It's one thing to listen to patients and understand what's important to them. But it's much more difficult to translate the patient voice into a science-based tool that can be used to measure meaningful risks and benefits of a drug in clinical trials.

"This is where we've fallen down the most," the FDA's Janet Woodcock said Wednesday in opening an agency workshop on developing clinical outcomes assessments (COAs). To date, the FDA has qualified only one COA based on what patients deem is important.

Yet a COA can make or break a clinical trial. When a trial fails, it could be that the drug didn't work. Or it could be the trial was asking the wrong question for the defined population. Even if a trial demonstrates a meaningful response, there could be a question of whether that response is meaningful to patients in the real world.

Hoping to take some of the guesswork out of using COAs that reflect patient needs, the FDA is proposing a public compendium of COAs that would encourage drugmakers to collaborate on developing patient-focused measurement tools that could be used in trials across diseases.

Initially, the compendium would include COAs that have been qualified, are in the qualification process, were previously labeled but not qualified or have been used on an exploratory basis. In its second phase, the compendium would be expanded to include unmet measurement needs that reflect what's important to patients, with the idea of encouraging development of tools to meet those needs. A mockup of the compendium is expected to be available later this year.

While academics and industry representatives participating in the workshop hailed the compendium as an important first step, many of them had questions. "It tells you what works, but does it tell you what matters?" asked Tom Sellers, the senior director of patient advocacy at Millennium Pharmaceuticals Inc., a subsidiary of Takeda Pharmaceuticals Co. Ltd.

Dennis Turk, a professor at the University of Washington, noted that historically, COAs measured what academics and clinicians thought was important. Now that the focus has shifted to the patient perspective, there's a difference both in what should be measured and how it's measured. That calls into question the use of COAs developed under the old paradigm. But getting a new patient-centric COA qualified for use in clinical trials can be costly and time-consuming.

Citing those hurdles, Stephen Coons, of the PRO Consortium, said if there is no viable development path forward for a rare disease, it will be hard to get industry interested in investing in developing new measurement tools.

While patient groups generally are willing to provide input, the time and cost involved in developing COAs would discourage them from developing their own tools, Sellers said.

But getting access to patients could help reduce industry costs, Quintiles' Jean Paty said. "You could definitely motivate a bunch of us if you said, 'Here's an unmet need and, by the way, here's a vehicle to get to the patients,'" he added.

While she welcomed the compendium as a way to bridge current gaps, Alicyn Campbell, Genentech Inc.'s global head of patient-centered outcome research for oncology, said what industry really needs are trial endpoints based on patient-reported outcomes (PROs).

Katarina Halling, PRO group leader for Astrazeneca plc, agreed. But there's no way currently for the FDA to review an endpoint until phase IIb, she said. That's too late in the drug development process.

Going forward, the FDA needs to re-evaluate existing COAs to ensure they are patient-focused, Coons said, adding that some legacy assessments shouldn't be included in the compendium. He urged the agency to seek a consensus on which COAs should be included. He also called for a formal mechanism for adding new clinical measurement tools to the compendium.