Rep. Rosa DeLauro's, D-Conn., hearing on the safety of GlaxoSmithKline's diabetes drug Avandia has ratcheted up the pressure on FDA to make broad reforms of its review structure, but the congresswoman has been careful not to urge a particular course of action for the agency.

The upshot is that FDA could have as much as two years - before Congress begins work on the user fee renewal program in earnest - to show that it has made meaningful reforms before some legislators attempt to make them for the agency.

Another prominent drug safety controversy could of course alter the political dynamic, but for now, FDA appears to have some room to allow its existing processes to run their course, even as DeLauro signals a lack of confidence in the agency.

In remarks after the hearing, DeLauro said that the next steps she would like to see were not necessarily a splitting of the Center for Drug Evaluation and Research in order to create a new FDA center devoted to post-marketing drug safety. Such an approach is among the more extreme reform plans offered for FDA, and is laid out in legislation sponsored by Sen. Charles Grassley, R-Iowa. Rep. Maurice Hinchey, D-N.Y., has introduced the companion bill to Grassley's legislation.

DeLauro Wants An Independent Process

Grassley was a panelist at the April 28 hearing of DeLauro's Appropriations Subcommittee on Agriculture, where he presented the findings of his recent staff report on FDA's handling of the Avandia data (¹ Also see "Grassley Wants An Rx Safety Reorganization; Can FDA Keep Him At Bay Again? Congressional Avandia Investigation Reopens Door On Splitting CDER" - Pink Sheet, 1 March, 2010.).

DeLauro demurred on endorsing it: "I want to take a look at that," she said after the hearing. "Clearly it's all about the process, how to make it stronger."

Even if DeLauro decides to champion an FDA reform bill, she can only advance the legislation obliquely. "I think we would like to address some of these issues and see, not being authorizers, but appropriators, how far we can go with addressing some of these issues to make the changes that are necessary to ensure the safety of these drugs," she said.

DeLauro noted, "There are IOM studies. There are GAO studies. My hope is those recommendations will be taken. There may be some new ones, but there are some that are there already, including making the drug review process an independent process ... Let's look at some of these things and put it into practice."

DeLauro isn't talking about making drug safety independent per se, she's talking about drug reviews themselves, reflecting her long-standing concern about what she perceives as the original sin of user fees. Even from her key appropriations perch, though, DeLauro is not going to be able to wean FDA from drug user fees, but if the Democrats retain control of the House in the mid-term elections, she will likely be a significant roadblock to any new programs (see ² Also see "Generic User Fees Face Opposition From Key House Appropriator DeLauro" - Pink Sheet, 3 May, 2010.).

Even assuming that drug reviewers remain under a single-pharma funded roof, panelists at the hearing offered a number of concepts that FDA could implement to mollify its congressional benefactor.

The agency's first step in showing it can successfully handle a drug safety crisis is the July advisory committee meeting to review the most recent Avandia data. Anything short of a recommendation to withdraw the drug will be met with skepticism by agency critics, but the event will allow FDA to show that it is working with outside experts on the issue.

Indeed, DeLauro's decision to not invite FDA or GSK to the hearing shows some respect for that ongoing process - even as it denies them an opportunity to defend themselves (see sidebar: " ³ How DeLauro Set The Agenda: Principal Avandia Players Not Invited To Hearing ").

Beyond the advisory committee, witnesses at the hearing made a number of suggestions for how FDA should handle safety issues, some of which industry and the agency could find even more disruptive than the rival drug review bureaucracy envisioned by Grassley.

Opening Up The Data Book

Under one scheme, FDA essentially would cease to be the final arbiter of clinical trial data. Harlan Krumholz, a Yale University School of Medicine professor, told the subcommittee that all clinical trial data, not just summary data, should be made publicly available to allow for independent review and interpretation.

Krumholz said independent evaluations of trial data by at least two non-industry expert groups should be conducted when concerns surface to verify findings.

Krumholz also said he wants the raw data files containing all data collected during a clinical trial to be published, along with summary findings, within a year of the study's completion.

"Once it's approved, I believe the privilege of selling in the marketplace comes with the responsibility of providing" the data, he said. "All the information that's available to know about the drug should be in the marketplace of ideas."

Krumholz said there likely would be different interpretations of the data when released, but at least the data would be available for physicians and patients to see.

"If you can't as a company pull together your raw data files, it makes me wonder about the quality of your internal research," he said.

The idea would be a step beyond the transparency requirements set out in the 2007 FDA Amendments Act, which requires only summary data to be released within a year of a study's completion.

Several comments to a question FDA asked last year about whether data from trials related to unapproved drugs should be made public said all trial data add to the global body of knowledge and should be released (⁴ Also see "Unapproved Product Trial Data: To Post Or Not To Post On ClinicalTrials.gov?" - Medtech Insight, 29 June, 2009.).

The best scenario would include releasing all clinical trial data upon drug approval, Krumholz said. He argued the most rigorous examination of a drug's efficacy requires the raw data. It also would enable the standardization of analyses across studies.

Having two independent groups evaluate the data ensures findings are verified without the possibility of influence, he said.

"We need to know what can be known based on what's already been done," Krumholz said. "There's no reason for the data not to be available."

In the case of Avandia, Krumholz researchers using pooled summary trial data showed increased cardiovascular risks in the famous 2007 meta-analysis by Cleveland Clinic's Steve Nissen that touched off the Avandia safety controversy (⁵ Also see "Grassley Wants An Rx Safety Reorganization; Can FDA Keep Him At Bay Again? Congressional Avandia Investigation Reopens Door On Splitting CDER" - Pink Sheet, 1 March, 2010.). Krumholz said today a lot still is unknown about the drug because no one has been able to review all the clinical trial data.

Right now there is an assumption that because a drug is on the market, it is okay, Krumholz said.

Krumholz also encouraged increased communication between FDA and the industry during the development process as well as dropping requirements that FDA and company scientists cannot interact.

"I don't believe we should be separating people in different corners where they can't talk," he said. "Sometimes (talks) have to occur proprietary because they're early in the development process, but you want the people to be talking to the FDA about what the path is ... "

Increased data access also should not give rise to problems with companies undermining competitors' new products, said Sidney Wolfe, director, public health research group at Public Citizen.

"If things were really transparent and independent, people were able to look at it, the ability of a renegade company or a company like that to be able to undermine something like that would be severely limited," Wolfe said. "It's because so much of this information is secret and not available to everyone that things like that could go on."

FDA Culture Crash

Wolfe sharply criticized the culture at FDA, including its apparent unwillingness to listen to some of its own scientists, and raised questions about the agency's commitment to public health.

When concerns are raised about a drug that has worrisome risks and no unique benefit, while other countries move to pull the drug from the market, FDA seems to acknowledge the risks, but will not take it off the market, Wolfe said.

Wolfe said some FDA experts appear to be ignored when they have concerns about drugs.

"They've changed the name of that office (the Office of Surveillance and Epidemiology) four or five times in response to criticisms, but the function, and the dependence, and the subordinate position of that office compared to the Office of New Drugs has stayed the same," Wolfe said.

Wolfe also criticized Janet Woodcock, director of the FDA Center for Drug Evaluation and Research, for not being tough enough on industry.

Woodcock is "just uncomfortable being a regulator," Wolfe said. "And I think that for someone who's the head of drugs at the FDA, being uncomfortable with being a regulator is incompatible with being the head of that division."

Even Delauro admitted the perception of FDA decision-making concerning drug safety is that it usually does not err on the side of public health.

"Whenever we get into the difficulty and there is a question as to the science, or the outcome, we always, we always appear to act on behalf of the industry," she said. "The notion is let ... this go on for another several years while more people die before we truly do address it... ."

Other divisions within FDA do not take a similar approach, Krumholz said. In food regulation, he said the agency is not afraid to ask for a recall if it finds a problem. "The thermostat is set as if there's a concern, let's be safe," he said.

Krumholz said the standard for drug regulation should not be to assume a product is safe until shown beyond any doubt that it is harmful.

"I think that's too stringent," Krumholz said.

"Emergency Response" Needed

One way to ensure that FDA strikes the appropriate balance is to create uniform process when questions arise about a drug's safety. FDA should have an "emergency response" plan for these situations, Krumholz suggested.

The response process needs to be clearly defined and should "occur rather quickly," Krumholz said. "I just don't think that three years down the line we should be still wondering 'what is the exact process by which this should happen?' We should know this is predictably occurring, and how we deal with it."

An FDA action plan would benefit drug companies as well as the public, he pointed out. Companies should want to know the trigger for a reassessment and the steps that follow, he suggested. "What are the expectations? It's not good for them either to be uncertain about this kind of thing."

- Derrick Gingery ( ⁶ d.gingery@elsevier.com ) & Cathy Dombrowski ( ⁷ c.dombrowski@elsevier.com)

Avandia Aftershocks: Will FDA Drug Safety Review Structures Crumble?

More from Archive

More from Pink Sheet