The FDA Just Said One Patient Might Be Enough to Approve a Drug
The FDA just proposed a framework that could let gene therapies win approval based on a single patient's data, no randomized trials required. It might be the biggest change to drug approvals since 1992, and it could unlock treatments for hundreds of diseases pharma has long ignored.
Imagine telling a pharma executive in 2020 that the FDA would one day consider approving a gene therapy based on a single patient's data. They'd laugh you out of the room. Then they'd check if you were feeling okay.
But that's essentially what the FDA proposed on February 23, 2026. And it might be the biggest change to how drugs get approved in a generation.
The Old Rules Were Built for a Different World
For decades, getting a drug approved meant one thing: run a big clinical trial. Recruit hundreds or thousands of patients. Randomly assign some to the drug and others to a placebo. Crunch the numbers. Show statistical proof that the drug works better than nothing.
This system is great when your disease affects millions of people. Got a new cholesterol drug? No problem; there are plenty of patients to fill a trial. But what happens when a disease affects five people on the entire planet?
You're stuck. You literally cannot run a randomized controlled trial because there aren't enough patients to randomize. The math doesn't work. The economics don't work. And so hundreds of ultra-rare genetic diseases have been sitting in regulatory purgatory, not because science can't help these patients, but because the approval process was never designed for them.
That's the problem the FDA just decided to solve.
Enter the Plausible Mechanism Framework
The draft guidance has a mouthful of a name: "Considerations for the Use of the Plausible Mechanism Framework to Develop Individualized Therapies that Target Specific Genetic Conditions with Known Biological Cause." We'll call it the PM Framework because life is short.
The core idea is deceptively simple. If you can show that a disease is caused by a specific genetic mutation, and your therapy directly fixes or compensates for that mutation, and you can prove the therapy actually hits its target, then you don't need a giant trial to prove it works.
Instead of comparing your drug against a placebo group, you compare the patient's outcome against the known natural history of their disease. Think of it like this: if a child with a fatal genetic condition was expected to lose the ability to walk by age three, and your gene therapy kept them running at age five, you don't need a control group to see something remarkable happened.
Get tomorrow's biotech intelligence before your competitors.
Join thousands of biotech professionals who start their day with our free, daily briefing.
The framework covers the heavy hitters of genetic medicine: CRISPR, base editing, prime editing, and antisense therapies, basically the entire modern toolkit for fixing broken DNA or its instructions.
What It Actually Takes to Qualify
Before anyone panics about the FDA rubber-stamping untested therapies, let's be clear: this isn't a free pass. The bar is different, but it's still high.
The framework rests on five pillars. First, you need to identify the exact molecular or cellular abnormality causing the disease. No hand-waving. No "we think it's probably this gene." You need precision.
Second, your therapy has to directly target that root cause or the biological pathway immediately connected to it. Third, you need solid nonclinical data (think lab studies and animal models) showing the therapy engages its target. In one notable case, mouse model data showing 42% liver cell editing was enough to support approval when testing directly in the patient would have been too risky.
Fourth, you need a well-characterized natural history of the disease, so there's a clear baseline to measure improvement against. And fifth, the clinical improvement has to be robust enough to rule out chance. Even with a single patient, the data needs to tell a convincing story.
Post-approval, companies aren't off the hook either. They're required to collect real-world evidence on safety and efficacy, monitor developmental milestones in treated children, and track potential off-target effects from gene editing. If the therapy turns out to be ineffective or unsafe, the FDA can pull it.
Why This Changes Everything for Rare Disease
The math problem in rare disease has always been brutal. There are roughly 7,000 known rare diseases, and about 500 of them are "ultra-rare," conditions so uncommon that conventional drug development is essentially impossible. Pharma companies have historically written off these diseases as commercially unviable. Can you blame them? Spending $2 billion on a clinical trial for a disease that affects twelve people worldwide doesn't exactly scream good ROI.
The PM Framework rewrites that equation. By allowing approval based on small-cohort or even single-patient data, it slashes the cost and time needed to bring a therapy to market for these conditions. Analysts at William Blair described the implications as "broad and positive" for the entire cell and gene therapy sector.
There's another clever wrinkle. The framework supports master protocols, which let companies test multiple therapy variants (say, different gene edits targeting different mutations within the same gene) under a single regulatory umbrella. Some diseases like CPS1 deficiency (a rare metabolic disorder) are caused by over 150 different mutations in one gene. Instead of filing 150 separate applications, a company could potentially validate its platform once and expand from there.
The Heavyweights Are Already Positioned
The biotech companies best positioned to benefit read like a who's-who of genetic medicine. Intellia Therapeutics is developing in vivo CRISPR therapies for conditions like transthyretin amyloidosis and hereditary angioedema, with Phase 3 data expected this year. Beam Therapeutics is working on base editing approaches for sickle cell and other blood disorders. Prime Medicine is pushing prime editing into ultra-rare territory, with programs targeting chronic granulomatous disease and Wilson's disease.
Then there's Aurora Therapeutics, co-founded by CRISPR pioneer Jennifer Doudna, which is explicitly building its strategy around this new pathway to deliver personalized gene surgeries. And Ultragenyx, a rare disease specialist, has a gene therapy for glycogen storage disease type Ia under priority review.
Vertex and CRISPR Therapeutics already proved the concept with Casgevy, their CRISPR-based therapy approved for sickle cell disease and beta-thalassemia. The PM Framework extends that precedent into far rarer genetic territory.
The Biggest Shift Since Accelerated Approval
To understand how radical this is, you need some context. The last time the FDA fundamentally changed how drugs could get approved was in 1992, when it created the accelerated approval pathway. That system let drugs reach patients based on surrogate endpoints (lab markers that are "reasonably likely" to predict real clinical benefit) with the requirement that companies run confirmatory trials afterward.
Accelerated approval was revolutionary, but it still assumed you could recruit enough patients for a meaningful trial. The PM Framework goes further. It untethers approval from statistical power entirely and ties it to mechanistic proof. If you understand why the disease happens, and you can demonstrate your therapy addresses that cause at a molecular level, the framework says that's substantial evidence.
FDA Commissioner Marty Makary called it a "major step" in removing barriers for ultra-rare disease patients. CBER Director Vinay Prasad went bigger, describing it as a "revolutionary advance in regulatory science." CDER Acting Director Tracy Beth Høeg predicted it would drive innovation, improve safety, lower costs, and expand access.
The Fine Print (and the Risks)
The guidance is still a nonbinding draft, open for 60 days of public comment via the Federal Register. It's not law yet, and the details (exactly how many patients, what level of natural history data counts as "well-characterized," how post-market monitoring will work in practice) will be hashed out case by case between sponsors and the FDA.
There are legitimate questions too. Approving therapies based on tiny datasets means accepting more uncertainty about long-term safety. Gene editing is still relatively new, and off-target effects, where the editing tool cuts DNA in the wrong spot, remain a theoretical concern that gets harder to detect with fewer patients. The framework tries to address this with mandatory post-market surveillance, but there's an inherent tension between speed and certainty.
Some observers have also wondered how far the framework could extend beyond ultra-rare diseases. For now, it's clearly scoped to conditions where traditional trials are genuinely infeasible. But once you establish the principle that mechanistic evidence can replace randomized trials, the boundary gets harder to police.
The Bottom Line
The FDA just told biotech something it's been desperate to hear: if you truly understand the biology and can prove your therapy fixes it, we'll work with you, even if your patient population fits in a minivan.
For the families of children with ultra-rare genetic diseases, kids who've had no hope of treatment because their conditions were too small for the old system to care about, this framework could be life-changing. For the biotech industry, it opens a regulatory path to hundreds of diseases that were previously untouchable.
And for everyone watching the FDA, it raises a profound question about the future of medicine: in an era where we can design a therapy for a single human being's unique genetic code, should we really still be demanding the same evidence we'd want for a mass-market blood pressure pill?
The FDA's answer, for the first time, is no.
Deals and Strategy
5 min read
The Cancer Drug That Plays Dead Until It Finds Its Target
Astellas just dropped $335 million upfront on a cancer drug that stays completely dormant until it reaches the tumor; then it wakes up and attacks. The early data on VIR-5500 is turning heads, and the deal's timing tells you everything about Astellas' post-XTANDI anxiety.
