One of gene editing’s hardest regulatory problems appears when a therapy is made for very few people—or only one. A treatment may use the same manufacturing platform and editing machinery as another product, yet change its guide sequence to match a different disease-causing mutation. Repeating an entire conventional development programme for every variant can be scientifically and financially impossible.
On 27 October 2026, the US Food and Drug Administration issued draft guidance proposing a framework for handling multiple individualized genome-editing variants within a single Biologics License Application, or BLA.
The document is a regulatory proposal, not a final rule and not an approval of any gene-editing treatment. Comments are due by 27 November 2026.
The key idea is an Individualized Feature Space
The proposal introduces an Individualized Feature Space, or IFS. This is a set of pre-defined boundaries describing how a product may vary from patient to patient while remaining part of the same licensed product family.
For example, the editing component might change to target different mutations in one gene, while the delivery system, manufacturing process, dose form and quality controls remain within specified limits. The applicant would have to define those limits in the BLA rather than treating “personalised” as an open-ended category.
If a later patient-specific variant stayed within the approved IFS, a sponsor could seek to add it through a supplement supported by laboratory data. The HHS announcement says the framework could rely on the clinical evidence supporting the original licence plus in-vitro data for the new variant, without automatically requiring a new clinical trial for each patient.
One licence does not mean one safety test
The draft document still expects sponsors to define pharmacology, biodistribution where relevant, immune effects and toxicology. It recommends pre-specified next-generation sequencing methods to look for unintended editing for each variant, using appropriate target human cells and other cell types informed by earlier biodistribution work.
Potency also matters. A guide that works efficiently against one mutation may behave differently against another because sequence context changes binding and editing. Manufacturing validation must show that the process can repeatedly produce material within specification, even as the individualized component varies.
The framework therefore tries to reuse platform knowledge without pretending that every edit is identical. Data and methods may be shared across variants, but each variant still needs evidence that it falls inside the accepted boundaries.
Labelling and naming become technical safety tools
Patient-specific products create a traceability problem. The wrong variant given to the wrong patient could be ineffective or harmful. FDA’s draft addresses naming, identification, labelling and indication statements so that manufacturing records, clinical orders and post-market surveillance can distinguish variants.
Post-approval management is equally important. Long-term follow-up is common for gene therapies because unintended genomic changes or immune effects may appear later. A platform licence would need a system that connects each treated patient to the exact product variant and manufacturing lot.
The proposal could change economics, not biology
A shared BLA could remove duplicated regulatory work and make therapies for ultra-rare mutations more feasible. It cannot solve the underlying challenges of finding patients, making bespoke products quickly, delivering editors to the right tissue or paying for one-off manufacturing.
Nor does the draft lower the biological risk of off-target edits. It proposes a way to organise evidence around that risk. Whether the framework is practical will depend on how broadly FDA accepts an IFS, what data are required for each supplement and how consistently manufacturing platforms perform.
What happens next
The Federal Register notice opens the guidance for public comment. FDA may revise it before finalisation, and draft guidance is non-binding even after publication.
The significance is therefore architectural rather than clinical. Regulators are trying to create a product category that sits between mass-manufactured medicine and a completely new drug for every patient. No patient receives a new therapy because this document exists—but it could determine whether future one-person edits can move through a repeatable approval pathway.
Featured image: Julia Koblitz/Unsplash. Contextual laboratory photograph; not an FDA-reviewed product.
Leave a Reply