CER's Place In Drug Development Is A Work In Progress, FDA Says

FDA has a lot of work to do before it can determine how comparative effectiveness research might play into its evaluation of new drug approvals or labeling changes, Chief Scientist and Acting Deputy Commissioner for Science and Public Health Jesse Goodman told a conference sponsored by the Friends of Cancer Research June 18.

He noted that CER is expected to draw from sources such as electronic health records, registries or health care claims in an effort to evaluate real-world use of health services and thus may look very different from the "gold standard" randomized clinical trials the agency is used to.

Asked whether such data might be factored into FDA decisions within the next several years, he replied, "These [comparative effectiveness] studies compare different things but the devil is in the questions you ask, the details."

"There can never be a study that compares every possible intervention at every possible dose and at every possible timing. It is going to be very important that we look at the information produced and say 'did it ask the right question, what is the quality of the data and what does it really show scientifically? How generalizable is it and how do we communicate it?'

"So one of the scientific areas we need to develop is our framework for evaluating this kind of data, for making inferences and conclusions from this kind of data and determining who it is applicable to."

CER And Pre-Market Research

FDA Center for Drug Evaluation and Research Director Janet Woodcock similarly emphasized the fundamental difference between CER and the clinical trial data required for drug registrations in a June 15 speech at the Drug Information Association annual meeting in Washington, D.C.

"Current FDA requirements for market approval versus the evidence base needed for [CE] are quite different," she said. "Trials required to get on market are of limited duration, don't evaluate long term, have restrictive entry criteria, leave out subgroups and there is a lack of comparisons of regimens or different types of therapy."

In summary, "pre-market style clinicals won't give you [outcomes] information because they are not done in a real world setting, so "in some ways the pre-market world is not the time to get this information."

Goodman suggested that CER "enthusiasts" may underestimate the difficulties involved in interpreting and using non-randomized clinical trial information. "We can only get so far with health care data that's non-randomly collected."

Woodcock commented that as the limitations of data and methods for determining CE are worked out, "it's going to be a very bumpy period."

FDA expects it will see more and more data on the comparative effectiveness of drugs, driven by interest among health systems, payers and patients, and the new funding and research initiative provided under the American Recovery and Reinvestment Act and the health reform law. One important development mandated by the law is creation of a Patient-Centered Outcomes Research Institute; public nominees for the institute's board of governors are now being solicited (see chart: " ¹ CER Institute Governing Board Nominations Due June 30 ").

"I see greater pressure on FDA regulatory requirements pre- and post-market to gather this evidence at some point during the drug lifecycle," Woodcock said.

However, she added, "the question is who in the U.S. is responsible for generating this additional evidence and what evidentiary standards will be applied."

Not A Good Time For More Data Requirements?

Noting that drug development is "in crisis," she observed "this is not a really good time to take on more requirements for the drug industry." And for FDA, "this is a huge problem. We're caught between conflicting societal expectations and demand." On one hand, the public wants greater certainty about product outcomes and on the other, wants access to innovative drugs, she pointed out.

She suggested three solutions to help meet demand for CE determinations. First, she said it is "imperative" for the U.S. to develop a clinical trial "infrastructure."

"We don't have the machine to generate all this evidence." It is "time to develop the capacity for finding out these answers in a way that doesn't cost hundreds of millions of dollars every time we ask a question." One solution might be clinical trial programs that are patient driven, such as those being operated in cancer treatment, she pointed out.

Second, Woodcock reiterated her support for progress in personalized medicine as the "best way to improve the value of medicines" (² (Also see "CER And Personalized Medicine Are "Synergistic" - FDA's Woodcock" - Pink Sheet, 2 Nov, 2009.)). She has maintained that CER and personalized medicine are synergistic concepts and "I think we're there on the science."

Finally, Woodcock pointed to the promise offered by large electronic safety surveillance projects, including FDA's Sentinel system and the drug industry-supported Observational Medical Outcomes Partnership project.

However, she acknowledged "that science is in its infancy, and there are lots of questions we're going to have to answer before those types of analyses and databases can be used to definitively answer the questions we have about outcomes."

In response to Woodcock's comments at the DIA session, Robert Giffin, senior research director with the Center for Medical Technology Policy, expressed concern with the prospect that adding CER to the current drug registration requirements will result in a lot of "parallel" studies. "We simply do not have the resources to do it, period," he said.

REMS Building To CER?

"There's got to be some way to bring these together," Giffin suggested. One way may be to "give some teeth" to the risk management evaluation strategies attached to drug approvals so that over time, "these can become comparative effectiveness studies," though he did not elaborate on that idea.

In addition, he said FDA could "start to really go to superiority trials, and have the regulatory approval based more on CER studies."

Founded by former Centers for Medicare and Medicaid Services Chief Medical Officer Sean Tunis, the Center for Medical Technology Policy is developing recommendations on increasing the use of "pragmatic clinical trials," in drug development.

Such trials would be designed to satisfy FDA approval requirements and at the same time provide outcomes data for payers, clinicians and patients. Former CMS Coverage and Analysis Group Director Steve Phurrough serves as chief operating officer and senior clinical director for CMTP.

- Cathy Kelly ( ³ [email protected] )