Blowing the Whistle on the FDA

an interview with David Graham

Dr. David Graham is the associate director for science and medicine in the U.S. Food and Drug Administration’s Office of Drug Safety, and the FDA medical officer who blew the whistle on Vioxx. Graham graduated from the Johns Hopkins University School of Medicine, and trained in Internal Medicine at Yale and in adult Neurology at the University of Pennsylvania. He also has a Masters in Public Health from Johns Hopkins, with a concentration in epidemiology and biostatistics. Over a 20-year career at FDA, he has contributed to the withdrawal of a number of drugs from the U.S. market, including Abbott Laboratories' Omniflox, an antibiotic that caused hemolytic anemia; Warner-Lambert's Rezulin, a diabetes drug that caused acute liver failure; Wyeth's Fen-Phen and Redux, weight loss drugs that caused heart valve injury; and PPA (phenylpropanolamine), an over-the-counter decongestant and weight loss product that caused hemorrhagic stroke in young women.

Multinational Monitor: Your research found that Vioxx causes heart attacks. How did the Food and Drug Administration (FDA) respond to your findings?

David Graham: When I began at FDA, I had no idea of the costs that would be demanded of my family or of me personally. Had I known the cost, and the extreme difficulty of working in an environment that routinely dismisses or twists the truth about drug safety, and punishes you severely for speaking the truth, I’m certain that I would have chosen a different path.

I did a study with several other people. It was a very good study. It found that Vioxx caused heart attacks.

I wanted to get that study published in the peer review literature, and FDA reacted violently.

FDA would not have pulled Vioxx, rest assured. FDA saw no problem with 100,000 people having heart attacks.

The week before Vioxx came off the market, I was in a meeting with very senior people from the Office of New Drugs, and my own Office of Drug Safety. They said to me, “Why on earth did you study Vioxx and heart attacks? We have no regulatory problem with this drug.” And, “Why are you studying Vioxx and heart attacks, we in the Office of New Drugs didn’t approve that study, we don’t want you studying that.” This from my own supervisors: “You’re work is scientific rumor.” My Center director called the study “junk science.”

The day after he made that remark, I had the lead article in the Journal of the American Medical Association (JAMA), and it was accompanied by a lead editorial, with the editors of JAMA calling for a complete restructuring of FDA. I found that rather amusing — so much for junk science.

In any event, a week before Merck pulled the drug from the market, FDA had no problem with the drug. They were perfectly willing to have it remain on the market.

MM: Should FDA have known about the problems earlier?

Graham: I think if one looks at the evidence, it is clear that prior to the marketing of Vioxx and Celebrex, there were strong theoretical reasons to expect that these drugs might increase the risk of heart attack. FDA was fully aware of these theoretical concerns. FDA also knew that these drugs, called Non-Steroidal Anti-Inflammatory Drugs, intended for pain relief and arthritis, would be used by tens of millions of people.

So here we have the set up: We have theoretical concerns that this drug would cause heart attacks; we know the drug is going to be given to tens of millions of Americans; and FDA does not insist that the drugs be safe, that there be clear, unequivocal evidence of safety.

Rather, FDA was willing to give a free pass on safety. FDA’s standard in this case, as it applies to all drugs, was: we can’t be 95 percent certain the drug will kill you, so we’ll assume that it won’t, and let it on the market. And then you and I, our parents and grandparents, our children — all of us — get to be the guinea pigs in that grand experiment — while drug companies continue to make profits, and FDA continues to receive its Prescription Drug User Fee Act fees for having put those drugs on the market in the first place.

MM: What happened when Merck withdrew Vioxx from the market?

Graham: When Merck pulled Vioxx from the market, it attracted the attention of Congress. This is a drug that is given to 20 million people, Merck is a big company, its capitalization drops $27 billion in one day — Congress says maybe something has happened, so they want to ask questions. Senator Charles Grassley was particularly concerned because he had seen the way FDA had dropped the ball with other drugs this year, involving antidepressants and the issue of suicidality in children.

When Senator Grassley’s staff began to ask me questions about Vioxx, I knew my goose was cooked. Because Senator Grassley’s Finance Committee was going to have a hearing, and I would be asked to testify. I was happy to tell the truth, but I wasn’t so happy at the prospect of being unemployed.

MM: How did your managers respond to your being asked to testify?

Graham: In the few days leading up to my testimony, managers at FDA — those above me in the “food chain” — sought to intimidate me. They did this in a number of different ways. They contacted Senator Grassley’s office, to try to convince Senator Grassley that I wasn’t worth his support, that I was a liar, that I was a cheat, a bully, a demagogue, and untrustworthy. At the same time, they contacted the Government Accountability Project, which is providing me with legal counsel, with the same line. In the third arm of this coordinated attack, my Center director contacted the editor of the Lancet, one of the most prestigious medical journals in the world, and accused me of scientific misconduct.

Scientific misconduct is the highest crime a scientist can commit. Scientific misconduct is a betrayal of all that science stands for. Scientific misconduct, if you have committed it, is career ending.

So here I am on the weekend preceding my testimony — the weekend on which I plan to write my testimony — and I get a telephone call from the editor of the Lancet, saying there have been serious allegations and what do I have to say. So I walked him through everything. There was no scientific misconduct. We had run into problems in our study, we addressed them and moved on. I had described all of these things to the editor of the Lancet at the time I submitted our paper. So he was very relieved.

My supervisors also knew that the allegations were untrue. They had been informed of all of the problems as they cropped up during the course of the study. They had a convenient lapse of memory — that would be charitable. Maybe it wasn’t a lapse of memory — that would make it intentional. We’ll let others determine which of the two it was.

By the end of the weekend, the editor of the Lancet had communicated to the Center director that I had handled myself in the finest tradition of scientists and that there was no scientific misconduct; there were no scientific problems with the paper; and that it should be published. The Center director agreed that it should be published.

It was supposed to be published online the day before the Senate hearing, so that the numbers in my testimony — the 100,000 heart attacks caused by Vioxx — would have been published in the Lancet and would have scientific credibility.

But my manager set a trap: If I allowed the study to be published, they could fire me. So I was forced to withdraw the article before publication. Fortunately, I was still able to give the testimony.

MM: The testimony made a huge splash, but it wasn’t enough to protect you from retaliation at FDA?

Graham: I had no idea that the testimony would attract the attention that it did. I had no intention of becoming a public figure, which to some extent it appears I have become.

When I finished the testimony, Tom Devine — my attorney from the Government Accountability Project, who is one of the two reasons I still have a job (the other is Senator Grassley) — said that if I did what I wanted to do, which was go back to my office and continue to do research and hope this goes away, I would be fired and be fired quickly. We needed to get into the media as quickly as possible, and I needed to make my case to the American people before FDA management had a chance to finish me off.

So it was a race. 60 Minutes came and said we really want to do your story. We asked, when will you be able to get it on the air? “Well,” they said, “we’d love to accommodate you, but we can’t get you there until mid-December.” In Tom’s estimation I would be a goner by mid-December. ABC News said we can get you on the air right before Thanksgiving, and so that is what we did.

MM: Is the Vioxx story an outlier, or representative of FDA failure?

Graham: The way FDA approaches safety is to virtually disregard it. FDA believes there is no risk that cannot be managed in the post-marketing setting. FDA has created the concept of risk management, which enables the continued marketing of unsafe medicines. If you do a careful examination of how it has managed drugs that it has identified as being unsafe, you’ll see that the techniques used to manage those risks are almost uniformly without efficacy. FDA knows this.

The case of antidepressants and suicidality is a perfect example. How does FDA handle this? With labeling changes. FDA knows that labeling changes don’t change physician behavior. Yet they act as if they are doing a great public good when they change the warning.

With the SSRIs [Selective Serotonin Re-Uptake Inhibitors, the class of antidepressants including Prozac and Paxil], I think FDA pulled a fast one on the American people. Because they said, “we are using our most powerful labeling, our most powerful medicine: we’re putting a black box around it.” But it doesn’t change physician behavior. People are as unsafe after the labeling goes into effect as they were before.

With the SSRI labeling, people are actually more deceived. Because the labeling says that the risk of suicidal thoughts and behavior in children is 1-2 percent. But FDA’s own senior managers admitted at an open public advisory committee meeting that our clinical trials don’t capture suicidal thoughts and behavior. So what we know about it is what has been voluntarily reported. There are lots more out there, we just didn’t measure them. They have evidence from an NIH clinical trial that the risk is more like 8 percent.

FDA doesn’t take safety as seriously as it should. They look for ways to serve industry, to enable industry to continue marketing unsafe products. The way the agency uses science guarantees it. Rather than ensuring with 95 percent certainty that a drug is safe, what FDA says is: We can’t be 95 percent certain this drug will kill you, therefore we will assume it doesn’t — and they let it on the market. That was the case with Vioxx.

That’s the dilemma we face with Celebrex now. We have this clinical trial that shows it increases the risk of heart attack. What is FDA going to do about it? It beats me what they are going to do about it. If they are concerned, I’d be very concerned. Because they weren’t concerned about Vioxx, and we saw what that got us.

MM: How have your colleagues responded to the controversy surrounding your testimony and statements?

Graham: My colleagues have been absolutely marvelous. Management has been typically hostile.

The day after my testimony, I came into the office, I was surrounded by co-workers who hugged me, kissed me, slapped me on the back, told me what a great job I did. It turns out a bunch of them had gone down to the cafeteria to watch the hearing on C-Span, and were cheering as I was stating one truth after another, and were really energized by it.

Subsequently, I learned that seven of my co-workers signed a letter sent to Senator Grassley basically saying, “He’s the heart and soul of what goes on here. Without him, we won’t have effective drug safety. And so you need to make sure that he is able to stay.” People who didn’t sign it because they are afraid of retaliation come to my office and say things such as, “Dave, we don’t know what we’ll do if you leave.” That has been very supportive.

Tom Devine, my attorney, tells me that the typical response to whistleblowers is that their colleagues abandon them, at least in public places, and tell them what a great job they are doing privately. But outside of that, there is no solidarity.

But that hasn’t been my experience.

MM: What do you recommend to reform the drug safety regulatory process?

Graham: I used to think you could create a separate center for product safety, located within the FDA. I am no longer of that mind. I have come increasingly to believe that a separate agency is needed for product safety, one that is insulated from the pressures that industry is able to bring to bear. Without that, we will continue to have bad decisions from FDA and hundreds of thousands people injured.

They say that physicians bury their mistakes. That is true. But they occur one at a time. When FDA makes a mistake, it is a mass grave.

I outlined three problems in my testimony. There is actually a fourth problem.

The three problems that I identified are a structural problem, a cultural problem and a scientific problem. The fourth problem that I would add is the fear factor: scientists need to operate with freedom. They can’t be subject to fear and pressure and intimidation.

The structural problem is locating the product safety center in the agency responsible for approving drugs. Big Pharma is already lobbying like crazy to prevent Congress from re-organizing FDA. The last thing on earth the drug companies want is strong, post-marketing drug safety regulation. Strong post-marketing drug safety is going to cost them money — it will cost more money to research drugs before they hit the market, and they are going to run the risk of experiencing the consequences of a drug that isn’t safe being removed from the market.

The culture needs to be changed. Industry can’t be the client. The culture is dominated by a worldview that believes only randomized clinical trials provide useful and actionable information and that post-marketing safety is an afterthought. This culture also views the pharmaceutical industry it is supposed to regulate as its client, over-values the benefits of the drugs it approves and seriously under-values, disregards and disrespects drug safety.

The third problem is a scientific one. Take Vioxx, Celebrex and Bextra as an example. Before they come to market, we know that there are good theoretical reasons that these drugs will probably increase the risk of heart disease. You can design a study that would allow you to say with confidence that the drug is safe to a certain level. You can pre-specify: how many extra heart attacks are we willing to accept, for the benefit of this drug. You can design a clinical trial, that will show yes, the drug does not increase the risk of heart attack more than 5 percent or whatever level you specify, and yes it is saves this many lives from gastrointestinal bleeding.

Those studies would have to be much larger and take much longer than current trials. And that means industry wouldn’t be able to get these blockbusters to market as quickly. If you are making $3 billion a year selling a drug, every day of clinical trials is another day you are not making $10 million.

From a public policy perspective, if we wanted drugs that are safe, we could have it tomorrow. It is easy to design those studies. But FDA is not interested in that.

Fourth, there has to be whistleblower protection. That is a start. But there also has to be strict enforcement of rules to protect scientists, so that they are not faced with losing their jobs if they arrive at conclusions that put FDA in a bad light.

FDA just had a survey done. Two-thirds of FDA scientists are not confident that products approved by FDA are safe. Eighteen percent say that they have been pressured to change their conclusions. Those are horrible statistics for an agency that is supposed to be evidence based and science based.