Another push by the FDA to streamline clinical trials leverages a statistical method that compiles previous findings rather than starting with a blank slate.
The agency’s draft guidance in January could particularly benefit pediatric, oncology and rare disease trials as part of a broader FDA trend to permit more flexible trial designs with increased use of real-world data.
The guidance encourages the use of Bayesian statistics in drug trials. Instead of starting from scratch and only adding new information from the trial, the framework allows researchers to incorporate preexisting data at the outset, including details about drug performance gathered from earlier trials, or predictions based on similar drugs.
The trial can build on that foundational knowledge, essentially providing a head start for faster progress.
“The Bayesian guidance is basically saying, let's not look at each clinical trial as its own isolated experiment,” said Meri Beckwith, co-CEO of Lindus Health, a clinical research organization.
These modified trials can more rapidly determine success or futility and make rare disease trials with tiny patient populations more feasible. With more external information to draw from, trials can be smaller, which is crucial for uncommon diseases that only affect a few thousand people worldwide, Beckwith said.
"Bayesian methodologies help address two of the biggest problems of drug development: high costs and long timelines,” said FDA Commissioner Dr. Marty Makary in a press release announcing the draft guidance. “Providing clarity around modern statistical methods will help sponsors bring more cures and meaningful treatments to patients faster and more affordably.”
The statistical model also helps sidestep thorny ethical issues, enabling late-stage cancer trials where a placebo arm is inappropriate, Beckwith said. By contrast, more traditional trials are viewed in isolation, making it harder to adapt to these complexities.
Drug trials have used a Bayesian statistical approach in the past, but only on a case-by-case basis.
“There were a number of gene therapy companies who got approval using a Bayesian approach over the last couple years. They were touted as the exception rather than the rule,” Beckwith said.
While the draft guidance hasn’t seen much fanfare, Beckwith calls it “a huge deal.”
Still, using this methodology isn’t a free pass to approval. The draft guidance introduces a number of guardrails to ensure that trials are rigorous. For example, if external data isn’t chosen carefully, it can introduce bias into the trial.
And implementing a Bayesian approach isn’t necessarily easy.
“There aren’t many biostatisticians who have really deep expertise and can oversee something as complex as a phase 3 trial using the Bayesian approach,” Beckwith said.
Beyond its technical limitations, the biggest barrier to adoption may be the industry’s reticence to move beyond historical practices.
“The challenge is more a cultural one,” he said. “This industry is extremely conservative.”
But if the Bayesian approach is adopted more broadly, it could help expand clinical research for underserved groups and pave a smoother path to market for new drugs, Beckwith said.
The FDA is currently accepting comments on the draft guidance until March 13.
