UNITED STATES OF AMERICA

P-R-O-C-E-E-D-I-N-G-S

8:25 a.m.

CHAIRMAN SHINE: Good morning. I'm Ken Shine, currently chair of the Science Board Advisory Committee. And I want to welcome you to this meeting.

We will be shortly joined by Dr. Cassell who had another commitment but will be with us shortly. I'm also pleased that Professor Barbara McNeil at Harvard has agreed to join the committee. She couldn't be to this particular meeting but she will be at our next meeting.

Before we actually begin the program, it would be useful, I think, for us just to go around the room and identify ourselves and our institutions for benefits of the group. And perhaps we could start over here with --

MEMBER PI-SUNYER: Yes, my name is Xavier Pi-Sunyer. I'm Professor of Medicine at Columbia University and I'm Chief of the Division of Endocrinology and Nutrition at St. Luke's Roosevelt Hospital in New York.

MEMBER ROSES: I'm Allen Roses. I'm the Senior Vice President for Genetics Research in GlaxoSmithKline.

MEMBER HARLANDER: My name is Susan Harlander and I have a consulting company called BIOrational Consultants.

MEMBER SWANSON: I'm Katie Swanson, Vice President of Food Safety with Ecolab.

MEMBER THOMAS: I'm John Thomas, Vice President and Professor Emeritus from the University of Texas Health Science Center at San Antonio but now at Indiana University School of Medicine.

MEMBER LAURENCIN: I'm Cato Laurencin. I'm a university professor and Professor of Orthopedic Surgery and Chairman of the Department of Orthopedic Surgery at the University of Virginia.

MEMBER RIVIERE: I'm Jim Riviere. I'm a distinguished Professor of Pharmacology at North Carolina State.

DR. CRAWFORD: Les Crawford, FDA.

DR. JOHANNESSEN: I'm Jan Johannessen. I'm the Executive Secretary of the Science Board to the FDA.

DR. WOODCOCK: Janet Woodcock, Acting Deputy Commissioner for Operations, FDA.

DR. ALDERSON: Norris Alderson, Associate Commissioner for Science at FDA.

DR. BRACKETT: Bob Brackett, Director for the Center of Food Safety and Applied Nutrition, FDA.

DR. SLIKKER: Bill Slikker, Deputy Director for Research, National Center for Toxicological Research, FDA.

DR. GALSON: Steve Galson, Acting Director, Center for Drug Evaluation and Research, FDA.

DR. CARBONE: Kathryn Carbone, Associate Director for Research, Center for Biologics Evaluation and Research, FDA.

DR. SUNDLOF: Steve Sundlof, Director of the Center for Veterinary Medicine, FDA.

MR. MARZILLI: Hi, I'm John Marzilli. I'm with the Office of Regulatory Affairs, FDA.

DR. SCHULTZ: Dan Schultz, Director CDRH, FDA.

CHAIRMAN SHINE: Good. I'm Ken Shine. I'm sorry. Did I miss somebody? Yes.

Again, I'm Ken Shine. I've Executive Vice Chancellor for Health Affairs at the University of Texas. And I sit at the fulcrum between FDA and the committee.

I want to thank the staff of the FDA with regard to the creation of the agenda for this meeting. A number of the issues that we're going to deal with are important and challenging issues and issues that the board is extremely interested in. And we look forward to a good overview of a number of these issues and some important discussions.

But I think it is appropriate to move quickly to hear from the Acting Commissioner of the FDA, Lester Crawford, who, with Janet Woodcock, will be giving us some opening remarks.

DR. CRAWFORD: Thank you very much, Dr. Shine. It's a pleasure for me to be here. And welcome you all back again. Thank you for the commitment you've all made to serving FDA in this way.

And we're grateful for it and we're also grateful for the new organization, Dr. Shine, that you have put forward on this board. And I think we're going to be able to do bigger and better things as the result.

I'd like to take this opportunity to briefly highlight some of the accomplishments over the past year at FDA. By creating innovative initiatives and improving existing processes, the FDA was able to accomplish many goals that enhanced the lives of Americans in 2004.

We introduced new initiatives to combat critical health threats such as obesity, counterfeit drugs, and medical errors.

FDA's Innovation Initiative has met some important milestones. First, a root cause analysis of multiple cycle reviews, a pilot program to test the effectiveness of earlier communication with product manufacturers, a cementing of our partnership with the National Cancer Institute, and the establishment of an Oncology Office within FDA to foster innovation in the treatment of cancer.

We also launched the Critical Path Initiative, which we have presented to the board on previous occasions. And Dr. Woodcock is going to update you on that initiative a little later in the program.

We approved a substantially greater number of products, including an unprecedented number of generic drugs.

We strengthened the security of the nation's food supply against potential bioterrorism attacks through the development of four important new regulations. We strengthened the food safety through measure initiatives and actions and we streamlined paperwork processes to reduce areas, reform outdated practices, and enhance new product innovation.

We're extremely pleased with these accomplishments and we will continue to advance these initiatives. But as you know, our successes are continually challenged by emerging health threats, changes in technology, and various global market forces.

FDA's responsibilities are growing in scope and complexity and we're responding by focusing on new and better ways to perform our core mission.

We have the following strategic areas for 2005 that I would like to mention. First, enabling technology development and innovation. Obviously we will be continuing our efforts in the area of critical path to pinpoint those areas of product development that could benefit most from innovative approaches in emerging technologies.

Second, protecting the homeland, counter-terrorism. The past year has witnessed some of the most significant enhancements to our Food Safety and Security Program in decades. Going forward, we will finalize implementation of our new food security laboratories, intensified inspections, and closer interagency collaboration.

Next I'd like to talk about improving manufacturing practices. Good manufacturing practices are not only vital to business success, they're also essential to FDA and they're essential to the public health.

The FDA's overhaul of the pharmaceutical cGMPs encourages manufacturers to modernize methods, equipment, and facilities that will help eliminate both production inefficiencies and undue risk for consumers. Our initiative also implements tougher inspection rules to make them more targeted and effective.

We're improving FDA's business practices also. We're seeking to create a stronger and more unified agency. The increasing complexity of our regulatory mission requires that we look for new ways to create efficiency, standardized processes, enhanced infrastructure, and improved planning.

Some highlights of our plan include full implementation of shared services, including consolidation information technology infrastructure, development of a common business process model.

And finally the thing that I'm most pleased with and I think most FDAs are, the continued development of the White Oak facility as we move forward with total complete occupation in the year 2010. By that time, we will have 8,000 FDA staffers in the White Oak facility and we're very much looking forward to its completion.

Under the area of patient and consumer protection, as you know the issue of drug safety has received a lot of high profile media and Congressional attention since we last met in November of 2004 which is why this session of the Science Advisory Board is scheduled at a very opportune time.

There has been a lot of public scrutiny over this issue with the recent concerns associated with SSRIs last fall and the inclusion of a black box warning on the label of Cox-2 inhibitors and the Merck withdrawal of Vioxx, the Pfizer withdrawal of Bextra, the recent recommendations of our Drug Advisory Committee on warning labels for over-the-counter NSAIDs as well as the prescription NSAIDs.

Silicon breast implants has just come up. And we will be processing over the next 90 days the recommendations of that Advisory Panel under the leadership of Dr. Schultz.

It is important that these concerns do not distort the fact that drugs are safer today than they have ever been before and that millions of Americans each day benefit from them. But in order to improve on current process, FDA has taken some bold steps to enhance the internal deliberations and decisions regarding risk and benefit analyses.

FDA also is developing new communication formats to better inform the public of the Agency's deliberation process.

Next I will discuss final risk minimization guidance. Just a few weeks ago, FDA issued three final guidances: pre-marketing risk assessment, development and use of risk minimization action plans, good pharmacovigilance practices, and pharmacoepidemiologic assessment. These guidances are evidence of FDA's commitment to transparency and risk management decision making.

We have announced. Secretary Leavitt and I along with Dr. Woodcock, the Drug Safety Oversight Board in February. The newly created Drug Safety Oversight Board will oversee the management of important drug safety issues within our Center for Drug Evaluation and Research under the leadership of Dr. Galson.

The board will comprise members of the FDA and medical experts from other HHS agencies and government departments as well as consultation with people from the outside world.

We have commissioned an Institute of Medicine study. In November of 2004, we asked the IOM to carry out a top to bottom study of drug safety. We've already transmitted two-thirds of the needed funding for this study to the IOM and they're moving forward in initiating the study.

IOM has assigned one of their most senior study directors to oversee the study and have staffed up. The proposed committee membership will be posted on the National Academy's website shortly, perhaps as early as Monday, April 18th. We look forward to their results and will act swiftly to improve standards for American consumers.

The second pillar of the Drug Safety Program is the Drug Watch webpage. FDA is proposing to set up a new Drug Watch webpage for emerging data and risk information. This 21st century electronic evolution will bring the power of information directly to consumers and increase the transparency of the Agency's decision-making process. This site will also enhance public knowledge and understanding of drug safety issues.

Now I would also like to posit a brief overview of the day's agenda, Dr. Shine, if I can.

Our agenda today is devoted to these issues of pre- and post-market drug safety, drug safety information for health professionals and consumers, and safety issues related to blood products, vaccines, and vaccine manufacturing.

You will be hearing detailed presentations on each of these topics throughout the day. We're seeking your advice on several issues today relative to our efforts to strengthen drug safety monitoring and communication.

For example, the Agency would value your input on the kinds of information needed by healthcare providers and patients about the safety of a drug and how FDA can best convey this information. More and more we believe patients and consumers and general citizens look to FDA for information, perhaps more than in the past.

We also seek your input on how the reporting of adverse events could be improved. Are there additional sources of useful data that might help with post-market surveillance? If so, how can FDA partner with the appropriate organizations to gain access to such data?

Perhaps most importantly given FDA's resource situation, we seek your input as to the appropriate areas of focus of these existing resources to most effectively assure drug safety. We look forward to your suggestions for improvement on this critical aspect of FDA's public health mandate.

I want to reaffirm that your independent views are important to FDA and to thank you for your commitment to the Agency by participating as a member of this board. I hope that your presentations today will give you a good understanding of FDA's commitment to this important issue of drug and biologic safety.

Now it is my pleasure as always to introduce my distinguished colleague Dr. Janet Woodcock.

DR. WOODCOCK: Good morning.

In the past, FDA and I have brought a number of initiatives before the Science Board. The first one that I was directly involved in was the GMP Initiative which has come to a satisfactory conclusion and is ongoing.

Today I'd like to update you on another initiative that we had gotten the input of the Science Board on and that is our Critical Path Project. We, about a year ago, published the white paper on the critical path.

And in that white paper, we promised to publish a report on the various challenges and opportunities, a project list almost of what had been identified as a work that needed to be done under the critical path. We are working diligently on that report and expect to publish it soon.

At the same time, we're identifying projects that FDA within its current resources can either carry out or lead in association with partners. And I would like to tell you that numerous partners have emerged and although I can't go into detail about all those offers that we've received and so forth, we're in the process of setting up how to sort through and determine the priorities and which projects we actually can work on with various partners.

And so I'd like to just talk about some of the areas of great need that have been identified and some of the things that we have done in the brief time that we have this morning.

As you know, we talked about one of the needs is to improve product characterization and manufacturing scale-up issues. And that was particularly identified by Dr. Carbone in the area of biologic, cellular, gene therapy, and so forth. We are continuing to explore ways in which to get some of that work done and are hopeful that we can move forward in some areas.

In the area of animal testing, the identification of the need for development of new animal models. Particularly the Device Center had some very innovative ideas about using in silico animal models and computer modeling to eliminate some testing and keep up with the rapid evolution of devices. And we are seeking to find out how we can move forward with those sorts of projects.

Other areas that we are still exploring but it looks like we will have partners willing to work with us include toxicogenomics and various data mining projects.

In the area of other broad projects, biomarkers, of course, were identified as an area that really needs to be developed and there has been considerable discussion about this. We've had numerous workshops and conferences run by the various centers or by outside organizations to talk about this.

And what we have identified is that the general conceptual development of how you qualify a biomarker for regulatory use needs to be further evolved. There is a considerable amount of concern, if you recall those of you who were here for Dr. Califf's presentation, based on the experience in cardiology and the Cass trial in particular, there is concern about the use of biomarkers as surrogate endpoints.

And that has, in my mind, put a chill over the entire field of biomarker development because it raises almost an insuperable barrier for the use of those, those types of biomarkers in regulatory decision making. However, we have been able to tease this out and I'm going to talk about that a little bit more. In addition, specific biomarkers need to be qualified further.

We have kicked off the discussion of the conceptual framework for biomarker development at an advisory committee that we held. And there is a general agreement that we should call this something else other than validation.

Perhaps we can call it qualifying biomarkers for various uses or evaluating biomarkers for various regulatory uses. But the term validation as always raises in people's minds some distinct barriers to accomplishment.

So we are putting together an initiative in regulatory policy to try to figure out for the various regulatory uses how you would describe and work up biomarkers. This is going to require public process and development of various documents and so forth.

I think FDA may have to lead this because it involves regulatory decision-making. But we hope to involve an academic partner or many partners. And this may require discarding some of the current ideas or framework about how -- the nature and use of these types of markers in development. And we hope it will encourage people to use them more freely.

I will say that there is a tremendous amount of enthusiasm now in the various sectors as the ideas have permeated. And there is a lot of interest in supporting these kinds of projects and moving them forward including, you know, financial support. So we think we can get this type of work done one way or another.

Now we also though need to do some worked examples. We need to pick some biomarkers and move them forward for regulatory decision-making because those concrete examples, I think, will be more persuasive than just a conceptual framework.

We're working on two efforts right now where we have identified good biomarkers that we think are very close to having substantial regulatory use. And one is in the area of imaging and one is the area of in vitro diagnostics of various types. And obviously the three medical product centers are going to have to work very closely together on this because these involve devices, drugs, and often biologicals together.

We have identified partners to work with us on these and I can't go into any more detail right now but we think what we will do, this would actually involve, at the end of the day, doing clinical studies to generate more information on the use of these biomarkers. And I can tell you there is a tremendous amount of enthusiasm in the relevant communities for getting this work done.

Now the general process we think we need to go through is we need to identify this type of opportunity for biomarker. We need to analyze the current available data on the marker. And usually what we will find is there are gaps in the data and how it correlates with clinical outcomes or how predictive it is that prevent it from being used in various regulatory ways.

We're going to identify those gaps in the marker performance -- in the understanding of marker performance and then devise trials that would actually answer those questions. And then have those trials conducted or add the biomarker on to existing trials so that additional information can be generated.

On the flip side of this, in the safety side which is actually the, of course, topic of much of the meeting today, there are many specific opportunities. And I love Dr. Roses' term. There are many forensic opportunities to use biomarkers, especially genomic biomarkers.

We are very interested in conducting research in clinical trials where the agent is known to be toxic to a specific organ. This is an opportunity to use new technologies such as genomics and proteomics and so forth to see if we can either predict who is at risk for such toxicities or pick up the evolution of the toxicity early.

We also would like to go back and we're very interested in partners to go back and identify and test people who have experienced severe adverse events and see if genomic or other patterns can be identified that would identify people at risk for these adverse events. We think, obviously, this won't completely abrogate drug safety problems but this could provide a great improvement.

Now in specific biomarkers, we are working, as I said, with partners. We had a meeting on Monday and Tuesday on pharmacogenomics drug development and in vitro diagnostic pharmacogenomic test development. We had more than 500 people there. And I would like to thank the Science Board.

We brought this project to the Science Board, I think, about 18 months ago and we've had over -- I think we've had 12 voluntary genomic data submissions where we have in, you know, regulatory context, had extensive discussion of the genomic tests and how they might theoretically be used in the development program.

This has also spilled over into the review side where these tests are then being integrated into the actual drug development program and in review. And I think particularly in the cancer field, we can see already this is an urgent need right now in the cancer field to be able to link the diagnostic tests along with a drug. And the Device Center and the Drug Center are working very closely together on this.

So we have many more partners who have biomarkers than we have resources that will be able to collaborate with them. So this is kind of an embarrassment of riches.

But in the safety area, we are going to focus, we're going to look at padatoxicity. We have many partners working with that, including NCTR is working with us on that. And we're also working on further evolution of the QTC prolongation issue with drugs.

Now in the clinical trial area, we're doing some worked examples of quantitative disease modeling and simulation. And this is a tremendous opportunity, which maybe we'll have an opportunity to share with the Science Board later about a way to quantitate what we know about a disease and what we know about how a disease responds to therapy in a way that then the effect of new interventions can be modeled.

And this can be used in the analysis of how you would conduct an additional trial. This applies across the board to vaccines, cell therapy, devices, and so on. And the Device Center is particularly interested in this, as I said, because of the rapid evolution of devices.

In early clinical trials, this week we have just published the exploratory IND draft guidance. This draft guidance is out there for comment and it is a document that talks about getting into people, not in the ordinary way of rapid dose, you know, escalation trials and so forth, or IND safety trials, but earlier using imaging, using micro dose techniques to look at metabolism, look at proof of mechanism for targeted therapy and so forth.

So we will be seeking from the drug development community and the academic community, in particular, and put on this guidance.

We're also doing a pilot on early meetings with sponsors that will allow us to talk about some of the quantitative modeling and also early strategies in development.

We had already talked, when we presented to the Science Board last time about the projects that we wish to take on in later clinical trials dealing with many of the analytic issues such as Bayesian designs and so forth. And we are working on those. We're moving ahead and trying to set up those projects.

In later clinical trials, there's also a regulatory modernization piece that we found has to fit with any given modernization of the science. So we move the science along, we have to move our mode of regulation along with it. And so we have started what we call an analysis into how we do bioresearch monitoring. We're moving along on that.

And we also recently had a Part 15 hearing, which is a way of perhaps initiating regulatory changes on adverse event reporting to IRBs. This is something the IRB community has brought up to us repeatedly as a problem. We had a very good meeting on that. And all of these efforts are intended to go along with trying to streamline the process.

In addition, and I'll hurry, we're working on -- we're continuing to work on the automation and standardization of clinical trials. We have gone a long way with our partners CDISC and working with HL7 on electronic data standards to be used in clinical trials.

We are starting to work more on the content now. Most of those standards have been worked on electronic interchange standards. Now we're starting to work on content standards. And this will be a Critical Path Initiative that we will work on.

And we're also partnering with the National Cancer Institute on electronic reporting of -- a repository for CVs of investigators, something that kind of plagues everyone who is involved in the clinical trial business, getting all this information about all the investigators around to all the parties who need it. And so the idea of an electronic repository is something everyone agrees with basically.

The final point I want to make is that training and infrastructure for this new kind of science or this enhanced kind of science continues to be a problem. We hear from the various industries that they cannot hire and find the multidisciplinary scientists that are needed for this kind of work. And that's true for the FDA as well. These folks are in very short supply.

These issues overlap with things that were identified by the NIH in their Roadmap Initiative. And so I think there are opportunities here to make sure that programs are set up and people get trained.

We have raised these needs with the academic policy sector and hopefully, you know, there will be a recognition of the need for more training. If we're actually going to move biomedical science ideas into the clinic and get them to be commercial products available to patients, we have to train people who are expert at that interface.

And the final note I would like to add before I close is one of the things we've all realized and we've learned throughout this year is the importance of diagnostics to this whole enterprise. We realize that better diagnosis is really the foundation of improved therapy. And perhaps we were all mistaken if we thought we could just determine better therapeutics without further advancing the science of diagnostics.

We're focusing on this issue with the Device Center and with our partners. And we hope this is something that can really be moved along.

Now the remaining barriers on this initiative, obviously we lack staffing at the FDA to staff many of these projects. And so that really slows down the rate of progress that we would have on them.

Various partners are still trying to work out how we could work together in public/private partnerships. And what the true, you know, arrangements would be. There are ways to do this but they take time to make consortia.

And we still, we're doing some analysis but we really need better incentive for diagnostic development and we really aren't sure how that is going to get done.

So the next steps are we are going to publish our critical path opportunities list, hopefully very soon. Subsequently, we plan to publish a description of the projects that we're actually able to be engaged upon at the FDA in the various centers.

We hope to further develop these consortia. In fact we will further develop these consortia, various types of consortia to get this work done. And the good news is there are people who are really willing to put resources into this to make this happen. This general recognition, this type of thing needs to be done. And we will try to gather some resources up internally to continue to support these efforts.

So the critical path concept has had a very positive response. We think publication of our list will galvanize people to realize that there are huge gaps that need to be addressed in their areas of interest.

We've identified many projects. And within our ability to do them have begun work on them. And right now the rate limiting steps are organizing these collaborations and obtaining the resources to get them done.

CHAIRMAN SHINE: Thank you very much, Dr. Woodcock. Excellent. I call to the committee's attention the large volume of guidance materials that we were provided in preparation for this meeting.

And one of them is the piece on pharmacogenetics, which this board discussed extensively in April of last year. And, Dr. Woodcock, I was very pleased to see the guidance that arose in terms of that piece. We'll be talking about other portions of that guidance as we go forward.

Are there comments or questions for Dr. Woodcock at this time?

(No response.)

CHAIRMAN SHINE: The Commissioner has a conference call and he will be back. And we do have some questions for him but we'll give him a chance to get some of his other business done.

Before we proceed with formal presentations, Dr. Johannessen has a statement to make.

DR. JOHANNESSEN: Good morning. The following announcement addresses conflict of interest issues associated with this meeting of the FDA Science Board on April 15th, 2005.

The Food and Drug Administration has prepared general matters waivers for the following special government employees: Dr. Kenneth Shine, Dr. Gail Cassell, Dr. Susan Harlander, Dr. Cato Laurencin, Dr. Xavier Pi-Sunyer, Dr. Jim Riviere, Dr. Allen Roses, Dr. Katherine Swanson, and Dr. John Thomas who are attending today's meeting at the FDA Science Board for the discussion of drug safety, safety systems for vaccines, blood, and tissue, and cGMPs for vaccine being held by the Office of the Commissioner.

A copy of the waiver statements may be obtained by submitting a written request to our Freedom of Information Office. Unlike issues before a committee in which a particular product is discussed, issues of broader applicability such as the topic of today's meeting involve many industrial sponsors and academic institutions.

The committee members have been screened for their financial interests as they may apply to the general topic at hand. Because general topics impact so many institutions, it is not practical to recite all potential conflicts of interest as they apply to each member.

The FDA acknowledges that there may be potential conflicts of interest but because of the general nature of the discussion before the committee, these potential conflicts are mitigated.

And I would just remind everyone to please turn their cell phones on vibrate and for the Members at the table to push the button when you talk. Thank you.

CHAIRMAN SHINE: Thank you very much, Dr. Johannessen.

I think we will now move directly to the first presentation, which is an introduction to drug safety by Steve Galson, Acting Director of the Center for Drug Evaluation and Research.

Dr. Galson?

DR. GALSON: Thank you. Okay. Let me see if I can make this thing work. Good.

I'm very happy to be here and to see all of you today. And we have a number of very, very interesting presentations about drug safety. I'm going to give a brief introduction just to set the stage. Don't have a lot of time in just ten minutes for much detail.

But I think you don't have to be working with the drug industry to be aware that the drug safety policy environment has become more contentious. This has been all over the news, as you know. Among other things, there's been a lot of fuss about the risks of the SSRI class of antidepressants, the withdrawals of the Cox-2 analgesics that Dr. Crawford already talked about a little bit.

There have been a lot of allegations thrown around, among them that the FDA doesn't care about drug safety, something that hits all of us very hard. There are numerous, in fact a record number of Congressional oversight investigations and hearings that we're trying to slog through.

Most recently, last month, Drs. Woodcock and Kweder engaged in a real testimonial tour de force in front of the Senate talking about our drug safety efforts and the history there.

And what's been going on in the press that all of you are aware of has been very, very interesting. It started out with a lot of very accusatory, very sharp articles and it has evolved somewhat. And those of you who are students of the press probably recognize this sort of pendulum effect.

Just last week, there was this article that came out in the Congressional Quarterly, which is a kind of low circulation, high impact periodical circulated on Capitol Hill. The thrust of this is that a lot of the effort that has gone on to nail down what the problem is that FDA ignores the fact the drugs benefit patients and, in fact, ignores the fact that Congress has been very involved in the last decade at trying to help FDA get more resources to speed up approval of new and innovative products.

And that when this is all over and the dust settles, Congress is realizing that they really set the stage for how the FDA is working today. And they're going to accept responsibility for that and probably not make that much change although, you know, I'm not making an editorial comment on that. Just noting that it has been very interesting to see how the changes have occurred.

And then just this week, perfect timing for you all, there was an editorial, an op ed in the Washington Post by Ann Applebaum, who has been closely following drug safety and who many of us have talked to to try to educate her about the realities at FDA.

And the interesting thing about this is it really describes the pendulum effect, how back and forth throughout the history of the last few decades at FDA, drug safety has come up and the pendulum is swinging now towards more concern about drug safety and expressing the real concern which is out there.

We know from patient groups and from the industry as well that if the FDA becomes more strict and more careful about safety, there's going to be a real impact for patients. So this side of the story is coming out as well.

I think all of you know, and this addresses one of the major misconceptions that's been out there in the press about drug safety that it is, in fact, a top priority across the regulatory programs at the FDA. It's not just an issue that we look at in post-marketing surveillance.

It crosses all of our work from our new drug program to the manufacturing, regulation of drug quality, to our generic drug program, to the regulation of clinical trials, promotional activities, and all the other things we do. Many of the resource decisions, the time that people spend on these problems are because of concerns about drug safety in various different realms.

So I wanted to very, very quickly talk about where this all comes from and it really is derived from the Food, Drug, and Cosmetic Act which very vaguely but at the same time very clearly says that we're allowed to require all tests reasonably applicable to safety.

And the way that we implement that is through our regulations, as you all know and among the millions of pages of those regulations that we have to assure that the product is safe for recommended use. And that's the way that we decide what data gets sent to us from companies.

And then after the regulations, we've issued a series of FDA and then worked very closely with the international community, as you know, on international guidance documents that specifically lay out in many, many different realms the level of evidence needed for safety evaluation.

But one of the issues that I think is very, very relevant for you all on the Science Board, we talk about it a lot, but is sometimes lost in the public or lost to the public is that even though we are scientists and we have hundreds of scientists in the FDA, many of the judgments having to do with risk benefit and regulation involve more than science. They involve judgment and they involve policy.

And so when people say, you know, the science is broken or we're not following science, what that really avoids is the fact that a lot of the decisions that FDA makes have to consider non-scientific societal factors.

And those of you that have been involved in regulation or involved in risk-benefit decision-making, which I know is a lot of the people in the room, already know this. That you can't just go back to your biochemistry book and figure out how to answer questions of risk-benefit. There is invariably judgment, invariably policy in there.

So, of course, this means there is a huge amount of room for disagreement. The most basic level of disagreement is how safe is safe? How safe does a product have to be demonstrated to be before we decide that it is safe enough to be approved?

I don't have to tell this audience that safety doesn't mean no risk. The risk-benefit for an individual drug or class has to be done separately because of different severity of clinical conditions, different situations, different availability of data for drugs as they go through the pipeline.

And so there is no formula that can be plugged in to say bingo, green light, safe enough, effective enough. There are always going to be individual decision-making and this is why there is always room for disagreement.

And then the fundamental double-edged sword, which Dr. Woodcock focused on very, very well at her testimony from the Health Committee, which is that if we are going to require more information be submitted to the Agency on safety, if we want to be more sure about safety, there are inevitable consequences.

It would drive up the cost of drug development. Or drive up post-market costs for companies and for the healthcare community as a whole. And this has -- can have adverse consequences for innovation and availability of new products.

So this zero sum game, double-edged sword concept is very, very important for you all and everyone to understand as we move forward. And it is very much behind the efforts that are underway in the Agency.

A lot of the questions that have been raised have been about whether the User Fee Program has changed how we balance benefit and risk in making drug approval decisions. And I don't have time to go into the detail of this but we have done a very careful examination of the rate of drug withdrawals and how frequently we use black boxes in warning about drugs.

And there is really no evidence of actual change in our risk threshold over time because of the User Fee Program. Of course, there are lots of people that claim this and they throw data around. But at least the careful evaluations that we have done, we haven't really been able to demonstrate that.

And this is just a list -- I hope you can -- yes, you can read it there -- of the drugs that have withdrawn -- withdrawing drugs from the market is something that has happened since the beginning of FDA's regulations. And as I mentioned, we've looked at, in what we think is the fairest way possible, the proportion of drugs that have been withdrawn as opposed to marketed before and after the User Fee Program was instituted. And there really isn't any difference in the proportion. And you can see throughout time it's been happening since way before the Drug User Fee Program.

But have our data requirements changed with science? Yes, they have. We are now able to avoid some well-known safety problems that plagued earlier drugs. We're far from perfect. We're going to talk about that a little bit later as is Dr. Woodcock again.

We have more data submitted to support applications than ever before and we do feel that the drug supply, because of this, is safer than it ever has been. But there will continue to be, as I've already pointed out, many issues.

And I'll just point out a few of these before moving on. The first is that the bulk of the studies that are conducted really have as their main focus efficacy. Science has moved ahead more with how to determine efficacy than it has with safety analysis. And many of you involved in drug development know very well that, for example, with toxicology, we're doing almost exactly the same toxicology tests we were doing probably when I was born for drugs. They haven't changed very much.

The size of the pre-market safety database is usually determined by the size for what's needed for the efficacy studies. Rare or time-dependent effects are not known at approval. They're just -- with the number of people participating in the studies for the period of time that we require the drugs to be studied, we are not going to be able to see those rare or time-dependent effects.

And we can't require additional safety studies after approval according to our regs. Our only tool, which is somewhat like hitting an ant with a mallet, is that we can take a drug off the market. And we can use that authority to get companies to do things. We do it. But we do have limits on our regulations.

As I mentioned already, major additions to preapproval safety database would have potentially adverse consequences. And then in the post-market area we don't, even as much as some other countries in the developing world, have an organized national pharmacovigilance system that automatically gets us information about drugs that run into trouble.

We have a system of looking at spontaneous adverse events reports but that's very, very different from an organized national sort of certainly required pharmacovigilance system.

So what we're going to do this morning is take you through a quick run through of our program. Of course not enough detail but you guys already know a lot about our programs. And then give you some time to look at the questions that you've heard about.

Very, very quickly, Dr. Oliva is going to go through Pre-Market Drug Safety. Rachel Behrman is going to talk about some of our very, very exciting initiatives on labeling and communicating with the public about drug safety which we think is one of the key issues for us.

I'm going to come back and talk a little bit more about the drug safety initiatives that you heard about from Dr. Crawford this morning. Dr. Seligman is going to talk about what we do in the post-market arena.

Theresa Mullin, Dr. Mullin, from our Planning Office is going to talk about some very careful work that she and her staff have done to really document the resources that we spend on drug safety. And then Janet will be back again to talk about critical path applications and applications of new science to drug safety.

So with that, I'll let Dr. Oliva kick it off. Thanks very much.

DR. OLIVA: Well, it's both a pleasure and an honor to address the committee this morning on the issue of pre-marketing drug safety. As a former clinical reviewer, it's no exaggeration for me to say that the FDA spends a tremendous amount of time and effort on the evaluation of pre-marketing drug safety.

And what I would like to do this morning is inform you on the contents of a new guidance that came out just last month on the review of pre-marketing drug safety.

This is a guidance that represents the good review practices for the clinical reviewer on evaluating drug safety. And it really represents the collective knowledge through many years of doing this in the Center. And it was a collaborative effort across many offices within CDER, including Office of New Drugs, Office of Drug Safety, and Office of Medical Policy. And I'm sure there are others that didn't get on this slide.

Now the guidance says very specific purposes and first of all, it's meant to assist reviewers in how to conduct the clinical safety review of an NDA or BLA. It does document good review practices, as I mentioned.

And we think it's important in order to provide standardization and consistency across all our review divisions on what constitute an adequate evaluation of drug safety as well as provides instruction on the format actual review -- of the written review of drug safety. It ensures that critical analyses and presentations of the data are included in the review.

This guidance also dovetails very nicely with the clinical review template that we issued last July. The two are completely harmonized so that a reviewer, once they're ready to document their review, has all of the headings of the drug safety guidance are included.

And the guidance, what it does is it expands the headings of the clinical review template so it serves as a very nice reference to reviewers on each heading, which I'll go into.

One of the important points that the guidance stresses is the need for collaboration. Although this is a document geared towards the clinical reviewer, we do recognize the need for additional skill sets such as analysis of event rates and so forth. And, therefore, it stresses the need for collaboration with biostatistical colleagues and other experts when necessary to get a complete picture of safety.

So what exactly is in the guidance? Well, there's advice on how to conduct and organize the review, of course. And there's also an annotated outline of what the safety review should look like.

The guidance identifies four important tasks for the reviewer when they evaluate a drug's pre-marketing safety risks. First and foremost is to identify serious adverse events that might do these three things: might prevent use altogether, in which case we would not approve the application, that could limit use, and that may require special risk-management efforts.

Secondly, the safety reviewer will estimate the frequency of commonly occurring adverse events, which will need to be communicated to the public.

Third, the reviewer evaluates the adequacy of the data in the application. For example, was exposures adequate and did they occur at relevant doses?

And finally, it's important that the reviewer identify any unresolved safety concerns that may need further attention either pre-marketing or post-marketing.

These are additional tasks for the safety review. We need to identify factors that predict the occurrence of AEs, both intrinsic and as extrinsic factors, ways to avoid adverse events through monitoring or other techniques, and ways to manage them when they occur. And finally we need to provide a comprehensive evaluation of risk information in support of labeling. And Dr. Behrman will get into that in more detail.

The guidance outlines the key sources of safety information. Now I'm not going to go through all this long list, but I put it up there to emphasize that there are a lot of areas that we ask reviewers to look at when looking at safety in an application.

We talk about causality determination which is heavily dependent on a comparison of event rates between treatment groups. And it's also heavily dependent on proper coding of the adverse events. And we ask reviewers to look very carefully at how adverse events are coded because this is the only way that we can get accurate numerators for event rate comparisons.

And there's also an emphasis on the individual case reviews of deaths, serious adverse events, and drop-outs due to adverse events or adverse drop-outs.

As far as the organization of the safety review, it's basically broken down into four broad areas that I show here. And I'll walk you through to each one. But this is what constitutes the documentation of the pre-marketing safety review.

The first one is the methods and findings. This is where the reviewer assesses the safety of the drug, describes what were the findings. And they describe the relevant data sources that went into this determination, what safety assessments were conducted along with the major safety findings. And it emphasizes the use of a very systematic approach.

This section of the safety review, 7.1, is quite detailed. And again, I won't go through each one of these. But I put it up again to emphasize the fact that reviewers systematically go through each of these domains with each application and documents the findings in these areas.

Section 7.2 is where the reviewer will assess the adequacy of the safety database. Was patient exposure adequate? Were sufficient numbers of subjects exposed for long enough duration at the appropriate levels? Were specific subgroups analyzed for safety? And what were those findings?

We look at the quality and completeness of the safety evaluation, including what animal testing of safety occurred, what in vitro tests showed, what long-term safety can provide, as well as any specific assessments that were performed. And we ask ourselves are additional safety testing needed either pre- or post-approval.

The third section of the safety review is the summary section where we provide a brief summary of the critical findings of the safety review. And this section also contains the adverse events that the reviewer considers are important and drug related that may need to be communicated post-approval. And here is where we document any important limitations of the safety database along with our conclusions.

The final section of the safety review is a general methodology section. It's sort of an appendix. This is where we get to describe the analytical methods that the reviewer used in reviewing the data along with a general discussion of any other methodological issues that were not discussed elsewhere, discuss the general idea of pooling, why was it done, how was it done, as well as how we explored the data for predictive factors and how we assessed and determined causality.

So in summary, I want to leave you with the notion that this final guidance is really a culmination of many, many years of collective effort throughout the Agency in how to conduct a pre-marketing safety review. It promotes good review practices. We believe it will lead to increased standardization and consistency of format and content of the pre-marketing safety review.

And it ensures the critical presentations of the safety data and the analyses are not omitted. And it's entirely harmonized with the Clinical Review Templates so that our reports that are posted on the web all have the same look and feel and are consistent and complete.

So with that, I'll end my talk and turn it over to our next speaker, Dr. Behrman.

CHAIRMAN SHINE: Yes, before we go further, we have a couple of minutes. And I would like an opportunity -- if any members, this is a very important document. If any members of the committee want to ask specific questions --

DR. OLIVA: Sure --

CHAIRMAN SHINE: -- about the document. One thing you might help me with is a number of these documents are headed, "Contains non-binding recommendations." And I wonder if you would -- or Janet or whatever -- help me with that. Given that this seems to me to be a rather important set of criteria by which our reviewers operate, why do we use this kind of language that kind of implies that they may or may not do this?

DR. OLIVA: Sure. Since this is a guidance and we abide by the good guidance practices by regulation, this is standard language. And that unless it is something that is mandated by a statue or regulation, then it's by definition non-binding.

And Dr. Woodcock, you may be able to elaborate further.

DR. WOODCOCK: Yes, that's true. That provides the flexibility, obviously. But we have -- the purpose of the supervisors here in the supervisory chain, if a reviewer decides that a different type of analysis is better, that's fine. And they should do that.

But then their supervisor has the right to say well, that analysis really wasn't -- doesn't get to the question that you were supposed to get to. So that is part of the purpose of having various experts look at these reviews and so forth.

But yes, we don't require in this guidance or any guidance that people do things a certain way. If there is a satisfactory scientific way or superior way to get to the same result of the analysis in this case, then it can be done that way.

CHAIRMAN SHINE: But doesn't that increase the ambiguity for people who are developing drugs and so forth to understand, in fact, how that particular product is going to be evaluated since there would be apparently this --

DR. OLIVA: Well, one of the principals in good guidance practices is that if someone is going to deviate from a guidance, that it has to be well justified and it has to be justified in writing.

So the assumption is that reviewers will do it this way unless there is a very good reason that has been fully vetted internally to do it a different way or a better way because we realize that guidances are living documents and that they -- and through knowledge and experience can change. And this allows the opportunity to improve on it.

CHAIRMAN SHINE: At what point in the process will the folks who are involved with the new product be notified that you're going to use a different methodology?

DR. GALSON: As -- I would say as early as possible, right?

DR. OLIVA: Sure.

DR. GALSON: I mean what we're trying to encourage is more intense early interaction between our reviewers and people from the companies, very early in the development. And as soon as we would detect that there was going to be something different that we were interested in seeing, we would say that. You know it's not a perfect system. But that's what we try to do.

CHAIRMAN SHINE: Dr. Pi-Sunyer?

MEMBER PI-SUNYER: Yes, I wanted to ask you a question with regard to time-dependent adverse events. You made the point that you cannot require additional safety studies after the approval of a drug so you don't have a way to do that after you've approved it.

But if it is possible that you're really getting safety data mostly two years, at the most, and say something in a drug makes you think that there might be a time-dependent adverse event, can you add some kind of qualifier that will allow you to continue to survey that drug?

DR. OLIVA: Well, let me elaborate a little bit because I, you know, somewhat used shorthand. We don't have the statutory ability to require something after approval. But we can ask and we do ask and we get things done that way all the time.

When there is an important safety question that has to be answered, we try to get it answered either by asking the company to do a study or working with other groups. And so I don't want to convey that we have no tools at our disposal. But it's not a you will do this and if you don't, X will happen kind of authority.

DR. WOODCOCK: Can I add to that?

DR. OLIVA: Yes.

DR. WOODCOCK: Can I add to that?

CHAIRMAN SHINE: Yes, absolutely.

DR. WOODCOCK: Yes, in addition, at the time of approval, if -- you know one of the important parts of the safety review is to identify the unanswered questions. Not just to review whatever has been sent in but to think about what don't we know about this drug? That's as important as what we do know about a drug.

And at the time of approval if there is a very burning question that hasn't been answered but might require much additional work and we feel the benefit-risk is satisfactory for the drug, we can ask that either registry or study be done as a post-marketing commitment.

DR. OLIVA: Another useful tool that we have to get that information post-marketing is the label itself. For example, if we have insufficient information to assess the risk of the drug in a special population such as hepatic or renal impairment, prior to approval we may contraindicate the drug in labeling in that population until that data are collected and submitted and reviewed.

CHAIRMAN SHINE: A key element, just to follow up, is the question that the reviewer is asked, namely is additional safety testing needed either pre-approval or post-marketing.

DR. OLIVA: Yes.

CHAIRMAN SHINE: My question is from a procedural point of view, who looks at that recommendation? What's the process by which we are sure that a recommendation of that kind gets evaluated at a fairly high level in terms of the follow up?

DR. OLIVA: Yes, once the safety review is written, then it is reviewed internally at several layers. There's usually a secondary reviewer as well as the division director and for a new molecular entity, the office director will evaluate those recommendations and look at the reasons for those recommendations and then act upon them and make a decision whether the uncertainty is great enough, whether additional work pre-marketing is needed.

DR. WOODCOCK: And whatever that decision is, it will be documented in writing.

CHAIRMAN SHINE: Yes, I just -- my concern is that we are very sure that those kinds of recommendations are evaluated at a high enough level in the organization so that we're quite comfortable that other kinds of interactions are put in the proper perspective in terms of doing that.

One of the interesting things, as I read the material, was you point out the problem with the definition of symptoms, complications, and so forth. I think one of the examples you use are the SSRI business, emotional disturbances versus suicidal ideation.

To what extent could you, in fact, have a glossary of terms explicit enough so that investigators would, in fact, have a minimum opportunity to categorize items in ways that don't truly display the nature of the problem?

DR. OLIVA: Well, this is an ongoing challenge. I mean there are numerous dictionaries and thesauruses available that will allow the mapping of investigator terms to standardized terms. And those are readily available to the reviewers and they're routinely used in the evaluation of the coding of the adverse events.

And if we see a problem, if we see that certain terms are being mis-categorized, then that would be a cause to have them re-categorized, if necessary either by an independent panel or --

CHAIRMAN SHINE: But the key is that the investigators are in the same thesaurus as you are, right? In other words -- yes?

DR. WOODCOCK: This gets to the issue of standardizing the case report form and the reporting of adverse events out of clinical trials. And as you probably well know, that hasn't -- but as I said in my introductory remarks, we are starting down that path.

The use, for example, of electronic case reporting forms with pick lists and so forth will help with this. But I think what Armando is reflecting is often we get, you know, like a company gets a narrative and an investigator has sort of decided what terms they're going to use for that particular adverse event or whatever.

And we need to -- you're right. We need to work on further standardization. But we're not quite --

CHAIRMAN SHINE: Yes.

DR. WOODCOCK: -- the field isn't there yet. Is that fair?

DR. OLIVA: Yes, no, exactly. I mean in an ideal world, the investigator when they were ready to enter the adverse event onto the case report form, they would have the pick list, the standardized terminology readily available to them so they could make that call.

CHAIRMAN SHINE: Exactly. And I think that's important.

Dr. Cassell?

MEMBER CASSELL: Yes, I think Ken, following up on what you said, to me in terms of pre-approval, the collection of adverse events and the investigators is one thing because you have selected investigators very knowledgeable, hopefully, in the area and so forth. But post-marketing when you're collecting adverse events, who is responsible for training of individuals, in communicating to individuals on the appropriate methods, terminology, et cetera, for collecting adverse event data?

And I guess, Janet, I was thinking back to your comment about the need for training and where in the medical school curriculum as tight as it has gotten today, we all know in terms of pharmacology period, it has shrunk considerably as pathology and molecular biology have expanded.

What are we thinking looking towards the future now as far as collection of adverse event data and the practicing physician? Who will be responsible for the education?

And again, just to add one more piece to that that I worry a lot about with all the changes and the way companies now are able to interact with the practicing physicians in educating in terms of drug use, again it just poses more limitations, I think, on getting to those individuals that are going to be collecting the data.

DR. OLIVA: Well, this is my own personal opinion, but I think standardization is the key. I think that we're now moving to the area where we have standard variables for different types of adverse events. And now, as Dr. Woodcock pointed out, standard terms for the actual adverse events themselves.

And the next step is encouraging collection of the adverse events in clinical trials using these terms. The next logical step will be the use of the very same terms in clinical practices and use in the electronic health record so that it is just one continuum from the doctor's office to -- every step along the way.

CHAIRMAN SHINE: Amen.

Quick comments from Dr. Thomas.

MEMBER THOMAS: I was just curious how and what process you used for pooling data. Whether it's at the pre- or post-surveillance stage?

DR. OLIVA: Well, my comments about pooling data were strictly relating to the pooling of safety data across clinical trials in pre-marketing -- during the pre-marketing review. That's the easiest place to do that type of work simply because there is a certain amount of consistency already built in in the way they are collected. And so they lend themselves readily for pooling.

In the post-marketing world, I'll have to let Dr. Seligman weigh in on that and others. But it's obviously much more challenging.

MEMBER THOMAS: There's probably a great adherence at the safety level --

DR. OLIVA: Yes.

MEMBER THOMAS: -- from the standpoint of protocol.

DR. OLIVA: Exactly.

MEMBER THOMAS: Than later.

DR. OLIVA: Yes.

CHAIRMAN SHINE: Dr. Laurencin?

MEMBER LAURENCIN: The guidance document that you put together appears comprehensive. But every page says non-binding recommendations. And so the two questions are number one, how much will it be utilized by the reviewers? And what are the plans to monitor to what the extent the reviewers actually utilize this? Because it is important for industry and for others who are looking to bring products through.

DR. OLIVA: Sure.

MEMBER LAURENCIN: That's the first question. What are your thoughts?

DR. OLIVA: Well, you know why it says that. Because it is a guidance. And we have to say that if it's a guidance.

But the nice thing about it, because it is a guidance to reviewers and it is an internal guidance, that we have other policy tools internally to make sure that reviewers adhere to it. You know, the supervisors and team leaders and division directors all can expect, and as part of internal policy, can say this is the way that it will be done.

And part of the secondary and teriary reviews will include a check -- does include a check to make sure that the reviews adhere to this document.

MEMBER LAURENCIN: Just as a follow up, what is the mechanism by which the changes in the guidance document will be relayed to all? Because so many times it is a draft guidance document. The phrase living document comes up and everyone smiles. But how will these changes be relayed to the greater community?

DR. OLIVA: Well, guidances are updated. A guidance can be modified and then that goes through the -- again, posted for public comment before it gets rewritten. So there is a mechanism in place to update guidance. And this one would go through that.

CHAIRMAN SHINE: We're going to have to move on. I think -- at least my sense is that the committee -- the board is quite impressed with the document. It really does show a good comprehensive kind of guidance with regard to it.

Our hope is that it will be applied fairly consistently. Dr. Woodcock has indicated there may be rare examples where a different methodology is involved which would require explicit kind of language.

But I think this is a good step forward. And I congratulate you on the accomplishment.

Why don't we move forward?

DR. OLIVA: Thank you.

CHAIRMAN SHINE: We're going to hear from Dr. Behrman.

DR. BEHRMAN: Good morning. Let me just make sure -- okay, thanks.

Good morning. I have actually the distinct pleasure this morning, unlike some others who have to discuss things that are in their infancy, to describe for you two initiatives that are nearing the end of development and will be rolled out shortly. And that's the Physician Labeling Rule and structured product labeling.

And just since we've had a discussion of standards, I'd just like to point out that Dr. Oliva and I, as members of the SPL team, have sacrificed our personal preferences and we present from a standard format. Not required but we do it.

(Laughter.)

DR. BEHRMAN: In terms of today's topic, which is primarily focused on safety, one needs to think about information. And FDA -- one can think about this in this way, FDA regulates information in two primary ways. We regulate the development of information through clinical trials and that is not the topic today.

And then we also regulate, to a certain extent, the dissemination of information. And over the years, we've probably focused on that -- or we have focused on that substantially less. And gradually have come to realize that this is clearly part of our mission and responsibility.

We have statutory authority to do this. The Federal Food and Drug Cosmetic Act clearly states that we regulate, we have authority over the content of labeling the way we speak.

So a drug or device shall be deemed to be misbranded if its labeling is false or misleading in any particular. And the Secretary cannot approve an application, can refuse to approve an application if the labeling is not adequate.

Similarly, we have implementing regulations, specific requirements on the content and format of labeling for human prescription drugs, 21 CFR 201.57. Those regulations were promulgated -- draft in `75 and the final in `79. Clinical trials, drug information looked very different in the mid-70s than it does now and clearly it's time for those regulations to be updated.

So the first topic that I'd like to cover for you is the Physician Labeling Rule, which I said is nearing completion.

Now anyone who has thumbed through a PDR recently may be surprised to learn that in fact there is a very specific order to the information. That's not obvious because there is no same look and feel and there's no application of the basic principals of communication in terms of white space and bolding and other means of emphasis. But there is a very specific order.

Unfortunately, the order is not one we would today say makes tremendous sense. Generally the principal is you put the most important things either first or last. Well, first is the description. That tends to be the chemical structure. Last is references. That, again, tends not to be the first thing one might turn to.

The safety information, as you can see, is sandwiched in the middle as is dosage and administration, for example. So the information that one wants most quickly is difficult to find.

So going back a number of years, the intent was to develop a tool that is more informative, that's more accessible, that's a better risk communication management tool.

And the development process was quite lengthy and quite comprehensive. It included focus groups. It included a national physician survey, physicians being the primary user of drug labels although the public meetings that followed included pharmacists, nurses, et cetera, and the public, others who use the drug label.

And finally, initiated a formal notice and comment rulemaking process with the review of, I believe, 93 written comments.

So what will -- in the proposed rule, what did we propose that the drug will look like -- and because the rule has not been finalized, I'm going to stumble a little bit and keep referring to the proposed rule. But it does give you the flavor of what we're thinking and certainly what was reflected in the comments.

Well, the first is to start with the most important information. And what is the absolutely most important information? A box warning if there is, what the drug is used for, and then the dosage and administration.

You'll notice that the safety information in terms of contraindications, warnings, precautions, adverse reactions, is consolidated. It's worth noting also that in the current label, there is a warning section and there is a precaution section. The distinction is not really clear to anyone.

And precautions, which sounds like it might be safety information, includes many other things including use in special populations such as patients with renal impairment, et cetera. So it's a mish-mosh or a potpourri. And that's been separated out with new sections having their own -- let's see -- for example, use in specific populations would be separate, clinical pharmacology has been pulled out, et cetera.

And then at the end, again putting something important last, patient counseling information. There's clear information that prescribers should be communicating to the user of the drug and we want that highlighted so it is easy to pull out, easy to access for both the prescriber and for the patient.

But perhaps the most novel and -- from the focus groups and from our point of view important notion would be of highlights, that a very quick summary of the most critical drug information up front where it can be easily accessed. And the limitation statement is -- we'll just be describing the fact that it is not the entire label.

But notice what one could see very quickly. That -- well, first of all, what product it is. But obviously if there have been any major changes since the last label. Or not since the last label but any major changes. Indications, dosage, contraindications, everything very important right up front. And I'll show you an example of that in a moment.

In addition, we feel it is important to establish minimum font requirements. The one -- my favorite comment to the proposed rule came from an eye care company that said that under the proposed eight point minimum font, they would no longer be able to publish their label in a 4.5 font. This is an eye care company.

(Laughter.)

DR. BEHRMAN: So we believe that no one should be asked to read such a label. And, in addition, that standardized bolding and white space, all of us -- I mean there are, I'm sure, many in the room who try to use labels to prescribe drugs, the information is there. It's very hard to scan it and find it. With standardized bolding and white space, that would be changed.

And as I mentioned, frequently referenced sections would be moved forward. And, in particular, safety information should be consolidated.

We heard pretty clearly that highlights not only being critical had to be limited to about a half page. And that we should bullet the information, again, making it easy to scan.

Moreover, that we should create some form of index. This will be particularly important in the electronic era so that you can hyperlink. But even if not, to get a quick, easy view of, in fact, what is the label, although what was proposed in the rule and what was discussed in the comments, in fact, is not an index. It's a content but that's another point.

And that we provide greater clarity in our requirements. Again, if you go back to `75 and `79 and read these proposed rules, they were appropriate to the state of information in the mid-70s but not in 2005.

So this is, we think, a thing of beauty. But this is an example of a highlight --

(Laughter.)

DR. BEHRMAN: No, quite seriously.

(Laughter.)

DR. BEHRMAN: I mean can you imagine, those of you that have been in emergency rooms in the middle of the night being able to just scan down a drug with which you are not familiar or for an indication that you don't know or a dose that you don't know or just to make sure that nothing has changed since you used the drug a year ago or what the most important warning is because you're using it in a population you've never used it before.

We think that in terms of importance of clear communication, safe prescribing, reducing medical errors, et cetera, that this will make a major contribution.

Similarly, the index I described to you -- and this would give you a feeling of, again, not only what the sections look like and how the more important information is reordered forward, but also give you a sense of electronically how you would link, how you would just jump to the section that you wanted if you were, in fact, not in paper, which you probably won't be pretty soon.

We're going to roll out, when this does roll out, a number of things with it. One thing is companion guidance. I know we've just had a bit of discussion about what guidance means but this is going to be a very new label for us, a very new label for industry. We're going to have to help people learn how to write them. It's not easy. Those of us that spent some time doing it, it takes some getting used to.

More detail, more clarity on what should be in the adverse reaction section and how it should be organized. Clinical studies, again, an important place for information about how to use the drug safely. And then warnings and precautions, contraindications, and boxed warnings, a section which we have never provided guidance and for which there is a lot of confusion about what the contents should be.

And then switching gears slightly, clearly to move forward into the 21st century, into the electronic era, the way I've been thinking about it lately is think about everyone who is walking around with an iPod and what they can tell you about their music but they can't tell you that about their drug information.

We want to support and are supporting the health information technology initiatives because the Physician Labeling Rule will be concise. It will be clear. And it will be easy to hyperlink.

So the other major initiative that I want to describe in that vein is Structured Product Labeling. And what is Structured Product Labeling? Probably most people don't know. But this is something that is going to roll out this fall so it's here.

It's the content of labeling in standardized electronic file format. So the blocks of text are tagged. There we go. So in pdf, you just have a lump of text. But in SPL, you have a tagged section. And, therefore, that means you can link it, you can move it. If you want to standardize it, you can do that. It's manipulatable.

This is crucial to support a very important and exciting initiative called DailyMed which will be an electronic repository of current labeling. It will be managed by NLM but it will be populated by FDA. So it will be the most current labeling, vetted, approved, the gold standard of drug information.

And I just wanted to add also it's responsive to the Medicare Modernization Act, which requires uniform standards.

So what else do we need to do in order to be able to spit this labeling out on a daily basis to NLM? Well, we need something called ELIPS, Electronic Labeling Information Processing System. And the primarily important thing to you is the last bullet which is that we will be transmitting SPL with one business day lag.

So this is really in real time. We're giving ourselves one business day so that we -- the oops phenomenon, the not sending something out to the public that we later regret, but we'll be able to receive SPL. We'll be able to reject it if it is missing something. We will be able to review it, manipulate it, and then we will be able to shoot it out.

And this is just portrayed graphically there for those of you that like that. So the applicant will send us their SPL. They will be required to do this starting in the fall. Right now they're just required to send us an electronic copy but it will have to be structured starting in the fall to our document room.

We will do what we need to do in terms of review and manipulation. When it is ready to go out, it will be released by a labeling authority. There will be a single labeling authority for each product. And then it will go to the NLM. And it will be accessible to whomever needs current labeling.

And anyone who has tried to look for current labeling on the weekends, for example, not at FDA, it's extremely difficult to find.

This is the most exciting part of this slide. We will meet our October `05 target date. We had our first demonstration. What's the right word? We had the first -- saw the first prototype this week. And we are on schedule.

And we will roll it out -- it will take a year to populate. Companies have a year to get their current labeling in to us. So by October of `06 -- and we may cut off at the end of the year just for convenience, the population -- the SPL ELIPS will be populated with current labeling for every product.

In one sense this is a pilot in that we're starting only with prescription drugs. And it's a pilot-approved concept. But in the other sense, it is, in fact, it is the beginning because we will have one -- at least one center up. And then we will expand to the other medical products.

Phase Two will include unapproved prescription and OTC drugs. That will -- the target date is August `06. It will require a number of other things to happen, ELIPS to be up and running, which will be on time, Substance Registration System, which will require that someone assign a unique code to each ingredient so that then they can be linked and tracked. And then we will need the Drug Listing System in place in August of `06.

So in sum, we are poised to provide and can provide up-to-date labeling information. We have, over the years, recognized that to support many of the things we do, some of the initiatives in providing information to consumers and to practitioners that Dr. Galson will discuss, initiatives such as providing the information with advertisements.

Right now, the "brief summary", which you often see, is just -- well, it's not readable and it's not intelligible. A year and a half ago, we promulgated draft guidance suggesting the companies might want to think about even before the Physician Labeling Rule was finalized, mocking up brief summaries that looked like highlights because ultimately highlights will exist. It will be vetted information.

We will have already decided it's the most crucial information. And why not make sure that's what consumers see with a drug ad as opposed to something that can't read?

So we're increasingly trying to educate ourselves in how to communicate. We sit on the gold standard of information. And we need to get a little bit better about, in fact, sharing it. So the Physician Labeling Rule will ensure that the actual label, be it printed or electronic, will be readable and will be accessible.

And that electronically, this will be available to anyone through NLM from any computer. And it will be accurate and up to date. And that will support, obviously, electronic prescribing, electronic health record, and decision supportware, really bring FDA in concert with others, into the 21st century.

CHAIRMAN SHINE: Thank you very much. We have time for a question or two.

Dr. Pi-Sunyer?

MEMBER PI-SUNYER: Will this be accessible to Palm Pilots? You know the average physician --

DR. BEHRMAN: Right.

MEMBER PI-SUNYER: -- now doesn't use a big computer.

DR. BEHRMAN: Exactly.

MEMBER PI-SUNYER: He uses a Palm Pilot.

DR. BEHRMAN: This will be -- we will take it -- we will send it to NLM. It will be available and accessible to anyone. So whomever makes either your Palm Pilot or your Palm Pilot software can access that. And then develop decision supportware or whatever search engine, if you will, that is appropriate for you.

For example, if you're an infectious disease practitioner, you may have very specific sets of queries that you want. And if you're a clinical pharmacologist, it may be very different.

CHAIRMAN SHINE: Dr. Thomas?

MEMBER THOMAS: I was going to ask you if the SPL would in any way replicate what we now know as the package insert with respect to information.

DR. BEHRMAN: Well, let me -- in case I didn't make it clear, let me walk through that. Package insert, that's the printed piece of paper that comes with your bottle. It will ultimately, if you will, replace that perhaps likely. The package insert is required -- every drug is required to have the package insert go with it. That's very expensive and it is not very accessible.

So we believe that once the information -- it's the exact same information just delivered in a different way -- is available either electronically or some other way, the package insert will no longer be that useful. But it will be the same. The content will not change.

So if you read the package insert or you read it off the computer, it will be the same information. Just how you choose to access it or how you able to access it.

MEMBER THOMAS: A follow-up question with regard to safety information, I didn't see quite as much safety information on your SPL. Is that going to be --

DR. BEHRMAN: Should be go back?

MEMBER THOMAS: -- accommodated by a hyperlink?

DR. BEHRMAN: Can we go back?

MEMBER THOMAS: No, maybe I just missed it.

DR. BEHRMAN: No, it's all there.

MEMBER THOMAS: It's all there?

DR. BEHRMAN: In fact, it's -- the really exciting part, I think, is that it is consolidated. And I probably didn't --

MEMBER THOMAS: Maybe that's why I missed it.

DR. BEHRMAN: Yes. Well, we heard that it should be together. So thinking about -- looking down, your box warning will likely to be short. Indications and usage, dosage administration. Then you get into what we consider the core safety information. Contraindications, warnings, precautions, adverse reactions, all together right up front.

So -- and if you go to the -- now remember this is just the highlights. So this is not going to be the whole thing. But you'll get your box warning. Any major changes which typically are safety related. And then you have all the safety information together. Contraindications, warnings, precautions, adverse reactions.

CHAIRMAN SHINE: Dr. Laurencin?

MEMBER LAURENCIN: Thank you.

I still have some concerns about the information being provided.

DR. BEHRMAN: Yes?

MEMBER LAURENCIN: And just as a preface, we've expressed this to the FDA. I'm Speaker of the House of the National Medical Association in one of my roles. And the NMA has expressed this that this information provides no information in terms of effects of drugs in the minority communities, specifically African-Americans --

DR. BEHRMAN: Well --

MEMBER LAURENCIN: -- in terms of one, degree of testing in the African-American and other minority populations, number two, adverse effects of drugs in minority populations. It doesn't present the information. Even trends, et cetera.

And the concern is with so many drugs that may be utilized by minority populations, to a great extent there is no information really describing, at least what I see here, describing possible effects in minority populations especially.

DR. BEHRMAN: Well, you are correct that this does not address clinical trials. So this will only address how to present the information that is developed in the clinical trial.

Although the guidance documents do talk about what the importance, in fact stress the importance of special specific populations, but in fact it is highlight. It will be easier to find information about specific populations.

MEMBER LAURENCIN: I mean my feeling is that --

DR. BEHRMAN: Yes?

MEMBER LAURENCIN: -- use in minority populations is so important that if it is not there, it should be specified that it is not there, it's not present, or it has not been studied.

DR. BEHRMAN: Right. So in other words, you would want to highlight the absence of information as well. And I'm just going back to the guidance documents that we're developing to -- I'm not getting there very quickly --

CHAIRMAN SHINE: But it would be appropriate in that section a use in special populations for that description to take place.

DR. BEHRMAN: Well --

MEMBER LAURENCIN: Agreed.

DR. BEHRMAN: -- but I wonder if you're saying should -- it was definitely not proposed to require the absence of information. It's to require a description -- and again, I'm talking about the proposal -- would require a description of the information and discussion of what is missing.

But I think that when you -- you have to think about the final rule plus the companion guidance documents. And as we have discussed publicly a number of times, there is much greater emphasis on describing different populations and how drugs behave and the safety profile.

CHAIRMAN SHINE: Well, this would seem to me to be an appropriate thing to follow up on in terms of Dr. Laurencin's concerns.

DR. BEHRMAN: Okay.

CHAIRMAN SHINE: Perhaps you can discuss that with him a little further.

DR. BEHRMAN: Sure.

CHAIRMAN SHINE: One final question I have is are you -- you've gotten lots of guidance and input into this. Are you doing marketing studies with the final product?

DR. BEHRMAN: Once the rule finalizes? Or for --

CHAIRMAN SHINE: I mean are you --

DR. BEHRMAN: -- for the labeling rule for SPL or for --

CHAIRMAN SHINE: For the labeling rule, have you got focus groups where other --

DR. BEHRMAN: There have been tremendous -- the research that supported this, which was --

CHAIRMAN SHINE: I'm not talking about the research. I'm talking about the final document being exposed to a variety of audiences in terms of whether it meets their needs.

DR. BEHRMAN: Our hope -- there is no plan to test the final Physician Labeling Rule in focus groups. Our hope is that as we develop different types of consumer information and practitioner information, that will clearly have to be tested, for example, the new kinds of information that we are putting up on our website because that is something that has not yet been subject to focus groups and other kinds of testing.

CHAIRMAN SHINE: Okay. All I was saying is that you go through an iterative process.

DR. BEHRMAN: Right.

CHAIRMAN SHINE: The question is is the final product carefully tested by consumers with a variety of backgrounds, age levels, things of this sort --

DR. BEHRMAN: Right.

CHAIRMAN SHINE: -- so we know that it works.

DR. WOODCOCK: I would just like to say it would be very nice to do this but under -- the rulemaking process takes about five years, okay, and so we try to do a lot more of this in guidance, which is more amenable to such an iterative process.

DR. BEHRMAN: And I just did want to add that we made use, in addition to our focus group testing and other testing, the work that CFSAN had done. They had done tremendous work. And also that the OTC regulations.

CHAIRMAN SHINE: I agree, Janet. There are lots of ways to do rulemaking.

Thank you very much, Dr. Behrman.

Dr. Galson is coming back and talking to us about this drug safety initiative.

DR. GALSON: Okay. I'm happy to be back here again. And what I wanted to do now is take you through some of the recent drug safety initiatives that have been announced.

The first is Dr. Crawford's announcement back last year in November and then our Secretary's announcement in February and a little bit more detail on those.

First of all, the announcement from November, and Dr. Crawford went through some of this quickly, is first the Institute of Medicine study, and I want to just emphasize that again because we think it is so very important. There are a lot of groups out there making comments about our drug safety system and investigations, GAO, Congressional committees, et cetera.

But there really is no peer to the Institute of Medicine, as you all know, in terms of getting neutral experts involved in a very deliberative fashion. And although the studies do take some time and require some investment of funds, we think this is going to be very, very important in the long run for helping us design a permanent system to improve, if improvements are needed in how we evaluate safety and what FDA's place is in the drug safety system.

The second component of this announcement was a program for adjudicating differences of professional opinion, very, very important within the Center for Drugs. As you all know, there are differences of opinion. We've talked about this before a little bit.

And two people looking at the same information can come to different conclusions. And historically, there have been flare ups every once in a while in the Center for Drugs when there are these differences and reviewers don't think that their opinion or their analyses are being adequately considered in a final regulatory decision.

So what we've done with this new policy, which was very, very carefully vetted internally, is create an organized way for a reviewer who has a difference of opinion to get that opinion examined and reviewed and moved up the hierarchy all the way to the Center Director so that if they really have a professional disagreement, it can be considered by others who can peer review it and look at the opposing views and help make a decision.

And it is very, very organized and lays out in a step-wise way what those reviewers can do. And we think this will help the sense among some reviewers that they have trouble getting their voice heard. And I'm looking forward to getting involved in the first one of those adjudications.

The third point here was that we've had a vacancy in the leadership of our Office of Drug Safety for some time with Acting Directors in place.

And we've sort of taken a huge step in involving the federal government's Office of Personnel Management in helping us design a really state-of-the-art search nationally, which involves getting a group of experts in to figure out what the ideal components would be for a director, attributes for a director. And that's moving along. And hopefully we'll be able to get somebody in place there within the next few months.

The fourth component of this was the acknowledgment that we need to do more in terms of getting input from outside experts in specific drug safety and risk management decisions. And so we'll be planning more outside meetings similar to advisory committees but there are other sort of fora that we can use such as Part 15 hearings that you've heard about.

And then the last part of this, the risk management guidances that Dr. Crawford talked about. We've already published so I won't spend any time on that.

And then a couple months ago, Secretary Leavitt, along with others, announced our Drug Safety Plan, which is very comprehensive and has as its overall goal the promotion of more of a culture of openness and enhanced oversight for drug safety within the FDA. And involves several different areas of change.

The first is more outside expert consultations feeding on to what I already mentioned.

The second is improving the management of drug safety within the Center for Drug Evaluation and Research.

The third is to change how we communicate about emerging drug safety risks to improve transparency. And this is very, very important for patients, for practitioners, for those in academia and other places.

And the fourth is to continue our work to improve scientific methods of improving adverse events.

So the initiative will give patients, healthcare professional, and other consumers better access to information about medicine. And this is all tied in as well to what you heard about from Dr. Behrman.

And second, to make FDA's drug review approval and monitoring programs as transparent as we can make them under our statutes.

We've got some new information outlets that we're putting together. And these are going to be implemented immediately as opposed to some of the rules that take a long time and the systems that Dr. Behrman was talking about which, although will be implemented fairly quickly, are not going to be implemented immediately. So these are things that we can roll out quite quickly.

The first is a proposed Drug Watch Program. I'll get into that a little bit more.

And the second are patient information and healthcare professional information sheets. These sheets are already being produced by the Center.

They're already on the internet for some high profile drugs that you've heard about in the news recently: Celebrex, other drugs that have been the subject of Congressional scrutiny in other areas. You can go on our website and see these sheets. And I have some examples to show you in a minute.

And the second big part of this is the creation of the Drug Safety Oversight Board. I'm going to give you more detail about that. The Drug Safety Oversight Board is going to provide independent oversight and advice to the Center for Drug Evaluation and Research Director on management of drug safety issues and policies. And dissemination of certain safety information through our website on an emerging basis.

So the oversight board will help select which drugs and which drug safety issues are important enough to be conveyed to the public on an emerging basis. They need to go up right away even before we can change the drug labels so that practitioners and patients are aware of this information.

And the second is when there are policy issues such as what is the threshold for putting a black box warning on a drug or other sorts of drug safety policy issues, this board will be convened and will help give important advice to the Center on it.

It will be chaired by the Deputy Center Director of the Drug Center. I don't know if Doug -- Dr. Throckmorton is here. He's the Acting Deputy Director and he'll be hopefully, without too much delay, chairing the first meeting of this board.

The Executive Director is going to be Susan Cummins, who I know is here. Susan, if you would stand up a minute so everybody can see you. She's agreed to take on these very, very important responsibilities. And we're excited to have her get in place starting next week.

The membership is going to include representatives from a number of different CDER offices, not just the people who were involved in looking at the drug. In fact the people who have made decisions about the drugs that are coming from the board won't be able to vote on decisions having to do with that particular drug.

What's really new about this, in addition to getting other offices across CDER involved, is that we're bringing in representatives from outside of CDER, membership from CBER, from CDRH, and even more significantly from outside FDA, from NIH and VA in the board. We'll also have consumer and patient representatives and we'll have consultants from our advisory board also helping the Drug Safety Oversight Board.

We do have restrictions having to do with the Federal Advisory Committee Act, which you are all aware of. We've received some criticism on this. Why aren't we having really completely outside people involved. And this would turn the board more into an advisory committee and really make it very difficult to do the work as facilely and quickly as we'd like the oversight board to be able to do.

We're going to be making our operating procedures available hopefully very, very soon so that all the questions that many people have about how we're going to operate in detail will be available to all of you.

A little bit more detail on our proposed Drug Watch web page, again, this would be a section of our web page designed somewhat differently than what we currently have up there about drugs that we think, again, that the information that is coming out, that's emerging, is important enough that we need to make the public, both patients and healthcare providers, aware of it immediately.

And I've got some examples here. Again, there are already examples up on our website but just to highlight some of the features, you can see it is designed to be very clear and crisp, to just have a very limited amount of information, specifically some alert language that really is the crux of what we want the public or healthcare practitioners to know about, whether there are any specific recommendations.

There may or may not be recommendations based on some of our dry runs with drugs. The details of data, particularly changes, and then links to more information if people want to get specific information on their links to our MedWatch and how to call in and get additional information. Again, this is the patient version of it.

Again, like the drug information initiatives that Dr. Behrman talked about, this is really an enormous step forward in making information about our products really available at people's fingertips so they can get them very, very quickly and find out key facts about emerging drug safety changes.

And this is just a proposal. This isn't what our website is definitely going to look like. But this is a proposal about how we would highlight, we could highlight specific and new drug safety information on our website.

So, again, you can go to our website now. It won't look quite like that but you can find some of this information very quickly. What we'd like to do is highlight it even more.

So in summary, we feel that we're being very responsive to the concerns that we've heard from all over about drug safety, decision-making, and communication on drug safety within the Agency and within CDER.

While comprehensive review is underway at IOM, we are going ahead and implementing what we see as very, very important changes to improve public knowledge, how we manage drug safety, and how we involve the outside world in our decision-making.

As you already know, and those of you who work for drug companies who are in the audience or on the board know, these changes are not free of controversy. They have, perhaps, unexpected adverse consequences that we may have to deal with -- we will have to deal with.

They certainly raise important policy issues in drug regulation. We're here to talk about them and we know that we will be having to deal with them over the next period of time. And we look forward to it.

So thanks a lot. I don't know, Dr. Shine, if you want me to stay up here for questions.

CHAIRMAN SHINE: Yes, please. Yes, we do have some questions.

Dr. Cassell?

MEMBER CASSELL: What criteria will you use to establish something as an emerging issue before it is posted? And then I guess along those same lines, how much will be done to try to validate the events that someone is reporting as adverse events?

DR. GALSON: Yes, both of those are very complex questions.

I think first of all, the board hasn't met yet. And when the board meets, that's going to be the first issue that it on its table is to really establish policies. In the meantime, we have, within the management structure of the Center for Drugs and in consultation with the Commissioner's Office, we have been making case by case decisions.

And just like our normal risk-benefit decisions about individual drugs, there isn't a formula. But in general, what we've -- the criteria that we've been using is when there is something that is emerging -- first of all, we think it has a very high likelihood of being real, it's not a specious kind of finding or something we think has another explanation.

So first if it is something real. And then if we think that patients and practitioners are going to use this information to perhaps change their risk-benefit analysis in using a product, that's when we have decided that this information would be available.

So that's -- it's not mathematical. But it is -- if this information can be used to perhaps change a decision, we think it goes out.

MEMBER CASSELL: And will biologics and vaccines also be taken into account by this --

DR. GALSON: Right.

MEMBER CASSELL: -- oversight body? And, I guess, thinking about that, if they were, one of the things that was striking about the membership of your committee was the absence of a representative from the Centers for Disease Control, particularly, obviously, with respect to vaccines.

DR. GALSON: Right.

MEMBER CASSELL: And I wondered if that might not be a valuable addition to the committee.

DR. GALSON: Right. As you know and you're going to hear more from Dr. Goodman, I think is going to be here this afternoon, the whole vaccine safety system which I know you are very familiar with is really very different from the one that is set up for drugs. So initially we have not sought to include vaccines in this board.

I can't say what's going to happen down the road. But we felt like we were biting off a huge chunk here and really representing a very, very significant change with drugs. So we haven't included -- the biological products that are regulated by the Center for Drugs will be included but not the other products that are in CBER.

CHAIRMAN SHINE: Dr. Thomas?

MEMBER THOMAS: Yes, I think one of the operative words in your presentation was threshold. And I'm referring to Drug Watch. And that, in part, has been elaborated on by our colleague. My question, however, is once a drug does get on the Drug Watch and suppose this is some idiosyncracy and can't be re-documented, how does one get their drug off this drug alert?

DR. GALSON: Sure. There is no question that a drug can both go on and go off the watch. As a communications tool, it's not going to be very useful if it just grows and grows and grows and grows. It won't be really -- people won't be able to highlight and prioritize the information.

So as -- the board will have to set criteria and policies for putting drugs on. We'll have to also set policies for taking drugs off.

DR. WOODCOCK: Can I make a comment, too?

CHAIRMAN SHINE: Dr. Woodcock?

DR. WOODCOCK: Yes, there will be a public document published, okay, for public comment on both of these issues that have been raised, the threshold for posting as well as the procedures and so forth for moving off.

DR. GALSON: Yes. And also public meetings, I'm sure, to accompany the comments.

DR. WOODCOCK: Right. So there will be a sense of discussion of this point.

CHAIRMAN SHINE: Dr. Galson, two questions.

One is one of the concerns has been the resources available to FDA --

DR. GALSON: Right.

CHAIRMAN SHINE: -- for working on safety.

DR. GALSON: Right.

CHAIRMAN SHINE: Will the IOM Committee be asked to look at the resource issue?

DR. GALSON: Yes. That's part of their charge.

CHAIRMAN SHINE: Thank you.

The second is you made reference to outside consultants from an advisory board --

DR. GALSON: Yes.

CHAIRMAN SHINE: -- with regard to this.

DR. GALSON: Right.

CHAIRMAN SHINE: What's that mean?

DR. GALSON: As you know, the chairs and members of our advisory committees are special government employees. And we can bring them on, as we do already in the Center for Drugs in some circumstances, to give us advice outside of their work with an advisory committee.

So if there is a specific drug safety issue that's coming from the board, we could ask either a chair or a member of the committee to come in and write a position paper or look at some data, prepare some materials that would then be presented to the board.

CHAIRMAN SHINE: I understand.

DR. GALSON: Yes.

CHAIRMAN SHINE: I understand that the organization of this board is designed to be operational, to have players and people who know about it to make decisions in a timely way. I applaud the inclusion of some patient or consumer representatives. But this is still very much an inside board.

DR. GALSON: Yes, right.

CHAIRMAN SHINE: And the question I would raise is whether in view of the importance of these issues, there ought to be some mechanism for some little more distance oversight of this process --

DR. GALSON: Yes.

CHAIRMAN SHINE: -- whether it is in the form of a separate advisory panel, whether it is in the form of a subcommittee of this board which interacts or oversees it, whether it's a liaison member from this board, or whatever.

DR. GALSON: Yes.

CHAIRMAN SHINE: What I'm concerned about is that we, as you use the word openness on many occasions, is that we have a process by which this board is scrutinized --

DR. GALSON: Right.

CHAIRMAN SHINE: -- regularly from a bit of distance --

DR. GALSON: Right.

CHAIRMAN SHINE: -- in terms of the issues. And I would just encourage you and the Commission to look very carefully at how we can structure. I don't consider the consultation from a chair of one -- that's a science --

DR. GALSON: Right.

CHAIRMAN SHINE: -- input.

DR. GALSON: Yes, yes.

CHAIRMAN SHINE: What I'm thinking about is some mechanism whereby we can assure that there is a real kind of oversight that goes beyond -- and, in fact, is designed specifically so that it doesn't involve people who have ever had anything to do --

DR. GALSON: Right.

CHAIRMAN SHINE: -- with approving of a drug --

DR. GALSON: Right, right.

CHAIRMAN SHINE: -- at any time.

DR. GALSON: Sure, sure.

CHAIRMAN SHINE: But who can represent --

DR. GALSON: Yes.

CHAIRMAN SHINE: -- the public interest.

DR. GALSON: Yes, I think that is a very, very important concept. As you know, we're just in the very early part of birthing this board. It hasn't met yet. We haven't put the guidance document on the web yet. We will hopefully shortly.

And I think that's a very important issue that we're going to have to address. And there are lots of different options. And I look forward to discussing with you --

CHAIRMAN SHINE: I would --

DR. GALSON: Yes.

CHAIRMAN SHINE: -- suggest that that is something that you folks ought to look at --

DR. GALSON: Yes.

CHAIRMAN SHINE: -- in preparation of a board because I don't know what their interests are and our interests are in terms of the public --

DR. GALSON: Right.

CHAIRMAN SHINE: -- oversight of this kind of activity.

DR. WOODCOCK: I will point out there is an existing advisory committee --

DR. GALSON: Right.

DR. WOODCOCK: -- on risk management and safety, drug safety.

DR. GALSON: Right.

DR. WOODCOCK: An established advisory committee or subcommittee, I'm not really sure.

DR. GALSON: Yes, no, it's a full advisory committee.

DR. WOODCOCK: Full advisory committee.

DR. GALSON: Yes. And, in fact, people have sometimes used too much shorthand because I'm so used to these issues. We've been asked multiple times by members of Congress and from the outside as well. Nothing that we're doing here in setting up the board is meant to diminish the role of our advisory committees.

In fact, I expect that we will have more advisory committee meetings on specific drug safety issues over the next few years as the board gets going. It's not meant to supplant the activities of our expert advisors on the outside.

CHAIRMAN SHINE: Well, there's a nice balance here --

DR. GALSON: Yes.

CHAIRMAN SHINE: -- because on the one hand, you don't want to supplant those functions. On the other hand, you want at least the oversight of this to be such that it does not include people who have had a vested interest in prior --

DR. GALSON: Right.

CHAIRMAN SHINE: -- drug approvals --

DR. GALSON: Right.

CHAIRMAN SHINE: -- which might in any way influence their decisions about what the policies are to be in terms of thresholds.

DR. GALSON: Yes.

CHAIRMAN SHINE: So I really would like to see a very close look at this even before the board organizes itself --

DR. GALSON: Yes.

CHAIRMAN SHINE: -- because you're going to see their philosophy reflected in whatever those recommendations are.

DR. GALSON: Right.

CHAIRMAN SHINE: Any other comments? Gail?

MEMBER CASSELL: With respect to the board, what do you include of your definition of consumers? In other words, what will that group be representing? What types of --

DR. GALSON: Right. We have an established a group of consumer representatives that are the same group that are members of our specific advisory committee. I'm sure you've worked with them if you've been on advisory committees.

And they represent all of the patient groups, the cancer groups, the arthritis groups. And they've all been gone through the SGE process and are vetted. So we'll choose from those -- that group.

CHAIRMAN SHINE: Any other questions?

(No response.)

DR. GALSON: Thanks.

CHAIRMAN SHINE: Very important activity, Dr. Galson. Thank you.

Let's take a ten-minute break. And we'll promptly reconvene in ten minutes.

(Whereupon, the foregoing matter went off the record at 10:23 a.m. and went back on the record at 10:38 a.m.)

CHAIRMAN SHINE: Thank you very much. We are now going to go to post-market drug safety and Paul Seligman is going to take us through his presentation. Paul, thank you.

DR. SELIGMAN: Yes, Dr. Shine, good morning members of the board. I have a slight case of laryngitis this morning so I'm not going to be my usual mellifluous self. Maybe I'll try to speak in a lower register.

But my objectives this morning are twofold. First of all to describe how we monitor drug safety -- how we monitor the safety of drugs marketed in the United States and second to talk about some of the challenges we face in implementing this Post-marketing Drug Safety Program as well as the challenges we face as a society in the ongoing assessment of the benefits and risks of drugs as information and science continuously evolves.

There are basically three pillars to pour Post-marketing Risk Assessment Program. The first is the analysis and interpretation of case reports of adverse drug events and medication errors generated by healthcare professionals or by consumers and submitted to the FDA either directly or via the drug manufacturer or drug sponsor.

The second pillar is the monitoring of how drugs are used or utilized in the marketplace. And third is a program to conduct population-based observational studies either in collaboration with outside researchers or directly by FDA staff.

What I will not be focusing on this morning in this short time is the full breadth of safety activities undertaken by the staff in the Office of Drug Safety.

These include participation in pre-marketing safety assessments of drugs, particularly for those products with notable safety concerns that may require additional efforts in the post-marketing arena, whether it comes to risk minimization or risk management or whether it requires additional post-marketing surveillance.

I believe all of you in your packages for today's meeting received the three guidance documents that we recently published on pre-marketing risk assessment, risk management as well as post-marketing pharmacovigilance. And though I won't be discussing these in my presentation, we're clearly here and happy to answer any questions about those documents.

Finally, as noted on my slide here, the Office of Drug Safety is also home of the MedWatch Program, which receives direct adverse event reports from consumers, healthcare providers.

But it's also responsible for the regular communication of safety information to over 45,000 members of our listserv as well as 130 partner organizations I think in a vital, important risk communication vehicle that we have within the Agency.

And finally as also noted in this slide, we play a major role in the prevention of medication errors by reviewing all of the proposed trade names to screen for look-alike and sound-alike names and for confusing packaging, which are a major sources of medication errors in the marketplace.

Our Adverse Event Reporting System, or our AERS System, is an Oracle database. It's a repository, an electronic repository that now contains over three million adverse event reports. As you'll see in the next slide, there's been a steady increase in the number of reports every year. In calendar year `04, we have now over 400,000 reports submitted.

This is a publicly-available database. You can access actually the last quarter of 2004's data on our website now. It was previously available, and is still actually available through the National Technical Information Service. We now provide the AERS data, redacted, of course, of personal identifier information, now directly on our website.

This slide just basically shows the trends in adverse event reporting over the last decade and a half. There are three important points, I think, here to note. First the line in the center of the graph shows the point where we moved from what we called the spontaneous reporting system to our AERS System when it became an electronic system.

And as you can see, there has been steady increase in reports but -- and I apologize. Although you can't read the little teeny box up there on the left, I think you might note at least by the colors that some of the colored bars are changing.

And one of the most important changes in the colored bars is the increasing proportion of serious adverse event case reports that we are receiving consistent with our emphasis with industry as well as in the medical community with reporting those serious reports.

In response to a question that was asked earlier by one of the panelists, all of the coding of adverse event reports is done via the MedDRA system, which is an internationally-agreed to coding language for all adverse event reports.

All the drug sponsors who code these reports and submit them to us use the MEDDRA system. We quality control them and assess this coding on a regular basis as well as those reports that are submitted to us.

One last point is that we employ via contract a staff that basically codes and enters these data 24 hours a day, six days a week. And then eight hours on Sunday. It's a huge expense for the Agency. And we are moving increasingly towards electronic reporting of these data. In fact, this year now over 50 percent of the serious reports are now sent to us electronically rather than in paper form.

The second pillar of our program is the tracking of how drugs are utilized in the market. This slide just simply points out that we use a variety of data sources, both to assess inpatient as well as outpatient use, looking not only at dispensed prescriptions, the sales, but data over time. And we use a variety of different kinds of databases that are either based on pharmacy benefit management systems or surveys in both the in and outpatient realm.

We use these data -- these drug utilization data in a variety of ways. First they permit us the opportunity to create a crude rate or a reporting rate, using the case reports that we receive in AERS as the numerator.

Second, these data in and of themselves give us a view of prescribing patterns by doctor, by patient, by diagnosis, by concomitant use medicine that provide us very important insights on how drugs are used both on and off label in the United States.

And finally for those drugs for which there are risk management plans that may call either for limited or restricted use of a product, these kinds of databases allow us to track the utilization of these products that can provide information on whether or not the prescribing practices that are recommended in the risk management guidances are being adhered to.

Finally, in the area of ongoing population-based studies, it's important to note that many of these studies are done by drug sponsors either at the behest of the FDA, as part of a Phase 4 commitment, and I'm sure you are all well familiar with that terminology, or are conducted on their own initiative, even sometimes without even the knowledge of the FDA being aware that such studies are being done.

We at the FDA have a modest research program, the Cooperative Agreement Program, that we're conducting with three groups. We have recently obtained access to the British General Practice Research Database or the GPRD.

And are working with the Center for Medicare and Medicaid Services, CMS, as they implement the Medicare drug prescription benefit to utilize their national database as it is developed and comes online for the implementation of that program.

We've also, to the degree that we can, have been opportunistic as the need arises in working with other groups such as the Veterans Administration, Kaiser Permanente of California, to conduct studies to clarify the relationship between drugs and adverse outcomes.

Now I just show this slide simply to point out that the size of the populations covered by these various databases, the Cooperative Agreement Program, covers the Harvard Pilgrim HMO Network in the Northeast. We use the Tennessee Medicaid Database and the UnitedHealth Group in the Midwest.

But also to emphasize that, you know, the way our healthcare system is currently configured in this country, we have to rely on a variety of different kinds of databases to answer questions and these are, depending on your point of view, have fairly limited populations and often make it, you know, on occasion, difficult to project the national populations.

Despite the limitations of adverse or spontaneous reporting as noted here, voluntary case reporting of adverse events still provides the bulk of information that results in label changes, warnings, safety alerts, and, on occasion, the withdrawal of a product.

Although it is difficult to rigorously test, we rely on the eyes and the intellect of physicians, pharmacists, nurses, dentists, other healthcare providers to make the connection between a drug adverse event -- or drug and an adverse event and to file a report.

But as we move into an age of bigger databases, linked databases, tools that can rapidly sift through huge volumes of data and records that are increasingly maintained on electronic media, I think we're all looking for ways in which to tap into these trends to improve both the thoroughness, efficiency, and timeliness of our ability to identify safety problems.

If you believe, and I certainly believe that the learned intermediary still needs to make the call, then we need to figure out ways to facilitate and make reporting either a more valued part of every healthcare practitioner's daily life and figure out ways that we can facilitate reporting such as, for example, using the electronic medical record and figuring out ways that one might be able to push a report button and populate certain fields of our MedWatch form and then submit it directly to the FDA.

But we also, you know, are very interested in the notion of being able to extract critical information directly from the medical record, key diagnostic, laboratory, drug use information, that may define an adverse event and then build the kind of algorithms to electronically scan databases and search for suspected cases essentially bypassing the role of the healthcare practitioner.

Clearly what we do is critically dependent on the type and the nature of the information that is available to us, both in our ability to conduct timely surveillance as well as in our ability to study the relationship between the drug and its toxicity.

What for me in my tenure at the FDA has been the most challenging is how to appropriately weigh evidence from different sources each of which has its own strengths and limitations.

There is certainly no easy formula for how to continuously weigh, assess, and reassess the benefits and risks of products as we try to translate data from randomized clinical trials, adverse event reports, case reports, observational studies into actionable information. And to me that synthesis can be, and I think will always continue to be, the greatest challenge for all of us.

Nevertheless, we do act. We act with label changes. We act with warnings. We act with alerts, medication guides for patients. And all of these with an eye towards limiting the risks of these products.

And what we sorely lack is a consistently applied regular program to evaluate the public health impact of our actions. There is some good data in studies that, you know, our organization has conducted looking at whether people read their healthcare provider letters or the impact of various alerts.

Think we're at least comfortable with the notion that while these efforts have influence, we really haven't carefully assessed how these products are received and what impact they have.

And finally, our assessment of safety in my mind will always be circumstantial. Until the science of drug-induced injury has advanced sufficiently for us to understand the mechanisms of injury, to identify markers for these mechanisms, and to help us identify which populations and even which individuals within these populations are at risk for drug-induced injury.

I suspect, Dr. Woodcock, you'll probably have more to say about this later.

And with that, I'm happy to answer any questions, Dr. Shine, from you and the members of the committee.

CHAIRMAN SHINE: Thank you very much, Dr. Seligman. I have to say that since the issue of packaging and names and drugs and so forth was one of the featured recommendations of To Err is Human, I'm delighted to see you taking a very close look at that and following it carefully. Is industry responsive to concerns about similarities in names and packaging? And have you had a reasonable response in terms of making sure that one can tell one Cephalothin from another Cephalothin?

DR. SELIGMAN: Well, you know, industry invests huge resources in choosing names. They know that we carefully scrutinize these. In fact, at present, we reject about a third of the names that are proposed to us because of sound-alike or look-alike or potential confusion issues.

We have held public workshops on these matters. We have, at present, a draft guidance in the works on good naming practices.

But there again, in many ways somewhat dependent on the size of the company -- or actually definitely depending on the size of the company, lots of resources being invested particularly in the larger manufacturers on carefully screening and doing the kinds of either analyses or focus groups that we -- the techniques that we employ to try to determine whether or not there might be a drug name that would cause that kind of confusion.

CHAIRMAN SHINE: Questions from the committee? Dr. Laurencin?

MEMBER LAURENCIN: How many people are working in the Post-market Drug Safety Risk Assessment Program?

DR. SELIGMAN: In the Office of Drug Safety, we have about 100 employees.

MEMBER LAURENCIN: And in terms of the adverse --

CHAIRMAN SHINE: They work on issues across the board in Safety. Not just in Post-marketing --

DR. SELIGMAN: Well, no, just pretty much focused on the post-marketing arena.

CHAIRMAN SHINE: Just post-marketing?

DR. SELIGMAN: You'll see in --

CHAIRMAN SHINE: Okay.

DR. SELIGMAN: -- Theresa Mullin's next presentation sort of the breadth of how, you know, safety work is done throughout the Center. But the 100 people that I described basically work on the adverse event, drug utilization, medication errors, sort of the limited activities that I just described in my presentation.

MEMBER LAURENCIN: Just as a follow up to that, 400,000 adverse events, fiscal year `04, how many drug label changes took place from those 400,000 adverse events that were reported?

DR. SELIGMAN: Oh, I don't -- someone else have -- I don't -- I know that there are lots of them. But I don't know the degree to which -- I don't have that information right at hand. Does anybody?

Theresa, does anybody have the information on how many label changes there were in this last year?

DR. MULLIN: I don't know how many label changes there were.

DR. SELIGMAN: All right. So what Theresa says, we currently have a study going on which is looking at black box warnings, in particular, and how many of those were done. But I don't know to my knowledge whether we have data on the total number of label changes that occurred --

CHAIRMAN SHINE: It would be interesting to know that.

DR. SELIGMAN: -- in answer to your question.

CHAIRMAN SHINE: Dr. Roses?

DR. ROSES: You mentioned the timeliness, efficiency, and thoroughness of those issues with the AERS database. What and if ever has anybody done to look at the accuracy and validity of the self-placed complaints?

DR. SELIGMAN: We look, particularly when we're looking at a case report, we look, you know, we often ask for hospital records, sometimes pathologic specimens, other data to help, you know, validate a particular report. I mean it's, although I don't have any specific data, I can tell you that there's lots of missing information in these reports. And missing information makes it difficult to interpret lots of these cases.

These are often cases which are complex, individuals with complex medical histories, on medical regimens with lots of different, you know, products that, you know, could be responsible for the adverse event observed.

Is that kind of what you were getting at?

DR. ROSES: Well, what I'm getting at is you said you act and I was wondering what you act upon. And whether or not we could design a much better database that would get prospective data in a standardized form from patients taking new drugs.

DR. SELIGMAN: Well, I think that's why I mentioned in my presentation, you know, the possibilities that, you know, electronic medical records or others might offer us or, you know, databases where we could actually query directly then get that kind of information which would fill in many of the gaps that we get currently -- many of the gaps that currently do exist in our voluntary reporting system.

CHAIRMAN SHINE: Dr. Pi-Sunyer?

MEMBER PI-SUNYER: Yes, I have two questions. The first is are there other countries that are doing this better that you can get information from? Or not?

DR. SELIGMAN: Better. That's one of those words.

CHAIRMAN SHINE: Yes, let me -- I was going to ask you a very similar question.

DR. SELIGMAN: Yes.

CHAIRMAN SHINE: Reference was made that we do not have a national pharmacovigilance program.

DR. SELIGMAN: Yes.

CHAIRMAN SHINE: And the question is what would be the ingredients of such a program? And does such a program exist somewhere else?

DR. SELIGMAN: Well, to my knowledge, practically all countries, at least in Western Europe, rely in voluntary reporting. France, for example, has a series of pharmacovigilance centers that are responsible for, you know, regions throughout France that are more active and aggressive in getting physicians and other healthcare providers to report to their center.

You know the British carry on a regular standard survey for new products that are introduced into the British market.

But many of these systems -- many of the Western European countries, because they have nationalized healthcare systems, have access to, you know, a national database, if you will, that allows them to do the kinds of things that we can't do in this country because we lack such a national database.

That's why that slide of, you know, a little bit here with Tennessee Medicare and a little bit with, you know, Kaiser Permanente, and the Veterans Administration and all of these various pieces to try to assemble sort of the picture of the elephant.

But, you know, I think many of the initiatives that are occurring elsewhere within the Department of Health and Human Services within the nation to try to, you know, standardize healthcare fields, data, electronic reporting, all of which, I think, will ultimately filter down to our benefit an be able to kind of access in a standard way the kind of information that will allow us to pick up on early signals.

Absent that, we have to do sort of purposeful kinds of things in a variety of different environments to get that kind of information.

CHAIRMAN SHINE: And you had a second question?

MEMBER LAURENCIN: My second question is -- it's a follow up to Dr. Roses. As you're trying to get information on reports, are you finding that HIPAA is a problem?

DR. SELIGMAN: No. I mean well I shouldn't say that. HIPAA -- there is, you know, an exemption or exception within HIPAA for, you know, public health activities and public health surveillance. And we can assure the providers who give us information about our ability to protect the confidentiality of that information within the current privacy statutes.

Nonetheless, you know, HIPAA has, you know, created a whole new level of, you know, concern and caution that we have to overcome on occasion. But I'm not aware that it has, in any meaningful way, impacted our ability to gather and gain information.

CHAIRMAN SHINE: Thank you very much. And your mellifluous tones --

DR. SELIGMAN: Okay.

CHAIRMAN SHINE: -- were well heard.

DR. SELIGMAN: Okay. Thank you.

CHAIRMAN SHINE: All right. Let's move on to Dr. Mullin.

DR. MULLIN: Good morning.

CHAIRMAN SHINE: Good morning.

DR. MULLIN: Let me know if you can't hear me using this. I'm going to present today the results of an internal study that we did last winter to try to provide an estimate of the total resources devoted to drug safety within the Center for Drugs.

We felt there was the need to do a study like this, because of the ongoing public discussion, and policy options being considered for dealing with and improving the drug safety approach within the agency. But the information that was publicly available was incomplete and providing an incomplete picture, and the perception of an imbalance ‑‑ a pretty severe imbalance ‑‑ in the resources currently being devoted to drug safety.

Well, what was available to inform those discussions? The most readily available resource numbers that we could provide in response to questions from the press, for example, were budget figures for, say, components within the FDA Center for Drug budget. Prescription Drug User Fee Program ‑‑ we have to report those numbers, so there were PDUFP numbers that could be provided.

We've had special additional needs requests we've had over the years for the Office of Drug Safety, so we could provide a number for that office. But these often were not tracking safety activities.

They were components within the organization, and, as a result, it would lead to statements sort of like this one I have here as an example. This year, the Safety Office received less than $24 million of CDER's budget of almost $500 million.

In addition, the reports in the media suggested that there was a limited understanding of the rather complex process of assuring drug safety within CDER, including the process of pre-market review. And here are some quotes that illustrate, we think, that limited understanding that was being conveyed.

FDA and industry have gotten very successful at getting drugs to market based on their efficacy. Safety has become a stepchild to the ‑‑ of the agency in the process.

Currently, drug review divisions generally oversee labeling and other decisions, even after medicines go on the market, with the Drug Safety Office in an advisory role. And these statements on the face are true, but you're missing all of the other background and effort around that, and so it ‑‑ we are ‑‑ it tends to convey a partial picture.

And so we wanted to do a study to try to provide a more complete account of the safety work and the resources devoted to that safety work throughout FDA, and answer questions like these. What activities are considered part of the drug safety effort? What level of effort does FDA devote to drug safety? And how much drug safety effort is spent on pre-market versus post-market?

Again, there was this notion being conveyed that safety was sort of a post-market concern for FDA, and mainly a post-market concern as opposed to pre-market and post-market. So we wanted to understand better how that worked out across the different components within the Center for Drugs.

The challenges that we faced in doing this included that the analysis of doing safety ‑‑ safety is integral to many of the functions within the Center for Drugs. There is no easy pullout of that information.

We had time reporting data that we wanted to use, but that was organized by stages in the life cycle of the drug, if you will, from pre-IND all the way through to special activities that we'd use to report.

And safety, again, is an integral part across those, and so we had to do something to get into ‑‑ dig deeper into that set of activities to pull out what was explicitly safety versus something else like assessment of efficacy and other issues. We wanted to really drill down to the safety questions.

So our overall approach included selecting a sample of experts. We wanted to take expert judgments from within those offices within the Center for Drugs, technical expert judgments, and combine that with the time report data that we had to try to make an estimate of this.

So we elicited those expert judgments on the percentage of the time they spent doing safety-related activities. And we used the time reporting data to extrapolate from those individual assessments where ‑‑ I'll tell you in a minute how we used the sampling of experts that we involved. And we used the time report data to weight the averages of that time across activities within each office and build up an estimate from that.

I'd like to add at this point that we ‑‑ this was a two-month effort. We wanted to get it done because of the urgency of the discussion. We wanted to try to come up with an estimate that would account for the bigger picture of safety work, and we couldn't have done that without the help of the Office of Management in the Center for Drugs.

They were able to provide us the time report data, break it down into the levels at the office level that we needed it, and by job series, and so forth, and the work of the Office of Policy and Planning.

And, in particular, in the CDER I'd like to mention that Richard Allen, Dong Kim, and Bob Linkous were pretty critical to our effort. And in the Office of Policy and Planning, Kara Morgan and Corrina Sorensen worked a lot on developing the methodology and the protocol, and Malcolm Bertoni was leading the effort on developing the methodology and the modeling, which was utterly essential to producing the numbers I'm going to share with you today.

Okay. Here you see the offices that we involved in this study. We had to look at the office level, because the percentage of time spent on safety work and what these offices did varied substantially. The scientific expertise, you see over in the right-hand column, populated many of the offices, but what they were involved in was very different.

And so here we've got offices ‑‑ new drugs, pharmaceutical science, the pharmacoepidemiology and statistical science, clinical pharmacology and biopharmaceutics, the Office of Compliance Medical Policy, the Office of Counterterrorism and Pediatric Drug Development, and executive programs ‑‑ have somewhat specialized functions within the Center for Drugs.

We took advantage of that, if you will, because there's a fair amount of homogeneity within those offices. So you can go to a medical officer doing drug reviews and, by and large, the way ‑‑ the way they're spending their time in one division was rather similar to the kinds of questions and the way they'd spend their time in another division.

So we were able to use the sort of homogeneity within offices to extrapolate from experienced reviewers and scientists within those offices. And that's really the approach we took on this.

So we developed a multi-attribute model using the time report data. Now, the time report data in the Center for Drugs has 400 categories ‑‑ pre-market and post-market distinctions. To do specialized studies, we often want to go down to that level. We didn't want to do that in this case, thought that would not be helpful.

But we used the most recent four-week time reporting data, which was from mid-October to mid-November of 2004. So that was ‑‑ people had just finished doing that time report cycle, and the technical staff do 100 percent reporting. So basically we had a very good snapshot of where people were spending their time.

We then, in order not to use the 400 categories, we had some of the experts in the Center for Drugs who are used to using that time reporting system, and performing the work within those offices, help us aggregate up to a smaller set that was still meaningful, and that the people using it could still connect what they're doing to those smaller number of bins, and we could still maintain that pre- versus post-market distinction which we wanted to keep as we got the estimates.

We also wanted to pull out overhead activity, and so we didn't include that in the estimate because we're trying to drill down to safety work. We conducted formal interviews. We developed an elicitation protocol using the principles of the Stanford/SRI approach, which is a motivation of ‑‑ to try to be sure that people are willing to participate in the effort and understand and not going to produce a biased estimate.

And that they ‑‑ in fact, we would ask a particular person that we interviewed, "Do you feel you are able to speak and represent the typical work of somebody performing this function in your office?" If they didn't think so, then they wouldn't be a good person to involve in this.

So we needed people who had some experience and who were very familiar with that function within their office, and we used that within our motivation phase.

Structuring ‑‑ we used the time report categories to help structure their thinking about activities that involved safety, so we would take a particular activity ‑‑ and I'll show you in a minute the list of activities that we used, and it's in your presentation, if you have the handout.

Within that category, what kinds of activities do you engage in? What would be ‑‑ if anything, would be explicitly safety? Not efficacy, not something else. This is directly related to product safety. We'd have them define that, identify what those items were.

Now, when you're reporting time to that category, on average, what percentage of it would be devoted to the things you just mentioned, you just described to me? And we did that systematically in each of these categories.

We conducted 51 of these interviews spread across those offices and those job functions that I showed you a minute ago. We used the 2004 time reporting data and budget actual data, which is the number of FTEs spread across the organization to apply our time reporting ‑‑ to apply those percentages that we got from the interviews.

And just quickly, although it's so small in your handout ‑‑ you may be not that quick to see this ‑‑ on the left you see the super categories that were used. And this is what we have given to people, and it's a brief description of the activities that that included, which again track to things that they had been reporting in a more detailed time reporting format.

So this is ‑‑ these are familiar terms to the people who we're working with. And it's important, of course, to have something that they can relate to and is familiar to them, in order to get a reliable estimate.

So we have a pre-IND category. You can see the meetings, of course, that may occur and the things before an IND is submitted. Here is the kind of activities you'd expect to see under the IND review, you know.

First, the review of it in the first 30 calendar days, and then beyond that, meetings on that application, and so forth, around the IND submission, and actually interactions that there might be that are of a more informal type throughout the development process would fall into that category.

The pre-IND category focuses more on the formal meetings prior to the NDA. So this would include the end of Phase II meetings and pre-NDA meetings ‑‑ time spent doing that, how much of that, what kind of safety questions come up, typically what part of that discussion and that interaction is explicitly on safety. That's the kind of question we ask people.

The NDA primary review ‑‑ and those activities, which are familiar to you all, the generics review process ‑‑ we included generics, of course, because generic drug review and generic drugs are just as important in this whole safety discussion.

Continuing, here are the other categories. We call out activities related to antimicrobial resistance, so that's in there as well. NDA supplement reviews, of course, pharmacovigilance and epidemiology work, and then general post-marketing activities.

And this would include a variety of things, as you can see ‑‑ Phase IV activities, promotional material review, compliance activities; the Office of Compliance, including inspections, to make sure the product is safe; as it's being manufactured, the ingredients are safe; and then pediatric studies, which include, you know, the written requests, and so forth, you know, the range of things there.

The result of this analysis, as we took the estimates ‑‑ and let me explain for a minute ‑‑ we talked to people in different disciplines and functions, and I'd like everybody ‑‑ now, there's a typo in the handout. So if you can take your pen right now and correct your handout, because the numbers got flipped. I don't know how, but it should be 32 percent in the pre-market and 18 percent post-market, adding up to 50 percent.

Well, what we did with these 50 interviews spread through those functions and those offices is take those estimates and apply them to the hours and other people who are ‑‑ who reported their time across those time report categories and the percentage of safety that, say, you know, a pharmacologist will spend within that category to the pharmacologists in that office, and build up those hours office by office, because the percentages differed and the hours differed by the office that they were in.

So each office built up and we aggregated up through the organization. You can imagine an org chart where we're aggregating up through the org chart to get the total effort for the whole Center for Drugs. And that total came to 50 percent.

And here's the ‑‑ a breakout for you to show you how that looked across the different offices. And what you can see is that the percentages varied, but, of course, the size of these offices varies, so the number of FTEs as a result reflects that. So in the Office of New Drugs, aggregating up 51 percent of the effort that they engage in over the course of their activities they considered to be explicitly safety.

And that translates into 352 full-time equivalent staff. That's what those hours translate into.

Office of Pharmaceutical Science ‑‑ 51 percent is what the workup, the buildup number, was for them for their safety-related work, and that translates into 252 FTEs. The Office of Pharmaceutical and Statistical Science as a whole, across all of the offices within that ‑‑ within that super office, 60 percent of the overall effort and 126 ‑‑ Office of Drug Safety, 100 percent of their time. So you get 96, as Paul said, about ‑‑ about 100 people there.

And you can just see the rest of the breakout there of how the percentage of time spent and what that translated into in terms of numbers of people, or FTEs is really a more accurate way to describe the level of effort.

How does that fit in the context of the whole agency staffing? This just shows you a summary of that, so the drugs ‑‑ you see it's about half of their folks, and here we have the FTEs on board for 2004 for the other components or the other centers within the agency. And so it's about ‑‑ it's 50 percent of drugs, it's about 10 percent of our overall FDA effort, our overall FDA resources.

So, in quick conclusion, ensuring safety involves the work of a lot of different offices and different disciplines, and they're doing different things, although it all is critical to safety.

So it begins at the earliest stages of drug development before early tests in humans, and it continues through the oversight of clinical trials and the marketing of applications, marketing application reviews, continuing through to post-market surveillance, risk management, epidemiological research, and it also involves oversight of commercial manufacturing to be sure that the product itself physically is safe, and proper labeling and monitoring of product promotion that assure that it's safely used, and overall about 50 percent of estimated staff effort.

CHAIRMAN SHINE: Thank you very much, Dr. Mullin.

Dr. Cassell?

DR. CASSELL: Thank you. Very interesting and very good to know, comforting. I have two questions, and one of them is related to the antimicrobial resistance. First of all, could you say a little bit more about the rationale for collecting that data? In this particular survey ‑‑ and then, I'd be real interested to know what the actual percentage was.

DR. MULLIN: We could probably ‑‑ I could ‑‑ the second question I'm probably better able to answer. The first one, we felt that any category where people have been reporting their time we wanted to explore.

And since there was enough time being reported there, we didn't assume that it was efficacy, just safety, either/or, it was just another way to frame our discussion systematically with the people that we were talking to throughout the organization. And it would be, we thought, an omission to not have it in there, and so we probably can pull that out.

DR. GALSON: You may be wondering, why do we even collect this? Is that part of what ‑‑

DR. CASSELL: Well, kind of. I mean, I'm really pleased to see it. I've not ever seen that data, but I ‑‑ it would be enlightening I think to know ‑‑

DR. GALSON: Yes.

DR. CASSELL: ‑‑ in terms of what we know is going on on the outside ‑‑

DR. GALSON: Right.

DR. CASSELL: ‑‑ how much time you're actually spending on the issues internally.

DR. GALSON: Right. The reason is very pragmatic. When Congress gives us specialized funds for a certain topic, we need to track and make sure that we can justify that we're spending that money. We have had targeted funds for antimicrobial resistance, not right now, but a few years in the past.

DR. CASSELL: But the percent you don't remember right off hand?

DR. MULLIN: I don't have it off the top of my head, but we have such vast and detailed spreadsheets that I suspect we could pull it out for you.

DR. CASSELL: The other question was with respect to the generics and the ‑‑ I guess the frequency with which you might have adverse events associated with the generic product versus the product still under patent life. And in the event that you do have problems, in terms of safety issues with generics, do you go back to those companies, then, also, and ask for additional studies to be done?

And excuse my ignorance, I should probably know this, but ‑‑

DR. MULLIN: No, no, no. Well, I'm going to have to say excuse my ignorance, because I'm the economist and I work with the data. I pull everything about the substance from those people over there, so I'm going to ask Dr. Woodcock ‑‑

DR. WOODCOCK: Yes. There is a therapeutic inequivalence process. The adverse events would be put through the same process that other adverse events and part of an analysis might be if, in fact, we found out whether it was a generic or not that was prescribed ‑‑ I mean, if we get the generic name, we may not know.

But we get reports of therapeutic inequivalence where people raise the issue, they believe a generic is not performing in the same fashion, and we have a whole process to deal with that.

CHAIRMAN SHINE: This appears to be a very careful, well-done study. It roughly shows that about two-thirds of the effort is in pre-market safety activities and one-third in post-market safety.

As an economist, what kind of methodology would be worth developing to decide whether that ratio is the right ratio?

DR. MULLIN: I think you can look at how many different ‑‑ the relative contribution to safety or preventing problems of different activities, and I think you wouldn't want to remove anything that's going on now.

I think there are areas which we have sort of chronically underresourced and maybe need to do a better job to catch up with the fact that ‑‑ well, we still ‑‑ we need better tools, I think, and this is what I hear from the scientists at FDA, to be able to predict safety in the pre-market period, and that's something that's I think in the works and of interest.

But the new systems that we need to keep up with the utilization in the marketplace and the multiple prescriptions people use at the same time are creating a more complex environment, so I think maybe that's the area we need more ‑‑

CHAIRMAN SHINE: Yes. It would be interesting to try to not necessarily ‑‑ it may not be about shifting resources. It may be about the necessity for additional resources; it may not be. But it might be interesting to develop some models which look at the overall rate of adverse events or new adverse problems that are identified in the post-market period, in comparison to what happens in the pre-market period.

There are a variety of ways that ‑‑ none of them perfect, but it would be worthwhile, it seems to me, to try to ‑‑ to get a better handle on what would be an optimum distribution of resources over time.

Do you have some ideas about that, Dr. Woodcock?

DR. WOODCOCK: Definitely. I will point out that some of the safety activities that are done in the pre-market period are overseeing the human experimentation, and that is a big effort. And a lot of safety problems are caught during clinical trials, and it's in the best interest of everyone to make sure that enterprise remains safe and trusted by the public.

So that's a chunk of resources. It is also well known that one of the main causes of late failures of products to be marketed is safety issues that arise late, or maybe the FDA turns a product down because the benefit-risk analysis is not viewed to be adequately toward benefit.

So I kind of agree with Theresa. I'm not sure that ‑‑ I think there are analyses that could be done that would be very interesting. It might relate to a question that was raised by the Board earlier about, actually, the rate of ‑‑ and magnitude of label change is very high in the post-market period.

There are large numbers of labeling changes that occur, and that accounts for a lot of the activity that goes on in the post-market period from those 400,000 reports that we get each year.

CHAIRMAN SHINE: When the fundamental rubric is balancing efficacy against safety, safety has to be, obviously, a major feature of everything that goes on pre-market. I would be interested in seeing if we can't use some rubric, which over a period of time could help us with the idea of what would be the incremental resources that would be required to do the post-market surveillance if we did it, as well as the pre-market surveillance.

DR. WOODCOCK: We agree.

DR. MULLIN: Absolutely right. And with better data we might even be able to have some marginal - you know, the marginal contribution estimate, and we can certainly be looking at cost effectiveness of interventions and new things that we try. And we are looking at an evaluation component on the new initiatives that we're talking about, to see that we do get the most for the money that we spend on this.

CHAIRMAN SHINE: Thank you, Dr. Mullin.

Oh, Dr. Roses I think has ‑‑ go ahead.

DR. ROSES: I was just going to comment on what you said. One of the characteristics of the pre-market component is that people are being studied by protocols, and they're being looked at regularly. The post-market component is a different system. It's haphazard. Medical care is haphazard. And it's not being studied by the same protocols in the reports, and that was by my question about accuracy. It can be very variable.

CHAIRMAN SHINE: Sure. But that doesn't mean we shouldn't be looking closely at whether, in fact, we're adequately resourced to, for example, determine the accuracy of those reports.

Let's move on. I think Dr. Woodcock has a presentation before the discussion period.

DR. WOODCOCK: All right. I'm going to talk about ‑‑ sort of pull all of this together and talk about sort of some of the ways we need to move forward I think.

What we've heard about the current drug safety system today is there is an extensive pre-market testing that is done of products and on protocol, and so forth, with frequent observations. And then, there's a rigorous FDA review, which includes an evaluation of what are the remaining uncertainties after that clinical evaluation and testing has been done.

Then, we have proposed implementing user-friendly communication with the drug label, a modified drug label. It will be compatible with e-prescribing and the electronic decision support that we all hope will come very soon.

We have implemented over the past five years a proactive risk management system whereby the risks that are identified ‑‑ and you heard this from Armando ‑‑ they have to be called out by the reviewers and then in the risk management plans around approval overt strategies for managing those risks need to be put into place.

And then, we have post-market, we have what Paul described a voluntary adverse event reporting system, and with the FDA having limited capacity to do additional population-based studies to gain additional information.

Now, the capacity of this system is it will generate a profile pre-market of common adverse events in the tested populations during the drug development process. We also nowadays ‑‑ and this is new over the past decade ‑‑ understand drug metabolism and the common metabolism-based drug interactions, many of which caused severe harm in the past.

We are able to develop plans for managing and evaluating these anticipated risks after marketing, and we can identify in the system the rare serious adverse events after marketing. And despite all the flaws of the voluntary reporting system, it's actually quite efficient in identifying many of these rare adverse events. The problem is determining whether or not they were linked to the drug.

Now, what the current system does not elucidate, as we've heard over the past several months with regard to the NSAIDs and the COX-2s, it increases infrequency of events that occur commonly to uncommonly in the population that is not taking the drug. In other words, if there is a background in the population that is not necessarily rare but not real common, it may not pick up that increase in frequency.

Time-dependent events, as we've talked about earlier, it may take a much longer time to evolve and occur. Events occurring more frequently in populations that are not tested in the trials, particularly the people who are very sick who weren't entered into trials, people who are in polypharmacy, for example, people with multiple medical problems, these we may not be able to extrapolate fully from the pre-market experience.

And events ‑‑ and this is very common ‑‑ that occur much more frequently with off-label uses, and we ‑‑ although we have an extensive program, we don't always pick up products that are implicated in medical errors post-market. We also don't always pick up products that people may abuse post-market, and this ‑‑ that's hard to predict in humans. So the current system does have safety limitations.

And as we've heard, we can utilize the emerging electronic medical records system better for surveillance, and that poses a great opportunity for finding things. Also, doing randomized trials or registry ‑‑ registries in practice settings after marketing. And, again, that also will be facilitated once we have a more coherent health care system in place.

We also need ‑‑ and I think Paul Seligman's people are putting in place or making plans for doing more surveillance in specialized settings, such as from the ER. The ER is a place you would expect to pick up a lot of adverse events, because that's where you go when you have one. Or in nursing homes, where a tremendous amount of adverse drug problems occur. So that's a very positive outcome.

And all of these approaches should be implemented, and probably whatever the Institute of Medicine comes up will be a very comprehensive list of opportunities. But the traditional focus here, I would like to remind you, is on detection, and then it's on communicating these problems through warnings, precautions, or whatever, and trying to manage the ‑‑ knowing the prediction, okay, knowing that these could occur with the drug.

But we need to add, where possible, really being able to predict who is going to get these adverse events. We need to move to a different level of management of adverse events. We need to try and prevent them rather than simply describe their occurrence in the label. We also need better measures ‑‑ and Paul mentioned this ‑‑ to monitor their emergence before they have fully occurred, and to mitigate them.

The goal of the new science, I think, would be to avoid treatment of individuals at high risk for an event, because serious side effects, even in drugs that are withdrawn from the market, are usually only occurring in an extremely small number of patients.

And when these withdrawn drugs occur, we hear from many people who have benefitted from those drugs who are very bitter that it has been removed from the market. And we need, as I said, to develop new ways of monitoring for emerging toxicity before it becomes severe.

Now, I'd like to go over very briefly ‑‑ I won't take much time on this ‑‑ why we do have to make these changes as a society. Right now, as Steven alluded to, drug development, including animal and human testing, is largely empirical in nature. And this tradition focuses on the outcomes and population means, and then the observation of outliers.

And if you look at the guidance that Armando presented, that's basically what it's going to talk about to a great extent.

This empirical approach directly translates into the trial and error approach used in clinical medicine, which I think Dr. Roses just said was haphazard. Something like that. And that's because clinicians don't have, really, any better information right now. And the way this is done creates, I think, a major loss of information that can be gained in the clinical development program and translated into clinical practice.

But ‑‑ and this has led to many calls, as you all know, for many more clinical trials and significant more work pre-market, larger clinical trials, trials in many more populations, and so forth, specific trials. But we all have to recognize there are significant limitations and a number of questions that can be answered via empirical testing.

There just aren't enough people in the world or money in the world to answer all of the kinds of questions you might want to answer about any given drug, much less the whole panoply of drugs. And these limitations are imposed by the availability of people, by the practices evolving, even when you're trying to do these trials, by cost, and so forth.

And the clear fact is despite the fact that any drug development program will cost hundreds of millions of dollars, we often lack information and approval of that drug that is important for prescribing. We do know that the drug is effective in a defined population, and we can describe its common side effects in that population that was tested in the trials.

But we know that many of the people who are subsequently exposed to the drug will not benefit from the treatment, and we know for any drug we approve that some people who are exposed to the drug will be harmed by the drug.

Now, let's switch to another imperative. Right now, the current data says that for drugs that enter into Phase I trials, they have already gone through perhaps five or more years of development work preclinically, and a huge investment of time and effort, their success rate of actually becoming marketed products is lower than 15 percent.

And there's no other real part of our technological infrastructure, say building bridges or designing airplanes, or whatever, that operates at nearly this level of risk. This is just an astounding level of risk ‑‑ of failure.

The same evaluative tools that will ‑‑ I would like to explain ‑‑ or tell you or argue that the same evaluative tools that will improve predictability will also support markedly improved decisionmaking in clinical medicine, and, therefore, will impact significantly on the drug safety problems that we are currently experiencing, and much of this information can be developed.

Now, I'm going to give you some examples to make this real. Drug metabolism is a good example. For example, there is a hepatic enzyme, which I have listed there ‑‑ and don't pay attention to all of the words and numbers and everything ‑‑ but this is an enzyme that is responsible for ‑‑ in many drugs for metabolizing it within the body, using it up.

And this is ‑‑ there's a typo in here. Drugs that are substrates of this enzyme may have up to 10-fold higher blood levels if you're a poor metabolizer. So there are probably a person or people in this room who is a poor metabolizer of these drugs.

It's a very bad choice of words. The clinical pharmacologists made up the poor metabolizer. But what it means is your enzyme uses a drug ‑‑ uses a drug up very slowly. And if you took the normal dose of the drug ‑‑ and we don't know who in this room is this person, that's the problem ‑‑ if you took the drug, you would have 10 times the blood level of most of the other people in this room from that drug.

And the distribution of alleles ‑‑ in other words, the distribution of really fast metabolizers, very slow metabolizers ‑‑ varies with ethnicity. And this gets to some of the points that were raised earlier about the different profiles in different ethnic groups.

The drugs that are substrate of this enzyme are frequently implicated and unavoidable ‑‑ they are called unavoidable right now ‑‑ side effects. And we think that the exposure ‑‑ in other words, the fact that these perhaps 1 in 300 people walking around are getting extremely high exposure to this drug ‑‑ is probably a very large contributor to this ‑‑ to the problem of these drugs causing adverse events.

Now, the Center for Devices recently approved a commercial test to determine ‑‑ you can determine your status, your enzyme status, and determine what kind of metabolizer you are. All right? But, of course, it's not being widely used clinically yet.

Now, drug development, though, recognized this some time ago. It used to be that these drugs were put on the market, drugs that were the substrate, and they got into a lot of trouble post-marketing. What happened was FDA worked with industry to develop tools that could use human cells and other models, but mainly human cells, to predict the human metabolism of candidate drugs before they got into people and before they began to be developed.

This caused the manufacturer to start eliminating candidates that were substrates of this enzyme system and not putting them into clinical development. Okay? Fifteen years ago, problems with metabolism was the number one ‑‑ and availability, and so forth, was the number one cause of late failures of drugs, if you can believe that.

In other words, drugs late in clinical development, they had already gone through many of the clinical trials and they had to be pulled from development, it was because of drug metabolism problems. Since this screening has started, that has dropped down to an extremely minor cause of late failures in development, because those drugs are eliminated early from the pipeline.

Now, this just shows the power of an evaluative, predictive tool that can be used early to remove this ‑‑ these problematic drugs from the pipeline and, thus, markedly improve the safety of marketed drugs.

Now, I'll tell you the critical path initiative is a great example. It's focused on development of additional tools that would actually do this. Unfortunately, the clinical test, a test just approved recently by the device center, for this metabolism has come much later.

All of the drugs that are on the market that are a substrate of this enzyme, the clinicians are not used to using this test and adjusting the dose of these products. So safety problems continue to arise related ‑‑ because of variable metabolism. One of the things we'd like to do is do some studies looking at can ‑‑ can metabolic testing like this decrease the frequency of adverse events of these marketed drugs.

But, in general, these new technologies, including genomics, proteomics, metabolomics, and so forth, can really help us predict both in drug development and in clinical medicine, and that's really my point. The drug safety problem is, to a great extent, related to the fact that so much of therapy is still empiric.

We could go and do these tests in patients who have had serious side effects and study them versus patients who haven't. And I would defer to Dr. Roses who has done a lot of this work. We can study this both in prospective trials and in patients who have had serious events that are reported under MedWatch.

We need to develop the ability to predict and avoid treatment of high-risk patients or monitor people for development of side effects before overt toxicity occurs. And while we need to do all of this other safety surveillance, because we'll never be able to predict all adverse drug event problems prospectively, this is where we need to move ‑‑ toward prevention, targeting individualization of therapy overtly, to really deal with the safety issues.

In addition, as we develop targeted therapeutics ‑‑ and this is beginning to happen in oncology ‑‑ we can identify subgroups of patients who have high response rates. This means we avoid treatment of people who have very low probability of benefitting. This is equally important.

Right now, we just don't know for many therapies, because they're empiric, who is going to benefit. This can help drug development, because it will remarkably decrease the needed power, how many patients you need to test, okay, before you find a positive result.

But on the other hand, out in the marketplace, it can significantly improve the benefit-risk of a therapy, because at least the people who are getting the toxicity had some chance of benefitting from the drug.

So for the future of drug safety, we really do need to continue to improve the current system, and we look forward to the committee discussion, but we also have to start really focusing on delivering the right dose to the right patient, and to monitor those people for emergence of side effects, and to really focus on preventing serious side effects by predictive measures. And that's one, as you know, of the very big focuses of the critical path initiative.

The issues for this ‑‑ how would we get this applied science done? Well, if you've heard some of this message before, it's because we're telling people this again and again to make sure that people really understand this is really part of drug safety. Unfortunately, there has been no entity really charged with this task. It has not really been seen as a job of the FDA, or any group really.

Public-private partnerships, as I alluded to earlier, may be an ideal way to get some of this work done. We need some venues to pull and itemize data from various trials, and so forth, and we need the ability to conduct studies or add-on modules to existing trials to develop some of these safety biomarkers. But we are very optimistic that we can get this work going and remarkably improve drug safety, as I hope the drug metabolism example has ‑‑ has shown.

Thank you.

CHAIRMAN SHINE: Thank you very much, Dr. Woodcock.

Why don't we begin the discussion period, with specific questions for Dr. Woodcock you might start with, but I would raise any questions that you would like to.

Dr. Riviere?

DR. RIVIERE: Yes. I've had one general comment and one general question, and your last slide pretty much nailed the comment. I think a lot of what has happened is that we have not got the resources and the base knowledge as to how to predict these side effects coming out that pick up in post-surveillance studies.

It needs to be really underlined somewhere in here, either as a charge to the IOM Committee looking at this, that ‑‑ how do we actually direct research to start determining what are the signals that we're looking for?

DR. WOODCOCK: Right.

DR. RIVIERE: Your very clear example on the drug metabolism ‑‑ and this has happened with other areas ‑‑ when we actually had mechanistic information, we find it.

DR. WOODCOCK: Right.

DR. RIVIERE: The other ‑‑ the other question is is that we develop these preclinical toxicology databases. I was a little bit disturbed in Dr. Galson's statement saying that the science of efficacy has advanced, but the science of toxicology has like sort of just been stuck and did not advance.

(Laughter.)

That's really not true. What I'm not sure of is how much of some of these areas ‑‑ and I know why it's not applied to preclinical studies. Because it might generate false positive signals, nothing ever goes forward ‑‑

DR. WOODCOCK: Right.

DR. RIVIERE: ‑‑ and, you know, in that constant pendulum battle.

Has anyone looked at these post-marketing adverse drug reaction databases, and then done studies to go back to the full preclinical data, not necessarily the univariate detection of a signal that we know cause and effect, but basically could we have come up with something to detect this earlier? And I think that's an area that the data is there.

We have a lot of information on both approved drugs and non-approved drugs.

DR. WOODCOCK: Right.

DR. RIVIERE: You could pick it up right now. If you knew this was a metabolizer of this specific enzyme, you'd know there was a problem. But is there ‑‑ any attention been focused to basically let's go ad hoc ‑‑ post hoc go back now and look and see if we could have done better.

DR. WOODCOCK: These are the data mining efforts I alluded to earlier, and we do have some folks who have come forward and are interested in partnering with us to conduct these studies, or retrospective data mining efforts.

Part of the problem is a lot of the data are not electronic, and so there's this huge data entry problem that occurs every time we have to go back and data mine the prior results. But you're right, FDA is sitting on this huge repository of animal and human studies, and then have ultimate human outcome data to follow that.

CHAIRMAN SHINE: Dr. Roses?

DR. ROSES: I can add something to that. I took the Lotronex database, the entire database, hired an outside firm, for privacy reasons, to track down every patient in that database to find out as much as we could about the adverse events that were happening. I'll just make three quick remarks.

The patients either came through self-referral, they came through physician referral, or they came through lawyer referral. In no case of a lawyer referral were we ever able to gain the identity or a DNA sample anonymously from any patient.

About half the patients were lost to followup in the physician referral. In the self-referral database, out of 186 ‑‑ I think the number was ‑‑ 168, one of those two ‑‑ we got about 30 patients that we actually got to. And of those 30 patients, three had never gotten a prescription for Lotronex, nor ever taken Lotronex.

That's why I'm concerned about the quality of the data that we get. And you can't go back. It's not just the information to do this genetically, to do it ‑‑ just even a SIP-2D, you have to have a sample, even if it's an anonymous sample.

DR. WOODCOCK: Well, one of the thoughts we have, if I may partly respond to that, is it's possible that if FDA asked for some of these and conducted ‑‑ partnered and conducted some studies like this, we may get more willingness by patients to give us the samples.

CHAIRMAN SHINE: Other questions for Dr. Woodcock? Oh, Jesse. Dr. Goodman?

DR. GOODMAN: I was just going to amplify a little on the comments several people and Janet made. And I'm not sure I completely got your original question, but I think part of the intent in some of FDA's pharmacogenomics initiatives, and the voluntary submissions, is to increase our power to do these kinds of things.

And even in some of our products like the vaccines, etcetera, or gene therapies, you know, we're encouraging archiving during ‑‑ during studies of ‑‑ of information. And, after all, DNA and genomics and results of microrays are just another form of information. But it's one that already is pretty automated and accessible.

A lot of the kind of concerns that somebody mentioned that you get all of these ‑‑ you get ‑‑ 2,000 genes go up, and some of them have ominous names. What does that mean? But the ability to then go back to them retrospectively, if a problem is discovered in a much larger population base and identified predictive tools, I think that's the kind of thing that can really move our ability to prevent some of these adverse events forward, much the same way it was done with metabolism.

DR. RIVIERE: And that's what I was really aiming at is ‑‑ is that it's not the real noise in the 80-odd databases. It's these really serious ones that we now know as this class effects, and maybe subclass effects, to go back to that original data ‑‑ I mean, which there probably were samples even collected some of these analyses ‑‑ this is the animal database and the preclinical database ‑‑ to see if we could have looked at that differently to actually pick that up.

CHAIRMAN SHINE: Allen, go ahead. Do you want to ‑‑

DR. ROSES: Yes. In predictive tox experiments that have been done, and they're in the process of being reported, written ‑‑ they have been reported, the experiment ‑‑ it's done in a number of companies. The experiment that we did was to ask the question: if we have chemicals that pass four-day animal tox, then fail at 28 days, can we predict which ones they were?

And some are using genetic-genomic kinds of techniques. The answer is we found 10 markers that would predict, with 98 percent accuracy, that a drug that passed four-day tox, looked pretty good, would fail. It's another way of asking the question: can we make predictions? The answer is, yes, we can, but now we have to get the database.

DR. WOODCOCK: Right. But that's a tremendous opportunity. Even there, it shows that you detected emerging toxicity before it became overt, and that's really one of the goals that we want to do in patients as well.

Sorry to interject.

CHAIRMAN SHINE: Absolutely.

Dr. Laurencin?

DR. LAURENCIN: I'm going to reflect back on some ‑‑ on the old science again. I mean, to what extent does FDA, in examining the clinical trials, examine the study population and the population to which the drug is going to be administered to? Because clearly, obviously, if the populations are different, then one would expect to have different adverse events. And that may be what we're seeing.

DR. WOODCOCK: The indication that is given to the drug relates to what was tested ‑‑ what population was tested in the clinical trials. So, yes, we look at that very closely. We do a lot of subgroup analysis and cut the data many ways for the pre-market database.

So whatever the company is granted reflects the patients who were studied. However, the clinical usage of the drug will always or inevitably be much broader than whatever narrow indications are put on the label. That's just how the practice is.

CHAIRMAN SHINE: Which leads to lots of questions about what, in fact, should have to happen in the post-approval state.

Dr. Thomas?

DR. THOMAS: I was going to ask about your remarks with regard to polypharmacy in this geriatric generation that we now have and multiple drug use. Do you have any way to specifically identify, say, drug-drug interactions or drug-nutrient interactions? Or is this mostly in the post-surveillance phase?

DR. WOODCOCK: No. First of all, we identify the drug metabolism prior to approval. So those interactions that can be predicted based on metabolic knowledge and what we know about all the drugs on the market, that is already identified.

In addition, companies are often asked to do specific drug-drug interaction studies with drugs that would commonly be used in that condition. Okay? But if you did a factorial design of all the drugs that are out there, and then added a new drug on, we cannot do ‑‑ or no one could do an empirical test of the interactions of all those drugs together and the pre-market period would never get done. So that is not done, and that has to be picked up, and often is, in post-marketing.

CHAIRMAN SHINE: Cato, did you have another comment?

DR. LAURENCIN: A followup question. NIH requires an assessment of numbers of ‑‑ contributions of women and minorities to clinical studies. And, in fact, when we study a section, we actually grade the ‑‑ grade it unacceptable ‑‑ either give it a U or an A, unacceptable or acceptable.

And it could actually mean that the clinical trial won't even be funded if it's got unacceptable numbers of women and minorities in the study. Does FDA do that? If FDA doesn't, why not? And what's the plan?

DR. WOODCOCK: Well, there is reporting of enrollment, and so forth, and that was done in regulations and guidance sometime ago. But since we do not fund clinical trials, we don't have the authority to tell people they can't do clinical trials, at the time of filing, then those subgroups are reported as part ‑‑ overtly as part of the NDA database, the representation of each subgroup.

DR. LAURENCIN: But you don't fund it if you ‑‑ if these ‑‑ if a guidance document states that these numbers are going to be important, and that they're going to be looked at, and that they will ‑‑ they will reflect whether a drug may be approved, then I think that ‑‑ that there may be a response corporately.

DR. WOODCOCK: Yes. We've done everything we can to encourage enrollment. As you know, we've worked with the National Medical Association, for example, in training investigators, and so