FDA Clears Beam Therapeutics to Launch BEAM-304: Platform Base Editing for Phenylketonuria

FDA clears Beam to begin testing BEAM-304: the therapy is designed to make precise DNA changes in the liver for phenylketonuria and function as a program for multiple mutations

Beam Therapeutics has received IND clearance for BEAM-304, an experimental therapy for phenylketonuria. The first Phase 1/2 trial will start with patients carrying the R408W mutation in the PAH gene, and the company will then add an editor for a second mutation. Beam wants to test a platform model: liver delivery and manufacturing stay the same, while the editors are swapped out for specific mutations.

On June 18, Beam Therapeutics announced that the FDA had cleared BEAM-304 for its first human trial. Phenylketonuria, or PKU, may sound like an obscure line from a school textbook, but for patients it means lifelong control over what they eat. The body cannot properly process phenylalanine, an amino acid found in protein-rich foods. When blood levels rise, the brain suffers. In the US, children are screened for PKU after birth, but early diagnosis does not eliminate the need for a strict diet, medical food, and chronic disease management.

BEAM-304 is aimed at the root cause. The drug is designed to deliver lipid nanoparticles to the liver carrying base editing tools. This kind of editing changes a single DNA “letter” without the usual CRISPR cut through both strands. In PKU, the target is the PAH gene, which encodes an enzyme that helps break down phenylalanine.

What the FDA has cleared now is not the sale of the therapy, but an IND. That is the application that allows a company to begin human testing. Beam is planning a Phase 1/2 study: first safety and tolerability, then reductions in blood phenylalanine and the possibility of easing the diet. The first group will be patients with the R408W mutation. After that, the company wants to add an editor for a second common mutation.

Beam is building BEAM-304 around the shape of the program itself. Rare diseases are notoriously hard to make work economically: one mutation may mean a tiny patient group, but the developer still needs separate experiments, manufacturing, regulatory filings, and clinical validation. Beam is proposing a different route: one clinical platform in which the editor changes for a specific mutation, while delivery, manufacturing, and part of the evidence base remain shared.

In early June, the FDA had already outlined reusing evidence for similar gene therapies: manufacturing, preclinical work, clinical information, and public scientific data. The regulator is not saying, “you can skip everything.” It is saying: if the product is genuinely similar to an earlier one, show which data can be reused and why.

Without that kind of framework, personalized genetic medicine remains a showroom for a handful of well-funded diseases. If every rare mutation requires a completely separate path, small cohorts will always lose to economics. If the regulator accepts part of the platform evidence, developers can build therapies in series rather than as one-off handcrafted projects.

BEAM-304 does not yet have human data. The company does have mouse data showing the therapy normalized phenylalanine levels, and it has said it will present updated preclinical results at the FASEB conference in July. The real test will come later, when the first human data arrive on safety, phenylalanine levels, and whether patients can actually relax their diet.