The Side Effects of Side Effects
The pharmaceutical industry thrives on side effects. At first glance, that might seem like a flat contradiction of the facts. Think Vioxx—was that withdrawal good for Merck? Of course not. But it was good for manufacturers with safer NSAIDS. In fact, the development of products with better safety profiles has, for decades, been an engine for driving industry growth.
Yet side effects persist. Just last September, a government study concluded that between 2004 and 2005 more than 700,000 Americans showed up at emergency rooms with harmful drug reactions. Conspiracy theorists blame the pharma industry for marketing dangerous drugs. The more accurate—and far more intractable—reasons include the misuse of drugs, off-label prescriptions for indications they were never tested for and the inherent characteristics of pre-market clinical studies. Not only is no study group ever large enough to be projectable to eventual mass use, but test groups, unlike post-approval patient populations, tend to be homogenous, carefully selected and monitored.
That's a truth our risk-averse society seems unwilling to accept: that there is no such thing as a 100% safe drug. Reducing—not avoiding — complications calls for more effective post-marketing surveillance, a goal much talked about but never achieved. Now the problem has been tackled yet again by the Institute of Medicine (IOM) of the National Academy of Sciences.
The Institute of Medicine recommendations
Widely publicized by the scientific and national media, the IOM report pointed its institutional finger at a variety of shortcomings, some internal to the FDA, some embedded in the underlying food and drug law, others the fault of Congressional cheeseparing, still others the fault of pharma companies. The basic thrust of the report is that we need a safety system that focuses on “a lifecycle approach to drug risk and benefit” in place of the current “striking disparities” between thorough pre-launch new drug application (NDA) review and relative neglect of monitoring marketed drugs. Here are some highlights of IOM's conclusions:
There are often “fierce disagreements” between the Office of New Drugs and the Office of Drug Safety (renamed Office of Surveillance and Epidemiology), and an external Management Advisory Board should be appointed to “transform the culture” of the Center for Drug Evaluation & Research (CDER) to which they both report.
CDER's emphasis on speedy NDA approvals and user fee deadlines have led to “subservience of the safety function.” And while this imbalance can't be fixed by “repositioning organizational ‘boxes' and changing unit names,” it would help if an ODS staff member were added to each NDA review team.
The upcoming PDUFA reauthorization deadline provides a “golden moment of opportunity” to permit spending of user fees for post-marketing surveillance, though basically it's up to Congress to fund FDA adequately. FDA should be given the power to make companies honor commitments made at the time of NDA approval to conduct follow-up post-launch studies.
Unfortunately, most media summaries highlighted the report's recital of shortcomings of FDA and industry, while neglecting some key observations that put these criticisms in perspective. Thus the report says: “Some observers believe that drug withdrawals…represent de facto failures of the drug safety regulatory system, or that newly identified unusual and serious adverse events indicate that someone made a mistake in approving the drug. That is not so.”
To find out what experts think is so, MM&M canvassed the opinions of those in the best position to judge, starting with Sheila Burke, chair of the Committee on the Assessment of the Drug Safety System at IOM and one of the three editors of the institute's report. She and Scott Lassman, senior
assistant general counsel of PhRMA, were jointly interviewed on National Public Radio.
The FDA watchers respond
Burke summed up the intent of the report as encouraging the continuous gathering and assessing of all relevant information about drugs both before and after NDA approval, seeking to achieve a better understanding of how to gain optimum benefits without incurring needless risks. As expected, Burke spoke favorably of the IOM recommendations; less predictably Lassman generally agreed with her. Yes, they agreed, FDA needs more money from Congress, and yes, in Lassman's words, IOM “did a very comprehensive job of looking at a very complex issue,” and industry “agrees with a lot of what they said.”
What's more instructive is to focus on the points on which they disagreed. Lassman sees a drug safety system which, while it can be improved, is “very robust,” while Burke points out that it was questions about the adequacy of safety monitoring that led FDA, the Center for Medicare and Medicaid Services, and others to ask IOM to conduct this study in the first place. If there was no problem, there'd be no study.
On specific issues, Burke supports and Lassman opposes using PDUFA fees more broadly to help correct the organizational emphasis on drug approval as against post-marketing surveillance. Similarly, they part company regarding the IOM recommendation for a moratorium on DTC promotion of new products. Such a delay might not be needed for every new drug, Burke believes, but FDA should have the power to require it. Counters Lassman: “We [PhRMA] called for companies to impose a voluntary moratorium until doctors could be educated,” but oppose government-imposed moratoriums. DTC, he went on, gets patients with undertreated conditions to see a doctor.
Another difference emerged when Burke and Lassman discussed the recommendation that labels of new drugs bear a warning symbol. Not only should all labeling be more transparent and easily understandable—right now they are “at times incomprehensible” said Burke—but unlike PhRMA she likes the UK system where the label clearly indicates that this is a new product that “should be given particular attention.”
No scholar has studied FDA more intensively than Daniel Carpenter, professor of government and director of the Center for American Political Studies at Harvard. As you read his publications and especially as you speak with him, you realize that his recommendations must have had some influence on the IOM conclusions they track so closely. Also he is acknowledged in the report as a reviewer and his published studies are part of the documentation.
System? What system?
So as might be expected, Carpenter is generally positive about the IOM report. He singles out two recommendations as being particularly crucial. One is “ending or at least ameliorating the stand-off between the Office of New Drugs and the Office of Drug Safety.”
The second is “giving the FDA the power to enforce the commitments companies make at the time of NDA approval to conduct Phase IV follow-up studies.”
As things stand now, Carpenter says, even though we call them commitments, the system is really voluntary. He recalls that there was a tacit understanding at the time review deadlines based on user fees were put in place that FDA was going to speed up the drug approval process; in return, drugs would be vigorously studied once they were on the market. But the pharmacovigilance part of the bargain has not been observed: as of September 2004, close to 1,200 Phase IV study commitments had been made but 68% had not even been started. “If FDA were allowed to fine drug companies that don't comply, that would fix this problem,” he says.
Carpenter, as we shall see later, is also in favor of lifting the restrictions on how FDA can spend user fees, and would like to see more flexibility regarding approval deadlines. (See sidebar page 50 for the latest about PDUFA renewal.) “I'm not down on deadlines per se,” he explains, but his research suggests that “some of the drugs that are approved right before the deadlines are somewhat more likely to encounter post-marketing safety problems.” Therefore he would give the reviewers the ability to take, not another year, but “a couple of months if necessary,” so that the deadlines don't become the “end all and be all” consideration.
One IOM recommendation that gives Carpenter pause is the suggestion that a management board—and if that doesn't work, a separate drug safety agency—be set up. He sees that as a potential contradiction to the IOM's emphasis on lifecycle drug monitoring. After all, couldn't the FDA then simply wash its hands of post-market surveillance and say, “That's up to somebody else?”
Concerns about post-marketing surveillance are hardly new. Way back in 1963, a Congressional hearing was warned that FDA was reluctant to revisit its approval decisions. Partly in response to such criticism, the agency in 1968 started to require the reporting of serious or unexpected adverse drug reactions. Manufacturers have, by and large, complied with this requirement, but physicians and hospitals are often remiss, so no one can tell with certainty just how common adverse events actually are.
It seems, therefore, that the post-marketing surveillance system is at best broken. A more pessimistic conclusion would be that there really is no system. There are some sporadic and uncoordinated attempts to improve drug safety, and once in a while a Vioxx rings the alarm bells. Asked whether it was really the Vioxx withdrawal that triggered the current concern, Carpenter says, “If it hadn't been for Vioxx, would we be having this conversation?”
“The FDA should have authority to punish companies that make promises to undertake post-marketing surveillance studies and they don't.” Those are the words of Peter Pitts, FDA associate commissioner from 2002 to 2004, so it's obvious that he agrees with some IOM recommendations—but not all.
An alumnus has some doubts
Pitts is senior vice president for global health affairs at Manning, Selvage & Lee and head of the Center for Medicine in the Public Interest. On the blog DrugWonks, one of his recurring pithy sayings is show me the money—his response to Congressional demands that FDA do a better job, or as he said in an interview: “The fact of the matter is that things can always be better and FDA can always do more, but you have to fund it.”
Pitts is against a DTC moratorium for new drugs (“I don't see how restricting information makes drugs safer.”) and that they carry a warning label (“All that does is scare people away from compliance.”)
But perhaps his most serious concern is that the IOM would further tighten the already stringent strictures on conflicts of interest. “Having been the former senior official in charge of advisory committees,” he says, “I can tell you from firsthand experience that the FDA takes conflict of interest issues very seriously.”
Barring “the best and the brightest” however—the people who know the most about the agents being studied—“actually goes in the wrong direction when it comes to drug safety.” He also takes issue with the IOM claim that “…there is an organizational culture in CDER” that makes the Office of Drug Safety subservient to NDA reviewers. “There is no pecking order,” says Pitts. “These people work together, they know each other…they're family.” Besides, safety and effectiveness must be judged in relation to each other. That's also why he thinks an independent drug safety board is a bad idea. “You can't view a drug through a one-dimensional view of safety.”
Jerry Avorn, MD, on the other hand, says “there are no recommendations that I vociferously disagree with.” Though the proposals for changing the organizational structure of safety activities are “something about which reasonable people can disagree,” he does think it's a debate worth having. Overall, “if we could just do what the report proposes, that would be a very important step forward.” Best known for his book Powerful Medicines, Avorn is a professor of medicine at Harvard, as well as chief of the Division of Pharmacoepidemiology and Pharmacoeconomics at Brigham and Women's Hospital.
His book is openly critical of FDA. For instance he says: “The (FDA) doesn't even see its role as monitoring how drugs are used; rather it's sticking closely to its authority to simply say yes or no to the initial entry of a given product onto the market. What doctors and patients do with it after that decision, FDA believed, was none of its business.”
It's no surprise, therefore, that Avorn comes across as more skeptical than Carpenter as to whether FDA is capable of reform. While not, he says, opposed to giving FDA the mandate and the resources to “try and finally get it right,” he adds that “FDA has had years to demonstrate that they can't do this,” so if they fail again he is willing to consider a separate safety agency. Yes, such a split would make him nervous, but he feels sympathy for those who say that “rather than take (a chance on) an organization that has demonstrated an appalling inability to get on top of drug safety,” it might be well to try something new. But he thinks much of the blame rests with Congress. “There are many reasons FDA has done this so badly for so many years, but a very major reason is simply that FDA does not have the money or the adequate person power to do good drug safety research.”
Unlike Pitts, he believes that user fees that are focused on NDA review are part of the problem. That doesn't mean, he continues, that there isn't also a need to pay better attention to safety concerns in the pre-approval process, citing Vioxx as a case where problems should and could have been spotted based on NDA data.
Avorn strongly endorses the IOM conclusion that the expertise of the drug safety staff is undervalued, and wonders whether having an ODS rep at the table at NDA review meetings is meaningful. What matters, he says, is “the relative influence and respect” that safety officials have within CDER. “It really doesn't matter whether someone with a little sign in front of them saying ‘Office of Drug Safety' sits at the table” as long as they're seen within the agency as “skunks at the garden party.”
Underlying Avorn's skepticism about FDA's ability to reform is his perception that the agency's mission has changed in the last 10 years. Today, “the mission of FDA is to perform speedy and efficient approval of drugs, and people who come along and talk about side effects or risks are seen as ‘not being with the program.'”
An even more basic problem, as many have noted, is that FDA has limited powers when post-marketing safety problems come to its attention. Sure, it can revoke the NDA, but such drastic action is only rarely called for. Bereft of legal authority to control how drugs are used, the agency has relied heavily on labeling revisions as a form of regulatory policy. That at least puts it on record that it has concerns about possible risks. Whether anyone pays attention is another question.
The gored ox responds
Let's let FDA have the last word. After all, they started this discussion by asking the Institute of Medicine to conduct its study, and turned out to be the prime target.
Andrew von Eschenbach, just before he was confirmed as commissioner, said in an official statement: “As the IOM report recognizes, much progress and reform of FDA's safety oversight enterprise is already under way.” But then he went on: “I am committed to taking additional steps and will look to the initiatives recommended by the Institute of Medicine to ensure we continue to fulfill our mission…Above all, we recognize a need to mindfully acknowledge any weaknesses in our organization's process or infrastructure or culture, and then aggressively design and adopt programs to strengthen the FDA enterprise.” Subsequently he told a Senate committee that he believes that “post-marketing surveillance is as important to FDA's mission as the approval of new drugs.” He also balks at the suggestion to split the drug approval and safety surveillance functions, saying that it would create a loss of continuity in the review of risks and benefits.
Steven Galson, director of CDER, conceded that the “IOM report offers a significant opportunity for CDER to re-examine how we address drug safety. The spirit of the report resonates with many of us, even if some have questions regarding specific recommendations.”
Scott Gottlieb, then FDA deputy commissioner for medical and scientific affairs, was less diplomatic—possibly because he realizes that some observers consider him part of the problem. In a speech at the Manhattan Institute he said: “I don't necessarily think improved drug safety is simply a matter of extending new legal authorities to FDA that serve only to add additional burdens to the development process.” Some of the ideas under discussion “will do little to make our drugs safe, but will do a whole lot to limit access to needed medicines…”
SIDEBAR: The padoofa kerfuffle
Remember the drug lag? That was the pharma industry's rallying cry when it was berating FDA in the 1980s for being too slow in approving new drugs. In response, Congress passed the Prescription Drug User Fee Act (PDUFA) in 1992 that made companies filing new drug applications pay for the review, while setting deadlines for FDA to arrive at a decision. Two stipulations in the original legislation are now making waves: the sunset provision that requires re-enactment in '07, and the provision in the original legislation (loosened in the 1997 renewal) requiring the agency to use fees for the initial NDA review.
Since then, there have been no complaints about a lag; criticism has focused on other issues: FDA has rushed approvals without due consideration of potential risk, and that the disproportionate funding of the Bureau of New Drugs has changed the balance of power within the agency to the detriment of post-marketing surveillance. Thus Avorn in Powerful Medicines says that “some worry that if industry helps to pay (the) evaluators it will also call their tune,” and he quotes an FDA official as telling a subordinate, “The industry is our client.”
Here is what the IOM says: “To restore appropriate balance between the FDA's dual goals of speeding access to innovative drugs and ensuring drug safety over the product's lifecycle, the committee recommends that Congress should introduce specific safety-related performance goals” when it reviews PDUFA . Carpenter, who was probably the first to call for user fees to pay for safety surveillance, agrees. Asked whether he favors PDUFA renewal, he says: “Not without reforms,” because FDA needs the flexibility to be able to use user fees for other purposes, including post-marketing surveillance.
Avorn says he is against the renewal. If the approval of the Cox-2 products had not been rushed, he feels, the Vioxx debacle might have been avoided. He says, PDUFA has drained resources from the Office of Drug Safety. “Pulling people over to the review process to meet deadlines not only did not help but also under-cut drug safety efforts,” he says, that “the expertise, orientation, and data that are required for drug safety surveillance are different from the expertise and data that are required for drug approval. In the case of approvals, you're dealing with clinical trial data and people who understand the review of clinical trials. That's different from looking at pharmacoepidemiological data.”
Hence Avorn concludes that when the PDUFA legislation comes up for re-enactment, “the very best thing would be to end it. If for whatever reason that can't happen, there should be a very strong requirement written into the renewal that the user fee dollars be used for safety surveillance, both before and particularly after marketing.”
Pitts occupies the middle ground. While in favor of PDUFA renewal, when it comes to safety funding he adds that “if you want a more robust post-marketing program, and I certainly agree with that, the money has to come either from taxes or from user fees.”
Daniel Troy, an FDA alumnus who served as chief counsel from 2001 to 2004, believes that letting user fees fund programs other than NDA reviews creates a “greater possibility of a train wreck.” He has suggested that the sunset provision in the PDUFA legislation be removed, pointing out that “having FDA and industry go through this every five years…is very, very stressful for all involved.”
As this article went to press, FDA was preparing to announce new PDUFA proposals.
SIDEBAR #2: What now?
When asked to predict whether the IOM recommendations will be acted on and, if so, whether they will solve the problems, we heard the following responses:
Carpenter: “I think the industry very much wants (PDUFA) renewal,” so if a few senators like Grassly, Enzi, and Kennedy demand concessions from the industry in order to get renewal, “there's a good chance the change [to add funding for post-marketing surveillance] will be made.” Will such changes help to restore FDA's reputation? “I don't think that the public's and medical profession's trust in the FDA has been lost entirely, but it has been damaged. Once there is greater transparency, the agency's reputation will probably benefit, but it's going to take some time.”
Avorn: “It's now in the realm of politics, and that is a little scary. There seems to be more concern about whether a particular Congressman sent some racy e-mails than whether thousands of Americans are dying because of preventable drug side effects.” Because a lot of what is required to implement the IOM suggestions require legislative changes on the part of Congress or policy changes on the part of the Administration, “that's where everything becomes murky…The most critical issue is whether this report, which is thoughtful and generally quite reasonable…just sits on the shelf, or whether there's going to be the political will in the nation to have Congress and the executive branch actually implement it.”
IOM itself, foreseeing that the key lies not in making recommendations but in having them implemented, used this Goethe quote as the epigraph for its report: “Knowing is not enough; we must apply. Willing is not enough; we must do.”
As the new Congress goes to work, let's see how willing they are to apply and to do. With a Republican veto-wielding president, a one-vote near-tie in the Senate, and a Democratic House, here is a bold prediction: There will be either a new spirit of bipartisan cooperation…or deadlock. Take your pick.