FOOD AND DRUG ADMINISTRATION



































                       Tuesday, October 19, 2004


                               8:30 a.m.










                CDER Advisory Committee Conference Room

                           5630 Fishers Lane

                          Rockville, Maryland





      Arthur H. Kibbe, Ph.D., Chair

      Hilda F. Scharen, M.S., Executive Secretary



      Patrick P. DeLuca, Ph.D.

      Paul H. Fackler, Ph.D.

      Meryl H. Karol, Ph.D.

      Melvin V. Koch, Ph.D.

      Michael S. Korczynski, Ph.D.

      Marvin C. Meyer, Ph.D.

      Gerald P. Migliaccio, Ph.D. (Industry


      Kenneth R. Morris, Ph.D.

      Cynthia R.D. Selassie, Ph.D.

      Nozer Singpurwalla, Ph.D.

      Marc Swadener, Ed.D. (Consumer Representative)

      Jurgen Venitz, M.D., Ph.D.



      Judy Boehlert, Ph.D.

      Gordon Amidon, Ph.D., M.A.

      FDA Staff

      Gary Buehler, R.Ph.

      Lucinda Buhse, Ph.D.

      Jon Clark, M.S.

      Jerry Collins, Ph.D.

      Joseph Contrera, Ph.D.

      Ajaz Hussain, Ph.D.

      Monsoor Khan, R.Ph., Ph.D.

      Steven Kozlowski, M.D.

      Vincent Lee, Ph.D.

      Qian Li, Ph.D.

      Robert Lionberger, Ph.D.

      Robert O'Neill, Ph.D.

      Amy Rosenberg, M.D.

      John Simmons, Ph.D.

      Keith Webber, Ph.D.

      Helen Winkle

      Lawrence Yu, Ph.D.



                            C O N T E N T S




      Call to Order

        Arthur Kibbe, Ph.D.                                      5


      Conflict of Interest Statement

        Hilda Scharen                                            5


      Introduction to Meeting

        Helen Winkle                                             8


      Subcommittee Reports - Manufacturing Subcommittee

        Judy Boehlert, Ph.D.                                    26


      Parametric Tolerance Interval Test for Dose

         Content Uniformity                                     53


      Critical Path Initiative


        Topic Introduction and OPS Perspective

          Ajaz Hussain, Ph.D.,                                  64


        Research Opportunities and Strategic Direction

           Keith Webber, Ph.D.                                 105


        Informatics and Computational Safety

          Analysis Staff

          Joseph Contrera, Ph.D.                               117


        Office of New Drug Chemistry

          John Simmons, Ph.D.                                  165


      Open Public Hearing

        Saul Shiffman, Ph.D.                                   192


      Critical Path Initiative--Continued


        Office of Generic Drugs

          Lawrence Yu, Ph.D.                                   204


        Office of Biotechnology Products--Current

          Research and Future Plans

            Amy Rosenberg, M.D.                                248

            Steven Kozlowski, M.D.                             282



                      C O N T E N T S (Continued)




        Office of Testing and Research--Current

          Research and Future Plans

            Jerry Collins, Ph.D.                               316

            Lucinda Buhse, Ph.D.                               338

            Mansoor Khan, R.Ph., Ph.D.                         362


      Wrap-up and Integration

        Jerry Collins, Ph.D.                                   410


      Challenges and Implications

         Vincent Lee, Ph.D.                                    419


      Committee Discussion and Recommendations                 428




                         P R O C E E D I N G S


                             Call to Order


                CHAIRMAN KIBBE:  Ladies and


      gentlemen--welcome.  I want to take a little page


      from the coach at the New York Times, who says that


      a meeting that starts that eight o'clock actually


      starts at five minutes before.  And to get us


      rolling in about 30 seconds, ahead of time.


                Do we know--


                [Comment off mike.]


                --he'll be here tomorrow.  All right.


      So--Dr. Amidon, my co-pilot here, will be here




                I'd like to call you all to order for my


      last go-round as Chairman of this August body.  And


      the first order of business, of course, is to read


      about all of our conflicts.


                     Conflict of Interest Statement


                MS. SCHAREN:  Good morning.


                The following announcement addresses the


      issue of conflict of interest with respect to this


      meeting, and is made a part of the record to




      preclude even the appearance of such.


                Based on the agenda, it has been


      determined that the topics of today's meeting are


      issues of broad applicability, and there are no


      products being approved.  Unlike issues before a


      committee in which a particular product is


      discussed, issues of broader applicability involve


      many industrial sponsors and academic institutions.


      All special government employees have been screened


      for their financial interests as they may apply to


      the general topics at hand.


                To determine if any conflict of interest


      existed, the Agency has reviewed the agenda and all


      relevant financial interests reported by the


      meeting participants.  The Food and Drug


      Administration has granted general matters waivers


      to the special government employees participating


      in the meeting who require waiver under Title 18,


      United States Code Section 208.


                A copy of the waiver statements may be


      obtained by submitting a written request to the


      Agency's Freedom of Information Office, Room 12A30




      of the Parklawn Building.


                Because general topics impact so many


      entities, it is not practical to recite all


      potential conflicts of interest as they may 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 discussions before the committee, these


      potential conflicts are mitigated.


                With respect to FDA's invited industry


      representative, we would like to disclosed that


      Paul Fackler and Mr. Gerald Migliaccio are


      participating in this meeting as a non-voting


      industry representative, acting on behalf of


      regulated industry.


                Dr. Fackler's and Mr. Migliaccio's role on


      this committee is to represent industry interest in


      general, and not any one particular company.  Dr.


      Fackler is employed by Teva Pharmaceuticals,


      U.S.A., and Mr. Migliaccio is employed by Pfizer,




                In the event that the discussions involve




      any other products or firms not already on the


      agenda for which FDA participants have a financial


      interest, the participants' involvement and their


      exclusion will be noted for the record.


                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 products they may wish to comment




                Thank you.


                CHAIRMAN KIBBE:  Thank you.


                And now we'll hear from the Director of


      the Office of Pharmaceutical Sciences, Ms. Helen




                        Introduction to Meeting


                MS. WINKLE:  Good morning, everyone.


                All right, I want to welcome everybody


      this morning to the Advisory Committee for


      Pharmaceutical Science.  This is, I think, a very


      important meeting, and I"m really looking forward


      to the discussion.  But before we get there, I want


      to welcome all of the members.  We have one new




      prospective member, Carol Gloff--Dr. Gloff--has


      joined us.  And we have two other prospective


      members who we're having a little complication with


      in getting on board.  So we're working on that.


                We also will have a number of SGE's here


      today; Dr. Boehlert, Dr. Amidon and several others


      who are going to participate with us in a number of


      things.  So I want to welcome everybody.


                I also want to thank Dr. Kibbe.  This is


      his last time as Chair.  It will break all of our


      hearts to see Dr. Kibbe go out of this position.


      He has been very, very enthusiastic as the Chair of


      this committee, and I think all of us have enjoyed


      working with him.  But he's not to go very far.


      We've already told him that we anticipate him


      coming back to a number of meetings and helping us


      with some of the discussion in the future.  So we


      really want to, again, thank him for all he's done.


                Dr. Cooney--Charles Cooney--has agreed to


      be the chair of the committee for the next two


      years.  Unfortunately, Dr. Cooney couldn't be


      here--after he accepted, he couldn't be here today.




      But he will be here at the next meeting.  So--he's


      been very gracious to accept this position.  He and


      I have talked at length about some of the issues we


      want to cover on the Advisory Committee, and he's


      very enthusiastic about moving ahead for the future


      of the committee.


                The agenda for the meeting today:  there's


      a number of things we want to take up.  I'm going


      to talk a little bit about next year--2005 being, I


      guess, this fiscal year--and some of the things


      that we plan to take up with the Advisory


      Committee, where we're going in OPS, just to give


      the committee a little feel about some of the


      things that we're looking at.


                I also want to give a quick--and I mean a


      quick--update of the cGMP Initiative for the 21                          




      Century.  We're also going to have an update on a


      number of the subcommittee and working groups.  Dr.


      Boehlert is going to talk about the Manufacturing


      Subcommittee meeting that we had several months


      back.  It was a very, very--we accomplished a lot,


      I think.  It was a very good meeting.  And Judy can




      fill us in on some of the highlights of that


      meeting.  Also Bob O'Neill is going to talk about


      the Working Group with IPAC RS, and some of the


      accomplishments--or the focus that we've had in


      that Working Group.


                We're also going to talk about the


      Critical Path Initiative.  And I think this is a


      really important discussion that we can have with


      the committee today.  Critical Path is, of course,


      one of the main initiatives in the agency now, and


      what we would like to talk about with the committee


      is give you some idea of our thoughts, as far as


      Critical Path; some of the things that we're doing


      in the Critical Path Initiative, in the office of


      Pharmaceutical Science in the various product


      areas, and get some input from you as to what


      direction we need to go; if there's other things we


      need to be thinking about; and if there's other


      types of topics that we need to be taking up, we'd


      like to do that.


                Dr. Woodcock talked about the Critical


      Path Initiative when she introduced it, saying that




      FDA was really in the best position to identify


      those areas, or those gaps, in drug development,


      and to work with others--collaborate--on how we


      could get the data necessary to fill those gaps.


                So this is really what we're looking for


      doing under the Critical Path Initiative.  And we


      need to be certain that we are identifying the gaps


      correctly, and that we are able to do the types of


      research that needs to be done to fill those gaps.


      Of course we can't do everything, so I think some


      of what we want to talk about and think about, too,


      is how we can prioritize some of that research.


                Tomorrow, we're going to talk about


      manufacturing, and moving toward the desired state.


      As I said, we had a very productive meeting of the


      Manufacturing Subcommittee.  A number of things


      were identified at that meeting that we need to


      discuss further; that we needed to look at and


      determine how we're going to do it.  A number of


      questions that we need to answer--and we're looking


      at possibly having a subgroup to do some of that--a


      fact-finding group.  So Judy will talk to that.




                But there are a number of things, too,


      that we want to talk about with the committee


      today; a number of--the gaps that we recognize that


      we have in OPS and the agency, in moving toward


      that desired state.


                So several of us are going to talk about


      those gaps.  We're going to talk about the


      organizational gaps, the science gaps, and the


      policy gaps--all of which are important if we in


      the agency are going to be prepared as the


      manufacturers and others move toward that desired




                So I think that will be a really


      interesting issue, and I think there are a number


      of things that the committee can help us with in


      identifying how best to address these answers and


      to address the gaps.


                We also have a number of bio-equivalence


      issues that we want to discuss.  We want to


      continue the conversation from the last Advisory


      Committee we had on bio-equivalence.  And Dr. Yu


      and some of his staff are going to talk about some




      recommendations from that.  And we're also going to


      bring up a new topic on gastroenterology drugs.


                So--moving on to OPS in 2005.  I think


      2004, we had an extremely busy year, mainly focused


      on the GMP Initiative, and all of the aspects of


      that initiative--especially the areas concerning


      manufacturing science and how wee were going to


      really address those issues and concerns, and how


      we were going to incorporate those into the


      regulatory framework.


                As we move into 2005, I think we still


      have a lot of issues that we have to handle under


      Pharmaceutical Quality Initiative.  We've already


      said that that's going to be some of what we take


      up with the Advisory Committee today.  But we


      really need to pursue those next steps.  In doing


      that, though, we also need to be looking at


      continuing to streamline the review processes.  We


      continue to get more and more products in for


      review, and there's got to be some way to offset


      that increasing workload.  And streamlining the


      review processes seems to be--we're moving in that




      direction, and it seems to be the answer to


      handling some of the enormous workloads that we




                Also, we need to incorporate best


      practices.  We've added the Office of Biotech


      Products in the last year.  They joined us in


      October of 2003, and they have a lot of practices


      in their review that I think can be very helpful as


      we move forward in looking at ways to improve--both


      in out office of New Drug Chemistry, and our Office


      of Generic Drugs.


                So we're going to be looking at


      incorporating best practices across the entire




                Supporting the Critical Path


      Initiative--I've already brought this up.  It's a


      very important part of where we're going.  I think


      much of our research is going to be done there, and


      I think we're talking about much more than


      laboratory research. I think there's a number of


      activities that we hope to take on in 2005 where


      we're looking at improving on how we do the




      regulation, and in actually working through the


      Critical Path Initiative to get some of this done.


      So we'll talk more about that as we get into


      Critical Path and some of those projects that we're


      looking at doing.


                We're looking at further integrating the


      whole Office of Biotech products.  There are still


      some things that need to be accomplished there.  I


      think there are still a number of questions that


      the Advisory Committee can be very helpful to in


      answering.  So you will hear more about this in the


      next fiscal year.


                And, last of all, I think there still


      continues to be a number of regulatory on follow-on


      proteins, as well as a number of general scientific


      issues that we'll want to discuss with the




                So I think we have a lot on our plate


      during the year, and I look forward to working


      closely with the Advisory Committee in the next


      fiscal year to help us identify some of the--other


      things that we need to look at, as well as help us




      with the issues that we already have identified




                Okay.  As I said, I'm going to talk real


      quickly about the CGMP Initiative for the 21st


      Century.  I think most of you all have probably


      read the background material, which included the


      report.  We've actually come to the end of the


      first two years of the initiative.  And I"d like to


      emphasize:  I don't think that's the end of the


      initiative.  I think it's just the beginning.  I


      think that the initiative helped us identify a


      number of things that we need to be looking at in


      review, that we need to be looking at in


      inspection.  We still have a lot of changes to


      make.  I think we've made a lot of progress--and


      I'll talk a little bit about some of that progress.


      But I think we've got a lot more that we have to


      focus on.


                So that was only, in my mind, the first




                But I thought it would be helpful just to


      step back real quickly and look at what the goals




      of the initiative were.  Because I think you can't


      really appreciate the accomplishments without


      really understanding what the goals were.


                So there were basically six major goals.


      The first one was to incorporate the most


      up-to-date concepts of risk management and quality


      systems approaches; secondly, was to encourage the


      latest scientific advances in pharmaceutical


      manufacturing and technology, ensure submission


      review program and the inspection program operating


      in a coordinated in synergistic manner; apply


      regulation and manufacturing standards


      consistently; encourage innovation in the


      pharmaceutical manufacturing sector; and use FDA


      resources most effectively and efficiently to


      address the most significant health risks.


                And you can see, when you look back at


      these initiatives, the role OPS has had to play in


      all of these goals.  I think they're very


      important, not only to the agency, but important to


      us at OPS, and important to the industry and others


      involved in the manufacturing of pharmaceutical






                So, quickly, through the


      accomplishments--again, you can read the report.


      You'll get a lot more out of the report.  But I


      just want to emphasize that there was an awful lot


      done in the last two years; a lot that will affect


      how we move forward in the future, in the 21st


      century.  So I wanted to highlight those.


                The first thing was Part 11.  We did a


      last in the last two years to clarify the scope and


      application of Part 11.  There were quite a few


      questions; quite a bit of complication in


      implementing Part 11.  And I think we've moved


      forward in trying to eliminate some of that


      complexity and complication.  We issued two


      guidances during the two-year period that have


      helped in that clarification.


                Technical Dispute Resolution Process--this


      was also a very important part of the initiative.


      And it really has had a very positive effect, I


      think, on the industry, and a positive effect on


      how the field has dealt with inspections and has




      increased the time and effort that the inspectors


      are putting into the inspections, and the time and


      effort that they're spending with industry when


      they go in and do these inspections.  And it has


      really been the basis of much discussion in the


      inspection process.  And the outcome--we have not


      had any technical disputes.  We have a very good


      process--as I said, the process has sort of set the


      framework for opening up the discussion.  And so I


      think that it has had a really positive effect.


      I'm actually a co-chair of that group. I kept


      waiting for disputes.  I thought we were just going


      to have tons of them.  We have a pilot program, and


      I thought in the 12 months of the pilot we'd be


      able to figure out how best to run the program.


      But not having any disputes, we haven't learned a


      whole lot of lessons.


                But, again, it's had its very positive


      effects.  So I think that it has really been useful


      under the initiative.


                The GMP warning letters--this was an issue


      that was handled very early on.  And we




      accomplished the goals that we wanted under this


      particular working group of the initiative; and


      that's that warning letters now are reviewed by the


      Center to ensure--in the Center before they go out


      to the companies--to ensure that they have adequate


      scientific input.  Many of the warning letters that


      went out in the past were not reviewed to make sure


      that the issues were scientifically sound.  So that


      has changed now.  And I think that's had a very


      positive effect.


                International collaboration--I won't go


      into that, but we have spent a lot of effort in


      ICH, and Q8, Q9, and hope to do a lot in Q10.  And


      also one of the things we are planning on doing is


      getting more involved with PICS, which is looking


      at inspections on a worldwide basis.


                Facilitating innovation--including doing


      standards and policies--we were very fortunate to


      put out a number of different guidances under this


      part of the initiative; the aseptic processing


      guidance--which industry is very familiar with.


      They've been waiting for this guidance for a long




      time.  And I think it addresses many of the


      questions that have been out there in industry's


      mind.  So I think it's a very, very positive part


      of the initiative that we were able to accomplish.


                The next guidance that was put out--I


      think many of the people--in fact, everyone on the


      Advisory Committee is very familiar with this


      guidance, because we did have a subcommittee on the


      PAT--the Process Analytical Technologies--under the


      subcommittee, and we were able to put, under Dr.


      Hussain and others in the group, we were able to


      put out a guidance to industry which has had an


      extreme effect, I think, on how industry and others


      are looking at manufacturing in the future.  I


      think it's been probably one of the best parts of


      the whole initiative.  It really has promoted the


      two--the team approach to doing work; working on


      standards.  We've worked with ASTM under E55.  And


      I think, all in all, this has been an extremely


      successful initiative under the GMP initiative.


                The last guidance that we've had, that was


      comparability protocol.  That guidance is still in




      limbo.  We're trying to make sure that before we


      issue the guidance that we're not increasing the


      regulatory burden--which I think many of us felt


      when we read the original draft guidance.  So we're


      busily working on that to make sure that what we


      come out of is very beneficial to industry and to


      FDA, and that we don't put any additional resource


      requirements on either part of the regulatory




                Manufacturing science--the desired state


      under !8 of ICH has become a very important aspect


      of where we're driving to.  And, of course, we're


      going to talk to that tomorrow morning; continuous


      improvement and reduction of variability have been


      an important part of manufacturing science, and


      areas that we need to explore more in the future,


      and assure that we can accomplish that, especially


      being able to open up in the agency and allow more


      continuous improvement for manufacturers.


                Product specialists--this includes


      enhancing the interactions between the field and


      the review.  We're looking at a team approach, in




      having our reviewers all out on inspections.  And


      we're looking at best practices from both the PAT


      team and Team Biologics.  I think there's a lot of


      best practices there that we can incorporate in out


      thinking in the future on how we handle review and




                Integration of approval and


      inspection--this is more of that.  We have


      developed the pharmaceutical inspectorate, and


      we're looking also at changes in pre-market


      approval program.


                Quality management systems--there's a


      number of things that we've worked on here.  They


      take a number of directions.  We've developed a


      standard quality systems framework; a quality


      systems guidance.  We've worked on GMP


      harmonization, analysis process validation, and


      good guidance practices--none of which are going to


      go into in detail, but I think all very beneficial


      to helping us in the future in the 21                                     

                                                        st century.


                Risk management--risk management, I had


      thought--we did introduce a site-selection model




      for inspection under this part of the initiative.


      I believe there's a number of other things that we,


      especially in Review, need to focus on as far as


      risk management, and have a much better idea of


      what the risk of products are, and how we're going


      to mitigate those risks.  And I think this is


      something that we will bring up in the future at


      the committee.


                Team Biologics was to look at a number of


      initiatives that were already underway, and adopt a


      quality systems approach.


                And last of all was the evaluation of the


      initiative, which hasn't been completed yet, but


      it's a very important part of what we've done.


                So that, in a nutshell--I mean, that's a


      lot of effort, obviously, that we've done.  And if


      you, again, will read the report I think you'll get


      a much better feel.  But I felt like, since we've


      talked about it so much during the last few years,


      that it was very important to sort of wrap up what


      has happened in the last two years with this






                So that's all I have to talk today.  I'm


      going to give it back to Art, and I look forward to


      very lively discussion on a number of these issues,


      and look forward to working with you for the next


      two days.


                Thank you.


                CHAIRMAN KIBBE:  Thank you, Helen.


                We now have a report from the chair of one


      of the subcommittees--the Manufacturing






                          Subcommittee Reports


                       Manufacturing Subcommittee


                DR. BOEHLERT:  Good morning, ladies and


      gentlemen.  Before I just get started here--I tried


      pressing down, and--aha.  I need an SOP for how to


      operate the slides.




                It's a pleasure for me to be here this


      morning to update you on the Manufacturing


      Subcommittee.  We met in July.  And I think you'll


      find that a lot of the topics we discussed tie in




      very well with what Helen was talking about this


      morning, and also with some of the topics that are


      going to be on your agenda.




                We met for two days in July.  Just a brief


      overview of the topics that we discussed:  quality


      by design--we've heard that this morning;


      introduction to Bayesian approaches--and we'll talk


      just a little bit about that; research and training


      needs--the industrialization dimension of the


      Critical Path Initiative--another topic we heard


      about this morning; manufacturing science and


      quality by design as a basis of risk-based CMC


      review; and risk-based CMC review paradigm.




                On the 21                                                      

       st:  introduction to


      pharmaceutical industry practices research study; a


      pilot model for prioritizing selection of


      manufacturing sites for GMP inspection; cGMPs for


      the production of Phase I INDs; and applying


      manufacturing science and knowledge, regulatory






                What I'm going to do is just go over,


      briefly, some of the topics that were discussed,


      and also the comments that were made by committee






                Quality by design:  topic updates.  This


      addressed three guidances that should be coming out


      of ICH.  The first of ICH Q8, which is a guidance


      on pharmaceutical development section of the Common


      Technical Document.  It's going to describe


      baseline expectations and optional information;


      requires FDA and industry to think differently.


      Industry needs to be more forthcoming with


      information in their submissions, and FDA needs to


      look at the review process; focuses on process


      understanding and predictive ability.  And if you


      really understand your process, you'll gain


      regulatory flexibility.  It's a framework for


      continuous improvement.  And Step 2 is expected in


      November this year.  That means it will be out for


      public review and comment.






                ICH Q9 is quality risk management.  It


      looks at risk identification--should link back to


      the potential risk to the patient, because, after


      all, that's what's important; risk assessment--what


      can go wrong?  What is the likelihood?  What are


      the consequences?


                Risk control--options for mitigating,


      reducing and controlling risks; risk


      communication--between decision makers and other


      shareholders.  And this may also reach step two in


      November of this year, although that was a bit






                And then we're going to talk about quality


      systems needed to recognize the potential of !8 and


      Q9.  And this is ICH Q10:  monitor and evaluate


      processes with feedback groups in a manner to


      identify trends and demonstrate control or the need


      for action; manage and rectify undesirable


      occurrences; handle improvements; management,


      implement and monitor change.


                This is currently on hold, not because




      it's not a good topic, but primarily because all


      the resources that would address Q10 are tied up


      with Q8 and Q9.




                We also talked about the ASTM E55


      Committee.  And Helen mentioned that this morning.


      Their involved in the development of standards for


      PAT.  And the important things here are consensus


      standards, with input from industry, academia and


      regulators.  There's an established process, with


      an umbrella set of rules.  And ASTM is recognized




                They have three functional subcommittees


      on management, implementation and practices and


      terminology.  But one of the concerns expressed by


      the committees is are they going to duplicate other


      initiatives.  There area lot of people right now


      working on PAT initiatives, and are they going to


      duplicate some of that.  So we need to make sure


      that everybody gets on the same page.




                All right.  Now, this topic I'm going to




      be reluctant to say a whole lot about, but we had


      an introduction to Bayesian approaches.  Dr. Nozer


      Singpurwalla was kind enough to give us an


      introduction to the topic.  So, Nozer, I apologize


      if I mis-speak when I summarize--[laughs].


                You know--so it's with fear and


      trepidation--he's threatened us a quiz--


                DR. SINGPURWALLA:  You've already done it.


                DR. BOEHLERT:  Yes, I know. [Laughs.]


      That's what I was afraid of.  But I didn't think I


      could leave it out, or you'd get after me then,




                Okay--Reliability for the Analysis of


      Risk."  Reliability--the quantification of


      uncertainty.  And I'm just going to say a few words


      here:  utility--costs and rewards that occur as a


      consequence of any chosen decision.  These are the


      things that Nozer talked to us about--risk


      analysis--process assessing reliabilities and


      utilities, including an identification of


      consequences.  We talked about scales for measuring


      uncertainty--for example, probability.






                Now this is a quote, so I have to be


      careful here.  "When the quantification of


      uncertainty is solely based on probability and its


      calculous, the inference is said to be Bayesian."


      I am not a statistician, so I'm certainly not a


      Bayesian statistician.  And then there is


      discussion of use of Bayesian approaches for ICH


      Q8, Q9, Q10 and the use of prior information.




                Industrialization--dimension, the Critical


      Path Initiative.  We heard about that this morning.


      We'll hear about it in the next two days:


      examining innovational stagnation.  Everybody needs


      to take a look at what we've been doing in the past


      and get things moving forward in a new environment,


      with new technologies.


                Critical path--has been inadequate


      attention in areas of new or more efficient


      methodologies and development research.


                Industrialization--goes from the physical


      design of prototype up to commercial mass




      production.  And Education and research


      infrastructure needs improvement.  And this


      education and research applies to industry; the


      education also applies to the agency.  We all need


      to learn how to go forward in the new environment.




                FDA has a strong interest in computational


      methodologies to support chemistry and


      manufacturing control submissions.  They're putting


      together a chemometrics group.  There's a new FDA


      research program focusing on industrialization


      dimension.  And there's training needs.  AS I


      mentioned before, particularly with the


      pharmaceutical inspectorate.  That's started.


      There is an inspectorate now of trained


      investigators.  There need to be more.




                Manufacturing science and quality by


      design--it's a basis for risk-based CMC review.


      Companies share product-process understanding with


      regulators.  And this is a new paradigm, if you


      will, that companies will share more of the




      information that they have available than they have


      in the past.


                Specifications should be based on a


      mechanistic understanding of the process; there


      should be continuous improvement; and real time


      quality assurance.  You shouldn't have to wait


      until the end of the process to know that your


      product is okay.




                Science perspective on


      manufacturing--define current and the desired state


      and the steps to go from here to there; define


      terms--and this is going to be important going


      forward--things like "manufacturing science,"


      "manufacturing system," "manufacturing


      capability"--what do they really mean?


                Real case studies will help.  This came up


      time and again in the committee discussions.  It's


      nice to talk about all these theoretical concepts,


      but give me a real case study that I can look at


      and see what it really means.


                Testing is mostly non-value added. 




      Quality by design is the desired state.




                Risk-based CMC review--from the Office of


      New Drugs--should provide regulatory relief by


      incorporating science-based risk assessment; more


      product or process knowledge shared by the


      industry--and I've said this several times; more


      efficient science-based inspections; focus


      resources on critical issues; and specifications


      are based on a risk-based assessment.




                Quality assessment rather than a chemistry


      review--in the past it's been a strict chemistry


      review:  go down the list and check off the boxes;


      conducted by inter--and I see some smiles on the


      parts of agency folks--conducted by


      interdisciplinary scientists--so it could be a team


      approach.  It should be a risk-based assessment;


      focus on critical quality attributes and their


      relevance to safety and efficacy.  They have to


      rely on the knowledge provided by applicants.  If


      industry doesn't submit the information, the agency




      has nothing to make their decisions on.  And the


      comparability protocols are an important part of


      this review.




                Role of process capability in setting


      specifications will need to be addressed.  Very


      often, those kinds of process controls that you


      have may have no clinical relevance.  The knowledge


      base at the time of submission can be an issue,


      because very often you don't have that much


      information at the time you submit.  It's a


      learning process as you go through early


      marketability and commercial production.


                Specifications should not be used as a


      tool to control the manufacturing process.  And we


      might need to expand the Quality Overall Summary


      going forward.




                AS I said before, the extent of product


      knowledge is key.  Risk-based decisions should be


      based on supportive data.  Voluntary--all of these


      new initiatives are voluntary.  And that needs to




      be made very clear to the industry.  These are not


      requirements that everybody drop what they've been


      doing in the past and start over with new


      approaches--strictly voluntary.


                Supplement need is based on the knowledge


      of the risk of the change.  And there should be a


      clear rationale for the selection of






                Identify critical parameters for product


      manufacturing and stability; train FDA staff and


      regulated industry--this came up a number of times.


      We all need to learn what the other is doing;


      should give us--industry--greater flexibility in


      optimizing the process; should lessen the


      supplement burden, which is good for industry and


      good for the agency.  And, once again, real


      examples would be an asset.




                In the Office of Generic Drugs--generic


      industry's focus is on producing a bioequivalent


      product.  Often patent issues--to design around. 




      They may not have the flexibility as the new drug


      folks.  Workload in OGD is a significant issue, and


      committee members made a number of comments on this


      when they heard how many submissions there are, and


      how far behind they are.  We were impressed by the




                Provide advice to industry on improving


      quality of DMFs--those are "drug master


      files"--very important to the generic


      industry--also to the new drugs, but to a lesser






                Desired state--include needed data in a


      filing; process and product design; identify


      critical attributes; identify process critical


      control points.  And this is the difference from


      the past.  Analyze data to produce meaningful


      summaries and scientific rationales; and reviewers


      assess the adequacy of the submission by asking the


      right questions.




                Okay--some additional committee comments




      that came out of the Day One discussion:  ICH and


      ASTM appear to be synergistic, but ICH needs to be


      very aware of the ASTM focus.  There was some


      concern they might not be tied into what's going on


      there; some concern that FDA, internally,


      themselves, may be getting ahead of what's


      happening on an international basis. So they may be


      a little ahead of ICH Q8, Q9 and Q10.  That's not


      necessarily a bad thing, by the way.


                Need concrete examples--that came up time


      and time again; need to clearly demarcate "minimum"


      and optional information--you know, just what do


      you mean by "this is the minimum you need," and


      just what is "optional" information?  And


      "optional" information comes in degrees.  The more


      you make the more you know.  So you may not have as


      much information at submission as you will down the


      road after you've been in commercial product for a


      number of months or years.




                Need to avoid implying there are two


      different quality concepts.  We don't want to say




      that products made in the conventional way---the


      way we've always done it--are different than


      products that may be made according to some new


      paradigms.  Bring in new training programs--and


      Helen mentioned we're talking about forming a


      working group under the Manufacturing Subcommittee


      to address some of the issues, particularly case




                We need to find better terms than


      "minimal" and "optional;" and focus on process


      first, and then the tools that we're going to need.




                We had some reports on an FDA research


      project that's being done by Georgetown University


      and Washington University, and their goal is to


      identify attributes that impact inspection


      outcomes.  They're compiling and linking FDA


      databases.  They're looking at variables for


      product-process, facility, firm and FDA.  Right now


      they're collecting data.  CDER is just about


      completed, and CBER is ongoing--although by now it


      may be even further down the road.  This was July.






                Focus--are cGMP violations related to


      managerial, organizational and technical practice?


      And then interviewing manufacturers.  They have an


      internet-based questionnaire that went out in the


      fall of 2003.  They're looking at U.S. and European


      manufacturers.  And their data collection is near






                There's concern with just looking at


      numbers of deviations or field alerts, particularly


      when investigation may have shown little cause for


      concern.  You can put in a field alert and then


      find out later on that--oh--you know, we figured it


      out.  It really wasn't a problem.  So if you just


      look at numbers, you get those as well as the ones


      that are true issues.


                Also it was pointed out that if you're a


      company with a very detailed SOP you have a much


      bigger chance for deviating from it than your


      company with a really poor SOP that sort of allows


      you to do anything, where you're hardly ever going




      to deviate.  But who's to say which one is better?


                India and China are not include in the API


      manufacturers.  And we saw this as a downside to


      that survey, because they are major manufacturers


      of APIs.




                We talked then about risk ranking and


      filtering, where risk ranking is a series of


      decisions to start to rank within a class or across


      classes.  Tools may be customized for each


      application.  And filters may be used to reflect


      resource limitations and/or program goals.




                There's a pilot risk-ranking model to


      prioritize sites for GMP inspections, using ICH Q9


      concepts to define risk; Site Risk Potential--a new


      term for us--SRP--includes product, process and


      facility components.


                Look at probability and severity


      components that make up harm; and look at other


      risk-ranking models, for example those used by EPA


      and USDA; and then using the CDER Recall database.






                Comments--from the committee--focusing on


      volume at a site may be misleading because, in


      fact, when you have a high volume your process may


      be better controlled than if you have small volume.


                We need to also consider the risk of the


      loss of availability.  If you're a single-source


      drug for a life-threatening condition perhaps that


      needs to come into the equation.


                Look at "hard to fabricate" products, or


      products with difficulty controlling uniformity.


      Investigator consistency will be--and has been--an


      issue, but with the pharmaceutical inspectorate


      that should be better.  And it was suggested by at


      least one member that maybe they should look at


      high personnel turnover in a plant, because that


      might be indicative of problems--although it was


      recognized that that might be hard information to


      come by.




                Committee members wanted to know if the


      sites are going to know how they are ranked.  That




      would be very useful information for management to


      know about.  Right now self-inspections are a


      critical part of the quality system but the value


      of these would be diminished if that information


      were to become available to FDA.  This has been a


      longstanding concern of industry.  You know, you


      don't want to share your self-inspections because


      then they lose their value to you.




                Next talked about GMP guidance that's


      proposed for the production of Phase I drugs.  CMC


      review to ensure the identify, strength, quality


      and purity of the investigational drugs as they


      relate to safety.  This draft guidance is in


      process.  It's a risk-based approach.  No regular


      inspection program, but these Phase I drugs are


      looked at on a "for cause" basis.


                I want to point out that it was noted


      during that discussion that for Phase 2 and Phase


      3, those drugs still fall under the GMP


      regulations--21 C.F.R. 210 and 211.






                Also had an update on the PAT initiative.


      As Helen indicated, that guidance was recently


      finalized, in September.  It should be expanded to


      cover biotech products.  And, of course, it


      requires continued training of FDA staff.




                We also talked about--we had a full


      agenda--comparability protocol.  We had an update


      on guidances, The goal is to provide regulatory


      relief for post approval changes.  It requires a


      detailed plan describing a proposed change with


      tests and studies to be performed, analytical


      procedures to be used, and acceptance criteria to


      demonstrate the lack of adverse effect on product.


      Many comments have been received from the public.


      That was FDA's comment on this.  We did not see




                But the committee had comments, as well.




                Single use protocol has limited utility.


      It's more utility if you're going to have


      repetitive changes--if you're only going to do it




      once it may not help.  Specificity of the protocol


      may limit repetitive use.  Just how much


      specificity is needed?  And for a well-defined


      protocol, an annual report should be sufficient.


      That really will lessen the regulatory burden.




                Some general conclusions from our two


      days--and we've heard the first one several


      times--general principles are good, but case


      studies are needed to facilitate understanding.


      That came up time and time again.  Case studies


      should cover all industries; for example, dosage


      form, API, pioneer and generic.


                The committee expressed concern on what


      appears to be understaffing in OGD.




                Failure Mode &Effect Analysis can be


      linked with risk-based decision-making wherein the


      results feed into decision trees; training and


      education of both regulators and the industry in


      the new approaches is going to be key; historical


      inconsistency in regulator findings may limit the




      utility of surveys.  In the past, you know, not all


      investigators have investigated in the same manner,


      so it's difficult to compare results.


                And that's the end of my presentation.  I


      thank you for your attention, and would be happy to


      address any comments, now or later.


                CHAIRMAN KIBBE:  Are there any questions


      for Judy?


                DR. SINGPURWALLA:  I have some comments,


      but I probably would wait until all the


      presentations are over, and then make comments.


      Would that be acceptable?


                CHAIRMAN KIBBE:  Whichever way you want to


      do it, as long as it's within one of the two tails


      of the Bayesian distribution we're all right.




                DR. SINGPURWALLA:  You are confused, Mr.


      Chairman. [Laughs.]


                CHAIRMAN KIBBE:  On a regular basis.




                You had a question?


                DR. MORRIS:  Actually, just one comment to




      add to what you'd said, Judy, about the Georgetown




                I think they had made sort of a plea that


      the reason that they hadn't been able to go to the


      Indian and Chinese manufacturers was strictly a


      resource issue.  It wasn't that they had ignored


      that as an area of concern.


                DR. BOEHLERT:  Ken, thank you for that




                CHAIRMAN KIBBE:  Go ahead.


                DR. KOCH:  I guess, looking around on the


      schedule, I'm not sure if we're going to talk any


      about training.  You mentioned it in several


      different ways:  the continuation, the inclusion of


      industry, etcetera.  But will that come up as a


      discussion topic at some point?


                DR. HUSSAIN:  Not in this meeting.  I


      think we will eventually bring that back at some


      other meetings, though.


                MS. WINKLE:  Actually, when I talk about


      some of the organizational gaps I'm going to bring


      up training as part of that gap.  So if you want to




      comment then, it would be fine.


                CHAIRMAN KIBBE:  Anybody else?


                DR. SINGPURWALLA:  Well, maybe I'll speak


      now.  I just--we--this is a question more to


      Ajaz--about case studies and specifics.


                We've been through many sessions of the


      Manufacturing Subcommittee meetings.  Has there


      been any concrete plan made to start seriously


      undertaking some case studies?  And, if so, would


      you be kind enough to let me know?


                DR. HUSSAIN:  Yes.  Dr. Boehlert's


      presentation to this committee--she's the chair of


      the subcommittee--and the decision was made to form


      a working group under that.  And after this meeting


      we'll start populating that working group and


      create a working group under that committee to


      start addressing that.


                In addition to that, I think we're also


      looking at other parallel tracks to create case


      studies.  One such case study has just started to


      take shape, with Ken Morris, and then Purdue is


      working with our reviewers to actually develop a




      case study also.


                So we hope in the next several months we


      will have examples and case studies to outline the




                CHAIRMAN KIBBE:  Anything else?


                DR. SINGPURWALLA:  Yeah.  One other


      matter.  After the subcommittee meeting, some


      minutes were released, and I had made some comments


      about the minutes.  I did not receive an update of


      the minutes--update of the revision.


                Has--is there any reason for that?


      Because the normal protocol--the normal protocol is


      you put out the minutes, people give comments on


      the minutes.  You either incorporate those


      comments--and if you don't, you let us know why.


      And then you issue a final document of the minutes.


      And then the entire committee, or whoever it is,


      says "Yes, we go along with these minutes."  And


      they should become a part of the record.


                I was wondering if this was done, because


      I did not have access to that.




                CHAIRMAN KIBBE:  I think the final draft,


      or the final copy of the minutes is posted on the


      web page--FDA website--so that after the draft goes


      out to the members of the committee and the


      corrections come back in, they update to reflect


      the suggestions from each of the members, and then


      they post it.


                So if you wanted to check the website you


      could see whether--you know, how well your


      suggestions were incorporated in the final minutes.


                DR. BOEHLERT:  I would just add, also,


      that I reviewed comments that were made to the


      minutes before I made this presentation, and I


      tried to make sure that they were all incorporated


      in what I said today.


                DR. SINGPURWALLA:  I thought so.


                DR. BOEHLERT:  If they were not well


      reflected in the minutes, they should have been


      reflected in my comments today.  So--


                DR. SINGPURWALLA:  I thought so, but I


      wanted to see what the protocol was.


                DR. BOEHLERT:  Okay.  Thank you.  That's






                CHAIRMAN KIBBE:  Okay?


                DR. WEBBER:  One quick question.


                CHAIRMAN KIBBE:  Go ahead.


                DR. WEBBER:  That will be okay?


                You mentioned the pharmaceutical


      manufacture and research study, and I'm looking at


      the dates there. It seemed like it was fall of


      2003.  And I just wanted to confirm whether or


      not--that was during the period of transition of


      products from CBER to CDER.  Were our products in


      OBP--the biotech products that transitioned


      over--were they--are they completed now within


      CDER?  Or are they considered part of the CBER.


                DR. BOEHLERT:  Yes, I think Ajaz


                DR. HUSSAIN:  No, Keith, that's


      not--that's an external study that's focusing on


      all of manufacturing.  So all products--CDER and


      CBER--products are under.  It doesn't matter




                DR. WEBBER:  Where they were--just all


      products--okay.  Thank you.




                CHAIRMAN KIBBE:  Anybody else?  Good, that


      will keep us pretty well on schedule.


                I have now a "Parametric Tolerance


      Interval Test for Dose-content Uniformity"--Robert




                 Parametric Tolerance Interval Test for


                        Dose-content Uniformity


                DR. O'NEILL:  Magic button.  There we go.


                Good morning.  I'm Bob O'Neill.  I came


      before at the last meeting--I was asked to be the


      chair of a working group that you all blessed, and


      I'm here to give you an update on where we are on


      this issue of addressing the specifications for the


      delivered dose--uniformity of inhaled nasal drug






                Just to refresh your memory, the folks on


      the left-hand side are the FDA folks who are part


      of this working group, and some are more active


      than others--some of them, in blue, are part of a


      sub-group that has been put together that is


      working on more specific issues that I'll address




      in a moment; and the folks on the right--Michael


      Golden, in particular, who is a colleague on the


      industry side, who is coordinating our efforts in


      that area.




                The objective of this working group--as


      you probably know--is to develop a mutually


      acceptable standard delivered dose uniformity


      specification--that's both the test and the


      acceptance criteria--for the orally inhaled nasal


      drug products, with a proposal to come back to you


      all.  And that's the time frame that I'm talking


      about right now.


                So there's been a lot of work going on in


      the past few months, and that's what I just wanted


      to bring you up on.




                There have been three full working group


      meetings, where the folks on that previous


      slide--and some others--have come together at FDA


      for two, three hour sessions, and to go through


      information that has been presented to--primarily




      by the industry--to us to chew on.  And we have


      spent a lot of time internally talking to


      ourselves, and coming up with some additional


      issues and proposals, and we met the last time with


      the working group, and FDA had a proposal that we


      felt was moving in the direction of what everybody




                Subsequently, there's been a working group


      that will now be chewing on what was presented to


      the last joint meeting, and they're meeting


      November 4                                                th.  And

there's a lot of statistical


      issues; there's data analysis issues.  But I think


      what we're all on the same page with regard to is


      that the need to reassess the FDA--the past FDA


      recommendations, and I think there's--as we


      indicated the last time we briefed you--that the


      parametric tolerance interval approach is an


      improvement in a value-added type of testing


      strategy, over and above the zero tolerance


      interval strategy that's been used for awhile.


                So the next steps are the following.






                This working group is meeting--the


      sub-group is meeting in November, and we hope that


      they will then come back to the full working group


      by the end of the year, and we will evaluated the


      iteration between the FDA modification to the


      proposals that have been made by IPAC--and this has


      a lot to do with the placement of the operating


      characteristic curve for the acceptance criteria.


      Essentially, there have been many operating


      characteristic curves that have been shown to you,


      some of which are more steep, some of which are


      more shallow.  But where the proposal is being


      evaluated right now is:  how good is it at getting


      from an acceptance or rejection perspective, those


      assays that essentially are off target mean.  You


      can look at the performance characteristic, or an


      operating characteristic curve of a testing


      strategy if you assume that it's 100 percent on


      target. But the more you move away from 100 percent


      on target, the more you look at how well does it


      grab that, and how robust is it to allowing you to


      be a little off 100 percent?






                And so we're in the stages of looking at


      the statistical performance characteristics of


      that, and we hope that the working group will


      evaluate this proposal in more detail, and come


      back to you in the spring of 2005, with a final


      recommendation to discuss with you.  So that's sort


      of the game plan.


                And Michael Golden is here.  He's my


      colleague on the working group from the industry


      side, and we'd both be willing to take any


      questions if you have them.


                CHAIRMAN KIBBE:  Questions?




                DR. SINGPURWALLA:  Well, I guess Jurgen's


      hand went up before mine.  So--


                DR. VENITZ:  Okay, let me go first.


                DR. SINGPURWALLA:  He may ask the same




                DR. VENITZ:  Maybe.


                In your draft proposal--or what you're


      considering so far to be a draft proposal--




                DR. O'NEILL:  Yes.


                DR. VENITZ:  --are you considering the


      intended use when you look at statistical


      characteristics of your operating curve, for




                DR. O'NEILL:  Well, certainly that has


      been discussed, both from an emergency--a


      one-time-only, a chronic use, a medical risk




                DR. VENITZ:  Right.


                DR. O'NEILL:  --so, certainly, Dr.


      Chowdhury is involved, and others are involved, in


      considering this issue.  So--


                DR. VENITZ:   And I would encourage you to


      do that because, obviously, in my mind, it is


      different whether you're looking at inhaled




                DR. O'NEILL:  Right.


                DR. VENITZ:   --and you're looking at the


      performance of a drug product, versus a beta


      agonist, for example.


                DR. O'NEILL:  Yes.




                CHAIRMAN KIBBE:  Go ahead.


                DR. SINGPURWALLA:  Dr. O'Neill, we had


      this discussion when you made the first


      presentation, so I'm going to back--


                DR. O'NEILL:  Right.


                DR. SINGPURWALLA:   --to the same point




                I agree with you that tolerance interval


      approach is to be preferred to the zero tolerance,


      or something to that effect.


                DR. O'NEILL:  Right.


                DR. SINGPURWALLA:  But in your description


      of the next steps, you have talked about operating


      characteristic curves, and performance


      characteristic curves.  Of course those are not


      indicative of any Bayesian thinking towards this


      particular area.  And while you're in the process


      of formulating your plans, I strongly encourage you


      to incorporate that into your thinking.  You may


      not want to adopt towards the end, but at least it


      should be evaluated.


                And the second comment I'd like to make is




      that--and I'm certainly not volunteering and, if


      asked, I would refuse--the working group members


      consists of individuals from the FDA and from the


      pharmaceutical industry.  It would be good to have


      some neutral people on the working group--people


      from industry or people from government agencies


      that are not connected with the FDA, so that you


      get some sense of balance.  Otherwise, it seems to


      be--you know, it seems to be a self-serving group.


                So I would like to encourage you to expand


      your membership.


                DR. O'NEILL:  Yeah.


                DR. SINGPURWALLA:  And I want to


      emphasize:  I'm not available.


                DR. O'NEILL:  Well--no, the last point--I


      mean, this is hard work.  The people who are doing


      this work are spending a lot of time, and there's a


      lot of evaluation--a lot of data evaluation going


      on.  We were presented with information from the


      IPAC group that consisted of a huge database.


                And one could look at, well, how much time


      do you want to spend on evaluating a huge database?




      I mean, it's an electronic database, and lots of


      different--and where I'm going to on this is the


      Bayesian argument.  The Bayesian argument is very


      much a sensible argument--or a sensible framework


      when you can look at empirical data that allows you


      to feel pretty comfortable about what your priors


      are, and what the distribution of information is.


      That is not always accessible to the agency.  It


      may be accessible to a sponsor.


                So the strategy of being in-process and


      out-of-process, and being in control, and what's


      acceptable variability is very much--very much--a


      Bayesian framework, and very much within the


      context of how you may want to be looking at this,


      in terms of looking at in-process validation, as


      well as acceptance criteria.


                The extent to which that carries over into


      the type of testing we have to be very clear about.


      And it's--at the point we're at right now, we're


      essentially most interesting, or most concerned


      about how far out can you push the acceptance curve


      so that it has a proper balance between accepting




      and rejecting--particularly when we don't have, or


      no one can show us empirically, what the


      distribution of off-target means are, for example.


      How far away from 100 percent does the mean have to


      be before you want to maybe ratchet in this


      operating characteristic curve?


                So, I certainly could see the value to


      external folks' helping us out.  The more the


      better.  And I believe that this is a


      time-intensive effort.  And just as, you know, you


      would not like to volunteer, we would have to go


      and find folks who could invest the amount of time


      that is necessary, in the time frame that we're


      talking about, so we can get where we want to be.


                That's not to say that more brains are


      not--and independent brains--are--but this is--I


      would say we're pretty much trying to meet in the


      middle of this whole thing with resources that


      we've thrown out it that we feel are fair and




                DR. SINGPURWALLA:  Let me clarify.


                I'm not volunteering because I'm making




      the suggestion.


                DR. O'NEILL:  Yes.  Yes.


                DR. SINGPURWALLA:  And that's the proper


      thing to do.


                What I would like to encourage you is to


      involve at least two Bayesian's on your group--two,


      because they need support--




                --from the point of view of simply guiding


      a framework, or guiding the concept, and things


      like that, rather than get involved with the




                And the two individuals--or perhaps


      more--need not come from two stratified groups.


      They should come from somewhere else.


                So I'm making two suggestions:  one is to


      have people with expertise in Bayesian statistics


      involved, and to have people from outside these two


      communities also involved--perhaps in a limited


      way.  This will give you a broader perspective and


      will not subject you to criticism two years down


      the line.




                And that's the suggestion.


                DR. O'NEILL:  Okay.


                CHAIRMAN KIBBE:  Anybody else?


                Ajaz, do you have something to say?


      Reaching for your mike?


                DR. HUSSAIN:  I think the point I was


      going to make was, I think, at this point in time


      it's going to be difficult to add more people to


      the working group.  But the point is well taken


      that I think you do need to bring that perspective.


      And I'm hoping this Advisory Committee, and some


      other format, could be sufficient to sort of bring


      that framework for that--that perspective to bear


      on the progress of this working group.


                CHAIRMAN KIBBE:  No one else?


                Thank you Dr. O'Neill.  Appreciated your




                Dr. Ajaz, perhaps you could begin our next


      topic, and then we can take a break, because we're


      running slightly ahead, and it will give us a


      little flexibility as we move on.


                And so we're going to talk about Critical




      Path Initiative.


                The Critical Path Initiative--Challenges


                           and Opportunities


                 Topic Introduction and OPS Perspective


                DR. HUSSAIN:  Yes, I think I'm pleased


      that we have more time, because many of the


      presentations here are very lengthy


      presentations--[laughs]--including mine.


                I'd like to sort of introduce the topic of


      Critical Path Initiative--the challenges and






                The goals that we have for the fiscal year


      2005--and the initiatives, and the strategic goals


      at FDA level and the Department level are shown on


      this slide.  And the slide is from the "State of


      CDER" address by Steve Galson and Doug




                Today, our discussions will primarily


      focus on the Critical Path, the cGMP initiative,


      focused on risk management and innovation.  And the


      goal at the Department level is to increase science




      enterprise research.  But also, I think the follow


      on biologics, follow-on proteins, I think is


      interconnected to all of these discussions.




                My focus today is to introduce you to the


      topic of Critical Path, and also outline a proposal


      that we are contemplating at the OPS immediate


      office level as an umbrella proposal for all the


      discussions you'll hear today by scientists from


      different parts of the Office of Pharmaceutical




                But at the same time, some of the


      discussions in here also impact, say,


      counter-terrorism effort and other efforts that are


      ongoing.  And not all projects that we'll discuss


      are Critical Path projects today.




                What is Critical Path?  It's a serious


      attempt to examine and improve the techniques and


      methods used to evaluate the safety, efficacy and


      quality of medical products as they move from


      product selection and design to mass manufacture.






                In the continuum of drug discovery and


      development, you really go from basic research to


      prototype design or discovery, to preclinical


      development, clinical development, to an FDA filing


      and approval.  You have a focused attempt, say, for


      example, at the National Institutes of Health on


      translational research.  The Critical Path research


      does overlap with some of the aspects of the NIH


      translational research, but it covers predominantly


      the drug development aspects of the entire




                In our White Paper, we identified some of


      the challenges for Critical Path.  The drug


      development process--the "Critical Path" is


      becoming a serious bottleneck to delivery of new


      medical products.




                Our research and development spending has


      been exponentially increasing.  And as an index of


      1993, you can see the exponential increase from


      1993 to the current 10 years--increase in both




      private and public spending on research.




                However, new product submissions have


      remained flat--or, some would argue, are on the






                Why is FDA concerned?  FDA's mission is


      not only to protect but also to advance public


      health by improving availability of safe and


      effective new medical products.




                FDA has a unique role in addressing the


      problem.  FDA scientists are involved in reviewing


      during product development--they see the successes,


      failures and missed opportunities.  FDA is not a


      competitor, and can serve as a crucial convening


      and coordinating role for consensus development


      between industry, academia and government.  FDA


      sets standards that innovators must meet.  New


      knowledge and applied science tools needed not only


      by the innovators must also be incorporated into


      the agency's review process and policy.






                The challenge is how do we proceed?  It


      should be a science-driven and shared effort,


      drawing on available data, need to target specific,


      deliverable projects that will improve drug


      development efficiency.  It cannot just be an FDA


      effort.  We can identify problems and propose


      solutions.  Solutions themselves require efforts of


      all stakeholders.  We have issued a Federal


      Register notice requesting input from broad


      stakeholders, and we have received a number of


      suggestions, and we are working through those


      suggestions as we formulate our strategy for a


      Critical Path research program.




                This is a significant initiative, and the


      Department of Health and Human Services' Medical


      Technologies Innovation Taskforce is providing


      broad leadership.  Dr. Lester Crawford is chair of


      this Medical Technologies Innovation Taskforce, and


      it includes CDC, CMS, NIH and FDA.


                This taskforce is working on finding




      additional funding to meet the needs of the


      Critical Path program.  It is meeting with external


      stakeholders to identify opportunities, enlist


      allies, and so forth.




                In summary, I think from a Critical Path


      perspective, the present state of drug development


      is not sustainable.  We believe FDA must lead


      efforts to question any assumptions that limit or


      slow new product development:  are these


      assumptions justified?  Are there more efficient


      alternatives?  If so, why are the alternatives not


      being utilized?




                As we sort of focus on the discussions


      today, I'll remind you that the Office of


      Pharmaceutical Science is predominantly focused on


      one aspect:  Chemistry Manufacturing Control--or


      the initialization dimension.  But the Office of


      Pharmaceutical Science also supports many other


      aspects, from pharmacology, toxicology to clinical


      pharmacology research and so forth.  So, although




      our review responsibilities predominantly are on


      the quality side, our research programs are


      interconnected to every aspect of the drug


      development process.


                So you will hear presentations coming from


      all aspects--all three dimensions of the Critical






                The three dimensions are:  assessment of


      safety; how to predict if a potential product will


      be harmful; assessing efficacy; how to determine if


      a potential product will have medical benefit; and,


      finally, industrialization--how to manufacture a


      product at commercial scale with consistently high






                Our discussions, to a large degree, have


      focused on the third dimension.  And I think you


      will see, today, many of the projects within OPS


      that also impact the other two dimensions.




                In our White Paper, we defined the three




      dimensions and the connections to the Critical Path


      as follows:  safety, medical utility, and


      industrialization.  An every aspect--every box that


      is there has a need for improvement and research to


      support that improvement.


                Applied science is needed to better


      evaluate and predict the three key dimensions on


      the Critical Path development.


                I just returned from Europe--spending a


      week there last week--and with respect to the


      industrialization dimension, I came back somewhat


      depressed.  The amazing work I saw coming out of


      the University of Cambridge in the area of


      industrialization of pharmaceuticals--the approach


      to new technology, in terms of manufacturing, novel


      drug delivery systems and manufacturing processes


      itself, was astounding.  I don't see any of that in


      the U.S.


                So my concern is, much of the R&D and


      innovation is going to come from Europe and Japan,


      probably.  And unless we really improve our


      infrastructure, we are going to be lagging behind




      in a very significant way.  And I think that


      concern keeps growing on me, and I think I do want


      to sort of emphasize that.




                Office of Pharmaceutical Science programs


      and Critical Path Initiative--the discussion today


      is to seek input from you and advice, on aligning


      and prioritizing current OPS regulatory assessment


      and research programs, with the goals and objects


      of the Critical Path Initiative.  Please note that


      not all research programs and laboratory programs


      are intended to focus on "Critical Path."  There


      are equally important other aspects--bio-terrorism


      and so forth--which may not be considered as part


      of the Critical Path Initiative, but they're


      equally important.  So all of our programs and


      projects are not likely--or should not be part of


      the Critical Path.  There are aspects.  So you have


      to distinguish that.


                We hope that you'll help us identify gaps


      in our current program; identify opportunities for


      addressing the needs identified by the Critical




      Path Initiative.




                What I'd like to do today is--before I


      introduce Keith Webber--he took the lead on putting


      this program together--I'll share with you an OPS


      immediate office project that Helen and I have been


      developing.  These are our initial thoughts of how


      an umbrella project, within the OPS office, will


      help to sort of bring all of this together.


                So let me share some of our thoughts on a


      Critical Path project that OPS--Helen and I are


      sort of developing right now.


                An immediate need in OPS is to ensure


      appropriate support of general drugs--the growing


      volume and complexity of applications.  That's the


      challenge.  You saw the numbers increasing.


                In the New Drug Chemistry, the new


      paradigm for review assessment and efforts to


      support innovation and continuous improvement goals


      of the cGMP initiative--Office of New Drug


      Chemistry has taken the lead to be the first office


      to sort of implement all of this.  So they have




      significant need for support.


                Biotechnology products--complete


      integration into OPS, and the evolving concept of


      "follow-on protein products"--although I have put


      follow-on protein products under this, we don't


      know exactly how the regulatory process will


      evolve.  It could be--let's say, a work in




                And, clearly, alignment of research


      programs in OPS to meet our goals and objectives.




                So what are our thought processes, from


      our immediate office perspective?  To develop a


      common regulatory decision framework for addressing


      scientific uncertainty in the context of complexity


      of products and manufacturing processes in the


      Offices of New Drug Chemistry, Biotechnology


      Products, and General Drugs.


                Regardless of the regulatory process,


      regardless of regulatory submission strategies and


      so forth, we believe we need a common regulatory


      decision framework--a scientific framework--for




      addressing the challenges.




                What are the motivations here?


      Uncertainty--whether it's variability or knowledge


      uncertainty--and complexity are two important


      elements of risk-based regulatory decisions.  A


      common scientific framework, irrespective of the


      regulatory path or process for these products, will


      provide a basis for efficient and effective policy


      development and regulatory assessment to ensure


      timely availability of these products.


                That's the overreaching OPS goal, is to


      provide the common framework.  Although the


      submission strategies might be different, the


      science should not be different.




                How are we trying to approach this


      challenge?  We know that there are no good methods


      available for developing a standard approach for


      addressing uncertainty.  That means you need


      different approaches for different assessment


      situations. [Laughs.] All right, let me complete my






                So what we are thinking about--a decision


      framework for selecting an approach for addressing


      uncertainty over the life cycle of products is what


      is needed.  So you may have different approaches


      and so forth, but a common decision framework will


      help us identify the right approach.




                Project 1 is to create an "As Is"


      regulatory decision process map for the Office of


      New Drug Chemistry, Office of Biotechnology


      Products, and Office of Generic Drugs.  Much of


      this work will be done through a contract--we plan


      to have a contractor come in and work with us on


      some of these things.


                We think a representative sample of


      product applications could be selected for mapping


      the scientific decision process in the three






                Determine regulatory processes efficiency


      and effectiveness, using metrics similar to that




      what we have learned from the manufacturing


      initiative; and identify and compare critical


      regulatory review decision points and criteria in


      the three different offices; evaluate, correlate


      and/or establish causal links between review


      process efficiency metrics and critical decisions


      criteria, and available information in the


      submission--that's the mapping process; and, also,


      evaluate the role of reviewer training and


      experience, and how it bears on some of these






      Summarize available information on selected


      products; collect and describe product and


      manufacturing process complexity, post-approval


      change history, and compliance history--including,


      when possible, adverse event reports that come


      through MedWatch and other databases; describe


      product and process complexity and uncertainty with


      respect to current scientific knowledge;


      information available in submissions; reviewer


      expert opinions and perceptions; and, if feasible




      or possible, seek similar information from the


      sponsors or company scientists on these same


      products that we might select.




                What we hope to do is aim for the


      following deliverables:  organize Science Rounds


      within our office to discuss and debate the "As Is"


      process map, and the knowledge gained from the


      study; identify "best regulatory practices" and


      opportunities for improvement--these may include


      opportunities for improvement of filling the


      knowledge gap, develop a research agenda for all


      OPS laboratories based on what we learn.


                What is, I think, missing today is a


      common scientific vocabulary.  There's a need to


      develop a common scientific vocabulary to describe


      uncertainty and complexity.  There can be--each


      come from a very different perspective right now.


                Develop an ideal scientific process map


      for addressing uncertainty and complexity; adapt an


      ideal scientific process map to meet the different


      regulatory processes.




                In the following--I think the three


      projects that we're thinking about are not actually


      fully independent.  They're all connected together.




                Project 2 is to sort of focus on a systems


      approach.  We believe that without a systems


      approach to the entire regulatory process--that is


      from IND to NDA--Phase IV commitments and cGMP


      inspection, the broad FDA goals under the cGMP and


      the Critical Path Initiatives will not really be






                So the team approach and the systems


      perspective that evolved under the cGMP Initiative


      only addressed a part of the pharmaceutical quality


      system.  Quality by design and process


      understanding to a large extent is achieved in the


      research and development organization.


      Pharmaceutical product development is a complex and


      a creative design process that involves many


      factors, many unknowns, many disciplines, many


      decision-makers, and has multiple iterations in the




      long life-cycle time.


                So we have to treat it as a complex system


      optimization problem.




                Significant uncertainty is created when a


      particular disciplinary design team must try to


      connect their subsystem to another disciplinary


      subsystem--for example, clinical versus chemistry,


      or CMC to GMP.  When you bring those connections,


      there's significant uncertainty.


                Each subsystem can have its own goals and


      constraints that must be satisfied along with the


      system-level goals and constraints.  It is possible


      that goals of one subsystem may not necessarily be


      satisfactory from the view of other subsystem and


      design variables in one subsystem may be controlled


      by another disciplinary subsystem.  Impurities is a


      good example.  Pharmtox, CMC, and how you bring


      that together.




                So the Project 2 proposal that we're


      developing is to use ICH Q8 as the bridge between




      the cGMP Initiative and the rest of the regulatory


      system, and to develop a knowledge management


      system to ensure appropriate connectivity and


      synergy between all regulatory disciplines.  Can


      that be done?  I mean, that's the feasibility


      project that we are trying to develop.  So--connect


      Pharm/Tox, Clinical, Clinical Pharmacology,


      Biopharmaceutics, CMC, Compliance all together.




                The current thinking is to approach this


      problem as connecting every section within the ICH


      Q8 CTD-Q, within the same document, but to all


      other sections in an NDA, in some way or form.  For


      example, each section within the P2 can have an


      impact on the other P2 sections and, similarly,


      other sections of a submission and to cGMP.


                By recognizing this as a complex design


      system that involves multiple attributes, goals,


      constraints, multidisciplinary design teams,


      different levels of uncertainty, risk tolerances,


      etcetera, we wish to find opportunities to identify


      robust designs and design space that provides a




      sound basis for risk assessment and mitigation.


                So this would be a scientific framework.


      It was a regulatory tool that could come out of


      this.  And with the case studies and everything


      coming together, this might be a way to bring and


      connect all the dots.




                What we have been looking out is outside


      pharmaceuticals.  We believe that a significant


      body of knowledge exists.  Example, in mechanical


      engineering, as it applies to the design of


      aircrafts, that addresses some of these challenging


      points that we have discussed.  These are three


      examples that I have selected as just illustrative


      examples of how multidisciplinary optimization


      methods and system-level problem solving tools can


      be thought about in the drug context.




                Just to illustrate this point, let me


      create an example here.  The applicability of


      multidisciplinary optimization methods for solving


      system-level problems and decision trade-offs will




      be explored in an NDA review process.  That's what


      we're proposing.


                For example, in the Common Technical


      Document for Quality--the P2 section, which is what


      ICH Q8 will define--critical drug substance


      variables that need to be considered in section


      2.2.1, which is "Formulation Development" are


      described in section 2.1.1.  So there's a drug


      substance, and there's a formulation.  They're two


      different sections.


                Information for "Drug Substance," has a


      bearing on that of the "Formulation Development."


      So how do you connect the two together?


                For example, the current language in ICH


      Q8 for "Drug Substance," states:  "Key


      physicochemical and biological characteristics of


      the drug substance that can influence the


      performance of the drug product and its


      manufacturability should be identified and




                So that's describing the information


      content in section 2.1.1. that we will hopefully




      receive whene ICH Q8 is done.  So how does this


      have a bearing on the "Formulation Development"






                I'll skip this and just show you a figure.




                You have the API--or drug substance


      manufacturing process.  The X(1.1) is the design


      variable; the f(1.1) is the objective function to


      be addressed; and the g(1.1) is the constraint for


      that manufacturing process that delivers the drug


      substance.  Okay?


                Since this is not part of ICH Q8, what


      will be part of ICH Q8 is section 2.1.1., which


      will identify what are the critical variables for


      the drug substance, as they relate to the


      formulation aspect.  But that becomes the input for


      what--how it connects to the "Formulation


      Development" aspect.  And that link is through a


      linking variable.


                Since my means and standard deviations


      have become finger-pointing and so




      forth--[laughs]--so you know--you have a design


      variable, you have a linking variable, you have an


      objective function, you have constraints around


      which you define your design space.  You have mean


      objective function--that's your target.  You have a


      standard deveiation that you sort of bring to bear


      on that.  And deviation range of the design


      solution, or the design space.


                So all of this sort of has to come


      together for this to be meaningfully connected.


      And, for example, if you start with a simple design


      of experiment, you may have mathematical models,


      which are empirical, but then they provide that


      connectivity.  So it's a start of a very formal,


      rigorous approach to dealing with uncertainty,


      knowledge gaps and complexity.


                So this might be a useful concept.  So


      that's the process right now, to see whether this


      could be a feasibility project that we could do.




                So the potential deliverables of using


      this approach could be significant. Since we are




      moving towards electronic submissions, in


      conjunction with electronic submissions, this


      project can potentially provide a means to link


      multidisciplinary information to imporve regulatory


      decision--that is, clinical relevance to CMC


      specifications.  We may not all have all that


      information, but the links--the structure--will be


      there as we grow, as we improve our knowledge base,


      or will it be refined, the links could get


      populated, and this might be an approach for


      knowledge management within the agency.


                Creating a means for electronic review


      template and collaboration with many different


      disciplines; provide a ocmmon vocabulary for


      interdisciplinary collaboration; create an


      objective institutional memory and knowledge base;


      a tool for new reviewer training; a tool for FDA's


      quality system--and, clearly, it can help us


      connect cGMP Initiative to the Critical Path




                So that's the project that we hope to


      develop.  We really want to get some feedback from




      you, and develop this as a project under the


      Critical Path Initiative.




                But the third aspect of this--it all could


      happen in parallel--explore the feasibility of a


      quantitative Bayesian approach for addressing


      uncertainty over the life cycle of a product.  The


      most common tool for quantifying uncertainty is


      probability.  The frequentists--the classical


      statisticians--define probability as "limiting


      frequency, which applies only if one can identify a


      sample of independent, identically distributed


      observation of the phenomenon of interest."


                The Bayesian approach looks upon the


      concept of probability as a degree of belief, and


      includes statistical data, physical models and


      expert opinions, and it also provides a method for


      updating probabilities when new data are




                The Bayesian approach may proivde a more


      comprehensive approach for regulatory decision


      process in dealing with CMC uncertainty over the




      life cycle of a product.  It may also provide a


      means to accommodate expert opinions.


                And I think there's a connection here.


      The evolving CMC review process may be a means to


      incorporate expert opinions.  And I think that is a


      significant opportunity.


                Using the information collected in Project


      1--that I described--you would seek to develop a


      quantitative Bayesian approach for risk-based


      regulatory CMC decision in OPS.


                So that would be a project that will run


      in parallel to the other two approaches that we are


      moving forward.




                So, I'll stop my presentation here with


      sort of summarizing, in the sense--I think OPS,


      from its goals and objectives, has to have an


      overreaching project that sort of connects all the


      dots together.  And the proposal--the first one


      clearly is a process map--"As Is" and so forth.


      But the two others are feasibility projects that we


      want to look at the Bayesian approach and a complex




      system optimization problem.


                The knowledge exists outside.  It's simply


      adapting and adopting it in our context.


                What you'll hear--after the break, I


      think.  Or--unless you want to start earlier--after


      the break, is other immediate office projects;


      Office of Biotechnology projects, Office of New


      Drug Chemistry project, Office of Generic Drug


      projects on Critical Path, and Office of Testing


      and Research.


                What we have done is Keith Webber will


      introduce the reset of the talks.  You will hear


      each group's perspective.  And we have requested


      Jerry Collins to come back and sort of


      summarize--after his talk on the Critical Path--the


      entire Critical Path Initiative from an OPS


      perspective and pose questions to you.


                And we have also invited Professor Vince


      Lee, who is now part of FDA--who used to be the


      chair of this committee--who has been with agency


      for almost a year now, to come with his perspective


      on how--what are challenges he sees.  So you will




      hear sort of presentations and some opinions from


      people who have been at the agency and been looking


      at this challenge for some time.


                So, again, the discussion today is to seek


      input and advice on ACPS; on how to align, identify


      gaps, and identify opportunities.


                I'll stop here and entertain questions on


      my part of the presentation.


                CHAIRMAN KIBBE:  Are there any questions?


                DR. SINGPURWALLA:  I have comments.


                CHAIRMAN KIBBE:  Okay.  Thank you.


                DR. SINGPURWALLA:  I just--what you say is


      music to my ears.  You have good vision about some


      of the things you want to do.  But I think it's now


      time that the dance should begin.


                We should get back--take concrete problems


      and address them.  I've said this before.


                But let me just make some specific


      comments on some of the things you've said.  And,


      of course, I'm going to question some of the things


      you said.


                The first argumetn I want to make on your




      slide on page 7, about efficacy and safety:


      generally, those tend to be adversarial.  Drugs


      that give you benefit may have side effects.  So


      the important issue is to do a trade-off.  For that


      you need to talk about assessing utilities:  what


      is the utility of the benefit, and what is the


      dis-utility of the harm?  That's a part of the


      whole package of thinking about these problems, and


      I encourage you to look into it.


                Now, I take strong objection to some of


      the things you have said.  You have distinguished


      uncertainty into stochastic and epistemic.  I have


      seen that distinction before.  I claim it's totally


      unnecessary.  Uncertainty is uncertainty, and one


      doesn't--one should not pay much attention to the


      source of the uncertainty--


                DR. HUSSAIN:  Right.


                DR. SINGPURWALLA:   --whether it is


      regulated allatoire uncertainty, or epistemic, does


      not matter.


                CHAIRMAN KIBBE:  Right.


                DR. SINGPURWALLA:  The Bayesian approach




      does not distinguish between the two.  And since


      you've been talking about it, I think--


                You also say that there are no good


      methods for devleoping standard approach for


      addresing uncertainty.  I think that's the wrong


      slide to put up.  That's liable to do more harm


      than good.


                DR. HUSSAIN:  Okay.


                DR. SINGPURWALLA:  There are methods


      available.  So I would not encourage you to put it.


                And the other thing is:  I don't like your


      linking uncertainty and complexity.  They're two


      different issues.


                And you also say that there is no common


      scientific vocabulary.  Well, I claim there is a


      common scientific vocabulary, and that is




                Now, as far as recommendations are


      concerned:  I'd like to suggest--and, again, I'm


      not volunteering since I'm making the


      suggestion--that you have your people exposed to a


      tutorial on Bayesian methods and Bayesian ideas, so




      that you get a better appreciation of what it's all


      about.  And the best way to do this is to take a


      simple example and work through it; work through


      your expert opinion notions that you're saying.


                Go through an example, and you'll get a


      better appreciation of what it's all about.  And


      once you get that appreciation, you'll be tempted


      to remove some of the other things you've said.


                Those are just comments.  Thank you.


                DR. HUSSAIN:  No--the point's well taken.


      And we actually have a project right now with the


      University of Iowa, looking at our stability data


      from a Bayesian perspective.  So we're just


      starting to put a real-life example on that.  So




                With regard to the utility, Jurgen and the


      Clinical Pharmacology Subcommittee has been sort of


      bringing that up.  So we will connect to the


      Clinical Pharmacology group.


                Jurgen, do you want to say anything about




                DR. VENITZ:  [Off mike.] Well, other than




      the fact that--other than the fact that we're


      discussing it.  It is a controversial issue,


      because you're really trying to map, then, a lot of


      different things into a uniform scale.  Personally,


      I don't see an alternative, and I think it's


      already done.  We're just doing it intuitively, as


      opposed to expressedly.


                So it is being discussed.  We have to see


      where it goes.


                DR. HUSSAIN:  And, regarding, I think, the


      common vocabulary, I think it's a common vocabulary


      in the context of when we speak from a pharmacist


      to a chemist to an engineer--we have very different


      interpretation--that's what was referred to.


                DR. SINGPURWALLA:  That's why you need a




                DR. HUSSAIN:  That's exactly--


                DR. SINGPURWALLA:  Put people together.


      Because about 15, 20 years ago, the Nuclear


      Regulatory Commission was facing similar problems.


      And one of the things they did is they had lots of


      tutorials to get everyone on board, talking the




      same language.  Otherwise, you'll have a doctor


      talk to an engineer, and those two talking to a


      lawyer--and you know what can happen.




                VOICE:  [Off mike.] Lawsuits.


                CHAIRMAN KIBBE:  Another question?


                DR. KOCH:  I guess, just to build on the


      last comment--when you get into all those


      multidisciplinary functions--and particular when


      the ICH Q8 is going to serve as a group, together


      with the implementing the cGMPs--there's a couple


      of organizations out there I think could serve as


      very valuable resources.  One we've heard about a


      couple times today in the ASTM 55, as a body to


      help at least standardize the terminology.  And the


      other one is the ISPE, which could serve as a


      multidisciplinary conduit that, working together


      with ICH, could probably facilitate some of the


      multidisciplinary issues.


                DR. HUSSAIN:  I think we do plan on


      extensive training and team building and coming on


      the same page.  If you look at the PAT and the




      manufacturing signs White Paper that we issued, we


      actually laid out a lot of these things in there,


      including the role of ISPE, ASTM, PQRI, and so




                So we have been thinking about this in


      that context, and at the ICS meeting in


      Yokahama--on Wednesday, I think, the date is


      set--we will be updating on that.  So I'll get a


      chance to talk about ASTM to ICH in Yokahama,


      Japan, also.


                So, we're aligning everything together.


      So that's happening.


                There was one point that I wanted to


      respond to:  the reason for keeping uncertainty, in


      terms of variability in knowledge--keeping the


      distinction, at least as we think about this,


      was--and the link to complexity, also--clearly,


      complexity and uncertainty are two independent


      things.  But, unfortunately--well, the challenge we


      face is this--in the sense we have a very complex


      product.  We have simple products--within the same


      office, in OPS and different regions.  Yet today, I




      think, from a variability perspective, we're not


      very sophisticated in how do we deal with




                And, for example, in our manufacturing


      science White Paper, we don't even deal with


      variability of our dissolution test method.  We


      don't even know how to handle it.  So we have


      challenges today where simple variability--we don't


      have a good handle on.


                So that was the reason for keeping


      variability and knowledge-based uncertainty on the




                CHAIRMAN KIBBE:  Ken?


                DR. MORRIS:  Just a quick question:  on


      your identification of the gaps in the current


      programs, are you thinking more in terms of


      technical gaps--as in science that needs to be


      done?  As opposed to logistical gaps within--


                DR. HUSSAIN:  Both.  Both.


                DR. MORRIS:  So, with respect to the


      scientific gaps, are thinking, then, to take it one


      level more--basically, are you talking more about




      new science that needs to be created?  Or science


      that needs to be communicated more


      effectively--within the agency?


                DR. HUSSAIN:  Well, I think the immediate


      need would be to communicate the existing science


      and bring all the existing knowledge to bear on


      that.  And, clearly, in the long term there are


      fundamental issues--and most of the new science


      would be needed.  So I think it's an issue of




                DR. MORRIS:  Thank you.


                CHAIRMAN KIBBE:  Anybody else?  Nozer, you


      wanted to--


                DR. SINGPURWALLA:  No, I just wanted to


      say that this distinction between allatoire and


      epistemic has been artifically created by


      frequentist statisticians.  And Bayesians don't buy




                DR. SELASSIE:  I have a question.


                CHAIRMAN KIBBE:   Yes, please.


                DR. SELASSIE:  You know, in your graph on


      R&D spending--has there ever been a breakdown in




      how much of that spending can be attributed to the


      "R" and how much to the "D?"


                DR. HUSSAIN:  I don't have that--I'm sure


      that information's--I don't have it.  So I'm not


      aware of it.


                DR. SELASSIE:  Because would one parallel,


      you know, the flatness?


                DR. HUSSAIN:  One was the public funding;


      one was more private funding, so--


                DR. SELASSIE:  Yes, but they're both going




                DR. HUSSAIN:  Yes.


                DR. SELASSIE:  But I'm wonder if, you


      know--because you look at your product submissions


      are flat.  Now, is that because there's not been an


      increase in development funding?  Or--


                DR. HUSSAIN:  I don't think so.  But I


      don't have an answer.


                DR. SELASSIE:  Yes.


                DR. HUSSAIN:  So let me say that.


                CHAIRMAN KIBBE:  Marvin?


                DR. MEYER:  Ajaz, you don't seem like a




      depressive kind of guy--


                DR. HUSSAIN:  [Laughs.]


                DR. MEYER:   --but you said you were


      depressed last week.


                DR. HUSSAIN:  Yes.


                DR. MEYER:  Can you give us just a real


      short synopsis of where you see Europe doing things


      right, and us doing things wrong?


                DR. HUSSAIN:  [Sighs.] [Laughs.]


                No, I mean, again, I'll focus on what I


      see happening in Europe--especially in the


      U.K.--and how they're translating academic


      research--academic finding research--into


      entrepreneurial business--in particular in


      manufacturing, in particular in dosage form


      design--the pharmacy-related ones.


                Look at Bradford, particle engineering.


      And the one I saw--I saw a beautiful manufacturing


      system for coating.  Forget coating pans.  This is


      electrostatic coating; precise, automated, complete


      on line, and so forth.


                Nothing of that sort is happening




      here--within my domain.


                CHAIRMAN KIBBE:  We have a couple more


      comments, and then we're going to have to take a




                Go ahead.


                DR. MORRIS:   Yes, just to follow up on


      that.  I think there's--I just came back from


      Europe depressed, as well, but I was in


      Scandinavia.  So maybe that had something to do


      with it.




                Yeah, it's pretty dark up there.


                But, in any case, I agree with Ajaz in


      that there are a couple of caveats and, in fact, if


      you look at our latest hires, they're one from--via


      Bradford, another one via Bath.  My post-doc is


      from Nijmwegin, another post-doc from Roger Davies


      Group in the U.K.


                And we're not training people--number one.


      So, aside from not transferring the technology


      effectively we're not training people to do it very


      much any more.  There are few places--represented




      at the table--that still do it to some degree.


                But that stems back to one of your earlier


      slides, which is trying to muster NIH and NSF to


      fund this sort of research.  Because some of you


      have been a lot closer to deanships than I.  If


      there's no overhead money, it doesn't get a very


      kind reception.  And the fact of the matter is is


      we haven't had it.


                So, this is--I'll stop here, because this


      is my old soapbox.  But I lay this at the door, in


      part, of NIH and NSF for not recognizing, in the


      face of overwhelming data, that there is a crisis


      that needs to be address.


                On the upside, there are some people in


      Europe doing some things--and Japan, as well.


                CHAIRMAN KIBBE:  Pat, go ahead.


                DR. DeLUCA:  Just a quick follow up on


      that, too.  I know from my trips to Europe, too, if


      you just look at the colleges--the pharmacy schools


      in Europe--I mean, they all have departments of


      pharmaceutical technology.  I mean you'll be


      hard-pressed to find pharmaceutical technology as




      an area of focus in an American college of pharmacy


      now.  Certainly you won't see any departments of


      pharmaceutical technology.


                So I think it's been--and it wasn't that


      way 20 years ago.  But, I mean, it certainly has


      changed, though.


                CHAIRMAN KIBBE:  Anybody else?


                Good--I think we're at a nice break point.


      And if we could take perhaps a 10 minute


      break--because Ajaz has managed to get us--use up


      all of our lead time.




                And we can get Keith to start his talk at


      about 10:22, that would be great.


                [Off the record.]


                CHAIRMAN KIBBE:  22 minutes after 10 has


      arrived, and one way or another we're going to get


      back on process.


                Dr. Webber, are you prepared to get on




                He's on the way to the podium.


                Those of you walking around with cakes in




      your hands, and sodas, you want to sit down.


      Nozer.  Here we go.  Good luck.  We gave you 10


      minutes to do that.




                You snooze, you lose, as the old saying




                So, Dr. Webber, shall we start our


      Strategic Critical Path?


             Research Opportunities and Strategic Direction


                DR. WEBBER:  Okay.  I guess we're about


      ready to get started on this session, regarding


      research activities and our strategic goals for the


      Office of Pharmaceutical Science.


      I'm Keith Webber, with the Office of Biotechnology


      Products.  And let me--




                --there we go.


                Ajaz went through a very good


      presentation, I think, on the Critical Path.  And


      I'm not going to really address very much about the


      Critical Path Initiative itself.  But, in my view,


      this--I've sort of summarized things into the Drug




      Development Path, which begins with discovery of


      potential targets--or potential new drugs; and then


      you have to have a period where one evaluates the


      candidates and makes a selection of what candidate


      you should carry forward into the pre-clinical


      study, where one looks for potential toxicities and


      potential efficacies in an indication of interest.


                If all goes wlel, one moves into clincal


      studies, and if all goes even better, into


      commercialization.  And then, once you're on the


      market, there's always the period of post-approval


      manufacturing optimization--or we would like to see


      that, from the FDA's perspective, anyway.


                And then, often, we get new


      indications--we see new indications being developed


      for drugs that are on the market.  And that


      essentially starts the process back up again--often


      at the clinical studies stage.




                The--I didn't bring a pointer.  Is there a


      pointer here?


                VOICE:  [Off mike.]  Just use the mouse.




                DR. WEBBER:  Just use the mouse.  Okay.


      That will work.  Right there is the mouse.  Okay.


                I guess, historically, FDA interactions;


      have occurred primarily ion this area here, from


      clinical studies on.  Prior to that, we have had


      very little influence, I think, until we receive a


      submission which contains information regarding the


      pre-clinical studies.


                But I believe we have opportunities to


      have an impact on this entire process in the




                Let's see.


                Essentially, I guess, sort of the essence


      of the Critical Path is the--in my mind--is the


      view from empirical versus guided drug development.


      And drug development has to be a learning process


      in order to make intelligent decisions regarding


      such issues such as your candidate selection; what


      dosage form you're going to have and what the


      formulation should be; in choosing clinical


      indications, you need to know what patient


      population is going to be the best selection for




      your product.  And then when you're evaluating


      clinical endpoints, one needs to know which are the


      most appropriate endpoints to evaluate in the


      clinical studies, and are there surrogate endpoints


      that are more appropriate than others, if you can't


      look at an endpoint which is directly related to


      survival or efficacy in the more normal manner.


                And, of course, with adverse event


      monitoring, any clinical trial is going to monitor


      particular parameters, and you need to have a good


      knowledge base in order to understand which adverse


      events we should be looking for, and the best way


      to evaluate those.


                And then, finally, the manufacturing


      method certainly is a major concern because that


      has to do with the ability to improve the


      manufacturing process post-approval and


      pre-approval, as well as avoiding issues that can


      come up with regard to safety and efficacy of your






                The goal of industry, as well as the




      agency, I believe, is to establish a knowlege base


      and the tools that are necessary to predict the


      probable success of any given product, and the


      manufacturing methods that are appropriate to it,


      and then to foster the development of products that


      are going to have a high likelihood of success,


      throughout clinical development and on the market.




                Now, for this late morning's presentations


      and this afternoon's presentations, we'll be


      hearing from a number of groups within OPS.  One is


      the Informatics and Computational Safety Analysis


      staff, which is in--essentially in the immediate


      office of the OPS; and then Office of New Drug


      Chemistry, Office of Generic Drugs.  And the first


      three here are the groups that do a lot of


      relational and database analyses as part of their


      research activities.  There are, in some cases,


      collaborative research going on with laboratories,


      per se.  But it's the groups on this--the last two,


      the Office of Testing and Research, and Office of


      Biotechnology Products, that have actual




      laboratories where research at the bench is going






                Let's see--within OPS's Critical Path


      Research, I think we can address--or can address


      the issues regarding candidate selection, based


      upon an understanding of the structure and activity


      of the relationships that we see, and the products


      that ocme down the line, as well as what's reported


      in the literature.


                Dosage form development and evlauation I


      think is an important area that we're working in.


      Toxicity predictions for products is--we're


      amenable to that, so our research can address that


      through, again, structure activity-type


      relationships and structure-function issues, as


      well as knowledge of the impacts that a particular


      disease state might have on physiological function


      that may lead to toxicities that wouldn't be


      present in all populations.


                Bioavailability and bioequivalence


      predictions are certainly important for all of our




      products, but particularly for the Office of


      General Drugs, they're quite critical.  And I think


      with regard to the follow-on products as well, it's


      a major area of concern.


                Metabolism prediction is something that


      is, I think, crucial because products, once they


      enter the body, as you know, they don't remain in


      their initial state.  And the metabolism can impact


      toxicity, it can impact efficacy, it can impact the


      bioavailability and biofluence of the products




                Immunogenicity is another area that is of


      large concern, particularly for protein products.


      And there we need to evaluate and understand, not


      only the caues of immunogenicity, or the impacts of


      various structures in the proteins on


      immunogenicity, but also the impact that the


      patient population has on immunogenicity; what


      impact the indication that's selected can have on


      impacts of immunogenicity as a safety concern.


                Often, as I mentioned earlier, you have


      biomarkers that you're looking at for




      pharmacodynamic parameters, or for surrogate


      endpoints.  And a good knowledge of the validity of


      a particular biomarker, and our ability to evaluate


      those, as well as industry's ability to select


      those, is dependent upon the knowledge that they


      have of the biology of the disease that they're


      studying, or that they're trying to cure or that


      they're trying to treat.


                The mechanism of action of the drug is


      certainly critical when you're looking at the


      potential.  One area is with regard to drug-drug


      interactions.  Oftentimes we've been looking


      primarily at metabolism for drug reactions, but


      certainly there's a concern that I think is


      building for utilization of multiple drugs that


      impact on the same metabolic--not metabolic


      pathways, but the signaling pathways, let's say, at


      the cell surface, which are getting the


      treatments--you know, getting a treatment into the


      cell, or that are resulting in the clinical


      effect--is what I'm trying to say, in a very poor






                Let's see--the pharmacogenomics is a new


      area that we're getting involved in, but it's very


      important with regard to patient selection, as well


      as the potential for certain populations to be


      impacted by drugs in a unique way, that can impact


      not just efficacy, but also the safety.


                And manufacturing methodologies are an


      area that we have research programs in within the


      office, and those are important for developing and


      understanding of the robustness of various


      manufacturing processes, and the ability to


      implement new paradigms, such as process


      technologies in the manufacturing process of






                Out strategy here is to coordinate


      cooperative research activities.  And, as I


      mentioned, we have predictive modeling programs.


      And these are generally based upon information from


      regulatory submissions that we receive, as well as


      from laboratory research that's going on within the


      agency, as well as outside and in the published






                One area which, I think, we need to build


      is our abilities to get information from industry


      that we don't get in a our regulatory submissions,


      and that they don't publish, and finding a means to


      have them help us to gain knowledge of that


      information so that we can implement it into the


      decisions we make and share that--basically the


      conclusions that come out of that with industry as


      a whole, to address the Critical Path.




                There's also laboratory research going


      on--you'll hear from the Offices of Testing and


      Research, Applied Pharmacology Researhc, and


      Product Quality Research, and Pharmaceutical


      Analysis--and also from my office, Biotech


      Products, from our divisions of Monoclonal


      Antibodies and Therapeutic Products--it should be


      Therapeutic Proteins.  Sorry.  Typo there.


                There's also research going on in other


      FDA centers that we can collaborate with, and do


      collaborate with, as well as outside, to gain




      information from academia, industry and other


      egoernment agencies, as well.




                Now, I think we can gather all this


      information, but it's critical with regard to how


      we're going to use it, and how we're going to


      disseminate it, such that we can have an impact on


      the Critical Path.


                There are a number of avenues to get to


      academia and manufacturers, and those include the


      public forums, where we can present the conclusions


      and recommendations.  We certainly write guidance


      documetns that can help in this manner, as well.


      And then, when industry comes to meet with us at


      the regulatory meetings, such as pre-IND, and


      pre-NDA meetings--pre-BLA meetings--we can interact


      with them at those points, as well.


                But we also need to change, to some


      extent, our review processes within the agency,


      and--so the information has to go to the reviewers,


      as well.  And we can do that via training programs,


      as well as the guidance documents that we do write.




      They're used a great deal by the reviewers.


                Then, again, mentoring programs, to bring


      up the new reviewers in an understanding of the new


      paradigms and new concerns, or lessen their


      concerns for particular issues that relate to


      pharmaceutical manufacturing, or clinical issues.


                And then all of this together should help


      to enhance the application of your process from the


      reviewer's standpoint, and with regard to the


      manufacturers should help to remove some of the


      hurdles and obstacles we see in the Critical Path.




                You'll hear the coming presentations.  So


      there are some questions we'd like you to keep in


      mind, that we'll be bringing up later for




                And first is:  are we focusing, within the


      office, on the appropriate Critical Path topics?


      And are there other topics that we should be


      addressing through our research programs?  And it's


      both the database relational type information or


      research programs as well as the laboratory






                And then, in the future, Critical Path


      issues may change.  So how should we identify


      Critical Path issues in the future.  And we'd like


      recommendations on how we should prioritize those.


      Because we're really--at this point, we can't do


      everything that needs to be done with the current


      resources, and so we're going to have to prioritize


      now, and in the future we'll need to prioritize, as


      well, and we'll need some guidance on that.


                That ends my presentation.  We'll move


      into the first talk--to stay on time--which is


      going to be--let's see, I'll bring it up here--Joe




             Informatics and Computational Safety Analysis


                             Staff (ICSAS)


                DR. CONTRERA:  Okay.  I'm the director of


      the Informatics and Computational Safety Analysis


      group.  Our main mission, really, is to make better


      use of what we already know; material or safety


      information, toxicology information that's buried


      in our archives; and also in the scientific




      literature and in industry files.


                Our group develops databases and also


      predictive models.  You can't develop models


      without the databases.  So they go together.


                We have develop our own paradigms for


      transforming data, because traditional toxicology


      data is textual, and converting into a weighted


      numerical kind of a scale that is amenable to be


      processed by computers, and also to be modeled.


                And we encourage, promote and also work


      with outside entities to develop QSAR--qualitative


      structure activity relationship software--and data


      mining software, for use in safety analysis.


                We don't work alone.  And you'll hear more


      about this in my talk.  We leverage, very much, and


      cooperate, and collaborate very much with


      outside--with academia, with software companies and


      with other agencies.  And we do this through


      mechanisms such as the CRADA--the Cooperative


      Research and Development Agrement--which is really


      a buisness agreement--and also we do it with


      Material Transfer Agreements, for an exchange, quid




      pro quo exchange, with software and other


      scientific entities outside the center.




                The Critical Path Initiative--you've all


      been, and you're going to be hearing more about it,


      and you've heard a lot about it.  I'm focusing on


      what is relevant to my group, and that is:  the


      problem is that we have not created sufficient


      tools to better assess safety and efficacy.  We're


      still relying on toxicology study designs that were


      designed 50 or sometimes 100 years ago.  And it


      doesn't mean that they're inferior, but maybe there


      are better ways of doing this now.


                So we need a process to develop better


      regulatory tools.  And it was really a controversy,


      to some extent:  whose misison is this?  And in the


      past, the agency didn't consider it as the agency


      mission to develop these tools--necesarily.  It was


      academia.  And academia said, "No, it's the


      industry."  It wasn't--it was vague as to who was


      actually responsible for developing new analytic


      tools that can be used for regulatory




      enpoints--especially in safety endpoints.




                So now how d we connect with the citical


      path?  I think we were doing Critical Path research


      well before there was a Critical Path Initiative.


      I mean, we've been in operation, in one form


      another, for over a decade in the Center, at a time


      when people were questioning whether this was the


      mission of the agency in the beginning.


                We developed databases and then predictive


      tools that are used by the industry--by the


      pharmaceutical industry--more and more to improve


      the lead candidate selection.  And the question


      was:  why should the agency supply industry with


      better tools to select lead candidates?  Well, it's


      in our interest that they develop lead candidates


      that have fewer toxicology or safety problems.


      Because when they come to us, in the review process


      and submissions, they can said right through with


      very few issues.  Otherwise, they'd bog down the


      system.  And we have multiple review cycles, and


      there are issues to be addressed.  And it would be




      wonderful if they could just slide through.


                And so also to facilitate the reiew


      process internally, by having reviewers having a


      rapid access to information that is usable for


      "decision support," we call information; that they


      can use to make judgments on a day-to-day basis.


      And we hope that also this could reduce testing;


      reduce the use of animals.  And also encourage


      industry--software companies--to get into the


      business of developing predictive modeling tools.




                And we see this three-dimensional diagram


      for the Critical Path.  Well, the computational


      predictive approaches are identified in two of the


      three pathways.  And so we feel we're right in step


      with what the future goals of the agency are.




                What have we accomplished already?  Well,


      again, we do two things:  databases and predictive


      modeling.  And this sort of summarizes some of the


      accomplishments; the first being we've developed


      predictive software for predicting rodent




      carcinogenicity, for example, based on the compound


      structure.  It's being used by the pharmaceutical


      companies.  It's distributed by small software




                We are also--obviously, we cannot screen


      industry's compounds in the agency.  That would be


      a conflict of interest.  But our software is being


      used.  We have an Interagency Agreement with


      NIH--NIH has a drug development program--we have a


      contract with NIH.  NIH sends us compounds that


      they're screening in their drug development program


      for treating addiction.  And so we are, in our own


      way, practicing what we preach, in terms of using


      our software in lead selection in drug development.


                We also--software is being used--and we


      lay a consulting role, within the Center, for


      evaluating contaminants and degradants in new drug


      products and general drugs, to determine--to


      qualify them, and determine limits.  So we feel


      that our software could have much more application


      in that realm.


                And decision support for review divisions.




      We collaborate very closely with the Center for


      Food Safety.  And, in fact, we're training their


      scientists, and have shared our software with them,


      and they're using our carcinogenicity predictive


      software to screen food contact substances. Because


      they're working under the new FDAMA rules that


      place the burden on the agency; in other words, the


      agency has to, within 120 days, decide whether


      there is a risk.  The agency has to give cause why


      a substance is a risk.  It's a reverse of sort of


      what drugs are.


                So in order to meet those kinds of


      deadlines, they had to go to predictive modeling to


      ascertain whether there's a potential risk of a


      food contact substance--within 120 days.


                EPA is looking at our--and we work with


      them.  And the software also can be used in


      deciding whether we have a data set that is


      adequate; whether there are research gaps that need


      to be filled.




                So we talk about the FDA information.  We




      get submissions, we review them.  There's an


      approval process, and then the post-approval


      process.  We extract information from this process.


      We extract proprietary toxicology data,


      non-proprietary toxicology and clinical data.  And


      we build proprietary and non-proprietary databases,


      so we can keep information that can be shared with


      the public through Freedom of Information and


      information that will not be shared--or cannot be


      shared legally--into two different databases.


                And we use these databases for a variety


      of functions:  for guidance development, for


      modeling.  And also for decision support fo the


      review; and also it feeds back on industry, because


      much of this information can be shared with the


      public, because it's under the Freedom of


      Information Act.




                We have leveraging initiatives in both


      realms.  We leverage to get support from outside to


      help us develop databases, so that we don't rely


      entirely just on FDA funding.




                And the objectives are to creat specific


      databases--endpoint specific.  They could be mouse


      studies, three month, 90-day studies, one year


      studies; the toxicology databases that people are


      interested in.


                These database initiatives are funded and


      supported through CRADAs and other mechanisms.  We


      have a CRADA with MDL Information Systems, which is


      a part of Reed Elsevier publishing company.  They


      are interesting in building a large information


      system, and so they're helping, supporting, our


      effort.  We have CRADAs in the works with Leadscope


      that has a wonderful platform for searching


      toxicology data.  And also we have a CRADA in


      process with LHASA Limited, in England--University


      of Leeds in England--that has a system also--an


      interest in these kinds of databases.


                What we--our databases are


      constructed--the center of our database is the


      chemical structure.  It is a chemical-structure


      based database.  And the structure is in digital


      form so that it can be teased--it's a




      chemoinformatic database.  And the digital form is


      called the .mol-file structure, and it's a common


      structure used in industry for over a decade.  So


      the chemical structure, as well as the name is the


      center search point.


                And then once you have a structure that's


      in digital form, you can not only ask a simple


      question about, "Can I find substance x," but you


      can also query and ask whether--"I'd like to know


      everything--all the compounds that are like it."


      And that's such a powerful tool--regulatory


      tool--that I think is another--puts us in another




                It's not that I want--"Tell me about


      acetaminophen," but I want to know compounds that


      are 90 percent like acetaminophen in a data set.


      And we're able to do that now--really easily--with


      the system.


                So once we have this system, then we tie


      in--the databases are linked to this search engine.


      We have our clinical databases that we


      model--post-marketing adverse event reporting




      system, and also the tox databases.  And we use all


      this--what we're really interested in is modeling;


      computational predictive toxicology.


                And the sources of that data on these


      databases come from reviews.  We extract


      information from the regulatory reviews and from


      other databases.




                So, now, getting into our modeling


      operation, we transform the data.  We supply the


      chemical structure data, and our collaborators and


      software companies supply the software.  And we


      work with them on an iterative basis to improve and


      make these things work, and develop software for


      these endpoints.


                We've also, I think, are probably the


      first group that have developed a way of using


      chemical structure to predict dose.  And so we have


      a paradigm for predicting what the maximum daily


      dose of a compound might be in humans, within a


      statistical, obviously, error bar, in humans.


                So, currently, in our prediction




      department, you might say, we have access to five


      or six different platforms.  And they represent


      very different algorithms.  And this is the


      point we want to have interactions with software


      companies that have approaches that are different


      from one another.  And then we evaluate and work


      with them to try to develop models, using our data




                So we have two CRADAs on board right now,


      with multi-case and MD/QSAR, and we have others in


      the works.  And we also have interactions with


      other prediction approaches from the statistical.




                In terms of the models that we're working


      on now, the objective is to model every single test


      that's required for drug approval.  And so we


      started with carcinogenicity, because that was the


      most--the highest profile, in terms of preclinical


      requirement; and teratology would be next.  These


      are endpoints that cannot be simulated in clinical


      trials; mutagenicity, gene tox--all these are


      models, either have been created or are in the




      process of being created and being worked on.


                We're also attempting to model human


      data--the adverse event reporting system;


      post-marketing human data.  This is an enormously


      difficult data set; very dirty data set, but it's


      enormous, in terms of its size.




                And we have had some success, preliminary


      modeling, of hepatic effects, cardiac effects,


      renal and bladder, and immunological effects in


      humans.  These are still works in progress, but we


      have made progress.


                And in terms of the dose related


      endpoints, we have made really good progress.  We


      were surprised, ourselves, because we didn't really


      think this would work.  We've been able to


      successfully model the human Maximum Recommended


      Daily Dose--you know, that's the dose on the bottle


      when you get your drug.  It says "Don't take more


      than 10 milligrams a day for an adult.  Well, we


      modeled that, because that comes from clinical


      trial data.  That is really human data.  And it




      represents an enormous scale--I don't want to get


      into it--but it's like an eight-block scale of


      doses, and we have 1,300 pharmaceuticals that are


      either--that we've modeled, in our database.  And


      we were able to successfully model this--and I'll


      get back to that in a moment.




                The other question that came up was


      proprietary data and sharing industry data.  It


      would be nice to get their data, especially in


      areas that we know the industry has a great deal of


      experience in, like gene tox data.  Right now we


      can't have access to data that was not in


      submissions.  And so we need a way of doing this.


      Chemoinformatics gives you a way of at least


      getting there partially.  We're able to share the


      results by not disclosing the structure and name of


      a compound.  You can disclose the results, but you


      say "What good is disclosing results, or using the


      results, without knowing where they came from?"


      Well, you can use descriptors--chemical


      descriptors--that can be used in modeling, but




      cannot be used to unambiguously reconstruct the


      molecular structure.  But they contain enough


      information to model.


                And so you're sort of at least halfway


      there.  You can share some information that can be


      used in modeling.  And so this is a feasible


      approach and, in fact, it's already being


      accomplished--legally.  It's gone through our


      legal--our staff at the agency and it's


      incorporated in some of these softwares.




                And this is an example.  This is 74 MDL


      QSAR descriptors for the compound methylthiouracil.


      Now, these descriptors are used in modeling, and


      ocntain a great deal of scientific information, in


      terms of modeling.  But all of these descriptors


      will not unambiguously recreate the structure of


      methylthiouracil, because there's a lot missing.


      It's like a pixel pictures.  You know, you have a


      photograph--a digital photograph--if you've only


      got 70 pixels, you'll get a rough picture of what


      it is, but you won't know it's your uncle.  It's




      just a person--you know.  But if you had 10,000


      pixels, you'd know exactly who it is.  It's the


      same idea.  So you can share this crude image.




                Getting back to modeling the human maximum


      daily dose--at present, we have to go through many


      steps to arrive at a starting, Phase I clinical


      starting dose, in a drug that's never been into man


      for the first time.  We start with animal


      studies--multiple dose studies in multiple species.


      So already that's a lot of cost.  Then you estimate


      the no-effect level--has to be estimated from this.


      Then you have to decide which species is closed to


      man by looking at the ADME and, you know,


      metabolism and everything.  And then you have to


      convert that to a human equivalent dose using


      allometric scaling.  And then, on top of that, you


      use a little--the uncertainty factors, dealing for


      inter-species extrapolations--finally come up with


      a dose that you might try for your first dose in


      human--in clinical trials.


                Well, if you could model, on the basis of




      structure, the maximum recommended daily dose, you


      get a predicted dose in humans--because that's


      human data.  You take one-tenth, or one-hundredth


      of that, just to be on the safe side, and you have


      a dose.


                And what's the benefit?  There's no


      testing in animals.  There's no lab studies.


      There's no inter-species extrapolation, because


      you're using human data.  And we think it's more


      accurate, because animal studies don't predict


      whether a drug is going to cause nausea, dizziness,


      cognitive dysfunction.  Animals can't tell you


      that.  But yet that appears in labeling for old


      drugs all the time.


                So we feel that this is a good approach.


      Everyone acknowledges that the estimation of the


      first dose in clinical trials is a bad--but it's


      the only thing we know how to do.  So this has got


      to be better, because it's better than nothing.


      You know, because right now what we're doing is a


      very crude approximation.






                What's another application?  And--in


      conclusion--the two-year rodent carcinogenicity


      study--in mouse and rat.  It costs $2 million. It


      takes at least three years to do.  And there's


      always controversy about the outcomes of these.


      Yet it has an enormous effect on the drug's




                Is it necessary to do these studies for


      all drugs now?  Can computational methods replace


      some of them?  I'm not saying we're getting rid of


      all testing.  But if we know a lot about a


      particular compound, based on the experience of the


      past, perhaps with predictive modeling there may be


      a subset of compounds in which we don't have to


      test as vigorously.  And those which we know very


      little about--and the computer can tell you that;


      that the compound is not covered in the learning


      set, and therefore you better do all the studies.


                But if a compound is another--you know,


      antihistamine, maybe there's a lesser path because


      a structure that's so well represented in the data


      set, that it's sort of silly to keep testing it




      over again, just to meet a regulatory requirement.


                So we're hoping that this would reduce


      unnecessary testing and put the resources where


      they're needed; testing things that we really don't


      know anything about, and that are new--that are


      really new compounds.




                So the challenges for accepting predictive


      modeling:  we need accurate, validated--and that's


      always--you know, what we mean by "validation" is


      always arguable.  But we need to develop that.


      That's part of our mission.


                Standardization of software; experience


      and training--it's not something that's going to go


      on a reviewer's desktop ever, because it requires


      interpretation.  It's a really special skill.


                We need more databases; adequate sharing


      of proprietary information; the bigger the


      database, the better.  But we need, also,


      regulatory mangers and scientists that are willing


      to consider new ideas--consider; don't have to


      adopt--consider.  That makes a big--you know, opens




      the door for innovation.


                And then the ned for an objective


      appraisal of current methods.  It's the emperor's


      clothes.  How good, really, is what we're doing


      now?  And that is something that's painful, but


      it's something that needs to be done.  Compared to


      what?  Is it better, worse--compared to what?




                In PhRMA 2005 meeting that occurred


      several years ago--and I think it was very


      farsighted--Price Waterhouse Coopers had a


      paradigm.  And they said, "Right now you have


      primary sciences:  the lab-based, patients--you


      know, clinical trials; and the secondary is the


      computational--what the call "e-R&D"--that there


      will be a transition where they'll reverse from


      primary to secondary.  And the primary science


      maybe in the next generation, will be the modeling


      and predictive science, and the lab and clinical


      will be the confirmatory science.


                So, with that, I'll end my talk.  We've


      published much of what we've done.  A lot of it is




      in press right now.  We have a web site:  our


      maximum recommended daily dose database is on our


      website, and a lot of people are working with it,


      and we're happy to say that they're getting the


      same results--which was nice.


                And I'll end my talk here.


                CHAIRMAN KIBBE:  i'll take the prerogative


      of the Chair and ask the first question.  And then


      we'll get rolling.


                Your database looks wonderful when you're


      dealing with toxicity.  Have you also done a


      similar thing with clinical effectiveness, or


      utility, of compounds?  Some way of looking at the


      structure, and then looking at the effect, and


      being able to predict how effective one structure


      is relative to another?


                And then follow up with that--if that's


      true, can we plug into the opposite end of your


      program and go back the other way, and just bypass


      drug discovery?




                DR. CONTRERA:  [Laughs.] Well--no fair. 




      I'll start with the last one--but you'll be only


      discovering what we already know.  There may be--


                CHAIRMAN KIBBE:  But I was thinking of


      plugging in different parameters--


                DR. CONTRERA:  Yeah.


                CHAIRMAN KIBBE:   --in the toxicity and


      outcome:  lower toxicity, higher efficacy--


                DR. CONTRERA:  Oh, yes.  Yes.


                CHAIRMAN KIBBE:   --and then go backwards.


                DR. CONTRERA:  Yes, that's possible.


                CHAIRMAN KIBBE:  Thank you.


                DR. CONTRERA:  But getting back to


      efficacy--yes.  In fact--I mean, industry is using


      it as an efficacy tool all the time.  That wasn't


      our mission.  But potentially--certainly


      applicable.  And sometimes we stumble on those


      things.  But that isn't our mission.


                And you know where research--we've got


      four people in this unit.  And then we have


      contractors.  And then we get students.  So we're a


      small, tight unit.  And you have to be very


      focused, in terms of your priorities, and doing




      what is feasible first, and less--and so we didn't


      get into efficacy.  No.


                CHAIRMAN KIBBE:  Who have I got down here?


      I've got everybody on the right side.


                So we'll start it at the end, and work our


      way down.


                Go ahead.


                DR. SELASSIE:  Okay.  I have a couple of


      questions for you.


                First of all, with your database, you have


      in-house data that you're generating for your




                DR. CONTRERA:  Yes.


                DR. SELASSIE:  Do you ever go to the


      literature and get information from it?


                DR. CONTRERA:  Yes.  Actually, that could


      be a much more complicated slide.  But we mine


      everything.  We mine other databases; the NIH


      databases; literature.  And, in fact, we're


      using--we're using our CRADA with MDL--because MDL


      owns almost every journal in the world


      now--practically.  Elsevier owns almost everything.




      And so--and they have access to data that's




                So, using the leverage with a publishing


      company, we have a pipeline now to the literature.




                DR. SELASSIE:  Okay.  I have another




                DR. CONTRERA:  Yes.


                DR. SELASSIE:  When you're inputting the


      structures, do you all ever use the SMILES




                DR. CONTRERA:  Yes, we use SMILES.  There