SUSAN DENTZER: Let's start by talking about what Vioxx is as a drug.
DR. JANET WOODCOCK: Vioxx is one of the anti-inflammatory drugs. We've had a number of those on the market for many years. The nonsteroidal anti-inflammatory drugs or NSAIDs. The NSAIDs as a class are somewhat toxic. They're some of the more toxic drugs for chronic conditions that we have on the market, and mainly they cause gastrointestinal problems, bleeding, perforations of the GI tract and so forth.
Vioxx was one of a new kind of NSAID called the Cox-2 inhibitors. These were hoped to be more selective toward the inflammatory process and spare the gastrointestinal tract and thus have a lower rate of gastrointestinal complications.
But when we approved Vioxx, FDA did not give it that claim for improved gastrointestinal safety because it hadn't been proven.
SUSAN DENTZER: As you said, these drugs are very toxic. Yet there are many of these drugs on the market available on an over-the-counter basis. Let's just talk a little bit about this broad family and what some of the obvious more familiar members of this family are.
DR. JANET WOODCOCK: Right. There are members of the NSAID class over-the-counter, and they're indicated for short-term use, for pain usually, or fever, things like ibuprofen or naproxen. These are not that toxic, and they're fairly safe when they're used according to the over-the-counter label for short-term duration at relatively low doses.
When these drugs are used chronically though, as they often are to treat arthritis symptoms, you get into the problems of gastrointestinal toxicity.
SUSAN DENTZER: And specifically that can mean?
DR. JANET WOODCOCK: Specifically you can have bleeding, you can have perforation of stomach ulcers, you can have obstruction of the gastrointestinal tract, and there are quite a number, probably thousands of people who die every year as a result of this toxicity. The estimate of hospital days is in the hundreds of thousands of hospital days that Americans spend as a result of toxicity from this drug class.
SUSAN DENTZER: So the Cox-2 inhibitors come out. It's believed, by some anyway, that they have this less toxic effect in terms of gastrointestinal effects, but FDA did not approve it on those grounds.
DR. JANET WOODCOCK: FDA did not approve the Cox-2 inhibitors, at the time of their first approval, as with a claim of improved gastrointestinal safety. And the reason was we required proof of that in clinical trials head to head with other NSAIDs of the class to show that these drugs would be safer on bleeding, perforations, obstruction and so forth. That required long-term trials to be done to look at the gastrointestinal safety.
SUSAN DENTZER: And at the time then of approval of Vioxx, similarly with Celebrex, that data did not exist?
DR. JANET WOODCOCK: That's correct. Both of those studies were done after the drugs were approved.
SUSAN DENTZER: And it was because of this that Merck undertook the study known as the VIGOR study
DR. JANET WOODCOCK: Merck did the VIGOR trial in order to prove that Vioxx was safer on the GI tract than naproxen, which was the comparison that was in the trial. And they did a year long trial, looking at bleeding, perforations and obstructions, and other GI complications. And that's because FDA required that to be done.
That trial was successful in that Vioxx was shown to be safer for the GI tract than Naproxen. Unfortunately, the dose of Vioxx that was used in the trial, which was twice the usual dose, was more toxic apparently on the cardiovascular system, and there were more cardiovascular events observed in the Vioxx treated people.
SUSAN DENTZER: And some of these events were serious events.
DR. JANET WOODCOCK: Yes, such as myocardial infarction or heart attack.
SUSAN DENTZER: How great was the magnitude of the risk?
DR. JANET WOODCOCK: It was several fold over the naproxen arm. So the folks taking Vioxx had perhaps two to three times amount of cardiovascular events. However, it isn't known--at the time at least--whether the naproxen was protective to the cardiovascular system or this was a toxicity of Vioxx. It also wasn't know whether the dose of Vioxx, which was twice the normal dose, was partly responsible for this.
SUSAN DENTZER: What was the theory behind the believe that maybe naproxen was protecting the heart moreso than Vioxx was damaging it?
DR. JANET WOODCOCK: Well, aspirin has effects similar to the nonsteroidal anti-inflammatory agents, and aspirin is also protective ont cardiovascular system, and people are supposed to take, as you know, low-dose aspirin if they're at high risk for a heart attack. Therefore, some people felt that naproxen could be having a similar kind of protective effect on the cardiovascular system.
SUSAN DENTZER: Was there any evidence for that or was it just a theory?
DR. JANET WOODCOCK: There was very little evidence.
SUSAN DENTZER: And FDA, how did FDA respond to that argument, that maybe naproxen was protecting the heart?
DR. JANET WOODCOCK: Well, FDA had a public advisory committee to discuss this trial, and the cardiovascular events were thoroughly discussed by cardiologists. The bleeding events were discussed by gastroenterologists, specialists in the GI tract. And it was decided that we didn't know, we didn't know what was the cause of this observation in the VIGOR trial, and it was decided what should be done is put information in the label about these results, as well as some precautions about use in people who are risk for cardiovascular events.
SUSAN DENTZER: And in fact at one point a big argument broke out among the gastroenterologists on the committee versus cardiologists. Can you describe what happened?
DR. JANET WOODCOCK: Well, the cardiologists felt that this was a serious finding in the cardiovascular system. The gastroenterologists, who have the experience with dealing with all the patients who have bleeding and perforations and so forth from NSAID induced toxicity said that's a very serious toxicity too. We have many deaths in this country from that. And so, depending on your perspective, there was a different interpretation of the severity of these two kinds of toxicities.
SUSAN DENTZER: And didn't they argue with each other about what was the worst kind of--a worse death to suffer?
DR. JANET WOODCOCK: They argued about which was a worse outcome to have, absolutely, to have a GI toxicity or a cardiovascular event. The cardiologists and the gastroenterologists who were at our advisory committee had a discussion or dispute over the relative severity of having a problem in your gastrointestinal tract, bleeding or perforation, versus suffering a hear attack.
SUSAN DENTZER: Which was the worse death?
DR. JANET WOODCOCK: Which was the worst, well, worse we'll say, yes.
SUSAN DENTZER: So the agreement is made, or at least it's decided within the committee that the thing to do--and the FDA concurs--is to put this warning on. How did Merck respond to that?
DR. JANET WOODCOCK: Well, we negotiated with Merck, and the label was changed, and warning were put--or precautions were put on the label, and the VIGOR trial results were described in the drug label. Merck at that time really believed that the naproxen was protective, and that was the explanation. The FDA, as usual, remained skeptical of any explanation until we had definitive data.
SUSAN DENTZER: And in fact, Merck was subsequently found to have [underplayed the risks of Vioxx], correct?
DR. JANET WOODCOCK: Well, FDA often will send warning letters out if we feel that advertising isn't balanced, and balance includes presenting the risks in an appropriate light, and so we will warn companies that they need to make risks more prominent if we think they're not being highlighted adequately.
SUSAN DENTZER: So what did the FDA say in this warning letter?
DR. JANET WOODCOCK: FDA said that the risks of the drug were not adequately highlighted given everything we knew so far about the cardiovascular risks.
SUSAN DENTZER: And consequently Merck then subsequently amended its advertising?
DR. JANET WOODCOCK: Yes. Merck had to amend its advertising to make it more balanced.
SUSAN DENTZER: So that happens. The label change occurs in 2002. Meanwhile, some other studies of a different nature, not randomized controlled trials, but observational studies and others, were starting to yield additional evidence about the cardiovascular risks. Let's just highlight some of them that came in.
DR. JANET WOODCOCK: Certainly. Well, we had data from a number of sources. Merck had submitted additional controlled trials such as in Alzheimer's disease, which did not show an increased risk. At the same time, people are doing studies out in the community, looking at population databases, in other words, studies in health care systems and seeing what actually happened to people who are taking different kinds of NSAID drugs. Some of those studies did not show an increased risk for Vioxx compared to the other NSAIDs. Some of them had a hint of an increased risk.
This wasn't definitive enough, and there were studies going either way to--for us to decide that there was a definitive answer about the risk of Vioxx.
SUSAN DENTZER: And meanwhile FDA was aware that Merck was going on to take another series of trials or a trial which was approved. Let's talk about that.
DR. JANET WOODCOCK: That's correct. Actually, Merck had ongoing at the time of the advisory committee, the approved trial, which is a trial that ultimately showed that 25 milligrams of Vioxx did have an increased cardiovascular risk after a considerable amount of time. That trial was a trial to try to prevent colon polyps from forming, so it was a prevention trial in normal individuals and it was against placebo, which gives you the best chance of finding something, if you're studying it, against a placebo.
So after 18 months on that trial, the patients taking Vioxx began to have a higher rate of cardiovascular events than the placebo patients. And Merck's data monitoring committee that was looking after the safety of that trial, eventually recommended the trial be stopped because of this increased rate. That was the first time we really had definitive evidence that 25 milligrams of Vioxx could cause excess cardiovascular events.
SUSAN DENTZER: And that is to say a normal dose of Vioxx?
DR. JANET WOODCOCK: That's correct.
SUSAN DENTZER: FDA had worked with Merck specifically to set up a portion of APPROVe so that it would look precisely at this question. If you could say that, and then let's talk about how that was done.
DR. JANET WOODCOCK: Yes. FDA had worked with Merck to make sure that this question of cardiovascular safety would be addressed in the approved trial because this trial, with its design against placebo, and having a fairly large number of patients, really had the best chance of answering this question about cardiovascular risk.
So what FDA did was interact with Merck during this process in how they set up their monitoring and safety checking during the trial to make sure that these events would be looked into very carefully.
SUSAN DENTZER: And Merck then, of course, went through its own internal process and consulted with a number of individuals about what to do. When Merck officials phoned up and said, "We want to come talk to the FDA because we're going to pull this drug off the market," what was the FDA's response?
DR. JANET WOODCOCK: Well, we were surprised, and we wanted to see the data right away, of course, which we did. We concurred with their decision. The APPROVe results showed that after more than 18 months of continuous treatment, about 1.9 percent of the control group, the placebo group, had either heart attack or stroke, a cardiovascular event, versus 3.5 percent of the people continuously taking Vioxx.
SUSAN DENTZER: So that was almost double the risk.
DR. JANET WOODCOCK: That's correct.
SUSAN DENTZER: In a case like that, is the evidence just airtight that there's a greatly elevated risk?
DR. JANET WOODCOCK: Well, you look at the statistics and the p-values and the probability this could happen by chance. Even something like this could happen by chance alone. However, the trends that we saw during the study, after 18 months, and the level of increase of the risk, it's very likely that this was an effect of the drug.
SUSAN DENTZER: We still don't know, do we, what the mechanism was that was causing this effect, correct?
DR. JANET WOODCOCK: That's correct. We still don't know what the mechanism of this adverse event is, and there's been a lot of speculation about it. It's felt that the Cox-2 selective drugs as a group may have some properties that predispose them to cardiovascular problems. It's also known that Vioxx in particular raises blood pressure and causes fluid retention, and that could contribute to a finding like this over the long term.
SUSAN DENTZER: And indeed, some people pointed to that early on, and some people said that Merck knew that quite early on and should have been more--should have had its antenna up sooner looking for effects just because of that.
DR. JANET WOODCOCK: Well, we knew from the VIGOR trial and other trials that Merck had done that there was increases in blood pressure and fluid retention that occurred with Vioxx, but the APPROVe trial was monitored very carefully during those 18 months, and no difference was found between the patient groups.
SUSAN DENTZER: And we, as we were saying, we have no explanation as to why this effect crops up after 18 months?
DR. JANET WOODCOCK: Yes, that's very unusual, but very few drugs are studied like this for a long term against placebo, and there's probably a lot we don't know.
SUSAN DENTZER: Now, Dr. Eric Topol, among others, have said that everybody here was derelict, that Merck was derelict and particularly that FDA failed to follow up on evidence and just waited passively, Dr. Topol says, for the evidence to roll in without doing anything about it.
DR. JANET WOODCOCK: Well, we don't agree with that assessment. We knew the APPROVe trial was going on. In the meantime we made sure that this information from the VIGOR trial was in the label of the drug. The additional studies from the population base that occurred subsequent to the VIGOR trial were not definitive. And so we felt we were right where the evidence was as far as what we knew about Vioxx at every point in time.
SUSAN DENTZER: Meanwhile, just this past summer, an abstract was published or shown at a conference with data that had been compiled by an FDA official, suggesting again higher cardiovascular risk. Let's talk about that.
DR. JANET WOODCOCK: This summer an abstract was presented in Europe by one of the FDA researchers about another population based study of Vioxx. We had started this in 2001 after the VIGOR trial because we were interested in additional data sources to help us understand this.
This result did not show anything new about Vioxx. What we saw in that was a very small number of people on 50 milligrams of Vioxx, twice the usual dose, had an increased risk of cardiovascular events. That's what we saw in the VIGOR trial. The chronically approved dose of Vioxx, 25 milligrams of Vioxx, did not show a really increased risk over the other NSAIDs that were looked at in that study except for celecoxib, and we really don't know what to think about that.
The 25 milligram recommended chronic dose of Vioxx did not show an increased risk over other NSAIDs in that epidemiologic study.
SUSAN DENTZER: So that in and of itself would not have set off any alarm bells?
DR. JANET WOODCOCK: That's correct.
SUSAN DENTZER: The higher dose did show this, but you're saying you already knew that from the VIGOR study.
DR. JANET WOODCOCK: We knew that from the VIGOR study and some other evidence that had come out.
SUSAN DENTZER: So in the end all that did was raise questions about the appropriate level of the dose.
DR. JANET WOODCOCK: Right. And we knew and had in the label that the 50 milligram dose should not--was not recommended to be used chronically because of the results in the VIGOR trial.
SUSAN DENTZER: There has been some, some suggestion that that data was somehow quashed, or that this researcher was not able to, to get a full airing of this study. What are the facts behind that situation?
DR. JANET WOODCOCK: Well, the facts are that was presented in France as an abstract. An abstract was also submitted by the researchers to the American College of Rheumatology back in May, and that those results were also presented there several weeks ago. So the data were put out into the public. The FDA has not had a chance really to look at that study in depth and see how the analyses were done, and we're conducting that review right now.
SUSAN DENTZER: If the FDA was not derelict here, was Merck derelict? Should Merck have done more sooner to establish the actual evidence on this question?
DR. JANET WOODCOCK: Well, Merck was conducting the approved trial which actually did give us an answer to this question eventually. It's difficult to imagine the kind of trial one could conduct after the VIGOR trial to really give more information about this. It's difficult to do a placebo controlled trial in arthritis patients because they're suffering from arthritis, and they can't go for several years without treatment while you're waiting to see the answer to a toxicity question.
On the other hand, to do a study in people at higher risk for heart disease, which in fact might have shown the answer quicker, has--you have to think that through, because we already put in the label a caution about using this drug in people with pre-existing hear disease because of the results of the VIGOR trial.
So the question is, what kind of trial could find out the answer in a speedy fashion and give results that were unequivocal? And the APPROVe trial did give us those results.
SUSAN DENTZER: And again, because it was a prevention trial it didn't raise some of these issues that arise about treating people who are already sick?
DR. JANET WOODCOCK: That's correct. It was able to be used in healthy people, and could have a placebo controlled trial and you wouldn't have a lot of dropouts. And if you're an arthritis patient and you're taking a placebo, you probably aren't going to stay in the study very long if you actually need medicine for your arthritis.
SUSAN DENTZER: Is this a way of saying then that when it comes to this whole question of studying drugs, you're taking into account all kinds of things, the ethics, the cost? How would you phrase that?
DR. JANET WOODCOCK: Well, it is very complicated to find out these safety questions about drugs, especially questions that relate to the long-term use of drugs. We know that from the Women's Health Initiative Study that was done that showed that estrogens that were prescribed in so many postmenopausal women actually had serious side effects. That study took a lot of investment and a lot of time to get those answers, but now we know some answers about the use of estrogen in postmenopausal women. That's true for many of the drugs that we take chronically every day.
SUSAN DENTZER: If this is not a case study in the FDA being derelict or a case study in the drug company being derelict, what is it a case study about? What is this story about?
DR. JANET WOODCOCK: I think the Vioxx story is an illustration of the fact that there's a lot of uncertainty about the outcomes of the use of medicines, especially the long-term use of medications. And as more and more Americans are taking more and more medicines, and actually we know it's improving people's lives, but we have to look carefully at the side effects and especially the long-term consequences of taking these medicines.
SUSAN DENTZER: How would we do that then, given that so many drugs now are being expressly developed for the chronically ill market with the expectation that they'll take them for many years?
DR. JANET WOODCOCK: I think we all have the hope that as we put into place the new electronic medical records and get the health care system more or less online, that we will be able to do more research on drugs after they're approved and learn more in the real use, real world situation out there about how--what's happening to people as they take drugs and what are the longer-term consequences, because the other mechanisms we have to learn about these longer-term outcomes are really not very useful or applicable.
SUSAN DENTZER: But that would really just be a way of capturing more population based data, which is useful, but can often not be definitive. So what do we need to do beyond that?
DR. JANET WOODCOCK: Yes. Well, we may need to really consider doing trials, more trials in people who are taking drugs after the drugs have been marketed, trials in the real world situation to find out answers.
These are almost societal questions now about large pop--parts of our population are taking medicines for a long time.
SUSAN DENTZER: So what we would be thinking then is really big, long-term studies that would be of the nature of APPROVe, randomized placebo controlled studies that would go on for 3 or 4 years, correct? Do we need to study all manner of outcomes of taking drugs, and if so, what kinds of--what are the outcomes we think we have to look at?
DR. JANET WOODCOCK: Well, the outcomes we're interested in are people's health improved? And in general, are we not causing more harm than we expected? You know that drugs have side effects, all drugs have side effects, and we expect that those occur. But the unexpected side effect such as a cardiovascular risk from Vioxx or from estrogens, that's something we need to know about so we can factor that into the equation.
SUSAN DENTZER: And how do we set up trials to capture that kind of information?
DR. JANET WOODCOCK: Well, the, the community that studies these things has a lot of suggestions. You need to do everything from randomized controlled trials all the way to observational studies that can answer some of these questions.
SUSAN DENTZER: So really a whole array of studies?
DR. JANET WOODCOCK: Yes. Or, or the tool kit that you can use, and depending on what question you need to have answered, you can pick a kind of study and it will shed light on that question.
SUSAN DENTZER: The abstract that we talked about earlier that was done by the FDA researcher, on the basis of the data that emerged from that, he made some estimates about the likely number of deaths that would have occurred as well as excess heart attacks and strokes from the broad population that was taking Vioxx over these last four or more years. Those are very, very large numbers that emerged, indicating a lot of people died from taking this drug.
DR. JANET WOODCOCK: Well, the estimates of number of excess heart attacks or perhaps even deaths from Vioxx are derived from a model that incorporated a very large number of assumptions. One of those assumptions in fact was that Celebrex has a lower risk than not taking any NSAID. In other words, that Celebrex is protective in some way. We have absolutely no evidence that that's the case.
So we can't really extrapolate hard numbers of how many people had heart attacks or strokes in excess from taking Vioxx from that kind of information. You might be able to do some calculation from the randomized controlled data from the approved trial. That might give you some indication.
SUSAN DENTZER: And indeed, people have done that, and just on a kind of a back of the envelope basis have said, well, clearly thousands of people died here.
DR. JANET WOODCOCK: Yes. Well, we know that thousands of people die every year from the gastrointestinal effects of the nonsteroidal anti-inflammatory agents, and the whole goal of developing the Cox-2 inhibitors was to try and mitigate some of that mortality and a tremendous amount of hospitalizations and so forth that result from GI bleeding and so forth from this class of drugs. So perhaps we were trading in this case one toxicity for another, and that's something we learned in the APPROVe trial.
But we can't take the attitude that if you weren't taking Vioxx you would not have had any adverse events happen to you, unless you actually didn't need a nonsteroidal anti-inflammatory drug.
SUSAN DENTZER: Does that mean that we'll never know the answer about how many people actually died from taking this drug?
DR. JANET WOODCOCK: I think that would be a very difficult figure to come up with.
SUSAN DENTZER: Except to say that it's reasonable to assume some meaningful number did?
DR. JANET WOODCOCK: It's meaning--it's reasonable to assume that some people died from taking this drug, but it's also known fact, say in the late '90s it was extrapolated that maybe 15,000 people a year might die from GI toxicity from the NSAID class of drugs.
SUSAN DENTZER: What about the other Cox-2 inhibitor drugs on the market, Bextra, Celebrex, and then of course Merck has in its pipeline and hopes to get approval in the U.S. for--or hoped at least to this point to get approval in the U.S. for Arcoxia. What is the FDA's attitude now toward those other Cox-2 inhibitors?
DR. JANET WOODCOCK: Well, of course there's quite a concern about whether there's a class effect of the Cox-2 inhibitors, and what the means is if Vioxx had this cardiovascular toxicity, do the other selective Cox-2 inhibitors also share that toxicity? And so of course we're looking at this very closely.
As far as Celebrex, we have not seen any signal, either in the controlled clinical trial data or the population based data that would make us think that it has excess cardiovascular toxicity over other drugs in the class. That doesn't mean that we're certain. That means that's what we know so far.
In addition, Celebrex has several ongoing placebo controlled trials that have data safety monitoring boards, and those boards have not stopped those trials. And we have inquired into that, and they are continuing those trials. So we know that that kind of data are being developed.
In addition, Pfizer has said they will conduct now a trial in high risk patients with Celebrex to get even greater certainty about any excess cardiovascular risk. So that's the story on Celebrex.
On Bextra, less is known. There was not a signal of cardiovascular toxicity in the pre-market database for Bextra. However, there also was not a signal in the pre-market database for Vioxx. All right? Bextra has not been observed for a long period of time, in particular does not have long-term placebo controlled trials. So that one we'll have to look for other sources of additional information.
SUSAN DENTZER: Will FDA require Pfizer to do some of those controlled trials for Bextra?
DR. JANET WOODCOCK: Well, we, we will have that under evaluation. We also, obviously, look very closely at any new Cox-2 inhibitors that are coming before us for market approval.
SUSAN DENTZER: The case in point being Arcoxia?
DR. JANET WOODCOCK: Being any, any of the Cox-2 inhibitors that people might apply to market.
SUSAN DENTZER: Just back on Bextra, you said--I think you said we will, we will be evaluating what to do. Is FDA currently evaluating what to do?
DR. JANET WOODCOCK: Certainly. I guess I should say we will be deciding what to do. We are evaluating all the available data on Bextra and potential sources of additional information.
SUSAN DENTZER: What would be the arguments, if any, the plausible arguments, against compelling the company to do these studies? Doesn't that seem like the obvious approach at this point, that FDA should require--
DR. JANET WOODCOCK: FDA doesn't have the authority to require additional studies once a drug is on the market. Our authority is limited to either taking a drug off the market or perhaps having the bully pulpit and saying these studies should be done, or something like that. But we do not have the authority to simply order additional studies to be done.
SUSAN DENTZER: So what about the bully pulpit option?
DR. JANET WOODCOCK: We might do that in a case where we feel that's what--that's a final option. Normally we will negotiate with companies over studies and get them to do what studies we think might need to be done.
SUSAN DENTZER: And just finally, why couldn't FDA require Merck to have done more studies at the point at which this evidence first became clear from the VIGOR, from the VIGOR study that there was this higher cardiovascular risk?
DR. JANET WOODCOCK: Well, as it turns out, we didn't have to require Merck to do additional studies because we knew they were ongoing. However, we did not have the authority to simply require Merck or any other company to do additional studies if the drug is already on the market. Our authority is limited to starting an administrative procedure to take a drug off the market. In addition we might use the bully pulpit, so to speak, to get a study done by, by saying this really needs to be done. We worked for many years on the, on the phenylpropanolamine, for example, and the companies, the over-the-counter companies banded together and did a study on phenylpropanolamine and stroke that eventually resulted in that drug being taken off the market.
SUSAN DENTZER: And FDA would also have the ability to get up and say, this is a dangerous drug; don't take it, correct?
DR. JANET WOODCOCK: FDA also has the ability to stand up and say we believe this is a dangerous drug. We recommend that people not take this drug. We can do that.
SUSAN DENTZER: Why was that not done in the case of Vioxx?
DR. JANET WOODCOCK: We felt, and continue to feel, that the actions that were taken, putting the warnings in the label, having a public advisory committee discussion of the issues, were what was warranted at that time.
Vioxx was the only drug in the Cox-2 class that had been demonstrated to reduce GI toxicity, bleeding, perforations, obstructions and so forth, and so it had a benefit that reduced the risk of one toxicity, and we had uncertainty about how much it increased the risk of cardiovascular disorders.
SUSAN DENTZER: So you're saying it, it in a way deserved to be on the market because it was the only drug that people could take for arthritis and other--for an anti-inflammatory purpose where they knew there wasn't a higher risk of these gastrointestinal events.
DR. JANET WOODCOCK: It was the only drug that had--was shown to have a lower risk than another NSAID, naproxen, of causing these gastrointestinal events, and particularly for people who had already had a GI bleed or had known stomach problems, it was a reasonable alternative.