DEAN KATZ: Well, now that our picture taking duties are over. I think we'll begin. Good morning. And thank you all for joining us today. I am Ruth Katz, the George Washington University School of Public Health Dean. It is my pleasure to welcome you to this forum this morning. Personalized Medicine from Promise to Practice.

We cosponsoring today's event with Research America, a group well known to many, if not most of you and certainly throughout Washington. Drawing on the involvement of some 500 member organizations Research America's mission is to make health and medical research a higher national priority.

I am delighted to be sharing the stage this morning with Mary Woolley, president of Research America. I will be telling you more about her and her organization in just a few moments. But I want first to jump into the substance of today's discussion. The concept of personalized medicine has been defined in many ways and Ms. Woolley will shortly be presenting some survey data to show how Americans think about it.

Regardless of the definition used, however, personalized medicine is by its very nature tailored to meet the needs of the individual. Now, at first blush, that seems to be exactly the opposite of what public health is all about. In our field, we are concerned with the needs of larger populations. We focus less on the diagnosis and treatment of disease and more on finding ways to prevent disease in the first place.

But prevention also turns out to have a fundamental role in realizing the promise of personalized medicine. Research America has been in the vanguard for making that connection for policy makers and the American public alone.

One of our goals for this forum then is to stake out a place for public health in that dialogue. And in pursuit of that goal, let me talk for just a few minutes now about some of our ties and how we can strengthen them.

Effective personalized medicine depends to a significant degree on understanding human genetics. By using genetic material for scientific research we can gather critical information about the susceptibility of populations to certain diseases and learn how they respond to treatment. That suggests that understanding similarities and variations within in and across groups is as much a building block of personalized medicine as it is of public health.

The field shares other perspectives as well. Obviously we will not be able to custom tailor medicine exclusively on the basis of genetic makeup. Social context, nutrition, tobacco use and many other environmental factors are also essential determinants of an individual's health status and treatment. It takes no stretch of the imagination to see how all of that fits within the framework of public health.

For personalized medicine to blossom, we'll need a sophisticated understanding of the interplay between genetics and the environment. Studies of the Pima Indians provide one of the better known examples of this.

Here in the United States, the Pimas are acutely vulnerable to obesity and diabetes, suggesting a strong genetic predisposition to those diseases. But the Pima Indians who continue to live in remote regions of Mexico, same genes, different environment are far less likely to face those very same health threats.

That points directly to the importance of prevention. If the Pimas are uniquely susceptible to diabetes in the presence of a high fat diet, we need to do much more than just test and treat individuals who seek medical care.

We also need to put the tools of public health to work, promoting good nutrition and physical activity, offering easy to access screening and reaching out to those without a medical home are the kinds of community based interventions that give us a broader reach.

Here and in countless other situations the value of personalized medicine extends beyond the individual to families, to neighborhoods and to the population at large.

The reverse is true too. Just as personalized medicine can promote healthier communities, public health can advance the science of personalized medicine. To build knowledge, enough people from enough diverse settings must donate the genetic research material that makes meaningful inter-group comparisons possible.

And ultimately individuals will need to make informed decisions about whether to identify their own disease risks through genetic testing. In the field public health we can support these kind of efforts by engaging and educating the community. We know how to communicate so the people of many different ethnic, racial and socio-economic backgrounds will listen.

We recognize that people are most willing to participant in research when they have been involved in setting the agenda and understand how they might benefit from that research. We appreciate the need to balance individual and community concerns. And all of that makes us the central partners in growing the personalized medicine knowledge base and translating that knowledge into practice. Public health professionals also serve the cause by guiding the policy discussions that must accompany the emerging science. Confidentiality of course, is among the issues of greatest concern. Genetic data need to be shared so that the research can move forward, but they must also be protected so they are not inappropriately identified with specific individuals. As well, wise decisions must be made about how or if employers and health and life insurance providers are permitted to use genetic information.

Congress of course, will have to provide statutory guidance here, but health policy analysts can do much to inform the debate. They can also help find a reasonable balance between safeguarding privacy and providing the benefits of information sharing as electronic medical records come into increasingly wider use.

Legitimate issues I think we all will agree might clash as we move forward and we'll need to be thoughtful about how to develop and use integrated technology. Inevitably, important decisions will also have to be made about access to care.

With the health care system already under tremendous strain and the ranks of the uninsured continuing to grow, we need to consider how best to share and pay for the critical advances we certainly anticipate. From the start, however, we must be explicit, absolutely explicit that personalized medicine will not become another measure of health disparities, that it will not create opportunities available only to a very select few in this society.

As the field moves beyond its infancy, other complex policy issues will surely surface. The promise is bold, but the challenges are indeed substantial. If we are to realize the full potential of personalized medicine, we will need the public health perspective, I hope you will agree, at every step along the way. And I hope you will agree too that today's conference is a great way to begin to make those kinds of connections.

We have a very full schedule this morning. So, I want to stop there and now introduce you to the co-host of today's event, Mary Woolley the President and CEO of Research America.

Under her 16 year leadership, Research America has grown in 500 diverse member organizations committed to increasing the nation's investment in medical and health research. Those public and private sector stake holders are involved in basic science, biotechnology and the behavioral critical health services, prevention and public health research. In supporting its mission, Research America conducts surveys, some of which you are going to hear about today, runs education and outreach programs and issues very very important reports.

A widely published and sought after speaker, Mary Woolley is a committed advocate for research and science. There isn't time this morning to list all of her board and committee activities, but they include service to the Institute of Medicine, the Harvard School of Public Health and I'd like to now say, the George Washington University School of Public Health, Children's National Medical Center and the Whitehead Institute for Biomedical Research.

Her writing has appeared in Science Nature, the New England Journal of Medicine, the Journal of American Medical Association and many other research periodicals. As an elected member of the Institute of Medicine and a Fellow of the American Association for the Advancement of Science, Ms. Woolley has often been honored for her work on behalf of medical research. She was named a Woman of Vision by the American Committee for the Weitzman Institute of Science and received a distinguished contribution to research administration award from the Society for Research Administrators.

She has also received awards from the American Hospital Association, Columbia University College of Physicians and Surgeons and the Federation of American Societies for Experimental Biology and the list goes on and on and on.

Prior to leading Research America, Mary Woolley was Executive Director of the Medical Research Institute of San Francisco. She earned her Master of Arts degree from San Francisco State University and a Bachelor of Science from Stanford.

I am delighted to be joining Mary in this very important program today. I'm delighted to welcome you all to George Washington University in particular to its School of Public Health. I know you will have a wonderful, wonderful outstanding group of speakers and great discussion. Thank you all for being here. Mary Woolley.

WOOLLEY: Thank you Dean Katz for that kind introduction and for your leadership of the School of Public Health at George Washington. Prevention and public health has long been a vital component of Research America's mission and with your leadership here, we look forward to taking that to even higher proportions.

I want to thank you, too, for making this facility available to us today. This facility as you can probably see is well suited to broadcast. And, indeed, this program today is being taped for television broadcast to a number of states around the country. So, we know that many people will have the benefit of the good discussion that we'll have this morning.

Well, even though we're here today to talk about personalized medicine, we know that we can't all agree as yet on the crisp definition that will make it come alive for the public. Nonetheless, everyone who's here today is familiar with the concept or you wouldn't be here. But what about the people who are not here? What about the public at large?

We are releasing some public opinion poll data today that I want to share with you and challenge you to think about in the context of the discussions that will follow.

You can see here in our first slide that 33 percent of Americans say they are familiar with the term personalized medicine. Now this public opinion poll is a national poll that was conducted very recently by Charlton Research Company. It's similar in its design to polls we see every day in the media.

It was conducted by telephone with a national sample of 800 adults, with a margin of error of plus or minus three point five percent. Now, this first finding illustrates I think the need for more discussion on this topic. Since Americans level of familiarity shows a relatively low level of immediate recognition.

But 33 percent is actually an overstatement about familiarity because when we asked those 33 percent who said they had heard of personalized medicine -- and we asked them what it means to them -- many simply repeated the phrase personalized -- or a version of it -- individualized. Another 12 percent said "don't know" again. So, just 17 percent or five percent of the total sample mentioned genetic makeup or DNA.

Well, moving on the next question we asked to all those polls included a brief definition of personalized medicine, as the ability to analyze a patient's genetic makeup, allows physicians and researchers to better predict, detect and treat disease.

So, having said that, we then asked people whether they expected the personalized medicine thus defined would have a positive or negative effect on several issues that had previously been identified as important to people when they thought about aspects of personalized medicine.

So, you can see here displayed people's responses and they can certainly say they were positive about all or none of them. Well, most say that personalized medicine holds great promise for diagnosing disease, not as many agree that personalized medicine can help reduce medical side-effects. And people are split as to whether it will increase or decrease health care costs. These are all topics I think worthy of discussion during this forum.

Well, Americans may not know what personalized medicine is, but they do have opinions regarding its components. Sixty-nine percent said they are willing to be genetically analyzed if it would help their doctor diagnose and treat them. So, this is a very personal view.

I think it's also quite positive given concerns that many of us have that people may be shying away from research -- it's positive that 66 percent say that they would be willing to donate genetic material for research even if it might not help them individually.

Although Americans say they are willing to be analyzed 95 percent say they have never had a genetic test to study their DNA, at least to their knowledge, and this is according to a similar poll to our own that was released last month by the Wall Street Journal and Harris Interactive.

It seems to me that it's not surprising to see that people approve of using DNA to identify criminals and to establish paternity. I think what's noteworthy is that they support the use of DNA for research almost as strongly. And you can see the range of other responses.

So, all of this said, it's not surprising that 93 percent of Americans say they believe the science of genetics is a good -- or a very good thing. Americans, however, are worried about privacy when it comes to personalized medicine. And Dean Katz mentioned this as a concern.

According to our poll -- back to our poll -- 76 percent believe that Congress should pass a law to specifically protect their genetic information. I expect that our distinguished panelist this morning will discuss the implications of what Americans know about and want from personalized medicine.

As they suggest, our panelists, what stake holders and policy makers should be doing to help realize the promise of personalized medicine.

We are again thankful and want to reiterate to the George Washington University School of Public Health and Health Services for their partnership today and we also extend a thank you to Roche, who has helped us make this program possible.

Now, I want to introduce our moderator for today's panel, Susan Dentzer. Ms. Dentzer is an on air correspondent for the News Hour where she leads an award winning unit dedicated to providing in depth coverage of health care, health policy and Social Security.

Prior to joining the News Hour in 1998, Ms. Dentzer was chief economics correspondent and economics columnist for US News and World Report, where she worked from 1987 to 1997. Before joining US News, Ms. Dentzer was at Newsweek where she was a senior writer covering business news.

Her work in television has included appearances as a regular analyst or commentator on ABC's Nightline; CNN Newsnight with Aaron Brown; the McLaughlin Group. She has been a Newman Fellow at Harvard University where she has studied health economics and related disciplines.

A graduate of Dartmouth, Ms. Dentzer served on the Dartmouth Board of Trustees from 1993 to 2004 and then became the first woman to chair the Dartmouth board. She's also a former trustee of Dartmouth/Hitchcock Medical Center and currently serves as a member of the Board of Overseers of Dartmouth Medical School.

While this is just an abbreviate bio, you can tell that with her experience, Susan Dentzer is the ideal moderator for our forum today. We are so please you could be with us, Susan. Thank you.

[APPLAUSE]

PANEL ONE

DENTZER: [picks up mid-sentence] about personalized medicine and we've also heard what public health thinks and hopes for personalized medicine. Now, we're going to hear from some experts in this field to talk more about what the state of the art is in personalized medicine and how we go forward. May I ask my panelists to come up now and join me on the stage.

Just to recap, we heard from Ruth Katz about the public health focus on the health of broad populations and how much this can be affected by what we find out -- by personalized medicine. She did an excellent job I think of synthesizing why in the end knowing in the end what's going on with black populations is really an accumulation of knowing what's going on with individuals and personalized medicine will open the door to do that.

As we heard from Mary, one in three Americans thinking that they understand the term personalized medicine -- but when you drill down really only about one in six understanding the actual link there to DNA.

What I'm going to ask this morning's first panel to do is talk about the state of the art of personalized medicine. You see that our title today is "Personalized Medicine: From Promise to Practice." Think about our first panel as focusing on the promise, or at least shall we say, the hope of personalized medicine.

Our second panel is going to get into the issue of how we turn all of this into practice, how we tackle some of the barriers to getting personalized medicine to become a reality, where we need to move public policy really to seize the fruits of personalized medicine.

Just a couple of housekeeping details. Our first panel will have a discussion among ourselves for about the next 40 minutes or so. Then we're going to open it up to Q and A from all of you. We'll do that for about another 15 minutes. Then we'll take a coffee break upstairs on the second floor. Please feel free to go upstairs and avail yourself to the coffee or tea available. Then we'll reconvene again at 11:25 for the second panel. Again, how we get personalized medicine into practice.

Let me start by introducing the members of this first panel. I'm just going to read really their names an affiliations. They have very extensive bios in your packets and I do urge you to read through them. We really do have a first class panel here.

Starting on my far end I believe, I think I can see all the way down there is Dr. Ken Buetow. Why don't you give the old Queen Elizabeth wave, just to make sure everyone can see where you are. Dr. Buetow who's the program director at the Center for Bio Informatics at the National Institutes of Health. Also with us is Dr. Stephan Chanock from the National Cancer Institute, Pediatric Oncology Branch. Head of genomic variation section there.

We're delighted to have with us Dr. Richard Frank, Vice President of Medical and Clinical Strategy of GE Health Care. Also, Evan Jones, Chairman and CEO of Digene. Dr. Bruce Korf, Professor and Chairman, University of Alabama at Birmingham, Department of Genetics. And Dr. Carol Kovac, General Manager, IBM Life Sciences. So, welcome to all of you and thank you for joining us today.

Let me start out by asking each of you to give me your brief definition of personalized medicine. Had you been on that call and gotten a survey of callers saying tell me what you think personalized medicine is, what would you have said. And, secondly, tell me how the entity you come from today is approaching this. Carve out for us why your entity, whether it be NCI, whether it be IBM is interested in this field and why you are in the personalized medicine business.

Let me start with you, Ken Buetow. Your definition of personalized medicine, what you would have said to the survey.

BUETOW: I probably would have hemmed and hawed a little bit, because I think as the other speakers have said today, there's not a crisp singular definition. But the one that best grasps for me I guess would be the integrated use of diverse information to -- about an individual and their disease to prevent and treat the disease.

DENTZER: That's very elegant. Thank you. That's great. And tell us where you're coming from on this.

BUETOW: Well, where I'm coming from is a large scale program within the National Cancer Institute to try to put in place the information technology infrastructure, to be able to perform this integration and on nationwide network to be able to deliver personalized information, both to researchers as well as to physicians and patients and participants in the research endeavor.

DENTZER: And since we're talking about personal information informatics obviously has a premier role in this. Stephen Chanock.

CHANOCK: I would probably also hem and haw. I would probably answer that I think that personalized medicine is really the opportunity to be able to look at the genetic composition of an individual, recognizing that each person is different. But look for the similarities that really allow us to really be able to assess risk.

Because I think personalized medicine is really about assessing risk. What is the likelihood that someone is going to develop a disease or that they're going to have a positive response to a therapy or toxicity to that therapy. So, I think that the key issue in my mind is thinking about risk. Because I think the thing that's so different about personalized medicine is that we've moved past the single gene, single outcome paradigm where we're really looking at most diseases and most circumstances as being a combination of genes. And, so, that's a key issue in this genetic paradigm.

DENTZER: Terrific. Richard Frank.

FRANK: Actually I'm not surprised at the results of the survey, because in the United States the delivery of health care has always been the result of a personal relationship between the physician and the patient. And, so, personalized medicine -- what's unique about that -- and I think the temptation is to equate personalized medicine with genomics, which of course, would ignore the not insignificant component of information technology as has been mentioned already and as I'm sure Caroline Kovac will note.

It's important to take this vast realm of information and convert it into insight. And we need information technology to do that. It's not only genomics as a technology platform that will deliver this information, but also many others.

And as we conceive of what is the definition of personalized medicine, we think through the process of health care delivery. Certainly genomics can help to assess risk and not what to do about it really only creates dilemmas for our patients. Do they initiate a treatment plan or even surgery based on risk even though they have not yet the disease?

The resolution to that dilemma includes a sequence of events that begins with a tailored screening program to look for early onset of that disease and then characterization of that disease in order to gain early intervention.

The early intervention with the treatment most likely to yield a good result that is a combination of good chance of treatment benefit and low risk adverse event. That is not yet the final step. But to follow on with early confirmation of the effective treatments. If the efficacy is inadequate, then perhaps up the dose. If the adverse events are too great, then perhaps reduce the dose. Or even change the treatment. And in this way will treatment become truly individualized for a particular patient or become shall we say personalized medicine -- that individualization being manifest as the best out come for that particular patient and therefore we might concentrate not only on life span, but actually health span, actually improving quality of life throughout.

And as a result, have more efficient delivery of health care across the patient population thereby resulting in broader access to those technologies. So, genomics is a very important initial component to that. The risk leading them to screening early detection, early intervention, early monitoring to truly individualize a personalized treatment for that patient.

DENTZER: So, you're saying personalized medicine is a really a continuum and they begin genomics and the genetic information. But it really has to move essentially entirely throughout the health care system.

FRANK: Yes, that's correct. I would suggest that genomics has painted for us a really wonderful promise and the manifestation of that promise is the sequence of events.

DENTZER: Evan Jones.

JONES: I would have to concur with my colleagues on the right giving quite eloquent definitions here. Had I been asked in the survey, I probably would have reflected I wonder if this is in the dictionary, quickly would have found out it's not and then I would reflected and said, well, personalized medicine today is the application of gene based information to help improve clinical outcomes, patient management and the identification of individuals with disease.

I do believe that it's the cutting edge of modern day health care and it's alive and well. You can think of many examples where personalized medicine is being used today to help improve patient outcomes.

One example would be in the management of HIV patients where not the human genes -- it's not the human genes that are being analyzed, rather the genes of the HIV virus -- help determine the application of any retroviral drugs, which drug is the best in the individual case, based on the resistance to that drug by a strain of HIV virus.

So, there's an example of personalized medicine today affecting clinical outcomes. Digene's business is in the field -- many ways, personalized medicine. We make cancer screening tests where we have a test for the cause of cervical cancer and it's a screening test that's approved by the FDA and there are literally millions of women each year -- get this test and based on the outcome of the test, they can be better managed if they're tested positive or if they're negative they can have the confidence that they're not at risk for disease.

DENTZER: Okay, thank you. Bruce Korf, how do you define personalized medicine. What's the University of Alabama at Birmingham approaching this -- or how is it approaching it and why.

KORF: Having listened to my colleagues on the panel, I would say I would sound like a broken record if anybody still knows what a record is. [Laughter]

I see personalized medicine as the mobilization of information about an individual to help enhance health, prevent disease and if an individual should be afflicted with a disease to improve the quality of the treatment and ultimately to try to achieve a better outcome.

Genetics is certainly one of the main -- maybe the main engine -- of progress in this area. And speaking as a medical geneticist, I certainly have been very anxious to see the message of genetics move forward. But I would also suggest as others have, I believe that genetics is not the only engine of progress here. That I see this fundamentally as information, mobilizing personal information to improve health.

And genetics is one very critical component of information about an individual. But environmental history I believe an important component of it and the ability to collect that information and to make it available across the spectrum of health providers with whom an individual may interact is I think also a very important driver.

The University of Alabama at Birmingham represents an academic medical center that I can tell you is thinking very hard about this. It is, in fact, right now engaging in a planning process trying to do what most I am sure academic medical centers are doing, which is not to extrapolate from the past in terms of how health is provided, but to anticipate the future to try to guess where or to make their best guess of where the field is moving over the course of these next several years. And to try to position itself to provide the best care that it can for the community and to contribute towards the development of health care models.

And I can tell you that personalized medicine, genomic medicine and information sciences are all converging as this planning process goes forward, I believe that will be a very critical component of the academic health model and for that matter the non-academic health model of the maybe next several years.

DENTZER: Terrific. And Carol?

KOVAC: There's always a danger in going last in a panel like this. I think I'll start with the second part of the question first, if that's all right.

Some of you may have been wondering what is a company like IBM sitting on this panel for. And hopefully in listening to my previous panelists, you've come to at least hear their ratification of the role that information and information technology can play.

At IBM we first got engaged in life sciences research in the mid 90's. We were inspired to do that by in fact the human genome project and Dr. Collins is here with us today. And that project inspired it because as we looked at what scientists were able to do specifically in sequencing DNA and the acceleration of the ability to sequence more and more DNA for less and less cost started to generate data of volumes that really required for the first time high performance computing to analyze those data to kind of make sense of it all, even in fact to create that first draft of the genome which was announced in the year 2000.

What we quickly came to realize though is that everyone was talking about the promise coming out of the genome project as being this vision of personalized medicine. And, so, we quickly came to see that it wasn't just about, you know, big massive computing engines, it was really much more around data analysis, being able to do comparative analysis on, first of all genetic data, but the more we learned about the need to develop these sort of markers, if I can use that term, to characterize patients and groups of patients on a more individualized basis, the more we saw that it was not just about genetic information -- I think all of the panelists have said that. But I'd like to sort of stress that, is that I think it would be wrong and a disservice to define personalized medicine as all or mostly about the genetic data.

Truly, I think, if we're going to realize this -- the promise of personalized medicine, we have to integrate many data types and there will be some kinds of personalized medicine that may not involve genetic data at all.

In the six years that have passed since the genome project, we've done many projects to start to build the data repositories for samples and -- and information coming from many, many, many patients working with government laboratories around the world, working with academic medical research centers here in the US and outside the US.

And what I would say is that there's really sort of five types of information that are going to be critical in building the basis for personalized medicine. One is the basic data that's in the genome. What's there to begin with when, you know, an individual is born. And, so, some people call that static genome information.

The next is what happens to that genome over time. So, it's kind of genetic information, but it's really looking at, you know, which genes are switched on, which are switched off in a particular disease state. And that is very, very complex, because with some 20 to 30 thousand genes in the genome, you know to look at all the combinations of turned on and turned off becomes a very complex pattern matching.

The third area that is tremendously promising, but also very challenging is looking at protein data. If we can start to look at very very minute amounts of proteins in the blood stream that could be diagnostic for the emergence of a disease, we will have a powerful tool, not just for personalized medicine but for the ability to actually quickly translate that into benefit in the form of diagnostics that can be blood or serum based diagnostics. Not everyone can get tissue -- in cancer you can often, but in other diseases you can't. Proteins: metabolic products are already being used as biomarkers for the prediction of disease. Think about things like cholesterol as an anticipator of likelihood for heart disease. With personalized medicine we can devolop many more of those metabolic biomarkers if you will to be more predictive from an early stage of disease.

Images are actually becoming used widely. My colleague from GE is probably very glad to hear that. But biomedical imaging is also being developed as a biomarker that can be used as a very early diagnostic and a predictor of disease and personalized to the individual.

And then, lastly, the integration of all of those data types is giving rise to what is really the promise I think of personalized medicine -- is unique signatures that can be ascribed to smaller and smaller groups of patients and correlated then to the outcome -- to medical practices that give the best outcome for those small groups of patients.

To summarize that long story, I think I kind of agreed with all of my colleagues that it's going to be very much about the development of many types of information and the integration of those using this convergence of computational methods with the very exciting progress in the life sciences.

DENTZER: Okay, thank you. I think we heard, as was mentioned, a lot of consensus around the definition of personalized medicine. That although it may start with the genetic information that is only the beginning. The individual genetic content has to be first of all married with our understanding of the disease and the disease process and then pushed throughout the system, so that we are collecting the data, we're analyzing the data. We are identifying the people who can be most in need of it. We can screen people actively. Then we can connect people up with the medical interventions that actually work.

So, that's a big, big process and thank you for basically establishing first and foremost what a continuum this really is.

I want to come back actually to a point that Steve raised. When the human genome project -- when the genome was unveiled, my impression in talking to the public was that to the degree people understood this at all, many thought the book of life. This is like having a deck of tarot cards.

I can now peel off exactly what has happened to me genetically. It's going to be clear and knowable and there'll be a simple intervention, like going to see a gypsy who reads your tarot card, "Do not go to that party tomorrow night." The equivalent genetically would be able to be told to me and that would be that.

I think, Steve, you used the phrase we've moved past the sort of one gene, one disease paradigm. We increasingly understand the complexity of this, the randomness of this. Speak to that and speak for a moment how we integrate our evolving thinking on disease processes into personalized medicine.

CHANOCK: Thank you. I think that this is one of the key issues that the world of personalized medicine has to take on and that is the complexity of any given disease. If we even just take some of our classical genetic diseases like cystic fibrosis or sickle cell anemia where we traditionally thought that just one gene mutation would give a particular disease, we can see that there's a wide spectrum of the disease process in the outcomes in those individuals who have that particular genetic disorder.

And it's really the other 24,999 genes are not falling asleep, but they indeed are in one way or another affecting that particular background.

Now let's move to the general public, where I think the question is how do we begin to really understand the complexity of many diseases that we can map to one particular gene, but rather 50 or 20 or 75 or six.

We're in this sort of very exciting period from a scientific point of view where we're beginning to discover these, you know, sets of genes that are related to whether it's prostate cancer or diabetes or hypertension, some of the very common public health issues. At the same time they're very important issues going at pediatric diseases such as Hirsch Brothers rare disease, sort of rare conditions of the gastrointestinal tract.

So, I think that where we are right now is in this very intense discovery phase of trying to put together sets of genes that we have then have to go back to it, I think my fellow panelists were making a very important point about environmental interactions, which lead us to them coming up with the kinds of decisions and processes that are going to be needed to really advise patients or to give individuals that opportunity.

I would just take an example - such as I'm trained as a pediatrician -- is we learn more about heart disease and think about heart disease 20 - 40 - 60 years down the line. We're now going to be faced with these dilemmas of do we take our one, two, and five year olds who are particularly high risk for certain types of heart disease -- do we take them to fast food places.

And these are the kinds of challenges that personalized medicine is going to bring to the public. Not only what doctor you go to or what test you have, but what things do you do in your own personal decisions that may decrease your risk for certain kinds of diseases, certain kinds of outcomes. And these are things that are certainly coming down the road and I think will be very daunting challenges. Because we're going to have to sort of re-educate the medical profession as well as the public to really think in these more complex terms.

DENTZER: Is the Big Mac intervention appropriate for you given your genomic characteristics. Evan Jones, would you build on that. How do you approach this from your perspective at Digene?

JONES: Sure, well, first just to add on the comments, I think there's a critical integration of diagnostic information and therapeutic approaches that's going to have to come together here to see success.

I've been in this business for now 15 - 20 years and seen a number of areas where, for instance, the diagnostic information is there linking genetic defects with chronic life long disease. An example would be Huntington's disease where I believe the genes were identified from 10 - 20 years ago.

And people haven't moved forward quickly with that, because it's very challenging to think what do you do with that information if there's not a clear therapy for a patient who's diagnosed.

This is a disease that affects young people and I guess lasts in their 20's. I'm not an expert in the field. But the point is people don't want -- industry doesn't want to spend money to make a test like that. And I don't think you see a lot of good therapeutic activity today because of the interplay there.

So, one of the tricks here is going to be to apply these tools in the right areas where we can really affect clinical outcomes in a meaningful way. And that's fraught with challenges. Just as an example, when you think about these long term chronic diseases how long does a clinical trial have to be to -- to find a good outcome and to show this information is clinically useful.

Which has led people to focus these tasks often times in the area of cancer where a patient might be dying or there's a recurrence from disease and you can complete a clinical outcome in a short period of time. Those approaches have their own sets of problems. But it's an exciting area and I hope that helps clarify.

DENTZER: We are developing our thinking in all these areas, obviously, but we already have a sense that personalized medicine may be more applicable to certain diseases than others and you mentioned of course, cancer, Evan.

Let's talk a little bit more about that though. Are we firm in our understanding at this point that it really is going to matter more in some diseases than other. Ken Buetow, let me toss that one to you.

BUETOW: I think we just don't know at this point. I think cancer has a lot of sort of technical advantages as to why we can explore it earlier, partially because of the way cancer emerges and it's identification as being a disease of genes. But I think as we start to peel back the veil on many other diseases, we're seeing that there clearly are strong gene based components to those as well that influence both susceptibility and outcome to these diseases. So, I think it's still too early to say that one disease will be earlier -- one disease this will apply to and another one won't.

I do think we're at the point where it's awfully clear that it's going to be critically important to specific diseases. For instance, cancer, as we have moved forward in understanding it's true definition at a molecular level, unquestionably being able to look at individual patient characteristics, their constitutional genome type, but probably more important in cancer at this point, being able to look at their tumor characteristics is going to be absolutely essential to us being able to have success in treatment and prevention.

DENTZER: And as a result we have the cancer genome atlas under development at NIH. Do you want to just say a word about that for people who are not aware of it.

BUETOW: I am confident that Dr. Collins will also mention it in his remarks later, the National Cancer Institute is undertaking a pilot effort right now that's part of a much larger program to explore -- can we systematically characterize and determine the underlying gene based alterations that are part of cancer.

We are starting with a series of cancers that are both of public health significance and going to begin looking and -- at both characterizing the mutations, the alterations, the gene based alterations that have occurred in cancer, as well as looking at some of these other ancillary types, other characteristics of the molecular components of the disease as well. So, very exciting and for the next three years we hopefully will find very exciting outcomes.

DENTZER: Steve Chanock, I'd like to bring you in on this question, too. How firm again are we in our sense that personalized medicine is going to matter more in diseases or one disease or another or various diseases versus others or as we just heard, do we just not really know yet.

CHANOCK: I would certainly agree that we don't know -- you know the genetic contribution to disease, the access to tissues, these kind of issues are much closer to the surface in some conditions right now than others and cancer would, as Ken said, I think be at the top of the list. But I don't think that that's because necessarily it's a unique entity in that respect.

I can easily imagine the application of this to many other areas of medicine, psychiatric illness, for example and neurological illness. But these have been much more difficult to study and as a consequence it will take us a longer time I think to reach a point where we will have products that are as accessible as they are beginning to be now in cancer.

So, I think that this is going to be a moving target and will remain so for a long time. And this not an area where we are going to be deluged, you know, like an avalanche over night with information that is going to literally change medicine before our eyes.

This is something I believe that will unfold probably over a generation or so there will be some results that will be relatively accessible and already are with us and will increase very visibly.

And others that will take a much longer period of time to accomplish and will only be accomplished if the research tools, the ability to actually get the information that we're talking about mobilizing to study large populations and to do genotyping and to maintain data bases, and so forth. This all has to happen and this is going to be a very I think long journey, a rewarding one, but one that is going to take a significant period of time to bear the kind of fruit that we're all anticipating. But I for one would be very hesitant to exclude any medical condition from potentially benefiting from this approach.

DENTZER: Steve.

CHANOCK: I think that both Ken and Bruce have been right on target. I would just add one thing and that is I think underlying this discussion -- I think for some time people have tried to come up with ideas of how much of a disease is genetic and how much of is it otherwise.

I think we're now at a new age where in a sense we have to rethink that -- think that all diseases are 100 percent genetic and 100 percent environmental. And it's really a question of how those interactions are taking place.

Is it one gene or is it 50 or is it 500 or is it 25 thousand that are important in disease A versus disease B. So, I think we have to sort of move beyond what this field of genetics for a long time used to trip over in asking the question, well, what's the inheritability per se. And these were I think very important tools in ways of beginning to look at these things.

But we are now at this very I think exciting time where we really look comprehensively at the genome and genetic variation, per se. But I think the harder question now that's coming up on the horizon is how do we measure those environmental exposures. How do we really assess what it really -- what really constitutes an exposure per se and how that interacts with particularly one's host or one's sets of genes or what genes are putting it in motion.

I think that that's in some ways an even more daunting task. We have lesser tools for those in hand now. I think there are some very exciting public and private enterprises going forward and looking at this. But we certainly have a long way to go to sort of begin to put together this equation that's really adding up to 200 percent.

DENTZER: Well, you have just laid out why this is such a dramatic information challenge. Not only do we have to figure out what's going on with 20 or 30 thousand genes, which as you said earlier are not all going to sleep just because a few go one direction or another.

We also have to figure everything that's happening in the environment and put those two things together. Carol and Richard I'd love to hear you address this from an IT perspective. How is that going to be possible?

KOVAC: I actually think that there's an important point -- and also listening to the previous comments -- to be made here. Because sort of listening to the panelists I think it possible to kind of look at -- to kind of ask the question -- to get a little confused about the way research has been done in the past.

This all sounds kind of interesting and new but it doesn't sound necessarily like a dramatically different way of doing research. We've done medical research for years on diseases that have genetic root causes. Even that have multi-gene root causes.

We have -- a typical medical research experiment might be done on hundreds of patients. That would be for some experiments a rather large set of number of patients. We've done clinical trials for drugs, devices, et cetera, for long periods of time and a typical clinical trial is run on, say, a few thousand patients. Maybe if it's a large trial it might be ten thousand patients.

When we're talking about the need for information to do personalized medicine and integrate not just what we see in a person's genome and what we see sort of active in a person's genome -- but also all of that other information that we all talked about in our first remarks being necessary to personalized medicine.

When we talk about that if -- and we're dealing with diseases where things aren't what a scientist would call deterministic -- so, although we can have knowledge about an individual's, you know, variations in the genome, if you look at those variations, people who have the same variation won't always get the disease so they won't get it at the same time or with the same severity. Why -- why is that? We understand variability and, yet, you know, some of that can still be good predictors.

To do the kind of research that we're talking about here, it's not the days of medical research that we were doing in the 80's and 90's. To do this research, we need to be able to collect data from hundreds of thousands, maybe millions of patients.

Now, you know, the good news is that we're treating patients every single day and could be, as we treat those patients, you know, getting samples and collecting consents to do research using those data.

The bad news is that it's extraordinarily difficult and a difficult challenge that isn't just a technical challenge, to actually start to aggregate that number of patients and the data and information around that.

And I -- you know although I come from an information technology company, I would say that we have a pretty good handle on the information technology that's needed to do that. I gave a talk to high schoolers not too long ago and we were talking about the wonderful thing that the human genome project was and they didn't get it, because they said it's only three billion letters, you know, that's only three gigabytes. To them that's a small number. It was kind of instructive. I don't want to minimize but the technology that we need to -- you know, to deliver these kinds of data bases is actually there.

DENTZER: So, what's the challenge?

KOVAC: The challenge I think is more going to be in the ability to start to get, you know, aggregate large numbers of patients to manage consents and to also -- I mean I think Mary talked about some of the challenges in her opening remarks. I think it's incredibly encouraging that the general public is very positive on this topic. We need to ensure that they stay very positive. That you know issues around the misuse of genetic information or use for discrimination in employment or insurance does not happen.

Because otherwise we're going to lose those people who are positive about this. But it's a tremendous amount of work and organization. Now, the good news is, is that, you know, places around the world are starting to do this. Countries have national heath services are through their national health services building data banks hundreds of -- I think the UK biobank will -- just launched several weeks ago -- started taking patient data. They're going to I think have a half a million, 500 hundred thousand patients in their data base.

I just came back from Australia and in Australia they're building something called the Australian National Cancer grid collecting data from patients in Australia, New Zealand and sharing that on an information technology based grid.

There are many, many, many projects throughout and here in the United States, we're I think not as far ahead as some other countries. But despite that factor at NIH we have a number of exciting projects to develop. Genome associations and biomarkers but that's going to be I think one of the real challenges.

DENTZER: Let's stay on this topic of what kind of research structures are needed in the public sector, in the private sector, bridging those two sectors to move this forward. We just heard Carol say the US isn't perhaps out in front on this. Other countries have taken the lead. What do we need, where do we move going forward? Actually, let's start with the private sector piece of this, because I'd love to hear, Richard Frank, how you're thinking through this from the GE perspective.

FRANK: Well, just to stick with Caroline's theme I think the magnitude of the problem can be grasped from a statement that that Amy [unintelligible] made at a conference at M.D. Anderson last March that only two percent of the cancer in the United States are actually contributing data to better understand the relationship between their genetics, their expression profiling of their particular tumor and the treatments they receive and the outcomes.

And, therefore, we aren't -- have only two percent of the potential data. And these are for the most part patients who are participating in what we might call adequate and well controlled efficacy studies as required by FDA.

There's a concept of doing studies that are more community based, which are called effectiveness studies and these allow not only to multiply manifold the numbers of patients contributing data into the data base but also the ability to discern the impact of other factors than just the genetics and the treatment and environmental issues, concomitant treatments, concurrent diseases, and so on.

And therefore, as Caroline rightly says it's not just a matter of infrastructure issue based on logistics or even cost it has to do with the American public accepting and industry, finding a way to work with the public sector -- so public private partnerships -- in such a way that we can gather these data in a way acceptable to the American public, they don't feel threatened by their data contributing in such a way that the FDA can accept these data -- so that we don't just do research and publish articles, but actually deliver new products and new treatment recommendations to the American public.

And, so, I think there's a huge opportunity and there are some logistical issues in the way. For example the interoperability among various electronic medical records provided by different vendors.

But the issue really goes beyond that, because the opportunity is huge. The opportunity of pulling data from electronic medical records is that you have an incredibly rich data base there. In contra distinction to what we get today, the best available in many cases is an event data base from the insurance companies.

And all they can tell you is that an event occurred. Someone got a diagnosis, they went to the Emergency Room, they got a prescription perhaps. But if we can find a way to pull data from electronic medical records, then you will know what their pulmonary function tests were and what the glucose was and even some family history matters, other demographics, and so on.

DENTZER: So a key barrier to be overcome in realizing the promise of personalized medicine is going to be a national electronic medical record or at least the moral equivalent of a national record.

FRANK: Solving the interoperability problem will be a key enabling logistical factor, but then we also have to understand how important it is that we actually take individual patient's data in such a way that they don't feel threatened. So, there will be anonymization [SIC], there will be consent issues as Caroline mentioned and then standardization in such a way that data can be pulled from different health networks, different providers in a way that we can make sense out of the whole.

DENTZER: Let's all jump over to the public side of this. Ken Buetow, what else do we need here in terms of creating the optimal research structure to realize the promise here?

BUETOW: I think we need to move -- I think we're in a natural evolution right now that -- that sees the transformation of what has been relatively small scale investigators working in individual laboratories, in individual institution and we need to come up with an infrastructure that facilitates creating larger or virtual organizations, larger virtual research activities that assembles across both organizational boundaries. But I would also argue the other key barrier that we have to overcome is sort of discipline boundaries. We tend to do science in individual discipline focus which has been very productive to date.

But if we're going to truly deliver on this sort of integration that many of the individuals on the panel talked about, we have to come up with ways that genomists integrate their information with [unintelligible] that comes into the clinic that actually has all of these components come together.

DENTZER: Which gets at the issues raised in the NIH road map -- how far along are we getting?

BUETOW: We're making tremendous progress. At the risk of giving a plug for what we're doing -- I mean we're piloting efforts right now to put in place a nationwide network of these interoperable nodes at the National Cancer Institute's cancer centers that have the capacity to bring together these [unintelligible] and at times desperate entities into an integrated whole, using information technology that actually allows us to connect genomics information with clinical trials and clinical research information and deliver it to patients.

But I also want reemphasize the point that there's real barriers to that. They're not just simply -- I don't want to say the technology is easy. It's quite hard if anybody's looking for funding, we still need lots of work for human technology piece of this.

But with that said, the true barriers -- many of the barriers we're bumping into are sociological, the cultural and how do we actually bring this technology to generate this new culture of virtual organization, of widespread sharing. And the connectedness of individuals.

DENTZER: Give us a more concrete example if you would of a sociological or cultural barrier.

BUETOW: Historically people have looked at -- researchers quite commonly have looked at data and information collected in scientific enterprises as their own property, as resources that support their investigations and their careers. And there's been incentives, actually much of the system incentivizes that performance and that sort of pattern of behavior.

DENTZER: Like getting tenure at an academic institution?

BUETOW: Like getting tenure, like getting grants, like getting all of the things that support our basic biomedical research. Infrastructure are all individual based, so maybe good personalized medicine. But they're all individually driven, so we need to drive the culture so that the bio specimens or the clinical trials that I collect are made accessible to people outside of my individual laboratory, outside my individual organization so that the whole can be more than the sum of the parts, so that everyone can build on that. And that's what we're attempting to do in the cancer, biomedical and formatics grid to see a big effort. Get the formal plug in some point -- to try to get that infrastructure in place that supports appropriate sharing with the appropriate protection for individual intellectual property, yet -- and individual human subjects protections -- it still allows the information to be shared across.

DENTZER: Well, if this is an issue in the research sector and the scientific sector, the sharing of information, I can only imagine it's an issue in spades in the private sector where the incentives are all around again, controlling information, patents, and so forth. So, how do we think our way through that, Evan Jones? And to the degree -- Richard if you want to address that too please, feel free to jump in. If realizing this -- the fruits of this means a new commercial paradigm are we going to get there?

JONES: From my perspective the promise here is almost unlimited. There are silos in areas where one company is focusing on another. But there's so much opportunity in terms of diseases that can be addressed that companies can focus in specific areas. There's capital that needs to be put to bear. And those issues can be addressed.

What I'd like to do is just give you a sense of what it's going to take to actually be successful here. And, you know, Digene's out there really on the cutting edge, the frontier of this personalized medicine and I think the lessons that we've learned can help us think through what it will take as we apply the resources and find other opportunities.

And just as an example we've identified the cause of the number two cancer among women. What did it take to get a product like that to market. And it was a ten year journey, just to give you a sense of how long it might take to see all the fruits from this personalized medicine.

That began with long term clinical trials involving hundreds of thousands of women. Getting that through the FDA -- and it's only at that point that you have an FDA approved product that you can really begin to make a difference here. And then you need the medical societies to get behind what you're doing.

Many medical societies that we're dealing with now have indicated that they won't accept the new patient management paradigm unless there's an FDA approval behind it that can give you a sense of how it slows things down.

And then once you've overcome all those barriers, you still haven't talked about physician adoption and one thing we haven't spent much time talking about here is the physician and his role. One of the things you see is physicians are driven today by clinical practice guidelines. Once you've got those converting physicians, driving forward a meaningful indication in its own right can be a five or ten year process.

DENTZER: So, we see by the fact that only half of patients appear to be getting beta blockers or aspirin after heart attacks even though we've known for two decades that was appropriate treatment. It can be a long time -- a longer time even than a few years.

JONES: I think the good news is we're making progress and through panels like this and discussions and collaboration between the federal institutions and industry. We can compress that time and really move forward some of these [Inaudible].

DENTZER: Please, Richard and then Caroline.

FRANK: I think this notion of practice guidelines I'd like to address briefly and then come back to your main question about investing in research. Electronic medical record can be useful not only for gathering data, but for ruling out practice guidelines. And they can be useful not only for ruling out practice guidelines in an educational sense like continuing medical education or something, but they can also be useful in monitoring the adoption of those guidelines and then looking across all the practitioners in a particular health care delivery network -- can determine the impact.

So, what outcomes were achieved prior to introduction of this new treatment guideline and then how do those outcomes improve immediately following, so that EMR can be used to roll out the guidelines, it can be used to monitor the adoption of those guidelines and they can be used to quantify the outcome benefit as a result. And, so, this is a crucial central component of the IT structure.

To go back to your original question I think there are two things that will be crucial in this regard. One of them is to understand what can be pre-competitive in terms of collaborative research and what necessarily remains competitive.

Ultimately we operate in the free market environment. The people invest in things from which they can profit and, therefore, ultimately, it's very important for us to make sure that intellectual property doesn't become a bad word in the United States or an unwelcome phrase, shall we say.

It's important for people to understand that the reason intellectual property was created in the first place was because the alternative is trade secrets. And the whole point is to reassure the innovator that they will have some protection, some potential to develop on their innovation and gain some profit from that. At the same that they agree to share knowledge about that.

So, the publication of a patent is precisely intended to obviate the need for trade secrets to get that information out there and actually enhances competition. It is a very constructive gesture.

So, accepting the value of intellectual property as a basis of competitive R&D what comes before that -- and what comes before that is the subject of public private partnership, certainly. Because what can be seen as pre competitive is something that can represent a platform from which competitive research can be launched.

And I don't know the details because I wasn't directly involved. But the pharmaceutical industry got together in a coalition, put together I think about 20 million dollars and really pushed the characterization of the genome. Because they saw that as pre competitive.

And I'll just give one other example. Today the National Institute of Aging is driving a protocol called the Alzheimer's Disease Neuro Imaging initiative into which the pharmaceutical has pumped another 20 or 30 million dollars.

Why is that? Because they see the knowledge about the disease itself characterized by imaging and in vitro diagnostics and family history and genomics, and so on -- represents a platform from which they can launch the disease modifying therapies that we so desperately need. So, I think accepting intellectual property and then looking for what is pre competitive are key components to answering your question.

DENTZER: We do want to open this up to questions and hopefully answers from all of you. Let me just make a note that we have two hand held mikes that will be circulating. Please try to grab one before you begin you question. It would be very helpful if you would identify yourself by name and affiliation for the benefit of our panelists, as well as for the fact that we are being televised.

So, let me invite people now to come forward with questions and perhaps while we are waiting, Carol, I think you wanted to add a few quick words on that subject.

KOVAC: On the topic of intellectual property and -- I think that we actually do have models and I think there is two models, one on the public side and one the private side that are useful and have really worked in the past.

The first is actually -- on the public side -- is the model of the human genome project which in fact did exactly what, you know, Ken was talking about -- is it really took down the barriers of -- and many of the -- it wasn't just about the technology. It was really about creating kind of the sociology of a new kind of research where researchers would contribute their data before publication into this sort of massive repository that took years and years and years to do the analysis.

And, you know that was really sort of a masterful -- and I think quite novel approach to scientific research that is a tremendous model, that we need to follow more in terms of making -- making all of this kind of information accessible to more researchers.

On the private side, you know, I completely agree with what Richard is that, you know, it is important because we are operating, you know, in a free market system. We do need to, you know, recognize that -- that allowing companies to create profitable businesses and enabling them to create profitable businesses, in fact helps to accelerate research while the National Institutes of Health spends an enormous amount on health research in this country.

In fact private industry spends even more than that by a substantial amount. And, so, we want to continue to -- to have those dollars for research and discovery being invested by companies that have a duty to their shareholders.

At the same time, I think this idea of pre-competitive information is very important and many consortia have been created to create that. The Snip [PHONETIC] consortium I think that looked at the single nucleotide variations in the genome was probably one of the earliest.

But today there's some very exciting ones being organized by the National Institutes of Health foundation which is bringing together private sector companies and public sector research to really create even more of these powerful pre-competitive data sources around genome associations. So, some very exciting work happening.

DENTZER: Let's take a question in the audience. I believe we have one.

QUESTION: My name is Jeff Allen with Friends of Cancer Research and I thank all of you for some great ideas and some of the overarching themes that you brought. My question is a little bit with regards to what's being discussed public private partnerships.

Friends of Cancer Research were involved with the foundation of the NIH and putting together the FDG pet project with Non-Hodgkin's Lymphoma. And I think that the reason I bring that up is among Dr. Frank's examples of the imaging technique, these are kind of things that we see moving forward at a fairly fast pace. With imaging then, my question is, is there a way that we should best prioritize some of these. We've heard some of the overarching themes but is this going to be a mass movement on all fronts or do you think from your different perspectives that prioritizing things like [unintelligible] patients selection for clinical trials are just overall drug benefit markers is the best way to go as a short term way of advancing.

FRANK: I'll start. I think it's important to recognize that just like twins don't die of the same disease and they don't live the same length of time, despite having presumably identical genetics so also do tumors with identical expression profiling -- not respond necessarily to the same treatments.

And this is the analogy to the environmental effect on the -- comparing the two twins -- systems biology is different between the two tumors, even though they have the same expression profile.

And, therefore I think the most direct answer to your question, what to prioritize -- it's important for us to take the best guess about what treatment this person should receive based on their expression profiling. But keeping in mind that they may different systems biology.

It will be important to follow up as soon as possible after the first course of therapy shall we say and whether it's in the gene or some in vitro [unintelligible] detection, somehow get a handle on whether that tumor is responding to that particular treatment.

Maybe they need a higher dose or shorter drug holiday between courses of therapy. Or maybe they need a lower dose because of the onset of some life threatening toxicity, as you alluded.

And, so, I think the single most important component for the imaging opportunity is to develop imaging diagnostics for each of the seven processes that distinguish cancer from normal tissue and, therefore we should be able to image not just metabolism as we do now with FDG, but also we should be able to image self proliferation, for example with FLT. We should be able to image epitasis, angiogenesis, and so on.

And by this mechanism we can have a formulary of agents that immediately characterize the treatment benefit using the tracer which is most appropriate for the system's biology of that tumor and for the mechanism of action of the treatment.

CHANOCK: I think that I would take a little bit broader view as well of your question and about the question of prioritization and I think we have to recognize that we are in a very exciting time where we have a tremendous amount of tools in front of us that we didn't have five years ago. And, so, there's been -- you know, with this exponential increase in information, there's a tremendous opportunity. But we have to be careful that we don't want to just sort of wipe clean the slate of things that had been coming along so far.

I mean, for instance, in the NCI there's been a tremendous investment in consortia groups and following perspectively [SIC] almost a million individuals across the world, looking for different types of cancer.

And I think the same sorts of things are going on in other institutes in the Heart Lung Institute, looking at the Framingham Study.

So, you really have pockets of expertise that you want to build upon and to be able to most important I think sort of break down the silos and have experts from different fields coming in and utilizing current resources while we really consider the best way to build the largest possible resource itself.

So, I think the prioritization is going to be partly based on the availability of things at the current time. But I also think we also have this sort of fundamental schism in science where you have, you know, investigator driven very creative questions and answers from particular individuals and then you have sort of more managed centralized studies, you know, run at the top which are crucial for particularly I think a lot of these large population based studies. And I think we have to maintain a balance in being able to be sure we have the best and the brightest coming up with creative new ways of looking at old data, or for that matter, formulating new ways to develop data per se.

So, I think the prioritization is probably the complex mix of using what we have in figuring out how to build new things, but not, you know, go in one direction completely or the other.

DENTZER: We have another question here.

QUESTION: Hi, Kate O'Brien from Roche Pharmaceuticals. And I appreciated hearing your definitions of personalized medicine and in light of the survey results that we saw, I'm wondering how do we get the message across to the public and how do we get the message across more specifically that it's not just about genetics. And also how do we avoid over promising what personalized medicine can deliver?

KOVAC: I have a suggestion about some of the answers to those and I think that the best way to get the -- get the message through is in fact to use examples of the benefit and real examples that people can understand.

DENTZER: Give us one, Carol.

KOVAC: Well, we have a few of them. The one that's always used is herceptin, which -- but I think it is a good one because it does help people to understand how does it work. I'm a woman, I have breast cancer, I get a genetic test. My doctor recommends that. Comes back to my doctor, my doctor says here are the results of your test. Now, you should, you know, take this medicine, because we now know that this is the right medicine for you.

And, so, it's a way to help people to kind of understand how it fits in the context of the normal medical care that people are accustomed to. I think it's also a way for people to understand why is this a benefit.

In the past when we didn't have herceptin, you know, you might have been in a -- you know minority of women who -- and your doctor would have been kind of flying blind with the best information that he or she had.

But now they can know more. Now they can become -- and this is a very powerful drug for you in a disease that -- where you know acting correctly really really matters.

So, I think the benefit is part of -- you know -- of the way you could communicate that example. I think you have to be able to talk in those examples about what happens to those data, how are they kept private and secure, et cetera.

And, so, I think that the more we have -- and then I think it also goes to your question about how do you keep it from being hype. Because -- and then you can say look today we have only a few of these that we can point to.

But with every passing day researchers in both universities and government laboratories and private enterprises are -- you know, working away, investing and this is why we need to do more. I think that's the story that we need to tell to, in fact, you know, continue to advance the desire to invest in this area and to -- and to move it forward. That would be my answer.

DENTZER: Bruce, do you have a word on how we speak to the hope, but don't get into the hype.

KORF: Well, you know, I think there is a huge goal in education and I would say it's not only education of the public, but also education of the medical profession. Most physicians in practice today were trained in an era where none of this was imminent and where the kinds of technologies we're talking about were not really in the curriculum at all.

I, however, am optimistic that over time, this will sort itself out. I think there has been attention towards professional education, first of all. I actually think we're going to see a generation being brought up in an educational system that will begin to integrate many of these concepts from an early point.

And over time this really will sort itself out. But at the end of the day, I think what's going to sell this is the combination of the I believe overwhelming benefit of many of these advances and the sort of priming of the population to make them realize through thoughtful and sober discussion where the pitfalls are and where the advantages are. But I actually remain optimistic that this can be done. But I think it's going to take some period of time.

And it's just about impossible to completely control the hype. And I think to a certain extent that occurs when there is a general perception that something very important is going on, but lots of people don't know exactly what it is.

And I think it's the responsibility of the scientific and medical community to do the best job it can with whatever outlet it finds to give a balanced message.

DENTZER: Let's take one more question.

QUESTION: I'm Lee Herring with the American Sociological Association here in Washington, D.C. I have to say I debated coming today, because I figured well it's going to be a bunch of geneticists talking about genetics only in personalized medicine. But part of my motivation was the fact that I know the Research!America better than that and Mary Woolley doesn't know this yet, but we have our application for membership on file and it's coming to you soon.

So, I was very pleased just to say that I heard the words social, I heard the word sociology a number of times from a number of you today and I don't think Dr. Buetow was just saying sociology in a casual sense. I think you were talking about actual, structural, real life social phenomenon that happened and are identifiable and can be modified.

And, so, I wanted to just make two comments and if anyone wants to reply that would be great. One is Ruth Katz had referred to -- she sort of implicitly said that we don't want to make personalized medicine out of the range -- sort of implied that it was an economic range of some of the population.

The implication there is people of lower SES or minorities or whatever -- you don't want to increase or aggravate the health care disparity situation which is already pretty bad. That can happen a couple of ways. It can be that medicine is simply financially out of the reach of the people or it can be the way the research itself is conducted.

And the former President of my Association -- Troy Duster -- some of you know him -- made a plea in Science Magazine just last February a year ago -- in Science Magazine -- about this kind of issue. If you only look around the lamp post where the light is those are the genes you'll find and those are the ones that will correlate with race. But this group seems to be more sophisticated than that. So, I'm pleased with that.

And the other point I want to make is that we have experts in sociology who specifically look at organizational culture and in fact the most recent prominent example is Diane Vaughn [PHONETIC] working with NASA right after the Columbia Space Shuttle disaster a couple of years ago.

And looking at the NASA culture that sort of contributed to that and apparently made some major inroads in helping overcome those kinds of organizational problems. Anyway, thank you very much.

DENTZER: Well, panel, you've been congratulated on your sophistication. Thank you for making the point that the sociological perspective is as important here as anything else. We are going to try to stay on -- do we have one more question here, okay. Two more. Okay, let's take two more quick ones.

QUESTION: Derrick Skols [PHONETIC], with the office of Senator Kennedy, I was just wondering what policy issues you think is most important to be addressed by Congress to bring personalized medicine from promise to practice?

DENTZER: And just let me emphasize again, that will be the topic of the second panel, as well, so we can certainly feel free to start scraping this big surface now. Go ahead, Steve.

CHANOCK: Oh, I think that there has already been some legislation put before Congress addressing the question of protecting genetic information. I think a crucial step I think in terms of being able to really assure the public that study and propagation of this information can be done on a safe and very suitable way. So, I think that's a very key issue to look very closely at.

DENTZER: And we saw -- I think 70 plus Americans saw that that would be to protect the privacy of their information.

KOVAC: I would echo that and I would also add that the role of the FDA is going to be very critical in bringing new personalized drugs to market.

And we need to appropriately support the work of the FDA to better understand how to utilize these personalized biomarkers in the regulatory process in the approval of the drug.

JONES: Another area that's critically important is to insure the quality of the diagnostic information that ultimately is going to help move personalized medicine forward. This is an area where there's a lot of debate today.

There are rules where clinical labs can basically create their own test and self-validate them. They have a much looser regulatory oversight than industry, where the hurdles can be quite significant. The benefit of the hurdles sometimes is, of course, you end up with generally a higher quality result and product that can be diffused on a large scale. On the other hand, it takes more time.

And, so, helping to evolve the clinic, the regulatory framework here in the United States as we look at this information, to balance the pros and cons of those two competing interests is going to be very important.

FRANK: My focus ultimately is always on the delivery of products to the marketplace that can actually help patients. There's never a publication in the history of mankind that ever diagnosed or cured a patient, it's only the products that we deliver to the market place that do that. And I think that the Food and Drug Administration over the past has shown a remarkable willingness to participant in that process, maintaining safety of the American population and, yet, accepting their crucial role as gate keepers, allowing new products into the marketplace.

And I think in this regard it's crucial to keep in mind that it's not only the data which are delivered to FDA as part of an efficacy clinical trials program. But there's also a huge potential in the post marketing environment that continue to monitor the new diagnostics and treatments not only to insure that they're safe as they go into a broader population with other diseases and other drugs on board, but also, for example, as Caroline mentioned to confirm the validity of biomarkers that may have been used and -- as part of the approval process.

So that when the FDA allows a new product into the marketplace that's not the end of the story. There's an ongoing assessment of safety and efficacy and validation of biomarkers which ultimately will serve the American public very well.

DENTZER: Great. Okay, we're going to hold that final question over to the next panel, really because we are going to take a break now. Let's just briefly summarize what we heard from this excellent panel already this morning.

First of all we heard that personalized medicine is a continuum. We heard very importantly that it's a continuum that's going to reach into not only individuals but across the whole public health sphere.

We heard that the beginning of the continuum is not just genetics it is the new paradigm that I think Steve espoused which is that all diseases are 100 percent genetic and 100 percent environmental. A terrific way of putting it. That basically that is the starting point for personalized medicine.

We heard that there are going to be many barriers to overcome in order to realize the fruits of this. And just again to summarize, briefly, we heard about the need for exist -- to take out the rudiments of what we have now in terms of electronic medical record and grow that nationally.

We heard about the need to beef up the Food and Drug Administration, because we heard from Evan about the importance of FDA approval. We heard about how difficult it's going to be to roll this out into physician practice.

And we heard very importantly just at the starting end the importance of breaking down silos that exist in existing research in order to move this field forward as quickly as possible. I think those barriers issues are a perfect set-up for our second panel. So, let's take a break now, grab a cup of coffee upstairs. We will reconvene here in 15 minutes. Join me in thanking our panel for a terrific initiation to our program this morning.

[APPLAUSE]

[END PANEL ONE]

PANEL TWO

DENTZER: [picks up mid-sentence] be put in place the promise -- realize the promise of personalized medicine. How are we going to achieve that? What do we have to do to get the job done?

And we heard many inklings of the barriers ahead of us from the first panel, again, ranging from lack of an electronic medical record, lack of sufficient resources, on the research side, on the FDA side, for example.

The need to go undergo some very important social and sociological changes to break down silos that have grown up in the research arena over the years, to forge more research partnerships and more public private partnerships on the pre competitive level.

And of course we also hard about some changes that may be have be undergone in terms of how -- not only how we collect data, but how we basically use that data through the FDA approval process and also what we expect the firms to show as we think about very long term studies that may be necessary to prove the efficacy of certain products that may come out of the personalized medicine arena.

So, with all of that as back drop, let me introduce now our second panel. Again quickly, you have their bios in your folder but I will simply introduce them now by name and affiliation. We're happy to be joined by Dr. Finley Austin, Finley again just acknowledge who you are, Director of Public Policy at Roche; Dan Perry, who is Executive Director of the Alliance for Aging Research; Thomas Wildsmith or Tom who is the Consulting Actuary for Hay Group; Dr. Philip Carney, President and Medical Director, Kaiser Permanente Mid-Atlantic Region and Dr. Janet Woodcock, Deputy Commissioner for Operations at the Food and Drug Administration; and, Stephen Ubl, President AdvaMed.

We are going to start the second panel by asking each of the members to tell me what do you think is the single biggest public policy barrier to the personalized medicine basically coming to the forefront. And, secondly, tell me why your particular organization cares about that, why you're concerned about this particular public policy barrier. And let me start over with you Finley.

AUSTIN: I think from all perspectives one of the biggest barriers is making sure that the public policy is science based and science driven. And that it looks at paradigm, not one particular technology or field within the area but looks at the paradigm as a whole. That this is not just a formal co-genomics based issue; and, therefore, what's necessary to foster the research.

And I think also that -- because we feel if you don't do that, if you don't take that science based perspective and don't have a realistic handle on what the science can and can't deliver, you're going to end up with policies that aren't going to be very effective.

And I would also echo that on the other hand, there's a lot of ideas that have been thrown out there around incentives, economic barriers. But to date there's been very little work to really step back and take a serious look at some of these issues.

And I think one of the things that the morning panel made clear in some instances we do have time. Some of these things we don't have to rush. Others we're confronting real patient issues today and we can provide better solutions today. And that's where we could be pushing things further in how we do that.

DENTZER: And why at the risk of asking an obvious question, why does Roche care about that?

AUSTIN: Well, our whole premise for being is to make products to help patients and then therefore deliver value to society, which will in turn deliver value to our shareholders, so that we will continue to exist and continue bringing this value, equation forward.

And, you know, Roche is over 125 years old and Roche wants to be around 225 and 250 years. And innovation is at our core in being able to bring this value. So we need policies that support innovation and commercialization.

DENTZER: Okay, thank you. Dan Perry.

PERRY: Let me split this in two. I think the public policy recognition of the power of personalized medicine is really somewhat underdeveloped at this point. We talk about the need to arm the FDA with more resource and NIH, but that -- we need to do that in any case, separate from personalized medicine.

I think probably the point where public policy potentially weighs in and could be a barrier is in the area of privacy. In recent years the concerns about big brother knowing your medical records and your medical history has really tied up policy makers on Capitol Hill.

I think they worked their way through that so far and I think it's very much something that can be addressed if we can lower the anxiety level that will necessarily come as personalized medicine matures.

The second part of that is not what the public policy inhibitions to the field are, but what the -- what I think the big economic barriers -- and that is we're talking about a tremendous transformation from the model of how to develop drugs and market them for the last several -- many decades -- is going to change to smaller markets which raises questions of whether the increased value of targeted medicine, but to a smaller market or patient population can bring the return on investment that allows those dollars to go back into research. I think that's the bigger issue than anything that looms right now on the public policy front.

DENTZER: And again, we can imagine many reasons why the alliance would care about this, but why do you care?

PERRY: Well, we're a not for profit organization that is looking at the swelling tide of chronic age related diseases. Forty percent of Americans have at least one chronic disease and the average 75 year old has three or more and are on five or more prescription medications.

We know that the side-effects that come with drugs that are in the one size fits all category are a tremendous burden and the cost associated with this silver tsunami that's headed our way have the potential to melt down all public and health care financing.

And, so, to be able to manage better, more effectively diseases like the type two diabetes, cancer, cardio vascular disease, stroke and Alzheimer's disease -- absolutely essential if we're going to get through the aging of the baby boom generation in a positive and healthy way.

DENTZER: Philip Carney, biggest barrier and why Kaiser Permanente cares.

CARNEY: You said I only get to pick one, right. The one I would pick would be the electronic medical record. And the reason I say that is Kaiser Permanente -- we have both the insurance side and the delivery side. We heard earlier this morning from people who are more aligned with the research component but there's a linkage with the electronic medical record.

In the past, and I'm one of the people who was taught in the old school, most of our recognition and our treatment of disease was based on observation, judgments, phenotypes. The genomic initiative give us the opportunity to really look at fundamental underpinnings of disease and go beyond phenotypes and as personalized care was talked about this morning.

But I don't think we can get there without the electronic medical record. The New York City phone book -- everybody knows how to use it, but nobody's been able to memorize it. The genomic material is maybe three phone books. We're not going to get there.

The doctor on the front lines trying to remember everything, it won't work. The advantage of electronic medical record on the front lines with individual patients is you can put strategic prompts in to remind the doctors, so that evidence based medicine can be upped beyond the 50 percent that was referred to this morning.

And also the coordination in populations -- how do you look at clusters of genes and link them with phenotypes and then prove in an evidence-based way that you can actually use that material to influence outcomes.

So, I think they're all linked and the challenge from a public policy standpoint in my opinion is two fold: One is interoperability, which was referred to this morning. You don't want to build the Tower of Babble and information highway with a bunch of dead ends.

But, secondly, it's the cost. Ten percent of medical groups in this country are aligned well enough and capitalized well enough to take a step forward on their own and say we're going there and promise to be part of one organization that does have that capability and we actually have --

DENTZER: And are spending several billion dollars to achieve that.

CARNEY: B with a billion, yes. And we now have 8 million members and we're building a data base. We believe out of that data base not only individual member value added, but also we believe there's an opportunity here to advance the frontiers of medicine.

Scale that up to the country and you can see what it might look like. Why does Kaiser Permanente care if we've already got the electronic medical record, then why do we still care? I think part of the answer there is really I believe we're heading toward a crisis in this country with the technology and the things that we can do -- is how can we afford them.

And I think the electronic medical record with full leverage of that tool will help us develop high quality cost-effective medicine and keep affordability within reach of everybody. We already have 46 million uninsured. We have to solve that one too. But without the electronic medical record, I think we sort of -- we're so slowed down in trying to make progress.

DENTZER: Okay, thank you. Steve Ubl tell me what you think the biggest public policy barrier is, tell us and why AdvaMed cares.

UBL: Since I was overlooked, I'll take two, if you don't mind. I'd like to echo previous panelists' comments and try to stretch the definition of personalized medicine. It's not just about genetics and [unintelligible], not just about imaging technologies. But we were talking earlier, it also refers to other medical technologies.

It wasn't too long ago that, for example, there was really a one size fits all prescription for an artificial hip or an artificial knee. And today we're seeing technology really evolve into mass customization. Is the patient a female versus a male, for example? Is the patient an athlete? Is it an early retiree, labor retiree, and so forth?

And, so, I'd like to stretch the definition of personalized medicine to really include the full spectrum of medical technology. I think the two greatest land mines to promise for personalized medicine from our perspective involved the regulatory pathway which was alluded to by Evan Jones earlier. And also something that really hasn't been mentioned so far, which is the payer community and its role in how to personalize medicine test and combinations, how are they reimbursed going forward. So, just a little bit more on both of those fronts.

On the regulatory side, as Evan alluded to earlier, you've got two very separate regulatory tracks for diagnostic test, those that are developed by manufacturers like Roche and those that are developed by commercial laboratories.

And I don't really want to get involved in which one is better. We've got numbers, for example, on both sides of the equation. But I think what is really needed is one clear set of regulatory pathway for these tests to come to market rather than two separate paths. We need one clearly defined regulatory pathway.

The FDA just waded in terms of this territory in the last couple of weeks by asserting its regulatory authority over certain laboratory developed tests. And I think that's a step in the right direction towards getting again to one pathway.

On the reimbursement side candidly the Medicare program has great influence not only on its own beneficiaries but on private payers as well. And the Medicare program really hasn't kept up with the pace of innovation with regard to personalized tests. It's basically a fee schedule approach to paying for diagnostic tests.

It was set up in 1984. It hasn't been updated for inflation for 11 of the last 15 years. And the two major ways the Medicare program pays for new diagnostic tests it either crosswalks a new test to an old test on the fee schedule and pays for it on that basis. Or it allows carriers to set the rates in what's called the gap filling process.

And some carriers do a better job of that than others. So you might have a test reimbursed at a very different rate in Miami than you do in Seattle. So this process really needs to be rationalized and it really needs to be based on the value provided.

We've been working with leading members of the Congress on the committees of jurisdiction to really try to combine the best aspects of the physician fee schedule along with the way procedures are reimbursed in the hospital outpatient department.

So, for example, when physicians get together through the AMA to talk about new procedures and the codes that are assigned to those procedures it's a conversation around the value being brought to that procedure.

We envision a similar process when a new diagnostic test is developed bringing together pathologists, manufacturers, leading physicians, and so forth, to go through a similar exercise. And then combining that with the painted methodology that really gives CMS flexibility to assign a test to what's referred to in the outpatient setting as a new technology APC, where they can monitor the test for a period of two or three years before assigning a final payment amount so it gives them a little bit more flexibility than a static fee schedule that's based on old technology.

So, those are the two land mines that we see that we're very focused on. I think our -- as a stakeholder, our role is pretty straightforward. We represent manufacturers.

I guess I would just close with one other comment, which is the positive role that Medicare can play in facilitating this and why I think we will succeed largely going forward -- is that in a deficit driven health care environment payers are going to demand the biggest bang for the buck.

And that really means focused on sub populations, which sub populations are generating the biggest bang for the buck. So, on the payer side of the equation to the extent that we can look at for example coverage with evidence development which allows CMS to cover technologies they may not have otherwise covered in exchange for registries and other longer term research to inform the process. I think the payer community has a strong interest in seeing personalized medicine succeed.

DENTZER: Okay. Tom Wildsmith.

WILDSMITH: My profession is fundamentally a financial one, so I can't speak to the science. What I can speak to is the way payers think about these issues. Fundamentally they only had two questions, how well does it work and what does it cost.

I think the reason we're the payer community perhaps is not more engaged is not more engaged than it is, is because as the first panel said, they're not that many new drugs or new procedures out there yet.

Our concern, of course, is understanding what's coming, what it means and what it's going to cost. And the biggest barrier for it becoming a more significant part of health benefit plans is really that we don't see that there yet to pay for.

Now, I do need to speak briefly to the issue of privacy and confidentiality. That has been a source of the debate around that has been a source of a great deal of concern to the payer community. Because take it on faith, we are as dependent on information as health care providers are.

We use it in different ways for different purposes, but we are as dependent on information as anyone else in the system. We do not -- payers do not feel that they're doing anything nefarious with it. I think there are more conspiracy theories about what insurance companies may do without information then reality would justify and fundamentally insurance companies and employer sponsors of health plans don't care whether a piece of information is genetic or not genetic, whether a particular procedure or drug or whatever you're dealing with is genetic or not.

What we care about is, is the information relevant, how predictive is it and what you're proposing to do, will it work or not and how much is it going to cost.

So, the greatest concern that the payer community has had with some of the legislation that's been proposed is that it will inadvertently sweep in information that payers have been using for decades. Information that Marcus Welby could have obtained before the genome was seriously discussed and the restrictions will be placed on it in a way that hamper our ability to provide the health benefits at an affordable price to the customers we try and serve.

DENTZER: Give us a particular example of information that they want to keep their hands that they think might be taken from them.

WILDSMITH: Let me set a little bit context. First, people with private coverage roughly nine out of ten will see the current employer and for all except the very smallest employer insurance companies honestly don't care about the health status of any particular person.

The easiest way to think about that is if you think about IBM's health benefits plan. If you know what they spent on health care this year and the rate of health care price inflation and whether there have been any benefit changes, you can pretty easily predict what they're going to spend next year without knowing anything about the health of any particular IBM employee.

It's very, very cut and dry. We do need the information though to determine whether a particular proposed surgery, whatever it maybe is in fact medically necessary, to be able to confidently detect fraud, to manage the health care and to make sure that we're paying properly.

Now set up a little bit of context. There is roughly ten percent of the market that's individually purchased health insurance. No one's forced to buy health insurance and the stock very candidly is way, way expensive. You could buy a good used car for what you might spend on individual health insurance for your family next year.

That makes it a huge financial decision and if I'm sitting thinking about whether I'm going to buy the health insurance or not, I'm going to think long and hard about what my likely health needs are next year.

Because it's a voluntary market and people are so price sensitive, this is one product that's totally different from cars or radios or computers or ham sandwiches in that the person who buys it determines how much it costs to provide that product.

That's not true with automobiles. That's not true with computers. It is true with insurance. How much it costs the insurance company to provide the product depends on who chooses to buy it. So, it becomes important in a voluntary market to be able to price your product, to know who's buying it and what their health status is.

DENTZER: That's been a terrific introduction and I'm sure we'll come back to some of those questions. I do want to move to Janet Woodcock. Janet, the gauntlet was thrown down to you a couple of people ago in terms of how FDA sees this moving forward. Obviously you heard a lot concerns about the regulatory pathway basically FDA being a partner in the development of this, as opposed to an obstacle to the development. Talk about that from your perspective. What is the biggest public policy obstacle? Of course, we know why FDA cares.

WOODCOCK: In my mind the greatest public policy obstacle is the need for committed leadership to move us forward. And the reason is, if you will, cultural or sociological. We have in a practice of medicine -- and here I'm speaking as a doctor not an FDA person. Practice of medicine has been basically empirical and has been intervention oriented. We're Americans, we like to do stuff.

So, reimbursement -- we pay surgeons the most, because they do stuff, they really do stuff to you and then we pay other interventions a lot and we pay, perhaps now, for medications.

Diagnostics, the whole area of diagnostics, which is actually the foundation of medicine, has been neglected in our policy thinking in my mind and our approach to how we practice medicine. As a result, we couldn't do it any other way, because we didn't have the tools.

But we're getting the tools now. Unfortunately, we have these huge institutions set up. We have medicine as it's practiced today, which is intervention and population base oriented. We don't realize -- we'll look back I predict in 20 years and realize just how bad it is doing this population base.

We intervene on people based on very few data points about them. We have that. We have the research engine that is going around. And what we heard from the panel was we're very interested in more mechanistic knowledge bridging all the links. Well, that's not what we need to do now for -- to make personalized medicine real.

We're going to have to get some diagnostic test that make us do better than we're doing now, which is not very well. We're not doing very well in picking interventions for people now. We have the regulatory systems and the reimbursement systems, they are also oriented around intervention.

And all of this, as we've heard -- I agree with everything all the panelists said -- we're going to have to change these huge battleships and move them in a different direction. And it's going to require a lot of leadership and commitment and it requires the vision of understanding the promise that actually -- understanding more about the individual patient in front of you will bring to the treatment of that person or prevention of disease and the benefits both economic and patient benefits and understanding that will bring.

So, I think the main thing is we have to all kind of come together and say we must go in this direction and from a public policy standpoint it must be that this is the direction we're going in, because this is the right thing to do for the public.

DENTZER: So, what you're all saying in different ways is that we have a world that is very well organized to deal with the challenges of the past, but not at all to rise to the challenges of the future.

And given, as Janet just said, that we're really talking about wholesale changes across an array of areas regulatory, reimbursement, et cetera, how do we begin to get this going. And most important how do we engage the people in Congress and elsewhere who will be critical to getting this going when they have the same problems everybody else does.

They've got entrenched bureaucracies, they've got silos, they've got jurisdictional issues, et cetera. How do we begin to push this along. Dan?

PERRY: Well I think the big trends are going to be demographics. We're looking at the oldest of the Baby Boom generation moving on to the Medicare roles in less than five years. When we're finished with that in 2029, there will be some 75 new -- 75 million new recipients on Medicare and with that population comes increasing vulnerability to chronic, life long diseases of aging.

And it's one thing to be taking a medication for something like pain or nausea or migraines that sort of short term and it's another thing to be just in a drug or having an implant in you or something for the rest of your life.

So, the critical, the clinical issues of this unprecedented large older population ultimately is going to force personalized medicine forward. And the second big trend is the increasing demand by those that pay for health care to see real value, real effectiveness, interventions that are proving their worth because they are so specific and that we can recapture some of the economic benefit by fewer lost work days, more productivity, fewer visits to hospitals and physicians because we really are preventing and managing disease, especially those big scary long-term chronic diseases that otherwise are going to pose such a threat.

Those are big trends -- looking forward to others on the panel talking about some nearer term reasons to leverage this. But that I think is the background.

DENTZER: That's helpful and I understand your point about why you think this will happen, why there will be changes, but how? How do we push the process along I think is the question and particularly, to come back to Philip Carney's point, if the electronic medical record is so critical to this -- and we now have had a health IT czar and we're I guess slowly working our way to standards of interoperability -- maybe or maybe not.

In the meantime, as Kaiser Permanente is many individual entities investing very heavily into electronic medical records -- but many others not - I mean, having so much difficulty making headway in that one arena. How can we begin to think about making headway in all these different arenas to push toward the promise of personalized medicine. Phil Carney and then I'd like to hear you Steve, talk about that as well.

CARNEY: I'll put my doctor hat on for the moment. My father was a physician, Northwestern class of 1922. Pharmacology was easy because you had only three drugs. I'm UCLA class of 1971. In my era, you had fuzzy brain scans and if you wanted to find out what was really going on you put a catheter up an artery and you squirted in some dye. And usually there was post operative pain.

Today, you can take a CT or an MRI and it takes more time to get the patient on and off the table than it does to actually take the image. From 64 a slide CTs, you can find out a lot about what's going on.

So, where I'm going with this is change. And you notice that each one of the two segments I told were probably about 30 years apiece. Physicians are pretty set in their ways. But they can change.

And we just went through, as Susan was alluding to, a tremendous change process in our own organization, actually nationally, with eight thousand physicians in our region 900 physicians going through electronic medical record.

Now, when I get in trouble with my computer, I look for a 13 year old. And you know, I think it's very telling -- how are we able to move our physicians over into this new era, many of whom are mid-career and sort of in some sense stuck.

I think what really carried the day and this is something that I think we need to give some thought to -- and that is the patient centered. If this was truly going to be value added for patients, physicians -- you know why did you go to medical school in the first place -- you got to remember that.

I think there's a way to rally the troops around that theme. And, so, I think we need to really figure out how we're going to tell the story of value added -- of better patient care. And, so, speaking from the medical community viewpoint, I think that's the way -- at least in our organization -- we we're able to engage them in change.

And change is a challenge, leadership is key, but at the end of the day you have to get that critical mass of people not only willing to change, but actually executing that change.

DENTZER: So tie this together with what Dan was saying, what we need to say to public policy makers is you've got this silver tsunami coming. You've got these individuals who need this care. We've got to make the care better and you've got to reorganize all these other systems to make it happen. Correct?

PERRY: That's correct and I think Dr. Austin will agree with this, I think what will also trigger this as a live issue in public policy is when the science begins to identify more interventions that can be traced to offsetting the actions of a gene or a gene sequence.

I mean right not we have herceptin and we have very little after that. But perhaps in a very short time, the science and technology will begin to force this issue.

DENTZER: Finley, do you want to comment on that.

AUSTIN: I agree and I can see, you know, we're changing already some aspects of the drug development paradigm. But we're not changing the paradigm. We now have information available to us we just simply didn't have.

But it's -- you look at an animal or you see a human disease and there's an increased protein present in the people with the disease and you don't see that in healthy people. So, I go in and I take a mouse and I knock in the gene and increase the amount of protein in the mouse and sure enough, the mouse starts looking sick like the human.

I give the mouse a bunch of different compounds and find some that make the protein go away and the mouse gets well and I've got proof of concept to develop a drug. And I knew that forward.

Now, as I'm doing this now, I start to say, well, is there a polymorphism in that target in humans or is there even a resource available. Has that [unintelligible] been sequenced in 30 or 40 different people from 20 different populations around the world. Is there a polymorphism in it that I think is going to affect how my drug interacts with that target.

We'll, if the answer is, yes, I can set up an experiment in my drug development program to see if the drug even works when I get into people, will there be variation. And will that variation correspond to their underlying protein or genetic variation.

But most of the time we still don't have that fore knowledge to do that test and, so, to me part of the issue is we need a road map of -- we're operating in a changing environment -- because what we're seeing now is you get a drug out on the market it meets FDA standards, it is a safe and effective drug in the population you set out to prove it works on. And it brings value to that population on average.

And you get it out there -- the more people it's in -- and you start to realize gee this drug seems to work a lot better in some people than other people. Then somebody comes along with a smart idea and says, well, maybe it's this biological parameter. Maybe that's predictive of who it works in.

Now we've go at situation where they could be right. That drug's still bringing the same absolute value to society just to a smaller group of people. You now have a new diagnostics player, if it wasn't Roche diagnostics with the great idea, who comes in to Roche and says, we think we can make this drug work better with this marker.

So, how do you say what's the value that now is apportioned to the original innovator, what's the value of the diagnostics is bringing into it and how do get the payment community to say well you already set the price, we're not going to -- even though you're getting extra -- you're still bringing us the same amount of value, added value, in fact, we don't want to let you change the price that you've already set.

So, I think what we need is a sort of a road map of how you deal with that economic situation, how you put the appropriate incentives in to deal with the fact that product values in health change over the life of the product.

DENTZER: We're going to open this up to questions in just a moment please and once again let me ask you to get a microphone or have a microphone brought to you. And also to introduce yourselves by name and affiliation but before we get there though Janet you were nodding furiously as Finley was speaking.

WOODCOCK: That's what I mean about leadership. I think the public sector is going to have to do some of this early work. This scenario that Finley described is already possible, for example with markers for drug metabolism. There are people in this room who take a drug and some will have ten times the blood level of others.

Some drugs you may have no effect, it doesn't affect you at all. You're a hyper rapid metabolizer. Now we have tests that actually can discriminate and those tests can predict what dose you should have to get an effect, or whether you should even take the drug at all.

But questions arise. What does that mean for the drug, what does that mean for reimbursement. Should the test be paid for. The payers say we're not going to pay for a test like this unless it's demonstrated through evidence based medicine that it actually improves the outcomes. Then they'll pay for it.

Who's going to do all this. That's the public policy question. Who's going to do the proof of concept to show, for example, that testing for drug metabolizing enzymes and making people get more correct doses -- intended doses -- is a good thing and for which drugs is it worth doing.

And actually that's why we're setting up consortia to do some of these things. Because I don't think any -- that the sectors I've describe have been discussed here. There's no given sector that has a motivation to do this. It's actually beyond the scope of a diagnostic company to do that kind of outcome trial usually.

But we need to take those steps in order to show people that -- show all the different sectors. This is worth doing. It is feasible in that we'll have benefits for patients which is where people said that's where the focus really needs to remain.

DENTZER: Let's open this up now to questions and answers from the audience. Let's take our first one please. Right down here in the front.

QUESTION: Francis Collins from NIH. I wanted to follow up, Janet Woodcock, with a comment you just made, because this is something all of us are wondering exactly how we usher this new phase of evidence based personalized medicine to an appropriate standard.

In an ideal world you wouldn't implement any of these new developments until you had proven their efficacy in a prospective, long term follow up, which is going to be tough. Because many of the tests we're talking about are ones that are going to predict risk of future illness. Maybe there's an intervention available. It may take a decade or more before you can figure out whether the intervention actually helped the people at high genetic risk.

Likewise a test that may be appropriate for predicting drug metabolism and whether or not a particular prescription is going to help you or hurt you. May be difficult to do in a short time without investing a lot of resources in a prospective way.

How is FDA going to struggle with this issue of what is enough evidence and can we, in fact, in many of these circumstances accept retrospective data, have a trial and you go back after the fact and do the genotyping and draw the conclusion. Or is that just not going to be good enough as the FDA steps more and more into this area of getting involved in regulation and genetic test, what's the standard going to be?

WOODCOCK: The standard is going to be different in different situations because there's always a benefit risk analysis. Personally, for example I think blood levels are extremely compelling because we approve generic drugs. Okay half the drugs that are taken in this country are generic drugs.

They're approved on the basis of achieving a specific blood level. But the payers have already say, it's not good enough for them. They want to see outcomes. And so have the clinical community. They don't care about blood levels, they don't know from blood levels. They want to see that the patients are benefited.

Prevention is a very -- if you want to focus on prevention you're starting at the hardest end of the spectrum. It's hard to do prevention any way leaving aside genetics, because you're taking healthy people and you're exposing them for a long period of time and you're taking on trust that whatever test you did or whatever risk factor you identified is right, number one. And that the intervention, be it Celebrex or whatever it might be, is less toxic than the beneficial effect it's going to have and that's a big thing to swallow.

But a lot of these are short term. A lot of the generic can be used -- we have to go away from the hardest scenario - let's go to easy scenarios. We can stratify patients based on these markers to interventions. We can look at their risk for recurrence of tumors, for example. We can look at their sub-type of lupus or pulmonary artery hypertension or whatever it might be and then during the intervention trials, which are going to be done anyway, we can see if the stratification has made a difference in treatment response, either for better or for worse.

So, there are a lot of things we can do right now that will -- and add into the current clinical trials that are going on that can show the benefits or lack of benefit of any given genetic or other characterization, be it imaging or [unintelligible] or whatever, that can really move this field along. I don't know -- nobody really knows how we're going to tackle these really tough things like risk stratification for prevention.

DENTZER: Let me try to pull this back a little bit. Because what it sounds like we're saying is that the FDA is going to work really hard in thinking through the issues on the FDA's plate and CMS is going to work really hard looking through all the issue on the CMS' plate.

And the new national health IT czar, when that job is filled again, is going to work really hard in digging through the EMR piece of this. All of these things are randomly going to kind of come to fruition, we hope, over the next two decades.

But if it proceeds along that course and things just kind of randomly work themselves out, are they going to move fast enough in concerts. The first panel did a very good job of connecting all the dots for us, telling us why all of these things were important and how they happen.

But these things all move forward on their separate track, maybe we'll get there in a couple of decades. Do we need to do something to move it all forward faster. And I guess this gets back to Janet your comment about committed leadership. Who's the leadership that pushes this all forward faster out in front of the silver tsunami. Steve, do you have any thoughts on that?

UBL: I guess the only comment I would make is that if the goal is more innovation I would agree with Dr. Woodcock that the diagnostic sector, just the way the infrastructure has been developed -- it's been developed for the railroads and sciences -- the airplanes.

And if you look at typical markers of interest whether it's venture capital company investments, and so forth, the diagnostic sector has been a laggard. And, so, I would say the most important thing you could do in the short term is send a signal to the capital markets that diagnostics is an interesting space. And that innovation will be rewarded.

DENTZER: And what is that signal, how does it get sent?

UBL: Well, I think it gets sent by policy makers taking initiatives like the critical path initiative at the FDA and if Medicare were to make some substantial change in the way it reimburses for diagnostic test, to move away from a static fee schedule, you need to recognize the differences between drugs and devices in this process as well.

I likened the pharmaceutical research model to sort of a big bang theory. You test a number of compounds for a number of years, you find one that works, bang. The device industry very different. It's more like the industry. Rapid incremental improvements.

So, you've got to build a regulatory model and a reimbursement model that recognizes that difference and recognizes the premium that incremental improvements are being made along the way.

The other comment I would make is that you need to look at the balance between pre-market and post market and the evidence levels that are being established. Obviously you have to have an appropriate evidence bar, but I think CMS has actually done some creative thinking in this area in terms of coverage with evidence development where things that they may not have covered at all initially they're now covering under the proviso that a company pursue a trial or registry so you can track patients longer term and inform the process on an ongoing basis as opposed to setting the bar so high up front that these products never meet the market in the first place. So, those are my comments.

DENTZER: Let me come back to you Janet. Janet, do you feel that the FDA is sending that signal that through things such as the critical path initiative -- in case there is anyone here who doesn't know what that is -- say a word about what that is.

WOODCOCK: The critical path initiative is an attempt from the FDA's standpoint to modernize with all these new tools we've been talking about. How we develop medical products, get them on to the market and what information they come out with.

And it became clear after we did a lot of discussion and research about this that diagnostics are going to be a key issue and certainly I've been trying to send this signal that diagnostics are the foundation of medicine. I believe the last 30 years have been the era of intervention. The next 30 years really need to be the era of figuring out what in the heck we're doing. [Laughter] Which is not a very elegant way of doing it.

But if diagnostics -- that's what the foundation of personalized medicine is of distinguishing individual characteristics and predicating your actions upon those rather than on this population base descriptors.

So, the FDA has been really trying to improve these -- the development of these tools and what we're doing is having consortia and trying to get public private partnerships in collaboration to bridge these gaps and to develop the tools because there's no given one party if you're talking about an intervention and a diagnostic or whatever in a drug or whatever, there's no one party that is responsible for any of this. So, they're trying to move it ahead through these public private partnerships.

But I would say people shouldn't despair. I don't think there's any there that we're going to get to. I think this is an incremental thing about improving the science base of medicine, of how we do medicine.

And any steps we take every year, we're going to get better I think or one hopes. So, maybe we won't have electronic health record completely implemented. I agree with what was said that probably the most important part of that is the decision support for the doctors.

The research needs electronic health records -- are going to have to await standardization of the phenotypic data, which is going to take -- which we haven't really started in my opinion.

Even though we don't have the electronic health record, though, we can make some of these steps, regardless and that's what we need to do. Once we get enough momentum by doing these things, then everything I think will come together and accelerate.

DENTZER: Dan.

PERRY: I think one way that the message needs to be delivered is to recognize that we're spending about one and a half trillion dollars on health care now. There is a lot of concern that we're not getting the value for it.

And -- but we tended in our -- in the debates on Capitol Hill and elsewhere to sort of demonize the fact that we're making such a large investment in health care and in health technology and in medical research.

And I think if we can move beyond that sort of assumption that if we're spending 15 percent of our GDP on health care and those parts that I just mentioned, that that's a bad thing and that we have to hammer it down to 12 percent or some other arbitrary figure. I think that would lock us into the current state of technology which misses the promise of personalized medicine and misses the potential for intervening and ultimately preventing of the some of the health problems that people are going to have.

DENTZER: I think we have a couple more questions. Let's take those now.

QUESTION: My name is Ed Abrahams, I'm Executive Director of the Personalized Medicine Coalition and I just wanted to take a brief and minor issue with something Dr. Perry said, when he said that after herceptin there's not many other personalized medicines out there. Later on this year we are publishing a paper called the Case for Personalized Medicine that will document that there are close to 15 products now on the market, many of which, or most of which have been approved by the FDA.

DENTZER: Without listing them all, give us a quick example.

QUESTION: There are a number of them. Not as well-known as herceptin, but there are a number of them. And I think it's very important to keep in mind that personalized medicine is being practiced today. It is here not. It's not just something on the horizon that we're all groping for.

But that brings me to my question, which is do you think that we could incentivize the development of personalized medicine as a public policy initiative on the grounds of it is safer, it is more efficacious by putting in place a reimbursement system that would reward, say, the co-development of diagnostic and therapeutic products or reimburse diagnostic tests at a premium over what exists today.

DENTZER: Steve, you want to take a stab at that?

UBL: Maybe I haven't been clear earlier, but absolutely I think that's critical. The current system is badly broken. It's a very static fee schedule that was structured in the early 80's and rarely even updated for inflation. And we need a new paradigm, a new model, so that there are greater incentives for capital flow into investment and research and for companies to commercialize these products.

We had a forum that AdvaMed sponsored on Capitol Hill a few days ago where even senior CMS officials talked about how badly broken the fee schedule is for diagnostics. So, in order of priority, I really do believe that need new ideas, new paradigms, new structures to reward innovation in this space and to pay a premium candidly for the incremental improvements that are being made along the way.

WILDSMITH: The one thing I'd like to add is that from the standpoint of most payers, they don't want to pay for a concept. They're faced with an individual with a particular set of symptoms. What I think you want to do is demonstrate to them that the treatment course of giving them these three tests and on the basis on those tests giving them one of a menu of interventions driven by the testing is more effective and lower cost than a one size fits all treatment.

And -- because that's really what they care about. And if you can get to that point where you can demonstrate that to them using language they understand, then they'll pay for the results rather than the concept.

DENTZER: Final question, please.

QUESTION: My name is Dewey Bennett and I'm representing the American Academy of Pediatrics. I'm a pediatrics resident at Washington University in St. Louis.

It seems like we were doing a lot of discussing about how to get from point A, where we don't have a lot of information, to point B where we have this cornucopia of diagnostic and prognostic information. My question is, what's being done right now to assure that once we have all that information it's used in a scientific and ethical and appropriate way.

As an example, the medical community doesn't recommend doing a fully body MRI scan for everybody because A, it's too expensive and B because you come up with a lot of clinically unimportant things that actually may be harmful to the patient if they're investigated. And then as far as pediatric medicine goes knowing prognostic information in young children has an impact on the parent's reproductive decisions, on the way the children are reared and all of those lines. I was wondering what your thoughts were on how we handle responsible use of the information once we get to the point we're talking about.

DENTZER: Please, Phil.

CARNEY: A couple of comments. One is evidence based guidelines and then getting people to adhere to them. Secondly is as new technology comes out, it does come out at a rapid pace at times. An example -- I mean our organization has a biotechnology committee which looks at evolving technologies to see what is evidence based enough that we might cover it in advance of sort of mainstream coverage.

We're trying to move upstream a bit on the issue. It is a judgment call. It is a gray area at times. But the effort to do so I think is another anchor to try and get everybody on the same page. So, I think it's a challenge and particularly when you get hit with so much information so quickly -- trying to stay on the right page -- and I think the issue of prompt -- or where are those crutches that realistically we as physicians moving forward into the next era are going to need. We're going to need to build those to stay on the right page.

AUSTIN: I would like to add, I think a couple of things are key in an appropriate regulatory environment, and, Janet and Steve they keep us on the right path. And we have a system where we weigh risk and benefit contextually for products that go through the FDA.

Of course the FDA can't regulate the practice of medicine and when products are out on the market and some people may choose to use them in different ways. That becomes another issue. And that's where I think public education is very, very key.

My grandmamma used to say a fool and his money soon parts. Now, if you educate the public, they're probably not going to cheek swab and pay 500 hundred dollars to send a DNA sample off to somebody whose going to come back and tell them that based on their genotype they should stop smoking and exercise regularly. I'll do that for five dollars. [Laughter]

So, I think these types of -- where you see these less desirable applications, where they seem to be serving much of the societal good -- I think the best way to address those is to have savvy consumers.

DENTZER: We have one more question.

QUESTION: Lee Herring again, American Sociology Association. I'll put on my psychology hat because I worked for the psychology community for many years. And the issue is there's kind of an inherent paradox in prevention and psychologists encounter this all the time. They come up with behavioral strategies that are based on science that cause people to either stop smoking or never start smoking to begin with, just to use smoking as one example. But there's no money in that in the sense that you don't have a drug to market, you don't have a device to build and market, and so forth.

So, I just kind of -- I mean I don't know what the answer to this is either. But the fact of the matter is a lot of very effective cognitive behavioral therapy are just general operant conditioning techniques or just self-monitoring. I mean in an ideal world if all patients were publicly educated about their situation and they were engaging in their own sort of personalized medicine, you'd have a lot fewer drugs, a lot fewer devices because people would be avoiding a lot of things that they should avoid or could avoid and engaging in a lot of things that they should engage in, such as exercise, and so forth.

I don't know if someone wants to comment on that sort of paradox that if you prevent so many things that could be prevented -- I mean in a world where you don't have marketing of tobacco, you don't need to prevent it, it's not an issue. But when you've got that kind of environment where there's a lot of unhealthy behaviors that are being encouraged whether it's addictive behaviors or whatever, obviously prevention can be very effective. But there's a lot less money in that as Dr. Austin just mentioned.

DENTZER: Dan.

PERRY: I don't have anything to say about the reimbursement for the service. But I will note that the older patients are, the more heterogeneous they are and aging is far more than a collection of what's going on in your personal genome.

It is an interaction with the environment like so many other things are. There's more than genes at work when there is a 40 year longevity differential between an Asian American woman in New Jersey and a Native American male in South Dakota, 40 years difference.

And that's not just genes at work, that's obviously environment, sociological, economic, geographic factors and a complex mix of those things with the genes.

So, you're right to point to behavior and behavioral research as part of the larger mix that we have to understand about aging, because that's what really is going to drive the cost and most of the activity for public health for the next better part of this century.

CARNEY: I would make one brief comment. Because I think it is a tough one, on paying for preventive medicine when Americans -- we're infatuated with technology. It used to be in England, if your teeth didn't come in straight, that's the way god made you. In America you went to the orthodontist.

So, I think there are some deep seeded cultural priorities that we have in our society. But I would say that there is hope around certain elements out there. Example, the [unintelligible] score, which are out there, large employers pay attention to them. There's more transparency with public report cards and some of the questions that you get your score on -- for instance, counsel the patient on stop smoking, the documentation of that. So, you can't force people to do the right thing. But you can try.

And if you can get credit in the process of trying in a competitive marketplace and it shows some value added connection for you, some of these preventive measures may get some traction.

DENTZER: Well, we've had a terrific discussion. And once again, pointed very closely to some of the barriers, and some of the things that need to be done to reorganize the way we approach these questions.

I'd like to close now by asking each of you quickly and briefly because we do want to get upstairs to lunch and hear from Francis Collins. Let's take this back now to the public. If there is one thing, one message that you would like to get across to the public to understand -- again, harkening back to our first panel -- not just what the promise is of personalized medicine, but what needs to be done to bring it about, what they have to understand, what they have to support.

We heard earlier about the importance of the public understanding that researchers will need their genetic material. Let's go beyond that though and say, what does the public really need to understand that has to be done, that has to change in order for them to benefit from the promise of personalized medicine. What would it be. Janet let me start with you.

WOODCOCK: I think that the public I've talked to, cab drivers, people all around the country, they understand this already. They don't know what they need to do or that they need to anything. I think the patient organizations, the public consumer organizations need to demand that their policy makers focus on this as an important health issue for them and for their families over the next 30 years and provide a commitment to bringing this about.

DENTZER: Thank you. Tom.

WILDSMITH: Again, I can't speak to the science. But from my perspective what we really want patients to do is taking charge of their health care and talking to their doctors, talking to their health care providers and engaging in that conversation and asking why, what does this mean for me. Why do you think that. And I think that demand is probably the most effective thing the public can do.

DENTZER: Thank you. Steve.

UBL: I think I would agree with what's been said. I think if there's a silver lining to some of the Viox and other adverse event stories in the press that someday we'll have the situation where a patient really gets more engaged, asking their doctor is there a test that will tell me how I'm going to react to this particular drug and demand it.

My daughter's five months old. She had a genetic screening at birth. That was important to me. Someday I'd like my daughter to have adult genetic screening that informs their clinical pathway. And I think to the extent that more and more of these tests become available it is being proactive, it is demanding more in the health care system. It's demanding more of your practitioner and demanding more of technology.

And I guess I would differ somewhat people from what Phil said earlier in that technology is a key part of preventive medicine. But it's diagnostics and it's just not as well understood, so we have to bring the forces to bear the technology can bring in the intervention side as Dr. Woodcock alluded and bring it to bear on the prevention side.

DENTZER: Dan.

PERRY: What I would want the people to know, to feel is do not be afraid, though we will gain vast new powers to be able to modify health longevity, avert disease but do not be afraid that you are going to lose your own personal powers in this or your privacy.

In fact they will be enhanced to the degree that medicine becomes more targeted and more personalized and more effective. So, don't resist and view your concerns and your fears rationally.

DENTZER: Philip.

CARNEY: If you believe that health care is a higher good and I do. Number one, accept personal responsibility for your health. That goes to lifestyle issues, it goes to compliance issues. And, two, support the higher good priority as part of society. And where I'm going with that is -- what percentage of our GDP should we spend on health care.

Well, if it's 16 percent now and if it's going to be 25 percent in 15 years or whatever the number is, is that good or bad. And the answer I think lies in value. And if truly that investment is value added for people and it's considered a priority, that's the type of thing we want to support.

DENTZER: And last word to you Finley.

AUSTIN: I'm going to echo some of what Dan said would be genes are not deterministic. We're looking at bringing in a variety of risk factors. This is one of them that's going to help inform you and your care giver on what's the best approach for you to manage your health.

And this is a good thing, not a bad thing. And that we're still going to be dealing in a world of probabilities, but hopefully what we can do is reduce the error bars and increase the certainty around those probabilities of what's going to work best for you as a patient, as well as what's going to work more broadly in the public health arena.

DENTZER: On that note first of all, join me in thanking this terrific panel.

[APPLAUSE]

[END PANEL TWO]

LUNCHEON/REMARKS BY FRANCIS COLLINS

 

WOOLLEY: I want to briefly introduce our key note speaker. We're quite privileged to have Francis Collins with us here today. I can tell you from the work that we do in public opinion polling at Research!America that Americans expectations for research and its ability to deliver on its promise have not abated.

There are extremely strong high expectations and strong expectations. That is to everyone's credit in the research community and it's also in all of our interest because it will continue to drive public support for research.

The American public also has the highest regard for leadership in research, a theme that we heard when the panel discussions were going forward, the importance of leadership. And there are in my opinion few leaders as respected as Francis Collins when it comes to research which is conducted in the public's interest.

He understands research. His own track record in research is second to none. But he also understands what the public's interest is really all about. He understands and has always made, put a high priority on the ethical, legal, personal and public interest side of the research equation and melding those things. Expertise in science and a real feel for humanity gives him the kinds of descriptions we read in TIME Magazine as the "soul man" if you will. So, it is my great privilege - that means he plays the guitar too and sings, but I don't think he'll be singing for this lunch today. He has promised to take us with him and looking into the crystal ball of what comes next in this field and research for health more generally. Francis, please come up.

[APPLAUSE]

COLLINS: It is a pleasure to have a chance to be at this forum that Research!America organized with so many interesting comments being made throughout the course of the morning and it's quite a task to try to wind that up.

I think many of the challenges of personalized medicine that I would like to touch on have been mentioned in the course of the morning. I'm going to try to focus on a couple of things. One is to point out some new areas of research that I think are going to provide additional amounts of the scientific evidence base that everybody has pointed to, which are critical if personalized medicine is going to become a reality.

It's not enough to sort of think it sounds like a nice idea. We really have to have the under girding of rigorous scientific data to provide the evidence of how to implement this and practice. But I'm also going to point out some examples where we're already there. I don't think we should assume, despite the fact that we're still on the early part of this curve that we haven't started up it, because there are exciting things happening all around us in that regard.

And I will not be able to resist also pointing out a number of the policy issues that are tied up with this that need attention, many of which have been mentioned, but I might highlight a few just the same. And then I'll finish up with a little bit of a scenario about where we might be going if we do this right.

So, just as a matter of a theme statement, I think it is obvious to anybody in the general public or to health care providers or to policy makers that ultimately the idea that we practice medicine in a fashion that doesn't take consideration of the individual is really quite irrational.

We don't think of buying shoes in a single size for heavens sake. So, why should we be satisfied with one size fits all medicine. There are very few other things we think about doing in our life that are generic and so why should our medicine be constituted in a way that doesn't take account of dramatic individual differences.

So, the goal here I think everybody has to agree to is both common sense, it's good science and it's likely to have substantial benefits in terms of personal health.

We argued a bit this morning, the first panel, about what personalized medicine is and we added a lot of different components and I agree that this is more than genomics. Although genomics I think is a significant driver, particularly right now on this agenda.

But some people would say, you know, I had a blood transfusion 40 years ago and it was darned important to me that that was a matched blood unit. So that was personalized medicine what are you guys doing stealing this term.

Well we are kind of stealing the term, but I think we're adding things that are substantial revolutionary influences: bio technology, bio informatics and genomics. Into that you could add proteomics imaging, public health and whole bunch of other things. But I think we all agreed this morning that potentially this really is a revolution in health care and a good one.

I will argue that genomics has played a significant role in this and the fact that we were able to complete the sequence of the human genome in April of 2003 is seen as a major driver of this transformation. But recognize what we had in 2003 -- it was a reference sequence which tells you a lot about the 99.9 percent of the genome that we have in common, but a particular focus now of course, is on the part that is different between individuals and how do we figure out how to study that in a way that plays out in understanding of disease and allows us to practice better prevention, diagnoses and treatment.

So, I'm going to take a diagram here one step at a time. The first thing we need to do with our genomic tools, if we're going to move in the direction of personalizing medicine is to identify the specific genetic variance that are associated with disease risk.

We've done fantastically well with Mendelian [PHONETIC] conditions. More than a thousand of them have had the genes discovered over the course of the last 15 years. And the genome project has enormously benefited that making something that used to be just brutally difficult, now something that a graduate student can accomplish in a couple of weeks time with access to the internet and a decent thermocycler and a DNA sequencer.

But for common diseases: diabetes, heart disease, asthma, hypertension, autism, these are still very challenging projects. And in fact until fairly recently were almost impenetrable and I want to tell you about what I see is really a major change in our ability to answer those questions in a couple of organized efforts that's going to speed that along.

So, how do we find these defects in a condition like say diabetes where you know there may be a dozen or more genes involved, no single one of which is responsible for more than a modest increase in risk and where the environment is obviously crucial.

Well, we really have to understand how to go and find those ticking time bombs that are there hiding in the genomes of all us and you need a strategy that can survey three billion letters and find perhaps ten or twelve of them that are playing an important role in a disease like diabetes.

A major step forward in that regard has been the success of the international HapMap project which was a six country collaboration that I had the privilege of leading and it yielded its data within the last year.

This basically now characterizes genetic variation across all of the human chromosomes at an unprecedented level of detail, allowing you to see not only where the variation is, but how it travels in neighborhoods.

And it means that instead of having to sample every single variant in the human genome, if you're looking for the risk factors for a disease, you can sample a much smaller subset and they serve as a proxy for all the rest. Because these variants do tend to travel in neighborhoods that are closely linked up together.

So HapMap generated the kind of information that we really needed to be able to do a systematic survey of the genome looking for those genetic variants that might play a role in common diseases, something we just didn't have the power to do before.

And there are now several success stories in the early days of application. Perhaps the most dramatic of which is the story of aged related macular degeneration, one of the most common causes of blindness in the elderly. A disease that caused my aunt now to be blind. And many of you know people in their 70's or 80's or even 90's who are afflicted by this application of HapMap even before the project was finished.

But because the data was all being placed in the public domain just as we had done for the genome sequence people started using it and this discovery of compliment factor H polymorphism that plays a major role in disease risk was really quite electrifying.

And within a short time after that discovery, two other risk variants and other genes were identified. And recent studies suggest that when you look at these three variants, they account for something like 74 percent of risk, which is really quite startling given this is a disease that many people did not expect would have that strong a genetic contribution, given that it comes on so late in life.

Don't get me wrong, there are obviously environmental factors, particularly smoking and they interact as was recently shown in a JAMA paper with the genetic risks. But we have gone from really having very little understanding of macular degeneration in terms of its cause to having the major causes now in front of us, which is really quite stunning in the space of just about a year.

And because the genes involved, one of which you see here, compliment factor H, are in the inflammatory pathway, at least two of them are. It's a strong suggestion then that one could begin to mount a preventive strategy for those who are at high genetic risk by implementing a program that's based upon what we already know about how to prevent inflammation. And many such programs are being contemplated even as we speak.

So, we'd like to see this happen a lot. I could tell you other stories recently about using HapMap that have discovered genes involved in type two diabetes, in prostate cancer and in a few other disorders. And one hears rumors that there are going to be a lot of this in the next year or so.

And in fact one reason there's going to be a lot of it is because recognizing this really is a unique moment in history to be able to make these discoveries, several of us have been working quite vigorously to try to be sure that we have the kind of resources in place so that investigators who have collected over the course of perhaps ten or 20 years case control studies with lots of affected people with a particular condition and lots of well-matched controls have the ability to apply this new HapMap based approach to identify those responsible genetic factors.

A particularly interesting and I think ground breaking partnership is represented by this Genetic Association Information Network, or GAIN, this is a public private partnership between the NIH the Foundation for the NIH, who is serving as the organizing entity for the whole thing and the private sector with substantial donations of resources coming especially from Pfizer, but also from Perlegen, Affymetrix and Abbott.

How does this work. Well, basically the goal is to encourage more of these what we call whole genome association studies of common disease where you take something like a thousand cases and a thousand controls and you apply this HapMap based approach to do genotyping across all the chromosomes looking for those variants that are associated with disease risk. The resources that are available through GAIN are sufficient to provide support for seven such studies.

And we went through an extensive peer review process over the course of the last six months with dozens of applications coming in from investigators who wanted access to this genotyping. We have as of a couple of weeks ago -- a meeting in Boston -- made tentative decisions about which case -- which studies to find and that will be announced probably before the end of this month.

And it's a very exciting group of diseases and studies that are going to be supported. In addition, there's another major initiative that will also I think play an exciting role here and which is focused not only on the genetics but also on the environment. People mentioned this morning about how critical it is that we not only develop better tools to identify those genetic risk factors -- and I've just been mentioning a few -- but also that we actually beef up our environmental assessment so that we can do a better job of measuring exposures that may play a role in triggering illness in a susceptible individual.

And that would include better measures of diet and physical activity which currently are still not measured in an ambulatory way as accurately as you would like. So GEI is in fact an initiative which is proposed in the President's budget in FY 07 and if we ever get a budget for FY 07 we expect that it will be in there, because it's been well received so far by the Congress.

This is a project that aims to accelerate both the genetic and environmental understanding of health and disease and it has two components. One is additional geno typing of these case control studies of common disease and particularly looking at health disparities.

And also the development of innovative technologies to measure environmental exposures, diet and physical activity, many of which are going to be ambulatory and which are based on cell phone technology. Because obviously that's a wonderful way to transmit data using an apparatus that many people are already walking around with any way.

I should say that there is much discussion going on right now at NIH about how to handle policies for these large, complicated, potentially very expensive studies that are going to generate very rich and very large data sets that no single investigator can possibly mine through in a short period of time.

And, so, we have with the leadership of Betsy Nabel, the director of the Heart, Lung, and Blood Institute, been engaged for several months in a discussion about a pretty groundbreaking approach that NIH intends to take which is to make data access to these kinds of studies much more broadly applicable.

That would include placing the results of these genome wide association studies into data bases that anybody with a reasonable scientific question can actually access and that they would be able to see immediately, not even waiting for publication.

But would also be a discouraging effort on the part of NIH against inappropriate intellectual property claims in order to keep this in the pre-competitive arena recognizing that patents are critical for further downstream research, but that applying them too early in the discovery pathway can actually be a disincentive.

And, so, if you're interested I would encourage you to weigh in, because we have put out what we call a request for information, asking the public -- and the public in the broadest sense -- to comment upon these policies that are being considered by NIH for implementation by some time early next year.

And because the URL is long and involved it's easier if you jut go to Google and type in GWAS which stands for Genome Wide Association Studies Policy. And there will you find a site that will allow you to see what the proposed policies are and that will encourage you to comment. And the comment period is open until October 31^st.

This is a major policy decision process that the NIH is going through and we really would like to hear from lots of people about their views on where we are proposing to go.

So, with all of that, I think it's fair to predict -- and this is a pretty bold prediction, but I think it's fair to make it now -- that the major genetic risk factors for common diseases like the ones you see listed on this slide and many others are going to be identified in the next two or three years. And that will be a menu of risk factors much broader than what we previously had.

And it will certainly begin to raise the drum beat of expectation, that it might be time then to begin to offer diagnostically the opportunity for people to find out their risks. And these being risks that would be at that point well-validated scientifically.

Let me say though that while these case control studies that I'm talking about are a very powerful way of doing discovery, that is of finding the gene variants that associate with disease risk, they're not a particularly way of quantifying exactly what that risk is. Because they're biased in their very nature. And they're a particularly lousy way of discovering gene environment interactions for reasons that all epidemiologists would know and I'm not going to go into. But trust me, they're not good at that.

And if what we really need for public health purposes and for implementing personalized medicine is not just to know that a certain genetic variant is a risk factor, but what environmental influences it interacts with and how much is that risk any way then what we really need are prospective large scale studies.

Those were mentioned this morning by a couple of groups who have already access to large patient populations such as Kaiser Permanente, the VA is another potential place to carry out such studies.

Ideally though, if we really thought this was a priority and I'm going to argue it's an incredibly high priority, we should figure out a way to mount in the United States a large scale prospective cohort study that represents as a snapshot of our population across the age distribution, certainly across socio-economic status, and geography, race and ethnicity, gender and educational level.

And then we might actually be able, if we followed such a cohort of half a million people and that's about the number it would take to have sufficient power to answer a lot of these questions, we might finally be able to say exactly what should we be doing as far as interventions.

Anybody who cares about the evidence base for personalized medicine I think has to look at this as a serious need, one in which I guess you heard this morning, other countries are investing in. The U.K. bio bank just now getting under way and certainly similar ventures going on now in Japan, in Taiwan, in Germany and, of course, the whole country of Iceland. But oddly not here in the US.

This is expensive. That's why you're not hearing about it, because it is not the time I think to be proposing a large complicated study of this sort. But if we invested in this, I think it would actually be such an engine for discovery that we would find in the long run it saved us money by not mounting a lot of separate studies on separate diseases. This would be about all diseases.

So, I wrote a little commentary about this a couple of years ago. There's been a lot of discussion about it. There's actually a study design that was put together by about 60 people who worked very hard on this that's up on the web. The Secretary's Advisory Committee for Genetics Health and Society has been considering at some length this whole question of large population studies. And have been quite positive about the need for it recognizing that this is a genome project kind of enterprise in terms of its complexity and its potential cost.

So, I put it out there without the expectation in the current budget climate that anybody's going to rush to embrace it, but to point out that if we're really serious about gathering that kind of potential base for information, we need to consider this.

Now, maybe we could cobble it together by taking advantage of various prospective cohorts that are already available including some of the ones mentioned this morning. And that might very well turn out to be a good balance between not letting the perfect be the enemy of the good. But, of course, you would want to be sure if you did it in that fashion, if you had a cohort that was representative of all of us not just some of us.

Well, okay, so I think one can say that the top part of my diagram here, the discovery of genetic defects that are associated with common disease is going to be happening in proliferative [SIC] ways over the course of the next two or three years. And that should be very exciting. And it will certainly increase the momentum for having conferences like this one about why personalized medicine needs to happen. And there will be a lot of pressure to see exactly how to do that in a fashion that's scientifically legitimate.

Obviously one of the first things that happens once you make those discoveries of variants associated with disease risk is the diagnostic possibility. People are already out there asking well, can I be tested for macular degeneration risk. It's most interesting of course to people to know their situation, if there's a preventive medicine strategy attached to it and there will not always be so.

But I think it's important as Carol Kovac said to have concrete examples of how we're already doing this, so it doesn't sound totally like pie in the sky. And in that regard I would remind that the paradigm has been set for a very long time about how diagnostics and prevention can provide a great benefit.

And in fact this is something that every newborn in the US is subjected to -- PKU screening and in most states lots of other conditions as well, early intervention is for metabolic conditions is essential for a good outcome.

We tend to forget that somehow. Oh, that's newborn screening. Well, you know, it is exactly the paradigm we're talking about applied in a way that's incredibly effective both in terms of science and in terms of cost.

In terms of such offerings to adults, I'll just give you a single example here, of a circumstance that I think is now very well validated where diagnostics can lead to effective prevention and, in fact, save money. An economic analysis has been done in this condition.

Here's a family where you can see there are three individuals with cancer at relative early ages. This is the kind of family that a medical geneticist or oncologist would recognize as potentially having this condition called heredity non-polyposis [PHONETIC] colon cancer.

It's possible to test the unaffected individuals now to find out that they are also at high risk and, if so, to offer them intense medical surveillance. And the point is that's not just to see what's happening, that's a point of discovering early onset neoplasms, particularly colon polyps, while they can be removed, and which will therefore prevent them later on from turning into a metastatic malignancy.

So, here's a condition that the data will tell you, knowing you're at risk, undergoing that kind of surveillance is absolutely beneficial, in fact is probably life saving in many such people.

In this family -- this is a real family -- they were tested and there were three individuals here found to be at risk, they are all now having colonoscopies every year and one of them has already found a few polyps that might ultimately have ended up causing a great deal of trouble.

So, this kind of paradigm is a useful one to point to. This is not diagnostics based on genomics for 20 years from now. This is now. It also raises about three different policy issues.

This is a family that struggled actually about whether or not to have this test. They were concerned about whether their health insurance or employment might be at risk. And interestingly they also came to light of course because of family history and they were surprised that, in fact, when they had talked to physicians before coming to attention -- that the family history didn't always seem to get much attention.

In fact, the woman who originally came to attention -- this woman here in the middle -- said she'd never actually been asked about her family history by any care provider until she brought it up.

So, we need to do something about that. Family history is after all a wonderful personalized medicine tool. It's vastly underutilized, it's free, it doesn't cost you anything to actually do this kind of particular diagnostic.

But you know who does it best is the person in the family who knows Aunt Ethel and can make some phone calls and find out what it was actually that somebody died from instead of trying to remember when they're sitting in the physician's office while a harried care provider is trying to take a quick family history just because it seems like it ought to be done. But is it really likely to be incorporated into that medical care plan.

A big step forward in this regard has been Surgeon General Carmona, while he was here, putting a lot of his time and effort into a web based family history tool and this is what it looks like. If you want to look at it more it's on the web site that you see listed here.

This enables anybody sitting at home in front of their own computer to put together a pedigree that can then be printed out in a standard format and then they can carry that to the care provider and say here's my family history you don't have to spend your busy time asking me about my grandparents, but let's talk about this particular condition that seems to be happening in my family at a higher than expected rate and what I should be doing about it.

This ought to be nicely integrate able [SIC] in terms of the electronic medical record. I think there's a real opportunity here to help in many ways, both to get care providers thinking genetically and to provide really useful information for people whose care could be positively influenced if this kind of information is available.

Of course, this does require us to point to the fact that I think we have a huge issue in terms of health care provider education. That was mentioned this morning. I want to mention an organization that we've worked hard with to try to provide that kind of useful information to busy practitioners and that's the National Coalition for Health Professional Education and Genetics founded by the Genome Institute along with the AMA and the ANA. So, physicians and nurses working together on this. If you want to read more about that organization that is its web site.

And, of course, as I mentioned this family almost didn't go through the testing at all because of their concern about genetic non-discrimination legislation not having yet been passed. And we're still there. So, in the 108^th Congress, as most of you know, a bill was passed through the Senate, S1053. It was not taken up by the House.

In the 109^th Congress, that's the one we're in now, S306 was passed 98 to nothing. HR 1227 is the House version of that. As you know we are down to very few legislative days in the 109^th Congress and barring some amazing sort of last minute effort it looks as if we may once again have to start over next year.

The bill covers both genetic discrimination in health insurance and in the work place. And in the view of many of us, this is a critical step, if you really expect personalized medicine to become a reality.

Because as was shown in the survey that Mary talked about this morning, the public expects this and we're actually fairly shocked that we have not yet provided it. And we need to I think get this done.

Besides the diagnostics and the preventive medicine, lots of interest in pharmacal genomics as another personalized medicine and one of the developments in this area that I thought was particularly interesting is the marketing of this [unintelligible] chip that allows the testing of the variations in the p450 genes, which play a significant role in the metabolizing of a lot of drugs.

But it's been interesting to see this has not been embraced by as many physicians as I initially thought it would be. I think in part because it is complicated, it's not so easy to interpret the output of this. And it will be interesting to see how that can be better developed and maybe what people are looking for on more explicit guidelines based on more data.

The specific application of P450 that NIH is interested in is to be able to do a better job of dosing a drug that causes a very large number of complications each year. The most common drug given for blood clotting problems -- namely Warfarin, which has been around a long time, lots of people are on it, including my own mother.

And, in fact, you will know if you've been paying attention that this is a drug which has got a very narrow therapeutic window. And if you're under a clot is likely to form and if you're over, a bleed is likely to form. And people's ability to metabolize this drug does vary quite a bit, so getting the dose right is a complicated and error prone process. And especially in that first month or two of trying to get the dose right a lot of bad things could happen.

Well, it's pretty clear that the gene CYP2C9 has something to do with the metabolism here. In fact, there's even been a careful economic analysis that I show you here on the slide that says that it would be neutral economic results to actually begin to implement testing for that in this model.

But it's actually better than that now, because we not only know about CYP2C9, we also know about variation in Warfarin metabolism based upon BKORC1 so, there are two genes now well characterized that play a significant role in what dose you need to get.

So NHLBI, the Heart Lung and Blood Institute and FDA are both talking about now mounting a prospective study to try to see if you could use this genotyping information to choose the dose of Warfarin instead of empirically guessing and hope you got it right. And that would be a very interesting outcome.

I'm not sure and a question earlier in this session pointed to this that we can afford to do that kind prospective pharmacogenomic study every time we come up with an idea like this. But this is such an important paradigm to set right now in a most rigorous way that I think Warfarin is a good poster child to try to undertake this sort of scientific study and see if we can show rigorously the inclusion of genotypes gives rise to better outcomes. And those studies we hope will get underway before very much longer.

Of course, ultimately all of the discoveries about genetics and genomics should play out not only in predictions, but also in development of new therapies of course, many pharmaceutical companies are banking on that and I think one of the main reasons, for instance, that Pfizer so willing to put their money into the GAIN study to discover gene variances - in many ways those are going to be the most exciting new drug targets.

And companies generally feel that those are pre-competitive discoveries and we ought to do them as quickly as possible so they can get on with what they do really well, which is to find small molecules that will work against those targets and ultimately get those into clinical trials.

And, of course, I think it's fair also to say that our list of therapeutics that will come out of the personalized genetic approach is not a long list. Ed mentioned something like 15 today. But it's certainly not a small list at this point and it's growing quickly and everybody's favorite -- at least mine -- is GleVac, not only because it was based on this strategy, but because it represents a almost cure. At least it puts many people into sustained remissions for a disease type of leukemia which was previously not well treated if at all.

We need obviously many more examples like this. And I might say one way we hope to find them is this project mentioned this morning, the Cancer Genome Atlas which is a bold new effort to try to achieve a comprehensive understanding of the genetic and epi-genetic causes of cancer, in this case largely looking at the mutations that occur during your life, although will undoubtedly discover some hereditary ones along the way.

And the expectation is that there are hundreds of yet undiscovered genes that are involved in various cancer types and that the discovery of those will point us towards better ideas of diagnostics and therapeutics.

The three tumors that are going to be studied in this pilot part of TCGA are going to be ovarian cancer, brain cancer; namely, [unintelligible], and lung cancer. That was just announced last week. For each of those about 500 different tumors will be studied by a variety of different high technology approaches, including DNA sequencing in the next three years.

I think we will learn a prodigious amount about what's going on with those types of cancers. And the consequences for clinical care will be substantial.

Well, let me conclude by trying to wrap this all together with a bit of a dream about where it might all go before we all go off to do what we are doing most of the time when we're not coming to conferences like this.

I do think it's fair to say that medicine has the opportunity here to be transformed in a way that's truly beneficial. And all of the other arguments we can make about economics and about reimbursement issues and about regulatory issues, and so on, should not eclipse the fact that what we're really trying to do is provide better care for people, to give them a better chance of staying healthy.

So, please consider, if you will, Betty. Betty is just a regular person who in 2015 at the age of 25 decided that it was time to find out about her own health care risks being at that teachable moment.

So, she found out about this Surgeon General's family history tool and she actually took the whole thing to heart and learned that she had some uncles with early heart disease that she never quite knew what happened to them.

So, she went to her health care provider. Fortunately this was somebody who had been a part of MICHPEG [PHONETIC] and made an effort to stay informed about genomic medicine. The provider suggests well, if you really are interested in this information, let's just sequence your complete genome. It only costs a thousand dollars, you only have to do it once.

This is not a completely unreasonable scenario by 2015. We may get well beyond the point of doing individual geno types for this gene or that gene and just do the whole sequencing.

Betty was a little worried about that. Gosh, that's a lot of information. Who's going to know about that and fortunately she could be reassured that back in the later days of 2006 [Laughter] legislation was passed.

So, she undergoes the gene sequencing and is found to have three gene variants that have been shown conclusively in well validated studies that were thoughtfully put together to increase her risk of early heart attack about four fold over the average person. So, here, again, family history kind of triggered the thought but the ultimate analysis here is much more quantitative.

So, given that information Betty and her provider sit down and design a program of prevention based on diet, exercise and medication precisely targeted to her genetic situation.

Betty's story continues. She does well, she's done the right things here. By age 75 she's still doing well, but she gets some left arm pain, which she assumes is do to gardening but her provider knowing her higher risk diagnoses an acute MI.

Referring to her genome sequence the provider chooses the drugs that will work best to treat her in this acute situation of a heart attack and she survives and is alive and well in the 22^nd century.

A good outcome. This is what we're all hoping for. Right. Well, let's just consider the flip side. If we don't play our scientific and policy cards right, could the dream become a nightmare. Betty's story gone wrong. [Laughter]

Betty never learns about her family history. Educational efforts for the public and health care providers were de-funded and Betty's provider thought genetics was irrelevant to practice. Many would say that today.

Betty hears about genome sequencing but after seeing her brother lose his health insurance from this information, she decides not to do this. She eats an unhealthy diet, gains weight, develops high blood pressure.

Tests might have been developed to figure out what was the right treatment for her high blood pressure. Those tests were never really validated and are not reimbursed. So, they're not part of medical care when she is interested in that.

And, so, she gets treated with a drug that causes a hypersensitivity reaction, so she stops treatment. After ten years of uncontrolled hypertension she develops left arm pain not at age 75, but at age 50. Unaware of the high risk, the provider assume this is muscular/skeletal, prescribes rest.

Betty returns to the ER a few hours later in cardiogenic shock. The absence of her genome sequence information prevents immediate optimum choice of therapy, Betty dies in the Emergency Room. So, this is really personalized medicine gone wrong, because it didn't happen.

So, I guess what I would have to say, without leaving you here in a down moment is we don't have to let that happen. We collectively -- people in this room - we have major leadership roles, many of you, in this future enterprise -- can if we martial our resources successfully together and get the word out about both what the promise is as well as the potential pitfalls, we can make the good story happen instead of the bad story.

And, so, my charge to all of you before you all go off and I do too to do other things is just two words: Save Betty. [Laughter] She's counting on you. We can do this together, all of us. I'm confident of it and I appreciate the chance to come and share this with you. Thanks.

[APPLAUSE]

WOOLLEY: Thank you, Dr. Collins that was really terrific. And I feel that we're all in very good hands indeed. I want in closing first of all to add one thing to the suggestions that Dr. Collins made that each and every one of us can do right now to advance the field and help realize the promise of personalized medicine.

He talked about among other things, checking in and creating a family history and promulgating that -- the ease with which this can be done is really an important thing to do.

A second thing was to Google GWAS policy and weigh in, because the NIH wants to hear from you, wants to hear how it is going to be appropriately formulating policy based on public input.

Now, I want to add another thing, while you're at your computer. Check out a new web site called Your Candidates-Your Health dot org. It will tell you what your candidates for Congressional office in November have to say on ten key questions that very much will determine what happens next. Will help determine whether genetic non-discrimination does become the law in this land, whether the NIH and the private sector have the policies and the resources to help realize the promise of research, so check that out. We can get there as a community of concerned stakeholders.

I want to conclude by saying thank you to Dr. Collins for his leadership for research, for health and for taking time to speak to us today.

[APPLAUSE]

I want to thank Dean Ruth Katz and her colleagues at the School of Public Health. It's been a delight to work with you here. I want to thank Roche, Dr. Austin and your colleagues of being with us and for helping us think through and make this program possible. Susan Dentzer for your superb job as an informed moderator.

[APPLAUSE]

All of our panelists, thoughtful individuals every one and people who will continue I know to play leadership roles in personalized medicine and in health, in better health, in many ways through research.

And also a big thank you to Dr. Christine Brown and all of my colleagues at Research America with whom it is a pleasure to work every day.

[APPLAUSE]

And finally to all of you a plea. Please fill out your evaluation forms, take just a moment to tell us what we can do better as your partners in assuring that medical and health research becomes a higher priority in this country. We need all your help in helping us get there with you in partnership. Thank you and enjoy the rest of your day.

[APPLAUSE]