FOOD AND DRUG ADMINISTRATION

















                          JOINT MEETING OF THE




                                AND THE














                      Tuesday, September 14, 2004


                               7:56 a.m.










                          Holiday Inn Bethesda

                         8120 Wisconsin Avenue

                           Bethesda, Maryland





      Wayne Goodman, M.D., Chair

      Anuja M. Patel, M.P.H., Executive Secretary




      James J. McGough, M.D.

      Jean E. Bronstein, R.N., M.S. (Consumer Rep)

      Philip S. Wang, M.D. M.P.H., Dr. P.H.

      Dilip J. Mehta, M.D., Ph.D., (Industry Rep)

      Lauren Marangell, M.D.

      Delbert G. Robinson, M.D.

      Daniel S. Pine, M.D.

      Barbara G. Wells, Pharm. D.

      Bruce G. Pollock, M.D., Ph.D.




      P. Joan Chesney, M.D.

      Deborah L. Dokken, M.P.A.

      Michael E. Fant, M.D., Ph.D.

      Richard L. Gorman, M.D.

      Robert M. Nelson, M.D., Ph.D.

      Thomas B. Newman, M.D., M.P.H.

      Judith R. O'Fallon, Ph.D.

      Victor M. Santana, M.D.




      Norman Fost, M.D., M.P.H.

      Charles E. Irwin, Jr., M.D.

      Lauren K. Leslie, M.D., FAAP

      Steven Ebert, Pharm. D.

      James M. Perrin, M.D.

      Cynthia R. Pfeffer, M.D.

      Robert D. Gibbons, Ph.D.

      Tana A. Grady-Weliky, M.D.

      Richard P. Malone, M.D.

      Irene E. Ortiz, M.D.

      Matthew V. Rudorfer, M.D.



      Gail W. Griffith



      Kelly Posner, Ph.D.



                        PARTICIPANTS (Continued)



      Samuel Maldonado, M.D., M.P.H.



      Robert Temple, M.D.

      Russell G. Katz, M.D.


      PARTICIPANTS (Continued)


      Thomas Laughren, M.D.

      M. Dianne Murphy, M.D.

      Anne Trontell, M.D., M.P.H.



                            C O N T E N T S


      Call to Order and Opening Remarks:

         Wayne Goodman, M.D.                                     5


      Conflict of Interest Statement

         Anuja Patel                                             6


      Opening Comments

         Thomas Laughren, M.D.                                   9


      Committee Questions and Discussion                        31



         Diane Wysowski, Ph.D.                                 152


      Committee Discussion of Questions and Vote               163


      Concluding Remarks

         P. Joan Chesney, M.D.                                 402

         Wayne Goodman, M.D.                                   404




                         P R O C E E D I N G S


                   Call to Order and Opening Remarks


                DR. GOODMAN:  Welcome to day two of this


      joint two-day session of the Psychopharmacologic


      Drugs Advisory Committee and the Pediatric Advisory


      Committee being held on September 14, 2004, here at


      the Holiday Inn in Bethesda, Maryland.


                We are convened to address recent concerns


      about reports of suicidal ideas and behavior


      developing in some children and adolescents during


      treatment of depression with selective serotonin


      reuptake inhibitors and other antidepressants.


                Our goal is to gather information from a


      variety of sources and perspectives to help us


      understand this complex situation and ultimately,


      to offer the best possible recommendations to the




                Now, I would like to turn the microphone


      to Anuja Patel of the FDA Center for Drug


      Evaluation and Research and Executive Secretary of


      this committee to read the conflict the interest


      statement into the record.




                     Conflict of Interest Statement


                MS. PATEL:  Good morning.  The following


      announcement addresses the issue of conflict of


      interest and is made a part of the record to


      preclude even the appearance of such at this




                The topics to be discussed today are


      issues of broad applicability.  Unlike issues


      before a committee in which a particular company's


      product is discussed, issues of broader


      applicability involve many industrial sponsors and




                All Special Government Employees and


      invited guests have been screened for their


      financial interest as they may apply to the general


      topics at hand.


                The Food and Drug Administration has


      granted particular matter of general applicability


      waivers under 18 U.S.C. 208(b)(3) to the following


      Special Government Employees which permits them to


      participate fully in today's discussion and vote:


      Jean Bronstein, Dr. Joan Chesney, Dr. Wayne




      Goodman, Dr. Lauren Marangell, Dr. James McGough,


      Dr. James Perrin, Dr. Bruce Pollock.  In addition,


      Dr. Philip Wang has been granted a limited waiver


      that permits him to participate in the committee's


      discussions.  He is, however, excluded from voting.


                A copy of the waiver statements may be


      obtained by submitting a written request to the


      Agency's Freedom of Information Office, Room 12A-30


      of the Parklawn Building.


                In addition, Dr. Judith O'Fallon and Dr.


      Victor Santana have de minimis financial interests


      under 5 CFR Part 2640.202 that are covered by


      regulatory waiver under 18 U.S.C. 208(b)(2).


                Because general topics impact so many


      entities, it is not practical to recite all


      potential conflicts of interest as they apply to


      each member, consultant, and guest speaker.


                FDA acknowledges that there may be


      potential conflicts of interest, but because of the


      general nature of the discussion before the


      committees, these potential conflicts are






                With respect to FDA's invited industry


      representative, we would like to disclose that Dr.


      Dilip Mehta and Dr. Samuel Maldonado are


      participating in this meeting as industry


      representatives acting on behalf of regulated


      industry.  Dr. Mehta is retired from Pfizer and Dr.


      Maldonado is employed by Johnson & Johnson.


                With respect to all other participants, we


      ask in the interest of fairness that they address


      any current or previous financial involvement with


      any firm whose product they may wish to comment




                Thank you.


                DR. GOODMAN:  Thank you, Anuja.


                We will be starting off this morning with


      a presentation from Tom Laughren who will give us


      an overview and also pose the questions, the five


      questions to this committee.


                Following his presentation, I would invite


      questions.  I also think it would be a good time


      before we get into the meat of our discussions to


      ask representatives from the FDA questions, to




      further interrogate some of the data that was


      presented yesterday.


                Before we get into the actual discussion


      of the questions, I would like us to think of the


      questions that were carried over from yesterday,


      pose those, and then we will take a short break,


      reconvene and start the process of discussing the




                Is that clear?  Okay.


                Tom, are you ready?


                            Opening Comments


                         Thomas Laughren, M.D.


                DR. LAUGHREN:  Good morning.  I would also


      like to welcome everyone back to the meeting today.


      I would like to do a couple of things in my few


      minutes here.


                First of all, what I want to do is to


      briefly review what I think are some of the key


      findings from Dr. Hammad's presentation yesterday,


      so that you have these in mind as you are


      considering the questions before you.


                Then, I want to talk a little bit about




      what I think the data mean and talk about what some


      of the regulatory options are as you are


      considering our questions, and then I want to go


      over the questions and the topics again.


                These are the 24 trials that we are


      considering. Again, 16 of them were in major


      depression, and the other 8 trials were in several


      various psychiatric disorders - OCD, GAD, 1 in SAD,


      and 1 in ADHD.


                Again, just for summary, I think these are


      the three contributions that the Division made to


      this effort. Again, we went to a lot of effort to


      make sure that we had complete case finding.  With


      the help of Columbia, we accomplished what I think


      is a rational classification of these events, and


      we both obtained and included patient level data in


      our analysis of the suicidality data again to try


      and understand some of the differences both between


      trials, within programs and across programs.


                These are the outcomes that we looked at


      again. The focus of the analysis was on two areas,


      the suicidality event data and also on the suicide




      item data.


                For the event data, could we have the


      other slide up that we had running yesterday?  Our


      primary endpoint, as you recall, was the


      combination of suicidal behavior and ideation,


      Codes 1, 2, and 6, where 1 was suicide attempt, 2


      was preparatory actions, and then 6, suicidal




                So, that was our primary endpoint, but we


      also looked at secondary endpoints, at suicidal


      behavior, in other words, Codes 1 and 2, and then


      suicidal ideation, Code 6, and then for our


      sensitivity analysis, we looked at this larger


      outcome including 1, 2, and 6, but also adding in 3


      and 10, where again, 3 is self-injurious behavior


      where the intent is not known, and 10 is not enough


      information.  Again, these are the cases where


      there is injury, but it is not possible to tell


      whether it's self-injury or other injury.


                With regard to the suicide item data, we


      looked at two measures about worsening suicidality


      on that item or emergence, and these again are the




      cases where the patients are normal at baseline and


      have some increase during the trial.


                In terms of our analytical plan, the major


      focus was on doing risk ratio analyses, both for


      the suicidality event data and for the item data.


      In both cases, we looked at individual trials, as


      well as for the event data, we looked at various




                We looked at both by drug, we combined all


      the SSRIs, MDD trials as a group, we looked at all


      of the other indications combined as a group and


      also did one pooling which included all 24 trials.


      For the item data, we looked again at individual


      trials and then a pooled analysis over all trials.


                Dr. Hammad put a lot of effort into again


      trying to explain the differences that we were


      seeing between trials within programs and across


      programs, and I just want to spend a couple of


      minutes talking about exactly what he did.


                He looked for confounding within trials


      using both the univariate approach and a


      multivariate approach.  There were a total of 17




      covariates that he looked at.  He was not able to


      find any evidence for important confounding in that




                He also did stratified analysis to explore


      for effect modification.  The three variables that


      he looked at were age, gender, and history of


      suicide attempt or ideation, so basically, what he


      did in each of these is to stratify on these


      variables within trials to look to see if there was


      basically an interaction.


                Again, he did not find any evidence for


      that, so basically, what that means is that on


      these variables, you find the signal both in


      children and adolescents, you find it both in males


      and females, and you find it both in those with and


      without history of suicide attempt or ideation.


                Finally, he looked at 12 trial level


      covariates, again, as an attempt to try and explain


      the differences across trials using a


      meta-regression approach.  Again, that approach was


      not able to explain the variability.


                Now, I would say that one of the problems




      in doing these kinds of explorations is that there


      is very limited power, you have a very small number


      of events.  When you use an eyeball approach to the


      data, you can't help but thinking that trial


      differences might have made a difference.


                I just use the TADS, the fluoxetine


      situation as an example.  The company had three


      trials.  There was no signal coming from those


      three trials.  If you look at the careful screening


      that was done to obtain the patients for those


      samples, and the exclusions of patients with prior


      histories of treatment resistance, and so forth,


      and then you look at the TADS sample, which is many


      ways was probably more representative of the


      community of patients who actually get treated,


      there is quite a difference.  Again, as you recall,


      in the TADS trial, you see quite a striking signal


      for suicidality.


                So, even though quantitatively, we weren't


      able to tease that out and to explain the


      differences using various quantitative approaches,


      it is hard to think that that may not have made a






                In my next three slides, I am going to


      present very briefly some of the data.


                What this slide is, is presenting the risk


      ratios for various poolings.  So, in this column,


      you have the risk ratios on our primary endpoint,


      which was suicidality ideation or behavior, 1, 2,


      and 6.


                In the second column, you have this


      expanded sensitivity analysis, 1, 2, 6, plus adding


      3 and 10.  The first row is all trials, so this is


      a pooling across all 24 trials.  In the second row,


      you have the pooling of the 11 trials with SSRIs


      and major depression.


                Now, there are two things I want you to


      notice about this slide.  First of all, in every


      case, the risk ratios are around 2.  They range


      from 1.7 to 2.2, but they are sort of in the


      vicinity of 2.


                Secondly, if you look at the confidence


      intervals on these risk ratios, in every case, it


      does not include 1, so in that sense, it is a




      statistically significant finding. So, this is the


      pooled data.


                What I have given you in this slide are a


      different set of poolings.  Here, what I am doing


      is pooling the individual depression trials in the


      7 programs that looked at depression, and these are


      the 7 programs listed here.  Every row is a


      separate depression program.


                What I have given you here, first of all,


      is the outcome on our primary endpoint a


      combination of 1, 2, and 6.  I have also given you,


      in the second column, the outcome on suicidal


      behavior, and in the third column, the outcome on


      suicidal ideation.


                There are a couple of things I want you to


      notice about this slide.  First of all, in every


      instance where we have events, and we had no events


      for Serzone, but in the other 6 instances where you


      have events, the risk ratio is always greater than




                Now, I want to turn to trying to tease


      apart where that overall effect is coming from if




      you break it apart by behavior and ideation.  Dr.


      Hammad made this point yesterday, in three cases it


      appears as if the overall effect is coming from


      behavior, in three cases it looks like it is coming


      from ideation.


                So, if you look at Celexa, here is the


      risk ratio for behavior, 2.23.  There is nothing


      happening for ideation.


                If you look at Paxil, again, it looks like


      it is coming mostly from behavior.


                If you look at Prozac, it looks like it is


      probably coming more from behavior than from




                For Effexor, there is a signal coming from


      both, but it is clearly coming more from ideation.


      Here, the confidence interval is almost




                For Remeron, it is all coming from


      ideation, and from Zoloft, there is nothing


      happening for behavior, it is all coming from




                I am not sure what this means.  As Dr.




      Hammad pointed out, this may simply be a small


      numbers problem, but we are not seeing a consistent


      finding in terms of where the overall effect is


      coming from.


                Finally, what I have given you in this


      slide is the data from the individual other 8


      trials in non-MDD indications.  As you get into


      these trials, the number of events you are dealing


      with is very small, and just to illustrate that, I


      have put the actual number of events in this slide.


                So, in each of these parentheses, the


      first one is the number of events for drug, and the


      second one is for placebo.  So, you can see the


      small number of events that we are dealing with.


                If you recall from the previous slide, for


      Effexor, we were seeing quite a strong signal for


      major depression.  These are two GAD studies.


      There is nothing at all happening here.


                For Luvox, again, Luvox was only studied


      in OCD, there was no depression trial.  Just one


      study in depression, only two events.  They were


      both happening in the drug group.




                For the two non-MDD Paxil studies, one in


      social anxiety, one in OCD, again, small numbers of


      events, but in both cases, they were happening in


      the drug group.


                The same for the Prozac OCD, just one


      event, but it happened in the drug group.


                No events for Wellbutrin.


                For Zoloft, this is the only case where


      the one event is happening in placebo, and not in




                It is hard to know what to make of all of


      this, although the one thing that you can't help


      noticing is that even though there are a small


      number of events, where events occurred, they most


      happen on the drug side.


                Just to summarize these data, again, if


      you look at various pooled analyses, the risk


      ratios hover around 2.  They range from 1.7 to 2.2.


      In all cases for those poolings, it appears to be a


      significant finding.


                The signal appears to be coming mostly


      from major depression, although perhaps not




      exclusively.  Despite those findings, there still


      are these inconsistencies in this risk, both across


      trials, within programs and across programs.


                On the other hand, my view is--and there


      isn't necessarily one consistent view coming out of


      FDA on this--but my view is that this is a


      reasonably consistent signal for risk.  You are


      seeing it in seven of nine programs.  We don't see


      any events in Wellbutrin.  On the other hand,


      Wellbutrin was only studied in ADHD, just one




                There is no signal coming from Serzone,


      which was studied in major depression.  I am not


      sure if that means that Serzone is free of risk or


      it simply may mean that the events, the


      ascertainment in those programs was not good enough


      to pick them up.  I don't know the answer.


                One other point that Dr. Hammad made


      yesterday, that I want to return to, is a way of


      thinking about this risk is in terms of risk


      difference, and if you look over all these trials


      and estimate what the risk difference is, that is




      the difference in the risk between drug and


      placebo, so you are subtracting the placebo risk


      from the drug risk, it is in the range of 2 to 3




                What that means is that again, out of 100


      patients treated--this is short term now,


      short-term treatment--you can expect 2 or 3 out of


      that 100 will have some excess of suicidality above


      and beyond what would be in the background that is


      due to drug.


                As a clinician, what you have to do is to


      balance that risk against the perceived benefit.


      The problem here, of course, is that we only have,


      at least from FDA's standpoint, a demonstration of


      benefit for Prozac, but if you take the TADS trial


      as an example of benefit, there, you can look at


      the benefit difference, and the benefit difference


      in the TADS trial, difference between drug and


      placebo in percent of responders, using that as the


      measure of benefit, it is about 25 percent.


                Again, you can interpret that in the same


      way, so that if you look at 100 patients who are




      treated with fluoxetine, you can expect that about


      25 out of 100 will have that benefit if you are


      looking at response as the benefit.


                So, you balance that against the risk,


      which again in that trial, the risk actually was


      greater than the 2 percent, it was probably more on


      the order of 7 percent, but you balance that risk


      against the benefit.  That is the kind of calculus


      that a clinician has to do.


                Finally, as was pointed out, there were no


      completed suicides in any of these trials.


                Again, we did not see the same signal in


      looking at the item data.  One exploration we tried


      to do to see if that could be explained by patients


      dropping out, and unfortunately, that was not an


      explanation.  The analysis of completers did not


      show really any difference from the analysis of the


      patients who dropped out.


                So, how should these findings be


      interpreted?  I think that this is an indication


      that there may be some increased risk for


      suicidality during short-term treatment, and I




      think this is probably a class effect.  Again, you


      are not seeing it in every drug that we looked at,


      Serzone and Wellbutrin being the two exceptions,


      but I think there is enough here to suggest that


      this is probably a class effect.


                The signal appears to be most compelling


      in major depression.  It may not be limited to that


      population, but again we are left with this very


      unusual variation in the signal across trials,


      within programs and across programs that we have


      not been able really to explain.


                What I want to do next is to talk about


      what some of the regulatory options are, and I


      first want to talk about possible labeling changes.


                As you recall, we already made a fairly


      major change to labeling back in March, and all of


      those changes have now been implemented.  There is


      a fairly prominent warning statement that directs


      the attention of prescribers to this possible




                Now, that language as it currently is


      written suggests that causality has not been




      established.  One thing that might be done to


      modify that, if there is agreement on this, we


      could say that causality has now been established


      for this risk in pediatric patients.


                In addition to that, we could go beyond


      that and provide specific suicidality findings in


      the labels for different products.  We could also


      provide more specific information about the


      efficacy findings for specific products in that




                There are other things to talk about in


      terms of that warning statement including things


      like bolding language or putting black boxes.


      These are all options that are on the table.


                The other option that you need to think


      about, and you heard many yesterday in the open


      session ask us to do this, you can think about


      contraindications.  The one thing I want to point


      out is that in this country, for our label, a


      contraindication means never.  It means that that


      drug will never be used in treating these patients,


      it is not an option.




                The other thing I want to point out is


      that the term "contraindication" has different


      meanings in different regulatory settings.  In some


      settings, it does not mean never.  If you read the


      fine print in the UK, for example, there is a


      suggestion that specialists may still use that


      drug.  So, you need to keep that in mind that in


      this country, a contraindication means that that


      drug is never an option.


                In addition to labeling changes, there are


      some other obvious actions that we can and almost


      certainly will take.  Our plan at present is to


      write a medication guide. This is basically


      labeling which ideally would be attached to the


      medication when it is prescribed in unit of use




                In addition to that, we will undoubtedly


      have another public health advisory when we decide


      on what needs to be done, and we will try and


      communicate these findings to our partners.


                Now, what I would like to do again is to


      quickly go through the questions and the topics. 




      The first topic is again we would like to have your


      comments on our approach to classifying these cases


      and to our analysis of the data.


                One of the questions for which we really


      need to have you vote on is do you feel that the


      suicidality data from these trials support the


      conclusion that any or all of these drugs increase


      the risk of suicidality in pediatric patients.


                If the answer to that question is yes, to


      which of these nine drugs does this increased risk


      apply, in other words, is this a class effect for


      all antidepressants, does it apply to certain


      subclasses within this broader class, or to


      specific drugs?


                If this is a class risk or if it applies


      to certain drugs, how should this information be


      reflected in the labeling for each of these


      products, and what, if any, additional regulatory


      actions should the agency take?


                Finally, there is this question about what


      additional research is needed to further delineate


      the risks and the benefits of these drugs in




      pediatric patients with psychiatric illness.


                At our last meeting, I suggested one type


      of study that you might think about, and I am going


      to make that suggestion again, because we think


      that this is one study that might get at one of the


      deficiencies here, and that is, not only do we not


      have enough information about short-term benefit,


      we also have little information about longer term


      benefit or risk.


                One way of getting at longer term benefit


      is the randomized withdrawal study.  Basically, the


      way the study works is that patients who are


      responders or appear to be responding to treating,


      at some point in the course of treatment, are


      randomized to either continue on drug or randomized


      to placebo, and one looks at time to relapse as the




                Now, I know there are concerns about that


      design. You know, one concern is the ethical issue


      of taking patients off a medication when they


      appear to be responding. I agree that is a concern,


      but I think there is a way of dealing with that.




                The usual randomized withdrawal trial is


      done after too short a period of time on treatment.


      I mean typically, they are done now after 12 weeks


      or so of treatment.  That is too soon.  No


      clinician would take a patient off of one of these


      medications at that point in time.


                On the other hand, at some point in the


      course of treatment, whether it is six months or


      nine months or a year, it seems to me that it is a


      reasonable question.  At some point, you reach


      equipoise where the clinician has to ask the


      question, well, is this long enough, you know, is


      there any benefit in continuing the treatment


      beyond this point in time.


                Now, that is a much harder study to do, to


      keep patients on treatment for nine months or a


      year before you randomize them, but that would be a


      way of answering that important question of whether


      or not there is continuing benefit beyond that


      point in time.


                The other concern that has been raised


      about these trials is the issue of distinguishing




      between withdrawal symptoms and relapse.  Again, I


      agree that this is a reasonable concern, but I


      think there is also a way of addressing that.


                In clinical practice these days, these


      drugs are tapered.  One doesn't stop them cold


      turkey.  I think that could also be part of that


      design, and that could address that issue.  So,


      that is one thing to think about.


                Before I end, I want to leave you with two


      thoughts.  We clearly have an obligation at FDA to


      inform clinicians and patients about the risks that


      are associated with these drugs, and we take this


      obligation very seriously.


                Along those lines, I just want to point


      out that our current regulations do not require the


      same level of certainty with regard to safety in


      terms of causality as is required for efficacy.  In


      other words, we can issue warning statements with


      somewhat lesser certainty about causality than is


      required to support a claim.


                Secondly, as I have pointed out several


      times, the lack of efficacy data in this setting




      for most of these drugs needs to be part of this


      discussion.  On the other hand, and I am not making


      your job easy, please bear in mind that depression,


      whether in adults or children, is a very serious


      illness that is associated with morbidity and


      mortality quite apart from whatever role


      antidepressants might have.


                As was pointed out yesterday, this is the


      major cause of death in this population, the


      depression itself, so please bear that in mind.


                I have very profound respect and gratitude


      for the clinicians who are out there on the front


      lines still willing to take care of these patients


      despite what has become a very controversial and


      difficult environment.


                I hope that as we discuss these issues and


      make a decision, that we not make it impossible for


      them to practice medicine.


                Thank you.


                DR. GOODMAN:  Thank you, Tom, for a cogent


      and clear presentation.


                I would like to ask committee members if




      they have any questions of Tom.


                   Committee Questions and Discussion


                DR. FOST:  This is for Tom or anyone else


      who has a handle on the numbers.  I know there is


      no precise answer to it, but it would be helpful to


      me to just hear you or someone else, maybe Dr.


      Shaffer, if he is still here and is allowed to


      talk, this question.


                Suppose there were no SSRIs, suppose they


      were contraindicated, that is, prohibited,


      approximately, let me just ask the question about


      suicides, about completed suicides, and I


      understand there is no suicides in the FDA data,


      but based on everything that we know,


      approximately, would there be more suicides, fewer


      suicides, or the same amount if there were no SSRIs


      in children?


                DR. TEMPLE:  There is not going to be any


      way to answer that, in part because you can't do


      rigorous studies of the kind that would answer


      that.  No one is going to let you not treat, not


      institutionalize, et cetera, someone who is getting




      worse and worse, and it would require long-term


      studies presumably against no treatment, and it is


      not easy to figure out how anybody is going to do




                So, you are left with the kind of data


      that people have pointed out is always uncertain,


      the data on suicide rates and whether they are


      going up or down, so it is very hard to answer that




                There were no completed suicides in the


      pediatric data, so that doesn't give you a clue.


      You can form your own judgment about whether


      increased suicidal behavior or thinking is going to


      lead to suicides in a certain fraction of cases.


      It is hard to imagine that it couldn't, but you


      don't know what that ratio is.


                The success rate of suicidal attempts is


      relatively low.  I gather it is higher in males


      than females, but I don't think there is going to


      be ways to put numbers on that.


                You have to form your judgment about


      whether you think the overall decline in suicides




      has got something to do with therapy or has


      something to do with other aspects of life in the


      United States, and nobody can give you a firm


      answer to that, as Dr. Wysowski said and as others


      have said.  So, it is very hard to answer that




                Certainly, some of the people who spoke


      yesterday, some of the treating physicians were


      quite sure that they were helping people with the


      drugs, and you heard families who said that their


      relatives were made much worse by the drugs.


      Putting numbers on that, though, isn't feasible


      based on the data we have.


                DR. FOST:  A related question.  To those,


      Dr. Shaffer and others who note a decline in


      suicides in the United States, in parallel with the


      increased use of SSRIs, and let's just say which


      should be an increase in suicidality, suicidal


      ideation due to SSRI, what is the hypothesis there,


      that there is fewer suicides, but more suicidal


      ideation?  That is what the data seemed to suggest,


      and I am confused by that.




                DR. TEMPLE:  Can I make another comment?


      The studies you are looking at are all the


      short-term studies. As Tom was pointing out, we


      have none of the long term sort of relapse


      prevention data.  It seems entirely possible that a


      drug could be causing early suicidality, but once


      you are over that period, it prevents relapse,


      which could have an impact.


                You know, there is just literally no way


      to sort that out with present data.  I mean it has


      never been my thought that any benefit these drugs


      have consists entirely of their treatment of the


      acute episode, because in adults anyway, we have


      lots of data showing that the likelihood and timing


      of relapse is affected by continued therapy.


                As Tom said, most of those studies go


      earlier than you would like to do in a pediatric


      population, because they consistently show that


      quite reliably.  Maybe that is where their


      importance is, it is very hard to know.


                DR. GOODMAN:  Dr. Pine is next.


                DR. PINE:  I have a question about some of




      the regulatory options.  In thinking both about a


      number of the comments that were made yesterday, as


      well as your comments at the end about how


      difficult the decision that we will have today,


      related at least in part to the dearth of data that


      we really need.


                Are there any options from a


      pharmacovigilance standpoint as far as regulatory


      actions that might increase the degree to which we


      are focusing over the next time period on the


      emergence of these events or bring, you know, new


      data over the next months to years based on a


      regulatory action?


                DR. KATZ:  There is the mechanism of Phase


      IV requirements that say we can impose requirements


      on sponsors to do various studies in Phase IV and


      postmarketing environment.  The question would be


      what those studies would look like.  I think that


      is the question.


                There are other obviously entities, the


      NIMH and others who were set up obviously to do


      large trials, and again the question is what would




      those trials look like.  You could do I suppose


      large long-term, and again, you have heard, I


      think, a lot of people say that there is a need for


      long-term data.


                I suppose you could do long-term


      comparative trials, you can't do long-term


      placebo-controlled trials, so other than the sort


      of randomized withdrawal design I think that Tom


      talked about.


                So, there is a mechanism to require




                DR. PINE:  I guess I am not so much asking


      about studies, and this maybe is a bit of an unfair


      analogy, but in New York, for example, as well as


      other states, whenever you write a prescription for


      a psychostimulant, there are a whole host of


      procedures that kind of go with that, that are


      designed to allow monitoring of the use of


      psychostimulants and the associated effects.


                Is there any--again, I realize I am


      thinking a little bit out of the box--is there any


      form of, I don't know, computer based or monitoring




      system that might give us a better handle on how


      many of these events are actually happening in


      regular treatment?


                DR. TEMPLE:  ODS should comment on that,


      but it is worth just looking at, say, the study Dr.


      Jick tried to do. There isn't any no-treatment


      group in that.  He is just comparing the risk with


      one group of drugs with another, and you can


      definitely do studies like that, but if you tried


      to compare treated people with untreated people,


      there will always be the concern of whether the


      groups are fundamentally different, a very


      difficult problem because people are treated.


                There might be environments in which


      treatment is not so common, where there is less


      likelihood to treat.  Maybe in those environments,


      you could do something like that, but Anne wanted


      to talk.


                DR. TRONTELL:  Just to expand briefly, you


      are talking about using observational data as Dr.


      Temple pointed out, where you don't have a control


      group, and although you might register patients, we




      have seen even in clinical trials that we have been


      discussing this past day, that the issue of


      ascertainment of these events is very complicated


      when you actually have a clinical trial mechanism


      in place to capture those events.


                The other challenge that you face with


      observational data, because people don't receive


      the drugs randomly, there is a phenomenon called


      "confounding by indication," in fact, some of your


      sicker patients you might presume are the ones who


      are getting the medication.


                We try and control for that, but it is


      very complex.  I think the better option is to


      think of some systematic way, and then you are in


      the realm of studies, as Dr. Katz was saying.


                DR. MURPHY:  I just wanted to follow up on


      one last thing.  Because we already know that using


      the system we have now for follow-up


      post-exclusivity because it is already mandated


      that we do one-year reporting once these products,


      whether they are approved or not, so we are looking


      at all-use.




                We do look at that and we report that, and


      we know that that is not going to inform us, you


      know, to answer the questions we need to answer,


      because of all the things that will impact that




                DR. GOODMAN:  Dr. Temple.


                DR. TEMPLE:  I just wanted to mention one


      related, but not quite on-point matter.  We talked


      yesterday about concern that the studies that had


      been done to gain exclusivity might have been not


      as good as we would like.


                We weren't particularly talking about the


      design of the studies, which we think is okay, but


      let's say the approach to them.  Maybe there was


      too much of a rush, and so on.  If we were to put


      out a written request now, it would be one that


      required a third arm to the study, namely, a Prozac


      arm, because we know that Prozac can be shown to be




                So, the study wouldn't count unless it had


      been able to show that it had what we call "assay


      sensitivity," the ability to tell effective drugs




      from ineffective drugs. We couldn't do that before


      because there wasn't anything at the time we wrote


      those requests that was known to be showable in


      children, but now there is.  There is three studies


      that all seem to show something.


                So, we should have much better information


      about what the pediatric population does in future


      requests.  That doesn't help the present




                DR. MURPHY:  I wanted to address that


      issue again, too, because I think I want the


      committee to be very clear on the fact that the


      Agency tells the company very clearly the type of


      studies that need to be done.


                We do give them, you know, a broader


      picture of the number of patients.  We tell them


      what we know will be the minimum, and, in general,


      I think Tom would agree that most of these studies


      have come in with the numbers in each arm that we


      have seen in other studies where they have shown




                So, the point here being that we do have




      control over the types of trials that are done, the


      number of patients, and the monitoring.  However,


      because there is a template up on your web that


      basically tells you what we ask for in depression




                When you look at what the safety is, as


      has been pointed out many times, these trials were


      not set up to answer that question.  So, I think it


      is those kinds of issues that we would like to hear


      more about today.  As Dr. Temple said, it is how


      better to do these trials in the future.


                Thank you.


                DR. GOODMAN:  Thanks for that statements,


      Dianne. I just want to make sure I understand it




                I think what you are saying, that if the


      conditions had been different at the time, that is,


      that the drug company was required to show, not


      only have a study, but a study that was positive.


      Then, the design would not have been any different,


      the sample size would not have been any different


      under those circumstances than the ones that




      existed at the time.


                DR. MURPHY:  I think what we are saying,


      that for the trials that we designed, they were the


      same for the one that did show some effect, which


      is Prozac, as those that did not, and that what we


      don't know is if a company is putting a trial


      together, and let's say we said that they had to


      have 300 patients to get their exclusivity, but for


      other reasons they really wanted this product


      approved, and they felt the enrollment was not


      going the way that they needed, would there be some


      other push within that company to then go out and


      get more patients, so that their enrollment would


      be better versus an exclusivity where all they had


      to do was meet that criteria.


                I am making that number up.  I think the


      issues that people were trying to get at is that is


      there a difference that affects behavior when you


      just know you have to do certain things versus you


      have another goal, which may be approval.


                DR. GOODMAN:  Dr. Temple.


                DR. TEMPLE:  The requirement for a third




      arm in evidence of assay sensitivity leaves it up


      to the company to decide how they are going to do a


      successful study.  They can look at the available


      data on Prozac and say, oh, here is the number I


      need, here is the kind of patients I need.  That


      succeeded in those three trials.


                They would then know that the trial would


      have to be one that can show the difference between


      Prozac and placebo.  That doesn't mean their drug


      has to show a difference between drug and placebo.


                That would be determined by the results,


      and there is no obligation that the drug be


      successful, but we would at least know we had a


      study that was capable of detecting effective drugs


      and distinguishing effective drugs from ineffective




                That would then become a requirement for


      meeting the terms of the written request because


      they would have to show that they had an adequate


      study.  Before there was an effective drug, there


      was no way to do that.  You couldn't tell whether


      the study was a good study or not.




                DR. GOODMAN:  Dr. Marangell.


                DR. MARANGELL:  If I could go back and


      address the question of what would the hypothesis


      be for long term, certainly, in the absence of


      data, there is some degree of speculation.  I do


      have a question directly to the FDA.  Is it okay if


      I respond?


                I think the number one hypothesis would be


      in the short run when you have depressed patients


      who are not yet stabilized, you may see an


      increased risk, and you do see certainly in this


      population an increased risk of suicidality.


                I imagine that what we would see with


      longer term data is a substantial decrease in


      suicidality over time, and that is what we are


      inferring from the cohort and the epidemiologic


      data.  I think that clinically makes sense, as well


      as mechanistically makes sense.


                The question for the FDA, can you give us


      a sense, I mean do we feel confident that we


      actually have all the available studies now in both


      children, adolescents, as well as in adults, and




      what is the FDA policy on requiring review of those


      studies including negative studies, when do they


      come to you and when do they become publicly




                DR. TEMPLE:  Well, let me start, others


      can comment.  When you submit to us an application


      to change the labeling, to add a claim, say, for


      pediatric use, you are clearly obliged under the


      law to provide every study, successful ones,


      unsuccessful ones, things that were interrupted,


      and so on.


                As far as we know, we are getting all


      those studies.  Of course, if there were something


      that were done that we didn't know about, well,


      then, we wouldn't know about it, but as far as we


      know, we are getting them all.


                So, most of the pediatric submissions to


      us were associated with labeling requests or


      something like that, so as far as we know, we have


      all those data.


                Dianne can tell you what is required under


      the best BPCA, and I think there, too, they have to




      provide them.  We have no rule that affects whether


      people have to publish results.  Congress is


      considering that, so are the journals and everybody


      is talking about that.


                Under BPCA, however, when we grant


      exclusivity, we provide summarized results, and we


      have done that for the drugs where the written


      requests were written after the BPCA, and we have


      gone back and asked the companies for permission to


      summarize our analyses for all of the others where


      it wasn't totally clear whether we could do it or




                So, the summarized result, that is not the


      same as a complete study report, the summarized


      results are now available publicly on all of those.


      I am sure between PhRMA's commitment to provide a


      registry between the journals insistence that they


      will get a registry, between congressional


      interest, I am quite confident that there will be a


      change in the way things get published.


                DR. MURPHY:  The only thing that I could


      add to that is that for the committee, for the




      routine practice within FDA, if a company submits


      an application, we review it, the studies are


      negative, there is no public acknowledgment of that


      unless the company for some reason wants to make


      that knowledge public.  We are not allowed to


      comment on that.


                Now, under BPCA, it said, it has a


      disclosure section that says you, FDA, will


      publish, as Dr. Temple is referring to, the


      summaries, the medical and pharmacology summaries


      up on the web--make them public, and actually, we


      have chosen to do that on the web--and we have done




                One of the issues that has happened is


      that between the enactment of the new legislation


      and the old legislation, legally, things were


      considered issues under the old legislation, so


      even though the studies came in, we had to reissue


      all those written requests to be able to say they


      now were subject to this new mandate.


                So, what again Dr. Temple was telling you


      is that unfortunately, many of the antidepressants




      came in, in that period when we had not yet issued


      that letter, but despite that, we have asked the


      sponsors to allow us to put those summaries up, and


      they have given permission to do so.


                That is why yesterday we said up on the


      web now are the summaries.  Again, this is not the


      data.  There is variations in, you know, some


      medical officers will put in more information than


      others in how much data is in these summaries, but


      they are up now, publicly available.


                DR. MARANGELL:  Is that true for adults,


      as well?


                DR. MURPHY:  No, adults are still under


      the same standard.  In other words, if the study is


      negative, we don't talk about it.


                DR. MARANGELL:  So, as an example, if an


      antidepressant manufacturer did a study in a new


      indication for a drug that is currently available,


      found increased risks of suicidality, no one would


      be under any obligation to make that public?


                DR. MURPHY:  That is a different issue.


                DR. MARANGELL:  But that is the question.




                DR. MURPHY:  The issue is safety, and the


      Agency always has the ability to make public safety


      issues that arise.


                Bob, do you want to say anything else


      about that?


                DR. TEMPLE:  We consider, for example, if


      someone with an antidepressant comes in for, I


      don't know, obsessive compulsive disease, and we


      don't buy it, we do not make those data available,


      they are considered confidential commercial


      information.  Obviously, a lot of the people, a lot


      of the public doesn't like that approach.  We think


      that is what we are required to do.  I can't


      comment on that, I am not the lawyer here.


                However, companies have a separate


      obligation for drugs that are marketed to report


      serious and unexpected, and any serious adverse


      reactions to us, and to do so promptly.  A finding


      of increased suicidality where that was not known,


      clearly meets that test, and they would be obliged


      to report it to us.  If we then thought that was


      true, we would add it to the label or do whatever




      we are supposed to do.


                So, safety data meets a different


      standard.  A new carcinogenicity study or


      something, those do have to be reported to us.


                Other studies have to be reported in the


      annual report, but they are not necessarily


      reported in detail, and not that much is


      necessarily made of them, and they do not


      necessarily become public.


                DR. GOODMAN:  Dr. Pollock.


                DR. POLLOCK:  Yes, the serious safety


      issue would have to be reported while the trial is


      ongoing to you, right?


                DR. TEMPLE:  Well, if it arises from a


      trial, it has to be.  Actually, the requirements


      for reporting serious unexpected events in a trial


      are more or less identical to the requirements


      before a drug is marketed.  They have to be


      reported to us within 7 or 15 days.


                A finding from an epidemiologic study,


      there is some judgment involved in whether that


      represents the kind of thing that has to be




      reported promptly, but they basically do.


                DR. KATZ:  There is also some judgment


      involved in whether or not an event is considered


      to be unexpected.  So, for example, suicide in a


      study of patients who are at risk anyway might not


      be reported to us in real-time, because it might be


      considered to be expected, the blind is still


      intact, you don't know if it's drug or placebo if


      it is in the context of a controlled trial.


                Afterwards, though, when the trial is done


      and analyzed, and it turns out that there is an


      increased incidence on drug compared to placebo,


      that is something we would find out about.


                DR. GOODMAN:  Go ahead, Dr. Pollock.


                DR. POLLOCK:  I actually wanted to explore


      your thinking a little bit about the recommendation


      for a maintenance trial.  I guess there are a


      couple of things. One is if there is this acute


      toxicity that we are concerned about, clearly, it


      doesn't address that because you are dealing with


      the children or the adolescents who have actually


      responded, and then are withdrawn.




                But I wondered if there was implicit in


      your request for that, a concern that still that


      the shorter half-life SSRIs seem to be, maybe not


      statistically, but certainly qualitatively more at


      risk in causing this phenomenon.


                I was taking that as implicit perhaps in


      your suggestion, maybe I am over-interpreting it,


      but is there a belief that somehow--I mean it just


      seems more than coincidence that signals seem a


      little bit higher.


                I know it has now emerged with Prozac, but


      certainly, Effexor, venlafaxine stands out at one


      end, then followed by paroxetine, and if there was


      kind of an implicit question that you were asking,


      assuming that people are still using after we are


      finished, you know, those medications, that you can


      require that those manufacturers actually conduct a


      serious maintenance trial as part of you were


      saying your Phase IV regulatory requirement.


                DR. LAUGHREN:  We certainly, you know,


      until we saw the TADS data, were entertaining the


      notion that discontinuation might be one




      explanation for the bigger signal, the apparent


      signal that we are seeing with Paxil and Effexor.


                The TADS finding certainly challenges that


      notion as a unitary explanation, since that is the


      single trial among the 24 that, by itself, has a


      statistically significant finding for that signal.


      That doesn't mean that the other explanation isn't


      possible.  I mean this could be a much more complex


      situation than one might seem at first glance.


                But a maintenance trial is not going to


      answer all those questions.  I mean a maintenance


      trial is only going to answer the question of


      longer term benefit, but the reality is that many


      clinicians, despite these concerns, are probably


      going to continue to use these drugs, and we have a


      dearth of information about what the longer term


      benefits are.  The maintenance trial is one way, I


      think, of getting at that.


                Now, there is this issue of how to


      interpret emerging symptoms in that setting, you


      know, when you take patients off the drug.  Of


      course, the drugs like Paxil and Effexor, that are




      known to have a stronger signal for


      discontinuation, obviously are a challenge in doing


      that kind of trial, but I think that one could, as


      one does in clinical practice, taper those patients


      to try and address that, and then look for what


      would be considered for relapse.


                DR. GOODMAN:  Dr. Temple.


                DR. TEMPLE:  There is another reason to do


      randomized withdrawal studies.  As everybody knows,


      in adults, the failure rate for conventional


      clinical trials of the acute episode is about 50


      percent.  That is, half the trials can't tell drug


      from placebo, and that is true even when you


      include a third arm of a drug that is known to


      work.  That appears to be the nature of the beast.


                Nobody really has a good explanation


      because if they did, they would fix it, but we at


      least think it has something to do with the


      environment and the discussions that go on even


      informally, even if it is not planned as part of


      the treatment.


                In the randomized withdrawal setting, the




      success rate for drugs that are known to work is


      nearly 100 percent. Very few of those trials ever




                There are at least two reasons.  One, only


      people who seem to be doing well are in the trial,


      so they are enriched with a responder population.


      You can make of that what you will.


                The other possibility, though, is that the


      environmental things that help people get better


      aren't really there, they are just out living in


      the community, there is nothing nurturing about it.


      They are just back in their usual environment.


                So, one of the attractiveness of these is


      to find out whether the drugs actually provide some


      benefit, even in people who seem to be doing well


      on them, which seems an important question here.  I


      mean, as Tom has pointed out repeatedly, the


      failure of most of the drugs to show effectiveness


      doesn't mean they don't work.  On the other hand,


      we don't have evidence that they do work, and that


      is not irrelevant either.


                A good way to show that, if they do, is




      the randomized withdrawal study.  At least that has


      been the history in adults, so there is a lot of


      attractiveness to it.


                DR. GOODMAN:  Dr. Chesney.


                DR. CHESNEY:  Thank you.  I have two


      questions.               The first one is for Dr.


      Murphy and Dr. Temple, and the second for Dr.


      Laughren.  The first question addresses the


      exclusivity issue.  I feel like in this case, we


      bypass the Phase I/Phase II stages that we would


      normally go through with new drugs, so we never did


      do the pharmacokinetic/pharmacodynamic dose finding


      in children that we would have done had these been


      new drugs.


                I wondered, I probably should know this,


      but could either of you explain, when we offer


      exclusivity with a new drug, if it is a new drug to


      children, do we require those studies, or do we


      not?  I am sure it is not that straightforward.


                DR. MURPHY:  We did required


      pharmacokinetic studies.  Actually, on the


      template, we outline three types of studies.  They




      have to do two randomized, double-blind,


      placebo-controlled, acute treatment trials with


      recommendation at six to eight weeks for safety and


      efficacy.  They also are to do a pharmacokinetic


      study to provide information pertinent to dosing of


      study drug, and they are to do a safety study.


                So, all of those were asked for.  Now, if


      you are asking do we go back and demand redoing


      dose finding again in these, no, they were not


      worded that way.  It was said that the PK study


      could be a traditional PK or, alternatively, a pop


      PK, and actually, I don't think that the study had


      any other information that would have, in essence,


      told the company that they needed to redo the dose


      finding, if that answers your question.


                DR. CHESNEY:  So, do we have dose


      information on all of these drugs?  Do we know what


      the usual ranges are, and what excessive ranges


      are, all those things?


                DR. GOODMAN:  Go ahead, Dr. Katz.


                DR. KATZ:  I think Tom mentioned this in


      one of his slides yesterday.  The written requests




      that we issue now are very different from the


      written requests we issued that probably generated


      most of the trials that we are talking about here


      yesterday and today.


                As Dianne pointed out, for example, in


      pharmacokinetics, we gave sponsors the opportunity


      to generate the kinetics in kids based on so-called


      population pharmacokinetic analyses, which is to


      say from data generated in the controlled trials.


                So, it was sort of after the fact.  It was


      just what is the kinetics of the doses you happen


      to give in the trials.


                In the earlier written requests, it was


      sort of the pediatric drug development was sort of


      tacked onto the adults, in other words, when the


      trials were designed even, the treatment effect


      size, for example, was used to calculate sample


      size was taken from the treatment effect size seen


      in adults.  We had no information, even preliminary


      information in kids.


                So, we didn't have a lot of preliminary


      information in those days that could inform




      adequate trial design in this population, in this




                Nowadays, we ask for different things.  We


      ask for formal PK, so we can learn before the


      definitive trial design, what the kinetics are,


      what doses give rise to what plasma levels.  We ask


      for dose finding studies, so we can determine


      before we design the definitive trials what the


      tolerated dose range is.


                So, the written requests are much


      different now than they were at the time that the


      requests are generated, these data were written.


                DR. MURPHY:  Just to reinforce that is


      that these were some of the earliest written


      requests that went out, so they really, as has been


      stated, and I think we tried to say this earlier


      on, we are learning.


                I mean because of the lack of prior


      research and some fundamental scientific questions


      haven't been answered, we are learning from the


      trials that we have now about how to do a better


      job on designing some of these trials, but these




      were some of the very earliest ones that were




                DR. CHESNEY:  Dr. Temple, did you want to


      comment on that?


                DR. TEMPLE:  Well, I just wanted to say


      there isn't any pharmacodynamic measure to allow


      you to do what is called PK/PD other than


      effectiveness itself.  In a lot of cardiovascular


      settings, there is at least something you think


      relates to the desired effects, so you can do


      relatively short-term studies and get a PK/PD




                Here, your only way to do it is to insist


      that every drug, every study be a dose response


      study, which is of considerable difficulty.  We


      have trouble getting really good data even in


      adults actually given the sample sizes involved,


      but there isn't any measure yet.  Maybe one of


      these days there will be an MRI measurement or


      something, but not yet.


                DR. CHESNEY:  Well, I don't want to


      overstay my welcome, and I do have a question for




      Dr. Laughren, but one does wonder about some of


      these children who didn't even express ideation and


      just suddenly, very early on, if they didn't have


      excessive levels.  I guess that is one issue I was


      getting to.


                Dr. Laughren, I wanted to come back to the


      point Dr. Pine was making.  I thought Dr.


      Reisinger's point in the open session yesterday was


      a very interesting one, which is that you would


      have to undergo some kind of computer-based


      learning program or some kind of program that


      authorized you to prescribe psychoactive drugs.


                Certainly, we have to do computer-based


      CBLs for all kinds of things in our hospitals and


      in other areas nowadays.  That had a real


      attraction to me, and I guess the question is what


      kind of authority does the FDA have in an area like


      that, can you say that anybody that prescribes


      SSRIs must do a computer-based learning program on


      line, or is that something that the professional


      societies take on?


                You offered several options, black boxes,




      revised label warning, but is this a potential




                DR. TEMPLE:  We can certainly recommend


      things like that.  Every labeling for a cancer drug


      says that this should only be used by people who


      are trained in oncology. That comes with no


      enforcement on our part except that people may be


      anxious about the consequences if they don't have


      that training.


                A labeling recommendation is certainly a


      possibility.  A step further to limit the drugs to


      people who have been given that way, those are very


      iffy questions, and it is not clear whether we can,


      in fact, do that.  There would have to be a debate


      about it.


                There are some examples of fairly


      interventionist activities.  As you all know, you


      can't get clozapine unless you have a white blood


      count, so no blood, no drug.


                There are not a whole lot of other


      examples like that, but there are other cases where


      patients must be given a form that lists what some




      of the adverse effects might be, and things like


      that.  You have to weigh the risk you are concerned


      about with the burdensomeness to the community and


      to the medical profession of those kinds of




                Putting something in labeling about what


      you should know doesn't carry those kinds of


      concerns, so if something sensible, suggesting that


      people ought to be trained in a certain way seemed


      like a reasonable thing, we could certainly


      consider that.


                DR. TRONTELL:  I would just like to add on


      to Dr. Temple's comments, because the FDA regulates


      drugs, but doesn't regulate the practice of


      medicine, and we walk a fine line in terms of


      dealing with some drug products where we may feel,


      as with clozapine, that only very tight controls on


      prescribing and dispensing and use of the product


      are allowed.


                There are a very small handful of drugs,


      they tend to be the exception rather than the rule,


      where training has been required as a condition of




      approval.  One product in particular is the drug


      product dofetilide, where, in fact, training is


      required for pharmacists or clinicians.  There is a


      highly structured way in which that product can be




                Again, those have tended to be reserved


      for situations where we feel the drug cannot be


      safely used without that very high level of


      precaution.  It is extremely difficult to put those


      in place for products that have already been


      marketed and used by professionals.


                DR. CHESNEY:  The public sees your role I


      think in a much broader perspective, as we heard


      yesterday, and I think that is something that is


      useful to clarify as to where your limits are.  You


      mentioned there is a fine line, and I think that is


      what we are all looking for, is where does your


      authority end and that of prescribing physicians


      begin, I guess in a sense.


                DR. TRONTELL:  I don't think we yet have


      an answer.  I think we always have the authority of


      our agency and hopefully, our ability to persuade




      individuals, but I think that the actual legal


      authority to do some of these is a matter of debate


      within and outside of the agency.


                DR. GOODMAN:  Dr. Nelson.


                DR. NELSON:  I would like to return to the


      topic of the incentives on the part of industry to


      perform well-conducted trials.


                There has been a lot of discussion about


      the evolution of the written request and about the


      improvement with three-arm studies and changes in


      the ability to request that, but my understanding,


      I am interested to know if this is accurate, is


      that there is still two potential linkages that


      don't exist that might decrease the incentive to do


      a well-conducted study, and that is, absent safety


      concerns, there is no tie to putting any efficacy


      information in labeling, so that they receive


      exclusivity if a labeling change occurs.


                Second, is that there is no link of


      exclusivity to a well-conducted study unless that


      has changed with written request, since I read them


      on the current web site, there is one asthma study




      where there was members of the drug group that had


      no drug level, members of the placebo group that


      had measurable drug levels, and the FDA concluded


      that the data was uninterpretable, but


      nevertheless, exclusivity was granted.


                I am wondering, is that a problem with the


      written request that is now fixed, or is there


      other solutions that would need to be put into


      place, such as legislation, to address those two,


      what I perceive as gaps.


                DR. MURPHY:  I think there was significant


      discussion about how exclusivity should work,


      should it be only if the product is approved.  I


      was not privy to those discussions, but I know they




                The reason for why it was put in place the


      way it is, I can't give you, Dr. Nelson, but I can


      tell you that one of the explanations I have heard


      is that there was such little data, and FDA was


      given the authority to define the trials, so again,


      as you have heard, we would like to improve, and we


      know we have to learn from what trials we have,




      that by providing FDA the authority to define the


      trials, that they hope that the trials would be,


      you know, of the best that they could be, and that,


      therefore, we would learn from the trials even they


      were failed, because that is important information,


      failing is important.


                So, I guess what you would say, you are


      asking if, and that is in a number of our labels,


      and that is a whole other discussion, but in


      situations, you know, we know that is the only


      study we are going to get and this is it, failing


      is put, that they failed to show effectiveness has


      been put in the label in a number of situations,


      and certain dosing or safety information.


                As I said, about a fourth of the time, we


      are describing, even irrespective of whether the


      study is positive or negative, we are finding


      safety signals, you know, important dosing


      information, and we are able to put that


      information in a label.


                The intent is that the information that is


      obtained, whether the product is proven to be




      effective or not is important, and that safety


      information, et cetera, would be obtained.


                So, that is the best explanation I can


      give you as to why it is set up the way it is right




                DR. NELSON:  I understand, but let me


      focus my question, I guess.  Right now the efficacy


      or lack of efficacy data is not in the existing


      labeling that we are discussing, so, for example,


      just to pick one, paroxetine, there is five


      studies, and the pediatric use just says it has not


      been established.


                Although that is a true statement, it is a


      bit misleading because many people interpret that


      to mean the studies hadn't been done.


                The other question is you could ask them


      to do a three-arm active control study, but if they


      do it badly, do they still get the money?  Even if


      they have done it, if they do it badly, do they


      still get the money?


                DR. GOODMAN:  Dr. Temple wants to respond.


                DR. TEMPLE:  If the written request says




      you need to do a three-arm study and need to show


      that the trial has assay sensitivity, that is, the


      ability to distinguish active drugs from inactive


      drugs, and the Prozac arm doesn't beat placebo,


      then, they would have failed to meet the


      requirement of the written request.


                We couldn't do that before, as I said,


      because we didn't have a known active control, so


      we wouldn't have known what to say.  So, in that


      case, the incentive to do a proper study becomes


      quite clear.  If they don't do a proper study, and


      succeed in showing that, they would not get




                In other cases, we have insisted that the


      variance be such that for, say, a blood pressure


      drug, an effect size of 3 or 4 millimeters of


      mercury could be detected, so if the whole thing is


      done sloppily and they could not have detected such


      a thing, then, they would not get exclusivity.


                Some of the other things, however, that


      you mentioned, don't have an obvious remedy.  I


      mean I guess following the example you said, we




      could say, oh, by the way, people should have blood


      levels showing that they took the drug.  Well, we


      hadn't been smart enough to think of that, and


      maybe that is something we should be adding, that


      is, some kind of compliance check.


                That, I don't think has been part of


      written requests to date.  That doesn't mean it


      couldn't be.  The test that Congress imposed is


      that if you comply with the terms of the written


      request, you get exclusivity.  That means if we


      weren't smart enough to ask a question, that is not


      considered their fault, and they are supposed to


      get it.


                DR. MURPHY:  And we have denied


      exclusivity where we thought the trials were done


      sloppily, and actually, sometimes when the sponsor


      said, well, we know you asked for this, but we


      didn't think it was correct to keep going, so we


      didn't do this for some reason, and we said, no,


      you should have come in and talked to us about why


      you weren't going to do it, you didn't do it, we


      told you, you need to do it, sorry, you don't get






                So, what I guess we are trying to say is


      if it's really sloppy, and they don't do what we


      tell them, we deny them exclusivity.  The problem I


      think we are dealing with here is that we all are


      learning how to better do the trials, and your


      other question about whether that should go in the


      label, the negative information should go in the


      label, is a whole other discussion.


                DR. GOODMAN:  I have a list of seven other


      committee members who wish to speak.  After we give


      them that opportunity, I am going to ask Dr.


      Wysowski to come up to the podium.  We had asked


      her to follow up on something from yesterday.  Is


      there somebody else that has a burning--we have one


      more and that is it--two more, that's it.


                Dr. Irwin.  His question has been


      answered.  Thank you.


                Dr. Rudorfer.


                DR. RUDORFER:  Yes, thank you.


                I would just like to revisit a couple of


      issues that concern me at the front end of these




      studies, and I recognize everyone from the FDA is


      pointing out that this is a learning process, on


      the other hand, we are faced with the dilemma of


      having these particular trials to deal with.


                The dosing question that was just


      discussed brings to mind a concern I have related


      to how the suicidal events we have been looking at


      were ascertained.


                As I understand it, for the most part,


      these were from adverse events questionnaires and


      surveys.  Is that correct?  I mean there was no


      particular suicidal scale?


                DR. LAUGHREN:  Well, all of these trials


      included standard depression rating instruments


      like HAM-D or CDRS, and so forth, and there is a


      suicide item in each of those instruments, and that


      is part of what we analyzed.


                But the problem is we don't really know


      how those were applied.  My guess is that most of


      the event data we are dealing with were spontaneous


      report or general questioning rather than specific






                That is really one of the areas that we


      are trying to explore with Columbia to try and work


      on a more specific instrument for improving


      ascertainment for suicidality, but no, in these


      trials, I don't think ascertainment was very




                DR. RUDORFER:  My question, as we deal


      with these data, would be this.  I appreciate the


      very dedicated and elegant work that both the FDA


      and Columbia have applied to these findings.  The


      question I have relates to the issue of the active


      drug versus placebo groups.


                Since it sounds as if much of the data


      were spontaneous reports or I assume perhaps


      discussion between the raters and subjects, or the


      investigators and the subjects, I am wondering if


      part of this is not dependent on the assumption


      that the blind was kept intact throughout the


      studies, and I wonder if we have any measure of


      that or any sort of quality control on that issue.


                DR. LAUGHREN:  No, we have no idea of


      that.  That is typically not something that is




      really ascertained.  It is the assumption, but how


      would you check on that?


                DR. RUDORFER:  In some studies, patients


      and raters are asked at some point.  I mean here, I


      am just wondering if, in fact, if a patient


      volunteered that, for instance, they were


      experiencing some side effects, they come in, the


      rater asks how are you doing this week, and their


      first comment relates to GI distress or something


      that sounds like a side effect, if they simply


      don't get more attention, in other words, maybe


      there is more discussion, maybe there are more


      questions asked as opposed to a patient that comes


      in and say, gee, I am feeling pretty good, I don't


      seem to have any side effects.


                Again, that would not obviate the fact


      that if we find signals, then, the signals are


      present.  I guess I am just concerned about the


      active drug versus placebo difference.


                DR. GOODMAN:  Let me interject.  I don't


      think I am as concerned about the unblinding, but


      your question raises at least in my mind the




      possibility that in the data, is it possible that


      we would see other somatic symptoms, more side


      effects reported in those patients, who also


      reported suicidality than in the opposing group,


      was there any attempt in the data to look at


      whether there were any other--was any other


      increase of adverse experience outside the target


      symptoms of suicidality in those patients who


      reported suicidality, the reason being that if


      there was, that would suggest it was part of a


      larger behavioral syndrome that was being induced


      by the medication.


                DR. LAUGHREN:  Our analyses had to be


      limited by what we had in our database, and we had


      to design this database late last summer.  We


      didn't anticipate all of these questions.  As it


      was, the database we had was a very time-consuming


      process to put together.  It took a number of


      months to get it.


                They are all good questions, but we don't


      have all those answers, but I agree that


      ascertainment for suicidality was not optimal here.




                DR. GOODMAN:  But the question is at this


      point, could you go back to that same data and look


      to see if there is a higher rate of other adverse


      experiences reported in those patients who were


      also identified as experiencing or exhibiting




                DR. LAUGHREN:  Not without designing


      another database and going back to the companies


      and waiting for a number of months, and I am not


      confident enough in the quality of the data we have


      here that that would justify that additional




                I mean again, these are all good


      questions, but we are faced with making a decision


      at this point in time with what we have, and we are


      asking the committee's advice on what you think we


      can do now based on what we have done.


                DR. GOODMAN:  No, I agree with that, I


      understand that, but we were also asked what other


      advice we would give in terms of future research or


      studies or data that we would like to see.


                DR. TEMPLE:  Tom, we did look at the




      association with certain kinds of things, the


      activation syndrome, things like that, right?


                DR. LAUGHREN:  We included in our database


      two other symptoms, hostility and agitation based


      on the preferred terms that the companies used, and


      again, we haven't looked, I suspect that there is


      variability across different companies in what


      actually got subsumed under those two things.


                There are the only two other events, and


      we don't even have the timing for that.  All we


      have is an indication of whether or not, at some


      point during treatment, a patient experienced


      agitation or hostility.  We don't have all the


      other somatic kinds of things that you are alluding


      to.  That would mean going back and trying to


      create another database.


                DR. GOODMAN:  Dr. Temple.


                DR. TEMPLE:  Let me just mention one other


      thing that has come up briefly and that Dianne


      touched on, and that is inclusion of negative


      results in labeling.


                As Tom has explained at the previous




      meeting and now, as a general policy, we don't


      usually put in labeling the fact that a study


      hasn't worked, because we don't think that proves


      that it doesn't work.  It just means that that


      study failed.


                But we are actively thinking about that


      policy for the pediatric part, because the whole


      point of doing the studies was the possibility that


      adults and children are different, otherwise, you


      wouldn't even think about doing that whole program.


      All I can say is we are actively thinking about it.


                It is not an easy to thing to do, however,


      because what you would want to say could depend on


      how good you thought the study was, and then there


      is the conundrum of what do you do if there is one


      study that says yes and one study that says no.


      That is virtually somebody a claim, which we really


      wouldn't want to do if it wasn't merited.


                So, I am not going to suggest that this is


      easy, but we are reconsidering this whole thing,


      because the whole point of the Best Pharmaceuticals


      for Children Act is to find the data and see




      whether drugs work in children, and not putting


      anything in seems funny, so we are reconsidering




                DR. GOODMAN:  Dr. Perrin.


                DR. PERRIN:  Part of my question Dr.


      Chesney eloquently asked before, but I wonder if we


      can get access to the wording that you used for


      cancer drugs as perhaps a guide to us for our




                My other quick question, I think back to


      one of the FDA group is am I hearing you right that


      if you have a drug that has been shown to be


      efficacious in a particular indication, that all


      trials requested in the future require an arm that


      includes that drug?


                DR. TEMPLE:  I am not ready to say that


      one would always do that, there are other ways to


      try to assure quality, but in this setting, it is


      reasonable to assert that we need to know whether


      your trial was an adequate test of whether this


      drug worked, and the only way we know to be sure


      that it is an adequate test is to have an active




      control, and to have that active control be


      distinguishable from placebo.  Then, you know this


      is a trial capable of showing things.


                We have determined that our future written


      requests will include a requirement for a three-arm


      trial, because that's the only way we know to be


      sure that the trial is a trial that is capable of


      showing what the answer is, and we want to be sure


      we get the answer.


                This comes up in written requests all the


      time, how much assurance do you have to have and


      how do you gain that assurance that the trial is a


      useful trial, and the reason it comes up is the one


      that everyone has alluded to, we don't think people


      are deliberately trying to mess things up, but the


      incentives to do a really good trial are greater


      when you have to win.


                DR. PERRIN:  I am a little confused.  As a


      clinician, you know, typically, if I am looking at


      a new medication, I want to know that it is better


      than current therapy.  I mean all of us are really


      interested in that.




                There are a number of pediatric drug


      trials, not in the area of antidepressants that I


      am aware of, where new drugs come on the market,


      approved by the FDA, where there are only


      drug/placebo trials, and not trials comparing the


      new drug with currently approved FDA medications.


      That is where I am confused.


                DR. TEMPLE:  Good question.  There are two


      possible uses for having an active control.  One is


      where you want to compare the two therapies.  Now,


      to do that, you would need a very, very large


      study, because you would be interested in even


      modest differences.  That is not what we are


      talking about.


                We are talking about the use of the third


      arm to show something about trial quality.


      Actually, a third arm is extremely common in


      depression trials now, because if the trial fails


      to show that your drug is better than placebo,


      there are two possibilities.


                One is that your drug is no good, and the


      other is that the study was no good, and it is very




      important to somebody developing a drug to know


      which of those two things it is.


                If the trial shows that Prozac, say,


      works, and your drug doesn't, you get rid of the


      drug.  If the trials shows that neither Prozac nor


      your drug works, you do another study.  So, it is


      extremely important.  But the two purposes of the


      trials are quite different.


                To actually do a comparison and try to


      detect a small difference, you would need very,


      very large groups. That is an unusual thing for


      people to do, and usually, the drugs can't be


      distinguished.  It is very hard to do that.


                DR. GOODMAN:  Dr. Temple, I heard you say


      before, if I heard correctly, that incentives are


      different when you need to win.  Were you referring


      to the conditions of the six-month exclusivity


      arrangement, and if you were, if the incentives


      were different at that time, would you predict any


      difference in the design of those trials or the


      conduct of those trials?


                The reason, let me say, I think that many




      of us keep harping on this point, is not so much


      because we think that the suicidality data would


      have turned out differently. I think it is the


      absence of a benefit, the absence of efficacy that


      at least I am concerned about, because that is


      mostly what we have in assessing benefit or those


      trials, and we only have 3 out of 15 that are


      positive, so if there was something about the


      conditions in which those studies were designed or


      conducted that might have negatively impacted the


      outcome, I would like to know it.


                DR. TEMPLE:  Well, Russ and Tom need to


      respond, but we haven't seen anything in the design


      of those trials as written in protocols that makes


      them look any worse than any other trials.  They


      seem to have reasonable size, so there is nothing




                But, you know, this is an issue that comes


      up when you do so-called "non-inferiority" trials.


      The incentive, you know, the point of such trials


      show no difference between treatments, and as I


      have written repeatedly, that is not a good




      incentive to give people.


                It doesn't stimulate the kind of optimal


      behavior that you want, which is stimulated by the


      need to try to show a difference between


      treatments, and that is a problem here if you don't


      need to win, to gain exclusivity, and I don't


      disagree with the idea that you shouldn't need to


      win, the point is to get the data.  That is why the


      BPCA was done that way.


                On the other hand, you do want a good


      trial, and one way to guarantee that the trial is a


      good trial, however the drug comes out, is to be


      sure that it is capable of showing something we


      need to be true, namely, that Prozac seems to work.


                DR. KATZ:  Can I just add?  One thing you


      need to remember about studies done in response to


      written requests is that they are very time


      sensitive, or at least it's possible that they are


      time sensitive.


                What I mean by that is you only get


      exclusivity if your study reports, your supplements


      containing the data come in while you still have




      some residual patent life left or exclusivity left.


      So, they have to be done within a particular time


      frame.  In fact, the letters that we send, the


      written requests include a date by which the


      studies have to be submitted.


                So, in some cases, there is at least


      potentially motivation to do studies rapidly, so


      that they are done and study reports are written,


      and the supplement, which includes these data, are


      submitted in time, so that they can still get their




                So, one at least potential question that


      has been raised is enrollment so rapid or does it


      need to be so rapid into these trials that maybe


      not all the patients are adequately diagnosed, and


      maybe they have something other than depression.


                It is very, very difficult for us, if not


      impossible for us, to be able to independently


      corroborate diagnoses in something, in conditions


      like these, so we, of course, take it on faith that


      they got the right patients, but maybe, for


      example, because of time constraints, they didn't




      get the right patients, and that might contribute


      to a negative finding even if the drugs were


      effective in a true population.  So, that is one




                DR. GOODMAN:  I think that is a fair


      answer.  Anybody that wanted to comment


      specifically on that?  Dr. Marangell.


                DR. MARANGELL:  Are you aware, is there a


      greater proportion of non-academic sites in these




                DR. MURPHY:  I don't know that we have


      looked at that.  I mean I know that there are


      definitely, in some of these studies, very, you


      know, academic sites that have been involved in


      numerous or actually well-known to us




                I do want to make one thing again.  One


      thing that every division within FDA is told, when


      writing their written requests, they are asked a


      number of questions - what is the public health


      benefit, what are the trials to get to that public


      health benefit, and you are not to take into




      consideration--and most of the time they don't even


      know because you would have to go into a lot of


      patent law--they don't know or are told to not pay


      any attention to when the patents expire or the


      exclusivity marking would go out, they must look at


      it only from what are the trials that they need to


      have done.


                Now, what is being told to you, though, is


      that--and we have written requests where the


      companies come back and say, well, that is not


      going to help us, because you put a date on here


      that it was due by 2007, and our patent expires in


      2005, and we have said, you know, we are sorry, we


      need these kind of trials.


                Now, would, in that situation, a company


      try to compress by getting more sites or, you know,


      whatever, would they try to do that trial in a


      different way?  Yes, possibly.


                I mean that is what we are trying to


      explain the balance between the way the process is


      set up, it is not to be driven by the time when the


      patents are expiring, the marketing exclusivity is




      expiring, the divisions are to determine what the


      studies are that are needed to the best of their


      knowledge at that time.  They are to design those


      studies to answer those questions.


                Do we try to be reasonable and say, gee,


      we would like a 10-year follow-up study, but we


      don't ask that for other--you know, we have to be


      reasonable within the realm of what we would


      normally ask for, for an approval product.


                Again, though, maybe we can be--we say we


      have to get this information because children grow


      and will go through a period where that might be




                DR. GOODMAN:  Thank you, Dianne.


                DR. MURPHY:  So, you have to ask for


      additional information, you might not, for adults.


      I am trying to explain the process.


                DR. GOODMAN:  I will accept some questions


      out of order if they are on this specific topic.


                Dr. Rudorfer, I think  you had one.


                DR. RUDORFER:  I just wanted to follow up


      on Dr. Katz's comment about whether we are looking




      at the right patients, which was an issue we


      discussed some yesterday.


                Just one point that I want to follow up


      on.  It is clear that in young people who present


      with major depression, there is a disproportionate


      number who go on to develop bipolar disorder, and I


      think one concern that we expressed yesterday was


      that the trials are very inconsistent in that


      especially in terms of accounting for family


      history, it sounded as if in some trials, a subject


      could literally be brought to the clinic by a


      parent who has bipolar disorder, and yet the child


      could be included in the trial.


                I realize this question might be, as we


      said, a little out of the box.  I would think that


      if there is any way to encourage the companies to


      actually try to find some of these thousands of


      young people and see what has happened to them in


      the 5 or 10 years since they were in the study, it


      could be tremendously informative simply in seeing


      whose longitudinal course has played out as what we


      recognized in young adults as major depression, who




      developed bipolar disorder, who developed some


      other disorder, and go back and re-look at, for


      instance, those who after the fact are confirmed to


      have the diagnosis that we thought they did on




                DR. GOODMAN:  Dr. Pfeffer.


                DR. PFEFFER:  Yes, I want to I guess


      continue on what Dr. Rudorfer is saying, because I


      think diagnosis is critical, and I think we can


      learn something about this that we have learned a


      little bit about depression in other realms, too,


      namely, that children are, in fact, different than




                So, what appears to be adult depression


      and what appears to be childhood depression may, in


      fact, be quite different, so that perhaps a lack of


      efficacy in most of the studies tells us something


      about the nature of the developmental course, first


      of all.


                I agree with what you are saying about the


      potential for bipolar.  That is one issue that is


      crucial, I think, in terms of maybe the adverse




      response that children are having, but also if we


      think of the number who had some suicidal thinking,


      that also might be a subgroup of the children who


      are in these studies also.


                The other part that I want to mention is


      that when I gave that talk last meeting, I talked


      about the complexities about what looks like


      depression in children, and not only course and


      family history, but life event circumstances, and


      that has not been looked at.


                So, for example, children who might have


      been having immediate family turmoil and looked


      depressed, that is an issue that might have led to


      some resistance in response, for example.


                The other point I would like to make is


      that we hear from some of our childhood


      psychiatrist colleagues yesterday who advocate to


      not ban use of these drugs, because they do see


      efficacy, and it may very well be that in their


      practice, with very careful assessment, careful


      diagnosis, they are selecting the subgroup of


      youngsters who potentially could respond, and




      respond very, very well.


                So, I think the question of diagnosis is


      crucial, which means that in terms of the study


      design, in a way, who has the most reliability to


      make a diagnosis, and what kinds of questions


      really are being asked and what data is being


      collected that might help us even look at


      predictability of response, and I don't think we


      have that, such as life events, such as family


      history, such as perhaps other issues that we would


      need to come up with and understand.


                DR. GOODMAN:  Thank you.


                Dr. Gorman.


                DR. GORMAN:  A lot of us keep saying that


      children are different, and I don't think it should


      come to us, then, as a surprise that children may


      respond differently to medicines than adults do.


                I think I would be more concerned about


      the efficacy of these trials if they were all


      unidirectional in the sense if they had all failed


      or they had all succeeded. I have heard nothing


      from the FDA to this point that says that the




      playing field has been tilted in any way since one


      of these drugs in this class, which may not


      actually be a class, but it seems like it might be


      a class, actually works for children in the bar


      that the FDA sets up.


                So, I am now going to address my single


      question to the rushing hypothesis.  After Monday


      Night Football last night, I like the rushing


      hypothesis.  There is one small question I have to




                Prozac was the first mover in this field,


      therefore, I assumed it came to market first, and


      probably then had the least time before its patent


      extension.  Is that a safe statement?


                So, it came to market first.  Did it have


      the smallest amount of time?  It was the first to


      go off patent, yes or no?


                DR. TEMPLE:  I believe it was the first


      one to go off patent by a little bit.  It is off


      patent now, and only, I don't know, are any of the


      others off patent?  So, we know it was the first


      off patent, which happened sort of a year ago.




                DR. GORMAN:  So, that would run


      counterintuitive to the rushing hypothesis, because


      Prozac had to get there first, and therefore, seems


      to have had the least time and would be the most


      likely to be rushed to get labeling.


                DR. TEMPLE:  Some of the trials were done


      before this program even started, I think, and they


      were done a long time ago.


                DR. LAUGHREN:  One of the trials was done


      by Emslie several years before, and the company


      obtained the data and submitted those data as part


      of that supplement.  It was done in the early '90s,




                DR. KATZ:  Right.  The studies that we


      asked for in the written requests don't necessarily


      have to be done or initiated after the written


      request is written.


                If they have a study that is very old, but


      that meets the criteria that we put into the


      written request, they can use that, so they don't


      have to be done specifically in response to the


      written request, they have to meet the criteria




      that we lay out, and it can have been submitted to


      us either before the written request.


                But they could have done a study many,


      many years in advance before we even contemplated


      written requests.  If they met the criteria, they


      can submit it in response.


                DR. TEMPLE:  But, also, remember it's a


      hypothesis.  We don't know why those trials fail.


      It could be that children really don't respond.  I


      mean we don't know the actual answer.


                DR. GORMAN:  Well, I would love to be in


      the position where I can say something nice about


      the pharmaceutical industry, because it sometimes


      seems to happen so rarely, but if Lilly did the


      trials before there was the potential for financial


      gain, because all they were doing was looking for


      labeling in children without the congressionally


      mandated reward for that particular behavior, and


      therefore had these studies in place, maybe the


      rushing hypothesis fails, but there is another


      hypothesis that could be generated from that.


                DR. LAUGHREN:  Again, in fairness, the




      Emslie trial was funded by NIMH.  This was not


      sponsored by Lilly.  They went back and obtained


      the data, and they did subsequently an independent


      trial that also succeeded.


                DR. GOODMAN:  Dr. Newman.


                DR. NEWMAN:  I think Dr. Temple did a good


      job of explaining why, if you have an active


      treatment arm, the trial is likely to be of higher


      quality when asked to demonstrate that difference.


                I wonder if another approach to motivating


      high quality studies would be to require that in


      order to get the exclusivity, that the trial be


      written up in a way that passes some sort of peer


      review and be published.


                That way, even published on FDA web site,


      that way, if the trial is sloppy and finds the drug


      doesn't work, those results would not be buried,


      they would become public and that I think would


      provide some motivation to do a good job.


                I am a little troubled.  I wrote down a


      quote from Dr. Murphy.  It said, "If a study is


      negative, we don't talk about it."  I think if




      that's the case, then, there is a lot less


      motivation to do a really good job on the study.


      Why not require the studies be published, be


      written in a way that it is of sufficient quality


      that they can be interpreted, and then maybe the


      quality would improve.


                DR. MURPHY:  But for peds now, we do.


      That is the difference.  That statement was for


      adults.  For pediatric studies, well, I should say


      it is for pediatric studies that aren't done under


      exclusivity, but for pediatric studies done in


      response to these written requests, we now are


      mandated to make them public whether they are


      approved, they are not approved, or even if they


      are withdrawn.


                DR. TEMPLE:  But you are also talking


      about a level of detail in the presentation


      sufficient for people to really get into was it a


      high quality study, and things like that.


                DR. NEWMAN:  Why not?


                DR. TEMPLE: it is a fairly good question.


      We don't believe we have authority to insist that




      things be published.  We get full details, we get


      all the data.


                DR. NEWMAN:  But you could peer review,


      you could peer review them.  You could send them


      out and say is this something that is publishable,


      and have people at FDA, who I am sure are very good


      at that, say no, this gets an F, you know, this is


      not good enough, send it back, or you don't get the




                DR. TEMPLE:  Well, our reviews, when we


      approve something, you get on our web site the


      contents of our reviews, so you get to see what we


      thought of all of the studies.  If we do not


      approve, however, we don't believe we have


      authority to make those data public, so you don't


      get to see our reviews.  That is our legal


      interpretation of what confidential commercial


      information is, and I can't rebut it or comment on


      it.  It's a legal determination.


                DR. GOODMAN:  Dr. McGough.


                DR. McGOUGH:  Just on that point, does the


      FDA now have the authority, if you wanted to, to




      put negative studies in the pediatric label, do you


      have the authority or does Congress have to do


      something for you to get the authority?


                DR. TEMPLE:  We have authority.  What I


      was saying before is--and we are, as I said,


      considering whether in the pediatric case, we


      should do that.  In other cases, we would, too, if


      we thought it was important to point out the


      negativity, and the negativity was a true bill.


                It is just the fact that if one study


      fails, doesn't necessarily say that something


      doesn't work, so we have been somewhat reluctant to


      just do that until it was convincing.


                But as Dianne said, we are actively


      thinking about amending that policy for the


      pediatric setting where the whole point of getting


      the studies done was to see how the drugs worked in


      children, for the very same things that they have


      been studied for in adults.


                It is a little different from novel use or


      something like that.  The BCPA calls for studies of


      the exact same things that have been studied in






                DR. MURPHY:  And we are putting negative


      information in some of the labels already for other


      types of products, but because of the complexities


      that you have heard, it has been the policy for


      antidepressants for children not to do that at this


      point, but I think, as has been mentioned, it is


      being reconsidered.


                So, we have, and I have got all the labels


      here that we have done, we are putting that


      information in some of these labels.


                DR. POLLOCK:  For the new approvals.


                DR. MURPHY:  No.


                DR. TEMPLE:  For where we grant




                DR. MURPHY:  Right.  In other words, if a


      product comes in and doesn't work, we have put that


      information in some labels where we think it is


      very clear-cut, you know, eight more studies is not


      going to change this for whatever reason, and we


      put that in here.


                We have also put in information where it




      hasn't worked where there are safety issues


      involved, and we are not clear what those safety


      issues are, but we want to tell you about them.


      So, those are in the label, too, even when it


      wasn't approved for that indication.


                DR. GOODMAN:  Dr. Santana.


                DR. SANTANA:  I have a comment and then a


      question that really relates to a point that Dr.


      Chesney made about issues regarding the boundary of


      practice and FDA regulation.


                My comment is that there was some comment


      related to oncology and issues, how we deal with


      some prescription and safety issues in oncology,


      and I think we have to recognize that pediatric


      oncology is unique in this country and that most of


      the trials, even the exclusivity trials, of which I


      have participated in some in oncology, are really


      under the umbrella of research centers and academic


      centers. Very little pediatric oncology is done in


      the private practice.


                So, by force, you are now dealing with a


      group of individuals that are more specific and




      more geared up to looking at issues that


      potentially could be relevant, whereas, I think in


      the other pediatric arenas that we are talking


      about, that doesn't occur.


                So, I think it would be a misnomer to use


      pediatric oncology, maybe it should be the gold


      standard, but I think we need to recognize that it


      is kind of unique in the way it is practiced in


      this country.


                Having participated in some of the


      exclusivity oncology trials, I can tell you that


      they are at the same caliber and at the same


      rigorous structure as any of our other oncology


      trials are in the cooperative group setting, et


      cetera, et cetera.  So, that was just a comment to


      clarify the pediatric oncology issue.


                I want to get back to patients and


      practicing physicians, because we have been talking


      a lot about study design and how to analyze data.


      I want to get back to the issue of patients,


      parents, and practicing physicians, and this


      boundary of what the FDA can regulate and cannot




      regulate in terms of the practice of medicine.


                I was struck yesterday by many of the


      testimonies from parents and families, and


      actually, there was even a gentleman who showed a


      slide, in which his child was given the medication


      as a free sample.  I am not sure that all these


      whistles and alarms that we put in labels are


      really going to work unless somehow that practice


      stops for certain medications that we think


      potentially have a greater risk.


                I wanted the FDA to address the issue


      historically.  Is there any ruling that you guys


      can impose or potentially think about, about how


      these medications are given without prescriptions,


      that is, either free samples or in the marketing


      world, so you could comment on that.


                Secondly, does the FDA have any historical


      data on successful programs?  There was a mention


      that maybe a med guide to parents and families


      would be a way to address some of the safety issues


      and bring people to a better level of






                Can the FDA comment on any successful


      programs that they can point where this has truly




                DR. TEMPLE:  Just on the free sample


      thing, my understanding, but I don't really know,


      was that a physician did use a free sample to, you


      know, like start the child out.  That is not


      without a prescription, it may not have been well


      done and may not have had adequate follow up, but I


      am not sure that it is the free sample that is


      involved, it is the lack of follow up that was


      described that seems like the problem.


                It is not easy to know how successful our


      various endeavors have been, and Anne Trontell may


      want to comment on that.  Some of them have effects


      that are not entirely what we wanted.  She


      mentioned the program on dofetilide.


                To start, dofetilide is a drug that is


      used to prevent recurrence of atrial fibrillation,


      but it is a drug that causes QT prolongation and


      Torsade de Pointes, and there is no doubt about it.


                The recommendation in labeling, and it is




      enforced by the need to give out various


      information requires that you come into a hospital


      or outpatient facility for a couple of days to see


      what your creatinine clearance is and to see


      whether you are a person who has QT prolongation to


      an excessive degree.


                Then, after that you can go out and be


      treated with it for long-term use in preventing


      atrial fibrillation.


                What appears to have occurred is that that


      is sufficiently difficult, so that people instead


      use sotalol, which doesn't have such a program, or


      quinidine, neither of which are an improvement of


      the situation.


                So, people can avoid some of these things


      if they are inconsistent across the drug classes,


      so you always worry about that.


                There is a very rigorous requirement for


      periodic measurement of liver tests with a drug


      called Bosentan, which is used for pulmonary


      hypertension, and although the drug was quite toxic


      when it was being developed, my most recent




      information is that we haven't had any fatal liver


      outcomes, perhaps a testimony to the fact that


      people are indeed following up these patients.


                Of course, this is a class of patients who


      are very sick and very closely watched.  You don't


      know if that is typical how we are going to do.


                Anne, you want to comment on some of the


      other programs.


                DR. TRONTELL:  Sure.  On the issue of


      sampling, first of all, I think in some instances,


      at the time of product approval, there have been


      informal agreements, but no FDA authority to


      restrict sampling exists to my knowledge, but there


      may be agreement, you know, certainly, we don't


      sample oncology drugs.  There are things that just


      don't happen.


                On the issue of what is a successful


      program, I think we struggle in the agency, because


      good evaluations have not been done on a standard


      basis.  We have the best information for those


      programs that are most restrictive, programs like


      clozapine or programs like the one for thalidomide




      to prevent pregnancy exposures.


                So, the available data to us to tell us


      what works tends to be only in those extreme cases


      where we have put, as Dr. Temple just described,


      for Bosentan, you know, very severe restrictions.


                If you are asking for specific information


      about medication guides, I think we have in the


      general literature evidence to suggest that good


      education certainly facilitates good behaviors, but


      I don't think we have any evidence yet that it


      guarantees that they do take place.


                So, if you had questions about the


      intermediate ones, I think for the most part, we


      don't have information about those specific


      educational programs or the reminder ones.  Not


      uncommonly, education is applied in the context of


      these very restrictive programs that I just


      described, and again, teasing out what the


      education alone does is very difficult.


                DR. GOODMAN:  Dr. Katz.


                DR. KATZ:  You are hearing the


      difficulties that we think we have with regard to




      imposing various sorts of restrictions although


      again, there are ways to do it although they may be


      very difficult to implement.


                But it occurs to me that it might be


      particular difficult in this case because the use


      we are contemplating in all but one case is off


      label, and I don't even know what the implications


      of that are.  Certainly, there are legal questions


      about what you can say and what you can restrict


      with regard to off-label use, and I don't think


      that we have thought through all the implications


      of that.


                DR. SANTANA:  So, as a follow up to that,


      since we last met in February and there was a


      recommendation to do something with the label that


      occurred and some warnings, has the Agency


      monitored the change in practice?


                I heard a number yesterday of 10 percent.


      That is prescriptions, but has that been rigorously


      looked at, that there was an impact of that


      modification that translated to something very






                DR. TRONTELL:  I will ask either Michael


      Evans or Judy Stafford from the Office of Drug


      Safety.  All we have really been able to monitor


      since the last advisory committee is volume of use,


      but they do have some information on how that has


      changed recently.


                DR. GOODMAN:  Ms. Griffith.


                MS. GRIFFITH:  I would just like to pursue


      this for a moment with respect to the patient and


      physician relationship.  When Dr. Chesney was


      proposing perhaps some sort of a computerized


      programming or education, or even with respect to


      these med guides, when Dr. Temple suggested that


      perhaps there would be, you know, a mechanism much


      like you have for other drugs, that the patient and


      practitioner would be signing a consent form


      outlining the risks and benefits, I want to


      understand the reason that you thought that that


      might be too great a deterrent to pursue, simply


      because from my perspective as patient rep and


      parent, it seems to me that in the course of any


      treatment process for any severe illness, which as




      we all understand depression is, you are often


      asked to look at the risks and to sign some


      statement to the effect that you understand what


      these risks are.


                You even have to do that if you get a shot


      of botox, not that I know, but it just strikes me


      you have put the parents now in the position of


      actually doing the risk-benefit analysis.  That is


      where we all are.


                If by providing the families with the


      statement that these risks are indeed serious, I


      think that what we heard yesterday was how little


      awareness there was on the part of the parents that


      these drugs could be lethal in certain cases.


                I am arguing for more information rather


      than less, not more restrictions, and I agree with


      the point that Dr. Santana made, that how often


      does a parent either open the box and read the


      information or understand it.


                So, if it is very clearly stated between


      the doctor and the patient or the parent, I think


      it goes a long way to satisfying the need to know




      for parents.


                DR. TEMPLE:  There are gradations of


      information, and we wrestle with how to do that


      without being an attempt to be informative, but not


      disruptive, so that, for example, a lot of drugs


      have what are called med guides.  These are patient


      labelings that are actually, under the law,


      supposed to be given out by the pharmacist.


                My own view is that if you don't make it


      part of the unit of use package, you might as well


      not bother, but in any case, we know that there are


      ways to get that information to patients either


      through the proper functioning of the pharmacy or


      by making it part of the package.


                An enormous additional step, which has


      been done in some cases, but, you know, thalidomide


      is a level of risk that is sort of in its own


      category, where there is a requirement that the


      patient and physician discuss all these matters.


      That is a very huge step, and what you might think


      is reasonable for thalidomide, something that is


      used infrequently, you might find more disruptive




      than you want or more difficult than you want for


      drugs that are much more widely used.


                As Russ pointed out, it is particularly


      tricky when the recommended use isn't even in the


      label.  How to write a med guide or something like


      that for something you are not really recommending


      and don't feel able to recommend yet, that may


      sound like a bureaucratic worry, but I think it's a


      serious worry.


                You don't want to warn people and


      simultaneously recommend a use that you don't think


      is recommendable, and any discussion like that is


      tantamount to recommending the use.  So, we will


      need to worry all of those things based on your




                MS. GRIFFITH:  Could I follow up?  I


      understand that and I understand that there are all


      sort of issues involved, liability on the part of


      the physician, but I am suggesting, from the naive


      perspective of the parent, I think of depression as


      every bit as serious as the use of thalidomide


      posing birth defect risks or, as Dr. Santana said,




      you know, the cancer example.  Parents need to be


      informed about those risks.


                I don't think that this is any different,


      frankly, and if it is an extraordinary measure to


      take, I think that it benefits both the


      practitioner and the patient parent.


                DR. TEMPLE:  For oncologic drugs, the


      label says you should be a properly trained


      oncologist.  There isn't anything in there that


      says what you need to discuss.  Patient med guides


      for oncologic drugs is by far the exception, I


      think because it is assumed that there always has


      to be such a conversation in the course of therapy.


                I guess what is being discussed is whether


      there is a common practice of having that kind of


      conversation in someone who is depressed, and


      obviously, we all think that there should be such a


      conversation.  The question is now to induce it and


      what to provide people to help them be sure they


      ask the right questions.


                DR. GOODMAN:  We have a representative


      from the Office of Drug Safety at the microphone. 




      Could you please state your name?


                MR. EVANS:  Michael Evans with the Office


      of Drug Safety.


                With regards to drug use, comparing the


      first six months of this year to the first six


      months of last year, the market rose with all ages


      about 7 percent.  Adolescents and children, in the


      first six months of the year, still comprised


      between 7 and 8 percent of that total.  So, they


      are still widely used in children.


                DR. GOODMAN:  How up to date is that data?


      There must be some sort of lag time, isn't there,


      between when the prescription--


                MR. EVANS:  It is according to IMS Health,


      and this is January through June is what we looked


      at.  This is outpatient prescription data, which


      comprises about 45 percent of all pharmacies in the




                DR. GOODMAN:  Let me make sure I


      understand.  You don't see any significant drop?


                MR. EVANS:  No.  I believe a woman


      mentioned yesterday that they saw a 10 percent




      decline.  We did not see that.


                DR. GOODMAN:  Dr. Irwin.


                DR. IRWIN:  Has the rate of increase


      remained the same or has it leveled off, or what is


      the direction?


                MR. EVANS:  In February, one of our


      colleagues, who gave drug use for 2002, that age


      group was still 7 to 8 percent of the total.  It is


      still the same this year, first six months.


                DR. GOODMAN:  So, there has been no great




                I will take again other questions out of


      order as long as they are directed to a


      representative from ODS.


                Dr. Marangell.


                DR. MARANGELL:  Actually, a comment


      directed to this.  I think it is really critical to


      this discussion that we keep in mind that our goal


      is to protect risk, but also that this is really a


      devastating illness, and I am not sure that I


      necessarily--I don't want to necessarily see


      prescriptions drop.  These people need to be






                What we want to do is make sure that


      people are educated of what to look for early on in


      terms of risk for those people that are at risk,


      children or otherwise.  They are not necessarily


      the same thing .


                DR. GOODMAN:  Other questions for our


      speaker?  Dr. Gorman.


                DR. GORMAN:  Is the data on the level of


      the class or is it on the level of individual




                MR. EVANS:  We looked at the class as a


      whole and then we looked at each individual drug in


      the class.  That was according to IMS Health


      National Prescription Audit, and we also looked at


      the National Disease and Therapeutic Index from IMS


      Health, and tried to apply those percentages to


      outpatient projected prescriptions.


                Only the players changed in that age group


      of children 1 through 17.  Paroxetine was knocked


      out of the top five, but sertraline still is the


      market leader, followed by fluoxetine.




                DR. GORMAN:  So, the information, there


      seems to at this point only be one drug that is


      efficacious in this age range, moved it up the


      ladder, but didn't make it number one?


                MR. EVANS:  Not necessarily.  This is what


      we observed when we were just looking at drug use


      in prescriptions outpatient, and the National


      Disease and Therapeutic Index is an office-based


      survey where a drug is mentioned during that survey


      and linked to a diagnosis.


                DR. GOODMAN:  Can you differentiate


      between the prescriber classes, whether it is a


      primary care physician versus psychiatrist?


                MR. EVANS:  We did look at specialty in


      MBA-Plus.  Psychiatry was still about 65 percent of


      the specialty, pediatrics, somewhere between 15, 20


      percent still.


                DR. GOODMAN:  Dr. Fant.


                DR. FANT:  Could you comment on the


      indications for the prescription, was it all


      depression or other off-label--


                MR. EVANS:  We looked at mood disorders,




      anxiety disorders, ADD, and other disorders.  In


      age group 12 to 17, it still appears there is not


      really any change.  It is still mood disorders,


      which includes major depression, is still


      two-thirds of the indications.


                It looks like in the 1 to 11-year age


      group, perhaps more shift to the ADD.


                DR. GOODMAN:  Dr. Pollock.


                DR. POLLOCK:  I heard rather consistently


      yesterday a lot of concern about the


      direct-to-consumer advertising and the role that


      that has played in this, and it may not be the


      purview of this committee, but I am asking if we


      can address this aspect and how that plays with the


      implications of labeling, that if we do put, as was


      suggested, a specific negative label in terms of


      the indications and certainly as a warning, let


      alone a black box warning, that the amplitude of


      these warnings are heard.


                I mean it is almost a penalty then for the


      intense direct-to-consumer advertising, which does


      play, as I understand it, a huge role in driving




      sales for some of these antidepressants.


                I just wondered if you could give us some


      indication, I mean is there a direct policy with


      D.D. Mack how these various gradations get


      translated into the few seconds that go on the tail


      end of a commercial.


                DR. TEMPLE:  They are certainly supposed


      to.  The presence of a strong warning or box


      warning should be reflected right in the major


      statements that are made.  The direct-to-consumer


      comes in two flavors, written and TV.


                In TV, you can't give as much information


      easily, but it has to be available readily.  You


      can argue about whether people make use of that


      availability.  But the major statement would have


      to reflect the balance between those two things.


                You know, I am sure people have views


      about whether that is done successfully or not.  If


      it is written, then, the written statements have to


      show that balance.  Any box warning has to be


      reflected in it.


                So, yes, it is supposed to reflect the




      balance of information that is in the labeling, so


      if the labeling changes, the direct-to-consumer


      advertising should change.


                DR. GOODMAN:  Dr. Perrin.


                DR. PERRIN:  Yes.  Back to the ODS person.


      A number of the anecdotes yesterday suggested that


      people were put on antidepressants quite off label


      and probably not for major depression, but rather


      for minor depression and acute depression.


                You said that you have the evidence on


      mood disorders that includes major depression.  Can


      that be disaggregated at all into other non-MDD


      forms of mood disorders?


                MR. EVANS:  We could look at that.  We


      didn't, we lumped them together just for a top-line


      statement at this time, because we wanted the focus


      to be more on suicidality than drug use.


                DR. GOODMAN:  Ms. Griffith.


                MS. GRIFFITH:  I wanted to know, since the


      data you got was January to June, and the Advisory


      didn't come out until late March, is it possible to


      look at the data that you got April to June to see




      if there is a decline?


                MR. EVANS:  We did look at it monthly in


      those months, and there was not any decline.  I


      mean it wasn't a change.


                DR. GOODMAN:  Are these new prescriptions?


                MR. EVANS:  These were total


      prescriptions, new and refill.


                DR. GOODMAN:  Did you separate out by new


      prescriptions in terms of the monthly rate?


                MR. EVANS:  We can, and we did, and we did


      not see much of a decline in those, as well.


                DR. GOODMAN:  Dr. Chesney.


                DR. CHESNEY:  On the surface, this looks


      not bad in the sense that 65 percent are being


      written by psychiatrists, but although I am not


      here as a patient representative, I do have a


      daughter who has been on these medications, and I


      know for a fact that most often psychiatrists do


      not prescribe these medications.


                My image is that at the end of the day,


      they take a whole packet of prescriptions--and I


      will be interested to have the psychiatrists




      respond to this--a whole packet of prescriptions


      that have been written by social workers,


      pharmacists, psychologists, and sign their name.


                So, I think when we are talking about


      educating primary care providers, looking at this,


      I am reassured, but I know that this does not


      represent who is actually writing the




                MR. EVANS:  Yes, this is a limitation of


      the data, too, the data is only as good as what the


      pharmacist inputs at the computer, and, you know,


      if that specialty is on there, hopefully, they will


      put that on there.


                DR. CHESNEY:  I am sure they don't, but I


      think this is very important in the educational


      issue, because the people who are prescribing this,


      on the whole, are not child psychiatrists, and they


      are family practitioners, they are ER physicians,


      they are nurse practitioners, they are pharmacists.


                I mean I was appalled at what happened


      when we visited one of these pharmacists, but no


      disrespect to pharmacists, but this is very much




      happening out there.


                DR. GOODMAN:  One last question for ODS


      representative from Dr. Wang.


                DR. WANG:  I was curious, has anyone


      studied what happened after the British


      contraindicated these in children, just to get a


      sense of what the impact of a labeling change, such


      as that, might be?


                MR. EVANS:  In February, they looked


      through 2002, the market between 2002 and 2003


      group, 15 percent with no change really in the


      adolescent and children population.  It was still


      around 78 percent, so I don't think there was a


      change much in this country.


                DR. WANG:  But you don't know of any data


      in the British--


                MR. EVANS:  We didn't look at that.


                DR. GOODMAN:  Thank you very much.  You


      may step down.


                We have six more presenters.  I am not


      taking any more, that's it.


                Dr. Leslie, do you remember your question?




                DR. LESLIE:  My question goes way back and


      is for the FDA.  I think reading through the


      materials that we received from the public, two of


      the major concerns about the data that was coming


      in were suicidality, et cetera, being captured


      under other labels, such as emotional ability, and


      then also the issue about dropouts were people


      dropping out of either the placebo or the drug


      groups, that were having the kinds of adverse


      events that were of interest to us, and then not


      being counted in the data.


                So, I had two questions.  One was do you


      feel confident that the data you have received has


      addressed those two issues for the analyses that we


      looked at yesterday, and the second was what steps


      could you potentially take to address those


      drawbacks that were raised by the public in the


      written requests that proceed from here on out.


                DR. LAUGHREN:  I can respond to the first


      part of that.  In terms of the data that we


      received, if you recall, we issued letters to


      companies in July of last year, which specified a




      very clear research strategy for looking for


      adverse events that might be related to




                It included both preferred terms and


      verbatim terms.  All these data are electronic, so


      it was a string search to look for events that


      might be possibly related to suicidality.  In


      addition to that, we asked companies to look at all


      their serious adverse event narratives, any event


      that had been classified as a serious adverse


      event, they would have to look at and make a


      decision whether or not that might represent




                Then, later in the year, we issued


      additional requests to basically ask them to give


      us all the serious adverse event narratives, so


      that we could have Columbia themselves look at


      those data, and also, all accidental injuries and


      accidental overdoses to try and broaden the search,


      to make sure that we could capture everything that


      might be related.


                Now, it is true, despite all of that, it




      is possible that certain events that didn't rise to


      the level of being a series adverse event might


      have been captured under some other either verbatim


      term or preferred term.


                The other question is whether or not the


      narratives that we received fully reflected the


      case report forms.  The narratives are created by


      the companies.  To try and address that, we have


      sort of a second level of this contract with


      Columbia that is ongoing right now.


                We have done a 20 percent sampling of the


      case report forms for the narratives we have, to


      have them check the narratives against the case


      report forms basically to see whether or not they


      fully reflected, the narratives fully reflected


      what was in the case report form.


                In addition to that, we have asked for the


      dictionaries.  The dictionaries, basically,


      companies, when the code data, they subsume them


      under preferred terms, and once they do that, that


      creates basically, a dictionary.


                So, we have asked for the dictionaries




      from all these sponsors.  Columbia is currently


      looking at those dictionaries to see if there are


      any other additional adverse event terms that might


      be of interest to look at, again to answer that


      question whether or not all relevant events have


      been captured.


                That is a very tedious, time-consuming


      process, but it is ongoing right now.


                Dropouts, Dr. Hammad addressed that


      yesterday.  We did look at dropouts, and as he


      suggested, it is true, many patients were dropping


      out for these events.  In a sense, it was almost a


      surrogate for that endpoint.


                DR. GOODMAN:  Dr. Posner, did you have a




                DR. POSNER:  I just wanted to say that, in


      addition, because we asked for all of the


      accidental injuries, and that would be the most


      likely place, that all of these events involved


      some type of injury or another, that you would find


      events that were missed, so we can feel reasonably


      confident that this body of data represents




      everything we would want to look for, but we are


      doing these additional steps, as well.


                DR. GOODMAN:  Dr. O'Fallon.


                DR. O'FALLON:  Yes, this sort of follows


      up.  We are back to what I am concerned about, the


      people who came here, they are worried about the


      side effects, the toxicity here.  Right from the


      beginning, when you told us back in February about


      this study, I was worried about what wasn't


      recorded in the drug companies' records, for


      whatever reason.


                I think that is still, no matter what we


      do going forward, we have got to address the issues


      there.  So, what I am wondering about, I would like


      to propose, and you shoot at and tell me that these


      things are not feasible, but it seems to me what we


      really need are somehow a standardized suicide


      monitoring procedure or whatever for future studies


      in mental illness, any kind of a drug that is


      targeted toward the mind, we should be looking for


      this type of thing, the suicidality side effect.


                Then, I think we are going to have to have




      some sort of standardized suicide coding.  They


      have done it in adverse events, not great, but it


      could be done better for suicide coding because of


      the work that has been done here.


                I think this has been a wonderful outcome


      in terms of the coding issues.


                Now, here is one that is going to kill


      everybody, but you can make suicidality a goal, a


      primary goal in, say, mental illness or maybe


      depression more specifically, where you really


      think that this side effect or toxicity, this


      adverse event is also a symptom of the disease.


                In cancer, they have had to struggle with


      the issue of distinguishing side effects from


      symptoms of the disease for decades, some way of


      going after that.


                Just one more comment.  This is a comment.


      I believe that there was a 40 percent--in the stuff


      I saw before I came out here--I think I saw 40


      percent placebo effect in TADS, the TADS study.


                If that is true, this is a major issue.


      That is one of the reasons why we really do need to




      have placebo arm in these different studies,


      because if 40 percent of this population will have


      a beneficial effect due to sugar pills, to try to


      tease out true effectiveness of these medications


      given their severe side effects, it is very


      important that we have a placebo arm even going




                I know that you are not thinking of it,


      but I think some of the people in the room are


      wondering why we have to go with sugar pills.


                DR. LAUGHREN:  Let me comment on the last


      point first.  We clearly agree with you about


      placebo, but it is not just the act of giving a


      pill.  In all of these trials, there is a lot that


      happens that results in improvement.


                DR. O'FALLON:  Yes, I know that.


                DR. LAUGHREN:  There is a lot of


      attention, the patients have a lot of interaction


      with staff.  That is really the placebo effect.


                But the other point you are making, I


      completely agree that ascertainment is ultimately


      the issue.  If you don't collect the information,




      it doesn't make any difference how carefully you


      search the database, if it wasn't collected, it's


      not there, so ascertainment is key.


                Again, that is one of the things that we


      hope to get out of this effort with Columbia is a


      better guidance document for future trials to make


      sure that suicidality is properly ascertained, but


      it is an evolving thing in the field.  I mean there


      is not at this point in time an optimal way of


      doing that, so we hope to get an instrument that we


      can apply for future trials.


                Again, I agree with you that coding of


      data needs to be standardized, and again that is


      one of the things that we expect to come out of




                DR. POSNER:  Could I just add to that?  We


      are very committed to addressing the question that


      you are referring to.


                DR. GOODMAN:  Would you bring the


      microphone closer.


                DR. POSNER:  I said we are very committed


      to addressing this issue in terms of suicidality




      adverse event monitoring, and Dr. Laughren


      mentioned guidelines that we are going to write and


      measures that we actually have developed that we


      can implement in all trials, that will help us


      collect the right data and then be able to use


      these consistent definitions to classify events, so


      we can make sense across all of these trials.


                What is important to note is that we are


      working on a National Institute of Mental Health


      study called TASA, Treatment of Adolescent Suicide


      Attempters, which is very focus on this issue of


      adverse event monitoring, and it is wonderful


      because it is helping us inform the process, so we


      have developed very, very rigorous standards of how


      we ask these questions, what measures we use, and


      how to do it consistently, and that will help


      inform the guidelines and the measures that


      industry and everybody else can use.


                So, we have made a lot of progress in


      that, I think.


                DR. GOODMAN:  Dr. Temple.


                DR. TEMPLE:  I just want to follow up on




      the last discussion.  It seems to me there are two,


      somewhat separate things, and I would be interested


      in people's views.


                One is to make the periodic routine


      question better than it is.  There is a suicide


      item on the score, but that didn't show anything.


      Maybe that will never show anything because when it


      happens, it happens abruptly and you don't happen


      to pick it up at the two-week period, but it does


      seem as if a better questionnaire on that question,


      done routinely, might be useful.


                The second part of it is how to


      characterize events, what questions to ask about


      those, what to write down.  Is that what you are


      thinking, they are two somewhat different things?


                DR. LAUGHREN:  I agree, they are two


      separate things.  There is the suicide item that is


      part of every one of these instruments and sort of


      standardizing how that routine information is


      elicited, but then when an event occurs, you have


      to ensure that the appropriate questions are asked


      to flesh out that situation, so that someone down




      the road who is looking at the data is able to make


      sense of it.


                DR. O'FALLON:  But I would like to point


      out that the monitoring procedures, especially for,


      say, the first two weeks or something like  this,


      should include a real collaboration with the


      children and their caregivers, their parents,


      whoever they are living with, to be on the watch


      for those and to report them immediately, and that


      those things would be part of the data.


                DR. LAUGHREN:  Let me just respond to that


      quickly.  Basically, you are switching gears to a


      clinical setting, I think, other than a clinical




                DR. O'FALLON:  You guys can write the


      regulations for the clinical trial, right?


                DR. LAUGHREN:  Right, but obviously, the


      points that you are making apply to clinical


      practice, as well.


                DR. O'FALLON:  Yes, but they would apply I


      would say in the clinical trial, because I think


      that you are not possibly getting all your




      information.  If you don't come in for two more


      weeks, people forget that they were scared 10 days




                DR. LAUGHREN:  Absolutely, I agree.


                DR. POSNER:  I just wanted to add that we


      are also working with the CDC just on this


      question, what are the right one or two questions


      that need to be asked in any trial or clinical


      setting to get this information to be able to


      classify it appropriately, which is exactly what we


      are talking about, and it is not necessarily the


      best questions on the measures that were used in


      this trials, but that is exactly the pertinent


      point in clinical setting or in research settings


      with the increased monitoring that you are


      referring to.


                DR. GOODMAN:  I am going to need to wrap


      up the remaining questions in the next five




                Dr. Irwin.


                DR. IRWIN:  The question I wanted to ask


      specifically was related to the one that is on the




      table right now.  Yesterday, we heard from several


      families and individuals about really homicidal


      behavior and more violent behavior outwardly


      directed, not internally directed.


                Of concern to me is that the focus has


      been so much on suicide, but what I wanted to know


      is what kind of monitoring or what kind of tools


      are in place to really measure that phenomenon in


      these trials, because it seems to me that we don't


      have any data that has been shown to us at least on


      adverse experience or events with the clinical




                DR. LAUGHREN:  Our focus here clearly has


      been on suicidality, and not on hostility and


      violence.  There was a lot less of that in these


      trials than we had suicidality, and we have not


      come to grips with that yet, but it is a whole


      other area that needs to be fleshed out and


      developed in the same way that suicidality has been


      fleshed out because again we have included in our


      database information on whether or not these


      individual patients at some time during the course




      of treatment were coded as having hostility or


      agitation as a preferred term.


                If you go back and look at what got


      subsumed under that, I am sure it is going to be


      quite different depending on different sponsors,


      and it really requires a parallel development to


      try and understand what that means.


                Again, all the things we have been talking


      about for suicidality apply to that domain, you


      know, how do you ascertain for it, what kind of


      questions do you ask to flesh it out, it is a real




                DR. GOODMAN:  By the way, the placebo


      response rate from the TADS trial is 35 percent.


      Looking at the paper again, I see 35 percent.


                DR. O'FALLON:  Okay.


                DR. NEWMAN:  Just to clarify, that is much


      or very much improved in the TADS trial.


                DR. GOODMAN:  I don't think that one is


      based on the CDRS, right, that is what you are


      saying, it is based on the CGI?


                DR. NEWMAN:  The dichotomous outcome was




      were they much or very much improved, so if you


      said just improved, reasonably, it would have been


      a lot more.


                DR. GOODMAN:  But that is the standard, I


      mean it has to be much or very much improved to be


      a responder.  That is pretty much across all


      clinical trials.


                Dr. Pine, please.


                DR. PINE:  I want to return to a point


      that was raised by Dr. Goodman and just call the


      committee's attention to a couple things that he


      said and then also raised a couple other issue


      relevant to the discussion about 10 or 15 minutes


      ago with FDA.


                That is the issue of both how perplexing,


      but also how important it is to think very


      carefully about the efficacy data and the


      difference between the data for fluoxetine, on the


      one hand, and all the other antidepressants, on the


      other hand, and how do we understand that.


                Number one, just to underline that I agree


      that the importance of that point cannot be




      overstated.  I guess there is a couple of issues


      that were not discussed 10 or so minutes ago in


      sufficient detail, and really two points to raise.


                One is that I do appreciate from a


      regulatory standpoint that it is very difficult to


      specify exactly how one is to do an appropriate


      study.  We talked about a lot of the details that


      we don't need to go over again except one thing was


      not discussed, and that was a discussion of the


      level of rigor that goes into both the training of


      the investigators who are going to ascertain the


      samples and document the diagnosis, on the one


      hand, but then also follow the response of the


      patients throughout the trial.


                Then, I think the last thing to say about


      that specific point is that when we look at the


      data that have been published, and probably the


      most extensive data are from the sertraline trial


      as opposed to the TADS trial, with the sertraline


      trial being a pharmaceutical-sponsored study that


      appeared in JAMA, and the TADS trial being an


      NIH-sponsored trial that was also published in JAMA.




                There are some fairly clear signs that the


      manner in which investigators were trained, that


      the criteria for enrolling patients for the process


      of evaluating the response as it was manifest


      throughout the trial was quite different in those


      two studies.


                Again, when we come to the issue of how


      important it is to compare the data for fluoxetine


      and the data for the other agents, I think we need


      to acknowledge that there are already signs in the


      data that have been published in the reports that


      have appeared in peer review, that the quality of


      the studies appears to be different.


                I think it is also important to note that


      if we were to look at the efficacy data by industry


      sponsor versus federally funded, there have been,


      to the best that I can recall right off the top of


      my head, two federally funded SSRI trials, both are




                So, we are 2 out of 2 on that score,


      whereas, if we look at all the others, we are


      basically 1 out of 13 or 1 out of 14.




                DR. GOODMAN:  Those two are both in


      fluoxetine, isn't that correct?


                DR. PINE:  That's true, so, you know, we


      have a confound between federally sponsored and the


      compound, but those are the data that we do have,


      and I think given the issues that I just discussed,


      you know, we are going to be very hard pressed to


      say this is a funding or design feature issue,


      which it might be, or that this is a medication


      issue, which again it might be.


                DR. GOODMAN:  I want to make sure I am


      clear on the source of the fluoxetine data for the