Event ID: 417068
Event Started: 10/20/2005 8:29:27 AM ET
[ A captioner is present. Just waiting for audio ]

Good morning. In keeping with our new tradition of reviewing military history, I will tell you that today on this day, in 1944, keeping his "I shall return pledge to the people of the Philippines, general Douglas McCarthy waded ashower and the SACGHS has yet returned on its second day of our October meeting. Yes, a lot less carnage. Sort of a cleanup item to begin with. Yesterday we mentioned the coverage and the reimbursement report and the recommendations that were edited and distributed to you. The folks were supposed to take a look at those and what we are going to do is not have a protracted discussion and try to edit and reedit because that just takes too long. What we'll have is essentially an up or down vote. Will we accept the staff recommendations for some of the -- for the editorial comments, the green line version that you have, good morning? If it's no, we will go back to the original language that the committee approved at our last meeting. So we'll just go around the room. Do we have everybody here? Yes, I think so. So the question will be on the approval of the edited version of the recommendations for the coverage and the reimbursement task force, the green line version that you all have. I guess we'll start with Jim? [ Inaudible ]

I vote for approving the edited version.

I will do the same.

Yes to the edited.

I would say yes to the edited. The only things I do have a question, sort of a concern that came up to me, especially in view of yesterday's discussion, and that is for item number 7, just a clarification point that where we say "Genetic counselling to provide -- to be able to provide full access of genetic counselling services for all Americans and then it goes on to identify a body who would oversee to determine who is appropriate to provide this kind of counselling, and I think that -- I don't know if the committee would like to look at using the word "Not just genetic but genomic counselling." I think that if we leave just genetic counselling, I think genetic counselling implies specifically genetic counselling for single gene disorders and I think the information that we heard -- and the specific profession, and specific specialty, and I think with the information that we heard yesterday, that we need to have prepared health professionals who will be able to provide and possibly bill for genetic and genomic counselling in education services. So I just think that we should be -- have a little bit of foresight and I don't know if the task force would want to address that or if, in fact, this was just intended for genetic counselling.

Suzanne?

We have a copy of the report, but at the beginning we had, at injure request, added some definition of some of the terminology used in the report. I don't know if you want to expand the definitions throughout the report whenever "Genetic counselling service" is used to say genetic and genomic counselling service and make sure that the definition included at the beginning of the report encompasses both -- what you just described genetic and genomic counselling services. So I don't know if anyone has --

I would think that especially in -- maybe if we, as you said highlight that in the beginning but I think in this specific area, since it involved an oversight body and since genetic counselling is a specific profession, I think it would be here better to expand it to genetic and genomic counselling.

Do you think we need -- just for consistency of terminology, do you think we need to clang the term throughout the report, though or just in this particular area.

Just in this particular area, especially if we are asking a secretary to assume a body to oversee this. I think it's critical to make sure that it's as expansive and not just the genetic counselling profession.

What if you used -- what if you said "Including genomic" would that -- I think because then it would be -- you are trying to remind, athe that point, at that key point to make a reminder that we are encompassing both.

But if you say "Including genomic" then it means there's genetic counselling for genomic conditions and I -- genomic conditions and I think as the feel evolves there will be other counselors providing the genomic counselling, not just genetic counselling. So I don't think we should put it and phrase it that it would just include genomic.

Because you have -- I don't want to get too bogged down in editing this at all. Do you feel strongly, Agnes that --

It's just the number one section, the first bullet that identifying appropriate entity to determine which health professionals are qualified to provide genetic/genomic counselling services. Because I think -- again, I think there's a specific connotation to what genetic counselling means and I think that's where we could get into problems with this. But I will --

I don't have the specific language in front of me for the definition that is used, but when we -- the term is defined, it is meant to capture, I think how you are defining genomic counselling and it does point out that, that these services are provided not just by genetic counselors but by a broad range of providers. And so I think -- I think at least the definition, as we have it set out in the beginning of the report does include single gene counselling and also how you described genomic counselling, as well as it's provided by a breadth of health professionals.

Would we be able to then just put a little -- you know how sometimes you say see table 1 with the definition again so that people know it refers back to then.

We could do that.

Okay. Thank you.

Emily?

I vote to approve it.

And do you?

Sylvia?

I approve it. I just had one point of clarification on 7a. When you took out bill payers and put directly bill for their services that's a little confusing because you can directly bill anybody, the family, the doctor, the hospital, you know, because I think we had put payers because it was specifically third-party payers that we were looking at. Wasn't it? I don't remember. 7a. You are deleting the word "Directly bill" you are deleting "Payers" and putting "Directly billed" for their services and it's like who? Who are we -- I don't know if that's a little confusing.

Do you have the --

Does anybody have feel strongly about that? I think it's implied, the word "Payers" is implied. If you are billing someone, you are billing someone with the implication that they are going to be paying the bill, hopefully. Any other thoughts on it? [ Inaudible ]

I haven't thought of it, but it's a good point. I mean, are you talking about billing the person directly or billing third-party payers or anybody? So --

Okay.

That doesn't make a difference in the reimbursement world.

I mean, anybody can charge. Anyone can go ahead and bill the person directly. The question is who will a third party payer pay to?

Okay. All right. Let's add that in there. Is it going to change anybody's recommendation if we add that in there? Cany. Let's -- oh, Joseph? We didn't get your -- I'm sorry. You snuck in.

No, I approve. All right. The committee approves the edited version of the recommendations for the coverage and the reimbursement report. That report will be finalized and then transmitted in short order to the secondary. -- second reteary. All right. -- secretary. All right. The next order of business is really to proceed to today's agenda. The issue of genetic discrimination will be the first item of business, and as you all know, those of you who have been following this committee's work the issue of genetic discrimination, of health insurance, has been our top priority. We have been closely monitoring federal legislative issues. And in May, they sent secretary Levitt, a DVD of testimony that we heard last fall, highlighting various public perspectives on the genetic discrimination issue and a legal analysis of the wildcaty is of current -- adequacy of current. They were disseminated to the public through the web site. This morning we will hear an update on the status of federal genetic nondiscrimination and report on public attitudes about the privacy and misuse of genetic information. Sharon terry is here representing the coalition for genetic fairness. And as many of you know, she's also president and CEO of the genetic alliance, ant international coalition -- oh, Sharon is not here today. Frank? Where is he? There he is, frank Swainy. He's already seated at the table. Frank will be standing in for Sharon, and with the genetic alliance, and we have got the -- you don't have slides? Okay. Well, thank you very much, Frank for joining us this morning and for standing in. We appreciate it. And with, that we will turn to you, and just ask the members of the committee that unless there are specific points of clarification, let's hold our questions until after all the presenters have completed their presentations.

Well, thank you very much. It's a real presidenture to appear officially before the committee. I have been able to monitor the last two meetings here at least, and we're very aware of the committee's strong interest in this legislation. I told Sharon, who, as is her want, emailed me about 11:48 the night before last, and said, could I do this for her. For some odd reason I was looking at my Blackberry when this came in. So that says something probably odd about both of us. I said, certainly. But I would be a pale imitation for those of you who know Sharon, she's a whirling tushish in issue in this area and it's a real honor to represent letter and her coalition. We will do a Q&A, in a little bit under the schedule, and anything that -- that I mention, I would be happeny to try to respond to the extent possible and appropriate. Let me review the bidding briefly. Our law firm has been working with a coalition for about eight months now. So give additional legislation which was mentioned has been concept around for quite a while. The legislation that has been introduced in the house of representatives is HR 1227. It is a bill that essentially has two titles. One of the titles would prohibit the misuse that is negative use of predictive genetic information about an individual in the employment context. The second title would do the same prohibiting the negative use of predictive genetic information about an employee in the health insurance context. This legislation, as introduced, is identical to legislation that was passed by the U.S. Senate in February 97-0. Unfortunately the shows not about to house it 97 or 497-0. There's significantly more controversy about the legislation in the house of representatives this year as there has been in each of the last few Congresses. The two titles of the bill are assigned to two different committees because the OEM ployment title is assigned to the committee that has jurisdiction over labor matters. The insurance -- the health insurance is assigned to the committee that has jurisdiction over health insurance and other matters. Additionally some parts of the bills must be reviewed by the ways and means because it affects indirectly the Medicare program and the ways and means committee has jurisdiction over the Medicare program. So we have a procedurally challenging project to get everybody moving forward, hopefully at the same time, and in the same direction. As I'm sitting here this morning, the bill has 150 cosponsors. The primary sponsor is Congresswoman Judy bigger from suburban Chicago and a more energetic and committed sponsor we could not find if we had perfect hindsight. She is marvelous! She is aggressively bringing in many of her fellow republicans who she corners on the floor of the house and essentially twists their arm on to signing on to her bill. She knows what she's talking B.she's a hour who has practiced employment law. She's also a member of the house science committee and has been significantly involved in the discussion of the science committee reviewing of progress to the Hume -- human Genome project and continues to go into that project and she speaks very directly and very sincerely when she says it's really silly for us to have invested hundreds of millions and billions of dollars of taxpayers dollars that we have in the human Genome and have everyone afraid to look at it. So I won't preach to the choir about the need for this bill. I will try to discuss some of the issues that we have come up against. First of all, the 150 cosponsors is significant. I recall back in March, the bill had not yet been introduced. Indeed, we were on the cusp of it, of having it introduced and you folks were pressing us fairly hard about what was going on and we couldn't say anything, because it's really not appropriate to do until the member sponsors. It was introduced on March 10th, with 40 cosponsors. When I attended your June meeting, it had 80 cosponsors. Today it has 150. On that basis, you know, about two more meetings we should have the entire house cosponsoring it. But unfortunately it probably won't just run geometrically from here on out. The issues that we are dealing with are primarily issues raised by the business community, who have significant -- some parts of whom have significant apprehension that the bill will make it unduly easy to challenge routine work place decisions using the book of genetic nondiscrimination or to be more blunt, if negative action is taken against employee, an employee is feeling litigious, he or she may say you are discriminating against me on the basis of gender or ethnicity or religion or age, and now add genetics into the alphabet soup of claims against the employer. That's the fear of some. Not all by any means but by some in the emLier community. Tush employer community and -- employer community and that's one of the issues we were trying to work through, to make sure that the bill is clearly written and it does what it's clearly supposed to do and it does no more than that. That it's not some vehicle to boot strap, every perceived issue with our health delivery system. There are those who say, well, if this bill is passed and everybody is bound to have some sort of genetic defect, then this will somehow inevitably March us towards national healthcare. I don't understand that logic, myself, but we hear that with some regularity. We hear some other issues that are more mundane and more solvable. For example, there are a significant number of individuals in this country who do receive insurance through single life policies. That if they are not members of groups, they buy their health insurance in the individual market. In the individual market, there's clearly a medical underwriting. You can't be in the individual insurance market without having medical underwriting. Part of medical underwriting in most cases -- and I think this committee is well on record as are many groups, in favor of family histories. You go to a doctor and if the doctor is a good doctor, that doctor spends sometime taking a family history. Somebody said, well, the family history is in effect, a genetic history. Does that mean that if we used the family history in some decision, that that is therefore a violation of an HR 1227? Well, we are saying clearly not. We want to promote family history. Well, what's the difference between saying that somebody had several siblings with cystic fibrosis and having a genetic test that says the cystic fibrosis gene is present? Well, those are the kind of things we are trying to work out at this point. They are, in my view legitimate issues and in my view, we are well on our way to working many of the important issues out. At some point, off ysly -- obviously, we'll draw a line in the sand and ask our sponsors to go ahead and get the legislation up and scheduled. We are involving in informal, but I think useful discussions at the U.S. chamber of commerce, the national association of manufacturers and some other groups that have -- that has had some traditional concerns about this legislation. It would be inappropriate for me and premature even if it were appropriate for me to say what the chairman will eventually do. I don't know what they are going to do but it's at least a positive that they have agreed to sit down and talk with us and see if something can be worked out on issues where there seems to be disagreement. Finally, I might add that it's very important the work that this advisory committee does, and candidly urging the secretary to take the positions that they must take. The administration thankfully is very strongly on record in favor of this legislation. But we need active advocacy at every level within the administration, within the scientific community and within the business community. Perhaps I'm being more candid than I should. One the disappointments is that the business community, and in particular, the biotechnology and the pharmaceutical community have not really stepped up to the plate on this legislation; although, they would have a very -- many participants in that community would be possibly impacted if we could remove the apprehension that's been clearly documented in the public at large over having genetic tests and having genetic tests used in product and -- and science development. Finally, I might add, there's been a very interesting case, situation, involving a professional basketball player, an NBA basketball player who was with the Chicago Bulls, and I'm not expert by any means on the physiology of his condition or his alleged condition, but the Bulls said that they wanted him to have a genetic test because there was a concern that he had some sort of coronary condition that might result in his dropping dead on the basketball court. And he refused to have that genetic test. He's since been traded to the New York Knicks where presumably he will play basketball without a genetic test or without publicly talking about whether he's had one or not. The interesting thing, and this is a classic case den -- and hard cases make bad laws. There are a lot of individual situations there that are not clearly generalizable to the public at large, but it has galvanized attention at least in the Midwest on this issue. The Chicago Tribune has come out with an editorial endorsing this legislation and has had a fairly interesting series of articles about this as a case study and what may happen and what employers might do in the future. I'm not sure how far you can push it, but it is a vivid reminder of the sorts of issues that are out there, and the reason that we need a federal law to clearly set out for individuals that live in every state, not just in the 33 states that have state laws on the book, what the rights and the responsibilities of employers and employees and insurance companies are in this very important area. And with that, I will close my PowerPointless presentation.

Thank you very much, Frank and don't go away, because we'll have some questions and we'll enjoy hearing your views on the case that you just raised and other matters. Is Christy here? Christy. Christy White is principal and co-founder of cogent research, a strategic research marketing firm and she will be presenting a report on cogent's research findings on public attitudes about the privacy and the misuse of genetic information. The role of the government in protecting that information, and government access to genetic information. A copy of Ms. White's biosketch can be found in the briefing book. All set?

Thank you. Thank you for having us here today. We're honored to be able to speak to you about this topic that we believe is very important, and, in fact, have been tracking for a few years now, and plan to do so into the future. Cogent is a market research firm. We're in Cambridge, Mass. We have been around for ten years. Mat jurority of the work is proprietary in nature. We do a lot of work with industry, a lot of work with the government as well. We saw this as an important issue to start tracking on behalf of both populations. The objective of the research was to provide a very comprehensive and actionable assessment of where Americans views were at that point in time, the snapshot, towards genomics so that we could help inform industry, strategy and policy development. the study itself looks at four main areas, awareness, favorability and interest towards genomics. So we start out very basically trying to understand what they know about genes in general, and the relationship between genes and health. Also, what they know about genomics in particular, what they think about it on a general level for the population as a whole, and how they -- how interested they are on a personal level for themselves. The study also looks at the specific catalyst and barriers to adoption and usage, trying to understand what are those things that -- what are the perceived benefits and drawbacks? What are their fears and hopes for genomics for them? And as -- for the public as a whole. We also look a lot at preferences for delivery. I won't spend any time on that today. That's my industry speech, but we look a lot and if anyone is interested in that, we do have some very interesting data on what does the consumer -- what will max mime con-- maximize consumer interest in terms of how they want to get the information, who they want to be involved in the decision whether or not to have a test or not, and how they want the information delivered to them, what -- how they want the information stored. I will touch on that to the extent that it relates to the topic here today. Lastly, we look at messaging and communications. So using some of the information that we know from qual tauttive researchD qualitative research, I will speaking to the qawnt study here. What are some of the potential messages and things we might say to the consumers to alleviate some of the fears that they have or address some of those concerns. The methodology for this study was a web-based study. We looked at U.S. adults 18 plus. The profile of the Sam; representative of the U.S -- Sam; representative of the U.S. population, on key socioeconomic, and region and gender. They are back to the 2003 projections that's the data we used for comparisons there. We conducted the data last January, middle of January. We're actually getting ready to one the study again. We'll probably run it towards the end of January this year for our third year. This survey, as I mentioned earlier, the majority of what we do is proprietary in nature. This study was ours so we had full control over it and we got a little out of control. We asked 205 questions. There's a lot to know and explore and look into on this topic. We talked to a total of 1,000 consumers it gives us a sample of plus or minus three at the midrange. We looked at a wide variety. A good chunk of those questions probably around 40 of them are demographic questions. We wanted to tease out what are some of the differences by some of these different key demographics and we also conducted a robust psychographic segmentation so if we clump people together by their belief systems, as opposed to a demographic like sage can we understand or -- age can we understand or tease out some different information in terms of what is happening? There's a few things that we will do differentyy next year when we run this. A couple of areas that we will look into and this is based on some meetings with both the government and the industry. We'll delve more deeply into people's awareness and usage of Internet-based products. This is something that the CDC is very inned in, in particular. -- interested in, in particular. We'll look at work place discrimination. Our focus was mostly on insurance government this past year but we'll be looking more closely at work place discrimination. And also, awareness of current legislation and protections that are in place. So we want to understand, you know, do people know what protections there are? What do they think of those protections and what are their hopes for where the protections might be. So that data as I imagine would be of interest to this committee would be available, I'm guessing late -- early February next year. We'd be happy to come back and talk about that. So I'm going to walk through just a few key findings from the study. Probably around 44 of the 200 questions. So we'll give you a good picture but clearly there's a lot more here. The first basic thing we want to know, you know, are consumers even aware of this -- of the issue? We asked them how much they have eard or read -- heard or read about using individual genetic information to understand and optimize health and you can see that awareness is very broad. Three-quarter of Americans saying that they have heard at least something about this issue. When we delve a little bit deeper, our into it, we understand that it's really not a very sophisticated understanding. Only 4% of Americans are telling us that they have heard a lot. A third of those who say they have heard something can't recall any specific details. And of those who can recall some details, they are mostly telling us that they understand is a relationship between genes and health, and not specifically how they can use that genetic information to potentially alter their health or improve their health. There are a few, however, that are getting at least to some extent a more sophisticated understanding of what's happening. They tell us things like genes can be manipulated for my health benefit, they can lead to early diagnosis and detection. They can help find cures for diseases and then, of course, can be used in the prenatal stage. So there are some that are at that more sophisticated level of understanding. We do know from qualitative research that this is classic. This is what they will say back to us. It is very easy to get them to make the leap. So although everyone is aware, and they are just aware in general about the relationship, it's not difficult to get them to make that leap almost on their own of how they could start harnessing the power of that information for themselves. When we looked specifically at interest around prescription drugs, using making informed choices about prescription drugs awareness drops to 50%. 75% get the general level of understanding, but 43% with prescription drugs. That's still actually not a very low number. What are the general attitudes? Their attitudes are highly favorable. You can see in general when we ask themN general how favorable are you to use genetic information to understand and optimize health, you can see more than after of Americans are saying that they are favorable. This is actually quite a high number. We might expect to see a bell curve here with a majority of people being in that sort of middle level, particularly since they don't fully understand it yet but we see that there are a lot of people -- people are very excited and they see a lot of promise in genomics for -- in general. And they are also as excited about it on a personal level. So when we asked them how interested would you be in using your genetic information for the purpose of understanding and optimizing your health, you can see that that interest is still quite high. How does that play out for personalized medicine in particular, 50% of the adult population say they are favorable in using their individual information to make informed choices about prescription drugs. 51% are interested in using that information for themselves personally. At this point in the survey we delved more deeply. We are trying to give them an understanding of how to use the information, the specific ways in which they might use their genetic information when it comes to personalized medicine or prescription medicine in particular. We looked at three areas, efficacy. So which drugs would be most effect miff your body? -- effective in your body? Safety, whether they would have an adverse drug reaction to the drug and prevent them, whether they would take the drug preventatively. Probably not surprisingly, these are very high and quite close, but safety is clearly the one that they are most interested in in terms of usage. But very close behind that comes efficacy and not surprisingly, the third would be preventative or proactive usage of prescription medications. So I think what these slides do is they present a picture for us that consumers, they are ready. They are interested in this. They can easily be brought along -- Therriened anyway, and they can -- they are interested anyway and they can be brought around in how to use it and when you express that to them, they say, yes. I'm interested in doing that on a personal level. What we'll look at now, is the how behind how -- what it is that needs to be in place for them to actually do this. So from a variety of data both qualitative and quantitative, and a variety of different types of questions, we see that interested solution driven and by that what we mean is that they are interested in reducing their risk of specific diseases. We asked other questions like, would you want to do this just to get a genetic profile for yourself and your family, and -- or for posterity sake, would they be interested in doing it for health benefits such as sleep, things like that? Really, where the sent greatest -- and interest is great across the board and that's one of the things I should have mentioned earlier. What we see -- there's not one key demographic where the interest drops below 40%. Unlike other areas or even when we are working industry we go in and we say these are the people to target, high income, it's this, it's that. In this particular case, interest is so broad that there are opportunities for all Americans in terms of products that could be created and they are taking advantage of them. But in particular, they want to focus specifically on reducing diseases. They also don't want -- they are most interested in finding out information if they are certain to experience that. And I know that may be something that's very hard to deliver on, but their concern is that they don't want information if they can't do anything about it. Again, they are still interested in that but they are more interested if you can tell them, they are certain to experience that. And lastly, they want to know that there is some treatment. So tell me, I'm going to -- tell me I'm at risk for this dice ease, I'm -- disease, I'm definitely going to get it and tell me that there's something you can do for me. There are three key things that we believe are critical to long-term and continued acceptance of genomics, and they relate to privacy concerns, emotional consequences, and moral issues. There's very high correlation between these three variables and favorability and interest. The privacy concerns is there at the bottom because it's almost the -- it's sort of the barrier that we have to get through to even be talking with them. It's a big concern for them. And it's something that we need to address and deal with. Right now, they are on the fence about whether or not there's an issue and we'll talk about that. Ethics are at the top because I think they have the potential to topple this whole thing ultimately. We'll talk about that too. So down at the bottom, privacy, I think, what privacy is result in for Americans is limited usage of genomics, what the ethical issues can do is there would be no usage of genomics. We looked at privacy from three issues, as I mentioned earlier, the employers, government, and insurance. Mostly focusing on insurance and government this time around. But from both qualitative and quantitative, these are the dominant issues that come into play when you ask them about things that might limit their usage of genomics. So we'll start with the moral issues. What we're showing here are the top two box in the blue for agreement with a few statements, which I will read to you. And what we have done here is we have thrown in those people in the midpoint, the three on the five-point scale and we are calling those who are the fence sitters. When you look at some of these statements that relate to ethics and moral issues in general, you can see that close to 50% of the public either agree with some of these statements or they are on the fence. They have the potential to be swayed to agree with some of these arguments. Of the first one is some of the recent advances surrounding the use of genetic testing make me uncomfortable from a moral standpoint. So about a quarter agree with that. And 55% total, we include the fence sitters. Meddling with our genes and DNA is trying to play God. Scientists researchers and doctors should stay out of it all together. A little bit less agreement but still we are getting close to 50%, when we consider the fence sitters and lastly genetic testing should be stopped because it will ultimately lead to cloning or altering human genes. 15% agree. 45% fence sitters. And I think that, you know, there are of examples of scientific advances that have either been stopped or at least the public has not been able to benefit from them fully because of moral arguments that have been made and this -- you can see that these are minority of people, potentially a vocal minority, and potentially a large percentage of Americans that could be swayed one way or the other. Talking about that middle part of the pyramid now, the emotional costs. So this -- this really related to the fact that knowing -- knowing my genetic profile is too great a responsibility and it ultimately affects my spouse, and my children's lives. There are close to a quarter of Americans agreeing with that. Slightly more saying it would be depressing to know that I -- or too depressing to mow that I was going to get a disease, particularly if there's nothing I can do about it. This is not as big of an issue as privacy, which we will move on to, but certainly it's something that needs to be addressed. Privacy. So 68% of Americans are telling us they are concerned, the four and five on a five-point scale about how their personal genetic information would be stored and would would have access to that information. A third of all Americans, not just out of that subset, but a third of all Americans are telling us that they are -- that their concern would prevent them from having a genetic test. So it's not a great -- it's not a great level yet but it could potentially be an issue. So trying to understand a little bit about why they are concerned, of course, everything we have talked about already, but specifically here. A third of Americans are saying they consider the results provided by DNA testing to be more sensitive than results provided by other tests. So there's an understanding there that that information is potentially more dangerous for them specifically, if it were to be misused. They also -- about half of Americans say that they are torn on the use of genetics and that gets back to that point I made about the ethics with the fence sitters but here what they are talking about is the potential benefits are incredible but the potential for a misuse is considerable. And misuse predominantly means that people are worried, half the -- half of the Americans are saying worried that their DNA sample may be used for tests other than the ones I have authorized. And we see this in qualitative all the time when we ask them about their concerns. This issue, what will happen to my DNA? Who will have access to it? Who makes sures that no one has access to it? So let's talk about misuse. There's a -- there's not -- not that this is probably a huge surprise, but there's a lot of concern about insurance companies in particular here. We had put in a few inflammatory statements with the goal of really trying to tease out who are those people that are concerned about insurance? And as you can see, even with a statement like "Insurance companies will do everything possible to use my genetic information to deny health coverage" 68% of Americans agreed with that statement. Insurance companies will use the information to deny coverage for drugs people need if the genetic profile indicates a low chance of responding. And 60% say that their increased -- their level of interest would decrease if the information became part of their medical record, and was there for obtain -- and therefore obtainable by insurance companies. However, they do want insurance companies to pay for these tests. 53% say that they would have a -- their interest level would be decreased if the insurance company would not cover the cost of genetic tests. The role of the government in privacy. We put in more questions about this the next time we run this at this point what we do know is that Americans do want the government to protect their privacy. 64% say that their interests would be increased on a personal level if they were assured by law that no one could access their DNA information without their consent. So what we want to do next time is understand, well do they think there are protections in place and they don't think they are adequate or do they not understand there's any protections in place? And what do they think of the current protections in place? We would be happeny to work with someone on this committee if they want to look at those specific questions before we field them. The government should establish specific laws and regulations to protect the privacy of genetic information. 71% are agreeing with that. and the majority were on the fence, otherwise. There was not a lot of people not agreeing with that. But they don't -- similar to what we saw with insurance companies, you know, not trusting them but wanting them to pay for, it we see a similar dynamic here. They don't want the government -- they want the government to protect their privacy but they don't want them to have access to their information. 24% agree, only 24% agree the government should create a national database of DNA information for the future health of all Americans and only 1%, when given a multiple choice that included a variety of different options collected the government as a place. Who besides you should receive a copy of the results besides your copy of the genetic tests. Only 1% thought that the government should have a copy of that. These are some of the key learnings that we took away from the research and some of these frankly have more to do with industry but I think government can certainly get involved in some of these important issues. First of all, we believe that there needs to be an effort to deepen American's understandings of genomics. We know there's a positive correlation between positive and awareness. So far what they have been reading is positive and it's leading them to be interested in this and I think that we know when we draw a connection to how they can actually use it in their lives they continue to be interested. So I think there's a need to deepen their understanding of how they can use genomics. There needs to be a focus on diseases where solutions exist. Because if we -- to my point earlier, we are telling people about diseases that they are likely to get. There's no solution for. We're just unnecessarily upsetting them. Advocating for privacy protections with limbed involvement on both levels. Protections government-related, insurance-related with limited involvement from those parties and concerns of access to information and addressing the moral concerns or framing the moral issues in a way that can help people understand how they can benefit -- how this can benefit them in the long term, and what the ethical benefits are to genomics. And that is it.

Thank you very much. Let's throw it open to questions. Members of the committee? Julio and and Sylvia.

I had a question about the legislation that is in Congress right now. Is this the type of test when you go to the -- just as an example, I went to the clinic last week for my own medical needs. And you have to sign so many forms waiving every kind of like -- you know, this is for access to your records from the insurance company and there's like four or five forms and you hardly understand them. A lot of them under very fine print. Do I really want to see this or not? And is this the kind of law that says it's best, could it just be something that you waive your right for genetic privacy for that purpose and so even there there is not an overall -- you know an overarching principle in the law, it's something that could be waived for -- for third party reimbursement?

I think the same people that design the home residence sale closing forms package are now working in the insurance and health insurance industry. It's not quite that thick. But individuals have a broad authority to waive protections and health insurance companies have many reasons to -- if you seek -- if you seek a health insurance company's reimbursement, they have -- they, under their insurance contract may demand details of why that reimbursement is being sought. So that's part of what health insurance is all about. Decide whether treatment is appropriate or unnecessary, or excessive, whatever the case may be. So typically, that's the basic waiver that the doctor is not supposed to share the information that they have about your health condition with anyone else, unless you allow them to. So you waive it and you say you consented that my insurance company is going to pay for this. I don't see anything in this legislation that will stop or alter that. The health insurance company will till be allowed to have that information. -- still be allowed to have that information. That's an important that I'm glad you asked. Everybody that is in the normal flow of genetic information would still be allowed to have that information. It's not the having of the information that's the problem. It is the misuse of it. If the health insurance companies turns around and increases your premium because something in the genetic information suggests that in five years you have a particular -- you may acquire a particular health condition that you done have now that would be a misuse under the bill. But for them to have the information would not be a misuse under the bill.

My question is for Christy and a comment. Could you tell me a little bit more about how you got the broad range of demographics for the study that was Internet based and then my plea to you is in the presentation you kept saying "Americans, Americans" and it's only 1,010 people, I think survey participants attitudes more than Americans.

Well, the data is projectable -- the data is projectable to the U.S. pop lake-effect snow, with -- population, with a sampling of plus or minus three. But, yes. That's the worst case scenario for sampling error because there's a lot of agreement. In most cases we are looking at a sampling for projectable to the U.S. population, I shouldn't even say Americans but to the U.S. population, of around -- you not worst case it's plus or minus three. In terms of how we collected the data, we used -- there are a variety of Internet panels that exist, web-based panels that exist. For this study we used greenfield online. Probably a lot of you have heard of them. They are one of the largest panel houses. Their pan sell somewhere in ex-- panel is in excess of 7 million Americans. Then generate their sample through a variety of means intercepting people off the web this he do a lot of in-home, calling into the homes. We know that the U.S. at this point, in excess of 80% of the U.S. population has access to the web. And -- and we are very cognizant of the fact that there are specific populations that are not on the web at this point. Or I shouldn't say on the web but underrepresented. Particularly they are over the age of 75, Hispanics and people with less than a high school education. So what we do is we oversample for those populations whether we go out and similar to what we would do with a phone survey or a mail survey or any methodology, we would look at the distribution of responses to the U.S. population, on key variables and census data and we weight the data accordingly. Typically with every web survey that's done, you have to weight to those populations because they are very difficult to get on the web. Now certainly there's a chunk of people that are not on the web in the U.S. and not represented in this study. We find that there are as many, if not more challenges in conducting phone-based research these days. What happened with phone-based research is you are underrepresenting the high income, high education population, that are savvy enough for quick enough to hang up on your predictive dialers or just tell you that they are on the do-not call list. And we just say thank you very much and we're sorry to bother you. So there are challenges with every methodology today. About a third of all the research we do is web based. We have been doing it for ten years. One of the first companies doing it and we are -- we are very rigorous to the -- probably to the upset of all the web panels in terms of how they pull the sample and what we get back in terms of results but your point is well taken.

Kevin?

I have a couple of questions, if I can do one for each. Thank you. First of all, Ms. White, you know, I was very appreciative at the end you were talking about in a sense how you asked the question, can certainly lead your group one way or the other. So to make it even more generic, what do you think would happen to your numbers if you substituted the word "Nanotechnolgy" instead of genetics?

Well, we always try to do qualitative research before qualitatives so we understand which terminology we can use and not use. For example, gene expression was something we wanted to talk about initially and from our qualitative research, it was clear that consumers could not understand that terminology. So we, through qualitative tried to understand what are the terms we can use and not use. I mean we would never use a term like nanotechnolgy. In fact, we try to keep all of our surveys at a sixth grade reading level to the extent possible, because that way we can ensure that all -- all the survey respondents are, in fact, understanding the question. We also -- we not only talk to them beforehand to understand what we can say, but we pretest the survey instruments. So once the actual questions are written, we have a small number of Americans or the superintendent population -- U.S. population go through the survey and if they respond to it we walk through and say every question, what did you think we were asking you there? Was anything confusing, hard to understand and we edit the survey accordingly from that.

Okay. Thank you. And then Mr. Swainy, one thing you mentioned, I think it's going to be very interesting to see how this shakes out. This idea of family history versus genetic information. In a sense, one could use either to get a sort of percentage risk. One could say, you know in predicting the future health of a particular individual or a family group or something like that. So how do you -- I would like to hear more about how this difference is going to attempt to to be clarified in the law.

I guess I'm tempted to say I would be happeny to further chent if I could have your assistance as we mediate this issue. It's obviously semantics. It's conceptual. I just have to go back to our starting point, which is -- which has been drummed into me my Sharon Terry, and that is that this legislation is intended to remove the -- remove the popular apprehension about receiving -- or undergoing genetic testing. It is not intended to change anything else that's good or bad about our healthcare system. It is not intended to prohibit or curtail the ability of physicians to take family histories. And if some lawyer got in there and read the bill, and said, oh, this would prevent taking or using family history, then it's our obligation to try to work -- work around that. We aren't at the end of that process yet, but conceptually that's where we are starting.

Thank you.

I would like to thank you Mr. Swainy for your presentation. We really appreciated the work of the coalition to help move this legislation along. And we would like to -- we hope that you continue why your efforts in back of Ms. Terry. Ms. White, I wanted to say that we are very happy to see that you're reporting on numbers regarding the public concern about both the privacy issues and the misuse of genetic information. I think this is one thing that the committee has -- has long been asking for to get further and further numbers regarding this issue. So I just have a question, based on that then for the committee, our committee itself, is whether we could make use of the numbers, the summary of the mention today to add as an addendum to what we have already sent to the secretary if that would be okay so do. I think it continues to add to the momentum of the need for this type of legislation.

We'd be honored if you did that.

Suzanne and Julio?

A question for Ms. White. You were very deliberative, obviously and scientific in your sampling and my question is, are you designing this as you go into the third year, that you will be looking at trend data?

We did actually trend the data to last year. There were not actually a lot of changes. Actuallyy was a little dishearted to see that awareness did not increase over the past year. We did go back and do a literature search to see if there was an increase it. Seems that way to me. It seems like everyone drives a red car, and I think gee, every time I turn around, I'm hearing a story about this. I was surprised that awareness did not increase. But, in fact, the literature research showed that there was not actually an increase in the number of stories. So -- but everything pretty much stayed flat from last year. And we will definitely trend that data again. And where there are differences, we will start to report on them and hopefully we'll see some of those differences soon.

I had a question for Mr. Swainy, just a clarification. These bills went through the senate last season, right? And now they are in the house. My understanding is that the biggest issue is the working community's fear, ba people can discriminate against them --

That's correct.

I hear the words but I don't understand how that could happen in practice. I mean what genetic background could someone discriminate? You have, say, this particular gene and you are a firemen because of that? I mean, it's so illogically that I can't follow.

Well, many -- many are in your situation. It -- the business community is a bit illogical. They say, of course we don't discriminate. We don't and we would never discriminate on the basis of genetic information. But on the other hand, we are fearful that we -- of this law that would say we could not discriminate. Putting aside that logical inconsistency, the discrimination to the extent that it is there or it is perceived to be there, but -- is less likely to be someone is going to be fired because -- or lose their job, but someone may not be hired because breast cancer has been -- has been a genetic markers is in their medical file for breast cancer. So they might not be hired. Why would they not be hired? Probably because the employer would be concerned that it will cost that insurance plan a lot of money when that person gets ill. So it's -- you know, I hate to be so direct about it, but it's not, generally speaking because employers don't like people who have a certain kind of gene. They are really probably looking at it from a dollars and cents Stanpoint and they don't want to hire somebody that they think is going to cost the health insurance plan a lot of money in some future years. Now, I say "They "In fact, I don't think most employers operate that way. There are some that do, unfortunately. And I suspect -- I've represented employers for a long time in various contexts and there are certainly outliars but there is a -- there is a concern about that, that even if I'm a good employer, for every 1,000 employment decisions I make, I know that 100 of those are going to sue me. And out of those 100, if 10 of those can add genetic discrimination, it will cost me a lot more time and effort to fight that off even though I'm a good employer and only make decisions on the merits. So it's that balance that we are trying to strike.

Joseph?

I want to thank both of you for the presentation. I appreciate that. My question is for Ms. White. In the -- the sort of next to last slide, had you contrasting example there, between the -- what is this -- I didn't -- I see it skin yep, independent of what subject you ask, is should the government work really hard to protect your rights? The answer is yes but should the government, you know, be able to have -- to tell you what to do or whatever? It should be no. The question I have, because it's relevant to the decisions we have to -- we have to make in terms of recommendations, is public engagement, yesterday we had a very -- a lot of conversation on that, about involving the public and helping us or advising us to make decisions in this way. Do you have -- or did you think in your survey, maybe not some of the questions you saw here, willingness of individuals to work with the government, in terms of making recommendations related to issues of privacy, issues of confidentiality, those sorts of things. Did you have any of that in your --

We didn't cover that issue specifically. And I'm -- I'm not sure -- my inclination is I would want them to be involved but you never know.

Well, I ask because of that -- it's going to be critical concern that this committee has. So thank you.

I mean one of the things we do on behalf of business awful the time is create -- we usually call them consumer groups, or consultant groups, advisory groups that are made up of consumers. If we are working with a pharmaceutical company and they are trying to figure out how to communicate something, you might get together a group of consumers, general consumers who are in their specific target audience and have them serve as an advisory board to the marketing team, or the -- or the public communications team.

Well, ma'am, as you understand additional questions can I sort of suggest that you may ask that, because that is going to -- that may be something important and --

Sure.

And I think this committee would be interested to see what the reaction is.

Absolutely.

I have a question for Mr. Swainy. I was wondering if the IBM announcement regarding its policy towards its employees and not using genetic information against them has any -- have you seen any impact on your negotiations or is that rippling through the employer community? Or is it having no impact whatsoever? What's -- what's your view of that?

Well, I think the IBM announcement, which was just a week or so ago is really singular. It is -- it's obviously the right thing to do. I believe it's an informal consultant for IBM, and working through some of the questions that came up internally. It's -- it's a leadership move, and it's certainly our hope that many other U.S. companies and U.S. employers adopt and articulate similar policies. I don't really have any information to respond to your question, because it's too recent. I can tell you that our -- our sponsors, particularly our republican sponsors in the Congress are very happy about the IBM announcement, because it confirms what they are telling their business constituents, that it's really in your best interest to go ahead and support this bill. I think it's going to have an influence, I think it is having an influence. It's too early for me to gauge what the level of it is.

You indicated at the beginning of your talk, Ms. white that you do have data on the preference for delivery method and who is involved in returning genetic data and information. Would it be possible for you to share that information with the committee? I think it's pertinent to the conversations we had yesterday about large population study and if and how such information should be returned to the participants if information is found.

Mm-hmm. Yeah. We do have information on that. and in particular, how involved the physician would be involved in the in addition, and who influences would be and what levels of influence different organizations would have on their adoption, for example, the government, pharmaceutical companies, shealt associations. -- health associations. We have a lot of data on that, which we are happeny to share. We are also getting ready to launch the first physician-based study which will be a compliment to this research, which will look at physician, both general practitioners and specific specialties. Very similar although you might think it is on the face of it. What is their awareness? What are their perceptions? What are the catalysts and barrier to their adoption? And we would get much -- of course more with them into usage and how -- how they would best like to have this delivered so that it's easier for them. We may remove some of the cost issues. That study will be launched early next year as well. So we'll have some data there too.

Thank you both very much for coming today and sharing your perspective and your comments and I hope you don't mind. I'm sure many of us will want to follow up with you. We'll have additional questions and thoughts and we'll enjoy working with you in the future. With that, it's time to turn to our pharmacogenomics session and Emily Winn-Deen, chair of our pharmacogenomics task force will lead that session for us. She will provide an overview of what will be discussed and a review of the task force's works since our June meeting. Emily?

Okay. So what I wanted to do to open this session today, on pharmacogenomics is to just give you a quick overview of what the session is going to be, and what the task force has been doing since we last met as a committee. We have a very broad representation of committee members, as well as ex officios on the task force and I'm not going to read everybody's name here because it's too long of a list. But I want to say that most of these people have been very active participants and that we really have appreciated all of the view points and the inputs that we have received. Today's session is to continue the fact finding on some of the issues that we identified at the June meeting and then to proceed with our work man in terms of trying -- plan in terms of trying to develop a recommendation on what this committee should or shouldn't do in regards to this subject. So in the June meeting, we identified a number of key issues and those are summarized in your briefing books. So I hope everybody had a chance to just quickly review that. And we also identified some areas where we felt we still had some gaps in our factual knowledge where we wanted to get a little bit more education and input. So in the R&D area, there were several areas we wanted to get some more input on and particularly, on the drug diagnostic code development, this is both happening in the industry side as well as the FDA side and we'll hear a little bit about theedA side, as well as some of the pharma perspective. How the whole concept of pharmacogenomics is impacting the way that research is done, the way that evidence is collected on effectiveness and safety. There's issues on particularly on existing drugs of how one might fund pharmacogenomics studies and who would be the right punting source -- funding source for that. And then suft sort of a topic that's out there waiting for us to decide how we want to address it. You know, to what extent does genetics segmenting of the disease or a response to a drug lead to some kind of orphaned disease status or orphaned drug status, potentially. In the infrastructure area, we'll hear a little bit about from FDA about their attempts to create some kind of data standards for pharmacogenomics data, at least, Taz comes into the -- as it comes into the FDA. There's been a lot of very active work in consultation with the pharmaceutical industry to come together on what kinds of data and how it should be submitted. We will hear a little bit about progress and regulation and also a little bit of feedback on how the first pass at trying to implement this is going. One of the key issues, of course, is, you know, it's all nice to have all of this stuff going on at the R&D level, and, you know, cool new science kind of level, but our real goal is to integrate this into clinical practice. And that raises several issues. We talked a lot yesterday about some of these issues. I don't know if I need to go through them again, but, you know, we definitely need to deal with the access and the education issues. The specifics of pharmacogenetics or pharmacogenomics leads to a need to develop some kind of standards for evidence and guidelines on how this data should be used in clinical practice, and inevitably when you get to some type of clinical practice guidance, you lead yourself down the road of if you are not following that guidance what kind of liability does that leave for the physician? There's also, in this area, again, a large number of ELSI issues, and I think they are really in several sort of big lumps. One of which is -- there are going to be some type of stratification that happens unintentional stratification that happens based on social economic status, or, you know, access to insurance, access to physicians, that will instead of improving health, will actually create more health disparities. You, again, with all genetic tests have the issue of informed consent but in this case for a test that's really only going to tell you about your response to the drug, does that informed consent need to be at the same level of both education and consent that you would if you had a -- you know, a very severe and genetic disease that you were talking to someone about? And how do you deal with those sort of new nuances of -- nuances of levels of Bowe education as well as -- both education, as well as informed concept. Of course any time you will be doing genetic analysis on someone, you get into the whole issues, as we just heard about patients concern about having their data in the medical record, and what that might lead to in terms of any disclosures which they feel would violate their privacy or confidentiality, or, you know, whether they might be discriminated against. I think it's less likely you would be discriminated against if you are a poor me tablizer for 2d 6 and likely to come down with Huntingdons but we still have to have these type of protections in place. The issue of race, I think comes into play here, particularly because of the recent approval of the BidiL for the first drug approved for a subsegment of the population, and the question I think arises which -- whether we could do a better job of segmenting the population based on the genetic market as opposed to whey think right now is really -- what I think right now is really more of a surrogate marker and how can we move that ahead? You know we don't want to do any undo harm. We want to make sure that the psycho traumas are minimized and then sort of outside of that, we have the issue of patents and intellectual property. This becomes extremely important when you get into the pharmaceutical side where the pharma industry commits a lot of money to developing a drug, and pretty much feels like they can't make that commitment without some type of intellectual property position. So, you know, how do we deal with that in terms of that influence on access, and availability of healthcare? So the task force had one intervening conference call since the last meeting and the goal of that call was to basical basical -- basically plan today's session. We asked the staff as well to survey the -- basically all the H.HS agencies to identify what ongoing federal efforts are already in place related to pharmacogenomics, and a summary of that survey was also in the briefing books. And then we discussed how we would develop a framework for committee recommendations. We didn't actually try and frame any recommendations at this point. But U.S. sort of talked about the process. So today we're going to have sort of a two-part statement. Before the break, we'll hear an update from FDA, both the diagnostic side and the pharma side to understand a little bit about what's going on, what's new in the ever changing world of FDA. There's been a lot of developments, I think since the last meeting. We're going to take a pause in the pharmacogenomics session to do public comments and then we'll come back with some presentations that address the economic and financial issues surrounding how this is implemented in both drug development and clinical practice and then finally, to finish the session with a talk by wily Burk on ELSI issues, and particularly, on how drug responses in different ethnic groups, you know, what are the ELSI issues surrounding that particular subsegment of the global scope of pharmacogenomics. So in terms of the federal efforts, I think I mentioned this already, the task force requested a review and the goal that was to inform an analysis of basically are there places where there's the same task being done by multiple agencies so you have overlap, or are there areas that we have identified as important for HHS to be working on but no agency is pears to have sort of picked up the ball and run with it? And I think that was, again, designed to help us with our recommendation to H.HS, about how the HHS agencies can best participate in this field. We've developed an outline of a Copp prehencive report and -- comprehensive report and basically this came from all the things that we discussed over the last couple of meetings, as issues. We've got very good feedback, I think, from the public, and all the folks that have had a chance to look at the coverage and the reimbursement report in feeling that it was useful to have both a sort of state-of-the-state summary, the definitions, you know, the basics of the field, and then it provided with us a framework to make some sort of recommendations. And so the task force at least at this point is thinking that we could do something similar in this area. With that, I will close my opening remarks and like to introduce our two speakers from FDA. So I don't know who is up first here on the schedule? Okay, it says Steve gutman who we know and love because of his service to S.PGM and this committee, will give us an update first on the pharmacogenomics pipeline.

Yes, Dr. Rudman and I will play team tag here and I will start and he will finish. I guess my remarks to a certain extent will both support and belie the notion that there's nothing new under the sun. FDA and my work, in particular, has brought to market in the last year the first two I think what you could characterize as pharmacogenomics tests. The first was approved in December of last year. That's the Roche ampli chip and the second was approved over the summer and that was UGT-1a 1 and there were common teams in the review process for both of those products. They were brought two as Class II under a new deknow voluntary classification that was created for metabolic enzymes and the deknow voluntary process provides FDA with increased flexibility when it Encounters a new test of this type, a lack of clear predicates. It does require some either low live risk or some ability to mitigate risk and FDA's assessment was actually somewhat parallel for both of these devices, in that what makes these products stand out as tests is their analytical strength and their clinical imperfections and their clinical imperfections being rather transparent. Both of these submissions were as a result of cedar drug labeling. In the case of Roche, that model product was Tra Terra and the model product was arino pican and the labeling changes made on the model product were advisory or cautionary, not strong required labeling changes. So there were modest changes but there were changes on which we were able to feel comfortable about anchoring the de novo claims process. And in all honesty, we do understand that the clearance of both products based on those models is a little bit like the Titanic approaching the iceberg, that we have looked at the tip and there may be a little bit more than meets the eye. And certainly, if there were specific claims and specific performance parameters that were to be generated on top of either of these assays, we would probably like to be revisited with more submissions. That would be probably be okay for UGT-1a 1 since it does not seem to be an infinite spectrum of possibilities. That might be more problematic for the Roche amplichip, many applications could be theoretically impacted perhaps dosage and selection decisions driven by recommendations by the amplichip. The medications that were best credentialed probably not now but at the time of the clearance were the neuro psychiatric drugs and certainly the use for toxicity seemed very straight forward, but as Emily sort of alluded to there, there are some outstanding education use and reimbursement issues that FDA did not resolve when it clears these two products and we were good for our word in that when we talked about dealing with these new tests we said we put these tests out trying to be as transparent as possible to let people know what we knew about the test, to follow Elliott's admonition to also be sure that we communicated what we didn't know about the test. More globally, as Emily alluded to, there is a very novel from our work ethic point of view, a concept paper which represents a joint environment for the center for drugs, the center for biologics and the center for devices on the co-development of DGICHics -- diagnostics as it might relate to drugs. It is a long document, maybe a little too long of a document but in my view -- I'm biassed, because I'm close to the document, it's rather reasonable limits in a preliminary manner the scientific issues on the plight for analytical validation of this type of test for clinical validation of this type of test, and for lieu sedating the clinical facility for this type of test ax ten makes a very important point that when a diagnostic is used to sleck a drug, the two -- select a drug, the two become inextricably intertwined and that I think most people are cog any zant of the fact that the DGICHic may -- cognizant of the fact that the diagnostic may drive the drug and a more arcane and missed point is that the performance of the drug may drive the performance of the diagnostic. Because the response to the drug behaves for the diagnostic, in the same way that prevalence behaviors for a diagnostic and can radically change the prevalence of a response, and therefore, radically change the predictive value of a positive and negative result. That's very parochial and arcane but a very important point. It is hidden in appendix C of this concept paper and anyone who actually is interested in that should look at appendix C because we actually tried to -- tried to make that simple. the comments -- and we did get wonderful comments, a wide range of comments but probably the most powerful two comments were the comment that the -- the document, which was certainly not intended to be precipitationive and -- prescriptive and suggest that one size fits all but was clearly aimed at an idealized development pattern that was not flexible stuff and recognize in a strong enough manner the need for addressing the non-congruence between the life cycle of drugs and diagnostics. I think that that wasn't the intent of the document, but it's actually the document read, and that's something that needs to be addressed. And the work plan is to take the concept paper which is really quite a preliminary scientific document, and convert it into a draft guidance that would allow for a second round of comments and then the draft guidance would become final guidance. And I -- I won't promise this time, of course because we missed the last one by a couple of months but there was the attempt that we would try to get the draft guidance out by the end of this calendar year. There's some specific guidance, guidance that Dr. Mansfield, who is sitting to my left, originally worked on when we were lucky fluff to have her -- enough to have her. It sorted as a multiplex document and is now more focused on genetic and pharmacogenomics tests. That is more specific, certainly more diagnostic specific and is in the final stage of review and I -- I, again, don't promise it will come out this year but I hope it will come out this year. And we continue to explore changes in guidance, and changes in regulars to clarify what is still a rather messy area, and actually, about two weeks ago -- and I want to thank Carolton Jones who is also in the audience, who actually submitted a frequently asked questions document that for us, and probably for them as well was somewhat unprecedented in that as the document -- and frankly, private entities both in the trades sector and if the professional sector, frequently provide us free advase, offer us guyance documents that we will sometimes laugh at and sometimes, in fact, embrace, longer and then use. But, the effort in the frequently asked questions one in which there was an effort to clarify the world of ASRs and very unique in that admin actually reached out to the laboratory community and tried to get input from the laboratory community as well. So it's a very interesting starting point. They put it in on -- officially on our docket, it's public. We do plan to steal the document, probably to launder it and to try to use it as a basis for draft guidance to help clarify -- clarify this colorful world. And it's a nuance document that's missing some pieces. We might plug in those missing pieces or try to at least. And there continues to be a background of confusion, and either inadvertent or misuse of offering tests some of the tests are actually going in ways I think even may not not have been predicted. We are interested in inning in a -- continuing in a modest way to explore whether there are incremental changes in the ASR exemption. There is as I suspect everyone in this room knows a new face to the FDA administration. We have a new acting commissioner. We have a new head of general counsel, and we have a variety of new leaders in the deputy position in the park lawn central commissions office. As far as I know they have not been asked the lit Tuesday test about abortion or ASRs. So we don't know how any of this will come out. Deborah Wilson is sitting to the left of me and was kind enough -- she's with the office of compliance and we operate as one unit in the CDRH. She took the lead in an ad hoc group to look at the colorful of direct to consumer testing of genetic tests and the notion was that FDA regulation of particularly the home brews but even the clear and the approved devices themselveses is perhaps not the strongest regulars that have -- regs that have ever been written in the history of U.S. regulatory authority and we'll try to leverage off of FTC and so we did put out a call for web sites and good debt interesting web sites where there was direct-to-consumer testing but we found this task to be much more challenging than we would have guessed. We were looking for two criteria. It had to be outrage withous but being -- jut rageous. But being outrageous was not enough, but to also be harmful because that's how we interpret F.TC's charge of being outrageous and fraudulent or weird is not tough. -- sufficient. We identified a couple of candidates, one in particular that we thought would be interesting to test the water we had argued about how to craft the language, so it would be friendly to a person who might not be a scientist. We went through several iterations when we noticed that either -- either because of the business realities that were in the background or maybe because FDA kept looking at the site, the site went out of business and put up a for sale sign. And then at least one or two other sites that we were actively interested in started to change the tone of their advertisements and in ways that made it seem to pus that they were becoming more -- to us that were becoming more alousetive. The work group or frankly the audience as a whole is if you have a favorite test that you think is outrageous but also dangerous, this committee does exist. We are anxious to give them a little bit of extra work and I -- I'm my staff and I were on an FDA computer and anonymous computer and we are inputting the right "Buzz" words. Maybe we don't foe DNA and test, or home use. We just couldn't find anything that piqued our interest. So I was surprised and maybe we're just looking in the wrong places. So this is an all points bulletin for help. Two final notes. Diagnostic industry from my perspective has either -- I think probably a little bit rightfully sensed a disenfranchisement, the dugs makes a lot of money and drugs is bigger than we are. And I had talked to earlier meeting about our efforts to set up some kind of work group that might represent the old pharmacogenomics round table and be a little bit more distinctive IDD in orientation. I do think that this committee and certainly us in drugs and diagnostics appreciate a match maker role in here in trying to bring culturally different companies together and Joe Hacka has set up a working group in our office and is having active negotiations with Carolyn Jones and others in the industry and the notion was -- I what have hoped that -- I would have hoped that we would have had this done. We hope to explore whether there are opportunities for industry and FDA in a very diagnostic-specific way to -- to describe the unique challenges that are -- that are on the table. And then, it is worth noting, and I apologize, I missed at least part of yesterday's activities, that there really are fertile areas of development and ongoing work that CDC has at least two initiatives e gap and also Joe bloom's work in the ear of trying to get material quality controls. They have an activity where they are looking at standarded Proteomics -- standards for Proteomicstesting. There is interest in trying to try to step and look at how outrageous evidence-based medicine and outcomes. Allen?

Okay. [ Inaudible ] okay.

You can do it from wherever you are comfortable if you are comfortable where you are.

I will go through these very quickly. This is actually a subset of the slides and we're going to go through it very quickly and part of this is you will have the documentation here so you can go back and review it. Sort off with the FDA's mission. It's really to protect and vad advance public health, to make medicines and foods more effective and safe and more affordable. This is in the critical path initiative. Next. Next slide. Thank you. Towards that end there was a publication, a stagnation, challenges and opportunities on the critical path that was issued under Dr. McClellan and Dr. Woodcock. Okay, this kind of give you the overarching description of really this initiative. In addition to that, there was an NCI and Pas -- FDA joint program developed by Dr. McClellan and others. Towards that end one was to develop markers and biomarkers for clinical development and evaluating new cancer medicines. Next. This is in November of 2003, a draft guidance came out, genomic data submissions which created a whole new paradigm for voluntary genomic submissions. The FDA had found really we had not received a whole lot of genomic information from the industry; although we knew it was going on. We there internal information, indicating this was a lot of research going on but we really had no access to it. The companies were not submitting it as I.NDs so we knew that there was something coming up and the question is how to get that information and work with it and figure out how to deal with it for the future. Towards that end we created this whole process, and really formed around the interdisciplinary form of the pharmacogenomics research group. And the goal is to get companies to commit to themic data -- genomic data submissions and meet with the companies and find out what they are doing and why. Process it and then there were three major goals. One was kind of public feedback, workshops with industry, with the public, NIH, and so on. Data and knowledge, and leading to guidances and policies. How do you -- you know what do you do once you get the information? And finally education. Okay? Both internal to the FDA, because there's not a whole lot of mass there in terms of education in there and also external. Next slide. Okay, and this is all under the construct of the critical path, really. The RPNG, I will use that otherwise we'll be here more than 20 minutes was created and represents the representatives from the entire FDA. Along with that, there's a group called the pharmacogenomics working group, which is actually just located in the office of clinical pharmacology and biopharmaceutics. They have numerous activities. I will just briefly go through them. One is the actually review of genomic data submissions of which we have a large number now. I will get into that a little bit later, required submissions, consults, policy development, which Dr. Gutman just mentioned. Education both interriblal and external, research there's a create ore or biomarker validation and clinical trial protocols and how to design them analysis and labeling of pharmacogenomics, and there's a research grant out there, I am the principal investigator for that. Looking into putting in all of this information. There are over 50 labels out there containing pharmacogenomics information, but some of it is more useful than others and we are really trying to put this to go into usable database. Finally there's the information technology area. The FDA is developing software. We're also looking at database development how many do you get to use this? This gets to the question of knowledge management. Next slide. This is a slide of the -- the IPRG organization, you will notice that at the top it's really driven Ben the FDA and actually -- driven by the FDA, and actually that's the main message. It covers all the centers that are relevant NCTR, office of commissioner, cedars, Seaver and so on. Okay. And then there's a whole organizational chart. The chair of the IPRG is Dr. Felix Drew. They have the center delegate at a high level. And then there's a whole area of activities associated around that. Next slide. I will not spend a whole lot of time. That's to say there is a whole process involved in this, which you could find on the web site. Okay. To date, we have received the final -- the guidance was actually finalized in March of 2005, and we have received 23 -- actually, 24 today. We see it's another one yesterday. Voluntary genomic data submission requests and have scheduled a meeting already. These include two joint FDA, Europe, evaluation meetings where we hold joint meetings and we'll talk about that in a few minutes including multiple different drugs and follow-up submissions on the initial study. So companies are coming back to us with follow-up information, and coming back to us with the different types of the submissions which is actually very helpful and very useful. And really tells us that we're doing something right. We expected to see a lot of cancer drugs. Okay we saw that. But as you can see, we have seen a lot of genomic information, a lot of different areas including different types of cancer, Alzheimer's, hypertension, hypoglee seemia, rheumatoid arthritis we have also seen a lot of discussion, a lot of different areas and I think this is probably not as clear. Okay? One is the whole question of the biomarker development. But then we get into the questions of genotype devic is, microrays analysis, analysis software, assumptions under them, metabolic pathways and the clinical declines and clinical protocols ever this is a fairly broad, fairly comprehensive discussion. And it covers a lot of areas and it's actually changing the way the FDA looks at how we do things. Next slide. A little bit about the Harmonization. Most of the large pharma companies are now global. Okay? And this is their -- this is really becoming more and more important and do we have this joint effort and consistency among both Europe and the United States? And eventually it will be Europe. I think this initial meetings now scheduled for ICH, okay? May 17th, we had our first joint FDA EME, meaning it was by video conference. I obviously would rather be in London than the Washington area. And as you might I mag thin's a lot of preparation -- imagine there's a lot of preen in this. Including the list of questions that was really being asked B.a lot of interaction before the meeting, and putting in-depth scientific evaluation, and response to questions and information, and dialogue between the FDA and EME. We obviously operate under different regulatory environments. Okay. We were very fortunate the sponsor provided excellent presentation. It was really good discussion on a lot of different issues including regular -- registries. One the novel things is we were issuing joint minutes, okay on these voluntary genomic submissions. Okay? Just as important, maybe more important is that the FDA and the EME eval U of Aated with only minor differences -- evaluated only with minor differences. It turns out, at least in this case, we were fairly consist enter. We both had to adjust our usual mat, EMEA was that parentally one of the first times they got issued written comments to the company. Very positive response, and clearly it's the first step towards Harmonizing. I should mention that there are three more meetings scheduled, being scheduled. One of them definitely in December and then two more in next year, 2006. So this is fairly positive development. Another area is what do we do with all of this information? Part of it is obviously we issue guidances and concept papers. This is the list. I will no the go through this. You will find this on the genomic web page, okay. There's a lot more information. I will not give you that information where to locate that later on. Next slide. This is not just the FDA, and reallies this is about building a process, with the industry and academia and NIH. There have been a number of different workshops, public workshops, it started off in 2002 with the workshop number one, that's what they named it. Okay? Creative. What am I going to say? Then there was a draft guidance on genomic data submissions. This has been followed by a whole slew of other workshops including the FDA, EIA, the pharmacogenomics workshop number two. This is where we discussed the draft genomic guidance and actually got a lot of input and this is done with the pharma, bioand pharmacogenomics, and others. And 2004, there was a meeting on the co-development on drugs and biologics and devices and I met with the cosponsor along with the MDAM A. It was a docket to get menes and we received quite a number of them. And actually this fed into the concept paper and ultimately into the guidance. This year, there was a third workshop in the series optimizing the benefit, risks of drug vent -- development and therapy as you see coming along more and more, this is coming from the research side into the regulatory side. Okay? And, again it was with pharma, bioand a lot of others. And finally, this October 7th and 6th, we had another workshop on the application and the validation of genomic biomarkets, for the use in drug development and regulatory submissions. I will talk about this -- next slide. Why don't we just go to this. The next slide. Okay. The handout from the DIA. Okay, these some are of the topics on keynote addresses. I will not go through this. I only have a minute or two to actually cover this, except to say it was well attended. Over 200 people, and, you know, I have to -- people, I today woodcock and others talking about really where this is going and ultimately, pharmacogenomics is kind of part of the whole biomark err oer question. Which -- biomarker question. And Proteomics is the next in the line, imaging and so on and so forth. How do you do this? How do you actually get to this stage where you can actually implement in this what do you need to do to validate these? And if you look at the topics that were covered, safety of biomarkers, efficacy biomarkers and I should add that more and more we are getting toxicogenomic submissions into the process, that is the one we received yesterday. We are seeing more and more toxicogenomic data. Updates on the HO 7s. It turns out that will be very important and the use of the word "Validation" is apparently somewhat controversial. Electronic submission working groups, standards for developing safety and efficacy of biomarkers, validation of these, introduction of the -- how do you actually introduce these into the drug development process? What are the regulatory implications TV? Developing and validating genomic markers and you can read this. Next slide. This is kind of an overview, and I like this slide, only -- it put it's in a context, you know rather than just listing all of these. But where we are and how we are doing, and the big arrow is really, actually been used in a number of different documents. It goes from basic research to FDA approval and all the different steps in between, and there are a number of different versions of this. But if you look in the bottom, these little -- these arrows, actually it turns out this is what we're really talking B.when you start doing the validation, what does it consist of? How do you do it? What are the criteria? The preclinical feasibility? The clinical validation and also the clinical utility? The fact that you have a test doesn't necessarily mean it will be useful for public health reasons. Next slide. At this meeting, actually, Chris Watson, I would like -- list Webster, I mean from pharma, and millennium presented this slide. I thank him for this. And he announced that pharma is going to present a consortia on biomarkers. You can read it for yourself, really and the goal is to develop, expedite biomarker development. Okay? They are expecting, hoping fully that the FDA will participate as well as, a lot of other organizations. Okay. We are looking forward to this. We don't know a whole lot more than what you see there but obviously it's an important outcome of the workshop. Next. Fine Raleigh concluding remarks. -- finally is concluding remarks. The FDA has been quite successful and the FDA has been in the regulatory lead including guidance development, analysis of pharmacogenomics data international collaboration and obviously the workshops. The VGE estimations I provided FDA with a wealth of significant NeoRxic data and informationD genomic data and technical areas which would otherwise be unavailable. So -- and in that sense, the guidance was really successful. The pharmacogenomics research needs to be seen in the context of bionarker development, and validation, as well as disease management to expedite the approval of new drugs and indications. It's not just about finding the drug. It's actually getting it to the public and it has to go lieu the FDA. Really -- through the FDA, really to do that. And needs to provide this data in a manner that FDA can readily analyze and expedites the review. This is about not just finding the biomarker or increasing public health but how quickly do you do this? How quickly do we get drugs to the market? Okay? FDA does not develop drugs, or pharmacogenomics tests but they can encourage them to be developed and finally, I would encourage this committee could help as a group by recommending the foration of a task force for pharmacogenomics assays. Thank you.

I want to thank you both of you very much for your very --

And finally there was the Joe sight.

The speed, overview of what's going on at FDA, we obviously should have allocated a little more time for the, quote, update. I had one question for you, Dr. Rudman. When you did this joint program with EMEA, was that at the request of a particular drug sponsor that you do that, or was this something where you sort of got to go and said we ought to pick a project and see if this would work?

This is actually -- we have an interest in this and so did the company and so we found through one of these interactions we found that we there a common interest and that's really how it came about.

Okay.

They were doing biomarker development and we were interested in this, and so actually the Kate crater came out.

Is there a mechanism for companies to do than "A regular base ?is are you working towards that.

They can submit them to -- we have a web site. We email addresses and people call us all the time, and ask these things. And so we're very receptive to them. We do have limited resources. We can't do anything in numbers quite clearly but, you know, we are very interested in continuing on with this process.

Okay. We have maybe 10 minutes for Q&A. So I've got Julio, and James and Kevin and Ed.

Hi. Just following on one of the things I got at the end on that, the need for the national standards for the testing which I think is crucial, because we discussed a lot both here and in different context, what do you do with the health result and the privacy and that, but the issue -- is it that acCoorate to begin with -- accurate to begin with, which is a very crucial issue. And we were discussing this informally before the meeting began. Not only is there no national guidelines or licensing board for that, but if such an entity were to exist, where would it belong? Is it in the purview of the FDA, of the CDC? Of -- who would be responsible for monitoring, for issuing qualitative control and ensuring that it meets that kind a standard?

That's a very good question and actually a very complicated one. Because you are really asking a number of different questions, I think. I will try to address them. One, it's not just about the validation, okay, and actually that's -- that's part of the issue. There are numerous types of validation. The text validated in vitro, are the labs performing it correctly? I mean this is a number of questions D. it work in humans? Is it clinically validated? And finally does it have clinical utility? I mean the fact that the test work, doesn't mean that the public health would necessarily improve by it. So these are all questions, okay, and so you have to kind of parse these out. Okay? The first step, before you can even get to enforcing these and there are processes in place, such as C.LEA is to define what you mean and that's what we need to start doing get some consistency on this and then I think we can assign responsibilities accordingly.

Okay. Let me just augment that response because at least for the diagnostic industry, there really is a premier standards crafting group that is C.LSI, the clinical laboratory standards institute, and so the most -- if you were able to subsume drug issues the most ladies and gentlemenical place to turn to for standards in this country, at least, is CLSI. CLSI is also the executive secretariate for the international standards group IC 02-12 12. So you have the capacity to kill two levels of standards with one stone in turning to CLSI. Again, they are focused largely on the DGICHic issues and so there's no reason they couldn't be a little more inclusive. They wouldn't start creating pharmacology standards.

I'm naive to what the FDA's purview is, and all of its manifestations. Do you have anything to do with guidance of how reports are -- results are reported? I Contaked Roch and I asked for a -- I contacted Roche and I Canada for a sample and it was incomprehensible and aim a geneticist and I do farm could genetics. And I -- pharmacogenetics and I worry about access to this information in an understandablyay to clinicians. I'm glad you brought up twice the issue of a difference between clinical validity and clinical utility and it goes beyond public health into the individual as well. And unless we have understandable types of reports, then the utility and the validity can really get blurred. I was just wondering if theedA has any role in that.

The FDA in general does not. Does not regular -- the FDA lass any role in that.

The FDA in general does not. We may be concerned about some aspects about the way the information would be reported out that will try to push the limits on it, but we don't have direct authority, unfortunately there's no one from the CLIA program. Does Judy Yost have the capacity to tell a lab that -- I mean there allegation the expectation -- CLIA has pre and some post analytical requirements. I -- I -- I actually don't know, but we -- we can certainly take that back -- back and find out what their authority over that -- I assume they have some. I don't know how strong it is.

I'm not sure how much they have authority, but I can tell you one thing, CDC, the C.LIA group that works with the CMS program, has looked at genetic test reporting in general and the variations of that not necessarily in pharmacogenomics, but in other areas. And there's quite a bit of variation in Ienetic test reports. That was a project that just got finished. I think the SACGT, I guess our parent group here, kind of took on these issues for at least a couple of years, tried very hard to sort of develop an overarching package for the oversight of genetic testing and the transition from practice to -- from research to practice. They came up with very thoughtful comments about the feed for at least three or four processes, one as an FDA-driven process which I think the FDA has kind of taken over the last couple of years, and struggled through it. The second is a CLIA-driven process, which includes a development of genetic testing, subspecialties but, again, these have different arms and the third is sort of a more perfunkory, what we call a public health-related effort for, you know, developing the kind of data that needed from a non-regulatory process and this is something that the CDC and others have worked on that led to the egap initiative that you heard about yesterday. So we are all moving in some directions, but I think there are just too many gaps right now in that process and I think pharmacogenomics is uncovering some of these gaps and this committee could sort of take it on again.

So Kevin passed his turn and Agnes is going to be the last question before the break.

And there's a good segue, you think some of what Wayne just said would cover some of the issues that I wanted to just bring up regarding the education and was really thank you very much, because it was exciting to hear all the numerous initiatives of the FDA in moving forward in this area of pharmacogenomics. But yesterday we did hear from some of the speakers regarding their concerns about launching the large population study, suggesting that one of the major barriers was the lack of genetic literacy from the public, and agencies such as the FDA, as well as in the health professional community itself. So I just wanted to say that I think that some of your response shows that there's a definite move in that area, sort of lessens some of my own concerns about that. But at our last meeting when we also had some updates on pharmacogenomics, there was one of the presenters talked about the use of TPMT testing for children who would be treated with 6mercapta purine and said there was a lack of that testing because the clinicians even with the labeling and things like that were not sure how to use it or maybe other concerns about liability. So my -- I know in your slide, Dr. Rudman, you had -- you had education initiatives both external and internal. And if you comment a little bit more both on what education initiatives are going within the FDA and possibly for the community, healthcare provider community can make use of some of these tests that would then come out?

This is a very good point. Internally, we have -- actually we have started a whole series, I think it's three or four of them now of internal FDA seminars and training sessions. We have also brought into the process, as part of this whole process, not only is there formal training in the FDA, in terms of teaching people about genomics, about different types of software and so on and so forth, it's numerous one of these types of formal trainings. But we are bringing the reviewers and their division directors and management, basically into these meetings more and more to try to get them to understand the issues and actually get their feedback on some of these issues. So internally, we have -- we have internal web sites. We have training programs. We presented, I don't know how many times already, okay, to different divisions and therapeutic areas to try to do this. Aim sure, you know, CDRH has a similar type of program in addition to this. Externally, a lot of our effort has been to really two mechanisms currently and we are -- because of resource limations, actually. One of them is workshops. We do a whole slew of workshops and we try to invite people, and try to get participation. We saw the list up there but a large part of it is also through our web site, which I had the web site address up there, to try and get people to really understand these issues. Some of these are fairly eclectic, and difficult to understand, frankly, as a layman. In terms of going outside, we have there preliminary discussions with some people, universities, about moving forward on this. We have also talked to, I think it was the American association of clinical science, okay, about setting up a program for these. We don't actually have the resources to go physician to physician, obviously to do that. But we do recognize the need and we are trying to move forward on those issues.

Can I ask a follow-up?

Sure.

Do you think that clear labeling and package inserts would be helpful to physicians? I think right now when you just say by the way, this is me tab rised by this en-- metabolized by this enzyme, it does not help the physician.

You actually note a very good point. It's a difficult issue. In some ways it is very sim. -- simple. We start off with this grant, to actually categorize this. When I tell people there are about 60 or more labels out there with genomic information, a lot of people are actually surprised at this. Okay? There's a lot of them. Most of it is the S.IPPA we are finding the labels that are more recent are more informative than recent ones. Some of these are fairly -- it's not always clear. For instance, the TPST issue has been brought before an advisory committee and that's part of how the outcome was generated okay? In terms of how physicians see this and how the FDA sees this, we have different proposals out there and so a lot of this is bringing these before these advisory committees to get their input. -- their input. Way?

Well, it's a glass half full or the glass half empty to realize that I'm a clinical pathologist and not a geneticists and I'm to survey what's going on in lab text or if you look at the recent activities at CDC's initial initiated with the institute for quality and laboratory medicine, the average physician doesn't know how to order a prothomban ittinE. The issue of egg no arance in laboratory medicine and an appropriate use of laboratory tests transcends the genetics issue. It's a core issue. If you look at medical school curriculums and clinical pathology, it is frightening, horrifying, disgraceful. If you, as a group -- I just don't think you can underestimate how -- I doubt the average practicing physician at a fine academic medical center would have understood my hurried explanation of predictive value because they wouldn't know a predictive value if it bit them in the nose. So this is an area that is replete with opportunities for improvement, genomics, or genetic testing would be a great way to start. I would certainly hope if you made recommendations or fixed this problem, you wouldn't top there. -- stop there.

Okay. I think that's probably a good time to pause for our break. And, you know, if people feel the need to talk further with Steven and Allen, they are there and before they escape out the door, we can probably get a few more questions answered. We'll take a ten-minute break and be back at 10 till 11. [ Break ]

I'd ask that everyone be seated, please, so we can begin the next section.

All right. Let's get started, folks. One of our critical functions here is to serve as a public forum for deliberations on the broad range of human health and societal issues raised by the development and use of genetic technologies so we greatly value the input we receive from the public and as you know we set aside each time some time to hear from the public. Today we will be hearing from Jeanne Jenkins the international society of genetics.

Good morning. I'm Jeanne Jenkins and I value the opportunity to be able to present to you the following public testimony on behalf of the international society of genetics. I also want to inform you about an effort of several federal agencies including the human genome research inns institute, health resource services information and in collaboration with the American nurses association and ISONG. We have provided the written testimony and at some points I may refer you to that document. As you know, genetic and genomic science is redefining and understanding the continues yum of human health and illness. Therefore recognition of genomics is essential. Options of care will increasingly include genetic and genomic information along with the continue wum of care for all persons including prevention, screening, diagnostics, prognostics, monitoring of treatment effectiveness. The clinical application of this knowledge has major implications for the nursing profession. And because essentially all diseases and conditions have a genetic or genomic con poent, recipients of nursing care will be from any stage of life and cared for in varied clinical settings. You will hear of a number of different examples where pharmacogenomics information will be used in determining how individuals respond to drug treatment. I won't outline some of those for you today because I know some of the presenters will be discussing those. But specific medications based on an individual's genotype has major implications for nurses and they must be able to implicate those into health care delivery so that all individuals receive equal access to genetic and genomic health care. The code of earths sdwoped by the international council for nurses and the American nurses association state that nurses have a shared responsibility with other health professionals and society to ensure initiation and promotion of community, national and international efforts to meet the health and social needs of the public. This includes the right to seek and rereceive genomic health care that is nondiscriminatory, confidential and private. As providers nurses must be able to advocate for and fulfill an essential role. In the say seszment, the policy development and regardless of genetic and genomic literacy. Socio economic or earth no cultural background. basic and advance practice levels. Genetic nursing is the protection promotion and optization of health and abilities prevention of illness and injury, alleviation of suffering through the diagnosis of human response, and advocacy in the care of genomic health of individuals, families, communities, and population. Nurses can fulfill these responsibilities by the identification of genetic and genomic risk factors, nursing interventions, information provision, services referral and promotional health behaviors to enhance the health and well being of the individual or the family seeking care. To fulfill the right of the public to access genomic health care without fear of discrimination the nurse's responsibility extend to the development with partnerships of stake holders such as patients, health care providers insurers government officials and legislators and listing of such outcomes of such partnerships are included in the written comments. The public will increasingly expect that the registered nurse will use genetic and genomic information and technology when providing care and these expectations have direct implications for the RN preparatory curricula as well as for the 2.5 million practicing nurses. The rate of progress for applying a genomic approach through the continuum of care depends not only on the tech any logical advances we've been discussing was also in nursing expertise. An expert emphasized the importance of integrating genetics content into nursing curricula in order to provide an adequately prepared nursing work force not only for today but also for the future and to assume this role with persons, families, communities and populations throughout the life span, the registered nurse will need to demonstrate proficiency with incorporating genetic and genomic information into their practice. An examples of such implications of practice are provided also in the written comments. Under tab 3 of your handouts there is indication of the clab ra tiff meeting that I had mentioned at the beginning which was an effort of several federal agencies including NHGRI and the American nurses association and ISONG where we provided an opportunity together for key nursing stake holders to attend the meeting in September 2005. This meeting was held in Washington, D.C. and representatives of key nursing organizations came to consensus which is a major advance in the profession of nursing. On a document whose purpose is to define essential genetic and genomic competencies for all registered nurses. This document is intended to guide nurse educators in the design and implementation of learning experiences that helps students, learners, practicing nurses achieve genetic and genomic competency. These compties are not to replace existing standards of practice but are intended to incorporate the genetic and genomic perspective into all nursing practice and education. The goal is to prepare the work force to deliver competent genetic and genomic focused nursing care and the essential compties -- including the professional practice do main, nursing assessment, identification, referral, provision of education, care and support. And the listing of these competencies are listed in your document as well. In closing, ISONG views genetics and genomics as integral to all nursing practice. Nurses will increasingly use the information such as pharmacogenomics and genomics to create, administer, individualize treatment care plans. All nurses care in the responsibility to ensure that all individuals have equal access to this kind of based health care and that it is nondiscriminatory. Thank you so much the opportunity to share with you the progress that's being shared in the education of nurses as being key to be able to translate scientific innovations that we are witnessing into health gains for all.

Thank you very much for your comments. Appreciate it. Emily?

Okay. So now, pharmacogenomics part II. It's my pleasure to introduce our first speaker of the next session. Met calf is -- Tom has the unique perspective of also having worked on the diagnostics side of the Roche organization and he's going to talk to us a little bit about how Roche as a exemplar forma company is using biomarkers in their drug development process. Tom, you can sit there if you want or would you rather -- okay.

You're actually much less intimidating if you sit down. I should disclose that I used to work for Tom and we were quite a pair when we went out in public together.

There we go. Sorry if it's intimidating for you but this is usual for me. So thank you for the -- thank you to the committee for the opportunity to present you on this topic. As Emily said, my comments today are based prince my on Roche's experiences in this field, but we believe that much of our experience is relevant to the pharmaceutical and diagnostics industry in general. What I'm particularly going to talk about is some of the process aspects and the feasibility of integrating pharmacogenomics and biomarkers into drug development from an economic perspective. I think this broad definition of biomarkers as an objective measure of normal biological processes pathogenic processes, pharmacologic processes or responses to prevebtsive or other health care interventions is generally broadly expected, broadly accepted and obviously as a part of that biomarkers can increase our understanding of drug metabolism, drug action, efficacy, safety, facilitate therapy response prediction and expand the molecular definition of disease and inform of the course of disease progression. Obviously this broad definition includes all diagnostic tests, imaging technologies and other subjective measures that occur in a person's health status and obviously all pharmacodiagnostic tests. My remarks are going to be on novel markers either newly discovered markers were markers being validated for further applications. So just a little bit about what's new, what's changing. Obviously genetics genomics protoyoemics, modern imaging techniques, under the current technology allow us to measure many more markers than we could before. There is an improved understanding of interventions, signally pathways and mechanism of to bes sisty and action and this will allow us to make more sense of the biomarker data that we're looking at and by statistics and allowing us to collect, store and interpret this data more effectively. Nonetheless, there is a tidal wave of data which we're generating and it's -- it takes a lot of time and energy to interpret this effectively. So what -- again, how is this affecting pharmaceutical development and particularly the integration of biomarkers into drug development these new data are allowing us to make considerable decisions in late development and and research and this is about which projects to move forward, what the comparative -- I suppose the comparative prospects of different molecules in the pipeline, but this is an essentially an evolution of what we currently do with pharmacodynamic and pharmacogenetic markets. There are yet as few validated surrogate markers which allow us to run considerably short at trials. And the -- I think quite rightly the status of a surrogate marker for a clinical end point is set very high and there's a lot of evidence needed for us to -- to reach the status of a validated surrogate and this will mean that it's going to take some time till we get considerably more surrogate markers and are able therefore to run shorter trials. And they're also as yet, very few highly informative response markers which allow us to run smaller trials or enrich our trials or potential responders. So we see this as being a steady evolution of the drug development process rather than a revolution. This is particularly because the first part is an evolution of the current paradigm and we have few examples in the -- of predicted markers in the surrogate markers. The various utilities that we see for biomarkers and novel genomic genetic markers and drug development are pharmacodynamic markers which confirm biological activity of our drugs allow us to make early decisions on progressing molecules and make optimumization of dosing more efficient. Prognostic markers which correlate with disease outcome and these enable or improve our ability to design informative trials and to interpret these trials confidently. Disease specific markers, which correlate well with the presence or absence of a disease in some cases these can be used to identify disease subtypes that they're more amenable to one therapeutic intervention than another and can be used to enrich our trials for those most likely to respond. I think this is a very useful utility which we are beginning to use more, but as you can see, this is not the directly related to the action or efficacy or safety of a particular drug. These are disease specific markers. And the last category, predicted markers, such as HER2 over expression in breast cancer which correlate strongly with the activity of our drugs. And depending on how strong this is, we can include this in our -- in our trial design. So if you -- this can lead us to come up with one possible classification of involve biomarkers with disease markers, form cological markers and predictable pharmacodiagnostics set out like this obviously there's an overlap. Some of these disease markers aren't predictive. Pharmacodiagnostics as we've heard too, but many are not and I think it's very important that we begin to tease apart these -- these different utilities and whether or not one disease marker can be used as a predictor pharmacodiagnostic. And in our experience in many cases these disease biomarkers lack the specificity to be used as predictive pharmacodiagnostics. So I'm going to focus most of what I talk about about predictive pharmacodiagnostics and especially with relation to the -- the impacts on -- on the economics of drug development. Pharmacodynamic and prognostic tests tend to increase the value of -- or if you integrate them into drug development they tend to increase the value of that drug development principally because of the size of the market that you're addressing is not really affected. Revenues do not decrease and they may increase because we're in -- we can improve dosing and dose scheduling and in investments and markers generally offset by improved decision making trial design, reduced attrition and therefore are beneficial in general. So I don't think pharma companies have any issues at all about -- including this work or in these sorts of markers and their drug development work because it's generally beneficial. The value impact of predicted markers is left there. They may reduce the size of the market. This reduction in the size of market may be offset by improvements in market penetration increased average duration of therapy and potentially in pricing. I think the pricing point is one in which we need to look at because obviously this is only relevant if this is done before the -- the pricing is set within the market. It may also improve the competitive position of the -- of the drug, but I think the -- the most critical thing about this is it depends upon exactly what we're looking at and requires careful case by case analysis. And I'll take you through some of the analytical components later on to show you the complexity of it. Now, looking particularly at the application of markers for response prediction, this is a -- a prototypic concept for the application of resource of markers and this -- it is schematic but I think it teaches us some important lessons. The first one has to do if one is going to include response markers in -- in a drug development program is to have a reliable understanding of first of all the biology of the marker and then also of the test that one is using to -- to test for that marker. So we generally have to spend considerable time in early drug development and biomarker discovery and then work up those biomarkers into reliable tests so invest time in biomarker test development and validation and then as we start introducing the drug into man to collect samples and to store those samples. And what's the most informative time in early development is going to be phase II because in early clinical development there's a lack of conclusions that one can draw and one generally begins to be able to draw conclusions about whether or not a drug is working or not at the end of phase II. So if one has -- knows which one biomarker one wants to look at, has been able to develop a test, then it's possible to do a retrospective analysis of these biomarkers on samples collected in phase II and correlate these in response or lack there of and perhaps with safety as well. And one possible thing one could do is prospectively recruit the patients using biomarkers found in phase II in your phase III program but in order to do this you'd have to have a very informative and reliable biomarker and in general, we find that it's -- it's very difficult to find biomarkers which are informative enough to have us take the risk of recruiting for a phase III trial based on the response marker. And and alternative might be to balance the various arms of a trial to make sure that both arms are equally populated with patients who have a good chance of response using this marker. So this is -- this is a -- I think one way of doing things, but it's currently we rarely run into the situation where we can do this at the moment. And this means that we are therefore not essentially not changing how we develop drugs. We are tracking more things but we are not changing the -- our trial protocols as yet. Or in very few cases. And I think the one drug where this -- this paradigm was followed was with Herceptin and this is the current diagnostic paradigm that's used to determine whether or not a patient is eligible for Herceptin therapy. There's a two particular tests which is used, one is a history chemistry which is basically testing a patient 's tissue and then if one has a result which needs further interpretation, one applies what's called a FISH test to that patient as well. Now, I think that worked with Herceptin therapy and one of the reasons that worked was because the marker was informative enough. But now I'm going to sort of come on to some questions which arise once you apply this paradigm broadly. First is, what's an acceptable response rate for a novel drug and when should one think about applying a stratification with a response marker? And if you go towards the end of this scale where you have a very low response rate, the first thing you start to question is whether or not you have a viable drug. And it's only when you move up a -- perhaps above the 10% in response rate that you think maybe you could apply a response marker in your clinical development. And when you get above the -- the 50, 60, 70% in terms of response, this is an excellent response array for a novel drug. Again, you start beginning to think whether or not it makes sense to apply a response marker. So clearly the sum window of opportunity and many drugs do fall into this category, but I think that's a -- one first set of questions which one has to ask. The second set of questions which one has to ask is whether or not you're looking at response or whether or not you're looking at your -- you're looking at safety markers and clearly, we believe that when you look at the -- the balance between the increasing efficacy of a drug and increasing safety of a drug, it's much more likely that we're going to be using this initially to be increasing safety of a drug simply because of the -- the practical issues of predicting adverse events reliably. And this is in terms of predicting adverse events. I think it's different if you're monitoring potential adverse events but predicting adverse events it's -- it becomes very difficult because these are likely to be infrequent events for a useful drug. And the third thing one has to take into account is what particular indication you're looking at and also what the -- the balance is between efficacy and risk in that indication and clearly in indications like cancer, there is a -- a high utility of this sort of approach because patients have a very great unmet medical need and one is prepared to take on perhaps more safety issues because of potential benefits than one would in other indications. A second set of issues that one runs into when you're looking at this is define whatting you call response. And that -- as I said at the end of phase II you're looking at the responders and you're looking to correlate these markers with with responders and if you don't have a clear idea of what a responder is and what a responder phenotype is you can run into issues and a recent example that we've run into is looking at patients suffering from rheumatoid arthritis where there are two accepted or broadly accepted the ACR is the most accepted test of whether or not a patient is responding and the one -- the ACR response measure includes a level of acute phase reactants and in a novel therapy which we're using we see it as being beneficial but we don't see any change in the level of acute phase reactants. So this again causes you to question whether or not you're looking at the right sort of responders particularly because novel therapeutic interventions are likely to change what we call response and what we call lack of response. There are many different test requirements if you're going to be using a reliable pharmacodynamic test and many of these are analytical. They're also practical issues. You basically want to test which is particularly invasive, which has -- where you get the information out to the test on a -- in a relatively short time and certainly where the value of the test information outweighs the acquisition costs of that information. And obviously availability of that test is also important and one practical aspect of using this is that in many cases some of the novel markets that we're looking at do not have widely distributed platforms established in the market and this is basically going to be a hindrance to getting wide uptake of these sort of tests in the market. There are also other considerations in terms of the predictive value of that test and the invasiveness. Certainly in our work we've seen this post genetic factors. It's quite easy to get information about post genetic factors based on pharmacogenomics tests but there are limitations currently that we see with the predictive value of -- of post genetic factors, principally because of the -- the -- the pen trants of many of the genetic factors that we're looking at. Other host factors such as proteins, serum proteins, metabolites and chemokines can be acquired active blood tests and infectious agents factors can also be acheirochoired out of blood tests but many tests that we would like to use require the use of tissue and in many cases it's very difficult for -- it's impractical to -- to require a patient to give you a tissue sample in order to -- in order to test for a predicted marker. One exception to this is obviously in oncology where in many cases one has access to samples from the primate human. And obviously one also wants to look at the clinical utility of response prediction and trade off the risks and benefits related to an empiric approach, whether or not the response rate as I said is closer to 100% or closer to 0%, the relative costs, the relative predictive value and issues related to market acceptance. So in summary, some of the challenges around this are identifying the right biomarker early enough. Just going to in a couple of seconds some of the challenges around pharmacodiagnostic tests within time lines and ensuring enough collection of the right samples and defining the sampling conditions at the right time, storing the samples effectively and also having a good protocol for preparation of a test out of the test sample. So I think if you have a validated biomarker, something where you know what the biological information coming out of that biomarker is and you have a good analytical test for that biomarker, essentially there are few issues to integrating this into a drug development. You basically just use the lab validated diagnostic test during your pharma development and you can use that in your -- in your filing together with the drug filing and this shouldn't be a big issue. It becomes more of an issue if you identify the biomarker, let's say in late preclinical work and you have to work up a biomarker assay yourself. Generally what you -- what you will -- what you will do is you will develop a -- let's say a prototype assay. You'll then use this prototype assay during your drug development and you would try and then work up a commercial test for this such that it can be introduced into the market and you'd do a validation crossover validation between the commercial test and the assay that you used in development and hopefully you'll get good enough correlation that you can use this in a data file. So I think this is -- this also works as long as you do the proprietary work in the right way. It becomes more of a problem if you -- if you discover the biomarker either in late development or post launch. One of the reasons is that you won't have a commercial IVD when you launch your drug and the second reason is that you probably won't have been able to integrate this in your -- in your regulatory submission and also in your submission to -- to payer organizations such that it won't be included in your -- in your pharmacoeconomic model. And this means that many issues arise when you run into this sort of situation and I'll go through some of those issues in a little while. So I think one of the -- the other things that we have to take into account about the incentives for pharma companies to do this is that this -- all this work costs money and pharma companies have to balance the investments they -- they make in biomarker work versus the investments they make in new medicines and clearly they'll do this as rational profit insented organizations, they'll do this based on the incentives which are laid out in front of them. So this basically summarizes what I've said, that in early development, pharmacodynamic markers help. Prognostic markers help, but in late development, we don't see many cases of currently of therapy response markers which are predictive enough to allow us to recruit based on those and so this isn't simplifying the situation currently. It's not making drug development any cheaper. It's not making drug xapment any simpler. And in few cases do we expect to be able to use surrogate markers. But we expect that increase of disease markers will allow us to run more efficient trials and differentiate our come pounds in the market more effectively and we expect that these activities will develop to the innovative diagnostics and improvements in the practice of medicine and this will feedback into improved drug development with time. So I don't -- I think I've -- I've been through most of the points on this slide. What I want to go on to is the impact of pharmacodynamics on key pharma value drivers. The value drivers that we see in drug development are the quantity of basically the quantity of clinical candidates, new molecules which we can put into a pipeline, and here we essentially see that the impact of pharmacodynamics is neutral. The quantity of molecules we have in the pipeline and this is measured in terms of success or attrition rates and here we see us being able to improve the potential of getting a specific molecule out of the end of the pipeline by using pharmacodiagnostic tests simply because there are some tests which we'll be able to bring to market which we would otherwise not have been able to do. Time also is an important value driver and this is the dwell time of a project phase and we see this as negative because this -- it takes time to integrate this test, to interpret these tests, and so this is going to be a negative impact on time and the cost -- the overall cost per clinical compound brought into discovery will also probably be negative. It might not be very negative but it will likely be negative. And we -- we think that the overall impact on project value for a successful project is likely to be positive. So taking this all together, adding predictive pharmacodiagnostic to drug development adds costs, uncertainty, complexity, and the potential for value creation currently varies from project to project and one has to make a project specific decision. And if you think in terms of this is just in terms of attrition rates, these are -- I think they're -- they might not apply to all pharma companies, but these are -- this is, I think, illustrative of the current situation. That you have a roughly a chance of somewhere between I guess 3 in 100 and 10 in 100 of drugs which you put at the front end of your drug development pipeline coming out of the back end of the drug development pipeline and one of the goals is to reduce late phase attrition because it relates to the phase. So I think one of the challenges we see is if you're going to do this effectively, you basically have to integrate biomarker work with all the programs you have early in the pipeline and that only a very few of these are going to be successful so all that work that you've put in early in the pipeline adds additional costs and you're not seeing any benefit in terms of the specific projects or you're -- you're basically losing something like 95% of those biomarker projects because the drug isn't getting to market. So this attrition rate that you have for drugs also applies to your biomarker projects and one of the -- I think one of the important things that companies are trying to do is they're trying to capitalize broadly upon their investments in biomarker work by applying these throughouts -- or to other come pounds in a particular therapeutic area. Now, this chart is deliberately complicated, but it's basically showing what the effects of pharmacogenetics and pharmacodiagnostics are on pharma value flows and you can see there are many different factors which one has to take into consideration and these have to be balanced up against one another. You have major factors, minor factors some positive and some negative but I think this shows the complexity from applying these on a specific project and determining whether or not you're likely to get a positive value or a negative value by integrating biomarker work into your project. So the -- the economic rationalE for personalized medicine, I think this is clear for pharma companies, the clear economic rational particularly from a societal perspective that if you have targeted therapies where a drug is linked to a pharmacodiagnostic, nonresponders or poor responders are removed from the pool of users and that costs monetary and negative utility for adverse offense are avoided. Better targeting can lead to a great adoption by good responders, some of whom may not have used the drug previously or may have discontinued use of the drug previously and good responders may have improved compliance and therefore additional net benefits. So the improvement of predict ability of outcomes creates additional value for patients as they essentially face less uncertainty. And these economic rational are very clear to pharma companies. What's the value of -- this is just looking at innovative medicine in general and this is a very unspecific, but obviously fully informed patients are willing to pay -- we see the value as what patients are willing to pay based on these various benefits. So this is -- the value of innovation for novel drugs is what patients are willing to pay, broad societal value and then obviously there's an innovative perspective so how are the incentives aligned with -- with working in this area. So the personalized -- the economic value proposition to patients to company, to pay and society as I say as a personalized medicine or a targeted medicine is to decrease the costs of adverse events faster and more complete adoption and hopefully improve compliance and greater predictability of outcome. And I just want to go through an example for you to illustrate this. If we look at a new drug which has a 20% response rate, this is toward the lower end of that scale I showed you earlier on, and we -- the company who's developing this has an initial price estimate that they can charge $1,000 per year for this drug, what are patients or payers willing to pay for this. How much is it worth if you know who will respond and who won't so what's the value of reduction in uncertainty and what if there are side effects? So again, taking this example, if you have for this new drug a response prediction test and this is, I think, again, we have to see that this is entirely theoretical because we're saying that we have a test, which with 100% sensitivity and 100% specificity accurately predicts the response to this drug and we know that's entirely theoretical and it's based on a readily detectable biomarker so if we were to screen all potential patients and only treat those likely to respond, what would be the value of this? And so it's clear to see what we expect to see happen here that we would get a targeted indication. We would hopefully get based on who's responding and who doesn't respond, faster uptake and improved competitive position. So what should the price be? So with no test we have -- if we're looking at 1,000 patients and patients are willing to pay $1,000 per patient per year without the test, then the value would be 1,000 to $1 million to the company who's selling this. So if we have a perfect test there we have 200 patients one could po sit that the willingness for each patient would go up six times. They have higher than that and the incremental value that we see here over and above the -- the situation without the test essentially comes from what we see as an additional value in reducing the uncertainty. This reduction in uncertainty has a value over and above the -- the case where one has to take a chance. So this -- this lays out what we see as the potential value for this. Now, this is as I said, this is a theoretical case but we believe the underlying concepts which we've built into this case are broadly applicable. One could always argue about what the value of the incremental value of the reduced uncertainty is but we certainly see that there's value you attached to that. Now, a loss of the value capture depends on the manufacturer's ability to set price and this comes into the economic incentives of companies to do this so if the pharmaceutical manufacturer cannot modulate its price, then what it essentially gets out of it even with the reduced uncertainty is it only gets $200,000 instead of $1 million. And who captures the other value basically payers and patients capture the rest of the value in this story. And so this really, I think, illustrates very clearly the impact of timing upon the -- the implementation of a strat fiing test in drug development. The price is nearly always set in -- with novel drugs quite early in the drug's life. Essentially, immediately after registration or around registration and this shows the -- if one isn't able to adjust price, which is, I think generally the case today, the company who's manufacturing this and who -- one is disinsented essentially from discovering and applying a strat fier although the total value for society stays roughly the same. Or perhaps increases with this reduction in certainty. So I think this is one very important factor to look at is that there is currently -- there are disincentives for a pharma company to do this sort of work post initial fixing of the price. What about diagnostic companies? There certainly seems to be some incentives for diagnostic companies to do this sort of work and if you think about the potential diagnostics, if they were able to capture the rest of this value, then there would be very big incentives for them to do this work because they'd be able to charge money for the test. However, current reimbursement schemes for diagnostics do not really -- or do not reward for value creation. They are essentially technically reimbursement schemes and this applies around the world. This isn't just the United States which has a situation. So there are -- I think insufficient incentives in many cases for diagnostics companies to invest in this. Particularly when we look at the attrition equation which I showed you earlier on, if you're going to invest in -- in predictive markers and you're trying to look at hundreds of potential medicines, which ones do you invest in and how do you deal with the attrition of all of those projects which never get to market? So this again underlines the fact that there are really not all that many incentives in place for diagnostic companies to do this work in a systematic way and this is essentially because of the loss of value which comes out of this because they are not able to -- to get reimbursement based on the value that they create. So as I say, the -- the capturing of the value depends on pricing and reimbursement conditions, obviously intellectual plays are all and timing plays a role. So the key messages are the value capture of a linked diagnostic depends on many factors including pricing and reimbursement conditions intellectual and property, et cetera, et cetera and the specific impacts of diagnostic and therapeutic. We believe it's related to who captures it. So unless you have a stronger correlation of value capture and value creation, those disincentives will remain. And our view is it would be wise to encourage value based flexible pricing and reimbursement systems to provide a level playing field that together with intellectual -- appropriately awards diagnostic and therapeutic innovation. And this is a summary of those points. I don't think I have to go through it in any great detail because I think I've teased out the main points there already. I just want to acknowledge a number of colleagues within Roche. Also lieu gar ris son who's corporated with this in terms of putting this model together from the University of Washington and some colleagues at Genentech and Boston consulting group as well. Thank you very much for your attention.

Thanks very much, Tom. We're going to go right into our next talk, which I think follows nicely on this where we're going to hear a little bit more about economic challenges of integrating pharmacogenomics into clinical practice. Catherine Phillips joins us. She's promes of of health economics and health services research and we've heard from her colleague in the past and I hope this will be a continuation on if education of health economics working in this area.

Good morning. I appreciate the opportunity to be here today and I know that I'm the only speaker between you and lunch, so I will keep that in mind. I won't be cruel. We all know that there's been a lot of hype concerning pharmacogenomics, some people saying it's going to revolutionize our lives, but where are we now in pharmacogenomics? Some people would say it's here, you better get on the bus or you're going to get run over. But others would say, well, where are the benefits? Where's the beef? I don't see anything good coming from this. And I realize it's cruel to put a hamburger on the slide at this point, but hopefully economics can help us figure it out. So today I'm going to go over some of the economic challenges of integrating pharmacogenomics in the clinical care. I'm going to go over some of the steps needed to maximize the value of pharmacogenomics and how economics can help us do that, and I'm going to go over three case studies. If you read the newspaper this morning you know that there's been more news regarding these. So why is economics even relevant? Well, it provides both a toolbox and tools. The toolbox is the conceptual framework and the tools are the message that we bring to bear. And economics can boil down to two things. One is incentives. In other words, here we're interested in why is pharmacogenomics adopted or not? And what type of incentives will maximize the value of pharmacogenomics? The other critical piece of economics is value. And by here I don't mean money only. What is the value of pharmacogenomics? How is value defined? How does value change by whose perspective we're looking at? And how can value be measured? I thought it was important to first tell you just briefly where I'm coming from in that I do wear three hats. I'm primarily an ak demission. I do research on drug safety and policy issues but I also do some work with the government which has helped me understand their view. I work with the FDA Gutman particular and I'm a member of E GAAP which I understand the company is already familiar with. So today I'm going to cover three steps. First of all, that we need to understand the importance of economic and noneconomic incentives. We need to consider value from multiple perspectives and we need to use integrative approaches to address new paradigms. I'm going to use three case studies which you're probably familiar with, Herceptin, Iressa and CYP450 drug metabolizing enzymes. First point, understand the importance of economic and noneconomic incentives. Pharmacogenomics adoption will only occur if there are properly structured and lined and built-in incentives. I often hear and I heard this morning that physicians need to be trained. Well an economist would immediately jump up and down and say no, that is never going to be enough. There needs to be built-in incentives for physicians to use pharmacogenomics or it will never occur. The problem is that incentives push in different directions. And that the incentives for adoption may vary based on the characteristics of the intervention. Here's a laundry list of some characteristics to provide incentives for adoption. And this is my own list. It's life threatening versus a chronic condition. If there's a strong advocacy group or interest and obviously the high reimbursement rates and as my colleague was just saying if pharmacogenomics is used early in the pipeline as opposed to later, if it's used for immediate versus future treatments decisions if it's used for focused narrow treatment decisions, a very important one that I've been hearing a lot about is that if pharmacogenomics is be for all label indications there's a lot more interest in using it. If it's used for on going monitoring versus one time use. If it targets an acquired versus inherited mutation. And then finally one that you might not have thought of which is pharmacogenomics is more likely to be implemented when it is not considered pharmacogenomics. We frequently call it targeted therapy, smart drugs. Now, why might that be the case? Well, first of all, the concept of personalized medicine is much bigger than pharmacogenomics. For example, it includes the use of family history, so it builds on existing approaches instead of appearing to have emerged de novo and it's easier for people to understand and support the concept of personal ietzed medicine versus genetic testing. One reason being because it emphasizes the drug as o poetzed to the person. Let's look at some case studies. Herceptin, this illustrates a very fast and successful adoption. It's one of the best known examples, although people don't consider it to be true pharmacogenomics because it targets a tumor. It has proved that targeting to small populations can be feasible and profitability for industry. Sales keep increasing. They're going to go up today based on the newspaper articles. In 2004 sales were 479 million, a 70% increase in one quarter alone. It's important to note that testing here is for gate keeping, not for dosage decisions. In other words, if you test positive, you get the drug. Iressa is an example of a fast but currently unsuccessful adoption. Here we have a case where the FDA accelerated approval of the drug, but the drug has been essentially with drawn from the market because post approval clinical trial showed no significant survival benefit. However the drug does appear to benefit specific populations but until recently there has been no diagnostic. There is one now developed but right at the moment there's limited availability, it's an expensive test and right now it's unknown what the benefits will be. CYP450 testing illustrates slow adoption. There have been many implementation challenges. The multifactorial data response the clinical outcomes, vary ability not only across drug classes but within drug classes as well. In this test it's the person and that raises more ethical issues and testing is not a strict gate keeper test. In other words the incremental benefit of the test is harder to measure. Second point. Consider value from multiple perspectives. All stake holders want evidence of value but they're going to differ in terms of their perspectives. Unfortunately from a societal perspective there's very little documentation yet of the value of pharmacogenomics. We did a review of all the studies to date and we only found 11 cost effectiveness analyses of pharmacogenomics interventions. A very limited range of conditions have been studied and the results were quite mixed. What are some of the challenges to determining the value of pharmacogenomics? Well, first of all, differences in perspective. In the value determinations are often made before the product reaches the clinical setting. I was talking to fa*i about my talk before I came and she said well, you're talking about a lot about the pipeline and before you get to the clinic and I said, but that's because the economic decisions are often made long, long before the product reaches the clinical setting. Therefore, we need to consider economic incentives throughout the pipeline and evaluations need to be conducted before the intervention reaches a clinical setting if the societal benefit is to be maximized. There are a number of technical issues in determining values. Lack of data I've already mentioned. Linking pharmacogenomics to outcomes. Comparisons to therapeutics and on the products themselves. So that economists like me can't get our hands on it. We have to evaluate some mreks multifactorial conditions and by definition diagnostic drug combinations are more complex to analyze in the separate interventions and there are a number of policy and political issues. There are few incentives to assess it from a societal perspective. We don't see those in insurers. That's not usually their role. In pharmacogenomics often has the benefit of preventing what is not occurred and it is always harder to measure the value of prevention. For example avoiding adverse events. It's hard to measure what the true value of that is. And it's hard to measure the value of diagnostics and I often hear people say well the up front testing cost is going to outweigh the downstream savings. Herceptin illustrates a successful adoption despite the lack of documentation of societal benefits and I'm going to pause on this slide because this is a very important slide. Many people do not realize this. Herceptin's expensive and the newspaper today said it's actually $4,000 a month. It increases median survival by a few months and there have been a few economic analyses, one was done by a group at Harvard considered to be well done. They concluded that Herceptin -- the important thing to understand about that is that anything that's over around 50,000 is usually considered we're not so sure at the benefits if the benefits are worth the cost. Outside of the United States the approval of the drug was slow because of concerns about the cost. Iressa, now Iressa illustrates how failed adoption has the potential to create large societal losses and sometimes we don't think about that in that a withdrawal of a drug from the market incurs large losses not to the industry but to society as well because of the patients do don't benefit the regulators having to spend all that time regulating and in general increases concern about drug safety. CYP450 testing is an example where widespread testing could have a huge economic and societal impact. But it's going to require some creative and complex approaches to assessing the value. We did a study back in 2001 that found that there was a linkage between adverse drug reactions in P 450 reactions. Bedid a more recent study and we found it had a large impact because many drugs are me tab lietzed by CYP2D6 and 12.8 billion dollars in expenditures annually in the United States particularly in the areas of mental health and heart disease drugs but this -- and this is a very big but, there's currently insufficient data to assess the impact of CYP2D6 testing. There's very limited outcomes on the clinical testing and only one package to the availability of the test. My third and final point, use of innovative approaches to address new paradigms. We've talked this morning about the role of diagnostics and codeveloped diagnostics and drugs that are going to play an increasingly important role and as we know, that requires the integration of historically acquired industries and regulatory mechanisms and early consideration of diagnostics which we just heard from my colleague. I'm doing a study for the FDA looking at barriers in the diagnostic pipeline where I'm interviewing a lot of key thought leaders and I've heard three major barriers mentioned. One is money both in terms of initial investment in biomarkers and the reimbursement rates. The second is availability of data in samples, and the third is the clinical utility tests are often not evaluated and thus it's difficult to demonstrate the value of diagnostics. With Herceptin and Iressa we've seen that it will be challenging to develop and determine the most appropriate diagnostic. With Herceptin several tests were approved but there's still debate over which tests to use and the development of diagnostics is often going to require multiple stake holders who traditionally have not merged forces academia industry and the FDA. With CYP450 testing it illustrates it will be challenging to adopt pharmacogenomics when it's relevant to multiple diseases and drugs because CYP450 testing is only done once in your lifetime but the results are relevant to multiple diseases, drugs and clinical specialties so it's unclear who's going to advocate for testing. And other critical shiesh u whether the test will be considered diagnostic or for screening. Medicare covers diagnostic tests but they do not cover screening tests and in this case it's a bit unclear which one the test really is. So it's unclear whether consumers are going to seek this out, providers will provide, industry will have incentives to continue to develop such tests and whether insurers will cover these tests if they are considered screening. So to summarize some of the next steps to address economic challenges of integrating pharmacogenomics understand the importance of economic and noneconomic incentives, incentives do matter, but they're often contradictory but they can be shaped by health policies. Consider value for multiple perspectives, the definitions of value will vary, but value must be determined one way or the other. If it is not done from a societal perspective then it will be driven by other perspectives and there are -- therefore I would argue that we need incentives for more economic research. Third, use innovative approaches to address new paradigms which will require truly multidisciplinary approach and innovative funding mechanisms. Unfortunately social science often lags behind basic science in this area. Why is that? Well, it's a risker area to do research in. It requires more understanding of basic and clinical science and it's hard to get funding in this arena. It requires development of development space. It's a good example. They are developing a database, however they expolice Italy do not include issues regarding application of the technology and then EGAPP once again is a good example but ultimately EGAPP will end in those issues and will need to be institutionalized if those evaluations are going to continue. So to conclude, pharmacogenomics is here now and will keep coming. I believe that there will be an inevitable push toward pharmacogenomics because it's part of the larger trend towards personal ietzed medicine and for that genetics information is one piece but it will be a critical piece and I believe the government therefore has a critical role in facilitating their role of appropriate pharmacogenomics in order to maximize its benefit by shaping incentives, by ensuring the value gets measured from a societal perspective and by facilitating innovative approaches. Thank you.

I want to thank both of our speakers for very good and inciteful presentations. We've got about 20 minutes for Q and A from the committee and maur renohas his hand up already so we'll let him go first.

I'd like to thank both of you this morning. I've learned quite a few tricks today but it's very interesting that we're embarking on what we're calling personal ietzed medicine and we still have to have a societal perspective on this and I fully subscribe to to this endeavor. I want to ask both of you sort of a question around this pipeline and the incentives or disincentives and maybe use a couple of examples because when things are out and we discover that there are suboptimum or have side effects they tend to be with drawn from the market. For example, the cox 2 inhibitors which are very effective. Unfortunately they double or triple your risk of heart disease. I don't know how much it translates into an absolute risk with respect to patients. In the area of vaccines for example and that's something we don't talk about or we haven't heard about. And a few years ago there was a vaccine that was very effective. Unfortunately few percent -- few per million babies had -- and as a result that vaccine was pulled from the market and we're talking about various side effects increase in some case which the ro to virus vaccine although the benefits were really proven to save a lot of lives as far as a major global health threat I guess around the world, much more than in the U.S. and in the case of the cox 2 inhibitors there were millions of people and I'm one of them that did sort of its back pain and other things depended on having a steady flow of that therapeutic. So can you kind of revisit those two examples? And I'd like both of you because these were sort of distal to the pipeline and by then I guess the incentive -- I mean, the drug is already on the market, the price has been set but side effects emerged and instead of studying why some patients are more susceptible to heart disease as in cox 2 inhibitors the drug was pulled from the market. So help me out a little bit through that example.

Well, that's very interesting. And it speaks to the idea that some people think of pharmacogenomics is going to be a panacea and I'll return to the newspaper article although I read it at what my time is 4:00 in the morning so I can't guarantee that I read it correctly but I thought it very interesting that they mentioned Vioxx. As far as I know pharmacogenomics couldn't have helped that issue at all, that there's some --

The investment in developing biomarkers that's the question I'm asking to find out why.

At this case we don't know anything that could have helped who finance going to benefit or who it was going to harm.

Could it be post mortem or is it too late to save those categories of drugs.

The other thing I thought was interesting they never mentioned the word pharmacogenomics. It was all grouped together in this idea of personalized medicine and it was all portrayed as we can now cure cancer through personalized medicine and so I get a little nervous when I see those types of meages going out to the public that are going to raise hopes that maybe it's going to take a while to -- to address.

Just to maybe answer, I think the -- the safety aspects of this certainly Merck was highly insented to try and find a predicted marker, but I think it's probably -- I mean, I don't know enough about the specifics of it from the Merck perspective, but it's -- it's practically a very difficult thing to do to reliably predict, so this is in terms of predicting and not monitoring, to reliably predict who's likely to suffer an adverse event because of a drug when you're looking at something that happens in a very few percentage and it's an extremely difficult thing to do. It's not only difficult to find the predicted marker, but it's also difficult to validate it as well. And you require -- you would require very big studies to validate or to demonstrate that effectively. So that's -- that's one thing and I think there are -- the other thing is that in many of those cases or in particularly in the Vioxx case, you still have a drug on the market which is an alternative and obviously if you have one drug which has a higher frequency of -- of safety issues and another and you're going to use the alternative which doesn't require the predicted marker, I think that's very clear. There are some cases as we've -- for instance, the use of the TPMT test where that is a safety marker, you're looking at something which is relatively frequent. Unfortunately with that marker or that test, you can reliably predict who's most at risk from safety issues so there are some exceptions to this rule that you're -- I think that it's going to be very difficult to do this from a safety perspective. And certainly if -- you know, in the ro to virus case you mentioned if one had been able to do that, then perhaps there would be a case to do that and I think the manufacturers of the -- of -- also reasonably well insented to do that at the moment if there was something which was feasible.

I'd like to thank him for his question because I want to piggyback right on top of that. Before I do I just wanted to mention one other thing. I'm fascinated by the fact that an economics researcher would see the need for more economics research. But in any case, what I'd like to ask you is both of you, the flip side of what he was asking, he kind of focused on if safety, I'd like to ask what the incentives might be, the economic struggles and challenges might be on the efficacy side. How -- what's the possibility here for rehabilitation of drugs products that have failed in a sense in the past because groups weren't identified necessarily that could be identified now that might benefit? So of course by dell would be the perfect example recently. Patent protection would obviously help and so you might have to distinguish when that might be possible and wouldn't be possible. And another drug is e thorn fleen which was considered to be a cancer drug, turned out to be one of the best treatments they found for sleeping sickness, of course there's no market for that but then it actually got out of the market as a facial hair remover. So I mean, it's these kinds of things -- the beauty of it being of course now that it's back active you can hopefully move some of that into Gambian sleeping sickness treatment. So what are the possibilities here to go back and look at different ways of rehabilitating old products?

With -- I think where you have a case where new incites would allow you to use a different compound in a different indication as long as the patent protection I think there is plenty of -- plenty of incentives to do that. And mren tiff of possibilities to do that. I think the -- from our point of view that the two biggest or the area where the incentives between what's out there at the moment and our view of what the societal perspective is most misaligned where you have a drug on the market already, which is being successfully marketed where there is the potential for you to say sharpening the targeting of that and as I -- I think I'd tried to lay out, there is -- there are few incentives for drug companies themselves to work on that. There are incentives for others to work on that. Incentives of payer organizations to work on that and certainly for diagnostic companies to work on that but there are few incentives for drug companies to work on that. And in general, I think the diagnostics industry doesn't have much incentive to invest heavily or speculatively in this area because of the -- it doesn't get value based reimbursement. There are cases where there are companies now -- diagnostic companies investing in let's say a test for improved prognosis of breast cancer, but that's because there are enormous economic general economic benefits from doing this and I think that they are reasonably speculating that they will get a good economic return on this. But as soon as you go away from the biggest opportunities, the incentives for diagnostic companies decline quite rapidly. As -- you know, for the cases that you mentioned I think each case has to be looked at on its merits upon -- but we don't see that as being a -- a massive misalignment between potential societal benefits and particularly with often drug legislation.

I will agree that's what I've heard. I'll add the related issue of off label uses where my industry colleagues tell me that the business models coming out all really heavily emphasize offlabel use. They expect to make their money off of offlabel use. The insurance company I talked to are shaking in their boots because they have a really hard time dealing with the reimbursement of off label uses because they don't have any evidence base to deny coverage and so they're caught in this in between of they don't want to ration care, but they don't have enough evidence on which to turn back these requests for reimbursement for off label use. When you're talking about these very expensive drugs they see it as a really critical issue.

Yeah, one of the things that has puzzled me is the lack of lawsuits that have been brought against, for example, physicians who don't use TPMT and for example most of my colleagues don't. And certainly lawsuits are kind of an incentive or disincentive obsession with physicians. I was wondering from your perspective, I didn't hear either of you mention issues of liability. It seems to me those might be important in bringing pharmacogenomics to the fore or not. My other question is unrelated and that came up in the survey results we saw earlier today. And that is in those gray areas and it's almost always going to be a gray area about whether a pharmacogenomics test predicts that a certain person might respond better or might respond worse, do you think that insurers will tend to try to deny coverage for that drug for individuals for which there's some evidence they would not respond as well? So those two questions.

So in the -- on the liability perspective, I mean, this is obviously -- I think it's a reality in the U.S. market, but I think where you have a clear standard of care and where that standard of care isn't adhered to then there's obviously much more risk or -- or chance, whichever way you look at it of litigation. We don't -- I don't think we have a strong view about that except that it makes sense to provide enough evidence to establish a clear standard of care where you have that evidence. So I think that's -- that, I think covers the litigation issue from our perspective. With respect to the -- sorry, the second question was?

Well, following up with that first one, do you ever think there would be a disincentive among companies to really do the research because they don't want to identify those in whom an adverse reaction might occur? Does that make any sense or is that more of a conspiracy --

I mean, certainly it's one of the things that one takes into account when you look at litigation risk in general. When you look at opportunities, but it's a reality which is out there and I think that we certainly we try very carefully, I mean, litigation, I guess, is the extreme end of the risk benefit continuum and we are very concerned about the risk-benefit continuum. We want to do as much good as we can with as little risk as possible. So you're going to shy away from the extreme ends of that and try and focus on where you're doing the most benefit with the least risk. You're --

The second question --

I remembered what your second question was.

Right. Do you think there's a chance that insurers would tend to deny coverage for drugs that were suggested that they would be less effective in an individual? Do you think that's a potential problem?

I don't know whether it's a problem. I think it's a reality that -- where if you have clear evidence that a patient is highly unlikely to derive any benefit --

Right. I'm talking about the more, what's going to be very common which should be the less clear-cut issue. So 60% chance of responding versus 30% chance of responding, is that -- which I think will be far more common with these tests than a really extremely clear-cut.

I don't think it's something that we -- most certainly we haven't debated this long and hard, so I don't think the -- there's no clinic attempts -- one of the things that we look at carefully though, let's take the -- I think this is a good camp l that we've just seen which does inform us a little bit is the situation with Trastuzumab and Iressa where you have a patient -- to deny -- and if you're going to require patients to undergo a biopsy, which -- which is invasive which has potential mortality and more bidty in order to extract that test information, we think that's probably unreasonable to do that lacking the strong evidence and even if the evidence base was there, that also might be questionable, so there are many questions which go -- which inform whether or not it makes sense to try and acquire the test information and then the interpretation of it and obviously the clearer cut is more likely I think it's going to be included in reimbursement decisions.

I'll add very briefly that patients did play a huge role in determining what pharmacogenomics intervention and in the case of Herceptin originally general entech was reluctant to move forward. Patients chained themselves to the gates of general enit can and demanded the drug. Of course they were talking about breast cancer and that's quite different than a lot of diseases where there are no advocates.

So I'd like as sort of a wrap up question for this morning's session to ask of you both -- you did a very nice job of helping us understand the financial decision-making and the value proposition thinking that a corporation goes through as it's making decisions about how to take things forward. It's our charter to advise health and human services. Do you have some specific things that you feel should be done at the HHS level that would benefit the industry in general? Advice, comments, anything that -- I just really like to give you a chance to make any pitches that you might have for government partnership in this.

Well, of course I have to say economic research.

Okay. Okay. Besides funding your grant.

I did try to cover that in my final slides and I do think the government has an important role here. Certainly in terms of the evidence base and helping facilitate the evidence base and ensuring the value gets measured, that the societal perspective does get included because otherwise that might fall through the cracks.

I think there's three or four things that we see. One is working on basically getting a clear regulatory framework and a clear set of standards so I think we've heard a little bit about that this morning that companies can invest with confidence in some of the opportunities around this because there's a clear frame work out there. That's one thing. The second thing is on -- in general and I think that we're seeing a movement in this direction more funding for translational research and happily we're glad to see that the movement in this direction, but it's -- I think it's slower than it could be. The third thing I think is clearly having the right incentives aligned for -- for profit oriented companies in this arena and in the particular area we see there is value based reimbursement and I think Kathryn quite right says it right in trying to interpret what the value is but if there's a clear misalignment of the incentives and the value creation it's unlikely that profit organizations are going to invest in these areas.

I'd like to thank both of you very much for your participation. I know that you came a great distance to be with us today and we really appreciate your time. We're going to take a lunch break now and we will resume the afternoon session promptly at 1:15.

Okay. I'd like to ask the committee to take their seats, please, so we can start the afternoon session. this afternoon we're pleased to have another committee alumna with us. Wylie Burke is going to talk to us about issues associated with particularly she's going to focus on the issue of race and genetics as it pertains to differential drug response. She's currently a professor and chair of the department of medical history and ethics at the University of Washington and is a faculty member in the public health genetics program and in the medical genetics training program. I think she needs little more introduction to most of us. Thanks for joining us and we look forward to hearing your presentation.

I must say that this is a topic and a set of questions that I was asked to address, they're difficult ones and so my main goal is to try and convey to you what I think is the complexity in these issues when we start talking about race and genetics and differential drug response. We can start with the complexity of race. Race is the term that's used generally to identify or at least it's assumed to identify groups with shared ancestry. That's the way in which we use the term and implicit in that use of the term is a strong belief, particularly, I think, in U.S. uses of the term race that -- that race has a lot to do with genetics. So there tends to be a tight alliance between those two terms and I want to talk about the complexity and why we need to think, I think, more in depth about that. It's currently understood to refer to five groups. African, European, Asian, native American and oceanic but as we think about the definition of race and the means of race I think it's important to understand that the definition of racial groups has changed over time. A con spik wows example in the U.S. has been a relatively recent allocations from the Indian subcontinent to Asia where they previously were not. And it has been used differently in different parts of the world. In addition, when we look at countries like ours where there's a strong tendency to categorize people by race, in fact, social factors are what determine what race you put down on your census form. So I think we have to be cautious about how the term race is used and really a variety of different meanings that might be incorporated in the term race. Now, that said, there certainly is a relationship, though I would argue it's an indirect relationship between race and pharmacogenomics. Self-reported race is correlated imperfectly, but correlated in a rough way with genetic measures of geographic ancestry so researchers who have made efforts to identify highly variable markers that are particularly useful in identifying ancestry of the five major groups that I've just referred to have developed marker panels that can be used for what's called ancestry testing and there is a rough correlation. As a result, the -- we do see some genetic differences and they're generally differences in prevalence so what we see when we categorize people by race or allow them to categorize themselves by race we see the prevalence of gene varies with ancestry and that of course includes variants that are associated with drug response. So I want to give you an example of that. The sip 249 has been identified to have many variants or several vary yanlts that are associated with reduced dose requirements for warfarin, the drug that's more likely used for anticooglation or throughing thinning and my colleague colated together a whole bunch of studies and you can see the results here that if you look to people that were of European ancestry, the prevalence of variants associated with reduced dose requirements looked to be about 36% with a fairly wide range but an average of 36% where as in African groups, it was 8% Asian, 4% -- and I've put the figure 10% for native American but you'll notice it's just two studies both done in Canada one where there was a degree of prevalence so we shouldn't generalize too much. The other point I would make sheer is that the preponderance of studies is of people of European descent and the result of the studies is in North America so what that means is our data comes from a small slice of the world's population. I think it's fair to say we know relatively little of the distribution of these variants around the world. But CYP 2 variants not only require lower doses of blood cooglation was have a higher likelihood of having bleeding complications. And there at least have been some proposals that this may be a clinically important genetic trait. What's interesting about it is that the one racial difference that has entered into sort of rules of thumb in clinical medicine is one that's not explained by CYP 29 variants. So Asian patients tend to require lower doses and if you go to the physician's desk reference, you'll find this paragraph and I've -- the highlighting is mine because I thought some of these sentences were particularly interesting. So they start by saying Asian patients may require lower initiation and maintenance doses of warfarin. They go on to cite their basis and it's one noncontrolled study of 151 Chinese outpatients and they go on to show the lower dose requirements, so I think we'll note uncontrolled study, small numbers and in fact, it's not a broadly representative population of Asians. It's actually one ethnicity within a large number of categorization. These patients were stabilize on warfarin for various indications so there might be a variety of clinical differences that also are of importance and patient's age was the most important determinant of warfarin requirement that's a known factor about warfarin. So it's a rule of thumb that made it to drug label so there's enough conviction in the clinical world that this is clinically meaningful and I certainly had docs tell me that this is the truth but the evidence base for it is not particularly strong. In fact, there are many, many sources of individual variability in P 450 expression of which the CYP 29 drug is won. This is a data summarizing all the factors that might summarize someone's response to drug and you'll see most of them are nongenetic or distantly related to genetic factors. So many environmental or life style factors might influence how one's P 450 is expressed or genes in that family is expressoed and how one might have differential drug response. It turns out that a subsequent study on a different gene actually provides evidence about a possible genetic contributor to this Asian -- this apparent Asian requirement for lower dose of warfarin. That's a study published recently by my colleagues from the University of Washington that was looking at the association between wor VIN dose VKORC 1 haplotypes. They were able to identify two groups. Group A associated with a low dose requirements and then group B, another set of variant haplotypes that were associated with higher dose requirements. And they -- the evidence is on this slide, when they looked at all the patients in their sample and they actually repeated this in two samples of patients at university of Washington and Washington university, you can see the correlation between what kind of haplotype someone had and what their dose requirements were. They also were able to take advantage of the fact that some of their patients also had the CYP 29 associated with lower dose requirements and you can see the mutual effect of the two genetic factors in reducing dose requirements for VKORC 1 for warfarin and what's really interesting is that there does seem to be a difference in prevalence between racial groups of these different VKORC 1 variant haplotypes. These data, the European African and Asian samples were taken from the data repository and what we have is that in Europeans there's just under 40% who have that low dose haplotype but close to 60% who have the high dose haplotype. With Africans you can see that fewer of the haplotypes within the African population are explained by this group A and group B, but it's a different distribution and then most interestingly, the Asian sample has a very high proportion of the VKORC 1 haplotypes associated with low dose. So this could be a genetic explanation for that clinical observation. Again, I want to note that the Asian samples were predominantly Chinese so we continue mostly to know about this particular genetic issue in Chinese individuals. This study also estimated the variants in drug response that could be explained by the CYP 29 and by the VKORC 1 variants and here's another important point. We certainly don't want to say we now know everything about why Asians might have a lower dose requirement than other racial groups, because VKORC 1 is only explaining 25% of the variants. CYP 2C 9 is explaining an additional 10%. It's very likely that there are other factors. This tells us there are definitely other factors and they may be genetic or they may be nongenetic and I think it's important to say at this point that either is a possibility. We await more evidence to understand this better. I think there are a few issues here, one is that even on the genetics side, granting that both genetics and nongenetics factors contribute to response, even on the genetics side we're going to see some mreksty so if you looked at CYP 2C 9 you only got a slice of the story. And there may be well other genes yet to come. There are many yeens that are involved in the metabolic processes by which the body responds to ingested warfarin. In fact, there was recently a study from Europe that identified variants in the APE gene as having an effect on warfarin metabolism. It comes from a group that's been doing a lot of epideem logic studies. And it's the very same one increased with associated risk of Alzheimer's seems also to be associated with lower dose requirements for warfarin and we may still have others yet to come. And there are many exactly as you would expect many well-established nongenetic factors so*ets associated with warfarin response. I've listed the ones that seemed particularly important in terms of clinical writing on this and the emphasis on nutritional status, GI disease has to do with the fact that one's vitamin K level is very strongly influenced by the status of bacterial colonizization so clearly the response to warfarin is complex. We're beginning to understand genetic contributors. They're very enlightening. They potentially will lead to pharmacogenetic tests that will help to prevent bleeding complications and not surprisingly we see some difference in the distribution of low dose across racial groups but whether we're getting beyond race I think is not clear at this point. It's hopefully I would argue we are. I want to get back to the point that races are genetically heterogeneous. Does it matter that all we know about VKORC 1 is in Chinese individuals by and large and I think the answer is yes we should care about that a lot. In fact, as I showed you with the CYP 2C 9 data, most of our data comes from Europe and the U.S. and most of it is on people of European ancestry. We're seeing an increasingly amount of research going on in Asia particularly in Japan so we'll see some more data from that sector but at this point we're very limited in the data that we have. One collection of data that I'm showing you on this slide shows why it's going to be important to do very broad sampling. This was a study that looked at the prevalence of the APOE epsilon 4 allele in populations around the world and the fundamental point I want to make sheer is that if you look at three different populations in Africa you find a range of 9% to 41 pkts. Five different populations in Europe, a range from 5 to 31% and so on. The only population where they didn't see that broad a distribution seems to be in the three populations that were of oceanic origin. And the point here is that we can't just take a sample of people that have four grandparents of a certain geographic ancestry and believe that we have sampled that geographic ancestry. Within racial groups I think it makes sense to anticipate and these days to support the notion that we should anticipate a lot of heterogeneity, another reason to be careful about generalizations. In addition to that kind of heterogeneity and most of the studies were taken in the different geographic locations, we know that race and geographic ancestry in this country Friday are related but not congruent. I'm just getting back to that point and I just want to show you a few data points that were cited by Mike in a recent JAMA article on this point. West African, most African Americans at least most African Americans that came here as a result of the slave trade were or whose ancestors came for that reason were originally from West Africa. And if you look at geographic ancestry testing panels in African American individuals, you find that west African markers account for about 80% of the ancestry on average. But the range is from 20% to 100%. If you look at people who are self-identified as European, a substantial proportion of them have less than 90% European ancestry. That mixture is even higher and more variable amongst people identified themselves as Hispanic. He cites data that people of Asian and African ancestry in other developed parts of the country tend to have more heterogeneous geographic ancestry than we see in the U.S. and of course, the other point I would make here is that this is a snapshot in time. In an increasingly global village and where there are many movements of populations. I read an article recently that said that immigrants from Africa who have come from Africa voluntarily now outnumber the population that is derived from people who came to this country from Africa involuntarily. And many of those people come from different parts of Africa than West Africa. So we have, for example, in Seattle, a large immigrant population from Somalia and Ethiopia and given that heterogeneity that we may see across people of African descent we may see the population identified as African descent. So we get to a fundamental question, is race clinically important in drug treatment? Well, race clearly captures many potential group differences as I've said. Even if you account for heterogeneity and you always have to have that heterogeneity in your mind, we do see broad differences when we group people by race in -- in -- not always but sometimes in issues of diet, issues of housing, issues of occupation, issues of environmental exposure. There's been a lot of work done, for example, looking at the higher risk that African Americans experience to be in substandard housing or to be exposed to medically significant environmental exposures and as I've shown, we also do see difference in prevalence of gene variants. So it's not unreasonable to expect that we might see differences in drug response when you consider that all of these factors might innuance how groups respond -- or how individuals within groups respond to drugs. I think the question really is, if we make such observations, what do we do with those? The story of the VKORC 1 and basically our growing understand of warfarin's response says that it's worth investigating and sometimes we'll find a specific genetic difference that may offer some explanation for a group difference and when we do, that genetic explanation will help us to identify individuals and move beyond the group. So we can now envision a day where it matters what CYP 2C 9 genotype you have and not what race you are in determining your warfarin response. Other times it seems very likely that an observation of group difference will get us to an extremely important nongenetic difference that -- that may be an important thing to address clinically. In a recent editorial in nature biotechnology, the editorial is described race as simply a poor proxy for the environmental and genetic causes of disease or drug response. Now, crude markers can sometimes be very useful indicators to get us to questions that are important to research. I think at this point it's very uncertain whether in the long run race is going to have sufficient predictive value to assist in drug treatment. What are the implications of these kinds of observations for pharmacogenomics research? Well, clearly, if we want to identify all the variants that are relevant to a particular drug response, we've got to study diverse populations in large numbers because we run a real chance of missing important variants if we don't. As we identify individual variants to always have in mind that multiple genetic and socio environmental factors are going to be important and so any one observation has to be sorted out for its place within that mix and obviously gene-gene and gene-environment interactions are going to be likely in that setting. I think also the issue of orphan genotypes is an extremely important one to think about. They're likely by their nature to be rare to be less studied and particularly in terms of the information we heard earlier today about the use of genomic information and drug development, there would be a real possibility that people with rare genotypes could miss out. Given the preponderance of research occurring in the U.S., the preponderance of drug development occurring in the U.S., I think we have to worry about a particular category of orphan genotypes and these are genotypes that aren't necessarily rare, but are genotypes that occur predominantly in minority populations and so they're not rare worldwide but they're rare within the U.S. population or relatively rare. And the -- the issue of loss of lactase leading to the condition called lactose intolerance represents a very interesting example of this. A group of researchers wrote an interesting article in 1989 in the journal of medical association claiming racial bias in federal nutrition policy and their argument was that federal nutrition policy has for a long time recommended milk product intake as a very important part of calcisum in the die yet but as this point points out the inability to digest lactose, that is to tolerate milk products and in fact, the rate of GI symptoms when one tries is quite high in virtually all racial groups except Europeans. I'm not sure there's data on oceanic groups so I should give that caveat. But native Americans, Hispanic Americans, it's a very common problem and milk is not a good source of calcium for people who have this problem. Now, again, it's a prevalence -- they run something like 70% versus something like 15% in Europeans and it's -- as we've been saying about other things, it's certainly not a simple racial trade. So for example, if you look amongst Europeans lactose intolerance is much more common of people of northern European di scent than it is of northern European descent, but what we have, we have something basically akin to an orphan genotype here because we have policy having been made and built upon the needs of a dominant group in the population who happen to have a genetic predisposition that is quite unusual worldwide. And I think it's a very important point that illustrates how careful we have to be about the orphan genotype issue. I've said that one of the implications of thinking about race and genetics and pharmacogenomics is that we need to do research in diverse populations and as we do so, we have a very important ethical concern to overcome. And that is the existence of mistrust about genetics generally and perhaps more specifically about genetic research among minority populations. I've pulled out a couple of examples but there are many examples. There is currently a lot of tribal mistrust in North America related to the use -- arelated to research misconduct in the past or alleged research misconduct. So there have been examples where tribes have agreed to the collection of genetic samples for certain purposes. The most famous example is that f of the HAVASUPI tribe which was very interested in the genetics of diabetes being investigated and then 10, 15 years later a member of the tribe comes to a research presentation and discovers that the very same samples that were collected for that purpose have been used for ancestry, for my grags studies, for studies of inbreeding. For studding of sketsphrenia, a variety of things that the tribe did not consent and that's one of the examples. The other example is one of many examples. We do have data that suggests there may be more mistrust and more worry about the misuse of genetic information amongst minority populations. This is a survey of premedical students had only minority students in the sample so we don't have a comparative statistic. And I would have liked in the survey data that we heard earlier to have seen a breakdown by race because I think that would have been informative. But certainly there's the worry there that -- that discrimination has occurred on other bases and this might be yet another basis for discrimination. So if we say that it's important in order for all groups to deroo*i benefit from pharmacogenomics, in order not to miss the orphan genotypes that are really important, then this becomes a tremendously important imperative to figure out how to develop partnerships, how to incorporate minority communities within the research enterprise, to move forward together so that the research is done in ways that minority populations endorse and feel are good for their community so that the -- and so that they have assurance that the information will come back to them in ways that can be helpful. I'm just going to end by talking about, in summary, what I think are the significant risks that derive from the use of race and genetics and the study of differential drug response and I'll make a couple of remarks about how I think these issues apply more generally to ethical concerns in pharmacogenomics. I've already made the point that we have inadequate research in minority populations or if you want to put it in a global scale, we have too much research in the U.S. and Europe and not enough in other parts of the world where the populations that are minority in this country are not minority. We need very careful attention to the side and the sampling methods they're used for populations. We need to get away from the situation where we say Asian but what we really mean is Chinese. We need to figure out effective ways to partner and address community concerns to have community concerns be incorporated within the research enterprise and obviously fundamental obligations of research integrity must be maintained if people give permission for examples to be used in one way they really can't be used in another way. That's fairly straightforward. As we go forward, of course we also have to think about multiracial groups. There's no simple bright line where Africa leafs off and Europe starts and one of the biggest things about ancestry testing is that it tends to focus on -- clear example. But we have many parts of the world where it's not clear. Where there's a lot of what ancestry testing researchers called a mixture but what is inevitable heterogeneity that becomes from trying to draw a bright line saying this group on this side and this group on this side. ? Addition, we have an increasing number of people in this country and many other countries who identify th*echlss as this multiparable just as we need to be concerned not o leave orphan genotypes behind we need to think in terms of a multiracial society. And clearly the issue is to avoid the -- the need to recognize that even when we find genetics predictors that help us to identify people that may have higher or lower likelihoods of either adverse effects or -- or effective responses to drugs that that's only one of many contributors and we shouldn't inflate it beyond what it is. And we need to be particularly mindful as we go into pharmacogenomics and do our -- do the job we need to do of bringing in lots of people from lots of different populations that we categorize in racial terms because that's how we're used to doing it, to make sure that we're avoiding if error of misrepresenting race as a genetic entity. We need to recognize that racial group differences may have many causes but most importantly from the genetics per speck tiff that genetic variations within racial groups is common. And let me just step back and say if we think about pharmacogenomics from an ethics and policy perspective, what are the concerns? Well, I think one of them is -- is that we've got lots of high tho*t -- hypothesis of investigation. perhaps given that some of those variants are for high dose, getting people quicker to effective therapeutic doses, but I think the question has to be researched. We need outcome data. The hypothesis is great. I'm sure at least some of these will work out but our experience from medical research is not all good hypothesis actually turn out the way we think we're going to. So we need to think about that.

I think we need to be very careful and defining and prosecting against risenings. One of the wrists is from ancillary inflammation. I mentioned that APOE 4 has been identified. So you can manage in that imaginary panel that you might put OPOE into the panel and it might predict who's going to have blooding complications and it may be a more significant risk than -- that outweighs the benefit that might be provided so I think we have to think very carefully about that. I think we have to think about the risks of the paradigm, the risks of oversimply fieing the predictiveness of the genetic information and thinking that it's simple. The TPMT example that Jim Evans -- oh, sorry. Can I get this back? The FPMT example that Jim mentioned did undergo FDA review in context of a -- of a drug labeling issue and one of the concerns that came up in the committee that considered that was not the issue that TPMT homozygotes would be identified in giving lower doses of chemotherapy that was dangerous for them. It was whether we would treat heterozygotes so I think we have to look very carefully at those issues and comb, as with any health care intervention, try and be as global as we can about anticipating potential issues and then evaluating them. And then finally I would say that there is a fundamental ethics and policy concern that goes across the board in health care, it certainly would go across the board in genetic innovations and that is, as we find these wonderful new ways that we think will help to personalize and improve the quality of health care can we be sure that everybody who could benefit from them can get them. I think that's an ethical certain. And I want to give a special thanks to my colleagues thanks.

Thank you very much Wylie. On our schedule now we have a little bit of time for discussion of issues. We'll let Wylie sit on the hot seat without the benefit of fellow panel members, so be kind, but I think we ietd fied a fairly long list of LC issues and I don't want to necessarily confine our discussion of LC issues to the narrow topic that Wylie brought up if we have other things that we'd like to -- like to bring up, I'm sure she may have some comments on a variety of issues as well that could inform our discussion. So are there questions?

Kevin?

Again. Thanks very much for the presentation and the issues that you raise. I'd just like to expabd upon one. You talk about about our need to be careful with our genetic reductionism or our geneticization of the issue. I'll also wondering in your experience with these various groups, when -- there's also a -- I might argue there's also the risk of a Jeanettization or a medicalization of risk and benefit overall. So when you mentioned risks and benefits you did tend to kind of be medical but that's I'm presuming because of the examples that you gave. But I was wondering if you could speak more of just that whole idea of defining risk period as a group.

I think the point is well-taken. Actually -- and I agree completely with it. I think with pharmacogenomics we are dealing in a more narrow area usually because we're talking -- on the other hand if we got to a point where there was some genetic trait that allegedly identified people who felt themselves to be perfectly healthy who somehow now needed a drug we'd have to look very carefully at the issue of medicalization and that goes more broadly across genetic risk information.

It was a really wonderful presentation. I had a question related to the issue of, like at mixed populations because here in the states for example people think in terms of African America or -- and you talked a little like in places like Hawaii, those concepts are very different and people really are a mixture of several of our countries in parts of Malaysia and India and that even this kind of clear concept is almost nonexistent and how do -- how can this type of approach, like let's say a medicine that's given for typer mention for blacks her in the states, how is it going to be doing Brazil or in Cuba? Do people take it or not take it or what do you think of this same issue in a culture in which the concept of race is very different from what it is here?

I think it's a very good point. What we can say and I think most of my remarks it's very specific as to how we think about things in the U.S. I think the answer is the fundamental answer is the same and your point speaks to the extraordinary importance of making sure that research occurs in other populations. And that we don't exclude population like Brazilian populations or Hawaiian populations where there may beless clarify about the racial categories that we've been thinking about. The movement should be in the other direction. Drop away from these racial -- and do vary yabt sampling and look to other factors that are relevant.

As always, I learn a few things from you every time. I was expecting to hear about Weidel but it didn't happen in your particular presentation, but as we think about sort of the public health issues and on pharmacogenomics and you being at the previous committees, I mean, you sat where these guys used to sit, trying to give advice to HHS agencies here, I'd like you for at least the next two minutes put that hat on and sort of help Emily and the group sort of think through the kind of recommendations to HHS and some of them, we're already in your slides. I want you to take over first and I'd like to see how it fits in the construct this afternoon.

Well, that's a very interesting dpafrp l. Weidel from my point of view has nothing to do with either pharmacogenomics or differential race response so let me just start by saying that. And that's why it didn't show up as an example in my slide. Let me explain why I -- why I have that view. The components of buy dell have been known to be effective in congestive heart failure which is the indication that buy dell is being promoted on, has been known since at least the mid-80s. THE FIRST SYSTEMATIC OF THOSE TWO DRUGS IN CONGESTIVE HEART FAILURE WAS PUBLISHED IN THE MID-80s CALLED THE V HEFT ONE TRIAL there were some minority participants and it resulted in a 36% reduction in mortal fi for people who got that drug combination compared to people who did not soy don't think there's any question that the components of it work in people other than African Americans. The story of how it came to be proposed as something that might be marketed as a drug for African Americans I think is an interesting story. Jonathan Kahn has written a lot about it and I think it has more to do with the regulatory procedures and the laws around getting past pa dents than it has to do with anything specific to race. But the one other trial I'll comment on is the ashlgs HEFF happened was the researchers who were involved in -- both of which showed the effect of this drug combination and preno, ma'am nantly white populations actually asked for a patent. There's a special kind of patent. I will stand recollected by the FDA folks, and this is a patent that allowed them to create a combination drug even though the two drugs would still be available generically and that was denied. And it was only after that denial that a subsequent reanalysis of V heft 1 led to the suggestion that there might be a great r response among African Americans in V 1. Now, the study was not designed to answer that question and it's merely a pie Pott sis. What was done then was the A he have trial. Enrolled only African American. So it did not adoctors the question of whether there's a differential response. What it did do which I think is very helpful is that it shows that the combination remains an effective therapy. This is for folks that are on standard therapy and the question is whether an additive adds additional benefit and the trial said it did. 43% reduction in mortality, so it was a very striking benefit and I think that's helpful. The trial was only African Americans but I don't see that those trial results say that it only works in African Americans. We have prior data that says that your penoAmericans benefit. I think we have a long history of testing drugs in one greep and then extrapolating those to other groups. I actually think the -- this particular drug combination -- I don't see it as evidence that only African Americans will benefit. And let me make one more comment while I'm on the buy dell story and that is the buy deal's a very expensive drug. The two components remained available. I don't -- speaking from the perspective of underserved patients I don't see it as being a break through. I do think they're helpful and useful and that physicians can use that information when they think about the patient on standard therapy who may need something additional. So that's my comment on buy dell. I'd be happy to answer more questions on that. Getting back to the other point that MUIN raised, the -- what can HHS do? I think the HHS perspective starts with the research perspective and I would say the tremendously important thing here is thinking very care cli about the inclusion approaches, making sure that broad intr*ekss of population is done. I think there's an organize gnat perhaps for trying to study outside of the U.S. and clearly all the ethical concerns that go along with approaching communities that haven't been involved in research and that may have some mistrust. I would say that's a leading issue from more of the clinical integration side I think there's also a research piece and that is how do we fund outcome studies, how do we do post market studies after a degreeing has become available to look farfully. Well, the -- and then I think finally there's an HHS imperative to think about access. In the newspaper article that was referred to this morning regarding Herceptin, they had a price tag something like $42,000 a year and it made me wonder whether poishts in this country who are underinsured or uninsured Herceptin when they need it.

Thank you for your presentation and you've actually -- you've actually answered part of my question, which with your last set of comments but the other part that I think would be really helpful is the understanding that in order to even get where you suggested requires education, now know, and education from -- well, from the student all the way up, so I was wondering if you had any recommendations of how that might occur. And you know what I'm asking is that it's an infrastructure question really and if you have any reck min dagss in that aria or any thoughts I would appreciate that. I'm not sure I have any thoughts that would be very more spefkt than you just made. I think we have to think carefully about how can genetics be integrated into our expectations of school curriculum right from the beginning on up so I think there's definitely a sort of public education curriculum issue that comes up. I would highlight two other issues. We have to sync together how we're going to provide this. I don't need it when I don't need it because I'm too busy and I have other things to do so I think we need to work hard to see what models are really going to work to help health care providers and then finally, I think we have a need for education about what research is and why it matters and -- and what atiernss we can get to all of the people were -- there's a lot of research that needs to be done and probably the most important resource in that research effort is the human participants that are willing to be part of it.

The other part -- I appreciate that, because that's your information. One of the th*er things I guess I was also getting at when one of infrastructure and that is those that make decisions on how that information gets disseminated, those who make decisions on how you structure the research process, why the study them, what populations to study, give information of that, because as you know, many of us who do research tailor our decision based on the grant and the guidance itself comes from the process so I was wondering if you had any thoughts about education, not necessarily about just the public or even providers, but those who are in the process of having to make decisions about what do we focus on because even in training of health care profession, this is something that was in the institute of medicine report on public health, and I know this is kind of outside of what you're talking about, but it to me structurally it fits. Ou do you begin ton even get those

What kind of things would they need to know? Because that's a -- I think it's important as well.

I would say I think you're making very important points and I think they're importantish su shoes that I'm not quite sure the best way to deal with because following the point you're making it seems that one of the issues is how do you convene leadership to think very carefully about how to -- how to help the public in the best possible most efficient manner. I can only say it seems to me that those are tremendously important things.

Thank you, ma'am.

Other questions for Dr. Burke while we've got her at our beck and call? Thank you very much.

Thanks for the opportunity to speak with you.

What I'd like to do now is actually ask people to turn to tab 5 of your books because we don't have slides and I think it's useful to just look this over a little bit. At tab 5 we have a couple things that we had identified. One was a list of key issues as summed up on the basis of discussions at the June meeting and what follows that is then a proposal for an outline of a potential SACGHS report on pharmacogenomics and I'd like to have a little bit of committee feedback on whether this is something that -- if you look at the scope of this potential report, it's a report that's going to be a nontrivial task to write even with some help. So I would like to know before we task the staff to start such a thing that this would be something that we feel would be useful and I'd particularly like to hear not just from the committee but some of the ex officios whether it would be useful to you in your own organizations to have some kind of internally written HHS report on sort of the state of the state right now in terms of what pharmacogenomics is and what are the issues and what are various agencies doing about it already, you know, where are we in good shape and then of course our task is always to identify places where we want to make recommendations primarily about where we see gaps although I think it is useful for us to also commend good work when we see it and to point out to the secretary not just the things that are going wrong, but also the things that we feel are going right within the organization. So are there any comments? I know this is a very extensive out line and I don't know that we -- I mean, we could go through it line by line but I'd like to get some since as a whole feels that there's value in going through the effort of doing a report.

I have a couple of specific comments but you don't want them right now. Right? You want the general idea of whether we should do this.

What I'd like to do is if we think it's a good idea I think we should go through and make comments about things that are missing or need work or specifically if there are areas where we think we might want to make a recommendation, even if we haven't totally formulated that recommendation I think it would be helpful to staff to know that this is a point where wre want to say something and then the task force can go back and do some work.

Yeah. I have a wonderful -- I think it's a wonderful idea and I think it's particularly relevant. I could be wrong but let's say within an agency, there are institutes and people like Rochelle Long. So it's very uneven. There is like, you know, some people think it's very valuable. Others don't think we should be investing in something else and so I think that to give like kind of a department wide like recommendation or like a list -- kind of a consensus of overview, white paper on the field I think would be very useful.

I'm not sure that it will level the playing field. I'd love to hear about them if you do. But I do think speaking for the NIH I think personally it looks very good. I think it's useful as simply of a summary of where things stand and where they ought to be and I think recommendations to the secretary also would be useful.

Any other agencies that have pros or cons that they'd like to chime in with?

I don't have any pros and cons, but I think as we think ability pharmacogenomics and the sublick health implications of that, it's very important to cover the whole spectrum from the bench to the fence, as somebody said yesterday and when I think of CDC and some of the other public health agencies, you don't dodge -- diseases have infectious outbreaks and things like that and you know, drugs and medical errors and drug adverse effects, I mean, with the exception of adverse immunization events because immunization is sort of a public health activity, and -- but drugs have a much more far-reaching implication as far as the health of populations and then access issues on some of the LC, things that we heard from Wylie so I would hope that the committee would take the whole landscape from the NIH research agenda on how to incorporate pharmacogenomics there all the way too the integration of evidence and some of the other processes to cover the FDA issues and then on the services side, just look at the whole department and basically I don't see where you've summarized what we all do. I don't see it here. Maybe I'm missing a page. Do we have some kind of a table that says what the feds are currently doing? I'm sorry. I didn't see the appendix.

There is in tab 5 behind the outline of the potential report is a one-page thing called agency feedback oncopharmacogenomics policy priorities. So this was the first level of feedback that we got. We are still trying to put together a more comp hence i table, but this got done in one night. It's pretty clear that the responding agencies are the ones that have the biggest stake here but I -- I don't want in any way to have the nonresponding agencies unrepresented as we go forward. These are just the ones who chose to respond to our request for what's going on.

I was just going to -- you know, to -- this first document was as it says the issues that the agency thought were most important and the subsequent work we're looking on is on the activities on the work that's going on. So there's a different focus on these two documents we're trying to put together. It's really focused on activities and that's what we can look to to what's actually happening. Not to what this is what we think should be happening.

Speaking from the perspective of -- I put it in the context as of yesterday where we really had the shift from the biological research to the broader context of the public's health and I think again, if this report can be done in that same way of looking up pharmacogenomics across all of these issues and that's the point of it to inform where we need to go for the public's health, I think we'll do the same type of discussion

And again, we agresed the -- our focus on the underserved population, every one finds to be a point of access.

Since we didn't do the reed Tuckson table we didn't get the chance to reiterate there are certain overarching subjects of which ak shes is one. The forum are basically important for all of these subtopics so I think to your point, those things need to be considered throughout any document that we would consider writing. Okay. So I'm going to ask then if people have any specific comments on the outline that they want to make now and also to invite if you don't have any specific comments because you need to think about it a little more that you use Faye as a central conveyance point to send any additional thoughts you have after we leave here today. James and then Joseph.

I -- one of the things I think is very important to have explicit in the report in the outline is the issue of real demonstration and a prodemonstration of pure efficacy. Also whether you include financial issues which I think are important too as we heard today. So for example, under 2D, factors innuancing uptake, I would think the very first thing there, if not, as a separate subheading completely but the very first thing under D though should be demonstration of efficacy because it's what really will drive all of those other things, at least in the long-term.

Well, in a way, but it also -- I mean, again, it should be -- it should be highlighted much more importantly and I think that under improved health status with the outcomes to regard to financial issues which would drive payers et cetera and we simply don't know if those things are going to pan out in the broad -- in the broad canvas. Yeah. Although we certainly hope they will.

Okay. Joseph?

Just on the side, I would agree with James that on -- on that issue, but on the other issue related to this, I mean, one of the things and this is clearly only the conclusion I can make because this is really not an area to me that everybody else is but my concern has to do with we're going to put together a document that has recommendations and be able to tackle some of the bigger issues it seems to me one of the issues is the infrastructure issue that I brought up earlier where the education is taking on the topic being providing a real -- some means by which a real good understanding of issues like things like ad mixture. Things that like the way -- something to the effect of what kind of recommendations can we make that a lot of us allow -- you know, look at the population, research that population or even look at ways the policy needs to be changed to better our service that population. And the focus is going to be not only on providers but also decision-makers in the process.

Let me ask you a clarifying question.

Reporter: I KNOW IT'S RAMBLING.

NO, NO, NO. NOT AT ALL. WHAT I'D LIKE TO KNOW IS CURRENTLY THERE'S SOME GUIDANCE WHEN YOU know -- that reflects the population you're planning to treat. Do you think we need to think about making some changes in the way FDA guides companies to set up the -- that kind of diversity in a clinical trial situation or is it -- I -- I'm just trying to get a little more granular on where you think we should be focusing that issue.

Well, my caveat would be -- I would say yes to that with the caveat that those who understand the functionality of clinical trials look at that recommendation real closely. I would suspect that given in ways that I participated in recruitment and other things related to clinical trials that it seems to me that more attention needs to be paid to that aspect of it than what it all usually is done. It should begin to just -- to focus a little bit more on how population are selected, what gets population involved in participation in clinical trials. Kind of more the basic fundamental things to achieve the integration and those things that are expected to have. But I think those who are much more familiar with that you should look at that bus that's what I should recommend. But in other research I would suggest the same thing. Somehow or another there should be a criteria that requires you to look more closely at the population and that representation not being a faith representation but also a representation of the pipeline because you could have African American groups that you're involved in that but evidence is the heterogeneity within that group. Paying attention to that heterogeneity. It also needs to be evidence of in any kind of application that you have, that you recognize that for what it is. Not everyone does that and I think that was some uniform way that we can make a recommendation that that is done, yes. That's kind of -- that's the direction I'm kind of moving.

Okay.

If that's feasible.

Kevin was next.

I would just like -- we can put this in the recommendations too because I think it addresses a variety of the points particularly in the ethical legal social area, but from what we've heard yesterday and even this morning on the genetic discrimination survey, I think we could put in there that we begin public engagement now on this issue. Why wait? It doesn't have to be on the same size. Your stale is talking about the large population studies because that's -- the impo fews for that is a little bit different, but still, I think some of o the response to survey data earlier people will saying this is the kind of day that we want to have so that we can put things into the bigger context. I think we could start on this too, the same sort of thing. Again, thing that we don't necessarily have to look into.

It's not like we would lay that whole public engagement process out but I think we would get to a point where we said there needs to be a process and whether it's related to, you know, sort of an early alert system from FDA that something's coming that might want to have an education of the public component to it or --

Right. But I'm thinking beyond that say what I've heard with certain interactions with native American groups and groups -- and again not necessarily groups categorized by even traditional sort of categories. Quotes unquote my forty or quote unquote racial but start this broad publication now to perhaps recognize for the first time that there are other such delineations of which we are not yet aware. That we wouldn't become aware unless we trip over them for some reason or another.

So you're talking about not so much a pharmacogenomics

So you're focusing on race is not race kind of thing?

It is pharmacogenomics in a sense because again, what we're talking about, you keep listening to what our targets are and we're always talking about the public h*ement and the benefits to the public t*elt. Well, what's public health except what the public thinks about its health? I mean, why are we always dictating to the public what its health is? I'm sure there were many researchers who were just shocked to find out that some small tribe in the grand canyon wasn't nested in what they were finding out. I'm sure they were stunned. That's the kind of concept chul thing.

That's the different between being an immigrant where you you came from versus --

They want to know why we came home? Right

There is something that could be also the own separate issue which is that if you study a group like if that goes on, if you study a group because of its ethnic capacity and characteristics, in other words, you recruit somebody because they're hyper tensive, but that they're hyper tensive and Chinese and I think that the relations to committed engagement is the least. I think there could be a commute engagement just shortly after that one because I think one would lead to the other, but you know, discussing the pros and cons without studying the different groups, I think there would be like a gap there.

Steve is making faces like he has something to say.

Yeah, I do. I -- I actually think this would be a fascinating report. It would be a pan my, I have no particular interest of any subject that would be in the report but I have a great interest if you step back and ask the question of what you plays off of this and it gets parts of NIH to be back on the same rate. Of course getting FDA and CMS or FDA and CDC to work together becomes even more challenging so I think it would be interesting to step back and actually ask if there's a lot of passion, there's probably not enough money, but there's a lot of money, there's a lot of intellectual capital being thrown at genomics probably in almost every -- every nook and cranny of the department and the request I would have is there some mechanism for better begrating or board nating that activity? And making sure that he's in fine public health, he seems to be headed in the right gr*eks. And one of the things was this outrageous suggestion that reed put on the table that what the department really needed was the equivalent to a drug czar for genetics. But maybe it wouldn't be impossible to have a drug czar for pharmacogenomics to have someone at higher level who actually mapped out for the department as a whole what it was doing, what it should be doing, where it was going. We ought to at least put that question somewhere and maybe -- maybe it's just dumb. I mean, it's culturally impossible or it's financially unnecessary.

Well, all suggestions are welcome at this point. I think what we'd like to do is probably -- and I'm sort of speaking for staff at this point, but to try and ask the staff to go ahead and do some of the things that are out lined here. We do have the offer of assistance, I forget which -- who I'm -- yeah. So you know, we're not going to be doing this alone, and so I'll recognize you in just one second so I just want to make sure that we just all think just sort of nods of heads. Okay. We are interested in your feedback. Please. Can you introduce yourself?

Certainly. I'm sorry. I'm subbing for Frances Chesley. He's my colleague and this is our report which is pretty near and dear to my heart and I wanted to discuss two points that I think maybe useful from at least our perspective. One is as I was quickly leafing through the proposed outline and I didn't see the two that are nearest and dearest to our heart, outcomes research.

We'll write those words down.

Outcome --

Okay.

Maybe we can elevate.

He's right. It's there. And just expand on that for two issues. One is I heard additional demonstration of efficacy and the outcomes that we talk about efficacy are usually learned from the outcomes we talk about in effectiveness and that actually is the biggest area that our agency is concerned with, so to the extent that we can focus on not only identifying what outcomes are useful, but also distinguishing the efficacy from the effectiveness iveness outcomes. When we're Faulking about launching new research project that a lot of our goals but I think they are rather long-term. Existing initiatives of on going collection and all that's required is linking them better and linking better and hopefully uniform standards of information so the clinical outcomes database in an HMO research capture the same kind of data as in another organization. Those are thorny issues so I hope that was going to be in the vote.

I think a point of emphasis in the report is to have agencies which fall under the HHS umbrella work as well as they can among themselves. It's a little more difficult for us to recommend to the secretary working with external agencies but we can certainly point out that there are significant external stake holders that need to be engaged, that this is not just -- it has to involve the significant stake holders that are out there, corporate and, you know, from the insurance world, the -- you know, all the different -- MUIN?

Just to add to what he just said and if you look at section 3D number 3, it shows under surveillance mechanisms needed and this is sort of the presentation you kind of heard from bob Davis a couple of meetings earlier and under 3A you have effectiveness and maybe what we should do and we can help you with drafting the report is add to surveillance mechanisms needed, surveillance and outcome research infrastructure needed because that's part of let's say HMO research brings and distinguishes it from efficacy.

I think it helps to say the word efficacy by itself and to say cost effectiveness which I think is what we mean when we say effectiveness. Is that what we're talking about?

Not necessarily. You can look at effectiveness in the real world in the absence of cost. We're all subscribing to the economic model here but suppose something worked in a randomized clinical trial, would it include people all of the same, you know, you throw it out in the real world where you have three people come in with other preexisting conditions and the efficacy of that drug may not lead to the same effectiveness in the population.

So you're talking about the translation from a small trial population to the general population.

It's a vague word. I think that's what he's referring to.

So this is again, one of the things that the staff is sitting writing madly because we're going to have all these definitions so we know what we have to do so that we can agree on terminology. Do I have other comments that people would like to make at this time? Because if not, we will be having a task force meeting between now and the next -- in fact, we'll have a couple task force meetings between now and the next full committee meeting because I think the next full committee meeting is not until February or something. End of March. Oh, well, we've got time for five or six committee meetings. Anyway, I think there's a lot to do if we're going to go ahead and do this and it's very helpful to the committee and to the task force in particular to have the help of as many brains and hands as we can get to -- to make this happen.

Emily, am I on the task force? I don't remember.

Yes.

Am I?

Sorry.

Okay.

You remember when the report starts to get formed.

On the 15th version of the report that you have to read and comprehend, you'll. . . Okay. So I think there were some other general committee business, but I -- I think as far as I'm concerned we sort of beat this horse to death at this point. I -- I yield the floor to our acting chair.

For the final blow?

Yeah, she's probably going to have another red line or something.

Okay. Do we have -- -- a ten-minute break? Oh. Ten-minute break.

Five-minute break but just to remind everybody again, to fill out the surveys that were in the table folders and to get them to abbey outside before you leave.

Sounds like a plan.

What's up on the screen you don't have, it's an attempt to summarize our discussion and deliberation over the past few days and the first issue that we tackled yesterday had to do with the large population research initiatives, resorsurce and you can just take a brief moment or two to review -- does everybody have this now or not? Just me? Okay. just read it on the screen and see if that accurately summarizes yesterday's activities and the thinking for moving forward.

In that first paragraph, sorry, Sarah. Back in the top paragraph, I -- I understand that it could potentially fit under ethics or even if one wants to talk method logically under scientific but I think the public engagement panel could even be set up there separately because in a sense, it is something new and it has its own kind of method dolling, its HAS ITS OWN SORT OF SCIENCE TO IT. SO WE HAVE SCIENTIFIC PUBLIC ENGAGEMENT AND ETHIC

Joseph?

I'm not sure -- sorry about that. I'm not sure what the first paragraph on the next page is going to say, but one of the -- one of the things that we agree to, I think, was that we would take into account the things that Kevin just mentioned, but going back to the -- if we could, go back to the report or the presentations from those folks and just include them in the conversations that -- I mean, in the deliberations and if we needed as the subgroup or the ad hoc group will then be able to take those in summary and use those as part of the additional work of that task group. I'm not sure if that is exactly reflected in that, but that -- I'm just expanding a little bit more on what was said in that paragraph. It wasn't if needed. It was just that was also something we said we would do. It was more definitive than that.

Is everyone in agreement that we are going to produce some sort of work product, some sort of report to the secretary on the large population research project resource initiative?

I thought we were yesterday.

Okay. So we're going to do that and the task force that already exists is going to lead that charge. And it will include some additional outside members who will help in putting this together.

In the content addition to -- the additional content areas will be those that are covered earlier around the research ethics and public engagement part.

The last item -- the last paragraph here in this section, this goes to the request from NIH for a formal statement on our part with regard to the public engagement initiative and -- well, not all of us were here at the bitter end. There was some discussion about whether that's within our purview, within our charge to do. So I -- it would be good to have everyone's input as to how far we should and can go. I think everybody felt that public engagement was critical and should proceed, but to what extent we can make a statement to that effect and to whom, I think that would be important to nail that down. Kevin?

Wouldn't -- I mean, I understand your concern and I agree with it. So could we not interpret what Francis asked us as would we be against or would we want to postpone or inhibit his looking into it? For some reason now before our report comes out? And we could just say, no, we don't have to then have any charge or whatever in order to -- to proactively tell him to do it but we could certainly say we responded in such a way that we didn't see any problem with them going ahead and doing it as far as anything we're trying to do. How's that?

Any other thoughts? Sounds good.

My sense of it again, just sort of watching you know, from across the web was it wasn't just efforts to engage public sun port for the concept, it was to engage the public about its views of the concept, that is that the public, it's not a question of voting yes or no, but pretty much how the public thinks such a thing should be done, what kinds of questions it ought to be able to answer, the priorities, those kinds of things.

That's right in the sense that it's not -- you don't presume there's support because there may not be.

And one would hope to find out what the on -- opposition is to it.

Okay.

I guess I have no comment on this paragraph, but if you all will remember, a lot of the discussion was about having the public health sort of voice at the table, the convening function, both from, you know, health departments and other take holders so I think you don't want to just inhibit NIH to do what it wants to do which is engage the public, but if you want the maximum bang for the buck, I think these efforts should start early on involving the whole department. And you know, I think that -- that would be a good thing that would come out from this group, but you're not ready to say it now I guess until the final report so you don't want to inhibit Francis from doing what Francis does but your recommendation will go out to the whole department.

That's it. Did you have --

I was going to ask Sarah to get rid of the word study in the second bullet just so we could make it initiative or whatever it's going to be.

Large population research initiative. LPRI.

Maybe we should go back to Hunt's offer of doing that interim letter with the recommendation for the community consultation before the report comes out. Because the report is going to take a while.

But would you do it to advise NIH or to advise HHS.

It would be advising the secretary that HHS should -- I think that was Hunt's offer -- well, Hunt's offer was to the NIH but I'm sure he would expand it to HHS and that would be a quicker document to come out than the report.

Well, are we ready to come up with recommendations in advance of completing the work for the report?

I think it was just a letter supporting that discussions with the communities should begin as soon as possible. It's not -- we're not telling them how to do it, we're not telling them who to ask, we're just -- I think it was just that concept.

Do we need to do that?

Yeah, I don't think we need that. I mean, you can believe -- we heard you. We heard you. We agree and that makes sense to us and we're going to do it even, I can commit.

Any other comments on this section? Does this add kwautly summarize what our deliberations and conclusions were? All right. Let's move on to gene patent. There is an upcoming NAS report. It's due November 9th if my memory serves me correctly and Debra graciously agreed to lead a group and we already have some volunteers and there may be more to follow to review the report when it comes out and advise us if there are any issues the report raises that we should consider.

Are we going to only send that report out to the subgroup or are we going to send that out to everyone for their reading pleasure?

Everyone.

I think everyone can have it, but the subgroup gets the extra duty.

I would like that clear so that everybody's expectations are the same.

Right.

Since I'm on the subgroup I get it either way.

And I guess the question is is Sarah going to buy a copy for everybody.

I've already been in touch with the committee and we'll get copies.

I was going to clarify it has been delayed to November the 17th the last that I knew anyway.

Oh, hot off the press. Any other items to discuss on that section on gene patents? Okay. Coverage and reimbursement, not much to discuss there, just summarizes the fact that we had some edits that were approved today by the committee and the report will be finalized and send to the secretary. Genetic discrimination? The first item, this was ago necessary's idea. Right? Am I remembering correctly to transmit the findings to the secretary. And the second bullet deals with the fact that there were some questions that we had as a committee that we thought might be useful the next time that they conduct their -- their surveys. Joseph?

Yeah. -- the audience was going to be also the public as well as providers and I think that -- I don't know if you guys need to be clarified but I think we would have questions for both audiences that they are looking at.

Right. Kevin?

Shouldn't we also mention that we get an update on the legislation HR -- I'm trying to find it here.

1227.

Right. 1227. Just so that we put that down that we've --

Just in the letter to the secretary you mean or --

No, this is just our notes. Right?

Yeah. These are actions. Next steps though.

Oh okay. I'm sorry.

Sorry.

There wasn't any -- or was there? Does anybody have any action items for or pertaining to the legislation and we can't really. I mean, we're constrained, so -- okay. Anything to add to this section on genetic discrimination? Okay. Pharmacogenomics, our last section. Give everyone a minute or two to look at those bullets.

Everyone in agreement that comprehensive report along the lines that we're similar to the outline that was present in our briefing books will be prepared, the task force will lead the charge. The goal would be to have a draft ready for our next meeting either in March or possibly in June. Any other additions to that? Hearing none, that's a wrap. Our next meeting is March 27th and 28th, 2006. We don't know where it will be yet. Location to be determined.

Cindy, just a question about, you know, would we be looking at other issues for the next meeting to discuss at the next meeting besides what we've listed there?

Yeah, we -- do you want to take some -- do people have time to --

Does anyone have any suggestions? I do think if both of these topics, you know, are ready to come back with them, I think that will take most of the two days. You probably want to have something -- hear something more about genetic discrimination again and perhaps research offered to come back. I think they're going to have additional data in February that would be of interest to the committee so we could consider that but if there's something else that's not on our priorities list or you feel there's something else that needs to come up, remember we're also going to have the patents academy presentation and dills cushion of that so unless -- but go ahead.

The only thing that I was thinking that came up during a few times during our presentations over the past two days was the issue of law enforcement use of genetics and that was one of the things in our original charter that was presented as looking at some of the legal issues on ethical issues related to forensics. And although we didn't have it as a priority area, I wonder, especially in view of moving forward with recommendations on the large population research initiative there were concerns that were expressed there whether that's a topic that we should look at formally as a committee.

Could I just remind the committee that the department of justice is an ex officio to the committee and we are -- the person who was appointed ex officio has left the agency, the department and we've not yet -- a new ex officio has not yet been appointed, but I guess -- well, I don't want to -- you know, it's up to the committee obviously if you want to add that to the agenda but it might make sense to wait until the justice department is back on board and confer with them and maybe have somebody from -- at some point, maybe not March or next year to have somebody from FBI come back.

And there is also legislation that's being considered, I know particularly over on the Senate side, I don't know what the House has done with it pertaining to DNA sampling and genetic testing and its use both by the defendant and by the prosecution and so if we have something like that on the agenda, it might be good to get a briefing from the hill as well on the progress of that.

Just a quick question for Sarah. How -- is there a limit to the appointment process or can justice just keep delaying and not give us an ex officio representative? Is there like a time line or time limit they have to send to us right away?

No, we would welcome them at the table at any time, of course, but I -- and we've been looking and checking and so on.

Okay, thanks.

We do check in with them periodically. We'll do it again.

Any other comments? Do we have a motion to adjourn?

So moved

Second.

All right.