🦠 The FDA said no, then yes, then maybe - all in 8 days

Remember when regulatory agencies were boring? The FDA just speedran a bureaucratic meltdown, MIT decided your vacation photos need a signature scent, and someone strapped a hydrogen sensor to their underwear for a week straight. This week in biotech: we're turning memories into smells, tracking gas with smartphone precision, and watching federal health agencies have very public disagreements with themselves. Let's get into it.

NEWS

Picture this: You're Moderna, you've spent years developing an mRNA flu vaccine, the FDA told you your clinical trial design was fine back in 2024, and then in February 2026, they refuse to even look at your application. Not rejected on safety. Not rejected on efficacy. Refused to review.

The issue? Moderna's Phase 3 trial of 40,805 adults used a standard-dose flu vaccine as the comparator for everyone, including seniors. The FDA argued that for people 65 and older, high-dose vaccines like Fluzone represent the "best-available standard of care," making Moderna's comparator inappropriate. Never mind that Fluzone itself was originally licensed using a standard-dose comparator. Apparently, the rules changed.

The timing is, shall we say, notable. CBER Director Vinay Prasad (a vocal critic of COVID vaccine mandates) reportedly overruled career FDA scientists who were prepared to review the application. HHS called this "categorically false." STAT News cited three agency officials.

Then, eight days later, the FDA reversed course and accepted an amended application with a decision date of August 5, 2026. The phrase "regulatory whiplash" appeared in multiple headlines, because of course it did. International regulators in Europe, Canada, and Australia have all accepted mRNA-1010 for review without drama. This is fine.

RESEARCH

Scientists at MIT just created a device that analyzes your photographs and generates custom fragrances to match. It's called the Anemoia Device, named after a modern word meaning nostalgia for a time you never experienced, coined by John Koenig in his Dictionary of Obscure Sorrows project.

Here's how it works: You insert a physical printed photograph. A vision-language model captions the image, you adjust three dials for mood and perspective, and the system generates a short poetic narrative. Then it performs what researchers call "cross-modal semantic-to-olfactory translation" (basically an AI sommelier for smells) and selects up to four scents from a library of 39 pre-blended fragrance oils.

The device comes from Cyrus Clarke, an MIT Media Lab researcher and former L'Oréal innovation lead, working under legendary human-computer interaction researcher Hiroshi Ishii. It was presented at NeurIPS 2025's Creative AI Track, which is more art showcase than a hard science proceedings.

Here's the catch: no formal user study has validated whether the device actually generates anything resembling synthetic memories. Sensory psychologist Avery Gilbert has pointed out that the whole thing reverses the Proustian phenomenon - real scent memories work because smell triggers recall, not because you generate scent after already seeing the image. Still, the prototype exists, it dispenses perfume based on your vacation photos, and someone is definitely going to try this with their ex's pictures.

NEWS

Dutch biotech VectorY Therapeutics dosed its first patient with VTx-002 on February 9th at Mass General Brigham's Healey & AMG Center for ALS. It's the first clinical trial of a therapy designed to target TDP-43 pathology, which is present in up to 97% of all ALS cases.

The approach is genuinely clever. VTx-002 uses an AAV5.2 viral vector to deliver genetic instructions that cause cells to continuously produce an antibody fragment. This antibody selectively binds the misfolded, aggregated form of TDP-43 while leaving the normal protein alone. One injection to the base of the skull, and your neurons become their own antibody factories.

The PIONEER-ALS trial will enroll 12 adults across sites in the US and Europe, with follow-up lasting up to five years. The FDA granted both IND clearance in December 2025 and Fast Track designation in January 2026.

To be clear: this is a Phase 1/2 safety trial. There's no efficacy data yet, and VectorY's preclinical work exists mainly in conference posters rather than peer-reviewed journals. But after decades of ALS trials targeting the wrong mechanisms in the wrong patients, a therapy that goes after the protein actually present in nearly all cases? That's the kind of rational drug design the field desperately needs.

RESEARCH

Here's how science is supposed to work: Researchers at Baylor College of Medicine just published a study in The American Journal of Human Genetics identifying mutations in a gene called ASTN1 as the cause of neurodevelopmental disorders in 18 children from 12 unrelated families. The twist? They solved it using mouse research that started in the mid-1990s.

ASTN1 encodes a protein critical for neuronal migration - essentially the molecular GPS that helps developing brain cells find their correct positions. Three decades of mouse knockout studies established exactly what goes wrong when this gene doesn't work. When clinicians sequenced these children and found ultra-rare ASTN1 variants, they already knew what to look for.

All 18 children had developmental delays or intellectual disability, ranging from mild to profound. Many had associated conditions, including autism, epilepsy, and movement disorders. Brain imaging showed structural abnormalities in most cases, particularly involving the corpus callosum and cerebellum - exactly what the mouse models predicted.

Lead authors Jesse Levine and Daniel Calame, working with senior author James Lupski at Baylor, assembled this international cohort from families across multiple countries. Before this paper, ASTN1 had no gene-disease association entries in clinical genomic databases like OMIM, ClinGen, or GenCC. Now, 18 families finally have an answer after years of diagnostic odyssey. Basic research: still worth funding.

RESEARCH

You knew this was coming. Researchers at the University of Maryland have developed a wearable hydrogen sensor that clips onto regular underwear and continuously monitors your flatulence. The peer-reviewed study in Biosensors and Bioelectronics: X found that healthy adults pass gas an average of 32 times per day - roughly double the textbook estimate of 14.

The device is about the size of a coin, runs for up to a week on batteries, and transmits data via Bluetooth. It specifically detects hydrogen gas, which is produced exclusively by gut microbes, making it a direct window into your microbiome's fermentation activity.

Lead researcher Brantley Hall and first author Santiago Botasini ran two sub-studies. The comfort trial had 19 participants wear the device for a week, averaging over 11 hours daily (95% reported no discomfort). The validation study gave 38 people gumdrops - either sugar (control) or inulin (a fiber that gut bacteria love to ferment). The sensor detected hydrogen spikes in 36 of 38 participants after inulin consumption, peaking 3-4 hours later.

The medical application is actually very real: roughly 20% of Americans report problems with excess gas, and current measurement methods involve either unreliable self-reporting or rectal tubes (which, no thank you). A nationwide citizen science project called the Human Flatus Atlas is now recruiting participants. Both researchers have co-founded a company called Ventoscity to commercialize the technology, because apparently, there's a market for quantified flatulence.

So there you have it. The FDA can't decide if it wants to review vaccines, MIT wants to bottle your childhood summers, ALS patients finally have a therapy targeting the actual problem, decades-old mouse research is still paying diagnostic dividends, and someone is tracking the nation's collective gas output one Bluetooth ping at a time.

Did the FDA's regulatory pirouette make you lose faith in institutions? Ready to smell your vacation photos? Curious how many times you've passed gas since starting this newsletter? (Statistically, probably twice.) We read every reply - hit us back.

Forward this to someone who thinks biotech is boring. We'll wait.

Keep questioning everything (especially your comparator vaccines),

Prateek & Jere

P.S. If someone offers you inulin gumdrops and asks you to wear special underwear, you're either in a clinical trial or a very specific kind of first date.