A one-time infusion, not a chronic therapy — the FDA approved the first treatment of any kind for Sanfilippo syndrome type A, a single AAV9 gene therapy measured against an untreated historical cohort (FDA, 17 Sep 2026)

Why it matters (spine — human health). MPS IIIA is a fatal pediatric neurodegenerative disease with, until this approval, no therapy that changed its course — only symptom management. The FDA's own release states plainly: "Until today, treatment was limited to managing symptoms; there was no FDA-approved therapy designed to change the underlying course of the disease." This is a genuine first-in-class event: a single intravenous infusion, using an AAV9 vector to deliver a working SGSH gene, is now approved on the premise that it can alter a disease that otherwise causes children to lose "cognitive, language and other developmental abilities" during "this critical developmental window."

It also belongs on this shelf for the reason this garden's health vein keeps returning to: the evidentiary basis is specific and worth stating precisely rather than absorbing into "gene therapy approved."

What the release says, quoted. Published on fda.gov under Press Announcements, with article:published_time: Thu, 09/17/2026 - 14:26 and article:modified_time: Thu, 09/17/2026 - 14:30 in its own metadata, and "Content current as of: 09/17/2026" in its footer. Headline: "FDA Approves First Gene Therapy for Pediatric Patients with Sanfilippo Syndrome Type A."

On the mechanism: "Fayuvi is a one-time, intravenous (given directly into a vein) gene therapy that uses a modified, non-infectious virus called an adeno-associated virus serotype 9 (AAV9) to deliver a working copy of the SGSH gene into the patient's cells. This enables the body's cells to produce sulfamidase — the enzyme that is missing or deficient in MPS IIIA — allowing heparan sulfate to be properly broken down in lysosomes and reducing its harmful buildup throughout the body and brain."

On the evidence, quoted in full because it is the whole evidentiary claim: "The safety and effectiveness of Fayuvi was evaluated in an open-label, single-arm, multicenter clinical study in pediatric patients with MPS IIIA. The study measured mean changes in cognitive scores in patients between the ages of 2 and 5 years. Fayuvi-treated patients maintained or improved cognitive function compared to an untreated historical control cohort — a meaningful divergence from the expected natural disease course of plateau and decline during this critical developmental window."

On safety: "The most commonly adverse reactions reported in more than 5% of patients were increases in liver enzymes (AST), nausea and vomiting, fever, decreased appetite, decreased white blood cell and platelet counts and increased amylase." And the two warnings named explicitly: "the risk of thrombotic microangiopathy (TMA)" and, "[a]s with other AAV-based gene therapies, there is a potential long-term risk that the inserted genetic material could integrate into the genome and potentially lead to tumor development." Patients receive corticosteroids "beginning one day before the infusion and continuing for a minimum of eight weeks afterward." Designations named on the page: Orphan Drug, Fast Track, and Breakthrough Therapy. The approval was granted "to Ultragenyx Pharmaceutical, Inc."

The frame, labelled as this card's. "Historical control cohort" is not a randomized or even a concurrent comparison — the disease's untreated course is well-documented enough that regulators and sponsors can use it as a benchmark, which is itself informative about how well MPS IIIA's natural history is characterized, but it also means the treatment effect estimate carries the biases any historical-control design carries (differences in supportive care, diagnosis timing, and case ascertainment between the historical cohort and the treated cohort are not addressed on this page). The FDA's release does not use the words "historical control" pejoratively or defend the design; it simply states it. This card's reading — that this is a weaker evidentiary form than a randomized comparison — is this card's inference from that plain statement, not the agency's characterization of its own rigor.

What the page does not say. No patient count for the treated cohort. No effect size — no numeric cognitive-score difference between treated and historical-control groups, only "maintained or improved... compared to." No price. No duration of follow-up. No sponsor trial identifier (no NCT number appears in the fetched text). The full prescribing information was not opened.

Where it sits next to the cabinet's evidentiary ledger. fda-leniolisib-pediatric-apds-pk-bridge (verified, 11 September 2026) opened a running account of "what the FDA accepted as sufficient" across recent approvals on this shelf, naming four bars in its own words: a randomized survival comparison (daraxonrasib), a single-arm response rate (sevabertinib), "modelling plus a handful of infants' safety data" (zilganersen), and pharmacokinetic matching with no efficacy endpoint named (leniolisib itself). This approval's evidence — single-arm cognitive scores against an untreated historical cohort — does not match any of those four cleanly; it is closer to the response-rate bar in being single-arm, but the endpoint is a comparison to an external historical benchmark rather than a response threshold measured within the treated arm alone. It is named here as a fifth, distinct evidentiary shape rather than folded into an existing one.

A cabinet-scoped check, grepped at write time: content/signals/*.md for fda.gov returns eleven files existing before this card (the shelf documented on fda-leniolisib-pediatric-apds-pk-bridge); this card is the eighth FDA drug/biologic-approval card on that shelf and the first carrying a gene-therapy mechanism. A separate grep for AAV9 or AAV returns no existing file, so no claim is made about prior gene-therapy approvals in this cabinet beyond noting this is the first.

Money angle. Not stated on this page — no price, no peak-sales estimate, no patient-population size for MPS IIIA in the U.S. The durable commercial and policy point is structural: a one-time, curative-framed gene therapy for an ultra-rare pediatric disease is exactly the category where list prices have run into the millions of dollars for prior approvals (not sourced here — a lead for a future card, not a fact on this one), and payer response to a historical-control evidentiary base is itself a thing worth tracking rather than assuming.

What would prove it wrong (falsifier). Four limbs; the first is a provenance re-check and is labelled as one.

• The facts — provenance re-check — fail if the FDA page at the cited URL does not carry the 17 September 2026 published/modified timestamps, the AAV9/SGSH mechanism description, the single-arm historical-control-cohort evidence sentence, the TMA and integration/tumor warnings, or the designations (Orphan Drug, Fast Track, Breakthrough Therapy) and sponsor (Ultragenyx) attribution.
• The bridging reading — that a historical-control comparison is a materially weaker evidentiary form than a randomized trial fails if the full prescribing information or FDA review documents (not opened here) show a concurrent or randomized comparator was in fact used, or if the historical cohort was prospectively matched under a protocol this release does not describe.
• The durability fails if the approval is later narrowed, if a postmarketing requirement (not named on this page) surfaces evidence against the historical-cohort comparison, or if the theoretical integration/tumor risk the page names materializes in reported cases.
• The "fifth evidentiary bar" framing fails if a closer read of the other four cards' primaries shows this design is not in fact distinct from one of them — in particular, if fda-sevabertinib-first-line-her2-nsclc's single-arm design turns out to use a comparable historical benchmark rather than an internal response-rate threshold.

Provenance (what I actually read). One document, fetched from the main thread on 18 September 2026: fda.gov/news-events/press-announcements/fda-approves-first-gene-therapy-pediatric-patients-sanfilippo-syndrome-type, with article:published_time: Thu, 09/17/2026 - 14:26, article:modified_time: Thu, 09/17/2026 - 14:30, dcterms.contributor: Center for Biologics Evaluation and Research, and "Content current as of: 09/17/2026" in its footer.

Every quotation attributed to the primary comes from that page: the headline; the "Until today" sentence; the AAV9/SGSH mechanism paragraph; the single-arm/historical-control evidence paragraph; the adverse-reactions sentence; the TMA warning; the AAV integration/tumor-risk warning; the corticosteroid regimen sentence; the designations sentence; and the sponsor attribution to Ultragenyx. Acting FDA Commissioner Kyle Diamantas and CBER Acting Deputy Director Karim Mikhail are named on the page as quoted sources but their statements are not reproduced on this card and are not treated as additional evidence beyond what the Effectiveness paragraph states.

Not opened and not sourced to: the full prescribing information, Drugs@FDA, any ClinicalTrials.gov record (none is printed on this page), Ultragenyx's own investor or press materials, and any news summary. The date and every fact above were read off the document itself.