TOPICS > Science

Dr. Robert Califf on Vioxx

November 3, 2004 at 12:00 AM EDT

SUSAN DENTZER: Let’s start by talking a bit about just some general concepts that we want to get across in this story, one of which is that people don’t always appreciate the fact that all drugs have risks and all drugs have benefits.

DR. CALIFF: I think it’s fair to say, the first principle of therapeutics is that, all drugs have benefits and risks and it’s always a balance of the benefit and the risk that we are trying to determine, as we develop drugs as therapeutics.

SUSAN DENTZER: Do you think this is commonly misunderstood by the public or, at least, that the import of that is not fully understood?

DR. CALIFF: Well, it’s a balance of benefit and risk is a very complicated construct if you really think about it. And I think it is not only misunderstood by consumers or the public very often, but even by health care professionals.

And if we think about it, the human brain is sort of constructed to make binary decisions. I’m going to do this or do that. And it’s only natural to think that, oh, if a drug is good, it’s all good or if it’s bad, it’s all bad. But the fact is, anything that we ingest chronically, particularly chronically over time, it’s going to have a number of effects that we don’t necessarily understand hen we envision what the structure of the drug is or what it’s metabolism might be. And we can only know that if we study it.

SUSAN DENTZER: Let’s move to just some basic discussion about studies. What do we gain, for example, when we study two drugs two drugs head to head, versus studying a drug against a placebo?

DR. CALIFF: You know, it might be worthwhile just to sort of start at the beginning. We’ve got to realize when a scientist envisions a molecular structure, we draw–thinks of a pathway by which it might work, usually first in a test tube and then in animals of various kinds. There are literally thousands of those for every one that makes it even into a human being in the early phases of study.

And so, you know, it’s really important to understand the work that has to go into getting to the point of even beginning to test the drug in a human being. And approximately, out of every hundred that make it into a human being in a phase one study, somewhere in the order of five, between three and five actually make it out on to the market. So, there is a huge winnowing process that has to go on, simply because our biology is so complicated, we probably will never understand it all. It’s just too much going on all at one time in a human being.

You know, when we developed a new molecular entity and we want to find out if it works in human beings, of course, there has to be a first person to take the drug. And in that circumstance, which we–circumstance we’re usually giving one dose or several doses and then we’re measuring things that happen after we give the dose. It’s unpredictable. That’s what we call a Phase I study.

Typically, this is done with a placebo control group, that other people are just giving a group of people the drug and measuring what happens. But this only gives us a very first, initial picture about the fact that nothing bad happens or that it behaves like we think it will when put into the human being in terms of its metabolism and biological activity.

Phase I doesn’t tell us anything about whether it’s really beneficial in a patient population, given over a period of time. So, most things get eliminated in phase one. The drugs don’t make it. They either have a side-effect or a problem with metabolism that wasn’t anticipated.

So, out of those that make it through phase one, then we have to begin to do studies in larger populations to try to figure out if we can give it to people and nothing bad happens, what might be the right dose to think about giving. Those are what we call pharmacokinetic and pharmacodynamic studies, where we give the drug to a population in different doses and we measure what happens to the drug in the bloodstream. Then we measure the side-effects and we begin to get a picture of what’s the range of doses that we might be able to give.

Again, out of every drug that makes it into phase two, out of every ten, maybe three will make it into the final phase of study.

In the old concept of how we do the whole drugs, phase three might have been thought really as sort of the final phase. These are studies that are done to get approval by the FDA for putting the drug on the market.

There was a great time, 20 or 30 years ago, when for most of the diseases that we have, we didn’t have an effective treatment. And so, the way to study the drug for all purposes was to compare it to a placebo. So, randomly selected, about half of people with the disease would get the drug and about half would get a placebo and then you would see what happened. If the theories were correct, the drug would provide benefit compared to placebo.

Now, that’s a fairly easy test in today’s world, because now for most diseases, there’s at least one effective treatment and for many diseases, there may be as many as a thousand effective treatments. Here, again, we have an issue of thought that needs to be carefully considered. And when we say effective, almost nothing that we do cures the disease. Almost everything that we do medically makes the disease better, but doesn’t cure it.

So, we think about heart disease, which is my specialty. Bypass surgery is a great operation. President Clinton just had it. It doesn’t cure the coronary disease, but it does enable one to live longer than if one had not had bypass surgery that is done to the right patient. In the same way, the statin drugs, the Lipitors, the Simvastatins of the world, Zocor, that we have prescribed, they lower the cholesterol. They provide a reduction in the risk of death, but they don’t eliminate strokes and deaths from heart attacks.

So, in the good old days, we could compare it to placebo. If you were better than placebo, you were put on the market and that was a great thing.

Now, we’ve got this very difficult question of not only are you better than doing nothing, but is this treatment better than another alternative treatment. This is actually a relatively new area of science. We don’t have a long history of doing direct comparative clinical trials and we’re still learning about how to interpret them.

SUSAN DENTZER: Let’s talk about, just briefly about some general classes, NSAIDS and the risk that [inaudible] of those and then we will move to the specific situation of Vioxx.

It’s not commonly understood, as you said, that all drugs have risks and benefits. It’s perhaps not even commonly understood that the entire class of non-steroidal anti-inflammatory drugs have cardiovascular risks. So, let’s talk about that. What are the risks of these drugs?

DR. CALIFF: We’ve known for a long time that non-steroidal anti-inflammatory drugs have a variety of risks, in addition to their benefit. First of all, let’s back up and recognize we live in a society where people are living longer and longer. If you just walk the streets of a town like Durham, North Carolina, about half the people that you run into will have had an aching joint some time in the past week. So, we are–or a headache or something like that and ware not taking non-steroidal anti-inflammatory drug that would relieve that discomfort.

It is very dramatic and obvious the symptomatic benefit that one gives.

On the risk side, from the very beginning, we’ve known that these drugs cause [break] in the stomach. And in fact, it’s a surprise to a lot of people to learn that thousands of people die from gastrointestinal bleeds [break] here in the United States from non-steroidal anti-inflammatory drugs taken chronically. Also, if you take them in very high doses, there’s problem kidney damage if you do it chronically.

So, these are drugs that are now over the counter, because we use them so much. But those are sort of really obvious risks that we’ve known about.

SUSAN DENTZER: What about the cardiovascular risks?

DR. CALIFF: What wasn’t really known until recently was the potential for cardiovascular risk, mostly because we thought of blocked up blood vessels as simply a problem of cholesterol being too high. We now know that inflammation is a key part of coronary-artery disease. So, these are drugs that alter the inflammatory pathways that we have in our bodies that are natural defensive pathways against infections and irritation and all sorts of things that happen to us.

So, when we perturb that inflammatory system, it could be either good or bad. And we don’t–it’s so complicated, we can’t really tell which is which. So, it really has not been known that there might be risks, because the non–that there were risks, but it’s been a theoretical possibility.

And now, interestingly, as this new class of drugs, the coxibs, has come about, the old-fashioned drugs, which are mostly generic now, are coming under scrutiny again. What’s been fascinating is that, one of them in particular, Naproxen, which was taken as a comparator to the coxibs, in head to head clinical trials, worked better in several studies now than the COX-2 inhibitor drugs.

It looks like now, in retrospect, that it’s entirely possible that Naproxen may actually be beneficial for the heart. Some are like aspirin. Remember, these drugs are a bit like aspirin in the way they operate. They affect the same biological processes in a somewhat different way.

On the other hand, a drug like Motrin, commonly taken in the United States or Diclofenac, which is the most commonly taken non-steroidal anti-inflammatory drug. It’s not taken so much in the U. S. They work a little bit more like the Cox-2 inhibitors in the studies that have been done [in that they] may have risks of causing heart attacks, not proven, potentially. : Commonly taken, non-steroidals like Motrin, known as Advil, potentially could cause heart attacks.

SUSAN DENTZER: Do we know that?

DR. CALIFF: We’re not sure. There’s a potential for risk. It’s not been adequately studied, believe it or not after all these years.

SUSAN DENTZER: So, what does that tell us in general about this entire class of drugs?

DR. CALIFF: Well, like every class of drugs, the NSAIDS have benefits that are clearly documented. They also have risks. The extent of the risk has not been adequately quantified at this time.

SUSAN DENTZER: And that includes these over the counter medications that we’re taking in droves every day?

DR. CALIFF: It’s perfectly fair, I believe, to say that, you know, at the time that the over the counter NSAIDS were developed, we didn’t know about all of this. So, I’m not laying blame on anyone, but I believe it’s perfectly fair to say, they’ve not been studied for the long term risks of cardiovascular disease.

SUSAN DENTZER: Let’s move to the situation of Vioxx and the fundamental question, where dif the wheels come off the cart here? What, in your view, was the single thing that went wrong; if went wrong is the wrong phrases, where we missed the boat?

DR. CALIFF: Okay. Fundamentally, we have a problem with our drug development system and it stems from a new understanding of biology and science that we didn’t have even a decade ago. If we don’t study drugs for their long term effects, we really don’t know whether, on balance, they provide benefit or risk. It’s important to understand that, in a scientific development program for a drug, we’re accumulating evidence over time. And when you cross the threshold to say that there’s enough evidence to say there’s a definite risk that outweighs a potential benefit, it’s very much still a matter of judgment. It’s not an all or none sort of phenomenon.

So, I don’t think we can point to a particular point in time where a single thing went wrong. I think we have a system, which in my view, doesn’t accumulate the evidence at as rapid a pace as it should.

SUSAN DENTZER: Was that the case in this instance, in particular, given that it was know as early as 2000 that these risks showed up, had shown up against Naproxen?

DR. CALIFF: Well, I would say that Merck played by the rules from everything that I’ve seen, including some of the recent things that have been in the newspapers about e-mails and that. Within any scientific development program there are debates that go on, which if we suppressed, we would be extremely unhealthy. So, we’ve got to have people communicating with each other about their beliefs, about risks and benefits and how development should be done.

I also really believe that it would be better, once we accept that these drugs can have long term effects, to go to the heart of the matter; do the studies that will get the answer the fastest possible way.

In the case of Vioxx, many of us recommended to Merck to take, take people who already had coronary disease with osteoarthritis, many of whom have been taking Vioxx, relatives of mine being among them and, go ahead and do the study on those people. They are already taking the drug. In a sense, randomizing to an alternative treatment, in a 50-50 allocation would be perfectly fine. And then we’d get the answer much quicker than the study that was done.

What many of us are recommending would be that Merck should have taken people who are already taking Vioxx, but half of them still on Vioxx, the other half on an alternative non-steroidal anti-inflammatory drug and then measure the outcomes in that population. By taking people who already had coronary disease, one would accrue the evidence that was needed at a much faster rate.

SUSAN DENTZER: And you recommended that to Merck?


SUSAN DENTZER: What did they say?

DR. CALIFF: Well, the way things were, many of us consult with the pharmaceutical industry, which I think is a very good thing. They need ideas and then the decision about what they do is really up to the person who is funding the study. So, they took our advice into consideration and decided to do something else. They chose to do another study, which in a sense, was cleaner, because it was a placebo controlled trial. But it studied a population with a much lower event rate, so it took much longer to accrue the evidence.

SUSAN DENTZER: Is that why you think fundamentally the evidence only showed up after 18 months?

DR. CALIFF: It’s a great matter of debate and we’re going to have a lot of fun scientifically wracking our brains about why there was no difference in the placebo controlled trial for the first 18 months. There are two theories about why there was no difference.

The first theory would be that it was the lowest population, so things just didn’t happen at a fast enough rate until later. The other would be that, the drug only has this negative effect over a very long period of time.

I would say, the implications of how you interpret the first 18 months are very important. The major implication of the first 18 months, if you believe that it’s actually an occurring detrimental effect, which doesn’t show up for 18 months–most chronically taken drugs we don’t study for 18 months. And so, there may be many others that either good or bad effects beyond that period of time that we simply don’t know about at this point.

SUSAN DENTZER: Can you offer some examples of some chronically taken drugs that we know large numbers of people are, in fact, taking, antidepressants, what have you that–

DR. CALIFF: Well, of course, one of the class of drugs there’s been the most debate about publicly lately has been the SSRI antidepressants. They’re all over the place. We had the privilege of coordinating the National Institutes’ of Health study in teenagers with depression. And you know, our interpretation of the data is that these drugs are very beneficial for teenagers with depression, but they also carry a risk in the short term.

It’s also fair to say the long term risk is completely unknown. And also, the long term benefits are completely unknown.

The SSRIs are a really fascinating class of drugs in this regard, because we have evidence that in people who have had heart attacks–many people don’t realize this, but about a quarter of people after heart attack have a serious depression. Those people have a four-fold increase in the risk of dying over the next period of time. It looks like the SSRI drugs may actually prevent cardiovascular death.

Unfortunately, no one has done the studies to prove it one way or the other, at this point.

SUSAN DENTZER: Let’s make, again, the fundamental point that you made just a moment ago, which is that we do have these large classes of drugs now that are taken for long periods of time for chronic illness and our system of studying them is not adequate to capturing the results of long term use of these drugs at the time they go on the market.

DR. CALIFF: So, we know now we have many types of drugs that are given to a population living a very long time with chronic illness and, I think it’s fair to say that, up until recently, we just didn’t understand the impact that this could have on the overall health of the patients that we’re treating. As a result of not understanding the issues until now, we have a system which is very much geared toward getting approval from the FDA, based on short term studies.

And what happens after that in terms of post-marketing studies for the most part is dictated by the companies themselves, because they have the responsibility of selling the drug and of measuring the outcomes related to that drug. By and large, this system makes it very difficult to do long term studies that measure outcomes.

SUSAN DENTZER: What do we need to replace it with?

DR. CALIFF: Well, how to change the system is a matter of great controversy. And of course, right now, right after an election, if the system is going to change, this is the time to do it, while it’s relatively safe to make changes and then repair it before the next election. So, many of us are hopeful that there will be system changes in how things are done.

The two main ideas that I would have, number one, I believe we should go more to the system of conditional approval for chronically taken treatments. The reason for conditional approval is that, we have to continue to encourage people with capital to invest in developing new treatments. We desperately need that. If we say that it’s going to take 20 years before you can get on the market, people will invest their money elsewhere and the industry that makes new drugs will die.

The other reason to make it conditional is that, if we do put it on the market with the right label that says, take it based on what we know now, if you want to, but realize that long term studies are being done. You need to understand that and consider that when you decide whether you want to take this particular drug.

The second idea is to get the government- funded sources more involved in doing human studies. Most people don’t realize the vast majority of the NIH budget does not go to human studies, but goes to basic science types of studies. So, we have a system where there are questions crying to be answered for the public health and there’s no mechanism to answer the question.

Medicare today is identifying literally dozens of multi-billion dollar, major public health questions and there’s no mechanism within the government for Medicare to say, we got a critical question that needs to be answered. Somebody go answer it. Instead, what we have is a bunch of unanswered questions and NIH was just not focused on this issue.

SUSAN DENTZER: Changes that would redress those specific problems would do what — create a new government-funded entity to conduct these trials?

DR. CALIFF: Many of us believe that we have to have public-private partnerships. Other than terrorism and war, which is, you know, a major thing that we all have to worry about, the chronic aging of the population, chronic diseases is really the number one sort of issue that our society faces, whether you’re talking about the Medicare budget, nursing homes, all the things that we worry about. We need public and private sources joining forces in terms of the finance to get us the evidence that we need to make the best decisions.

Remember, that a decision to give the wrong drug now for a disease like coronary disease for 16 million people at risk can have huge financial and health consequences for our whole population. We got to take that into account.

So, in response to the question, how should this evolve, many of us would say that the questions ought to be determined in a public forum. That is, what is it that needs to be answered? What do we need to know when we compare treatment A and treatment B? That should be a matter of all the parties. The consumers, the doctors, the funding agencies, the people that pay for health care should get together and decide the priorities.

The funding itself should come through a combination of public sources and the companies that sell the products. Remember, that they already do a huge amount of research in the post-marketing period. It’s just that the questions are determined by the companies, not by the public health need.

SUSAN DENTZER: So, what is this forum? It doesn’t exist. I mean, do we have to create it?

DR. CALIFF: Well, we’re having a meeting at the Institute of Medicine this month actually to begin a discussion about what the forum should be, about which priorities might be set. Remembering that no one has said we’re going to have such a system, but if it were to evolve, you might think of it as a national problem list.

You know, in medicine, we think of things in terms of problem lists. You’ve got heart disease and cancer and problems that you need to solve. If we had a national problem list that said, what are the most important questions that need to be answered through research? It might be how to say coxib drugs stack up against an NSAID. It might be, what is the role of bypass surgery versus angioplasty. It might be, how do we image the brain to detect Alzheimer’s disease early.

All these things are competing. So, the priorities need to be set.

As we think about this, it’s a dizzying array of issues that would need to be addressed. On the other hand, we got to remember that Medicare is a trillion dollar issue. If just a very small part of the Medicare money that went into health care went into figuring out which treatments really worked and we stopped using the treatments that didn’t work, it might actually be a financial positive.

SUSAN DENTZER: Staying on this question of how we create a more rational system, actually, I want to go back to a comment you made a moment ago, which is that you believe Merck was operating by the rules, as they exist now. It’s just that the rules were wrong, in effect, is what you’re saying. Could you say that in that compressed way?

DR. CALIFF: The only thing that I’m saying, I think Merck acted by the rules. I just think we need a new set of rules.

SUSAN DENTZER: Again, to flesh out the thought, this new set of rules would…?

DR. CALIFF: And the new set of rules that we need would really mandate that there be long term comparative studies of treatments that were being offered for chronic diseases.

SUSAN DENTZER: You say that it was up to Merck to decide that it wanted to do the APPROVe study [in hopes of getting] the new indication [for colon cancer prevention]. They decided what it was that they were going to do and they drove the study process. So, what happened in this vein with respect to Merck and how Merck made these calls about what would be studied?

DR. CALIFF: You know, in a situation where there’s so much money at stake and so many people’s lives, we have to have a system where if people operate by the rules, that they do okay. The rules that we had that Merck was dealing with essentially said, we only need short term studies. If you want to do long term studies, it’s up to you to make that decision.

I believe we need to have a different system in which part of the process of approval is a concerted effort from the perspective of public health to design long term studies that compare the new drug with the old alternatives head to head.

SUSAN DENTZER: You said earlier that one way of dealing with this would be to grant the particular drug conditional approval. So, what would that look like? What would the FDA actually say to a drug company about conditional approval for a drug?

DR. CALIFF: Well, there’s a mechanism for conditional approval now at the FDA that’s used particularly for innovative drugs that don’t quite meet the standards we would normally have, but for a disease where there’s no other alternative, where people would say it’s worth the risk not knowing what we normally know to get this drug on the market. So, conditional approval would be a statement from the FDA that says, it’s okay to market the drug by a particular label and then you have to commit to do a long term study of a certain type.

Now, under the old rules, the design of that long term study was really up to the company. I believe we need a system where the design of the long term study is a matter of public health, where consumers, physicians, the FDA and payers are involved in designing those studies.

SUSAN DENTZER: Had that system been in place and had Merck had to operate under those new rules for Vioxx, what study would have been asked for and when?

DR. CALIFF: One of the fascinating parts about this whole industry is that, under the current rules, the company is wanting to do the right thing almost always, but they have to ask the question. How much risk are we going to take that we’ll get a bad result, when if we did a different study, there’s a lower risk and it might actually show an additional benefit of the treatment?

And they have to answer to a board of directors of the company, which in the end, has to allocate the money. Of course, in a publicly owned company, from the way public companies are funded, which is largely our pensions, places like Duke University. So, it’s a fascinating system.

I believe if looked at from the perspective of the company, you might choose a lower risk strategy. It might take longer to get the answer. You might have additional benefits to the company. From the public health perspective, the question was, how can we get the answer in the quickest possible way. It would be two different perspectives.

I think if everyone played under the rules where once you’re on the market, there was a mechanism for a public health perspective, we’d all have the same ground rules and I think it would work quite well.

SUSAN DENTZER: In this case, would that study have been basically what you suggested that they do, that they didn’t do or would it have been a shorter term, head to head study against another NSAID? What would it have been?

DR. CALIFF: One of the beautiful things about clinical medicine and clinical therapeutics is there are many opinions. I can only give my own opinion. But I believe that taking people with coronary disease, who had aching joints and needed to be on a treatment for that anyway, randomizing them would have gotten the answer very quickly about the risks and benefits of Vioxx. And I believe we would have won the argument, those of us advising that, if it had been a different forum.

SUSAN DENTZER: And we would have gotten those results sooner than these results that came out of the APPROVEe study?

DR. CALIFF: Yes, I believe we would have gotten the results sooner with a strategy of studying people who already had coronary disease.

SUSAN DENTZER: As it is, again, as you said, playing by the rules, what was in Merck’s interest was to mount the APPROVEe study to get a new indication for colon cancer prevention, which had it succeeded, would have been a phenomenal new use for this drug.

DR. CALIFF: Right.

SUSAN DENTZER: They had every reason to set the trial up the way they wanted, as opposed to answering this other fundamental question.

DR. CALIFF: It’s a rock and a hard place for a company under the current set of rules.

You know, if you make tennis shoes, no one asks you to go out and do studies that show that you have lousy tennis shoes as a design of a study that you did. And in a sense, that’s why we need an FDA that’s strong, that has a different set of rules that regulate the commerce of getting drugs to people.

And it’s important that those rules are also applied fairly across the spectrum so that no single company is disadvantaged.

SUSAN DENTZER: So, in your view, the FDA was playing by the rules, which are inadequate. Merck was playing by the current set of rules, which are inadequate. It’s not culpability here so much as there is a systems failure. Is that a fair phrase?

DR. CALIFF: One of the things that struck me as interesting is someone said to me recently, you know, if there’s going to be a congressional investigation, it’s Congress that should be investigated.


DR. CALIFF: The rules for the FDA are set by Congress. And I do a lot of work with people at the FDA. They’re constantly being called to hearings for show, often without a full understanding of the complexity of what needs to be done.

So, if the FDA rules are going to be changed, it’s really a political process that makes that happen. And if there is a good thing to come from the situation that has occurred with Vioxx, it would be to get the momentum to create a new set of rules. I believe that both Merck and the FDA were doing what they normally do by the set of rules that exist. The rules are just not adequate for what we know now scientifically.

SUSAN DENTZER: One that I would like to cover as well is, the tension that was clearly underway within Merck between the marketing side and the scientists and even among the scientists themselves about what was really going on [with Vioxx].

First of all, could you put that in context for us? Is that typical of the debates that go on within drug companies?

DR. CALIFF: Everyday in every medical product company, whether it’s drugs or devices, there’s a conflict between the scientific arm and the marketing arm of the company. The scientific arm has a job to do the studies that depict the risks and benefits of the treatment that’s being made by the company. The marketing arm has a job to sell the product within the confines of the rules of the game that define the ethics and rules of commerce that they need to live by.

It’s every day activity to have discussions back and forth between marketing and science in interpreting the potential risk and advising those who are on the ground in the marketing arm, talking to doctors about what to say.

SUSAN DENTZER: When the plaintiffs’ attorneys and others say the marketing arm is driving the show here; they were selling this drug even if the allegation is almost beyond its indication of playing up pain relief aspects that exceeded what even the evidence showed and underplaying the cardiovascular risks, which they had been told to put on the label, what is the accuracy of that statement? Does that charge stick in a case like that?

DR. CALIFF: The marketing arm drives every company that sells medical products. That’s how the revenue comes in that, of course, funds the research and the development of new products. Without selling the product, you have no company. So, there’s always a tension between the marketing arm and the science arm.

The marketing arm is charged with depicting what’s in the label. We know the way the human brain works. If someone comes in and in a friendly way tells you about all the positive things about a drug and maybe mentions the negative things, but not as prominently, your mind is going to seize on the positive things. That’s the art of marketing.

I believe that we do have a problem with the medical profession in this regard also. The medical profession is entirely too dependent on sales representatives to inform doctors about the risks and benefits of therapies. This is a very difficult problem, particularly given all the great, new treatments that we have. It’s way too much for an individual doctor typically to keep up with. And our profession has not kept pace, unfortunately.

SUSAN DENTZER: Was that a problem in the Vioxx story as well?

DR. CALIFF: I believe that the inability of the medical profession to draw its own conclusions, you know, in office to office transactions is part of the problem that occurred with Vioxx. It sets up a circumstance where an effective salesperson can convince you that it’s good to give an expensive treatment when a less expensive one may be just as good.

SUSAN DENTZER: In that context, was this drug oversold?

DR. CALIFF: Well, it’s been said and I don’t know the actual data myself. But it’s been said that, I guess, it was three years ago, the number one spend on direct to consumer advertising was for Vioxx. And so, here we have a drug heavily hyped in direct to consumer advertising that has to be pulled off the market due to later data. I think that tells us there’s something wrong with the system.

SUSAN DENTZER: I want to also ask you to paraphrase a comment that you made in an editorial you wrote back in 2002. You said: “Our patients, the public and volunteers for human experiments deserve better than the current state of confusion about Cox-2 inhibitors.” And you said that in 2002, two years before this happened.

DR. CALIFF: This information about Vioxx came into focus several years ago. It was really obvious to many of us that, our patients really deserved better information so that we could make choices in their interests.

Besides that, the fact the studies are controlled for the most part by the companies brings in the question that, those who volunteer for the human studies, as to whether all the information is getting out to the public [sic] in the way that it should. Now, I believe that Merck did get all the information out that they had, but it’s filtered when it comes out from a company in a way that’s a bit different that–than if there’s a more public airing.

SUSAN DENTZER: One example of the filtering, at least, in the views of some was, when Merck came back after the VIGOR trial was also published with a pooled analysis, it left those Vigor results out of the pooled analysis.

DR. CALIFF: Yes. You know, I’m the editor of a journal called “The American Heart Journal,” which is the oldest cardiovascular journal. It’s not the biggest now, but it’s a lot of fun. We publish clinical articles and we got a review article that was authored by an academic with a number of co-authors at Merck.

Our reviewers initially rejected the paper, but when I looked at it, as the editor I said, you know, we really should publish this because the information needs to get out there and we need to air it out. So, we then asked several academics to write editorials about that article.

It seemed clear to me that the information was being filtered through the eyes of Merck, not suppressing the data, but in–by drawing conclusions about what should be included in each publication, if you believe in the product, you’re going to see it differently than if you’re totally unbiased.

SUSAN DENTZER: And do you think that is more what was going on here?

DR. CALIFF: I think it’s always the case, if you work for a company that makes a product or you’re the inventor of the concept, frequently an academic or your reputation is made by writing papers or giving lectures about how good the product is, it’s very hard to be unbiased in that circumstance. That’s another critical aspect of clinical trial methodology that I think we’re just coming to understand.

Clinical trial often involves thousands of human volunteers for a human experiment, literally dozens to hundreds of investigators, health care providers who get the consent and put the patients in the study. I believe the interpretation of that data should be a public affair. It’s entirely legitimate to have the same set of data and for different people to draw different conclusions looking at the data. This happens everyday.

It’s only by getting it out in the open air and having people express their opinions that you begin to get more of a consensus about what the real meaning of the result is.

SUSAN DENTZER: Which ties this back to the issue of having a clinical trials registry.

DR. CALIFF: Right. It’s been a matter of great public discussion recently about the fact that many human experiments–and that’s really what a clinical trial is, is people volunteering for human experiment. Many of these never see the light of day, at least, the way it’s been up until now, often because the company, with all integrity that they have, will say, well, it’s not really important. It doesn’t really matter in the context of medical care.

But my question is, who should make that judgment?

SUSAN DENTZER: You were one of the consultants who the top people at Merck called as soon as they got the data on the safety issues in the APPROVe study, and asked what should they do. Tell us the advice that you gave Merck at that time.

DR. CALIFF: Well, when the result of the clinical trial was made available to Merck by the Data Safety Committee, they immediately got together a number of us as consultants to advise them about what to do to get–not to tell them what to do, but to give them our opinions. My opinion, on looking at the data, was that they should go to the FDA and get a warning label on the, on the label that said, here’s the risk that we observed. Be aware of it; make your decisions based on your judgment about the benefit you get from relief of joint pain versus the risk of a heart attack down the road.

SUSAN DENTZER: Why did you think a warning label was, in effect, enough?

DR. CALIFF: My idea about the warning label is that, almost all of these drugs have an unquantified amount of risk. And so saying, we’re going to take Vioxx off the market, in a sense, is pushing people to take another drug, which may be no more proven to be safe.

In addition, many people have told me in my clinical practice and relatives of mine–and in fact, when I was on a national talk show, the person who put my makeup on practically begged me to go to the FDA and convinced them to keep it on the–Vioxx on the market, because this person said, it’s the only thing that works for me for my chronic joint pain and I’m not functional without it. So, I will never advise, based on the available data, that Vioxx be a first line treatment, but to have it not available based on the amount of risk that’s been quantified, I think it’s really a shame.

SUSAN DENTZER: Your own mother was taking Vioxx. Ttell us the advice that you gave her.

DR. CALIFF: I don’t think my mother would mind there if I said that, she was taking Vioxx and osteoarthritis. My brother is an orthopedic surgeon. He is prescribing for her and it worked great.

My advice was switch to a generic and say, Naproxen, which is cardio-protective. Take a proton pump inhibitor with her and then you’ll be fine. My mother, of course, said, well, Vioxx works better than anything else. Maybe at my age, I’m willing to take the small risk of a heart attack in order to be able to move around and do what I want to do. I think it’s also fair to say that, like many people, my mother has a supply ready that she squirreled away so that if she needs it, she can take it.

SUSAN DENTZER: Back to that advice that you gave to Merck, which was to go ahead and go to the FDA [and propose], “Let’s have a higher level of warning than has been the case previously.” Why do you think, in the end, they didn’t take that road?

DR. CALIFF: I actually don’t know why Merck didn’t take my advice, as good as it might be. (Laughter). I believe though a major factor is the risk of liability. And I think we have a very serious problem in many aspects of American life now about dealing with liability. Everything we do in life has a balance of risk and benefit. And I believe that Merck felt that if they kept this on the market, the number of mounting lawsuits would just be something that couldn’t be tolerated, even if a limited number of people would benefit from taking the drug.

There are many other drugs that never get developed because of concern about liability that might do quite a bit of human good.

SUSAN DENTZER: We spoke with a woman yesterday who is a now 48 year old woman, but several years ago, she was taking Vioxx for a bone spur. She’s a runner, a pleasure runner more than anything. She took Vioxx for two and a half months and, out of the blue, with no preexisting cardiovascular risks, had a heart attack and ended up with three stints and is now essentially going to have to retire on disability from her former profession as a medical researcher.

Is that a plausible story to have been, in effect, caused by Vioxx?

DR. CALIFF: It’s plausible that anyone who has heart attack who was taking Vioxx that it could have been caused by Vioxx. However, it’s critical for people to understand that, the relative risk of Vioxx is very low and the absolute risk is very low. So, out of everyone who has a heart attack on Vioxx, maybe Vioxx plays a small, contributory role, but Vioxx as a sole culprit would be almost implausible.


DR. CALIFF: Well, if we, if we–again, if we look at people on placebo and people on Vioxx, people on placebo had almost as many heart attacks as people on Vioxx. The increase was something like 1.5-fold over period of 36 months, which means that two-thirds of the heart attacks couldn’t have possibly been attributed to Vioxx, because they were equally likely than half a known placebo. We also have to keep in mind that there was no difference in the death rates in the two arms in the APPROVe trial.

SUSAN DENTZER: Some of the criticism has been that Merck never attempted a definitive cardiovascular safety trial of Vioxx. They did have this secondary end point of cardiovascular risk in APPROVe, but they never mounted the overall, definitive cardiovascular risk trial. Is that a fair criticism?

DR. CALIFF: I believe it’s unfair to assert that Merck definitively had to do a cardiovascular risk trial. But I believe it would be the right thing to do, to do a cardiovascular risk trial. By the rules that Merck was operating under from the FDA, mandated by Congress, there was no imperative to do it from a business perspective.

SUSAN DENTZER: So, we’re back to the issue of what the rules, what the current rules require, as opposed to what better rules should require.

DR. CALIFF: Right.Wwhile I wish the rules were different, under the current rules, Merck was not required to do a cardiovascular risk study.

SUSAN DENTZER: Even the FDA didn’t necessarily have the power to require them to do it?

DR. CALIFF: I would say the FDA’s interpretation of their mandate was that they could not require Merck to do such a trial. Otherwise, I believe they would have.

SUSAN DENTZER: One final question, which is that, an issue has been raised about the VIGOR trial in particular with respect to excluding patients who would have normally have required aspirin. Let’s try to put this into English. DR. CALIFF: Not only did Merck not do a mortality trial in patients with heart disease, in the Vigor trial, they actually excluded people with a serious history of heart disease. All but the lowest risk patients were excluded in terms of heart disease. So, there were some people with heart disease in the trial, but it was pretty carefully constructed so that intermediate to high risk patients were excluded.

SUSAN DENTZER: And there was this additional issue about whether any patients who were also given aspirin could be in that trial.

DR. CALIFF: One, one of the issues that I found fascinating in reviewing the VIGOR trial is that, taking aspirin was a contraindication to being in the trial. That, essentially, should exclude everyone with heart disease, since aspirin is an imperative for these people.

Now, it turns out some people with heart disease did get into the trial. And you wonder who were their doctors, because they probably should have been on aspirin.

Another fascinating aspect of this whole issue is that, if you combine a cox-2 inhibitor with aspirin, you essentially get the same effect as a generic non-steroidal anti-inflammatory drug as far as we can tell. So, right now, most of us would say there’s not a great benefit of taking a cox-2 inhibitor if you’re taking aspirin in terms of prevention of gastrointestinal bleeding.

SUSAN DENTZER: You might as well just take Ibuprofen.

DR. CALIFF: Well, don’t take Ibuprofen. Take a Naproxen, is what I would say, at least, based on the available data.

SUSAN DENTZER: So the way the rules were written to get into the VIGOR study, Merck said, in effect, if you’re a patient who is taking aspirin to reduce your risk of heart disease, you can’t get in this trial. Why did they say that and what was the effect of it screening those patients?

DR. CALIFF: I wasn’t there to design the VIGOR trial, but my understanding of what happened was, the goal was to compare Vioxx with the non-steroidal drugs. And the question was, can we reduce gastrointestinal bleeding. If a patient was on aspirin plus the Vioxx, that probably would have not created any difference in the GI bleeding compared to taking a non-steroidal with aspirin.

So, I believe the goal of Merck was to demonstrate there was a benefit in terms of gastrointestinal bleeding. The result of that was, that higher, high–medium to high risk patients with heart disease were excluded, because they should be on aspirin.

SUSAN DENTZER: So, essentially, it gets back to the point that you raised. They should have addressed this by putting a lot of people at risk for cardiovascular–at high risk for cardiovascular disease on the drug to see what happened.

DR. CALIFF: As a cardiologist, I know that many of my patients have been taking drugs like Vioxx. They also really need to know whether there’s a benefit to risk balance, which is favorable to them. So, I believe that by doing a high risk cardiovascular study, they would have not only answered the question about the general risk, but would have helped the 16 million people in the United States today who know they have heart disease already, most of whom have aching joints, because they are over age 65.

SUSAN DENTZER: A lot of people have said that the Vioxx episode was a true, unmitigated drug disaster.

Do you agree? And if the word isn’t disaster, what would you call this whole episode?

DR. CALIFF: I believe the Vioxx episode was a signal to us that we have multiple drugs out there that may be doing a great amount of benefit or a great amount of harm that we just don’t know about at this point.

SUSAN DENTZER: Was it a disaster or was it just a wake up call like none we’ve ever seen before?

DR. CALIFF: I believe that Vioxx was a wake up call, but no more so than hormone replacement therapy, which was a much more widely used drug with about the same amount of harm.

SUSAN DENTZER: So, it’s just the latest in the wave that was, perhaps, most recently hormone therapy–

DR. CALIFF: SSRIs and suicide, it’s all a balance of benefit and risk. And unless we study it, we don’t understand it.

I mean, let’s go back to the first principle here. Americans are living longer than they ever have. If you go back to the year 1900, every year, there’s been an absolute steady increase in life expectancy. For every two girls born today in the United States, one is going to live to be a hundred years old.

Not only are we living longer, but we’re feeling better. The rate of disability, that is, the inability to do things that we want to do is improving faster than the increase in longevity.

So, when we call something a disaster, let’s put it in the context of all the benefit that’s accruing from the great advances that we have in therapy today. Having recognized that we have a great amount of benefit, then let’s go back and say, we got some real problems here we got to fix. Many of these are problems that have been there forever. We’re just now getting to the advanced state of scientific knowledge and information and technology where we can, where we can address it.

I mean, who would have ever thought one of our greatest problems in the United States would be what to do with people between 80 and 90 years old who develop heart disease.

SUSAN DENTZER: If I asked you to fill in the blank: “This episode as a case study in …blank… what is it?

DR. CALIFF: The Vioxx episode is a case study in the need to change public policy based on the latest scientific evidence. SUSAN DENTZER: Meaning?

DR. CALIFF: Our understanding of therapeutics has changed to the extent that the old FDA system is no longer the system that we need. One of the things that many of us have enjoyed studying are the unintended consequences of new policies. Well-intentioned people make new policies and often, just like we give a drug to someone and we’re not sure what’s going to happen. We institute a new policy. Sometimes it does more harm than good. So, with all humility, I would put forth some ideas about what might be done. First of all, we now really are evolving rapidly in the United States to a two-tiered system of therapeutic approval, is the way I would say it. That is, the hurdle used to be get it through the FDA. If that happens, you’re on the market. It’s up to your sales force to convince a variety of payers to pay for the use of that treatment. Now, because of the aging of the population, CMS has become the dominant payer for health care in the United States. The combined cost of the Medicare and the Medicaid system dwarfs anything else that we have in terms of paying for health care. And so, we now have the second hurdle, which is, will Medicare pay for it. This used to be only an issue with devices, but as of the Medicare Modernization Act, it now becomes a huge issue for drugs. So, imagine that you’re working at Merck and your product is not approved for payment on the Medicare formulary system. You’re going to lose your job. Your job is to get it paid for. And so the question is now, how might Medicare use its power of being the dominant payer to not obstruct innovation, but to focus the industry on doing the right thing, to align the profit-seeking motive of industry with what the public health needs. Now, people like me believe the way Medicare can do that is to not mandate the studies that are going to be done for Medicare, just like I wouldn’t mandate the studies that are going to be done just from the FDA, because the FDA has a limited number of people. They don’t see all the issues.

I believe that we need more of a public forum that would say, out of all the things we could use Medicare money for are some allocation from the government to stimulate the right studies [break]. One of the critical [break] questions that need to be answered. And that’s not a matter of any individual to dictate or any small group to dictate. I believe that’s something we need to have a public discussion about.

SUSAN DENTZER: So, you would envision, in effect, some kind of consortium where the various payers, NIH, others could come together and, as you say, create this national problem list?

DR. CALIFF: Including patients. I mean, one of the great things that’s happened in combating the AIDS epidemic is the involvement of the community of people with AIDS in the design of the studies, what’s best for them. You might get a very different view than you would get from the company or from the academic people, who might be interested, questions that the patients don’t see as very relevant to their health.

In addition to this, we need to look at what Medicare is currently doing today. The last three devices to be approved for payment by Medicare have had a contingency that, along with approval, had to come some other form of study. So, in the case of cardiac defibrillators, a tremendous development–and believe me, I know it now, because someone very close to me just had sudden death here in the halls of Duke and was resuscitated successfully and she’s alive and well, you know, thanks to the ability to do this.

But it’s a very expensive technology. And so, the question is, how should it be applied? All the answers aren’t in. So, what Medicare said was, we’re going to reimburse for people that meet the criteria that were in the clinical trials, but you’ve got to collect a registry that is public information that allows us to really understand better how to apply the technology.

Another case is PET-scanning for Alzheimer’s disease. Many of us are afraid of Alzheimer’s disease. It’s nice to have the problem to live long enough to get it, but it’s a terrible problem if you get it. It might be useful to know ahead of time before you had symptoms that you had Alzheimer’s; it might not. Do you really want to know through a PET-scan that you’re going to start losing your memory two years from now. I don’t–that’s a debatable question.

A number of people came to Medicare and said, you need to pay for PET-scans for diagnosing Alzheimer’s disease. Then the answer to come back is, okay, well, pay for it, but only if you put–only if you volunteer to be in a clinical trial that will ultimately define whether the technology is useful.

So, that would–that’s the world that might be the future of drugs. It won’t be enough to get through the FDA. You’ve got to get over the hurdle of will Medicare pay for it and who’s in a better position to define the key questions than those that are putting out billions of dollars based on people coming, saying, you need to buy our treatment?

You also have to keep in mind, we’re in an era where scientific discovery is at an unparalleled rate. Medicare is being bombarded with new technologies and advocates who say this is the greatest thing since sliced bread. Sometimes, it is; sometimes, it isn’t. And just relying on judgment, without the data to make that decision, I think, is a bad policy. But that’s the current state.

SUSAN DENTZER: What is underway now is the filing of lots of lawsuits and there will be much litigation, in effect, a congressional investigation underway, mainly targeting the FDA and whether there were slipups at the FDA on this, as opposed to any discussion about creating a kind of systemic reform in the system that you described. Are we about to go do all the wrong things in response to the Vioxx crisis? Do you feel comfortable saying that?

DR. CALIFF: The greatest tragedy of the Vioxx crisis will be if we went on a witch hunt for individuals who may have been involved. What we need to do is redesign the system so that future drugs can be given safely to the American public. That’s what I really believe.

The poor people at the FDA–imagine that you’re on a job where you have very little in the way of resources. Every decision that you make is going to make somebody upset and people are a bit concerned about telling the public that we don’t really know if the balance in risk and benefit in the long term of most drugs that get on the market. And there are many people, including me, saying that now. That’s a little unsettling.

And the big picture that, on balance, we’re doing so much better with health than we ever did partly due to drugs, you know, it’s a risk worth taking, in my opinion. But if a drug gets on the market and it goes bad, they’re called on the carpet. If they take too long and they’re too cautious, they’re called on the carpet. It’s a no-win situation.

So, I think people that volunteer to work at the FDA and also for what they get paid, particularly if you look at the salaries of FDA physicians, compared to what they would make in practice, they’re sort of unsung heroes, in my view.

So, it’s good to examine what the FDA is doing. I’m all for that, but the sort of witch hunt mentality that somehow punishing people because of Vioxx is going to fix the problem, I think, it would be a tragic mistake.

In addition to that, we all know that multiple lawsuits is not going to make the system better. It’s just going to make people more cautious and that’s not going to be good for people who have serious diseases that need to be treated. I think if you talk to patients who have serious diseases, they’re willing to take a lot of risk if there is a chance of benefit. And we need to have a system that allows them to do that, if that’s what they want to do