Bitter Pill

After Vioxx, the FDA faces public demands for more information about dangerous medications.

"When the FDA approves a drug, it's considered a Good Housekeeping Seal of Approval," said Sen. Charles Grassley, R-Iowa, at a November hearing of the Senate Finance Committee. But Grassley said the September 2004 market withdrawal of Vioxx-the blockbuster arthritis and pain medication produced by Merck-suggests the agency may have "lost its way."

Two million people were taking Vioxx when it was yanked from the market because it increases the risk of serious heart problems. Merck's 2003 revenue from the drug totaled $2.5 billion. Medicaid alone paid more than $1 billion for Vioxx during the five years it was sold, according to Grassley.

In February, an FDA advisory committee voted 17-15 to allow Vioxx back on the market. But committee members recommended that the FDA issue strong and prominent "black box" warnings on the labels of Vioxx and similar drugs, and ban marketing to consumers. The FDA usually follows the recommendations of its advisory committees, but is not required to do so. As of late February, the drug's future was uncertain.

Like Enron, the Vioxx blowup implicated not just the company, but also the regulator. A panel of three scientists told the senators that the FDA should have taken stronger and swifter actions before and after Vioxx came on the market.

Scientists at the FDA's Center for Drug Evaluation and Research make decisions about drugs based on numeric information. Still, FDA scientists must make judgment calls. Drug reviewers examine the results of clinical trials and other data provided by the pharmaceutical companies to determine how well medications work and what side effects they cause. No drug is perfectly safe, but the reviewers decide for which patient populations and conditions, if any, the drug's benefits outweigh its risks. In the case of Vioxx, the three scientists testified, FDA failed to make that determination.

Gurkirpal Singh, a professor at the Stanford University School of Medicine, told the committee that Merck scientists were concerned about the cardiovascular side effects of Vioxx as early as 1996, three years before the FDA approved the drug. But the company never conducted a clinical trial to assess whether the risk would offset the drug's benefit-less stomach damage than other pain medications. "The trade-off of heart attacks for the rare instances of stomach bleeds is not a reasonable one," he said.

In June 2000, Merck submitted to the FDA a clinical trial that showed a significantly increased risk of heart problems with Vioxx as compared with naproxen, another common pain medication. It took until April 2002 for the agency to add a warning to the Vioxx label, but it had little effect. "No black-box warning about adverse cardiovascular events, the most prominent warning, was added to Vioxx," said Bruce Psaty, a professor of epidemiology and medicine at the University of Washington, during his testimony.

The harshest criticism came from within the agency. "The FDA has let the American people down," David Graham, associate director for science in the Office of Drug Safety at the FDA's Center for Drug Evaluation and Research, told the senators. In August, Graham completed an analysis that showed a link between Vioxx and heart disorders. He said FDA senior managers attempted to suppress his findings and refused to acknowledge the danger that Vioxx posed until the company pulled it from the market the following month.

Pictures of the gaunt scientist speaking at the hearing-his eyes wide, jaw set and hands gesturing-soon appeared in newspapers and magazines nationwide. Graham estimated that Vioxx caused 88,000 to 139,000 heart attacks and cardiac deaths, "the rough equivalent of 500 to 900 aircraft dropping from the sky," he said.

Grassley launched a congressional investigation and proposed an independent drug safety office. Last fall, acting FDA Commissioner Lester Crawford asked the Institute of Medicine to assess FDA's drug safety efforts and suggest improvements. And in February, Health and Human Services Secretary Michael Leavitt announced the creation of a drug safety oversight board to direct safety monitoring and provide more information to the public.

Symptoms and Diagnosis

Fourteen drugs, including Vioxx, have been withdrawn from the U.S. market since 1990. "These drugs got through the approval process, and we don't think they should have. What about other new drugs?" says Larry Sasich, a researcher and pharmacist with Public Citizen, a nonprofit consumer advocacy group.

The scientists in CDER's Office of Drug Safety look for serious side effects in drugs after they are approved by the Office of New Drugs, which also does some safety monitoring through clinical trials performed by the makers. The Office of Drug Safety has two main tools: epidemiological studies such as Graham's Vioxx analysis that look for clues in data from large populations, and the Adverse Event Reporting System, a database of side effects reported to the FDA. The agency knows the database does not catch everything because only a fraction of drug problems are ever reported. "It's a signaling system," says Robert Temple, CDER's director of medical policy.

The Office of Drug Safety lacks the authority to issue warnings or remove dangerous drugs from the market. Only the Office of New Drugs can do that. But because the Office of New Drugs is responsible for approvals, some say its scientists are reluctant to take regulatory actions that might indicate they made a mistake in their original review. "When they approve a drug, if safety problems occur, people take that as criticism," says Curt Furberg, an epidemiologist at Wake Forest University and a member of the FDA Advisory Committee on Drug Safety and Risk Management. Putting the scientists who approve a drug in charge of issuing warnings is "an enormous conflict of interest," Furberg says.

The Fear Factor

"There's another problem," Graham charges. The agency has an industry-friendly culture, where drugs are safe until proven dangerous "beyond a shadow of a doubt." This fall, he heard from colleagues that there were plans to transfer or fire him because he concluded that Vioxx never should have been approved and had spoken out about the agency's inner workings. Graham enlisted Grassley's protection. The FDA denies the accusation.

No other FDA scientists have been as publicly outspoken as Graham, but a survey of CDER employees conducted by the Health and Human Services inspector general in late 2002 suggests that some share his concerns. Twenty-one percent said the center does not allow differences of opinion, and 18 percent said they had been pressured to "approve or recommend approval" of a drug despite safety concerns. This is more significant than it seems, Graham says, because the majority of drug reviews are not controversial, and so only a fraction of reviewers might encounter such pressure.

The FDA chapter of the National Treasury Employees Union added a professional difference of opinion clause to its contract six years ago. Disagreements usually arise when managers override reviewers' concerns about a drug, says Robert Young, the chapter vice president who wrote the clause. "The reviewers don't factor in the economic consequences of not approving a drug or taking it off the market. The supervisors higher up have to think about what the papers will say, what Congress will say, what the boss will say, what the White House will say," Young says.

Graham says he was pressured to change his conclusions about drug risks eight to 10 times. "What I did was seek to find words that would say what I wanted and that were easier for management to live with," he says. But the first time it happened, in 1988, Graham wasn't able to find the right words, and his message got lost. "Afterwards, I felt so guilty," he says. The drug was Accutane, an acne medication that Graham named in his testimony as one of five dangerous drugs on the market. It causes birth defects.

Industry Fees

The Vioxx withdrawal revived an argument over whether industries should help fund their regulators. The 1992 Prescription Drug User Fee Act allowed the FDA to collect fees from pharmaceutical companies for new drug applications. The user fees-which will bring in close to $300 million this year-come with conditions, including collaboration between FDA scientists and drug companies during the review process and time frames for drug reviews.

The user fees brought an immediate surge of funding, and the act accomplished its goal: It got drugs approved faster. But the act also fostered what some have termed a "cozy" relationship between the regulators and the industry. Agencies for drug regulation in the United Kingdom and the European Union also collect user fees, but they are not required to meet procedural guidelines. In a 1999 article rating FDA's management capacity for the Government Performance Project, Government Executive reported: "While the FDA is happy to get funds unfettered by legislators' whims and requirements, the fees have created a complicated set of incentives that raise the possibility that the agency could be driven to focus increasingly on industry priorities."

Graham says the collaboration with drug companies (which FDA managers often call "partners" or "sponsors") makes it difficult for the FDA to take regulatory actions. "Time and again in internal meetings, you hear FDA managers saying things such as, 'We can't ask the company to do that because they won't go along with it.' " This makes FDA managers reluctant to warn the public about side effects, Graham says. Pharmaceutical CEOs don't hesitate to call on friends in Congress or in the agency's upper echelons, and when that happens, the agency better have definite proof that a drug is dangerous.

Safe or Sorry

Temple, CDER's medical policy director, says it's not always better to err on the side of warning the public when there's uncertainty about whether a drug poses hazards. "I believe that telling the world that a useful drug does something bad when it doesn't isn't doing the world a service," he says.

"This came up very much when we were thinking about antidepressants," he says. The agency had reports that a class of medications called selective serotonin reuptake inhibitors, which includes popular drugs like Prozac and Zoloft, causes suicidal thoughts and behavior when given to children. The agency wanted to verify the data before taking action. "Most of us thought the reports needed to be looked at by someone who understands pediatric [suicides]," Temple says.

The FDA issued health advisories but didn't allow a researcher to present his analysis of several widely prescribed antidepressants at an advisory committee meeting. Why? "We were highly concerned about overstating the implications of the problem before we knew," Temple says. That researcher was Andrew Mos-holder, and it turned out his analysis was right. The British agency for drug regulation warned against the use of SSRIs-except Prozac, which it deemed effective-for minors in December 2003. The FDA, using the same data, allowed children to continue taking the drugs without warning until fall 2004, when additional research confirmed Mosholder's conclusions and the FDA issued a black-box warning.

In Temple's view, the FDA was acting in the public interest by making sure to verify dangers before sounding alarms. FDA managers say they're willing to adjust this approach. "[FDA] culture, which has received some criticism in recent months, is not to alarm the public when we get a signal," Crawford told The New York Times in February. "That era is sort of past. What the public, we think, is demanding is to know as soon as we know what's going on."

But in the case of drugs that might cause children to think about killing themselves, it seems obvious that parents would want to know about the risks as early as possible. Why would the FDA believe otherwise?

Mixed Messages

The answer lies with another group-the public, which has been telling the FDA for years to speed new drug approvals and let consumers decide for themselves whether a drug is too dangerous.

On Oct. 11, 1988, more than 1,000 people traveled to the FDA's headquarters in Rock-ville, Md., to protest the slow delivery of AIDS treatments. The angry crowds shouted and carried signs saying "Federal Death Agency" and "Red Tape Kills." They claimed the agency's slow and overly cautious procedures were killing AIDS patients.

The FDA got the message. "They'd never had people chaining themselves to the doors of their headquarters before," says Arthur Levin, director of the Center for Medical Consumers, an advocacy organization based in New York. "I think that scarred them a lot." Other patient groups learned from the effectiveness of the AIDS protesters nearly 20 years ago. "The mind-set has become, unless we have really clear and convincing evidence that there is a problem with a drug, people need it and want it and we want to give it to them," Levin says. "FDA sees the agency's role as getting out of the way unless it can't get out of the way." Unlike AIDS treatments, though, most new drugs are not breakthrough therapies, just additional choices in existing classes of drugs.

Pharmaceutical companies used to complain about the so-called "drug lag," resulting from faster drug approvals overseas than in the United States. "Foreigners got newer and better drugs first," said Randall Lutter, chief economist in the FDA's Office of Planning, at a February forum on drug approvals at the American Enterprise Institute, a think tank in Washington. Now, roughly half of all new drugs are approved here first. And still critics press for more speed, arguing that the FDA impedes medical innovation. "Bigger and longer [clinical] trials mean more expensive drugs and fewer drugs," says Alex Tabarrok, an economist and fellow at George Mason University's Mercatus Institute, a market-based research center.

Walking the fine line between protection and innovation has been a perennial challenge for the FDA. In a 1985 inter-view published on its Web site, former FDA Commissioner Alexander Schmidt expressed the dilemma. "There's an old saying in the FDA that if you're being criticized just about equally from opposite sides of a question, you're probably about where you ought to be," he said. Even in the case of Vioxx, which has been called the FDA's worst drug-safety failure, some patients wish the drug hadn't been withdrawn. "I would go back on it in a heartbeat," one woman told USA Today in January. There probably are some patients for whom the benefits of Vioxx still outweigh the risks.

As NTEU's Young points out, determining how restrictive FDA should be is a political decision. "The electorate should decide where they want to draw the line," he says. Framed in those terms, the uproar over Vioxx and the SSRIs raises a larger question about the FDA, and government in general: Is there a way to stop making trade-offs between the seemingly contradictory demands of safety and access, and do a better job with both?

The advent of electronic medical records likely will make tracking drug safety far easier. The real key, though, will be more effective communication-clearer and more frequent-so consumers can understand the drugs they take and decide for themselves whether the benefits of Vioxx or any other drug warrant the risks. "The public has spoken, and they want more oversight and openness," HHS Secretary Leavitt says.

This is the goal of the new safety board, which will post information on drug risks to the Drug Watch Web site even before the agency takes regulatory action. The FDA also plans to issue one-page information sheets with safety facts for the public and health care professionals. Critics of the safety board say it won't be independent because it will consist of government scientists. They contend that the agency's 2006 budget lacks funds for significant reform. The board will not have the authority to remove drugs from the market or add warnings to drug labels. But Amit Sachdev, the FDA's deputy commissioner for policy, said at the AEI forum that the board is "the first step in a multistep process."