Pharmacy & Therapeutics, December 2007
Avandia, the type-2 diabetes drug manufactured by Glaxo-SmithKline, is the most recent poster child for drug safety hysteria. A meta-analysis published last May by a noted cardiologist
fueled concern and headlines over possible cardiovascular side effects from the drug (rosiglitazone maleate), forcing the FDA to revisit Avandia’s labeling in the absence of any new, convincing clinical trials. The result: GSK’s agreement in mid-November to expand Avandia’s existing black-box warning with confusing language saying “a risk of an increase in myocardial
ischemic events was not confirmed or excluded in three long-term clinical trials.”
It is that kind of decision-making in the absence of facts that the Food and Drug Administration Amendments Act (FDAAA) of 2007 aims to help eliminate. The bill is a political steroid shot in the arm to the beleaguered FDA, long considered a 98-lb. weakling. As a result, its Center
for Drug Evaluation and Research (CDER) will flex new muscles in 2008. But don’t expect CDER to become Hulk Hogan overnight, if ever. In many ways, despite its size and provisions, the legislation is hardly radical. “It is not a rewrite of authority; [nor] is it a huge addition the
way the 1962 amendments were,” explains Bill Hubbard, who left the FDA in 2005 as Associate Commissioner. He now serves as senior advisor to the Coalition for a Stronger FDA.
“It is a tweaking, but a big tweaking,” he says.
The bill gives the FDA authority to impose restrictions on the distribution of new drugs that have significant benefits—like Avandia—but whose safety hasn’t quite been buttoned down at the time of approval (again, like Avandia). Those up-front agreements are called Risk Evaluation and Mitigation Strategies (REMS). The agency will be able to force companies to perform postapproval clinical trials to pin down unresolved safety problems. If those trials turn up troubling information, the FDA will be able to act quickly to force a label change. None of this had happened with Avandia.
Concerns about Avandia hit the headlines last spring when Steven Nissen, M.D., chairman of the Department of Cardiovascular Medicine at Cleveland Clinic and immediate past president
of the American College of Cardiology, dropped a bomb. The meta-analysis he published in the New England Journal of Medicine in May 2007 combined the results of 42 previously completed small studies to show that Avandia posed a 43% increased incidence of heart attacks. At the FDA’s insistence, GSK had added additional language about heart attacks to Avandia’s
warning label 15 months earlier. The Nissen meta-analysis, however, implied that the company
was soft-pedaling the risk—even though he conceded thathis meta-analysis was “less convincing” than a clinical trial, even though the New England Journal of Medicine itself had seconded the fuzziness of the doctor’s analysis in an accompanying editorial, and even though FDA Commissioner Andrew von Eschenbach, M.D., had noted “serious concerns.” He said that
patients taking Avandia and their health care providers were confused about the drug’s safety because of media reports surrounding the journal article.
Confusion and hysteria are what Congress hopes to eliminate in 2008 and beyond with the provisions of the FDAAA. A more accurate title for the bill might have been “The No More Vioxx, Ketek, Paxil, Avandia, et al., Act of 2007.” The legislation weighs in at more than 400 pages of technical verbiage and contains at least 200 provisions. Many of these provisions have deadlines for implementation and are identified within the bill itself. Although the bill’s drug safety provisions were the subject of numerous headline-producing congressional battles, they were actually an add-on to legislation reauthorizing the FDA to collect user fees from drug companies for the five years starting in fiscal year 2008, which began on October 1, 2007. Those user fees, althoug h controversial, are a critical component of the CDER’s budget.
In fiscal year 2008, according to Alan Goldhammer, Deputy Vice President of Regulatory Affairs of Pharmaceutical Research and Manufacturers of America (PhRMA), companies will pay a total of $459 million in user fees. These fees will accoun t for 62% of the CDER’s budget. The $459 million is an increase of nearly 50% over the $305 million the companies paid in 2007, and it will increase in each of the next four years over the five-year term of the reauthorization. Most of that money has been, and will continue to be, used for the review of new drugs. However, the bill sets aside $54 million for drug safety activities in fiscal year 2008; that figure will also increase steadily over the next five years, and much of the $54 million will be used to hire
staff in the area of postmarketing surveillance. By the end of 2008, the CDER’s staff will have grown by about 500 employees, a 20% increase. Congress expects to see a bigger, stronger, better-staffed CDER win more wrestling matches with pharmaceutical companies over drug safety, but the companies aren’t so sure that a larger CDER will necessarily mean fewer drug safety headlines.
Dolly A. Judge, Pfizer Vice President of Federal Government Relations, says: I think more interactions are better for drug safety management purposes. But I know the optics of the industry and agency meeting more frequently is not necessarily better politically. Senator Grassley and others have been critical of the relationship between industry and FDA, so more meetings could further enhance a false perception that the relationship is too close or somehow it is unholy, so to speak.
Certainly, the FDA has met with drug company representatives in the past, both before and after drug approvals, but the agency’s leverage has been limited. In the past, for example, the CDER’s negotiations with a company like GSK about adding or upgrading a warning about heart attacks to a drug like Avandia could go on for years, as happened with Merck and its pain relieve r rofecoxib(Vioxx), for example. It has been happening, apparently, with Avandia in the wake of the publicity around Dr. Nissen’s article and an FDA advisory com mittee meeting last summer.
In the past, the only power the FDA had in ordering a label change, after a drug went on the market, was to threaten to force a company to take the drug off the market. The FDA rarely used that authority, because such an action might have hurt patients already taking the drug. The FDAAA sets up a quicker pathway for making label changes. With the new step, the company and the FDA would move quickly through a process of proposals and counterproposals.
If there is no agreement, there is a time limit for ending the company’s appeals and
forcing it to make changes dictated by the FDA. Normally, these negotiations would last no longer than 120 days and might be substantially shorter. Why have these labeling discussions
taken so long? Part of the reason is the absence of clinical trial data; this lack has fostered
much uncertainty. Again, when the FDA approved Avandia in 1999, the drug was considered a real benefit for patients with type-2 diabetes. At that time, the FDA knew that adverse reactions might be a problem, but it had little leverage to force GSK to conduct a major postmarketing clinical trial directed at pinning down the risk of a heart attack. Because of the cost of clinical trials, GSK had little incentive to undertake such a trial on its own. Subsequently, many small studies were performed on Avandia, but heart attack was not an endpoint in any of them. This series of events led to Dr. Nissen’s hysteria-promoting article in May 2007.
Under the FDAAA, the FDA can force a study or a clinical trial in conjunction with a new drug approval if the scientific data are deemed appropriate. That authority has been absent in the past. A study can be ordered (1) to assess a known serious risk, (2) to evaluate signals of serious risks, and (3) to identify an unexpected serious risk when the available data indicate the potential for such a risk. If any of these three objectives can be achieved by the active
postmarketing risk-identification and analysis system, which this bill charges the FDA to set up, the agency cannot order a study. A clinical trial can be mandated only when a study cannot accomplis h the three objectives. The pharmaceutical company sets the timetable for completing the study or trial. If those deadlines are missed, the FDA can levy civil penalties.
The FDAAA allows the FDA to fine a company $250,000 for missing a deadline by 30 days; that fine is doubled every 30 days until it reaches $1 million. The FDA has this same authority to impose a penalty if a company fails to make post-approval label changes after the FDA mandates them. The FDA’s authority to order post-approval studies and clinicaltrials is constrained somewhat by the need to consider whether it can obtain the information it needs from the new active active postmarketing risk-identification network. With this new
network, the FDA would link up with databases currently established by the Department of Veterans Affairs, the Centers for Medicare and Medicaid Services, and by private health insurers. The new network will supersede MedWatch (the current system for reporting adverse reactions); however, this system is considered almost useless. MedWatch depends on reports from health care providers and individuals. Unfortunately, duplicate reports are often filed. Moreover, the FDA must then go back and interview those who sent in reports. All of this takes an enormous amount of time. Even after this step, the agency has no “denominator” (hard facts on how many other patient s had the same adverse reaction, much less how many people are taking the drug).
With the new postmarketing database, those data will no longer be available from MedWatch; the FDA will be able to go into the database and look for trends (e.g., cardiovascular
events tied to a particular drug). Setting this network up is going to take time, though. The FDA will have two years to develop the data sources. By July 1, 2010, at least 25 million patients must be in the database; by 2012, 100 million must be included. On its own, the FDA had already awarded contracts in this area. The new user fees for drug safety, flowing into the CDER budget starting in fiscal year 2008, will allow the FDA to accelerate its efforts to meet the
timelines in the FDAAA. Although FDA officials claim that they will be able to meet those deadlines, this new network might cost considerably more than what the FDA will have in user fees to pay for it.
Given the failure of Congress to appropriate adequate funds for drug safety, no one expects future FDA appropriations to add congressional money to the user fees that will be available for
the database. That is why the agency hopes to also supplement user fees with funding provided by the Reagan–Udall Foundation, also being established by the FDAAA. The fund has a
much broader purpose, however. It is a private foundation, modeled on the ones already used by the Centers for Disease Control and Prevention (CDC) and the National Institutes of Health
(NIH). The fund will raise its own operating budget. That money will be spent on collaborative research in which drug companies, academic centers, and others join together for research that
would be dispersed broadly and whose results generally would not be patented. The kinds of research projects to be undertaken would be those on the FDA’s Critical Path Initiative (CPI) list.
The FDA established the CPI in March 2004. The CPI list includes 76 projects, but the FDA has funded only half a dozen or so. These projects cover biomarkers, animal models, inter-individual
variability in drug response, data analysis technology and methodology in drug development, and improvements in designing and conducting clinical trials. Some are concerned that because the Reagan–Udall Foundation will depend on private funding and because most (if not all)
of that funding will come from the drug companies, it will serve the interest of drug companies, even though its 14-person board of directors will have representatives from widely diverse
groups. Only four directors will be from the pharmaceutical, device , food, cosmetic, and biotechnology industries.
Steve Walker, co-founder of Abigail Alliance, a patient advocacygroup, says,
“The Reagan–Udall Foundation is the only provision of the bill with real potential to improve the FDA, but there is a real chance it won’t accomplish that goal.” He argues that the foundation has no authority to impose change on the FDA. It can fund only research, the results of
which the FDA can choose to accept or reject. “The FDA is a profoundly change-resistant agency and will likely reject or weaken any recommendations for real change from the institute,” he contends.
The Republican Senate staffer who worked on the FDAAA says that some of the concerns raised about the foundation are “misconstrued,” but she does acknowledge that the legislation allows drug companies to make “directed” donations based on bylaws to be developed by the board of directors; some of the research would be patentable, again based on board policies.
Just as there are skeptics of the Reagan–Udall Foundation, however, some are more optimistic. Mary Richards, Deputy Chief Executive Officer of the Parkinson’s Action Network (PAN), says that the development of a biomarker for Parkinson’s disease would be of “huge, huge importance” to the Parkinson’s community. It might not only hasten development of a drug that would be effective in the later stages of the disease—a drug that does not exist today—but might also allow for better diagnosis. Although this is the type of research we expect the foundation
to fund, two questions remain: • Can the foundation’s board of directors ensure that money
will come in from multiple sources? • Will the money be applied on the basis of strategic public
health needs? Ms. Richards states: “I think there will be a lot of scrutiny around
Patients with Parkinson’s disease and many other conditions are hopeful about another provision in the FDAAA that would require drug companies to post more complete information
about a clinical trial in progress on the NIH database (www. clinicaltrials.gov). Right now, that Web site has only limited information posted by the sponsor when the trial gets under way before the drug is approved. The FDAAA substantially increases that early information; mandates new search features for users; and, perhaps most importantly—and most controversially—forces companies to post, for the first time, the results of Stage II, III, and IV clinical trials after a drug is approved. The post-approval information must also include links to FDA reviews of the clinical trials and to medical journal articles in which the trials are discussed. This must be accomplished within 90 days of the FDA’s approval. The bill allows the FDA
to levy those $250,000 and $1 million civil penalties against recalcitrant companies.
The House bill had included a provision forcing companies to submit summaries of their clinical trials in “layman language.” PhRMA fought to have this provision kicked out of the bill because
of concerns it would expose drug companies to liability for making promotional claims and because it felt clinical trial results were useful only for medical professionals. Eventually, Congress eliminated the lay language provision. Instead, House and Senate members agreed to language requiring the FDA to finish making its rules in three years. They also debated
whether to require a summary of the clinical trial and its results, written in nontechnical, understandable language for patients , “if the Secretary determines that such types of summary
can be included without being misleading or promotional.”
Even patient activists have disagreed about whether it is good to add clinical trial results to the www.clinicaltrials.gov Web site. According to Mary Richards of PAN, the detailed additional clinical trial data that will become available will allow Parkinson’s patients to make better
decisions about new therapies. Many disease communities, including those involving
Parkinson’s disease, are experiencing low participation rates in new trials. She explains:
By providing greater data transparency and access to results information, we believe that will
enhance awareness and understanding of trials as well as secure greater trust in the clinical trial process. . . .
Ultimately, adequate enrollment in trials is essential for bringing newdrugs to the market, and this provision may have a positive impact on the current participation issues. However, Steve Walker believes that tossing huge masses of data from clinical trials into the public domain will give the Steve Nissens of the world more fodder for creating fiascos. He says:
I asked Congress to go back to the drawing board on posting clinical trial data and figure out how to make it available in a manner that would effectively limit irresponsible snipers from doing statistical drive-by hits on the integrity of our drug approval system. Obviously,
they didn’t do that, and I think that was a big mistake.
The FDAAA will give the FDA more money and authority in 2008 and beyond, but it doesn’t do everything that many inside and outside of Congress had hoped for. The bill gives the CDER
statutory authority to require REMS for the first time; however, in the past, the agency has demanded that drug companies produce Risk Minimization Action Plans (RiskMAPs) for some
drugs for which questions about safety rivaled their benefits. Those RiskMAPs required the companies to provide medication guides for the drugs, for example, as REMS will do. Many in
Congress wanted the FDA to require REMS for all drugs. The authority in the FDAAA pertains only to drugs for which REMS are needed to ensure that the drug’s benefits outweigh the
risks. Neither does the bill give the FDA new authority to restrict the advertising of new drugs, a power that many outside groups and many Democrats in Congress had favored. Of course, the
passage of any significant new legislation is subject to all kinds of compromises. In the case of the FDAAA, the finished product clearly improves the FDA’s bargaining position with drug
companies on matters of safety. In the end, having additionalleverage is one thing; using it effectively is another.