Interview with Dr. Robert Temple

SUSAN DENTZER:

Let's start by talking about exactly what the FDA said in its most recent warning about this class of medications, and use of them for adolescents and children.

DR. ROBERT TEMPLE:

We said three main things. The first thing we said is that when you initiate therapy for depression, you need to be particularly careful because it's a risky time for people. They may get worse. It's not uncommon for people who are started on antidepressants, or who are not started on antidepressants, to get worse after their initial diagnosis, and it's crucial that the physician pay close attention to the patient to see if that's happening. So that's the main thing, observe.

We explicitly said we don't know that the drugs make this worse, but in any event, whether they do or not, you should be paying close attention. So I think that was the main thrust of the warning.

We also reminded people that … as a general matter, people with bipolar disease should not be put on these products alone, or antidepressants alone. There are other drugs to be taken with them, and they may be part of therapy, but they deserve particular care. Just putting somebody with bipolar disease on an antidepressant may be quite risky.

The third point we made is that all of these drugs cause certain other kinds of psychiatric things — anxiety, something called achethesia [ph], which is a sort of overall nervousness and movement disorder, and you need to watch carefully to see whether these patients are getting that, and it's getting worse, because if that happens, you may want to cut back on the dose, or stop the drug entirely.

We don't know whether that relates to certain acting out behavior for sure, but it's something to watch closely for.

So what it really says is you don't just drop a person on an antidepressant and lose them. You really pay close attention, and we know that, whether the drugs have a risk of increasing suicidality in some people or not. We don't know yet whether they do, but whether they do or not, it's a dangerous time.

SUSAN DENTZER:

Let's go back now and talk about why the FDA first became aware of these concerns. What happened?

DR. ROBERT TEMPLE:

Going back into the early '90s, there has always been a concern that some people given an antidepressant may get worse because of it. There were isolated reports of people who seemed to be that way, and it's always hard to interpret an isolated report because you never can know whether the drug did it or whether the person was going to get that way anyway. It's very hard to know.

So we have over the years, and other people have over the years, looked at the accumulated clinical studies in adults now of acute treatment of depression to see whether there was an excess of suicidal thinking, suicidal behavior, and actual suicides. These reports up to now have all said that there isn't, and we're in the late stages of yet another analysis of our data. Obviously, we'll let everybody know what that shows.

So we don't think that, at least for adults, there is an overall increase in suicide or suicidal behavior. Does that mean that these drugs couldn't make some people worse? No, it doesn't. They could. It might improve some people and make some people worse. But if you look at the overall results of many tens of thousands of patients now, you don't see an excess of suicide…

What you see when you look at the collected data, some call it pool data, is that there are not more suicides in the people who got the drug than in the people who didn't, there isn't more suicidal thinking in the people who got the drug than who didn't and so on, and that for adults, we think that's pretty solid.

Now coming back to children, in 2002 we received studies of the drug Paxil, or Paroxetine, in pediatric depression and some related diseases, and we noticed on reviewing them that there was this category of emotional lability — that's where certain kinds of adverse reactions were collected that seemed to have some suicidal thinking in them, and we asked for more details of that. We wrote to the companies and said we're not sure why you included suicidal behavior under the category emotional lability, and could you tell us more about all of these things, and the company responded by doing an analysis of suicidality.

SUSAN DENTZER:

Let me just stop you there. This was a phrase that the company used?

DR. ROBERT TEMPLE:

That's right. We did not tell them specifically to do that. That's how they chose to make sense out of it.

SUSAN DENTZER:

And what did you understand them to mean by that?

DR. ROBERT TEMPLE:

Well, that was the problem. Let me make it clear. When doctors report a lot of adverse effects, you have to collect them under some name or otherwise you'll have one of each and you won't be able to interpret it. So they put a wide variety of things which could include suicidal thinking, but could include, you know, fighting and things like that. They included them all under the head emotional lability.

But our reviewer noticed that included among those were some cases of apparent suicidal thinking, so we asked them to look into it further and they provided a response, and actually then we asked all the other companies with drugs of a similar class to do the same thing and to look into it. So that's the database that we're now looking at. This involves close to 20 studies in about 4,000 people — children and adolescents. There were no suicides in any of those patients. There were no actual suicides, but there were some episodes of suicidal thinking or maybe suicidal attempts.

SUSAN DENTZER:

What was actually said in these case reports that constituted suicidal thinking?

DR. ROBERT TEMPLE:

That's a terrific question because what we know is that these reports, or whatever happened, was classified by the company as suicidal thinking. Well, you don't know what they actually meant by that until you look at what the doctor reporting it said. And it could be anything from screaming at a parent, "I'm going to kill myself because I hate you," to a more serious statement, that "I've been contemplating suicide." And the implications of those things are different.

Similarly, suicidal behavior could be serious cutting of wrists or something like that, or standing at the edge of a building and threatening to jump.

SUSAN DENTZER:

Now, it's important to say that somewhere in here you all concluded that it wasn't just Paxil that needed to be looked at, so you asked a number of companies with a number of drugs to report similar data.

DR. ROBERT TEMPLE:

Once we got the results, our initial request went back to the people who make Paxil because that's where this emotional lability term came up, and in response to that, they produced a report that looked at possible suicidal behavior. Looking at that, we then asked all the other companies to do the same thing — to go back to their studies and look and do the same thing, and those reports came in, and we analyzed those, too.

Those are the results that were actually shown at the advisory committee which, in a numerical sense, seemed to show more of those events, whatever they mean, in the group that got the antidepressant than in the group that didn't. And it's very important to say that we have not concluded that this really means that there is an excess of suicidal thinking or behavior, but that these events need to be looked at closely to see what they are.

It remains possible that it does represent such an increase, but we need to find out before we tell people what it says.

SUSAN DENTZER:

So we have several layers of uncertainty here. We have first of all not consistent definitions of what was suicidal thinking across these studies. Then on top of that we have some studies of the same drug showing differences and some not.

DR. ROBERT TEMPLE:

Well, that's why we're uncertain. We don't know for sure that when we get a better fix on what these events on, whether it will all become clear or not. But we just feel that's what you have to do to start. Counting these cutting episodes or slapping your head as suicidal thinking and making conclusions about that doesn't seem sensible to us. We want people who are independent minded to look at these things and decide what their view is as to whether they're really suicide related or not, and we just don't think we can do much until that's done.

SUSAN DENTZER:

Now, in the meantime across the Atlantic, the British are looking at similar data. Let's talk about what they have done. The British have essentially now concluded, or at least told providers that the only drug that should be prescribed for children and adolescents of this class of medications is Prozac. Why are the British coming up with such a different conclusion?

DR. ROBERT TEMPLE:

I'm not sure entirely why they reached a different conclusion, but we agree on a number of things. The only one of these drugs that has been shown to work in children is Prozac. They have two properly controlled studies, and they both show that the drug works in children, and it's approved in this country, just as it's approved there, for that use.

So my advice to someone who is thinking about starting an antidepressant in children is it makes a lot of sense to go with the drug that's been shown to work. But we're not prepared to conclude that the other drugs don't work. One of the facts of life of any depressant trials is the drugs that work don't always show that they work. In adults, we know of only about half of the studies of drugs that definitely work are able to distinguish the drug from placebo. That's partly because people get better spontaneously, because the environment is nurturing and it improves people. We don't really know, but that's the way it is.

So the fact that so far most of the studies in children have not shown effectiveness and there are a couple of others that did, but only a couple, doesn't really prove these drugs don't work. What you saw was huge improvement in the placebo group, and maybe that kept the drug from showing anything.

So we haven't really concluded that you shouldn't use these drugs, but we do point out in our warning statement that the only one that's been shown to work is Prozac. But we're not prepared to say the others are contra-indicated. Drugs are all different. Prozac has particular side effects, another drug might not have the same side effect. We don't think you can get away with necessarily limiting it to one drug. We would obviously like to see data that show these work, and we don't have data that show they work. But we didn't feel we should tell people, don't you even think about using them.

SUSAN DENTZER:

Many people are crediting the British now for having divulged the information that there were all of these negative studies involving these drugs; that that information had not been published previously either by the companies, or have obviously been made available by the FDA. Do the British owe a vote of thanks from people for outing the notion of these negative studies?

DR. ROBERT TEMPLE:

The question of what people should produce when they do a study is an important one, and I want to distinguish between what we can do as a federal agency and say what my opinion is. We don't have the capacity — we don't think we have the capacity anyway — to force people to publish studies. We make the results of our review available and that would include any studies that were successful or that failed. Also, when we grant a period of pediatric exclusivity to a drug for doing these studies, the results of those studies become known at least in an abbreviated way.

My own view is that the results of human studies ought to be made known to people. It's a privilege to study human beings, and I think the results should be available, however they come out. But we don't have anything in our law that says that's a requirement.

So I think it's fine that this question was raised by the British. We, of course, caused the data to be collected in the first place, so we certainly were going to be looking at all those data and the results would become known eventually anyway, but I don't think it's a bad thing that they put the discussion out.

As I said, we showed the results of the British data at the open advisory committee, so we certainly made use of the fact that they had revealed the data.

SUSAN DENTZER:

The FDA does have greater authority now, as I understand it, under new legislation to make this kind of negative data public; correct?

DR. ROBERT TEMPLE:

Companies can get patent extension by doing the pediatric studies we ask them to do. When they get that extension, then the results of the studies they do are made public. So the results of these studies would have gradually become public, at least in abbreviated form.

SUSAN DENTZER:

So eventually the information would have come out.

DR. ROBERT TEMPLE:

Some or all of it would have come out.

SUSAN DENTZER:

So back to the question of what might happen, as you get (a second analysis of antidepressant data) back from Columbia, what is the timetable for this assessment by the Columbia experts, and what do you expect would come forward after that?

DR. ROBERT TEMPLE:

Well, we expect to get the results in the next several months, or we say sometime during the summer, and we're optimistic about getting them. Once we have the events classified, we can do the same analyses that people have been doing up to now; namely, look and see whether there are more of these suicidal events in the people who get treated than in the people who don't get treated, and we're interested in the question overall for all of the drugs. But we're also interested in the question drug by drug.

There was no hint of this problem in the Prozac data, so is that a true difference, is it just an accident? We don't know yet. So we're going to be looking at the results drug by drug, as well as overall, and once we have them classified properly, we should be able to do that. We will definitely go back to the advisory committee on that question and show the results of the studies.

SUSAN DENTZER:

At [the advisory committee meeting] there were a number of parents who got up and described in heart-wrenching detail what had happened to their children who had killed themselves. … What did you make of those things that those parents were saying?

DR. ROBERT TEMPLE:

Well, I was horrified by how bad the things that they described were, and how much pain they must have induced. The trouble with any of those is there's no way to know from a report like that whether the drug caused it, or whether it was part of the person's underlying disease. You know, these kinds of reports are sometimes treated dismissively as anecdotes. I don't think that's entirely fair. They're real enough. The trouble with an event is that you can't know without more information, or perhaps without a controlled study whether it's what would have happened anyway, or whether it happened because of the drug.

There were suicides before there were antidepressants, and we know that over the course of a person's life, if they have depression, there's a very substantial chance that they will at least try to take their own lives. I mean, that's been true all along, and it's true now. So you just can't tell.

It's tempting to believe that if a person didn't commit suicide and then got the drug and did, that it must be the drug that did it. But you can't know that. For one thing, they weren't being given an antidepressant before, and now something triggered in a physician or in a patient a desire to go on an antidepressant. Maybe they're getting worse. You just, in most cases at least, you just can't know. And that's the trouble. That's why we're looking to the control trials to give an answer.

I'm sure that's frustrating because the cases seem so real. My child was okay, and then my child was not okay. But parents don't always know everything that's going on in a child's mind. It's just very clear that people commit suicide in unexpected ways sometimes, and they don't necessarily tell people they're going to do it. They don't necessarily describe their troubles. So it's very hard to learn about causation from those kinds of reports. That doesn't mean they're not very moving, because they are.

SUSAN DENTZER:

These parents are utterly convinced in many cases that the drugs caused their children to die.

DR. ROBERT TEMPLE:

People can be convinced of things that aren't true. There have been a wide variety of remedies that were promoted as beneficial that have turned on reflection and on the conduct of good studies not to be beneficial. You can't always know. People get better and get worse on their own schedule sometimes, and you just can't really know what was going to happen. That's just unfortunate. It would be a lot easier if you could, but you can't usually.

Sometimes when a drug is taken away, put back, taken away, put back, that's a kind of study in a way. Sometimes you can learn about whether the drug really did it, and once in a while you have that kind of information but not usually. Most of the time if people are doing badly, no one wants to restart the drug when they're better.

But it's true, people are convinced that the drug did it. People are equally convinced that the drug made them better sometimes, and that conviction isn't true necessarily either. The way you know almost always is by doing a controlled study, which is why we're looking at the 25 controlled studies of the antidepressants for any hint that there is an increase in suicidal thinking.

SUSAN DENTZER:

Let's make the point that the Columbia folks (studying the drugs) are making — apparently, that there has been, apparently, a 25 percent reduction in suicides among the pediatric population attendant this period of use of the SSRIs. So what is it that they're saying or statistically that they believe is happening?

DR. ROBERT TEMPLE:

We saw some data at the advisory committee — this is epidemiologic data, this isn't a control trial — suggesting that the rate of adolescent and pediatric suicides is actually declining over the last 10 years or so, roughly a 10 percent (decrease).

David Shaffer, a psychiatrist from Columbia, presented data at the advisory committee that looked at adolescent suicide in the United States, and what he found over about a 10 year period was about a 25 percent reduction. Now, these aren't control trials, and you've got to be careful with epidemiology. There may be other reasons for this, but it certainly suggests that there may be a benefit from treatment, because the last 10 years is the period in which there's been much more treatment of adolescent depression, and it certainly doesn't suggest that there is an increase in suicidal activities.

You always have to be careful — what I think it really reminds you, or should remind us of is that depression is a potentially fatal disease. It's very important to get the right answer here, and we don't want to discourage use of drugs that may be life-saving more than there's reason to do so. And similarly, we don't want to ignore the possible adverse effect. I mean, we've got to come to grips with both of these. But we think it's just terribly important that on a matter like this we come as close to being right as we can.

SUSAN DENTZER:

And let's make that point as well. Since the FDA has issued this warning, we're told by physicians that parents now are very fearful if they have not previously had their children on medication, or if they themselves have not been on medication, they're extremely reluctant now when they are asked to consider putting their kids on medication. Has FDA overreacted here in such a way that it may deter people from being on medication who should be on it?

DR. ROBERT TEMPLE:

Well, believe me, we have been extremely worried about the effects of our actions. It's hard to not say anything when the British are making announcements, and when we're seeing these data that suggest there might be a problem, and when we see that the drugs don't seem to be able to show that they work very well. So saying nothing didn't really seem like an option to us.

We tried in our warning to make it as clear as we could that it's the situation that deserves close monitoring, not only the drugs. Has everybody understood that? The answer is no. I've seen published reports in newspapers that didn't seem to understand that either, that it said oh, no, we're declaring the drugs cause suicide, which is absolutely not what we did. We tried to say we're looking, we don't know, and that you need to be very careful here.

It could be that the drugs actually cause some increase in suicide when you use them early, but they might even have a long term benefit. It's very hard to know those things. Nobody is going to let you do a placebo control long-term trial to see if there are more suicides in one group than another because that would involve leaving people who are grossly depressed off therapy. I don't think anybody would do such a trial…

It's hard to imagine, given what we know about the likelihood of suicide in people who are depressed, it's hard to imagine that treating their depression isn't at least in many patients going to improve the likelihood that they will not commit suicide. But getting rigorous data on that is extremely difficult, so we are very concerned that people will read what we've said and say, "I'm never going to use these drugs again." We don't think that's a good public health outcome at all.

We tried to be very careful in our warning statement to make it clear that we don't know whether the drugs are responsible for this, and that you owe these people a close look while you're treating them. That's the message we'd like to convey, not that the drug shouldn't be used anymore. Certainly one would hardly apply that to Prozac which has been shown to be useful. But we are worried about that.

The problem goes in both directions. You wouldn't want to underestimate the risk that the drugs have, if they have one, but you wouldn't want to overstate the risks either, because there's a lot of reason to think that treating pediatric depression is important.