The system that the Food and Drug Administration uses to ensure drug safety is badly in need of overhaul, according to a report by current and former members of the FDA Drug Safety and Risk Management Advisory Committee. The report, published in the October 9th issue of Archives of Internal Medicine (2006;166:1938–1942), notes that in 1994 there were more than 100,000 fatal drug reactions among 33 million hospital admissions, and that “there has been no meaningful change in our regulatory approach to drug safety, and that when recommendations for improvement have been made, they tended to be only small changes in the status quo rather than a fundamental effort to fix the problems.”
In addition, the report said, the new Drug Safety Oversight Board is unlikely to succeed, because of its lack of authority, resources, and independence from FDA. The solutions must come from Congress which controls FDA's legal authority as well as its funding, said the authors, led by Curt D. Furberg, MD, PhD, Professor of Public Health Sciences at Wake Forest University School of Medicine.
Dr. Furberg said in an interview that even obtaining appropriate authority will not be enough. “You need a dynamic leader who understands the problems.” Does FDA have such a leader now? Dr. Furberg's response was quick and unequivocal: “No.”
Even if FDA could make major changes, does it have the will to do so? “No,” he said again.
Richard L. Schilsky, MD, Professor of Medicine and Associate Dean for Clinical Research at the University of Chicago and Chairman of Cancer and Leukemia Group B, asked for his opinion, said there is nothing new in the report. “Some things I agree with and some I don't,” he said. “The big problem is that most of the suggestions don't apply to oncology—and may be a waste of time.”
8 Major Problems
Dr. Furberg and his coauthors (Arthur A. Levin, MPH; Peter A. Gross, MD; Robyn S. Shapiro, JD; and Brian L. Strom, MD, MPH) cite eight major problems with the way FDA assures drug safety.
▪First, initial reviews for approval often fail to detect serious adverse drug reactions (ADRs). For example, a nine-year study conducted by the General Accounting Office found that 51% of all approved drugs had at least one serious but unrecognized ADR during the approval process.
The authors said that this is the result of short-term trials with too few low-risk patients. FDA does not require long-term trials with patients likely to use the drug.
This is Dr. Schilsky's major bone of contention. “In cancer drugs, the most difficult parameter to evaluate is efficacy, not safety. In fact, when you're dealing with a serious life-threatening disease, almost any degree of safety is considered acceptable. The risk-benefit calculus always leans toward efficacy—the opposite of non-cancer drugs.”
For example, he said, Novartis recently circulated a “Dear Doctor” letter about a small risk of heart failure with Gleevec. “Gleevec is used for serious leukemias and GIST, both potentially fatal. No patient is going to forego the possibility of remission because of a small risk of heart failure. The risk of cancer death is far from small,” Dr. Schilsky said.
- Second, the agency's primary source for identifying safety issues after approval is the Adverse Event Reporting System, which receives about 400,000 reports a year, primarily from pharmaceutical companies. Very few come from physicians and patients, which results in “massive underreporting,” said the report. Only about 10% of all ADRs and 1% of serious ones are reported.
- Third, the report said, setting a threshold for action in response to ADR reports is subjective and slow and when combined with underreporting, far too many patients are exposed to harm over a long period of time.
- Fourth, FDA has almost no control over postmarket safety studies. The majority of such studies are never initiated by pharmaceutical companies and are not pursued by FDA. As of September 2005, 1,231 commitments to conduct postmarket studies were unfulfilled. FDA has no authority to take legal action against violators nor a policy to grant conditional approval to drugs with an incomplete safety record.
Dr. Schilsky commented that FDA does indeed need more legal authority, but postmarket studies are not the way to go. “As soon as you use the word ‘study’ or ‘trial,’ people back off. They don't want to get involved with eligibility criteria, recruitment, and all the other requisites of a full-fledged clinical trial. It would be much better to establish a postmarket registry of toxicity that reflects real-life practice.”
But will physicians report adverse events? Dr. Schilsky acknowledged the logistical problems (taking the time to do it or hiring someone, coping with the added paperwork, etc.), but he said that a registry is much more workable than a postmarket trial.
- Fifth, FDA cannot take legal action against companies that suppress or delay reporting unfavorable trial results. Only the Department of Justice can do this, but FDA rarely asks it to do so.
- Sixth, the structure of FDA itself creates potential conflicts of interest because its Center for Drug Evaluation and Research, which reviews and approves drugs, also is responsible for taking regulatory action against the drugs it approves in the first place.
- Seventh, the agency has become too financially intertwined with the industry it regulates, specifically its dependence on user fees—the system that requires drug companies to partially pay the costs of evaluating the drugs they present for approval.
Moreover, the Prescription Drug User Fee Act (1992) requires that FDA perform something of value to the pharmaceutical industry in exchange for the fees.
Dr. Schilsky said he believes that user fees are a bad idea: “FDA should get rid of them, and Congress should appropriate the funds necessary to evaluate new drugs.”
- Eighth, the report said, FDA is short on expertise in drug safety, public health, and ethics—in the agency and on its advisory committees. Only the Drug Safety and Risk Management Advisory Committee has this talent.
But this is not an issue for oncology drugs, Dr. Schilsky said. “In all the years I've been treating cancer patients, I've never heard of a cancer drug being taken off the market because of adverse events. If a drug works, we have always found ways to mitigate the toxicity.”
OT asked Dr. Furberg: Since FDA has so many small “cosmetic” reorganizations, does he think that the agency really cares about drug safety, or is it merely a knee-jerk reaction to public loss of confidence.
“FDA doesn't have the ability to protect the public against harmful drugs,” Dr. Furberg answered. “They simply can't do it on their own. The current leadership is ineffective.”
Recommendations for Improvement
Dr. Furberg and his colleagues proposed a series of recommendations that would fundamentally change the way FDA operates. The recommendations fall into three categories: authority, independence, and operations.
FDA needs more direct legal authority than it has, which can come only from Congress. The agency should play a more active role in design and approval of clinical trials, especially for chronic conditions. Moreover, the report said, “a higher priority must be given to drug safety, including adequately powered safety trials to be completed prior to approval. Safety evaluation should be given the same priority as efficacy evaluation.”
The group also said that FDA should be given the unilateral right to suspend marketing or mandate immediate withdrawal of harmful drugs.
A less drastic but equally important recommendation is expanded authority to change drug labels as a result of safety concerns. And the agency must have the authority to hold drug manufacturers to their commitments for postmarket clinical trials—with clear consequences for failure to comply.
Regarding independence, the authors recommend establishment of a Center for Drug Safety within FDA, separating pre- and postmarket review and evaluation.
Special Circumstances for Cancer Drugs
“But not for oncology drugs,” Dr. Schilsky said. “For us, an independent safety review is not a good idea— unless everyone on the panel is an oncologist and understands that with cancer, the balance between safety and efficacy is not the same as with other drugs. Ours should always be tipped toward efficacy.”
Startling Recommendations about FDA Operations
The authors had the most startling recommendations about FDA operations:
- There should be a new category of conditional approval for drugs that appear to have clear benefits but where there are also unanswered questions about adverse events. This would put pressure on sponsors to conduct and report safety studies.
“Conditional approval is an easier solution than going through all the legal steps that it takes to pull a drug off the market,” Dr. Furberg said. “This would put the industry on notice to pay attention to safety.” Dr. Schilsky said he sees nothing wrong with conditional approval: “It holds no threat to oncology drugs.”
- The Center for Drug Safety should take advantage of large public and private databases—for instance, from the Department of Veterans Affairs, Department of Defense, Medicare and Medicaid, and large health plans such as Kaiser Permanente and Group Health Cooperative.
- Dr. Schilsky liked this idea too. “FDA needs to work in partnership with these large databases,” he said. “They are there to be used.”
- A permanent network of collaborating drug safety centers ought to be established across the United States, building on the existing Centers for Education and Research in Therapeutics sponsored by the Agency for Healthcare Research and Quality. Funding for this should come from Congress rather than from the pharmaceutical industry.
- Other countries use large population-based, disease-specific adverse event registries to identify suspect drugs. The registries work well and should be used here.
- The composition of all FDA advisory committees should be changed to eliminate members who receive significant support for clinical trials from pharmaceutical companies. Experts who specialize in the disease under consideration should make up no more than half the committee. The rest should be proficient in risk management, drug safety, public health, epidemiology, and ethics. This would create a more balanced and objective committee, the report said.
Despite the almost entirely negative tone of the report, Dr. Furberg was somewhat optimistic about the future: “Momentum is building, and there's a groundswell of popular support for big changes at FDA. And if Congress gets involved in the effort on a bipartisan basis, I think we have a chance.”
Congressional Bills Concerned with Drug Safety
▪S 930, the Food and Drug Administration Safety Act of 2005, was introduced by Sen. Chuck Grassley (R-IA). It requires FDA to conduct postmarket risk assessment, determine whether a drug presents an unreasonable risk to the public, and take corrective action if a risk exists. The law allows for significant penalties for violations, as well as withdrawal or suspension of a drug. The bill has been referred to the Senate Committee on Health, Education, Labor, and Pensions (HELP).
S 470, the Fair Access to Clinical Trials Act of 2005, was introduced by Sen. Christopher Dodd (D-CT). It requires that NIH establish a database of clinical trials that test drug safety and efficacy in serious or life-threatening diseases. The information gathered must be made public. The bill also has been referred to HELP.
HR 870, the Pharmaceutical Research and Manufacturers Accountability Act of 2005, was introduced by Rep. Pete Stark (D-CA). It provides for penalties if a CEO or other executive of a pharmaceutical company knowingly conceals evidence of a serious adverse drug reaction. The bill has been referred to the House Subcommittee on Health.