhill course is predictable enough to serve as a yardstick, and that giving a placebo to dying children would be unethical[2]. Critics of this evidence standard argue that a single-arm trial in a small group cannot fully separate the drug's effect from differences between the two sets of children. A second, separate argument is about the price and who pays it[3]. The label also carries warnings for liver injury, low platelets, a blood-clotting complication and infusion reactions[5]. Neither disagreement was settled by the approval.
The Event
The FDA announced on September 17, 2026 that it had approved Fayuvi (rebisufligene etisparvovec-hopf) for pediatric patients with mucopolysaccharidosis type IIIA[1]. The therapy is a single intravenous infusion using an adeno-associated virus serotype 9 vector to deliver a working SGSH gene[1]. The same day, Ultragenyx said it had received standard — not accelerated — approval, along with a priority review voucher, and set the U.S. wholesale acquisition cost at $3.95 million per patient[2][3]. Ultragenyx shares rose roughly 10% to 13% on the news[9][7].
Undisputed Facts
- The FDA approved Fayuvi on September 17, 2026, as the first treatment for MPS IIIA, also called Sanfilippo syndrome type A[1].
- Fayuvi is a one-time IV infusion that uses an AAV9 viral vector to deliver a functional copy of the SGSH gene[1].
- Before this approval, care for MPS IIIA was limited to managing symptoms; no approved therapy targeted the underlying cause[1].
- FDA had granted the therapy Orphan Drug, Fast Track and Breakthrough Therapy designations[1].
- In July 2025, FDA issued Ultragenyx a complete response letter — a formal rejection — citing chemistry, manufacturing and controls issues, not efficacy[8].
- The approval rests on the Transpher A program, in which treated patients scored 23.5 points higher on the Bayley-III cognitive raw score than a natural-history comparison group, p<0.0001, measured between 24 and 60 months of age[2].
- The Fayuvi label carries warnings and precautions for hepatotoxicity, thrombocytopenia, thrombotic microangiopathy and infusion-related reactions; elevated liver enzymes were seen in studies[5].
- Ultragenyx set the U.S. wholesale acquisition cost at $3.95 million per patient[3][4].
- Ultragenyx estimates about 3,000 to 5,000 patients live in what it calls commercially accessible geographies, with a median life expectancy of about 15 years[2].
The Pressure
Strip away the moralizing and blame. What structural realities persist regardless of which narrative wins?
- Tiny patient pool, huge price
- Ultragenyx counts roughly 3,000 to 5,000 patients in reachable markets[2]. Development and manufacturing costs for a gene therapy do not shrink with the patient count. So the only way to recover them is a very high price per child. The $3.95 million figure is a direct consequence of that arithmetic, not an add-on[3].
- The clock inside the disease
- MPS IIIA damages the brain early, and the trial signal was strongest in the youngest children[6]. Brain cells already lost are not restored by adding the gene. That makes diagnosis speed and insurance-approval speed medically load-bearing, not administrative detail.
- One-time cost, annual budget
- U.S. insurance is written year by year, but this therapy is a single infusion. The payer who funds it may not be the payer who reaps decades of avoided care costs. That mismatch, not stinginess, is the structural reason one-shot gene therapies get slow coverage[3].
- Ethics constrain the evidence
- A randomized placebo arm in a uniformly fatal childhood disease is very hard to run and hard to justify. So the strongest available evidence is usually a single-arm trial against natural history. Both the approval and the criticism of it flow from the same constraint.
- Regulatory turnover
- The division that reviews gene therapies has changed leadership twice in 2026 — Prasad left CBER at the end of April, Makary resigned as commissioner on May 12[3]. Approval standards for this class are in motion, which raises the stakes of each individual decision as a precedent.
Material realityA child with MPS IIIA who is not treated loses speech, thinking and movement, and typically dies around age 15[2]. That does not change based on how this approval is framed. What is now true and was not true last week: a licensed product exists, it costs $3.95 million at list price, and it carries label warnings for liver injury, low platelets, thrombotic microangiopathy and infusion reactions[3][5]. The evidence behind it is a single-arm trial compared against untreated children's records, with a 23.5-point cognitive raw-score advantage and follow-up running nearly eight years[2][5]. Whether individual children actually get the infusion will be decided by newborn and early diagnosis, by insurer and Medicaid coverage decisions, and by how fast both move — not by the FDA, which has finished its part.
Narrative as a weaponUltragenyx is the most active shaper here, and its press release supplies most of the numbers in circulation — the 23.5-point effect, the patient-population estimate, the 'first-ever' framing[2]. It wants you to read this as a scientific and humanitarian milestone, with price as a reasonable trade for a lifetime of care. Financial media, from Stocktwits to Boursorama, want you to read it as either a stock story or a cost story, and each drops whatever does not fit[9][4]. Patient advocacy groups want the frame to be access and speed, because for them the fight shifts on approval day from the FDA to insurers. The evidence-standards camp — quieter in day-one coverage but institutionally real, given who ran CBER until April — wants you to notice that no child in this trial was compared against a matched, randomized control. The FDA's own release is the thinnest in rhetoric and the most useful as a record: it states the mechanism, the indication and the designations, and does not characterize the size of the benefit[1]. Note also what is missing from the first-day coverage: no major U.S. right-leaning outlet appears to have reported the approval independently, so there is no conservative editorial frame to weigh here yet.
How Each Side Sees It
Each major actor’s view — how it frames things, its underlying incentive, and how it’s materially affected. Tap a side to read it.
Frames it asThe agency's case is that MPS IIIA is fatal, has no other treatment, and follows a course predictable enough to be measured against[1]. On that reasoning, a well-documented natural-history record is a legitimate comparison when a randomized placebo group would mean assigning dying children to nothing. FDA also points to the strength of the result it accepted: not a lab marker, but a cognitive score, with nearly eight years of follow-up behind it[5][2]. The agency granted standard approval rather than the accelerated kind — its own signal that it considered the clinical benefit shown, not merely predicted[2].
WhyFDA must hold a line that is credible in both directions. Approve too loosely and the agency is accused of rubber-stamping; approve too tightly and it is accused of letting children die waiting. Its 2025 rejection of this same product on manufacturing grounds, followed by approval once those issues were fixed, is the evidence it will point to as proof the review was real[8].
Impact on themThe decision becomes a reference point for future ultra-rare gene therapy reviews that rely on external controls. It follows a period of turnover at the top of the agency: Vinay Prasad, a vocal critic of loose gene-therapy approvals, ran CBER before leaving at the end of April 2026, and Commissioner Marty Makary resigned on May 12, 2026[3].
Frames it asThe company's argument is that it finished a job others abandoned. The underlying science came out of Nationwide Children's Hospital and Ohio State, was licensed to Abeona Therapeutics, and was out-licensed to Ultragenyx when Abeona ran short of money despite positive data[5]. Ultragenyx argues that a $3.95 million one-time price should be judged against a lifetime of round-the-clock care for a child who would otherwise lose speech, mobility and thinking[3][2]. It also stresses that the price is a list figure, not what payers finally pay, and that it has set up a program with dedicated gene-therapy guides to help families through insurance[2].
WhyUltragenyx needs commercial proof that ultra-rare gene therapy can pay. With perhaps a few thousand patients worldwide, the only lever available is price per patient. The priority review voucher it received adds value: such a voucher entitles its holder to a faster, six-month FDA review of some other drug, and vouchers can be sold to other companies[2].
Impact on themThe stock rose about 10% to 13% on the day[9][7]. Abeona, which originally licensed the asset, also saw its shares gain on the approval of a drug it no longer owns[9]. Commercial launch was signaled for roughly 30 to 60 days out[7].
Frames it asTheir central claim is about time, not statistics. MPS IIIA moves fast, and the trial data suggest the benefit is largest in children treated youngest — under two, or earlier in the disease[6]. So for this community, a demand for a randomized placebo arm is not neutral caution; it is a decision with a body count. They argue the natural history of this disease is unusually well documented and unusually uniform, which is exactly the condition under which an external comparison is defensible. They also warn that approval is not access: a $3.95 million list price and a slow prior-authorization process can keep a child from being treated during the narrow window when treatment helps most[2].
WhyTo convert a regulatory milestone into actual treated children — which means newborn screening, fast diagnosis and payer coverage, not just a label.
Impact on themFamilies face the coverage fight next. The therapy is one-time, so the cost lands on whichever insurer or Medicaid program happens to cover the child that year — a known problem for one-shot gene therapies[3].
Frames it asPayers do not dispute that the disease is devastating. Their argument is about who absorbs a one-time seven-figure bill and on what evidence. A single self-insured employer or a small state Medicaid program can see its budget distorted by one child. They argue that when benefit is shown against a historical comparison rather than a randomized control, the size of the benefit carries real uncertainty — and they are being asked to pay a fixed, certain price for it. Their preferred tools are outcomes-based contracts, where some payment is refunded if the child does not respond, and installment payments over years.
WhyBudget predictability, and avoiding being the payer that happens to be on the hook in the month a child is infused.
Impact on themCoverage terms, prior-authorization rules and any rebate deals will decide how much of the $3.95 million list price is actually paid — and how quickly children reach treatment[3][2].
Frames it asThis camp is not anti-treatment; it is anti-uncertainty-sold-as-certainty. Their argument runs like this. In a single-arm trial, everyone gets the drug. The comparison group is a record of other children, often from other places and other years, diagnosed by other doctors. Children who enroll in a trial tend to be the ones whose families found a specialist early — which can make them look better than the historical group for reasons that have nothing to do with the drug. They also note that the headline effect comes from a small number of patients: an earlier analysis presented to neurologists showed a 23.2-point effect in 17 children who were under two or earlier in the disease[6], close to but not the same as the 23.5-point figure in the approval package[2]. When the label also carries warnings for liver injury and a clotting complication[5], they argue the size of the benefit needs to be pinned down, not assumed. Their fix is not to block the drug but to require rigorous follow-up after approval.
WhyTo keep the approval bar from drifting, on the view that a weak evidence standard eventually produces expensive drugs that do not work — which harms patients most of all.
Impact on themThis position had an institutional voice at FDA until recently. Prasad, who had publicly attacked the agency's earlier approval of Sarepta's Duchenne gene therapy, ran the division that reviews gene therapies before leaving at the end of April 2026[6][3].
Like this article?
The Bias Ledger average rating 3.3
The same story, as framed by outlets across the spectrum, ordered least to most biased. The bias score (1 = straight, 10 = heavily spun) is an AI assessment of that framing — click an outlet to see its track record. The tell is the word choice or omission that reveals the angle.
| Outlet | Vantage | Bias | How they frame it | The tell |
|---|---|---|---|---|
| Reuters | U.S. center / international wire | 2 | "US FDA approves Ultragenyx's gene therapy for rare disorder" — short wire copy, approval plus the share move, distributed widely to local radio and regional sites[7]. | The framing is clean but compressed. The wire lead treats the approval as a corporate event, so the price, the label warnings and the single-arm design all sit below the fold or drop out of the republished versions entirely. |
| STAT News | U.S. health-policy press, center-left readership | 3 | "FDA approves a new gene therapy for Sanfilippo syndrome, an ultra-rare disease" — reports the approval, then the $3.95 million price and the FDA leadership turnover[3]. | Uses the emotive shorthand "childhood Alzheimer's" while also foregrounding cost. Both moves are defensible; together they steer the reader toward a cost-versus-desperation frame rather than an evidence frame. |
| Endpoints News | U.S. biotech trade, industry readership | 3 | "FDA approves Ultragenyx's gene therapy for Sanfilippo syndrome type A" — straight regulatory reporting for a professional audience[8]. | Trade framing assumes the reader already accepts single-arm natural-history comparison as normal for ultra-rare disease. That assumption is exactly what critics dispute, and it goes unstated. |
| NeurologyLive | U.S. clinician trade publication, pharma-advertising funded | 4 | "FDA Approves UX111, First Gene Therapy for Sanfilippo Syndrome Type A" — leads with the first-ever milestone and the efficacy numbers[6]. | Leans on company-supplied efficacy data and expert quotes. The label's liver-toxicity and clotting warnings get far less space than the cognitive-score result. |
| Stocktwits | U.S. retail-investor platform | 4 | "RARE Stock Jumps 13% — Extends Gains After-Hours As FDA Clears First Sanfilippo Gene Therapy"[9]. | The children are the setting; the trade is the story. The $3.95 million price does appear, but only as a standalone fact ahead of retail-trader chatter about upside — never connected to coverage, insurance or who actually pays it. The label warnings and the evidence debate are absent entirely. |
| Boursorama | French financial media (Société Générale-owned) | 4 | "Ultragenyx fixe le prix de sa thérapie génique contre une maladie rare à 3,95 millions de dollars" — the price, not the approval, is the headline[4]. | A European frame where a national health system would face the bill. The medical first is demoted to a subordinate clause; the number leads. |
References
- FDA Approves First Gene Therapy for Pediatric Patients with Sanfilippo Syndrome Type A — U.S. Food and Drug Administration · U.S. federal regulator — primary source; institutional interest in defending its own decision
- Ultragenyx Announces Approval of FAYUVI Gene Therapy, the First-Ever FDA-Approved Treatment for Sanfilippo Syndrome Type A (MPS IIIA) — Ultragenyx Pharmaceutical · Company press release — the manufacturer; directly financially interested
- FDA approves a new gene therapy for Sanfilippo syndrome, an ultra-rare disease — STAT News · U.S. health/biotech news outlet owned by Boston Globe Media; subscription-funded, center-left readership
- Ultragenyx fixe le prix de sa thérapie génique contre une maladie rare à 3,95 millions de dollars — Boursorama · French financial news site owned by Société Générale; market-investor orientation
- FAYUVI — Highlights of Prescribing Information / product page — U.S. Food and Drug Administration · U.S. federal regulator — primary source, approved product labeling
- FDA Approves UX111, First Gene Therapy for Sanfilippo Syndrome Type A — NeurologyLive · U.S. clinician trade publication, pharmaceutical-advertising funded
- US FDA approves Ultragenyx's gene therapy for rare disorder — Reuters · International wire service; center, market-facing
- FDA approves Ultragenyx's gene therapy for Sanfilippo syndrome type A — Endpoints News · U.S. biotech trade publication; subscription-funded, industry readership
- RARE Stock Jumps 13% — Extends Gains After-Hours As FDA Clears First Sanfilippo Gene Therapy — Stocktwits · U.S. retail-investor social/news platform; trading-oriented
- RARE Stock Jumps 13% — Extends Gains After-Hours As FDA Clears First Sanfilippo Gene Therapy — Asianet Newsable · Indian commercial news network (Asianet News); market-desk republication