A first-ever gene therapy, a rare lung cancer variant, and a panel that decides what your insurance covers

A note before I start. I write these to be accurate, not exciting, so where a study is small or a finding is being overstated in the coverage, I say so. There is a real gap in this week's digital health news and I explain it at the end rather than filling the slot with something I could not verify. This is a news summary, not medical advice.

The FDA approved the first treatment of any kind for Sanfilippo syndrome type A

On September 17, the FDA approved Fayuvi, generic name rebisufligene etisparvovec-hopf, from Ultragenyx Pharmaceutical, for pediatric patients with mucopolysaccharidosis type IIIA, better known as Sanfilippo syndrome type A. Before this, there was nothing. No disease modifying therapy at all for an inherited neurodegenerative disease where children lose skills they had already acquired and typically die in adolescence (Source).

The approval rested on an open label, single arm, multicenter study in children roughly 2 to 5 years old, in which treated patients maintained or improved cognitive function compared to an untreated historical control cohort. That design matters and I want to be honest about it: this is a single arm trial with a historical comparator, not a randomized controlled trial. In an ultra rare pediatric disease with no existing treatment, randomizing children to nothing is ethically fraught, so the FDA accepted a weaker evidence design. That is a defensible tradeoff, but it is a tradeoff. The safety picture is not trivial either: the label carries a warning for thrombotic microangiopathy, and adverse reactions above 5 percent included elevated AST, nausea, vomiting, fever, decreased appetite, and reduced white blood cell and platelet counts. The FDA also notes the theoretical long term AAV risk of genomic integration and tumorigenicity, which is an open question across the whole AAV gene therapy category.

Beyond the families directly affected, this is a test case on two fronts. It is AAV gene therapy aimed at the central nervous system, which is much harder than treating a single organ. And it is the FDA signaling what evidence it will accept in ultra rare pediatric neurodegenerative disease, which sets the path for every similar program behind it.

A rare inherited variant sharply raises lung cancer risk in never smokers

Published in Science on September 17, a study led by Pasi Janne at Dana-Farber Cancer Institute analyzed genotype and health data from roughly 3.37 million 23andMe research participants. They found that an inherited germline EGFR T790M variant, carried by about 1 in 15,850 people, is associated with roughly a 62 fold increase in odds of lung cancer among never smokers and about 25 fold among smokers. For scale, smoking itself carries roughly a 4 fold increase in the general population. The variant was associated only with lung cancer among the 17 cancers and conditions screened, and appears to trace to southern Appalachian US ancestry (Source).

Now the caveats, which matter more than usual here because the coverage has been sloppy. Janne himself cautioned that the analysis "does not give the full picture of the lifetime risk," and that unmeasured genetic modifiers and environmental factors may change individual risk substantially. Some outlets headlined 25 fold and others headlined 62 fold for the same paper, and those are not competing estimates: 25 fold is for smokers and 62 fold is for never smokers. If you see them presented as a disagreement, the outlet did not read the paper. And the most important framing point: a large relative risk on a small absolute baseline is not a large absolute risk. The variant is rare enough that it explains only a small fraction of never smoker lung cancer overall (Source).

Why it matters anyway: lung cancer in never smokers is a large and badly understood disease burden, and current screening guidelines are built almost entirely around smoking history. A specific, testable germline variant gives a concrete path to screening relatives in affected families, which is a real clinical action rather than a theoretical one.

Eight new members joined the USPSTF

On September 17, HHS announced eight new appointees to the US Preventive Services Task Force, bringing the panel to 16 members alongside eight holdovers from the prior administration, with Seth J. Corey, a pediatric hematologist oncologist, named chair. The other appointees are cardiologists Ronald Karlsberg and Venkatesh Murthy, health economist Stephen Parente, gastroenterologist Louis Wilson, radiologist Dennis Wulfeck, pediatrician Patrick Hunter, and family physician Goldie Stands-Over-Bull (Source).

This is the story on this list that touches the most people and will get the least coverage. USPSTF "A" and "B" recommendations legally determine which preventive services private insurers must cover with no cost sharing: screening colonoscopy, HIV PrEP, cancer screenings, statins. If the task force moves a recommendation, your out of pocket cost for that service can change. The composition shift is toward subspecialists rather than primary care generalists. The American Medical Association responded that "maintaining a strong primary care voice must remain central to its work," and Aaron Carroll of AcademyHealth raised conflict of interest concerns about specialists evaluating services they themselves perform and bill for. Context worth knowing: the panel has been largely inactive for over a year, and its supporting agency staff fell from roughly 300 in 2024 to 84 in 2026 (Source). If you want one health policy thing to watch over the next year, this is it.

ARPA-H launched a program to speed up autism diagnosis

Also September 17, Secretary Kennedy announced SPECTRA, a research funding program run by the Advanced Research Projects Agency for Health under program manager Russell Shilling, aimed at developing technology and tools to shorten autism diagnostic timelines, which currently run months to years largely because of assessment capacity shortages. No dollar figure, grant count, or application timeline was disclosed. Helen Tager-Flusberg of Boston University's Center for Autism Research Excellence called the technology focus "timely" while saying she hoped the review process would be unbiased (Source).

The reaction was mixed, and the reason is worth stating directly. Autism advocates objected to Kennedy's characterization of autistic children and to his public claims linking autism to vaccines and to acetaminophen. Those claims are not supported by the evidence base. Researchers described cautious optimism about this particular program alongside real wariness about the broader federal autism agenda it sits inside. Diagnostic wait time is a genuine, measurable bottleneck to early intervention, and shortening it is a real good. It is still worth watching where the money actually goes.

WHO reported the DRC Ebola outbreak past 7,200 cases

In his September 16 media briefing, WHO Director-General Tedros Adhanom Ghebreyesus reported more than 7,200 cases and more than 3,500 deaths in the Democratic Republic of the Congo Ebola outbreak across seven provinces, with more than 1,700 people recovered. In the preceding week, Ituri province alone recorded about 300 new cases and 160 deaths, while North Kivu's weekly caseload roughly doubled over two weeks from about 100 to over 200. Tedros said transmission is declining in the hardest hit parts of Ituri and that South Kivu has had no new cases since May (Source).

A case fatality ratio of roughly 48 percent across more than 7,200 cases makes this one of the largest Ebola outbreaks on record. WHO's outbreak documentation indicates the causative agent is Bundibugyo virus, for which there is no licensed vaccine or specific therapeutic, unlike Zaire ebolavirus. That detail explains why this one has run so long and so wide. Growth in North Kivu while Ituri declines means the outbreak is migrating rather than resolving, and it has already crossed into Uganda. Separately, for World Patient Safety Day on September 17, WHO stated that roughly 1 in 6 people with noncommunicable diseases experience preventable harm during care. One caveat: the WHO regional outbreak page describing this outbreak was last updated June 8, 2026 and covers a narrower geography than the September 16 briefing, so treat the provincial spread as approximate.

What is not in here, and why

I could not find a genuinely new, verifiable digital health, wearables, sleep, or metabolic study published in this specific week. Rather than stretch something to fill the slot, here is what I looked at and rejected. A Florida International University study finding smartwatches substantially overestimate calories burned, with accuracy degrading as body fat rises, is real and interesting, but it was reported around September 2, outside this week.

Two ScienceDaily items dated this week are recycled older papers, and they will almost certainly show up in other roundups you read. "Nearly 10,000 patients reveal what works best for knee arthritis pain relief," dated September 13 to 14, is a writeup of a 139 trial network meta analysis published in PLOS One in June 2025. "Scientists reversed biological age in older adults in just 4 weeks," dated September 15, covers a 104 participant University of Sydney randomized trial that ScienceDaily already ran on May 11, 2026, and which uses estimated biological age as a surrogate endpoint rather than any clinical outcome. Also worth noting for calendars: EASD 2026, the major diabetes and metabolic meeting, runs September 28 through October 2 in Milan, so that data drops after this window.

Everything here reflects FDA and HHS announcements, the WHO Director-General's briefing, the published Science paper, and same-week reporting from STAT News, linked inline. Study limitations are stated where they exist. This is a news summary and general information, not medical advice. Verified September 18, 2026. The studio's app lineup is at jcmobileappstudio.com/apps.