Can Personalized Gene Editing Become a Repeatable Business?

Kiran Musunuru helped build a gene-editing treatment for a single Philadelphia-area infant last year, a child known as Baby KJ, and the field has spent the months since asking whether that can be done for anyone else.

"The larger question became, could the same kind of customized therapy created for Baby KJ's mutation also be scaled for many patients with different mutations?" Musunuru said. "The only scientifically valid way to tell is through research and a clinical trial."

That trial now exists. Beam Therapeutics, the Cambridge, Mass.-based base editing company, cleared an Investigational New Drug application with the U.S. Food and Drug Administration in June for BEAM-304, a treatment for phenylketonuria, a rare inherited metabolic disease. Beam plans to test one editor against the most common PKU-causing mutation, then a second editor against another mutation, inside the same Phase 1/2 trial.

The science was already there, but the regulatory structure wasn’t. Until now, each disease-causing mutation meant its own application, its own trial, and its own manufacturing package, which is why bespoke gene editing has stayed closer to a research result than a product. Beam is testing whether regulators will let a company build once and reuse, and the FDA has since signaled it will go further.

Picking Mutations by Headcount

Beam went to the agency with two products rather than one.

"When we initiated a discussion with the FDA, we proposed the fact that we had two products that together could cover about 45% of the PKU patient population, or about 7,000 to 8,000 patients," said Beam president Giuseppe "Pino" Ciaramella, PhD. "The FDA suggested that we should file both products, Beam-304A and 304B, for the two different mutations under the same IND."

On Oct. 27, the agency published draft guidance that would extend the same logic past the trial stage. Under the proposal, a single biologics license application could cover multiple patient-specific versions of a genome editing product, within defined limits the agency calls an individualized feature space, without a new clinical trial for each one. The same day, the Centers for Medicare and Medicaid Services said it plans to open its Cell and Gene Therapy Access Model to conditions beyond sickle cell disease, which is the first real signal of what reimbursement might look like.

"The important thing is to have a platform that you can rely on," Ciaramella said. "Once the regulatory paradigm has been established, we're cautiously optimistic that what we are doing in the liver will be possible to expand to different diseases and other tissues as well. Many companies are thinking about developing platforms for different diseases using lipid nanoparticle delivery for different technologies."

One Delivery System, Many Guides

The reusable part is most of the therapy. A lipid nanoparticle carries an mRNA-based base editor to the liver, and a guide RNA directs it to the target. Swap the guide RNA and the same delivery system and editor can be pointed at a different mutation in the same gene. The safety and manufacturing work that supports one editor supports the next.

PKU is a useful test case for that. It occurs in about one in every 10,000 to 15,000 newborns in the United States, according to the National Human Genome Research Institute, and Beam estimates about 20,000 people in the country live with it. They cannot break down the amino acid phenylalanine and manage the disease through a restrictive lifelong diet. Untreated, it can cause intellectual disability and seizures. Hundreds of mutations can cause it, but most patients carry one of a handful.

Musunuru, the Barry J. Gertz Professor for Translational Research at the University of Pennsylvania's Perelman School of Medicine, and Rebecca Ahrens-Nicklas, MD, PhD, director of the Gene Therapy for Inherited Metabolic Disorders Program at Children's Hospital of Philadelphia, had been studying PKU in a mouse model and talking with the FDA about an umbrella trial, a design borrowed from oncology that tests several drugs against several mutations at once. Musunuru's lab showed that the tools used for KJ could correct a PKU mutation in enough liver cells to bring phenylalanine down in mice. Beam is now collaborating with their team, starting with R408W, the most common PKU-causing mutation.

The Most Common Mutation First

Prime Medicine, also in Cambridge and working with a related editing technology, dosed the first patient in October in a Phase 1/2 trial of PM577a for Wilson disease, an inherited disorder in which copper builds up in the body. The candidate targets H1069Q, the most prevalent Wilson disease mutation in North America and Europe. Prime is planning to expand its treatment to patients with another disease-causing mutation, and eventually many other less frequent mutations, within the same trial. Same structure, different disease: start with the mutation that covers the most patients, reuse the delivery system for the next one.

The government is funding the same idea. In July, ARPA-H committed up to $160 million over five years to seven academic and industry teams through its THRIVE program, which asks grantees to build platforms that can test multiple treatments for multiple diseases in a single trial. Musunuru and Ahrens-Nicklas are among the grantees.

Nobody Has Priced This Yet

None of this has been tested at the point where it gets expensive.

"The regulatory pathway is untrodden to a large extent, so while FDA seems very amenable to highly effective personalized medicines, we still have to see how this will play out in the real world," said Mohammed Asmal, MD, PhD, Prime's Chief Medical Officer. He noted that while platform approaches have great potential to be both scalable and eventually reimbursable by insurance companies, there is much work to do.

Ciaramella said the clinical trial "opens up the opportunity to start thinking about whether the approach could be deployed to gradually more unique mutations."