FDA proposes a single license for gene-editing therapies tailored to different patients’ mutations, allowing new mutations to be added based on laboratory data, usually without a new clinical trial

FDA proposes a single license for gene-editing therapies tailored to different patients’ mutations, allowing new mutations to be added based on laboratory data, usually without a new clinical trial

On October 27, the FDA’s Center for Biologics Evaluation and Research published draft guidance. Under the proposal, a single license would cover multiple versions of a therapy within an “individualization space,” defined limits on how the therapy can vary from patient to patient.

A single inherited disease can be caused by hundreds of mutations in the same gene. Some occur in only a handful of patients or just one person, and the DNA editor must be tailored to each mutation. This often involves changing the guide RNA, a short molecule that directs the editor to the intended site in the genome. Under the conventional approach, each version of the therapy needs its own development program, clinical data, and license application. “That approach is not feasible or necessary,” Andrew Byrnes, director of the center’s Division of Gene Therapy 1, said during a webinar.

In May 2025, the New England Journal of Medicine published a report on an infant, KJ Muldoon, for whom a personalized DNA editor was developed within seven months of birth. As BioPharma Dive reported, former center director Peter Marks suggested in an editorial published that same day that approving a therapy’s “general approach” and changing a single component, such as the guide RNA, could “turn a therapy for one patient into a therapy for many, making such treatments commercially viable for rare diseases.”

The applicant defines the boundaries of the “individualization space” and supports them with clinical data. Within those boundaries, components such as the guide RNA and an editor selected from a specified list can vary; the delivery system, formulation, and dosage remain the same. The FDA illustrates this with a hypothetical table for a base editor that corrects a single DNA “letter.” The first four rows list mutations in one gene that have been evaluated clinically, each with its own guide RNA. The final row sets a rule for future mutations in that gene: any C-to-T or G-to-A substitution.

If a new guide RNA falls within these boundaries, the mutation is added to the license through a supplement as a manufacturing change. A new therapy requires evidence of safety and efficacy; a manufacturing change requires evidence that quality and activity have been maintained. There are two laboratory checks. The first establishes that the ability to make the intended edit in cells meets a threshold derived from batches that were effective in clinical use. The second uses sequencing, which reads the DNA of human cells, to look for unintended edits elsewhere in the genome. Each guide RNA has its own set of potential off-target sites. If the relationship between laboratory and clinical findings is understood and predictable, the FDA generally does not expect clinical data.

The clinically derived threshold applies only to therapies that use the same editing mechanism. An editor based on reverse transcriptase, an enzyme that copies RNA into DNA, cannot be added to a license for a base-editing therapy. It has its own risks of unintended edits and immune responses and requires a new application. A mutation in the same gene that causes a clinically different disease constitutes a new indication and requires clinical data.

An analyst at investment bank TD Cowen described the FDA’s approach as “plug and play.” Over October 27–28, shares in DNA-editing developers Prime Medicine and Beam Therapeutics rose by 48,5% and 21%, respectively; on October 27, Beam presented the first clinical data for BEAM-301. A principle similar to the FDA’s proposal underlies BEAM-304 for phenylketonuria: delivery and manufacturing are shared, while the editor is tailored to the mutation.