FDA's Woodcock: 'The clinical trial system is broken'
The Food and Drug Administration's top drug regulator sharply criticized the current clinical trial regime, calling for industry to embrace more collaborative studies.
"I believe the clinical trial system is broken," said Janet Woodcock, who has worked at the agency for more than three decades, at an industry panel on Nov. 14. "I do not believe it serves the interests of patients."
Instead, the director for the Center for Drug Evaluation and Research pitched platform trials with master protocols, which she argued can more efficiently answer clinical questions. While this design can require competitors to work together, Woodcock said she has seen a "groundswell" of recent support.
Despite the dim view of the current system, Woodcock said she was optimistic that trial research is already in the process of changing. She noted industry and advocacy voices have approached her in recent weeks about starting such studies.
In July 2017, Woodcock made the argument for master protocols in a New England Journal of Medicine review article, which are overarching frameworks that can evaluate multiple treatments and diseases in the same trial.
Three main study types use master protocols: basket, umbrella and platform trials. Generally, a basket trial tests one therapy against many diseases while an umbrella trial evaluates many therapies on one disease.
A platform study focuses multiple drugs on a single disease like an umbrella trial, but uses a decision algorithm to let therapies enter and exit the platform. The CDER head remarked the algorithmic design has led to some skepticism and resistance.
"People have horror and anaphylaxis over this," Woodcock said. "The statistical community believes this is a very valid statistical approach, but this isn't how things have always been done."
One example of the concept in action is I-SPY 2, a Phase 2 platform trial that simultaneously tests five therapies in certain subtypes of early-stage breast cancer. The study, started in 2010, has enrolled more than 1,300 patients and is expected to test 17 drugs by year's end, according to its sponsor.
Some of the tested arms include a combination of Roche's Perjeta (pertuzumab) with Herceptin (trastuzumab), Puma Biotechnology's Nerlynx (neratinib) and a combination of Bristol-Myers Squibb's Opdivo (nivolumab) with AstraZeneca's Lynparza (olaparib).
The investigators gave its first update on the trial in September. Anna Barker, the chair of the I-SPY 2's oversight group, called the design the future of clinical trials.
'We must have patient- and disease-focused trials in the future, and they have to be learning systems," Barker, also former deputy director of the National Cancer Institute, said at the event. "We can't any longer afford these single experiment trials which are actually using up so many patient resources and taking so long and costing so much."
But such designs face a tough road to widespread adoption, especially due to market dynamics at play.
Industry must be persuaded to collaborate when they have been historically competitive, Jim Reilly, a vice president at Veeva Systems, a life sciences technology company, said in an interview with BioPharma Dive.
But Reilly, who works on technology for clinical data management and operations, said he's also observed a positive trend of sponsors being willing to work together, particularly in data transparency and standardization.
Another critical challenge is that platform trials are the most useful in highly competitive research areas, where comparative data and combination results are especially valuable, Woodcock said.
"So those things have to be navigated, but I believe we'll get over a tipping point," Woodcock said. "People will realize the strength of [these trials], then there will be opportunities to go into the community and enlist the community practitioners, which then will result in a tremendous increase in ability to enroll patients and answer more questions."
And other recent reports support the notion there is room for improvement in clinical trials. An analysis by Tufts University's Center for the Study of Drug Development released this July found the typical study has changed dramatically since 2000.
Comparing average Phase 3 pivotal trials from 2001-2005 and 2011-2015, the report found the number of endpoints and eligibility criteria had respectively increased 86% and 61%, while the number of randomized patients decreased 18%.
At the time Ken Getz, a Tufts associate professor who led the analysis, explained that "protocol design scope and complexity have steadily increased" and predicted the trend was likely to accelerate with the focus on rare diseases and specific patient populations.
These trends feed into clinical trial challenges that Reilly observes today, including difficulties in finding patients due to lengthier eligibility criteria.
"There's always a harm in the status quo," Reilly said. "If we operated exactly how we are today, I think it means that we would see the same kind of level of disengagement and lack of coordination."