Accelerated approval—controversial, yes—but absolutely necessary for small, pre-revenue biotechs trying to find new treatments for small patient populations.
That’s according to the FDA’s Peter Marks, M.D., Ph.D., who was speaking at the American Society of Gene and Cell Therapy annual meeting Wednesday.
Marks, director of the Center for Biologics Evaluation and Research (CBER), provided his thoughts on how to reduce costs for the many small companies working in gene therapy; an area with great potential, but also risk.
Using a biomarker to obtain accelerated approval is key to get an initial approval for therapies that address devastating diseases, rather than wait for a clinical endpoint to show efficacy in a longer trial.
“The wherewithal to do a three-year study or a four-year study without having a revenue stream, it’s just beyond many companies that are startups. So having the accelerated approval process is a way to get there,” Marks said during the meeting.
This is particularly important for neurologic diseases, where a traditional endpoint of how a patient feels, functions or survives could take many years to parse out in a clinical trial, he added.
As long as a biomarker is scientifically validated, Marks said, more products can get to the patients that need them.
“As a matter of policy, I would much rather take the chance that we’re occasionally going to make an error and give something an accelerated approval, than have people so desperate that they’re going out and either going overseas to get unproven therapies—which happens—or using other pathways in which the FDA doesn’t have much regulatory oversight to be able to get therapies,” the regulator said.
And the FDA does, sometimes, have to withdraw treatments that have been granted accelerated approval. For gene therapy, Marks says it’s a much easier process because they are one-and-done. You don’t have to deal with patients tapering off or finding other treatment options, particularly if they believe they are seeing benefit.
Marks also cautioned against calling a failed confirmatory trial the end of the road for a treatment.
“There’s the idea is okay, well, if it fails, the products failed, but that’s not necessarily the case. Unfortunately, clinical trials are a uniquely human endeavor, right? We make clinical trial constructs, and humans are less than perfect. Maybe AI one day we’ll design always perfect trials,” Marks said. “Sometimes we design trials that it’s not the drug that’s the problem, it’s the fact that the trial doesn’t have an endpoint that is either sensitive enough, or that the endpoint cannot be measured in a timely manner such that you don’t get an answer.”
So the FDA takes a careful look at a gene therapy confirmatory trial to see if it was an inherent problem with the study design or if the therapy itself is bad.
Marks continued: “That’s what we should be doing for every product, but I think it’s particularly true if we think we’ve leveraged the science correctly for the initial accelerated approval just to make sure that downstream there isn’t something in the confirmatory trial that led us astray. Because there’s nothing wrong with having to do a second confirmatory trial if you think it’s the trial design, and not the product itself.”
As gene therapy turns towards more common diseases, Marks said the FDA and payers may prefer traditional approvals. For companies, confirmatory trials cost extra money on top of an already expensive clinical program.
Elsewhere in gene therapy, Marks envisioned a regulatory future where gene editing companies can get authorization to treat all the mutations in a gene without having to come back to the FDA for each individual use case.
“It may be that we would be able to find a way to reduce the burden here of having to come back to the agency once you’ve shown that your editor is doing what we’d like to see it do in a given disease gene,” Marks said.
That was music to biotechs companies’ ears at the ASGCT conference, but Marks cautioned, “this is not in stone yet.” He urged regulatory executives at these companies to be thinking about the safety differences between each different guide used and the initial mutations to tackle to prove safety. There may be 30 different mutations for the gene you’re going after, but Marks said they’d want to see a more streamlined case made for three or four.
“Look, If every base pair change, we’re going to have people come back to the agency, there are a lot of nucleotides out there that can go wrong in the genome. And I don’t think we can have 3 billion visits to the FDA a year so we’ll have to figure this out,” Marks said.