On Nov. 18, an unassuming safety researcher from the U.S. Food and Drug Administration sat down before a Senate committee and tore his bosses to shreds.
The FDA, he said, had ignored warnings that the pain pill Vioxx was killing people by causing heart attacks and strokes -- and he said the agency was incapable of defending the public against another drug disaster. "I could have given a very mealy-mouthed statement," says David Graham, the scientist. "But then I would have been part of the problem."
For his steadfast advocacy of drug safety and his willingness to blow the whistle on his bosses, we're naming David Graham our Face of the Year.
Without Graham, the Vioxx debacle might have been seen as an isolated event. But because he was willing to step into the spotlight, the withdrawal of Vioxx from the market looks like part of a systemic failure to properly weigh the risks and benefits of drugs. To hear Graham tell it, this is part of a systemic failure to address drug safety on the part of the FDA, a story that reaches back over the entirety of his 20-year career at the agency. That could kick-start a broad debate over what risks we're willing to take every time we swallow pills. In the long run, change would be good for regulators and drug companies.
Graham is an unlikely whistleblower. For years, he has toiled at the FDA's offices, presenting data to advisory committees but not putting his face before the public. He would seem out of place in the often secular world of academic science. A father of six, Graham is a deeply religious Catholic and taught himself biblical Greek so he would be able to read the New Testament in its original language.
During his medical residency, Graham trained as a neurologist. But though he loved solving puzzles, he didn't enjoy day-to-day patient care. After his residency, he began studying epidemiology--the science of conducting experiments on large groups of people. That led him to public health, and the FDA. "Instead of an individual patient at the bedside," he says, "I've got 290 million patients."
Graham's research has been unique in its ability to overcome problems in side effect detection. Drug risks are notoriously hard to track. It's easy to find isolated cases, but hard to know how often a problem is occurring.
Graham has overcome this problem by using only large samples of data collected by managed-care organizations. When big, controlled clinical trials designed to compare two medicines emerge, the results often resemble what Graham managed with a supposedly less-sensitive study. "You don't have a better standard than that," says Bruce Psaty, a safety researcher at the University of Washington.
This technique has allowed Graham to raise the alarm early and often about potentially dangerous drugs. He has rarely cried wolf--most of the medicines he flagged were eventually pulled from the market. Graham can be credited to some degree with the withdrawals of Abbott Laboratories' Omniflox, Wyeth's Fen-Phen and Redux, Warner-Lambert's Rezulin, the over-the-counter drug PPA and, of course, Merck's Vioxx).
In his U.S. Senate testimony on Vioxx, Graham expressed reservations about Roche's drug Accutane, Sanofi-Aventis' Arava, Pfizer's Bextra, AstraZeneca's Crestor and Abbott Labs' Meridia.
But Graham says that for his entire career, he's been getting heat from his bosses. "You don't get rewards for doing the work that gets a drug taken off the market," he says.
At a 1999 advisory panel for Rezulin, a Pfizer diabetes drug that was eventually pulled from the market for causing liver problems, Graham's slide projector broke. His data helped identify a dangerous medicine, but he only received a letter of reprimand for not having backup slides on acetate. "Nobody ever said to me, 'Oh, good job,' " Graham says.
The FDA does not comment on personnel matters. While the agency has said it believes the safeguards on drugs are more than adequate, it has commissioned a report from the Institute of Medicine to analyze the issue in more detail.
But the response to Vioxx, which was taken by 4 million Americans at the time of the recall and which Graham thinks caused 140,000 heart attacks, has been more of the same. Vioxx's maker, Merck, conducted a big trial to prove the drug caused fewer ulcers than an older medicine did. There were more heart attacks and strokes in people taking Vioxx than among those taking the other drug, but the trial's safety monitoring board let the study continue because the potential heart risk had to be balanced against the ulcer benefit.
But the labeling changes for the drug minimized the cardiovascular risk while stating the ulcer benefit--exactly what one would expect if the FDA were going easy on Merck. Alastair Wood, associate dean at Vanderbilt Medical School, says Merck should have been forced to conduct a trial in higher-risk patients sooner. Instead, during Senate testimony, the FDA's Sandra Kweder, deputy director of the office of new drugs, said that the deaths Graham pointed to were just the predictions of a mathematical model.
"What's a theoretical death?" asks Wood. "I guess it's someone else's."
Kweder said during her testimony, "Detecting, assessing, managing and communicating the risks and benefits of prescription and over-the-counter drugs is a highly complex and demanding task. FDA is determined to meet this challenge by employing cutting-edge science, transparent policy and sound decisions based on the advice of the best experts in and out of the agency."
With all this against him, why doesn't Graham just bolt to a comfy job at a university? For one thing, he says, he wouldn't be able to get some of his data. For another, he would have to apply for funding with the National Institutes of Health or, worse, get it from a drug company. But he does not see change coming fast at the FDA. "It's an agency in denial," he says.