FDA proposes approving individualised gene therapies without a randomised trial

The Food and Drug Administration issued draft guidance on 23 February setting out how sponsors could establish substantial evidence of effectiveness and safety for individualised therapies when randomised controlled trials are not feasible because the patient population is too small [s1].

The guidance, from the Center for Biologics Evaluation and Research and the Center for Drug Evaluation and Research, is titled "Considerations for the Use of the Plausible Mechanism Framework to Develop Individualized Therapies that Target Specific Genetic Conditions with Known Biological Cause," and comments must be submitted within 60 days of publication in the Federal Register [s1].

It applies specifically to genome editing and RNA-based therapies such as antisense oligonucleotides, while leaving open that the framework may extend to other tailored therapeutics provided they directly address the underlying specific cause of the disease [s1].

The problem it is trying to solve

Randomisation is the tool regulators rely on to separate a drug's effect from everything else that changes over time. It requires enough patients to form groups.

For an ultra-rare genetic condition — sometimes a handful of people worldwide, sometimes a single patient with a private mutation — there are no groups to form. The evidentiary standard that protects patients in common diseases becomes, in this setting, a standard that cannot be met at all. That is the gap the draft addresses.

What would substitute for randomisation

The guidance focuses on therapies that target a specific genetic, cellular or molecular abnormality and are designed to correct or modify the underlying cause of disease [s1]. FDA lists key criteria [s1]:

• identifying the disease-causing abnormality;
• demonstrating the therapy targets the root cause or proximate biological pathway;
• relying on well-characterised natural history data in untreated patients;
• confirming successful target drugging or editing.

For traditional approval, therapies should demonstrate improvement in clinical outcomes, disease course, or biomarkers if those biomarkers are established to predict clinical benefit [s1].

Two of those criteria are doing most of the work. Natural history data replaces the control arm: if the untreated course of a disease is well characterised, an individual patient's trajectory can be compared against it. And confirmation that the therapy actually reached and modified its molecular target is what distinguishes a mechanism-based argument from a hypothesis.

FDA acknowledges the statistical exposure directly. It recognises that an adequate and well-controlled clinical investigation in this context will involve a small sample size, and therefore says results should be sufficiently robust to exclude chance findings [s1]. When determining effectiveness, the agency says it considers the specific disease, the strength of the evidence, and the challenges of conducting clinical investigations for individualised therapies [s1].

The platform provision

The most structurally significant part of the draft concerns how one approval could extend to variants of a product that were never tested.

Because genome editing technologies are designed to be highly specific to unique DNA sequences, FDA says a product targeting different mutations in a single gene could be included in a single product application and potentially evaluated through master protocols that assess these product variations in a single trial [s1]. A highly supported plausible mechanism of action may then be used to support adding other genome editing product variants intended to treat patients with mutations that were not included in the clinical trial supporting the original approval [s1].

That is the reusable-platform idea applied to regulation: approve the editing approach and the mechanism once, then extend it to new mutations in the same gene without repeating the clinical work. It is also where the framework carries the most risk, since the extension rests on the assumption that mechanism transfers cleanly across mutations that differ in ways the original trial did not test.

What the agency said about it

The announcement carries statements from four officials [s1].

Health and Human Services Secretary Robert F. Kennedy, Jr. framed it politically: "President Trump promised to accelerate cures for American families — and we are delivering, especially for children with ultra-rare diseases who cannot afford to wait" [s1].

FDA Commissioner Marty Makary described the guidance as "a critical step the FDA is taking to tailor our regulatory approach to patients with ultra-rare conditions" [s1].

Chief Medical and Scientific Officer and CBER Director Vinay Prasad said that "after 25 years the FDA has, for the first time, outlined a framework to facilitate these approvals," calling the Plausible Mechanism Framework "a revolutionary advance in regulatory science" [s1].

CDER Acting Director Tracy Beth Høeg said the agency anticipates the draft "will inspire industry to place increased focus on individualized therapies, thereby driving innovation, improving safety, lowering costs and offering more patients with ultra-rare diseases a unique shot at a life-saving treatment" [s1].

What is not settled

This is draft guidance, not a rule, and guidance does not bind the agency or sponsors. It describes current thinking and is subject to change after the comment period.

Several things the announcement does not specify will determine how the framework behaves in practice. What threshold of natural history data counts as well characterised. How "sufficiently robust to exclude chance findings" is operationalised when n is one or two. What post-approval evidence obligations attach, and what happens when a mechanism-based approval is followed by outcomes that do not match. And how far beyond genome editing and RNA therapies the framework extends.

The tension is straightforward and is not resolved by the draft: a mechanism that is plausible is not a mechanism that is demonstrated, and the history of drug development includes therapies whose mechanism was correct and whose clinical effect was absent or harmful. The framework's value depends on whether the four criteria are applied strictly enough to catch that case.

What to watch

Who comments, and on what. Rare disease organisations, gene therapy developers and clinical trialists have different exposures here, and the comment docket will show where the disagreements are.

Whether the final guidance retains the platform extension provision in its current form. That is the provision with the largest downstream effect.

And whether any application actually uses the framework. A pathway with no traffic tells you the criteria were set where sponsors could not reach them.

Sources

• [s1] U.S. Food and Drug Administration, "FDA Launches Framework for Accelerating Development of Individualized Therapies for Ultra-Rare Diseases," 23 February 2026. https://www.fda.gov/news-events/press-announcements/fda-launches-framework-accelerating-development-individualized-therapies-ultra-rare-diseases