Confirmation Number:371000
Event Started: 6/16/2005
Good morning, everyone.We need to start on time just in case Reed is at homewatching us on the web.Good Reed.And good morning to everyone.Welcome back.First order of important business, of course, since welike to look after everyone's stomach is to remind themembers and the ex officios you should do so at the there in the lobby next to the registrationdesk no later than 9:00 and then as yesterday, yourlunches will be delivered here.Let me also acknowledge and welcome Jodi brown who ishere from the division of health Canada, we're delightedto have you with us.Hope you learn something and hope we in chance willlearn something from your activities north of the borderasSo, welcome.Let me point out to the committee, you have in front ofyou the clean copy of the final recommendations that wevoted and approved unanimously yesterday on coverage ofgenetic tests and services.This is simply for your information so you have a cleancopy to take home and look over.We have another full day ahead of us.Today we'll be hearing a number of perspectives on thecurrent state of the field pharmacogenomics and theimportant policy issues that we identified as acommittee when we went through our pry ortizationprocess. entire day will be voted Topol si issues.We have a number of outside speakers and of course ourstaff.Bio sketches for today's speakers are found in yourtable folders and at this point I'm going to turn itover to Emily Winn-Deen who will lead the discussiontoday and we'll begin by giving us an overview of the forces work and the goals that they've identifiedfor us today.Emily?
Thanks.We're going to start today with an overview of the workthat led to having this session on pharmacogenomics.Pharmacogenomics was identified as one of the fourissues warranting in-depth study during our prioritysession last year, and since then it's been increasinglyapparent that this field has the potential to have alarge impact on health and health care and needs to beconsidered carefully.Pharmacogenomics testing may offer more individualizedapproach to medicine through the identification ofgenetic variance or bio markers that may help to targetthe correct pharmaceutical interventions and theirindividual genetic variations.The field of pharmacogenomics will allow furtherintegration and transferp of the human genome data fromthe human genome project into the practice of medicine.There's been a lot of data on the number of deaths thatoccur.The latest figure is about 100,000 deaths per year thatoccur due to adverse drug reaction and there is the hopethat pharmacogenomics will also play a role in reducingthe number of deaths.During our priority setting discussions within the taskforce, we focused on physicians' need for relevant andpractical advice on the application of pharmacogenomicsdata in the clinical setting.I'd like to acknowledge the task force and all themembers who contributed both the folks within the SACGHScommittee as well as our ex officios, Kevin Fitzgerald,When the task force began to develop a committee webegan in four areas.We tried to put everybody on the committee on art of alevel playing field and get everyone oriented and that'sthe goal of today's session.The four areas was the state of the field ofpharmacogenomics today, where are we with translationalefforts in pharmacogenomics, what are the ethical legaland social issues that this branch of genetics mightraise, and what is the role of government agencieskeeping in mind our charter as an advisor to HHS.The translational issues that were identified andincluded regulatory funding of -- regulatoryissues, funding of pharmacogenomics research andtranslational research, the potential to create neworphan drugs or diseases through patient differentiationvia genetics, we wanted to include the perspective fromdifferent sectors of the -- both the community as wellas the industry that are affected by this and try andfind some cooperative approaches in the spirit ofpublic, private partnerships that might help move thisfield forward.In addition, pharmacogenomics may pose some uniqueissues and we wanted to make sure that we did notoverlook some of these and we're most concerned aboutnot having any exacerbation of health care disparitiesFinally we wanted to make sure that we get a goodoverview of what's going on already within HHS andhopefully today's discussion will give us an idea of are today as well as where we'd like to be interms of any gaps that we identify.Prior to this session we sent out a request to thevarious HHS agencies and asked them two questions.The first was, what is your age and what is the mostimportant policy issues concerning the voids in thefield of pharmacogenomics and from your particularagency standpoint what are the specific questions thatour committee could address through each policy issue?The issues identified by the agencies included thefollowing.Applying pharmacogenomics knowledge in the drugdevelopment process, assessing clinical validity,analytical validity and integration of pharmacogenomicsinto clinical and public health practice.The full summary of the input from the agencies can befound at tab 6 of your briefing book.The first category was suggested by NIH and althoughthis will remain largely a private sector endeavorprimarily within the PHARMA industry it's important forus to understand how pharmacogenomics knowledge will beused in drug develop tment.The second category was highlighted by four agenciesCDC, CMS, cur sa and MIH.Under integration the need to educate providers andconsumers as well as privacy and promoting wide accessto clinical trials were noted by CDC, FDA, and NIH.In the public health arena considerations of ethnic andracial variation and the effects of diverse population,the potential use of pharmacogenomics for screeningpurposes and the need to monitor pharmacogenomics impactwere identified as important issues.Again, CDC, NIH and her sa all contributed to theseissues.Access and costs remain important concerns that willneed to be considered and addressed.The need to understand the direct and indirect costs andpotential for reduction of overall health care costsrelated to pharmacogenomics is important for us to tryand understand in a little more depth.Adequate focus -- I'm sorry.Adequate access from her sa while cost was highlighted CDC, her sa and NIH.The feedback from the agencies largely parallel the mission and will be very helpful.It was suggested that our discussion this afternoonwould initially focus on an explicit statement of whatwe expect pharmacogenomics to do for people's health.We welcome more explicit suggestions from any of thespeakers and any of the ex officios as we move forwardin our discussion.Additional issues that we're were identified throughother outreach efforts included barriers and theseadditional issues that we identified were doneby our task force discussion as well as some conferencecalls with key individuals within the private sector.We consulted with dale Clark who's the chief technologyofficer and chief medical officer for GE health care aswell as with mar ra who's the president of enzymegenetics and her colleagues.The barriers that were identified by bill Clark andreally echoed from folks from general same included thatthere are really no uniform reporting standards forThere needs to be an appropriate approach for the valueof pharmacogenomics testing, there are issues of robusttechnology and reasonable costs that need to beaddressed, and whether FDA approval will be required inorder for reimbursement to take place forpharmacogenomics.On that same strategy, there's really a lack of paths forward in particular home blew andwhile there is a lot of data available on correlation ofgenetic variation with different drugs, there's stillnot the body of data required to actually give gooddosing guidelines for many of these drugs.So we're still one step away from being able totranslate it intoThe other barrier was, you know, really what is thecatalytic event that is going to be required to movepharmacogenomics out of academia and into standardclinical practice?You know, what is the driver here?Is it better medicine?Is it legal liability? what are the issues that are going to make thishappen?Because I think we have good evidence from severalarenas for things where we understand the science andyet the science hasn't really translated into a newstandard of care in the practice of medicine.And we need further clarification from the regulatoryagencies on what is actually needed to changes in labelling and how that's supposed to match.General sign suggested some additional strategies topromote pharmacogenomics.They felt that it pharmacogenomics was a paradigm shiftand that all key constituencies within the health caresystem need to understand it's all part of ourprogramming today was to try and begin to bring togetherall of these different types of constituencies.We recognize that due to time limitations we were notable to have every single piece of the puzzle presentedto us today and that some of these things will probablyhave to be deferred to our next meeting, but we aretrying today at least to make a start in bringing theseWhat are the financial incentives that are needed inorder to encourage companies as well as physicians tomove forward in the practice of this new typeA couple of other things that they were concerned about.They felt that there was a need to address both the --the home brew, the laboratory developed tests as well asFDA approved tests.My knowledge there's only one FDA approved test.I think it's the most of the work being done in thisfield today is tests and we need to recognizethat and find ways to address it.And that the government in their role as both theregulator and a payer needs to be looking at how theycan put in place policies that would result in betterdrug efficacy and improved state.So the purpose of today's session is to really provide acommon understanding of the fundamentals of and the state of the field today, toidentify issues that will be critical to move usforward and to determine if there's a specific role thatthis committee can play in facilitating this translationinto the practice of medicine.I want to remind the committee that our goal is toadvise HHS.We can't solve all the problems of the field, but Ithink that there are a number of agencies within HHSthat are involved in this field.We need to assess whether we feel they've got everything in hand or whether there's some specificrecommendations that we'd like to make going forward for they could do more actively or more cooperativelyamong the agencies.So with that in mind, I'd like to give you a little bitof an outline of the session today.We are very pleased to have a panel of speakers who I to say are all experts in their fields and wegreatly appreciate their willingness to come and sharetheir knowledge with this committee.We're going to start with the fundamentals, what theheck is pharmacogenomics perspectives from boththe diagnostics and the PHARMA kinds of industry.Finally, we'll have a talk on the LC issues.At the end of this long session, I hope you're alltaking notes because at the end we'll have a fullcommittee discussion about really what we heard, what wewould like to do a committee moving forward and thetask force is looking for guidance from the committee onwhere you would like to see usSo that we can be prepared if we need to do somespecific activities in the interim between this meetingand the October meeting.With that I would like to introduce our first speakerwho really needs no introduction, because he is, if Idare say, it's a grandfather of pharmacogenetics.Rich Weinshilboum joins us.He was intimately involved with the methyltransferaseresearch and actively teaches pharmacology as well aspharmacogenetics.
First of all, let me thank the committee for theinvitation.Someone who's been doing this sort of stuffer fordecades to be introduced -- I am a grandfather but to beintroduced that way is a little disheartening early in morning so what I thought I might do to be helpfulto the committee and actually I think really our rolehere is to be helpful to you is to do pretty much what Idid with a group of graduate students yesterday morningabout same time.I was asked to begin with origins and concepts, inessence quick overview of where we are.Let me begin with a disclosure.Although for years I wasn't invited all of a sudden I'vebecome very popular since the FDA guidelines come out soI'm invited to pharmaceutical and bio tech companies butMayo is in the upper Midwest where with scannedknaveians settled so all of the honorary induck tees donot come to me.They go back to Mayo to support our research andeducation.There's a flip side to this.I've spent my entire life in an academic environment andthat's why it's so important that we have eric andWalter to give you an upclose and personal view from thefor profit industry side because their view will bequite disht than mine.I should also in a matter of disclosure point out that Icurrently have the honor to chair the nationalinstitutes of health pharmacogenetics research the PCRMwhich you'll see down in the corner of my slides andeach one of these slides represents one of thesecenters.As of next week, the next chair of that group will takeover, the stars will move around a little bit so I'll beback here next week.So let's begin.Sort of pharmacogenetics 101.You all know that what we're talking about is the studyof the role of inheritance in variations amongindividuals in their response to any genome buy ticsthat we write a prescription for, the patient takes tothe pharmacy and takes the medication thinking that Iknow what I'm doing.So drugs are just a subset and we're talking aboutgenetic variation in the chemical and drug response inmany ways this represents a revolution.That is the genomic resolution.In beginning the 21st century there has been a paralleltherapeutic in which we have gone and I liketo demonstrate this in this fashion.I was actually around in 19 # 1 but rumors around thestudents the contrary I was not reading this then.Here is the 10th edition.The books are the same size.There's virtually nothing in this book.That is, there is morphine and digitalis.Alpha and sulfa drugs but no antibiotics noantipsychotics, no antidepressants.Roosevelt was the president of the United States and itwas treated with feenl bash toll and so as a matter offact there has been a dramatic change in the therapeuticagents which we a matter of fact, it's been earth shaking.We talked about paradigm shifts in your introductorycomments and bring that together with the genomicrevolution and those are the ingredients that havecreated we are talking about today and it's thereason that we're sitting here because the concept ofpharmacogenetics and pharmacogenomics dates backcenturies.Every time I'm called up by a matter offact, these concepts, these concepts very around for a century but they have been accelerateddramatically by the technology that came out of thegenome project.So my definition is the convergence of the advances incogenetics that have occurred over the last decade andto bring that volatile mix together and I think that'sone of the reasons that we're sitting here.The clinical goals are obvious and in the introductorycomments we mentioned avoiding adverse drug reactionsand I'll use an old chestnut to illustrate that in amoment, but let's don't forget that we're alsomaximizing therapeutic efficacy.Selecting those patients who might responds best to the and one of the impedments and I'm speaking fromthe view of the academic world through the involvementof pharmacogenetic process limits the market to thedrug. sure I'll hear something quite different in just we need to get the issues out and at leasttalk about them.The scientific goals are also obvious.The structure with variation of the drug responsecorrelation.Now, I never thought in my lifetime and I've been doingthis stuff for over three decades that I'd be standinghere talking to you about a matter of fact, the lab, I walked in the other day I said how many base pairs did you sequence thisweekend and he said 5 million.When you think about it, this is a mom and pop store,folks, and when you stop and think about that that'struly an amazing revolution that has occurred.Let's immediately say I mentioned that I'm an internist,that all of us who write those prescriptions is thatgenetics is only one factor that plays a role.The patient's age, adrenal function plays a role.We are increasingly sensitive to the fact that males andfemales respond differently to drugs.So this is only one factor but it's one where objectiveinformation may now be brought to the physician and thechallenges which you mentioned in your introductorycomments, how do we help the practicing physicianintegrate this information into the therapeutic it's going to be -- let's don't forgetbecause my medical students, they focus on what does the do to the patient, but the patient is doing a lotof things to the drug.That the drug must be absorbed and we know thetransporters play a role in this process, get to itssite of action, interact and metabolize, all of theseprocesses we now know have significantly and relevantgenetic variation.Most of these grew out of the field of drug metabolismbut that's only of the demonstration project because ofpharmacogenetics that has impacted unhomogenized humanbeings.I to think of this as a scientific evolutionanalogous the way we have approached the applicationof genetics to diagnostic tools.Let's begin with some rather dramatic monodrawmatic.They were necessary to make the point.I tell you how many years I would go aroundtalking about pharmacogenetics and as soon as I'd saythe words polymorphism their eyes would glaze over andthey wouldn't pay any attention. And they would tell mewhy don't you get a -- I would say I'm studying thevariation and so I had to make the point.That will not be probably an example of the major way inwhich genetic variation will manifest itself andincreasingly we're talking in terms of PD pathways andI'll define those in just a moment and adding widescreens to gain insight into the myriad of ways in whichgenomics can play a role in that.Pharmacogenetics are those factors that influence thefinal drug concentration at its target predominantlytransporters, drug metabolizing enzymes.Pharmacodynamics are the factors that influence thetarget themselves.Not just the target themselves but all the factors thatinfluence the target itself. It's gun to be awfullyhard to make an end run around genetic variation in thepharmacodynamics.Let's use a couple of what eric turned to me and said Iassume you're to talk about the old chestnuts.I yeah, sure, of course I will and so let's usethese and I like to use them because they're bothwell validated and because in the graph pharmacogeneticguidance that they put out in 2003 and in March of thisyear these are no longer drafts they -- as valid biomarkers meaning they're old fashioned and we know a gooddeal about them.So let's use TPMT.Here's the drug that was developed in the welcomecompany by -- they shared the no bell prize in 1998 andpart of the main stay in the treatment of childhood.A disease that was uniformly fatal when I was in medicalschool and today we cure 85% of these kids with drugs,no no radiation therapy, that's what I meanwhen say the therapeutic revolution was a quiet one.These drugs are also used as a
These are very powerful cytotoxic agents and everynow and then you would treat the child with leukemia andthe child would die from the drug therapy, not anythingthat anyone wanted but we would call it an id owe sinCATTic reaction which means we didn't understand whatcaused it.This shows you data 25 years ago on TPMT and in case I to say it what you see here reflects the level ofthe enzyme activity for reasons that will become clearwhen I show you the gene in just a moment and these are298 randomly selected northern European blood donors inMinnesota.There's an important reason why I say that. is everyone in Minnesota is named Anderson andstuff like that so -- so -- but there's a scientificreason for bringing that up.90% of this population had high activity, saw 10% -- hisdaughter had zero ensoo*irm activity.People will stop you walking through the mall and askyou how your mom's enzyme activity is.So using very sensitive molecular techniques developedby a monk in a monostair -- what are the kids and youcould just as easily determine if this was a genetictrait using that approach you can say that this womanhas two copies of a gene below activity and these areheterozygous with immediate activity and that's true inevery case.This just shows you the consequences of having twocopies.This is long after we had described if you have low TPMTactivities you are at serious risk.This is a heart transplant patient in Germany treatedwith standard doses.Here's the white count.Here's dose, notice that the white count is stoppedthe drug is stopped goes up drug is started white countgoes down to zero.They then measured the TPMT in the blood.These cases by the way are not reported.Do they occur?Tragically, yes, because I get many of the telephonecalls.I one just two weeks ago, again, exactly the same. you have low TPMT activity on a genetic basis,this can be life threatening.Mary at St. Jude has demonstrated this is also a riskfactor for the plasm. in which he isfield and at our place we havebeen doing the TPMT phenotype and we do about 5 to10,000 of these tests per year.About half on our own patients and about half referredin from physicians outside and we are individualized intherapy clearly for these people.We treat them with one tenth to one-fifteenth thestandard dose and that's been our wayBecause of the dramatic clinical concept and becauseit's relatively well validated this was one of the firstexamples that the food and drugconsidered for possible information labelling at twopublic hearings, I testified at both of them.Through a service here I saw before we began and thatwasThese are old examples and that's why eric asked me, oh, am I going to have to hear about CD 6 again and sothis shows you that CD 6 metabolizes 40 or 50 commonlyused drugs including beta-blockers and antidepressantshere you're looking for the -- which was neverintroduced on the market.It undergoes CD 6. the way we have represented this is the waypharmacogenetics pursues this these are the forms thatmetabolizes up here.It's about 25% of a Caucasian or European population andagain I say that because it metabolizes and low numbersare also -- that obviously is also or not so obviouslybut is also of clinical importance.This shows you data with the previous slide came fromthe instep this is also from these clinicals.So we're looking at the antidepressant or crypt lean andwhat you're looking at is pharmacokinetics, that isplasma levels depending on the 62D 6.Most of us have copies of that active gene.By the way, this slide unites the two cop picks whichare the least favorite of the male medical students.They find pharmacokinetics eternally boring so you cansee you can either have gene deletion or you can havepolymorphism that result in activity, a map *p muchlarger area to serve.This lady, who was herself a nurse had 13 copies of theactive gene.Look at her pharmacokinetics.Metabolized way up there, way off scale.So these are active.In most cases 62D 6 terminates the action of the drugbut for codeine it converts it into morphine so if youare metabolizing 62D 6 you will not get the an el geeziceffect from codeine but if you're a rapid metabolizer, a62-year-old man hospitalized for pneumonia, treated withstandard doses of codeine right after the PDR.The patient stopped breathing.He had morphine levels 20 times the expected levels.I just show you, this is a preview of Walter. have no stock in any company and certainly not inWalter's, but let me say that all that we're doing hereis using the metabolic ratio to give us insight into thelevel of DNA.In today's world and we'll be talking about this later,devices like the one that comes from diagnostics givesus direct insight.I finally want to give us a peek of the future ideal ofwhy we live in Minnesota we're right no*eks to Wisconsinthis is pall link and if you haven't read the story ofthe war farren is the kind ofcaptioner disconnected from
It is so solid, so old, why if you just go to theacademic medical xer X, in a good medical center in agood city, why don't they test for the D 6 before theymetabolize that enzyme?
Well, of course he's asking one of the many questionsthat I've asked over the years and I have been goingaround overdosing audiences on this sort of information,particularly for the more dramatic examples.For some of the well established examples in TPMT and 2D6 ewed as examples because they are relativelystraightforward and dramatic.That's why I said demonstration products merely to makethe point you'd have to invent them.They didn't have to invent them we're fortunate to havesomebody stumble across them early on.Part of the difficulty is at the level of the practicingphysician understanding this kind of information andthese concepts. I will -- we'll talk about that later and actually,I'll mention this when I make my later presentationabout practice of medicine.At our place, we have a genomics education program whichfocuses on therapeutics and diagnostics which we havefunded about $1 million a year merely to continues wallyraise the conscience of the physicians and educate them.Physicians are talented and want to do the best fortheir patients but vocabulary is the barrier here.We have to make things user friendly and easy for thephysicians.Number two, you're right in regard to this age ofinformation and the Internet that the patients arebeginning to drive the process and we need to be carefulabout not having inappropriate patient education. it's going to be an interesting challenge.I get the opportunity to present something calledinternal medicine reviews, which for the upper Midwestmeans a lot of internists like myself come in and wantto hear what's going on and even dental reviews.A dental reviews, which are dentists from the upperMidwest, they're telling me that their patients arecoming in, having done just what he said, having beentested over the Internet and they'll know their 2D 6gene type because they don't want to get Tylenol number3 with codeine if they can't respond to it.So the patients -- the dentists, I found thisfascinating that dentists are now seeing it.So the patients may be ahead of the profession in someways.There are a lot of other barriers that we'll have totalk about when we go into the further discussion but Ithink is a great challenge and you mentioned it inyour introductory comments with regard to the barriersto the introduction of this science across what I' ferto as the translational boundary.
Thanks.We've got time for a quick one more. I think you mentioned that TPMT that there has beenconsideration for labelling by the FDA.Was that included in labelling with pharmacogenomics?
I'm just this guy from Minnesota who invited into testify.It is my impression that the labelling has been changedto make information with regard to the existence ofgenetic polymorphism and the availability of testing.There was no man date for testing to make the physicianaware of that information.
Okay.I'm sorry.We're going to try and keep on time to get some responseto the next topic. next focus will be on public health perspectives andspeaking with us today is Robert Davis, the departmentof technology Washington school of public health.He's currently on sabbatical.CDC's office and he's going to give us an overview ofwhere we are in
Thank you very much for inviting me to be -- the conflict of interest disclosure I'm onsabbatical.I want to start by showing our house.This is the celebration that occurred when the -- waslicensed.We're base fans and I came home and was celebrating withmy life having get the license I promptly found that went into -- the yiefrt of Washington -- my wifestopped talking to me and my partner got mad at me soI'm a bringer of bad news today or the bringer of asobering outlook and I've already suffered theconsequences.I just want to introduce that -- that it was atremendously exciting talk when we heard about it andabout the fantastic students that they've given us butthere's a big step between the clinical level and how itcan be applied at the public health.I want to walk about how we have to get from here andthese are my kids.They share my genes.I am the biggest fan of the genomics revolution.I wanted to talk about how we get from this degree ofexcitement to an understanding of how it actually workedat the macrolevel of public health.Let me go back to the start.As we've heard, the goal of public health is really thesame goal as -- as the goals -- the right person at theright time.I think you'll be amazed that people with asthma probably not the best thing for thosepeople.Ron s*imer have really
Where people are still receiving telephone callsabout the proper or improper use of -- of therapeuticsfor leukemia.That is in essence, why are we still in the year 2005receiving case reports of -- of people who are not the evidence in the proper way.And the question, how can we -- response tomedication, safety, and we do clinical trials to -- we'dhave to document the outcomes among patients and toexpand those in terms of larger patient populations.Public health approach clinical application effectivegeneralized ability of these real world applications tounderstand the full implications of what happens when weactually take this stuff and try to apply it.So here's an example that I think is perhaps not as old.It's probably got a year I can -- it has to dowith increased evidence without data -- and the mostcommonly used medication for asthma, just regular useas -- I've noticed that adverse effects can have effectsin some cases.Albuterol works wonderfully in most of my pediatricpatients but it's been clear to me that in some of mypatients it doesn't have the same effect.Plays a role in the responsiveness of patients andpatients come arguing that the B 2 AR 16 homozygousrespond differently than albuterol and there's onereference and many others documenting this at thepatient level. the basic science then is really addressingthe evidence about how albuterol and how genes worktogether to affect lung function.I thought that maybe before I retired I would see thistype of information and here weOur knowledge of polymorphism doesn't increase -- Iapologize.Oh, does this increase.In other words, with our knowledge of that polymorphism,the decrease cost of health care and increase qualify oflife and so the public health approach given that -- can we measure its so that's the first step.And then the public health approach really expands even to say when you release this, when you licensethis and when it's used with everybody and people areactually being screened perhaps for this and then beingput on this before the -- excuse me -- they're pgscreened for this polymorphism before they're being puton this albuterol, what happens when you study thiseffect in terms of prednisone or -- what happens in theother leagues for those who may suffer from diminishedlung function.It's actually somewhat a different disease than asthma adults and what happens in different ethnic groupswho carry different genes that may impact the effectof -- that may actually modify the effect of the sector.So in essence, can we study -- can the public healthapproach to say, we need to understand all of this.In addition to understanding how the polymorphisms inalbuterol work together in the global macrosense.So how -- given -- that's a pretty large charge to thiscommittee and so how would we go about collectinginformation in measurable clinical outcome in a divorcepopulation set including different ethnicities?Well, there are really three major options that I couldtalk abouttoid.One is observational studies, clinical trials and largepractical trials and they all have different strengthsand weaknesses and that's what I'm going to walk throughnow.Now, it turns out that observational and co --observational sites can basically be broken down intocase control studies and this is in essence a one stepabove the -- the very compelling case reports that weheard of from the previous speaker and that talks aboutamong asthmatics, so we could basically say among thosegiven albuterol or those not given albuterol, what's therate of a good or versus bad outcome in people givingalbuterol compared to people not given albuterol andthen if you satisfy them according to the gene status, set up how we would look at this in the studyin an observational setting and you could also -- thoseseem to be very large and very expensive but they dogive you information as to whether people on albuteroldo better depending on the gene -- on their gene status.You could alternatively just nest the case controlstudies and pick, say, a couple hundred people who havegood outcomes and a couple hundred people who have badoutcomes among those who have asthma and look at thepercent who slb on albuterol in terms of the proportionsthey make up and patients with that outcome andadditionally satisfy them according to the gene statusand you would actually have evidence that suggests thattells me whether or not the effective -- whether or notalbuterol improves asthma outcomes according to status.It's actually very hard, takes a long time and veryexpensive but it's out there already.We could actually begin to get this information todayand in fact, people are. -- smoking and alcohol, the nicething about genes is that they distribute themselves ina fairly unbiased situation and we get good informationgood evidence about the effectiveness about albuterol indifferent groups.The difference is the sample size of the limitationsreally -- you're set up by elderly, by other children,by other ethnic groups until even somewhat largeobservational studies will run into limitations termsof how much information they can give us.Randomized clinical trials allow you to go out and infact find a couple hundred people with who are ho mozygotes and allow you to randomize albuterol among thetwo different groups of people, among the two differentgroups of gene strata and that would allow you to address whether or not albuterol worked betteramong one or two -- am I shouting or -- oh.I'm not shouting loud enough.That's the first time I think anybody's ever said thatto me.Okay.And the nice thing about it is that you couldadditionally strat fi according to other genes.So you could additionally do in essence a 2 by 2pictorial design or you could say you could random mypeople to albuterol and do a pictorial design that way.So the nice thing about randomized clinical trials isthey allow you to very directly allow you to address thespecific question with the very high quality.The disadvantage of a randomized clinical trial that inhealthy patients and often limited to those on monotherapy and they have very limited generalized ability.I'm 48 and on three medications already. that happens I don't know.I'd like to blame somebody but I always blame my genesso I would not be considered a healthy patient for mostof these trials and most of these trials have limitedgeneralization ability for me.I'm a white male.Usually most stuff is generalizable just to me.The nice thing about random mied trials is that you canstratus fi additionally by elderly, by pediatrics etc.This has led to something very exciting which is theconcept of large clinical trials.Many patients over 100,000 in trials that are randommied patients are at the clinic and provider level andthis allows for head to head comparisons of mostcommonly used medications so it allows us to ask notonly does statin A work better than statin B but it alsoallows us to ask that specific group where they werestatin B were assessed for happen tiet group B and itnot only allows you to enroll enough people to studyvery small different groups that -- that may actuallyhave minor clinical impacts at two of the health impactsit also allow us to utilize the naturalexperiments among the large number of people, if youroll 100,00020,000 of them might be pediatrics andthat's still a fairly large sample size so you can lookat the drug effectiveness by gene status according todifferent risk groups and you could also look at otherfairly common gene polymorphism to look at geneinteraction and then you could look at the modifyinginfluence of other medications so there's really a lotto be said for really strongly considering IErecommending that we integrate genomics into largeclinical trials.I that's one of the more exciting things on thehorizon.The other thing is they not only look at the drug effectthe effect but also the system effect.That is given that we -- we know what's going on becauseit's how well does the system respond to thatinformation.And that's really an underappreciated real worldgeneralizeability feature.So what is the -- what are the needs of the UnitedStates in terms of setting up a network that couldactually address these issues?Well, in yellow in the subsequent slides, you'll seethat I've outlined what -- what I think we need for --for this kind of evidence of effectiveness to becreated.We need clinical researchers, epidemiologists and triallists as a network of researches and I guess what I'mgetting at is that this is a full-time occupation to dothese kinds of studies.This is nothing you can do 10% of your FCE because itreally requires a complete mind set, a mind change. paradigm shift in how you actually think about doingyour studies and who you are going to talk to and so weneed actually dedicated clinical research, dedicatedtrialists that are looking at genomic -- that arelooking at pharmacogenomics and pharmacogenomic tests.And we also need organizations that are willing toaddress these because the problems here is that thesetypes of issues can either be tremendously helpful or itcould show up on some page of USA today.And as you know, in a derogatory or a rather fare smtitle about a large organization studying the attributesof their population so we really need to align ourselveswith mass organizations medicare, the VA, medicaid, totalk how we can actually network our researcherstogether with them to do these large practical clinicaltrials and large observations and randomized clinicaltrials. will need to be brought up to speed and many ofthem will require a tremendous degree of reassurancethat we'll do the right thing at the right time.I'll talk later about the types of data sets for thesetypes of studies. we get this evidence, it will come in a big mixmatch that we call published medical evidence and that all grapple with on a routine basis.And so what we also need is a system somewhere aroundhere that talks about a systematic analysis of drug and effectiveness.And this relies primarily on the format of systematicreviews and formal META analyses.And I'm very pleased to say that there's already beenmovements here where as the E gap project whichevaluates the genomic applications is already conveningand this committee knows quite about this so I won'ttalk about this in any further detail.Now, we had a question, why -- from one of the panelists to why are we still not able to integrate thisevidence?And I think it's clear to say that the enterprise hasfailed miserably in integrating it into clinicalpractice. said this and I'm just reiterating this -- thisopinion that I actually believe that there -- we reallysimply has not paid nearly enough attention to thescientific approach to intre grating evidence.The United Kingdom has already begun -- as for at leastone decade been leading the way toward collectingevidence and to integrate and translating researchprojects but we are still as of June 2005 really onsquare one still in terms of any fundamental successin -- in systematically integrating evidence.So let's assume that the evidence is strong that knowingdata 2 receptor status among asthmatics improveoutcomes.Let's say we do the study that shows it makes thedifference.What's the best way to get this into practice?Still I think in the United States, we -- we are doingit the old way still.The old way was that if we could only educate doctors,this would solve the problem.I -- I'm going to say something very politicallyincorrect.I this is not a waste of time because it 'snecessary and people get mad at me to say it's a wasteof time but what we do is when we educate doctors isthat doctors test better.That's a far cry that they actually apply that evidence.There have been a number of studies indicates that theytest better and their practice doesn't change a bit.I think we can educate the patients and the patientswill have better knowledge but if the doctor doesn't doit that's really not money well spent.We could do academic detailing, high costs and temporaryeffects.Private detailing is not a bad idea except that it's adirected change in terms of what gets sent to thepatient.So I don't think that any of those are really the way we should be integrating evidence intopractice.There is a new movement that is long overdue and that'sto do experimental trials as a means to test the bestway and to prepare and here is one example which is theeagle care for asthmatics versus the electronic reminderwithin the health record IE epic that's being now --ordering based on diagnosis or prescribing behavior andfor example somebody comes in and you give them thediagnosis of asthma and the electronic medical recordsays that's their first diagnosis ever, it wouldautomatly order the receptor assuming that this evidenceis effective that it effects strong clinical outcome.I think it should automatically write the rightpreskrapgs in the right dose.We're hoping to do a similar trial for warFARIN.It's taken out of the physician's hands but in essenceit's not dependent on me to remember to order the testand then look at the test results beforeI write the prescription.So what kinds of systems are necessary to get thisevidence integrated into practice?Well, to do that kind of study that actually requires adifferent kind of pesh.Not necessarily an epidoo*ed deemologist anymore itrequires health services researchers and those are adifferent breed.It also requires substantial EMR development.It takes a lot of time to develop these popup screensthat could automatically order tests that areconditional on a disease being diagnosed and it couldorderI'm not saying it's a bad thing, I'm just saying welack -- we lack this right now.We are not doing that.So finally I'm going to talk what I mean bysurveillance.I've talked about how we could collect the evidence, howwe could integrate the evidence.I still don't think that's the full range of things thatit is incorporated by the public health approach.The approach has always incorporated some degree ofsurveillance that we would need to do.One has to do with quality measures one has to do withethics and what do I mean by quality measures?This would be standard publications.Just like the MWR so the standard publication of howwe're doing with vaccine coverage.I think it would not be an unreasonable approach thatyoung people around the country with asthma, how manyare being --'m assuming now that the state is now solid.I'm saying we should not be dependent on individualpublications that spore radically get published.I think we should have a national system that says whatpercent of asthmatics are being tested before they'rebeing treated and what percent are being placed onappropriate medications to bring results.
I think we need some sort of surveillance mechanismso that we are on the outlook for this.Also the inappropriate use of tests that is you know,these tests being used on the wrong population orincomplete counseling.I think it would be a horrible idea if we just served tests and didn't have any institutionalizedapproach conveying that information to the patient.And then unintended outcomes, whether it be once heunderstands the drug metabolizing effects, I mean,things that we can't possibly conceive of will happenand I think there has to be some sort of surveillance.I also want to talk for one second about the safetymodel that I think is -- is something we -- we whouldreally consider in the vaccine model we currently have apassive reporting system for the effects of vaccinationsand we also have a data link that puts together apopulation that looks at vaccine safety among 5% of theUnited States.I think the pharmaceutical model has something similarwith that recording system that's passive in nature.The ser that projects and a couple of other projectslook at other things.I think in the future we will have a registry of theseadverse event reports of people who have unintendedafter vaccinations and it will be easy, IE possiblewhere we go get DNA and we'll get a candidate genegeneration approach.We'll begin to form a registry of people who haveunintended effects and that these will allow us to thenstudy new candidate teens for their role in predisposingcertain people to adverse effects following vaccinationsand there's no reason why we can't do the same thing ina registry of adverse effects in the pharmaceuticalarena.Well, we need safety researchers.Those are different than epidemiologists and healthservices researchers.They're trained to grapple with these varioustroublesome issues.And finally, I want to talk about the developments ofthe electronics health record.Everything I've talked to you today has assumed theavailability of data in electronic format to collect theevidence, to conduct trials of integrating evidence intohealth care, to provide information that guides andmonitors clinical care, either topic alerts when you'represcribing medication, pop up alerts that may pop upwhen family history is collected or pop up alerts thatpop up when high risk conditions are noted.In fact, none of this existed today and there is atremendous need to develop this type of electronichealth record.Research actually has to be done in each one of thesefive areas.How we collect the information, how we process theinformation, how the data is actually structured in ourdata file so we can study it and secure the transmissionof that data.It's funny to think that when I used to put in RO 1s ORWHATNOT WITH HAD TO address it.We do not have a predominant microsoft industry and soright now we're still at the election that most systemsare home grown systems even in the larger players ofthe -- of the clinical arena.So the -- you could see that I -- I guess what I'msaying is that we need a systematic approach to createthe automated files of the medical records, the networksof providers who are willing and able to grapple withcollecting the evidence of effectiveness, networks ofresearchers who are willing able to do studies ofhow to integrate the evidence into clinical care andwilling able network and researchers who are able todo the surveillance that I think would be necessary forpharmacogenomics.To create this system will take a lot of work and a lotof money and it's not clear who's actually going to leadthat charge.I think that funding could come from these players.Far ma and insurers I think would all have a role forcreating such a system that would allow this to occur.I think that's also a role for ledge stragss andstandards that they could man date some of these things.This is clearly out of my field and I don't want toaddress that.I do want to leave you with -- with one thought.Again, I am the biggest fan of the ability of us to dothis type of work.I think that some of you might have been thinking boy,this guy really lives in the land of fairy tales, wheredoes he get the information from?
Well, it isn't texts.They're only solutions.Everything I told you is a challenge but thatwe have within our power to -- to -- to solve.Thank much.
I want to thank you to being extremely responsive toour charge of tell us what the issues are and the thingsthat we could potentially consider as a committee forareas that we could maybe make some real task force kindof recommendations.Are there questions from the committee for Dr. Davis?
What you defined for us was an infrastructure whichdoesn't exist at this time.The first speaker mentioned that there's the likelihoodthat this may be driven by litigation.And I teach about pharmacogenetics to our medicalstudents and I maintain that the diagnostics will bedriven by litigation.And so that's going to happen much more rapidly, Ithink, than we will have time to develop theinfrastructure that you've -- you've discussed. how -- how would you develop a rapid response when --when the -- the medical legal industry recognizes thatthere's a large vein of gold out there that they hadn't and now create the new cottageindustry against us?
That's a great question.I -- I think there are two things that can happen.One is there's this pharmacogenetics research net work,I think I got the name close, and that's a wonderfulnetwork, one I'm actually very jealous about but whatreally sort of struck me is that there is nothing --there is no network like that for what I was justdescribing.There is a network for what I was just describing forvaccines and it was created because in the late 80ssomebody actually -- well, number one, there were onlythree companies left in the United Statesproducing vaccines and the liability that they werefacing in the court system, the total dollar amountactually exceeded their total net assets and inresponse, the CDC actually formed the safety data linkthat actually does exactly what -- not exactly butpretty much what I've shown you on 5% of the UnitedStates.So we have shown the capability of setting up thesenetworks.We have something -- in response to these litigationconcerns.The certain networks are -- were formed, I believejoined by the FDA and HRQ specifically to look at issuesof patient safety and I think they actually have theresearchers and the network that will be able to addressmany of these issues.Why aren't we doing it?Honestly, it's a matter of money.I there needs to be a substantial allocation ofresources.Right now we're -- well, how about if I stop there.I don't want to start moaning about the small amount offungding that we're able to get for these studies butthey substantially less than the amount that we needto do this in a
I wanted to sort of follow up on that question.You described a system of large population basedclinical trials, which, you know, I really enjoyed youroutline, but as I started to think about, if you had tomake a 100,000 patient clinical trial to answer everypharmacogenetic question that might be posed, what thecost of that is to the health care system and I -- I'mnot going to say which part of the system, whether it'sU.S. government or private should pay for that, but howdo we even begin to grapple with the thought of doing for all of the drugs that are out there?Do you have any thoughts on how one might prioritizewhich things you would start with?
Would no suffice?No, yeah, so that -- that was the honest answer.But you flew me up here so I -- you know, just simply tosay that, you know, I that what I see coming ingenetic testing and pharmacogenomics it has two things.It's really caught the public's imagination and thesesorts things are offered to patients alreadyand it has sort of the stunning ability to bankrupt thesystem.To either bankrupt the system or to dramatically improvehealth care and I think that if you look at it way,then actually the cost of these studies is actually notas much as one might think.I think a lot of the cost of setting up theinfrastructure.I mean, most of these patients in these large clinicaltrials are being seen already.The technology to run their gene tips and collect theinformation already there.It's a matter of putting those pieces together andsending that network to exist.And then you could actually, you then have to actually -- a group of people bho are -- who are wiserand far more experienced than I to prioritize that and Isay that with my pediatric heart shrinking because whogets out in those priorities doing -- they're driven bymorbidity, mortality and costs.These are middle age people or elderly people.I think that, you know, the priority setting needs toalso look at gender specific effects of pediatrics, thevery elderly.I should have just stopped with no.How's that?
Is there some agency within the government that youwould see taking the lead in trying to develop such anoverarching plan?
You know, I've actually wondered about that a lotbecause we don't really have a single agency that sort used that -- that puts public health on -- that haspublic health as its mantle.I think that there is a very clear role for the FDA, avery clear role for HRQ and actually for many ofthese -- well, I'm talking about there's a very clearrole for the CDC although it would expand its -- okay. think
I want to follow up with a question. the -- one of the expenses is thesequence.We can get the testing, we could get the sequencing butif there's an RAS that decreases the price, I was what's the -- knowing that --
Right.So in other words (inaudible)
My question was a follow up seeing population(inaudible)
We really have to step back to a research kind ofstudy.So the drug companies are usually not willing -- and NIHis the exception -- a little reluctant to this kind oflarge study because it's high and they don't see the -- do you have any ideas for --
Maybe we can get you started or at least get your slides up. Can't do much without Emily. There she is. Can someone shut the doors in the back, please? While we're waiting to begin, let me acknowledge Sandra Howard, who's joining us today from planning and eVal nation and hhs. Thank you for being here. I look forward to your participation.
Yes. Thank you so much. I'm very pleased to be here. I do work in the office of the secretary. My office provides analytic policy support to the secretary, who's very interested in the issue of personalized medicine among other aspects of this particular project. My office also provides analytic support to some of the advisory committees to the secretary. If we can assist you in your deliberations, we certainly would be happy to look into that. We've already been discussing this with Sarah and other staff. Thank you.
Terrific. Thank you very much for being here. Just adds a word of -- everyone here who's taking advantage of Reid's absence. He did tell me the only thing he didn't want me to do today was to embarrass him. Please protect me and while try to keep on time as we go forward. Emily?
So we're now ready for part two. Now he's going to focus a little more on his role as a physician and talk to us about pharmacogenomics in the practice of medicine.
And what I'd like to do now is I have a reading stance here is to move beyond the sort of pharmacogenetics one-on-one and begin to talk about the issues which we appropriately have already begun to talk about. That is the translation of this information into the clinic. I think we need to step back and I think this talent and opportunity, Dr. Davis, had something similar. As I thought about how to organize this, I think it's important to talk about it in terms of the science, and I've divided it into basic and translational science and regulatory science and ethical, legal and social science that I as a pharmacologist. I think it's important to put up a diagram like this in which we already have implicitly talked about. And that is eventually what we want to get through is the therapeutic encounter claims the physician and the patient when either the physician writes or Dr. Davis says how the computer writes, and so whatever we end up with so that the patient has the right drug at the right dose. In general, those of us in an academic sense tend to think in academic thoughts like whatever it personally happens to be and a relationship with our funding agency can be the American heart, et cetera, and that we will be able to influence this. That's a sure sign of the approach because frankly, drug development in the United States has focussed on the pharmaceutical and just as the NIH is the place that's predominantly academics. We need to think in terms of regulatory. Interestingly, the amount of interchange between these, that is between say the NIH and fda speaking totally as an office so just as I made the point initially that I spent my life in the medical center, I clearly know nothing about this area other than what I found before it, but it struck me that these two agencies didn't talk to each other that much in the past. What you are going to hear is that dialogue is also important in moving forward with regards to those kinds of interactions that's already been mentioned in the previous presentation. So let me begin by pointing out that although our focus has been on translational pharmacogenomics, Dr. long would point out nigms has been supporting it and clearly we need the basic pharmacogenomics research in order to get to the translaiptional research, and they feed off of each ear. I think it's important to make that point because Dr. Davis was talking about putting a team together. Frankly, we found that our team, clinical investigators that having basic sciences is critically important. Basic runs rite right by what you're doing. We need to be sure that the latest developments are incorporated and the whole team really includes all aspects of health care research. I want to come back to the scientific goal because we were just talking about research institute and what they can offer. Obviously, our understanding of the gee Knolls -- genomes keeps changing. So the nature of stee sequence and structure differences and dna that can have tactical differences, this is a slide I keep adding with regards to the nature of the source of genetic variation that will be important and is important in pharmacogenomics. Obviously the polymorphism -- [ Inaudible ] I already mentioned with regard to [ Inaudible ] Increasingly, we are finding large segmental things. I will show you an example in one second. As Dr. Davis said, you need the basic scientist sitting right there in person in the flesh at the table because your essays will be out of date manana. Whole new areas of genomic sciences are opening up. Or what I like to call pharmacoenpi genetics. What this is showing you is on chromosome 16, a duplication of 145,000 bases, one of the genes we were studying and when they finally got the idea of the genome project being complete is an interesting and ever changing target that that area has one of our genes at 99.9% duplicated in the middle of this duplication of this big chunk of dna. That really messed up our Gino type. The comment was made what about sequencing? Sequencing, if you've got instead of two copies of that, have four copies and you are trying to interpret your sequence phases. That's a real mess. I won't bore you with the details other than to say the science is changing out there. We need to remember the basic science is going to drive this, too. At the NIH, and I put this within the context of the program, the director of the NIH and NIH has gone through the strategic planning ex are size in which they have given it the unusual strategic planning catchy phrases but the concept is pretty simple. The two pathways to discovery means biology is complicated. No one has the way to see all aspects of it so you need what Dr. Davis is talking about. The research teams of the future means that you are going to have to organize the way in which we gained the new knowledge and test the knowledge in new and different ways. Now, I've never done any knockout mikes but if I could do a human knockout, there's really only one gene I want to knock out, the gene for the human ego stressor. Frankly, the biggest barriers is putting these groups together is who's in charge here and we need to find ways that we can adequately reward teams and social interactions in ways that our current system Frankly discourages. And finally, re-engineering the clinical enterprise basically is a need for multisense and multigroup organizes because it's just what Dr. Dave sis talking about. The power calculations are going to kill you. No place, the mayo clinic is a big place, but we know that we have to team up with others in order to be able to have adequate numbers of patients to test these high hypothesis and determine how to move forward. What has happened as a result, though, and I got in a little trouble about my comment about not thinking about Pharmacogenomics but what's happened is the dramatic changes that have occurred that whereas the examples [ Inaudible ] Begin with genotype shoalder to shoulder across the world marching out. They set the next columns and called the Dean. I even told you the names, got the polymorphism and that's in 15 to 20 years. In today's world you put it in a web browser and then you have the gene sequence. That's what Dr. Hansel spent a year of-life to get. Now we can begin with genotype and one of the strategies being used in this area is to rapidly determine gene sequence variations in ethnicity once you have the common variation then to use which of that variation is significant and then the rally hard part, which Dr. draiiveis is talking about, to determine which of the common variation is significant is of clinical importance. And those are among the challenges. This is not the only way to do it. They are complimentary approaches. Let's take a different example. I made an interesting observation myself when I put these examples together. Ppmc, [ Inaudible ] And that is, I said, where has this information come from? And there's an important point here. I'm challenging Walter and I'm challenging Eric because all of this information, all of these chestnuts have come from academics. They have not come from -- [ Inaudible .] We need to find ways to partner with our mutual friend in order to be sure in the future industry is making and being a little provocative here, and that's unusual for me, but let me do it anyway. That industry is making these kinds of contributions. So [ Inaudible ] Is a derivative and its toxicities are predominantly diarrhea and mild depression. This diarrhea is not something you take immode yum for. This is life-threatening diarrhea. Here's the way that now going back to the metabolism. The run on tea can itself is met tab liesed to Form S m-38 which is the active drug which is conjugated [ Inaudible ] And that gene, I have to show you is being stressed because I love them. This is a really nice gene I love to tell the graduate students about. That is it has a whole bunch that are alternatively sliced into conserve four downstream Exxons. Then you get the substrates depending on which you set in. The one that met tab lieses tea can is usually responsible for metabolism and gilbare's syndrome. We now know that's predominantly due to repeat. If you have seven, you have lower level of activity being the promoter. If you have six, which most people do, you have higher level than people who have these like myself. If I go into my physical exam, you are unconjugated Billy Ruben is off and it always is. That doesn't make any difference. But with the tea can,makes the difference. If I'm ever treated with that drug, which I hope I never need to be, I know that I will need a somewhat different dose, a low are dose of the drug. This is to get us through the pathway. It's also to do something else. I'll expose. Here's -- [ Inaudible ] This is from the pharmacogenomics stage which is sponsored by that research effort that I mentioned, and what we're doing is putting pathways there all of the little squares that are sort of this purple color are drugs or met tab liesed all of the drug things that are either metabolized the drug or transport the drug. Now this begins to give you some idea of the degree of complex did I that you will find ourselves dealing with most drugs or the metabolic and transport pathways look like an explosion at the spaghetti factory. So you are going to find that this will become extremely complicated and the examples that we've used are examples of immissty. The real world is going to be much more complex than that. I showed you that because I wanted to be sure that I brought to your attention that the NIH is sponsoring in knowledge-based [ Inaudible ] Where all of the data from the network and we hope from outside the network eventually come together in one place and that kind of a database is a tremendous challenge. Try to combine geno type and fheno type makes this look fairly simple. Having taught medical students and graduate students forever, I've learnd that reiteration is an important part of the [ Inaudible ] Sigh Epps. Let's talk about metabolic activation because imuran is a drug converted which can be menthol lated or ox idizedp that's what I showed you a moment ago. [ Inaudible ] And that's a major mechanism probably for the [ Inaudible ] Drug. I show you this because this is kind of a bow wow pathway, really. I'm showing you the simplified pathway. When we first published our data on ccmt, I will tell you that everyone knows that this is the major metabolic pathway. This is actually a minor pathway. And I thought about bringing along the line from the review for cancer research that says this these dumb pharmacologists aren't smart enough to understand that this minor pathway couldn't possibly influence individual variations and responses. Everybody has those. I didn't bring it along. What was going on at that time was Lynn Leonard had demonstrated that by measuring [ Inaudible ] He can forget who was going to get toxic on these drugs. It might be an international meeting. She said dick, what I can't figure out is you see these kids with exactly the same dose of exactly the same drug pip said Lynn, maybe it's because this pathway can pump more of the drug down here. So she sent us blood samples from 395 con -- 95 convective children from the uk, the united kingdom. We measured the enzyme activity. She measured the niewrk Leo tide level -- nuke Leo tide levels. We also had samples from 1989. Samples from individual figures, standard doses to the drugs to develop life-threatening toxicity. Half of them died. She said us other samples and a group of patients with [ Inaudible ] Diseases were up in the thousands. With the active metabolite. This was 26 days after the drug was stopped and still above any of the controls on the same doses of drugs. When we published this, we said if this can be confirmed, we can prevent and confirm this toxicity and it's been confirmed over and over. That's to make the point that pathway analysis is extremely complicated. What you think are priority, just because something is a major pathway doesn't mean that's going to explain the variation. So the translational lessons are the importance of having intermediate phenotype. Kids with leukemia have a large number of toxications. There are a variety of reasons they will become mildly depressed. If they have a viral infection, they will. But by having the active metabolite, we can control those. In addition it emphasized the difficulty of pathway analysis so when we design these things, 100,000 page study. And please understand, it's going to be extremely difficult to fish out what a given genetic variation might be doing of importance. This is just to make the point that the modified central diagnosis is not because the protein goes to metabolite. That means that the assays that we have available will have to be very different kinds of assays. So the clinical assays will involve pheno types and by that, I mean the intermediate phenotype. It may well be a metabolomic signature. To so it may be using [ Inaudible ] Which at first they won't even understand. But we need to know that during the discovery phase, we'll be looking at all kinds of phenotypes between the dna and what we see in the patient. It's going to become very interesting, but I think we're going to need those different phenotypes that at the clinical level, we'll be measuring not just Smith but also appetites and eventually he was already talking about 3 billion nuke Leo tides. I'd be interested to see how doctors deal with that when the patients come in. Obviously, we'll be talking to Walter in a moment. Significant balances we do know a great deal more about than I do. This is just to make the same point I made before. Walter will be talking about it. I knew he was going to be here. So I used his device as an example. The scientific evolution, let's think about what I've been saying and what we all know. And Dr. Long who's in the audience will be saying we've gone to a complementary genotype which Frankly has accelerated the process ten fold at least so due to functional genomics. Before we Anne have the paper off on the resequencing data, we'll be dealing with our clinicians because they have the dna to test hypothesis. So the clinical science in theory we ought to be breaking down those barriers. With the right organizational structure and with the diminished ego structure, we can actually get there. We've gone from monogenic to say we need to be thinking polygeneralically and we've gone from single genes and proteins to entire pathways from single polymorphisms to extremes and then it will eventually give us all 3 billion nucleo tides and from the mom and pops which is what I've done through most mief career. And we've already talked about all of this. I'm just reiterating things of Dr. Davis. Regulatory science I feel obliged to put this up so poor Eric can respond to it. This is not my comment. It's from surviving the blockbuster syndrome last year talking about pharmacogenomics and there has been some skepticism with regards to segregating out different patient populations to respond. When I do my clinical work, even the mayo medical students know if beta blockers, ace inhibitors and channel blockers. That's not the question. The question is, which one will respond? We're talking about turning the system. We know half the patients didn't. So that brings us back to the little diagram clearly with regard to the drug development process. The role of the food and drug administration and representingtry science has become absolutely critical. And I have clearly made a joke about this since the beginning, but as a matter of fact, it was not a joke. It was true. I have noticed it's taken an interest in pharmacogenomics. We'll be hearing about this later on today from the food and drug administration that since the food and drug administration has been interested in this area, pharmaceutical industries have been increased a tremendous differences. You can generalize but as a matter of fact, there was and remains some resistance to thinking about issues of segmentation of the market as a result of knowing at the front end, which patients will and will not respond to a given class or specific drug agent. At the translational science, we already talked about this. The involvement of this science and drug development process is already going on and you know that. It is increasing and what that says is that all of the examples I've given you, buy yo curie [ Inaudible ] That's the 1930s. These are all examples of drugs out on the market and academic scientists studied them and came to the conclusion there were large genetic variations in their side effects or therapeutic [ Inaudible ] Eventually, a great deal of this science has very significant regulatory implications which I might qawl -- which I'm not qualified to deal with but which I'm sure we need to address. Clinical trials are going on. Go into your web browser and go and look at the clinical trials. Tens of thousands of them and how many of them have pharmacogenomics built into them at the front end. Remember, you already spent the money. That's the point the doctor was making to create the infrastructure to get the clinical data together and send them off. So why don't we make dna a part of that so that you can either prospectively or retrospectively go back and ask the question. Part of the road map is public partnership within the network. We have been grappling with that. There are very significant issues of proprietary interests to expand its barriers. We'll put all of these issues out on the table and talk about them through the course of the day. We need to find ways we cannot just talk about this but actually find ways to deal with this unique problem of each side so we can deal. Finally, legal social and ethical issues, you know much more about this than I do. Confidentiality is just as big an issue here as it is with all other areas of dna testing. Insurance, perhaps a little less, though, because nobody knows, and although we tried what ppmt is there for. It's found in bacteria. But we don't have any disease that if you are like that lady whose daughter works in the mall and comes up and asks about mom's enzymes, we don't know that this means you're at risk for any disease. If you ever find that out, this becomes an issue. For many of those, that's less of a problem here, although it's still a problem. And finally, what do I mean by therapeutic acting? This is not like brta one or two. If I find that a patient has -- is below ttmt, I want to lower the dose of the [ Inaudible ] I can do something right then. So that in this situation, it will be therapeutic active, it is in the area of Pharmacogenomics. Finally, the issue that was raised a few moments ago from the "New York Times" October 10, 2004, the gee Joel -- genome in black, white and gray and what it was entroorly. The issue related today was the hearing for the drug that is being evaluated for the possibility of being approved for only one ethnic group, for African-Americans is being discussed right here. I heard Francis Collins interviewed on public radio about that and heard his comments which were that this is undoubtedly a skin color disease but the underlying genetic variation, so there's striking differences that I mentioned. This keeps coming up. This is 2001 and the New England journal of medicine where there were articles about ethnic differences and responses and converting enzymes and two edit torials taking the kinds of opposed points of view and this which the knows much more than I do. And it was 23003 -- 2003 and day gentleman view all over again having the exact same discussion. And I find this common variance is not found in Asians when I was visiting the professor in Singapore where the population is 80% Chinese. They said this is a problem we see only with the European kids. What's the deal anyway? They were -- they actually have developed a testing to use for Europeans. They were devoted. They came to Minnesota in February to learn the technique. Finally, this issue of health care professionals, I heard what Dr. Davis said. The implication was clear, and I will have to say that in a review that we wrote, we said this would be an important part of what we need to do. We were roundly telling by the sociologist I continue to believe because what I have seen is that the gastroenterologist, you see the patients for years and totally embrace the ppmt and ol colg, resist eabs is one that in that community, toxicity is their business. So we need to realize that the sociology differences within medical subspecialties, too. But if gastroenterologists are hopeful, I think there's hope for everyone. Finally, I want to end where I began by pointing out that this is only one factor among many factors that influence individual variation of drug response. The clinical goals are ones no one can argue. No physician wants to harm his or her patients. We all want to maximize efficacy of the drugs that come out of the therapeutic revolution. It would be much, much cheaper if at the front end we could select a responsive patient. Genetic inheritance is only one fact thert drug response phenotype but in the case of our understanding, the case of increasing understanding is increasing dramatically and the goal has already been demonstrated. You have examples out there that make it very clear that this will Ben fight -- benefit our patients. Thank you very much. I hope I haven't gone too far off time.
I want to thank you very much for that enlightening talk, and throw the floor open for questions from the committee, and I recognize Deb as the first.
This actually isn't directed, it's inspired by our talk, but it's a question to the fda. Why doesn't the fda require gpmg testing before the drug can be used in a patient? Is that within the purview of fda to have that kind of labeling requirement?
Felix, you want to try that one?
Felix, can you come to the mike? Feel free to sit at the table.
I guess I would not be at the fda at the time this was discussed in the advisory committee. It was the first case that this came to the fda from the prospective of personalizing medicine in a drug label. And it's my understanding that at the time, although the evidence scientifically was pretty solid, the advisory committee didn't feel compelled enough that actually attest -- a test needs to be done and is required. So we settled to provide the scientific information in the label so that I would say an educated physician could test the information and move forward and do the testing. Moreover, the issue at the time also was that there was no commercial test available. So that was another consideration that the committee felt was an issue that needs to be addressed for information that is going to be in the label if a test needs to be done. An example for it would be like where a test is required for the prescription of the drug and at the time, thats with approved, a test had to be commercially available.
But it's kind of a chicken and egg problem. Until the fda requires it, then no one's going to develop it. and I don't think since fda is directed to look at safety and efficacy that it's right, if you want to use the term right, for the fda to make excuses why not to protect the percentage of patients who get this drug and die from it.
Maybe I can comment since I had the opportunity to be at both of the public hearings. And I think it's fair to say that the committee attempted to approach this in a measured and judicious fashion. Tpmt was, I think, the first example that had been brought forward probably because of the dramatic effects of the toxicity in the population in which they were looking in which in this case, purely children with acute leukemia. They were not examining the off label applications in inflammatory bowel disease. We need to be quite clear what was being discussed. The concerns that were expressed, and I want to be very careful because it probably must be clear to you that I can be enthusiastic about things so I want to be measured. Were those of the heel ma tolg on colg community that they were balancing the possibility of worrying the physicians and remember we can now cure a previously uniformly fatal illness, and they were worried in trying to express what they expressed. It's not a position I agree with, but I'm trying to be balanced here. The majority of the patients being treated that the physicians might cut back on the thyapurine dose and it wouldn't be increased mortality. So I think that was a reasonable perspective. I did find it interesting because there is this concern that the public won't understand or resonate to these sorts of issues that the most -- I think it's fair to say, the most vigorous advocate for testing were the parents of the patient advocates, the parents of the children with leukemia. One of the moms there had a child who had mild depression. I think it's fair to say she was vairly vociferous in her position, but where the committee came down finally was to recommend informing in the label the information would be included in the label but to not mandate it.
But we've already clearly demonstrated that the physicians don't understand genetics. That's published in the literature repeatedly, so you are putting out there information, you know in the dark. Hoping that someone will do something with it, and that doesn't seem to be a very effective approach.
I agree with you to the point that we also need to make sure that what we put out there can actually be applied in the clinic. Though it's not just about providing the information, but it's about providing a consequence of the information. So in other words, and Dick mentioned the example for which we have an advisory committee meeting in November last year where we under the midst of updating a label because there is actually toxicity that is prevalent in a much higher frequency than for tpmt than people who have a certain genotype with prevalence of 10% this the population have a 50% risk of experiencing toxicity. The question is, however, what are you going to do about the other 50% that do not and might benefit from the drug? So you need to be very careful of not excluding patients that are willing to take the risk of treatment because they have a disease if they want to do so. So at this point in time, I think it's about providing information and making an educated treatment about the physician. So I don't think we have enough and sufficient information to actually determine what the actual treatment should look like.
Can I ask Dr. Winchell a follow-up question? Are there actually in the on colg community clinical practice guidelines that have been put together on how to use tpmt testing and how to adjust those based on those results?
Of course, this committee was a pediatric committee so what we were hearing there was their perspective. It's my understanding that those sorts of guidelines and people taking a leadership role here are at St. Jude's through the community that those dpliens either are being developed or certainly being discussed with regard to exactly how they should move forward. And I think in fairness, it was AAALAC of clearly defined guidelines, and the kind of systematic clinical trials that might guide the practicing physician that was another of the concerns that was expressed. Going from the basic through the translational to actually developing tactical information for the physician has proven to be a barrier even for some of these more well-developed examples. I think that we need to be fair and realistic here and realize that we're just feeling our way into the traps laition of this information into the clinic.
But didn't you say that mayo has guidelines for how to dose in response to the tp --
Mayo has the test available and the homozigos low individuals either are not treated or are treated with one-tenth to 1/15th standard dose and are monitored. The bigger challenge and the one that remains controversial are the 130% -- 10% of the European population. And there is no consensus that I'm aware of, Felix may be aware of one, with regard to the appropriate algorithm for dosing those patients. In general, the clinical studies have been the outcome. They said actually these patients do a little better although they have a little more toxicity for most diseases that are being treated, but it is that intermediate stage between demonstrating that the polymorphism is important for the star 28utga1-a-1 and developing clinically useful practical guidelines. And that's not the sort of study that in the past the national institutes developed with all that enthusiastic about supporting. These are generally old drugs that the drug company are less than enthusiastic about supporting those studies, also, and we come back to what Dr. Davis was talking about. How do we actually develop practical, useful information in the real world? I think that's going to be an interesting challenge for all of us. I would assume we'd be talking about that for the rest of the day.
Okay, Julia.
Dick, I may be misquoting someone horribly, but max [ Inaudible ] Was not going to understand it and they just said the value of the new group would come.
My graduate students say that about me every day.
Do you realist acheological -- realist achelely think, and I'm not sure about this, that people out there in the trenches practicing will be requesting the disease or whatever test it is, to adjust their therapeutic decision. Do you think it's like the current generation is trainable and able to make that -- [ No audio ]
Captioner has no audio. Please turn the volume up ]
Please stand by. Captioner has no audio.
I can tell I will get in big trouble with the ceo of mayo who doesn't know what it is, either. These are very useful to roll out on the road to stimulate the kinds of discussions we're having here. I put it up there for a very good reason. It's beginning to be mufn more complicated than that. Probably there is a pree protein that transports vitamin k. So we'll probably have three or four different genes we'll have to examine to narrow down the beginning doses. If we could do that, though, we would salve a lot of money for the system and a lot of morbidity and mortality. So the fact of the matter is, we need tpmt and 2-d-6 to make the point. They are the cystic Phirosis equivalent on the Pharma -- cystic fibrosis equivalent on the pharmacogenomic side. It will be scatter add cross the genome and eventually and 20 or 30 genes. For many drugs, that's why I made my explosion spaghetti analogy.
I teach medical students and graduate students every day, God bless it. I really have great confidence that if this information will eventually be made cost effective because of the kinds of technology advances that Tim and thinks colleagues do that it will find its way into medicine. We have to find a way to validate it, to prove to our colleagues that it truly will help them pay for their patients and I have every confidence that it will become a standard part of medical practice. What we want to do is accelerate that practice. We're having to learn as we go.
Just a brief follow-up. I think to the extent that these examples are good ones, I think the ample chip example is a great one because it's a wonderful chip and it's gone through lie censor. I think there will be a lot of resistance to its use because I think a lot of clinicians wail Wyll say show me the evidence that my use will improve outcomes. That's what we really need. The buy yo logic underpinnings are very well known. It's tons of fown read about, but I think the clinicians will hold us to the standard of show me that it either cuts costs or makes my patients happier or improves outcomes or some mixture of those. And there's nothing ongoing to do that right now.
Okay. I want to thank everyone for the lively discussion. I think we need to mrve on or we're never going to get through the whole realm of perspectives that we're trying to cover today. The next section is designed to give us some perspectives from industry, and it's my pleasure to introduce to gentlemen I worked with in the past and I know they are experts in their field and will provide us with really good insight into the way the folks in industry look at this issue and what they are trying to do about it. the fist talk will be from Eric Lye. Dr. Lye joins frus flax so he SmithKline. Set vice president for research and has been involved heavily in the genomics efforts within gxk to integrate it both into the discovery process as well as looking at how to integrate it into the clinical trial process. Dr. Lye? [ No audio ]
Thank you. Good morning, everyone. First of all, I would like to thank the committee for inviting me. Second disclaimer, I certainly do not speak for the industry nor do I speak more gsk in general. These are the slides that myself and a few of my scientific colleagues put together, and third, after Richard's excellent talk this morning, the two talks, I think I can go home now. So in the next 10, 15 minutes, what aim going to do is instead of sticking to my talk to cover some of these areas, what I'd like to do is try to focus on some of these topics that either were not covered in this morning's talk or answer some of the questions that have been brought up. So first of all, just a quick introduction for genetic research and gsk. In 1997, gsk formally established genetic research as a separate functional line in r & d. What that means is out of all of the major pharmaceutical companies, we're the only ones that has a separate genetic division within r & d and Allen roseson the head of that. That has a major impact on the research because we have about 600 people worldwide that are dedicated to genetic research. The important thing that was mentioned a few times ago and also this morning at Dr. Davis' talk is that in order to do Pharmacogenetics, you have to have the phenotype and the dna samples. And at gsk, we collect individuals in all of our clinical trials. Phase one, two, three, post marketing surveillance, okay. A number of other pharmaceutical companies have started to do this but not all of them. This is important. Without the dna, you are not going to be able to do the Pharmacogenetic studies. Right now, there are about 20 plus Pharmacogenetic projects at gsk at different stages from phase one all the way to post marketing surveillance commitment. Now, before we talk about Pharmacogenetics, it is important to understand the current drug development process and how it affects Pharmacojep netics and why it's important. Currently, in Ard to get a drug approved, do you phase one stawd I -- study, phase two is to demonstrate that it's effective, and phase three, a much bigger collection of patients to demonstrate it indeed you can replicate this in a large population, meaning in the neighborhood of 1,000 or a few,000. And that's how you approve a drug. Now, most drugs are effective only in a majority of patients, not everybody. This is not something that, you know, new. It's been the public domain and published way back in 2001. These are just a general groups of drugs in the different diseases. And with respect to their percentage of patients that would be effective. And more importantly, all drugs have side effects. I mean, there's no drugs that I can think of that if you take the wrong dose or in certain individuals that do not have side effects and some drugs indeed produce a major adverse reaction in a small, very small subset of individuals, this is reality. Okay? So what I've changed -- so here, I'm trying to demonstrate what types of Pharmacogenetics are we talking about? This is very important because everybody talks about Pharmacogenetics. What exactly are we talking about? Here I show a number of hypothetical response versus drugs with major adverse reactions, so and this is a percentage of patients who respond to a certain molecule, a certain drug. And on the x axis, on the percentage of patients with major adverse reactions. Now, the first group would be up here. This would be everybody's dream drug and that it would be effective and no side effects whatsoever. Unfortunately, as far as I know, nothing like this really exists in reality, okay? And then the second group are down here, okay? These are the drugs that fail in that either they have no efficacy whatsoever or that have some efficacy but their major adverse reactions are so high that you would not carry on into the phase two or phase three. And as a matter of fact, most of the molecules that we put forward, 90 to 95% belongs in this group. This is the group where pharmacogenetic studies is not really necessary because they are effective in the majority of the patients, and they have very little or very low percentage of patients with major adverse reactions. Okay. A lot of the over-the-counter drugs fit into this group. So most people do quite well on Tylenol. Some people that are using Tylenol does not work too well. They have to use Ibuprofen. For myself, Tylenol works great. If I take two Ibuprofen, I'll be on the floor and I've done it. Certain people rea act nicely to certain drugs. Pharmacojep netics is not necessary because basically you can take it, it's cheap and a couple cents and if it doesn't work, it's okay, a few hours of stomach upset. In the a major deal. This is the group where efficacy pharmacogenetics is important. So in this group where you have a subset of patients that are very effective and the side effect is in a percentage that it's okay for the general population but it would be very important for that subgroup of patients. A lot of the cancer drugs fit into this group. So for example, accept it. Lastly, this group were drugs that are effective in the majority of the population, but they also have a pretty high percentage of adverse reaction. And this is the reverse reaction pharmacogenetic studies. So when you talk about pharmacogenetic work, there are basically only two groups of studies. The efficacy or adverse reactions. And these two groups are the Pharmacogenetic studies that we're talking about. Now, what we are dealing with basically is looking into the risk versus the benefit ratio. Okay. What we're saying is that this group, the risk benefit ratio, the benefit is so high and the risk is so low that it is okay. We're trying to use pharmacogenetic studies to increase the benefit risk ratio so that it would go up this way or go down this way to get into this Ideal situation. That's what we're talking about. to address one of the questions Richard brought up in the last talk about market subsetting and how pharmacogenetics is going to kill the idea of blockbuster. Now, I think that is a myth. In that when people talk about it, they don't talk about 100% of the drug market share. No drugs have 100% of the market share. You don't need to have 100% of the market share in order to be a blockbuster which is by definition, I guess, a billion dollars. For example, Receptin is a blockbuster yet it is only effective in 25 to 30% of the patients. It is a myth that you need to have all of the market share in order to achieve that. Now, just a quick slide on how do we exactly do Pharmacogenetics. We have to start off with a bun of of markers and you have to collect well characterized patients with samples from samples from the patients and the controls from all of the clinical trials so that you can have tissues and dn apt and depending on which phase you are in, you are talking about a few hundred to a few thousand and you determine the differences. You do the experiment, the genetic or genomic experiment to determine the pro pile of the patient and control and analyze the data to compare the differences, and then you come up with your answer. In response to one of the questions earlier, I think scientifically, we are there, in that I do not believe that we need to get down to the thousand dollar genome in order to achieve this. Scientifically, we're there. The problem is that there are a lot of other factors that affect the application of Pharmacogenetics to medicine. So these are some of the benefits that will increase the impact and change this benefit and then we can target the group of individuals most likely to benefit from the drug and not experience adverse reactions so for example acceptance. And as a Pharmacosuitical company, we think it will lead to a more evident base drug development approach because for the ones that do not respond to certain drugs, it would give us a need to go into the pathway to ask why did they not respond. And fill the gap between the current drug development practice and to increeps the safety and efficacy of medicine. Now, I'm just going to go through three very quick examples, and in looking at the agenda before I start, I think exam -- I picked examples that I thought would be covered by the time I gave my talk, and indeed two of them were already covered extensively. The first example is how to testing. Her2 is an uncle gene in 25 to 30% of breast cancer patients. Recepton is specifically to this target. So you want to test first to make sure that your patients that overexpress her her2 and then you treat it. So it's a standard approach of using it. Example number two, tpmt, to test and not to pest. This is already coverd so I'm not going to go through this but I have the same question that was asked a while ago in the last q & a session. I was not in this public meeting, but scientifically as a scientist, you look at this information, it is so compelling, you have like why are we -- you ask why are we not testing this? We cannot get this as an example to test. What, you know, hope do we have in coming up with 20 snips after type profiles in order to get it to test? Scientifically it is a great example. So these are some of the things that we can think of will cause for availability and a comeacial world. I think that's already now commercially available. I don't know the cost of it. This could be one of the factors. Change in practice? Well, that could be a factor because no longer are you asking the doctors to tell the patients take two of these in the morning and take two of these and call me in the morning. You can't do this anymore because you have to do the test first in order to prescribe. Lack of physician awareness. Well, if you just put it into the drug label I don't know how many of you have actually read the drug label for tpmt. It is enormous. How many doctors will actually read that label and say oops, on line 39 it changes. Now it tells you that we're recommending testing first. And this is one of the questions that you address this morning talk about is it really AAALAC of knowledge in the physician? The last example is the p-54 testing and it's been around for 50 years now as far as the chemistry's concerned. The molecular basis has been known since the 1980s. A few examples that's been talked about this morning. So why have not they really been taking in Pharmacogenetics and clinical practice? It could be it's a complicated gene family and it's difficult. It's a limited awareness in the doctors. I think that most importantly it is how to get it. You know, you have to have a place for people to order these tests and more importantly, what do you use as a prescription decision? Meaning that in order for p-450 to have a good, clinical application, you have to have interpretations. This I just took it out of the quest diagnostic of the report on 2c6 and 2c19 and this is the one from left court. Now they basically tell you if you test for 2-d6 or in this case what other drugs are affected and how you should deal with it. So you have to have this kind of comprehensive information for the doctors. Without this, it will be very hard for it to apply. And another disclaimer. My wife actually works at Lepcort. So just to make sure everybody understands that potential conflict of interest. So lastly, what I want to talk about is that in order for pgx to be adrapt to be useful, you really have to look at the scientific part and that is what other decisions perceived as a benefit and then for the rest of the general public to be ready to adoptpton and you go through basically from a scientific discovery to a validation to a demonstrated utility into a routine clinical test. With the three examples that I've talked about, Receptin would be up here where it's perceived to be at high benefit by the physician. Everybody's ready to adopt it and it is being used to test fist and you treat later. P-450, I would think that it would be somewhere around in the middle. Tpmt, I this I that scientifically is very high, yet there's a barrier. Now, as far as barriers are concerned, it does not take a lot of people in order to kill this. All you need is a very small percentage of individuals to come up with other factors that can inhibit the application of novel applications. So in summary over the next ten years, we think there will be increased genetic information into the prescription of some of the med caigs. Not often. And integration of pgh in the medicine will help to identify people to respond better than others and do eliminate or decrease adverse reactions. And definitely there's one consideration by the policymakers to increase health care. Now, these are the areas that we can think of for the committees to focus on. The first thing is we have to change the perception of prescriptions. No medication is totally safe, and that is of major problem -- that is a major problem in the general public in that if you tell people that everybody in the United States that 100 individuals die in the United States because of auto accidents, nobody will raise their hand and say, well, we should ban all of the automobiles. That is just too dangerous. Yet, we have drugs that have been pulled out of the market with as few as three or four individuals with adverse reactions. Okay, so this is an education. We have to educate people nothing is totally safe. And pgx will increase and improve the benefits to risk ratio, but that is not going to totally eliminate it. We cannot promise that this is going to be individual medicine for every patient. We can only say that this is going to increase or a targeted population. The next person that you test will have very different genetic backgrounds. And that person might have a side effect. Fear of genetic testing is an important thing in that pgx does not change the patient, does not change the respondent for the disease, okay? You are just trying to predict by giving it a better chance of the prediction. So we need people to understand this and need protection & insurance on the discussion yesterday. Finally, this is we need a support of the research and health care environment in order to make this happen. So the last slide I listed a number of stake holders in this in order to get this to happen. In summary, this is a big dance. Everybody has to be a part of it and play the role in order to make it happen. A farmer can develop the molecule and do the scientific discovery but in order to make it into practice, a lot of the other bodies have to come involved. Thank you. [ Applause ]
We'll have questions after both speakers have given their impressions here. The other speaker in the session is Dr. Walter Cox who is the head of research. Walter has a long history in the area of Pharmacogenetics and was the project leader for the Roche chip. So I'm hoping he can give his perspective. I also want to point out while he's getting his slides up that the committee has received some additional information. Eric was kind enough to bring some of the gsk literature that they've put together to help with education of the community on human genetics. Walter has brought a paper, a nice review for Pharmacogenetics assays which you now have for reading on the plane on the way home. So we thank them for providing those additional materials, and I'll let Walter begin.
I appreciate very much the opportunity to bring my perspective as someone who's from the diagnostics industry to this committee. You'll see from my slides I resisted the inclination to gratuitous think promote the chip. There's not a single picture in there, nor did I pay anyone to put them in other slide sets. Now that it's been introduced, I will use the test to provide you some examples of what some of the challenges were and how this will affect us going forward with various types of tests. I wanted to broadly cover areas that really had more policy implications in where we are today, where we're going and what those challenges are. So the first of those would be developing Pharmacogenetic tests of the sort that we've been discussing earlier this morning for drugs that are already on the market. The new world is, of course, as we've also heard the opportunity to develop drugs and diagnostics together and various concepts around that, that we can talk about. I have personally believed there is a need for large scale clinical studies of the sort that are challenging for an industry to take on by themselves and I'll address that. Health care provider education has already been addressed and then reimbursement, I believe you covered yesterday extensively. I'll bring it up once more. So thankfully, Dick made my job easy and printing all of these well-known examples, the fact that we have many genetic dea term nants. I would like to say that Gino type fee know type ail -- genotype phenotype are jnly not perfect. They are as Dick said one component of an entire picture. So the idea that we'll be able to prescribe a very specific dose based on a genotype is maybe asking a bit too much. I will say, however, if you look into package inserts for a large number of drugs that are on the market today where there is a drug-drug interaction that leads to phenotypically the same types of lacking the genetics. So adjust the dose to the low end of a therapeutic range. So presumably, a physician could use this same sort of information, which they cannot determine than any other way than a genetic test and then adjust the doses accordingly. I think physicians are very well used to adjusting doses and titrating them in their patients. I think clearly having guidance are available. There are papers in the literature starting to provide that based on pharmacokinetics. The particular situation we had was something look a p-450 test is that these drugs are on market and the companies responsible for the drugs are typically sponsoring studies to show what the impact would be to have a marm ma -- Pharmacogenetic test. This that sense, the burden of clinical validity and utility falls on the diagnostics. For people 50, we were fortunate enough that the fda felt these were valid buy yo markers. Clearly, they are being used throughout drug development today and have been for ten years. In fact, the reason new drugs are far less impacted by the polymovic drugs is because those are weeded out. If they have this liability, they often don't make it through the pipeline or there are chemical means of modifying the structure so it becomes less important. Clearly, the fda's expressed a strong interest in some of these examples. I might just take this opportunity to tell you a little about what goes into developing a genetic test and I'm using Pharmacogenomics to cover both genetic and.
ne expression, although I will not talk about gene expression based tests here at all. We just don't have the time for that. Clearly this is another opportunity to use patterns of differential gene expression to predict drug response. For 2-d-6 without showing all of the slides, it's one of the most polymovic close eye you can work with. The number of alleles known and reported doubled so it went from something like 30 to now over 60. So it was a bit of a moving target even as we were developing a test. It was challenging because it had all of those kinds of vair yaigs Dick showed before, and just a plethora of variations. How to get all of those with one test was not easy but was need probable with new, novel technology. I think it's opening doors for all kinds of multiplex assays that we never even contemplated before. Other challenges that I can't resist to mention, there are intellectual property challenges. There was at least one allelial variant I cannot report because there was no amount of money that would allow me to get access to a license for that variant. Analytical validation was challenging for allelic vairance which are not common. Although we worked with many around the world to find the samples we would use to validate performance in some cases, we simply couldn't find a bone that fied sample. So what we did was to make those variance by mute toe genesis and proved that you could detect them. Those are the kinds of things you have to do. Having said that, even noa now as we've gone into larger populations abroad, in China and Japan, we found new variance with the test that we had not had the opportunity to see before. So this starts to be a little bit like drug development in that near phase three trials you've got 5,000 or however many subjects, but when you go into 20,000, you start to see things you haven't seen before. Perhaps it's not so important. We found some that are not as rare as one might have thought and will lead to a second generation test. As more and more variances are discovered there will no doubt be updates. So I think -- one other thing, then, to address is points that have been made about clinical utility pip -- utility. We are actually sponsoring over a dozen clinical studies in various therapeutic areas the larg et of which is 4,000 psychiatric patients to try to bolster the clinical utility that many have seen in case studies and smaller studies that only have 100 or 200 subjects. It is a large enderve ter take on for a company like ours, and so the need for ultimately prospective clinical trials where this information is used to make a differential drug or dose decision or show an outcome difference, those are ones where one could imagine that a public/private academic partnership might be a good way to do those rather large studies. Now, going forward, we're increasingly considering buy yo markers during drug strel development and in some cases finding that these markers can stratify a patient and predict who is likely to respond. So the fda, Wye are pleased to say, has put a considerable amount of effort into providing guidance both in the terms of workshops and public meetings as well as guidance documents for the analytical properties and multiplex tests, for how data of this sort would be submitted by the pharmaceutical industry and how drugs and diagnostics might be developed together. The most recent one coming out in April. There's still a lot of details to be worked out around those. When Felix shows the slide later on this afternoon, I think it's number 14. Think back to what I'm going to say now in terms of the challenges of timing. Those two endeavors so they are in sin cronny with one another. There's certainly some basic process questions about how the review processes go on within two different organizations but most importantly, the guidance document suggests you would be able to make an annal it acheological I -- analytically validated test that when you go first time into man, you have a test ready to go. One frequently doesn't know what the marker is that predicts response. Until later phase two studies. Therefore, in order to demonstrate the phase two in the pivotal trial, you are unlikely to ever have a fully validated ivd test. I can tell you one reason why right off the top of the bat, one-year stability study takes one year. I doubt very many pharmaceutical companies want to wait to hear for that to be done, let alone all of the other development work which is a minimum of 18 months for a simple test. So the sort of questions we ask ourselves is if you have a well validated from an annal it achele -- analytical point of view and you use that to demonstrate the clinical utility and retain samples, can you then cross validate the ivd so that the two can actually merge and launch at the same time? Absent that sort of an approach, it would be very difficult to have these two and parallel without delaying one or the other substantially. Not to mention the risk on the diagnostic side. In phase three, a lot of these drugs don't make it. You will have developed a test that never gets used. The notion that you might have to do two independent phase three trials, I think, will make it very, very expensive to ever introduce Pharmacogenomics into routine practice and would certainly hamper. I didn't mention so much, but I should, that humans are genetically rich. Our dna reflects or ancestry. It is a beautiful thing to see but it is challenging from a diagnostic perspective. People from different geographical origins have different variation in their dna. You need to be broad and encompassing in that genetic variation so that when a test is used in a country as diverse as ours, everyone is helped by this information. In fact, we put a great deal of effort into that chip to make sure that it covered all peoples. It's important as well, we're starting to see even in gene expression differences in somatically acquired mutations and cancers such as egfr where it looks like Asians have different responses. It's not only the genes you inherit from your parents but how they develop. The cdc has provided these statements about the need for large clinical and epidemiological studies. Gervin what I have told you as you go into larger and larger populations, you find larger studies that I think would be enorm A.ously helpful and provide additional background for parm suitical industry. The NIH heard about the research network and there is translational clinical research there. I would hope that we would do more of that and that may be a pivotal case such as the warafarin decoy might be used as an example to show what the real validity and utility of these tests were. Warfarin is one of the most litigated drugs and I understand there is as many as 250 to die from it, itself. This is such a situation that where having search a test could be eenrmously useful. The drug that had 20 million prescriptions in 2003. So it's not something that's going away despite it. It's stel a much-used drug. We talked about education needs. Maybe I shouldn't beat that horse to death. I'm remindd that package inserts have a lot of information for physicians in it if they are able to take the time to readp. How reprovide what vehicle we use to make this information more user friendly and clinically actionable for physicians is a challenge that we all need to face. The one thing I will say is that in area where's it makes a big difference, the physicians get it. I was at the asco meeting. The overwhelming message at that meeting was molecular diagnostics are driving molecular targeted therapies. And in areas of disease where life-threatening disease exists and therapy choices are crucial, this information is used and taken up very quickly. HIV drug resistance is an example of Pharma quo genetics of a viral agent. On on colg, this is driving it. So I think when there is a need, the education comes more rapidly. Nevertheless, we still have challenges ahead of us. Finally, I think I would just like to say the current system is not Ideal for reimbursing these types of tests. When you are trying to find perhaps 10% outliers who have genetic variation and therefore need to be treated differential whereas 9 and 10 are standard, the models for reimbursement aren't there for that kind of preventive action, if you will. Initially, my guess is it will be used more when something untoward happens to understand why it did but we are not yet at a point we can readily incorporate this pro sprektively. Although, it would -- prospectively. Although it would make great sense that when the genetic test done once indpliew ens 15% of the market. You could benefit for life with other agents. Finally, I would also like to make a plea as Dick did for the partnership opportunities that exist in this area between academia government and the private sector to try to bring natural ma Coe genomics to the clinic and provide patients with better health care siewn ir. Thank you. [ Applause ]
In keeping with trying to keep us on time, what we're going to do is take about the next 15 minutes for questions and answers for the two speakers who we just heard from, from industry and then we'll move directly to the public comments and on to our lunch break. So I'd like to ask if there's anyone for the committee or ex officios that like to kick it off, ceiven?
Just to get a better sense of where both companies are coming from. I'm not asking you to speak for all of industry or anything like that. One of the comments, and I think both of you referred to this, is when you are looking at developing various either diagnostic tools or drugs or whatever. There's this argument that keeps coming up about the size of the subgroup. Eventually, of course, with genetics, you can break it all down to wear all individuals except identical twins and even then you might find enough differences. So what cutoffs do you use in your industry for, saying, okay, we've got x-amount of market out there potentially to develop this product. I only ask because again, in these sorts of partnerships that you are looking to develop, the question will be to know, all right, what are your cutoffs, what are your bottom lines and how does academia, how does government come in to help with those kinds of partnerships. The example that comes to mind. Currently, we heard about the testimony going on today about the buy bill, drug and the use of that particular group. Let's say somebody discovers that native American populations after they cross the brid frch Asia developed some sort of -- bridge from Asia developed some sort of variant and no one's going to be running around developing drugs or products for native American populations because it's just not that big, I would presume. So it would fall in an orphan drug category. So I'm interested in getting from you where you would see your sort of cutoffs or limitations.
I'm a scientist so I'll answer the question scientifically. I'm not aware of any hard cut off percentage number. On the other hand, you can look at the history and look at the record was 25 to 30%. Erisa is 30%, something like that. So there are questions that give you percentage.
You said it was about a $1 billion market.
Yeah. So is Erisa simpler?
I don't know the number. I think it would help to as we go forward on who has top nush what directionp.
Glevac used to treat a particular leukemia. Not a huge number and nevertheless the drug is doing well and there are diagnostics available for that. And just this last spring found out when drug resistance arises, there are now follow up therapies for that. So when there's a real medical need and benefit for both therapy as well as diagnostic, I think it's going to be used because the science is driving it.
It's always interesting to hear your comments that diagnostic therapeutic combo guideline that has come out of the fda is not really very feasible. I haven't heard the corporate perspective on that. I've only heard the fda's perspective. I assume that's feedback on what the fda has gotten. Do you have any hopes of ever seeing a diagnostic therapeutic combination coming to the fda? It's more directed at Joe.
Do you want me to go first? We're assuming that they will come in. We don't know what the frequency will be. We have to remember for that combination situation where there is such a risk with a drug itself if there must be a diagnostic test as with Receptin. It's too early to tell at this point in time how frequently that's going to help.
But that combination didn't come in together. I don't think the Herceptin came in together.
They were on panels on the same day.
Oh, really?
Yeah.
They were both developed the same time. I've heard the history wasn't quite so smooth but in any case, I mean, going forward, you would like to do it in a concerted way together. I wouldn't say that it's infeasible. I would say that if you don't know what the marketers are that are informative until phase two, often that's what I see in the real world of pharmaceutical companies that I deal with, including our own, then there's no way to have an ivd final product ready for the pivotal phase three. That's one conundrum about how you align those processes.
So are asr's and lab-developed tests discounted in the ability to bring drugs to market without the diagnostics that's, you know, needed?
Asr's are a possibility, but our position is that mike crow waves are not asr's.
I wasn't referring to mike crow raves. I was referring to asr's. Many of the Pharmacogenetic types of tests, you piblished the variant and we can do it in the laboratory so it doesn't require an fda-approved cleared test in order to be able to do that kind of testing. Does the fda take that into account?
We're looking at that as we go along, but the main object is communication. Really communication. The earlier the better so we can start working out these problems and try to develop them, including how are we going to deal with asr's.
Tim?
To shift subjects a bit. I want to go back to your discussions about the reimbursement issues. If you could give us more background about the reimbursement around the ample chip and where that standsp.
I'm no reimbursement expert but I laid out for our reimbursement folks what the steps in the test were and typically with the cpt codes are used for dna extraction and so on. So the thing that I think is misaligned is using technical steps to put value on a test. My view is, it's what the clinically relevant information is that you are providing that should drive the reimbursement for the test. So if I perform the same procedures and can predict nausea and vomiting from a drug versus whether you are likely to respond to a chemotherapeutic agent and cancer, I think those two tests are predictive informations have very different value associated with them even though they might use exactly the same steps. That's sort of where I'm coming from.
Okay. We've got Barbara and then Maureen.
Jft a follow-up on Kevin's comment from before. I was just wondering when these Pharmacogenetic and jean Knollic studies are undertaken you used the example of tmpt in the literature, you mentiond that the allele of concern is present one in 20 people northern European descent. That was when you mentioned that is not necessarily present in Asian population that you studied. Is there an expectation, you know, maybe not necessarily a cutoff, but some kind of expectation that there would be a diverse population studied before there's a, you know, guideline that's put out about this is to be watched out for or not?
Maybe I can just tell you that for example, in the Pharmacogenetics network, I medicationed in all of the sequencing studies, studies from many subjects are a standard part of what we do. No surprise to a sophisticated audience like that, we find striking differences in the allele frequencies and types in the different pop layings. Remember, these are large studies. I think the point Walter just made about going to China and seeing in anes a yn population some different variants that are functional important is a lesson that we all understand and cloorly that was the implied message with regard. In fact it's what I heard Francis Colin say on public radio with regard to the 42% decrease in mortality. The African-American population treated with the drug where,s nothing in the Asian-American population. What Francis said is we need to understand the underlying molecular mechanisms that are responsible. But the answer is, yes there is a great sensitivity to examining of diverse populations as possible.
Maureen?
I wonder if we could putp slide number five from Eric's presentation? I'd like to kind of talk around that. Obviously, the promise is sort of that balance that we all talk about, and on that slide, you had sort of on the percent of patients with major adverse effects versus the percent of respondents. And your next one, I think we had just finish it up, because it has sort of the balance where you had everyone's dream drugs where almost everyone responds and no side effects on the population. On the other hand, you have 930 to 95% of the drugs that are failed because of large side effects and low response. Now, if you put a third access which is the potential thing that's coming pack to your thing earlier. The target audience. If you treat children with lieu keel yarks you have the drug and you have tpmt, that is a very limited segment. I don't know what the incident is, but it's not the same as heart attacks in middle aged men. So you have that sort of access of the potential pop laipgs to be targeted, and I wonder if we could have a little bit more discussion about those gray zones? For example, I mean, go back to tpmt. Again, I don't want to beat a dead horse where the percent response is very high and you have the percent of patients with major adverse effects is less than 1%. It's somewhere, 1 in 300. Where is that? That as not -- that's not your dream drug. It's almost saying Pharmacogenomics is not necessary if I read this chart correctly. I mean, can you elaborate on that? And the second question is, the pipeline of new failed drugs, the 930 to -- 90 to 95%, is that an order for Pharmacogenomics there? There is a lot of stuff that's being discarded without being studied. And I mean,that a way to save some of these drugs?
So with respect to your first question on tpmt, I think that you have to understand this graph is basically used for illustrations. So how big those circles are sometimes they could overlap. So it could potentially for the adverse reaction go a little bit to the left to, you know,.5,.25%. It really depends on a particular drug and how bad the side reaction is. It could be just, like I said, you could be a stomach discomfort for half a day.
I guess my question is, what is the decision analytic framework here if there is one? Is this just in the hands of the practice of medicine to figure out those pros and cons or is there something more overarching in terms of, you know, devising evidence-based decision analysis model here some.
Well, that's what I'd lake to bring up. I think that's for the committee and fda to discuss. I mean, basically, my understanding on the tpmt is they are saying that percentage is not big enough. That's my understanding in that it does not quite get to the circle to the right. It might be a wrong interpretation. But there are overhappens and a lot more factors than just signs. Now, economic definitely needs to play a major role in this, not just economic of the disease and how much a market is, but also I think that we need to keep coming back to this benefit in that it's not just the side reaction or the adverse reaacts -- reaction that you see on day one that you mentioned. It's actually a long process. When somebody has to be in the hospital for three months because of one dose, that's very costly. So you actually have to develop Pharma quo economic models for add -- Pharmacoeconomic models for adverse reactions. In Europe, they are ahead of us, because the government, the ones actually paying for the drugs. That's why they developed these models and figured out, well, for certain drugs it's indeed worthwhile to prevent the reaction even though they are much less frequent. Because in the long run, that makes sense. It's just like, you know, preventive medicines in dental care. Now, insurance companies pay for preventive care and dental because they figured out that it's cheaper. And until you develop a major problem. So that's the answer to the fist question. The second question is, on the drugs, I did cover that a little bit, and that on the benefit of pgx and this a lot of those failed pause either they are the wrong target, because they have high toxicity, they get into the wrong p-450 and so forth. By doing studies, you actually can figure out some of them why they fail. That's why in one of my subsequent slides, I said provide you more evident-based drug development process.
Okay. We're going to take one more question from Deb and then we have to move on to the public comments.
I realize I have a gap in my knowledge. Doctor, can you explain to me what the Pharmacogenomic research network does? Do you do Pharmacogenetic testing for clinical trials? Is it like a core facility question?
I'm sorry I sort of threw that up here as a map. This is a network supported by multiple NIH institutes. It has approximately a dozen research centers and one knowledge-based/database that stand for the research centers do both basic Pharmaco-- that's why I had the balance between basic and translational both basic and translational studies. Generally translational studies which are related to the nature of their laboratory-based activities and includes in the same way Dr. Davis was pointing out molecular epidemiologists statistical geneticists, laboratory-based investigators so that in our center we're resequencing genes as I pointed out during functional genomics but immediately translating that into studies of breast cancer and psychiatric illness that is drug trairp. In other centers, the focus is on cancer, card yo vascular disease, on asthma ranging from laboratory-based studies of polymorphisms and functional characterization and testing in translational studies whether this information will help us to better either enhance efficacy or decrease toxicity. You'll hear this afternoon, you'll have an opportunity this afternoon when Dr. Long is here, she is responsible at the administrative level for coordinating the network to perhaps ask additional questions. I don't know whether I've answered your question or clarified anything. There are a series of manyarch centers supported by ul-one. It's been going for five years. We've just been through a competitive renewal phase. Next week in Bethesda, the isn'ters involved in the next five-year period will be meeting.
I was wondering if it was a thing like NIH nci had set up sort of core facilities to provide certain kinds of analysis, very proodly across, you know, many research programs. I was wander fg that's the kind of function that this had that could interface with clinical trials in doing sort of blanket Pharmacogenetic testing as clinic ale trials.
It's interesting you should mention that. As part of the road map, there is a senter that's now gone by the board. You are looking at someone who was given the opportunity to write for the network to do with clinical trials why do you think I mentioned clinkingal trials.gov exactly what you are proposing as you know, the NIH stepped back from the original -- we proposed that a region be the United States. We were told that in some cities, in the northeast that alongwood avenue would be a region but I won't go into that. As a matter of fact, that was the concept that you are proposing is exactly the type of concept, which within the network is one of the things we're thinking about in terms of raising the profile of the discipline throughout all of buy yo medical science.
What would it take to do that? It would be nice if there were a competitive arena.
I have to cut off the discussion because we have to reserve the time that's been allotted for the public comment Terry. I'd like to thank the morning panel industry have you much for the information, education and more importantly for your many comments of the things we address. I hope we can come back to you to sort these comments out into bins that we can manage and try to prioritize our work as a committee for additional advice and comment.
Thank you, Emily, for taking care of the morning for us. We now have our public comment session, and as Reid Tuckson noted yesterday, one one of our clinical meetings is to serve on the whole range of health and society eatal reasons. We set aside time each meeting and each day to hear from the public and that's what we'll do now. We have two speakers and I'd ask the commentators to keep their comments to five minutes and if you have written comments, give us a copy of those so they can be entered into the public record. Our first speaker is Joanne from the American clinical laboratory association. If you would just come to the front, there is an open seat there. Welcome, thank you for joining us.
Acls is an association of independent clinical laboratories, national, regional and local laboratories. Our members include small focussed esoteric labs. Independent laboratories and the laboratory-developed tests develop and perform represent a key constituency in the development of this exciting new technology. We look forward to working with the committee as you continue your consideration of the issues associated with Pharmacogenomics and its promise. Thank you.
Thank you. I applaud your brervety. Any -- brervety.
Y questions or comments?
I want to make a comment on behalf of the group that trayed to put the program together. We didn't in any way mean to slight reference laboratories that are doing lab-developed tests. We've recognized the valuable role that you are playing in this field there just simply wasn't enough time on today's program to hear from all Quint sis. We would -- constituencies. We would like to reserve the rate to call on you for a future meeting.
Thank you.
Other comments from the committee? If not, thank you very much. Our second speaker is vice president of tregtry affairs. Welcome and thank you for joining us.
Thank you. Thank you for the opportunity to comment on the exciting topic of Pharmacogenomics. We believe we are uniquely positioned to discuss this as a buy yo technology company and who develops a unique therapeutic comment for all medical needs and as a laboratory service provider of genetic tests and clinical technology. The age of Pharmacogenetics has started. The process is truly iterative. While there's been a handful of notable successes for the drug companies, it's really only the earliest few drops out of the pipeline. Most of the fruits of our efforts will not be realized for seven to ten years from now. The agreement on the systems and understanding of what the requirements are for the realization of targeted therapeutics which are now confirmed need to be in place now. Therefore, genzime believes first, we believe there needs to be a broad coordinated effort necessary integrating Pharmacogenomics as all key constituencies need to understand the role of Pharmacogenomics. It should be education to physicians and other providers to get them on board and thinking about it. There theeds to be education of payors. Education is necessary in a number of levels for the foundation of Pharmacogenomics as a concept, as a benefit to patients and benefits to pairs. More importantly to this committee, there needs to be education and coordination of agencies throughout the hhs, fda for the drug and test development, cdc and cms for laboratory services, cms, cdc for education and NIH for the design of experiments that will be necessary to lead these development technologies. It's critical that the efforts between the agencies are coordinated especially as new rules and recommendations are created. We cannot have new rules in one agency which are not consistent with other agencies. For example, the buy yo marker is deemed valid by the fda. It should also be accepted by cms as valid. Some other examples. There needs to be a shift in thinking about population means evidence-based medicine to targeted populations and cohort outcomes. The whole classic drug approach has been on centrist large populations. Now we're looking at truly just the outliers. There needs to be new methodologies developed. For instance, a perspective analysis of retrospectively collected samples in buy yo banks and "validation" at the recent dia/fda meeting. We came to no agreement on the process on how to do that. Terminology must also be agreed upon upon organizations. Dr. Janet woodcox stated in her workshop on April 11 this year that further exploration of concept of the framework is needed and reassessment of the ideas of validation and perhaps even adopting now gnomen clay chore for validation. We also believe that there needs to -- the government needs to pay to encourage innovation a. Innovation's critical to moving the health care system forward with a fast pace of medicine today, laboratory-developed tests are considered the state-of-the-art diagnostic tests and are often the way innovation occurs in the laboratory. In many cases, manufacturers will not seek fda approval through five or pma's for the products or devices because the roots are not viable because the populations are too small or especially because since the technology is changing so rapidly and the pipeline is so long that by the time you get your test approved the technology has passed you by as was mentioned this morning. For drug manufacturers, it's important to provide incentives such as labels and exclusivivety to to justify the additional development costs and time lines. In doing sork the regulatory pathways must be clear, predictable and easy to implement. Pharmacogeneral Knollics to work, drug manufacturers must recognize the benefit of creation of drugs to be more targeted to the right patients and show better efficacy and safety. We need to bolster the support to diagnostic access. Especially inclusion of laboratory development tests which right at this moment are not discussed in the early fda models. We have submitted more details in writing to this committee but we've covered many of those topics this morning, and we stand here ready to help assist you to volunteer for your efforts going forward.
Thank you very much. Questions for the committee or comments?
Are you going to make your written comments available to us?
They have been provided already.
Okay.
Thank you very much. Appreciate that. We are now at our lunch break. An announcement, first, for those who will be headed to the airport at the end of the afternoon. You should sign up for airport transportation at the registration desk to facilitate getting out in a timely manner. For the lunch break, committee members and ex officios, the lunches we ordered will be outside as they were yesterday. For members of the public, lunch is available in the hotel restaurant as well as other restaurants in the area. And we will reconvene promptly at 1:30 P.M. and continue the session on Pharmacogenetics. Thank you very much. &%C1 &%C1 he first part of the afternoon session we're going to hear three short presentations representing the different agencies within health and human services that work with farm gems. The -- pharmacogenomics. She as oversight over the pharmacogenomics research base. It's a unique position having looked at all the applications that have come in as well as looking at the funded researchers within the network to talk to us a little bit about the state of the art in that part of the world. Rochelle?
Thank you.
I thank the organizers for inviting me. I'm the first of three panelists talking about the department of health and human services. I'll be talking to you about nih, the National Institues of Health. Then I'll be moving on to tell you about the fame the pharmacogenomics research. I looked up and found over 400 different awards supported that have pharmacogenomics. For today's talk I will be talking about grants, community outside of. In NIH. 24 institutional training grants and fellowships. The area shows 50 cooperative areas for pharmacogenomics. This includes some of the large multimillion dollar awards through the department but also clinical trials, any kind they're collecting materials from people and plan to use it for the pharmacogenetic studies. There are two facilities and centers. There are nearly 200 individual research grants. Normally this is the bread and butter from the NIH. I think it shows how they take multidisciplinary teams to approach it. There are a small few business awards and grants.
There are many institutes in NIH. It eye den nice the genetic contribution to complex diseases. Many are banking DNA samples. This is one thing, by the way, that is not done as a pharmacogenetic network, but I'll get back to that they're promoting sharing tool tones able researchers to do better quality research and promoting data sharing activity. The idea being the federal government funds are being used to support the work. The results should be shared. Yet, data sharing is a concept that NIH supports. When I surveys the national institute of mental health, there are treatments to relieve depression. They also have tissue depos tore rise and collecting materials for human genetic studies. It sparse the pediatrics, farm farm a cooling. The national heart land institute is one of our major participants. They fund a significant number of multimillion dollar awards themselves. They also have had a large program called pgas that support tools be they mice or statistical method. The emphasis is on tools and getting that out there. The heart and lung, blood institute support sequencing services. These are often sequences and Gino type and they support research grants. Not all good research takes place at major universities on the east and west coast of the United States. The national cancer institute has multiple large adult and child clinical trials. They're thinking about doing pharmacogenetics of samples. They also have a cooperative human tissue network. They also bank samples and support individual research grants. The national institute of diabetes, digestive and kidney disorders have several trial groups. They have the drug induced, network studying protocol from people with severe drug induced liver injuries. The national institute of ageing has an Alzheimer's correlation, a classic predictor for complex disease, one component. The human genome uses blocks as a tool to look at the genetic contribution that contributes to drugs and vaccines and compounds in the environment. The big effort is collecting and identifying so that investigators can go on to do these sorts of studies. The human genome is also the study on molecular libraries. The national institute of drug abuse has tissue and cell repositories. They're part of the consortium available through the nur rope signs blueprint, trying to raise the research level for all. . Where I am based most often studying metabolizing enzymes because the enzyme systems are related to different classes of drugs. It would be common to cancer, heart disease or depression. Starting around 2,000 we started the pharmacogenetic research. This is the way that the pharmacogenetic research looks from approximately 2001 to 2004. At this time there were six institutes participating. The initiative is undergoing renewal. I'm pleased to say we will have nine institutes and offices contributing. It's really becoming an knishty. They have supported research in drug metabolism. You heard the speaker Earl lier. He looks at metabolizing endisie. another longstanding grant tee looks at the membrane transporters. Each of these groups was charged with putting together interdisciplinary teams. They are paired with genetics backgrounds. People of pharmacogenomic and pharmacogenetics are brought together and they look at samples from clinical studies. You need a large team to do the research. We have had groups looking in the cancer area, both at breast cancer and colorectal cancer and leukemia. We had a number of groups, as I mentioned, a good supporter of ours, they're looking pat both cardiovascular and pulmonary diseases, looking at drugs that lower cholesterol levels in the blood, antiarrhythmia agents, antihypertensive agents as well as looking at drugs used to treat asthma. It's interesting, many of the investigators come fling this side of -- coming from this side would have genotype to fenotype studies. They had families of proteins of interest. They're looking at variations and trying to see what that meant functionally. When we had the first competition, a lot of people came who had interesting patient samples. They wanted to look at the genetic contributions. I call these more of the fenotype or genotype. The network is united by farm gkb. I'll tell you more about that. Farm g standing for pharmacogenetic or genomicks. What the medical, decision making point ultimately might be for predicting responses for drugs, but it is still a research tool. It is not yet a place where a common practicing physician can log in to find out what drug to give to the patient. We're not there yet. There's a lot of research that needs to be done to accurately predict the drug responses. We support the local awards to help groups get started and we supported and award that looked at the pharmacogenetic studies for minority populations. This is a recent slide. It shows any researcher can come to it, can browse through genes, can look at primary data. Can look at pathway pictures. You saw this earlier, can enter simple queries and can start to pull up data. As soon as data becomes human data, you need to have a password to access this site. You need a research purpose, a valid research purpose. You have to describe your research program. None of the information is individually identified. If it gets down to the granular level, a person with red hair in Chicago who came into the study with a certain time, so a lot of thought has gone into this to make sure it is ethically and legally compliant. The farm gen the fame -- the pharmacogenetic research is published in both clinical areas and journals. They're working on actively depositing the correlations between the two. They're working to develop pathway displays that can easily display mechanisms. There are almost no drugs that I can think of that encounter one single gene that goes through the body, one single protein. It's that spaghetti diagram concept. I want to emphasize this is open for the scientific community, sub missions of data. It's not a network only tool. I think this group is still learning as a network. Early on they worked to device policies. For example, what should you put in informed consent. Is that different from just the scientific publication. They work to develop the properties that are not enCumbering. They were asked to deposit their data early on. People want what is important and meaningful to be able to be commercialized. That doesn't mean this can't be shared with others. Developing principles. Looking at ways and comparing ways to do clinical study designs, ways to do more and more experiments. This is more interesting and active area of the network. Another area of the network is for them to share their work with everybody in the research community. They're working in the area of four papers, the first one being an overview. We'll discuss the issues and have recommendations in that paper. The second paper is looking at pharmacogenetic testing and for research purposes what needs to be known. What are the considerations and how will this fit into an ethical framework, how will it fit into a regulatory framework. But the emphasis for this group is results. The third paper will deal with guidelines for educating professionals in the area of pharmacogenetic genomics. That may include others as part of the team. Each of these papers will be targeted to the appropriate journal to get the word out to the community that should be hearing some of this thought and discussion process. The fourth paper tentatively is in the area of doing association studies in pharmacogenetic genomics, but I expect -- I see draft papers, draft outlines. I expect them to be hitting the streets in journals over the next couple months or so this network works to generate sample sets to the depository. I want to credit JULIO. There was extensive community consultation that took place and a real effort in getting samples right and letting them know it would be used for research purposes. Finally many members of the yet work are members who testify at times in front of FDA hearings. They have the knowledge. They have conducted the studies. It will continue. So I will conclude my talk just by pointing out that it was our institute that commissioned and actually had two publications that you have as bro shores. One called medicines for you. The other called genes and populations. These were developed to encourage people about the purposes of research and helping making decisions about joining research studies. They're available free. I encourage you to take copies and request more. That concludes my talk. I would be happy to take questions or delay them to the panel however the organizers think it's appropriate
We'll have the three and have an open Q and A to all of you on the list. Next is Felix Frueh. He's going to talk to us about the specific efforts within FDA to develop guidance documents in this area and we apologize in advance for putting you on the spot for all things related to FDA but you're the chosen victim, I guess, the sack cri fission lamb. &%C1 &%F0
I would like to thank the committee for giving me this opportunity to present an update on the FDA's guidance as it relates to pharmacogenomic. Three days ago I was at a summit. The person that introduced me had a graphic of all the stakeholders who have an interest in pharmacogenomic. With the writing upside down, with the regulators. Then the FDA was on the bottom. I was surprised that the regulators were on the top. So I think we're making progress. I would like to give you a little update. The role of the regulators. Pharmacogenomics has one of the key opportunities to products. What we need to realize is this is really a play of two partners. They need to work together. So pharmacogenomics combine drugs with diagnosisic and the regulation of both to adequately reflect this thinking. I think FDA made it very clear over the past couple of years that we take pharmacogenomics seriously. We have a couple series of guidances that illustrate the current thinking, and I would like to go into this. This wasn't meant to be read. It was meant to illustrate that we have a website up that deals with genomics at the FDA where you'll find all the information on guidance and additional background information that we have. The talk is going to be split into basically three sections. I'll talk on the pharmacogenomics admission guidance, the device guidances. Then I would like to combine them into a concept paper as it is now that was also addressed earlier today. Earlier in March this year after about an 18-month gestation period the pharmacogenomic data was published. We've gotten since a very good response. We continue to receive guidance, which is very useful. Why is guidance important? It illustrates the FDA's approach to genomic information. It's a guide to drug development. It empowers the FDA to make drug development more efficient. We provide several new ways to interact with the FDA. It's a means for Fostering targeted therapy and a new communication tool to share information on a voluntary basis for the first time with the FDA. We have gotten very good feedback on that. It's also an outreach to stakeholders who have expressed a great interest and support in this guidance. It really was the guidance that wasn't just shows up somewhere on an FDA website but has many headlines, so it was a very powerful tool for us to start the communication with stakeholders that otherwise wouldn't have got involved in that dialogue. It clarifies what types of genomic data has to be introduced. It encourages the industry to use it. There's guidance shows how genomic information can be conveyed to the FDA and if one desires to do so on a voluntary basis for certain type of data. It introduces a group and it clarifies how the FDA deals with the data. The guidance does not provide information how to Val Dade genomic biomarkers or how to use biomarkers. We limit the guidance with intention to genomics at this point, although, if you read the guideians and you replace the word pharmacogenomics, I think many of the concepts would still apply. It requires data submissions, which is the main focus of T most importantly for industry, it does not create new processes. So it used to use the framework that we have and put the genomic data in that existing framework. The submission pathway is a submission pathway for exploratory data regardless whether it's part of an existing or active investigation, drug application or new drug application. It's intended to build expertise for developing scientifically sound regulatory policies. So we want to learn with these submissions. It create has forum with the FDA out of the review process. The data that we discussed in that voluntary forum is not being used for regulatory decisions. It's an interaction between the scientists of the FDA and the scientists at the industry or at the company without the regulatory overhead. We received the first submission in March '04. We have about a dozen specials received since. I would say the program is women underway and-- is well underway. We have an evaluation of pretty complex data that we are engaging in. The dialogue has been critical top understand. I think the success was illustrated by the fact that the two companies who submitted the first two are coming back. One of them has come back. They want to see they've been doing work in the meantime, and they want to get our input again. It has been an outreach. We had the first meeting with the Europe mean regulatory agency in may of this year. Japan has published pharmacogenomic guidance. The interest is definitely growing. They have issued a guidance on the instrumentation for clinical multiplex systems which we're moving now to the device arena. The definition is coming from the guidance. It's a device intended to sort signals from clinical samples, so it as targeted at what we've been hearing a lot about and for giving us Sam peels. Now the technology is through a component system. The second one talks about the device. This specific guidance goes into detailing, providing information on such devices that are intended for use in testing DNA to identify the presence or absence of a human marker. The device itself is used as an aid in determining the treatment choice for therapeutics. We've seen that before. The point I want to make is this really for the first time has set a new par a dime in how the FDA is looking at new devices. These are highly complex devices, and we no longer have the option to look at every single data by itself but in a combination and with that the complexity of this becomes a new challenge. For the three bullet points we have heard a lot about so i don't need to go into the detail of that. Now to put it all together we need a strategy to combine devices. In April of this year we publish add drug test concept paper. We are planning on issuing a draft guidance later this year. What this concept paper does is put into perspective a couple of things. If we're talking about biomarkers, we have the identification of the target, the target validation and move it along the drug development pathway to all -- to an approval. We visit that often. We have a label that reflect was we actually see in clinical trials. That becomes a strategic issue. We touched on this earlier this morning. What is critical in this process is this is an interaction between the device area and the drug development area. This, again, puts in perspective what is going on through the process and provides tools and information, exchange opportunities between sponsors and the FDA. If we're talking about how to do these things, I think it's critical to overlay these so we have a smooth process in how to develop a drug test combination. The submission process is a process that can be used throughout the entire development pipeline to discuss novel and exploratory findings that at some point may actually help in the area here and to characterize them. The benefits of this approach are, I think, obvious to us. We can use it for patient strait phication as well as a safety issue. We can use it for enrichment purposes in clinical trials. The labeling becomes a critical component of it and it can be crucial for a company to bring the product to the market. I think the example of Herceptin indicates that the product could be approved. It has the potential to save drugs from being withdrawn from the market and rescue Canada drugs that would be stopped in the drug development process. Strategy, competitive advantages, timing, costs, availability, therapies, the platform choice and the complexity of the platform itself are all critical issues that need to be addressed during the process. Ultimately, whatever is come together market needs to be clinically useful. Otherwise, why develop it in the first place. Often that create as bottle next. So creating a usefulness is critical. In summary, the FDA encourages the use of pharmacogenomics and provide tools to support the translation of pharmacogenomics into clinical practice. The combination of drug therapy and use of devices is critical. We're developing our guidance with this in mind. Pharmacogenomic data submission guideians, the one that was issued in March this year has been well received and is currently being implemented. Regulatory agencies around the world are interested in pharmacogenomics. I think it's fair to say that the U-SFDA is leading the way. I would thank all of whom have been visionary and critical in making all this happen. There is a website where you can find all of these documents in writing. At the end I put up a couple of questions for the committee, perhaps discussion that we have at end of these areas of talks. Thank you very much
Thank you [ APPLAUSE ]
Finally, we'll hear from muin Khoury. He is our representative on this committee from cdc. He will give us an update on the EGAP project. &%C1 &%F0
thank you. I'm the last speaker in the long list of speakers, probably everything that needed to be said has been said. I need to apologize to some of the members because you heard about EGAP. The context is pharmacogenomics. Seems that the word EGAP keeps coming up. I want to tell you what it is and is not and how this works in the world of pharmacogenomics. All these points have been made before. It is a public health issue because it can affect a lot of people. The potential for targeting prevention efforts and avoiding side effects we heard this morning about 100,000 people die yearly from the side effects. Clearly, it's a population issue. Transition from research to practice, you heard Dr. Davis talk about that transition, tran lags. -- translation. Providing public education, et cetera. So pharmacogenomics -- I may be a bit bias, but I think pharmacogenomic is moving more quickly than other fields. At the CDC we have a role in protecting the public from bad things like infectious disease outbreaks. We also want to use whatever technologies are available to improve the public's health. We do a lot. For example, the consumer advertisement in four cities, we talked about this briefly yesterday. We also have our finger on the pulse with respect to the potential public health implications of public health in general. A couple years ago some of us did the paper for medicine which seems a long time ago. There were only 751 genetic testers. We deemed a small fraction had impact. So I wanted to discuss where we are with EGAP and how we got here. Sometimes it seems like an uphill sort of struggle to get to where we were. On the right hand side you have all these committees that have been meeting over the last few years that have been essentially asking for HHS and CDC to do something in this area. Our responses over the last few years, on the left hand side, early on after the task force report, we put together a number of interagencies, HH-S data working groups to figure out what type of data is needed and how to monitor the impact. After the report in 2000 we he started the aids project, which really, I don't have time to go through this, but it laid the foundation for the kind of questions that we could query old genetic tests all the way to the issue. Most recently this year we started the initiative which we hoped would be a more sustainable effort because we learned a lot collectively, both at CDC and in collaboration with our HHS agencies as well and a lot of folks from the ake deem ya and private sector. So we are launching into this project whose goal is to establish sustainable process for assessing genetic tests and genomic technology from research to practice. You've seen this complex diagram when the doctor from our office presented this, maybe not last time but the time before, but to cut a long story short here, the basic infrastructure behind the EGAP is the EGAP working group, which is a nonfederal multidisciplinary working group that interacts with stakeholders and there are a wide variety of them from healthcare providers all the way to regulation labs, industry, et cetera and request evidence, done by evidence-based sensors. They identify gaps in our knowledge. Some of these, depending will report to audiences. The two immediate target audiences for us are consumers and providers. This is not a grass root process by any stretch but more of an educational leveraging process. For those few tests that will emerge, we could refer them for more direct appraisal. Now those two committees, those existing task forces that have been sustainable and have demonstrated their usefulness over time have not been taking on too much genetic tests. They have a lot of applications. They've been reluctant to take on genetic tests for two reasons, first the volume, second, the framework does not have clinical benefits to persons. For most of them, I'm told by members of different committees, that they would return uncertain or incomplete evidence for most genetic tests that exist right now. We don't want that to happen necessarily. We want to describe what we know, what we don't know and leverage land do the pilot project that would allow us to essentially round up our knowledge so that we can move genomic applications faster. We don't want this to be a bottle neck that says don't do this, but this is what we know. This is what we don't know. In order to do it right, more research needs to be done in this area. The EGAP objectives from the previous committees, tremendous knowledge has been gained from the project, which I can answer questions. The existing processes that already exist help technologies from various groups and international experience. The UK, Canada have a lot of evident on the way. Want to create a transparent process, developing and publishing the method and provide clear linkage. We want to develop and disseminate information that's useful to healthcare providers and consumers and e eengs -- eventually the policy make pers. The key objective, which is only a three-year experiment right now, is to evaluate and develop a sustainable process. In January of this year we had a meeting on genome being applications where we had 21 invited par tis pants from around the world. We considered existing and pope tension methods for genetic tests and genomic applications. We have established the working group after solicitation of nominations in March with great response from both professional organizations and individuals. We have a steering committee, and the process was completed late in March. The EGAP working group, represented here, let me tell you we have a world class slate of wonderful people here. The committee is chaired by the member from the university of Washington. Not only do when vast ex-chair of the task force but the current chair of the task force from the Colorado department of public health. We didn't have to twist anybody's arms. We have evidence-based people, clinicians, laboratorians. And we have economists and public health people, so the working group was established. We had our first meeting may 18, 19, a few weeks ago. They're scheduled to meet two or three times a year. They talk about potential topics that they want to take on. Notice, the federal government has no real influence on them. There are lots of stakeholders that can suggest topics. We can take pharmacogenomics to that table. They might want to tackle the pharmacogenomics. They're working on finalizing the framework that was started in the January meeting. The third, and that's very important, they have one working on outcomes. Most of the task force isal health outcome model whereas in genetic and genomic applications, they might want to consider patient family related outcomes. The second meeting would be July 18 and 19. What was also done already is we want to begin sort of they decided as a matter of priority with respect to the application of genomics, look at the ones that are common and important like screening tests, those used in clinical scenarios, prevention, including pharmacogenomic tests and move the focus towards prevention. Newborn screening. There are existing processes in the federal government on disorders. There is a separate process at NIH to deal with rare diseases. The conduction of evidence based reviews would be essentially started in July. The evidence-based process would start in August and September. Throughout the last few months we have been engaging stakeholders. The contractor that's working with us, the list seems to be growing. We have feedback. One news letter appeared on May 6th. This is really, so far, has been a model partnership with our sister agencies. I can say that with no reservations. One of the things we want to do, we want to influence the funding process. First, retrospectively to look at relevant data, some of the ideas of network. All of these things can be leveraged that you heard from the pharmacogenomic research. What we are also doing is developing and implementing a plan to evaluate the process and the impact and value to the health community. So there are two overall types of products, both from the working group. The published methods would be out there. The approved evidence-based reviews, accommodations and lessons learned. We want to disseminate the working group product. Also, derive information from stakeholders on the impact of this process and then data from the pilot studies. Again, I went through this very quickly. Because of the lack of time, I think I'm going to leave you with this image of sort of an interactive process that I think is going to be on pharmacogenomics. One thing to leave with you, this is sort of a step in a long-term process that I'm hoping that the public sector and the private sector and academia will come together to apply to pharmacogenomic and other genomic applications. Thank you [ APPLAUSE ]
Thanks, Muin, for the update. Because the talks have run longer, I would like to take one or two questions while our next speaker is getting set up for her talk, if I can put you on the spot, to come up and get your slides going. Then we'll take Q&A immediately after her talk. Anyone with an urgent question that you would like to address to the HHS speakers? Kevin?
In the FDA presentation but when you're talking about clinical benefit or therapeutic benefit, is there a definition that is implied in that. I'm thinking of children who are projected to be of a certain height or less. I presume when we get into this kind of thing more of those controversies will come up. Is there a definition
There's no general definition. I think the definition is looked at on a case by case basis. Anytime you're approving a drug, for example, so you're basing it on an estimate on what this present time makes the most sense to approve a drug or not. I think that applies for codevelopment situations as well as the drug application that we have today. &%F0
Did you have a question or comment?
A suggestion
Okay
So, which is actually, at NIH, there's a large program of CRC's, some have pharmacogeneticpharmacogenetics. Is there any movement to increase it in the area ya of research eye think to coordinate with other groups that are dog activity in the same area make good scientific sense. As far as the efforts are possible we are trying to identify different groups. The research grant application you're asked to see have you formed the right team. Bebeyond that, it's a matter of networking, getting the right people together. They usually do want to start talking
We'll pause in the Q&Q. I'd like to introduce Patricia Deverka. She a fellow in the genome ethics law and policy. She'll talk about some of the issues we may want to consider
Thank you Dr. Winn-Deen
I thought I might preface my remarks. I was gratified to hear Dr. Davis this morning talk about the need for large observation studies and large clinical trials to be conducted to more clearly study the association of polyMoorism and asthma. I agree with that and aptly put together an outlean for a large study when I was working at a large pharmaceutical company. About four years ago medco had requested me to evaluate this field of genomics and what this might mean for medco's client base and business model. I developed start-up companies, both in an attempt for me to learn more about the science as well as to understand how new pharmacogenetic tests would be brought to market. It was clear what was missing that strong evidence that it was worth doing pharmacogenetic testing in a real world sense. It seemed medco would be a good real world laboratory to study it and asthma was a disease that was highly relevant to medco's clients. They're essentially farm suit cam plan -- pharmaceutical plan sponsors. I proposed the study and took advantage of the fact that they had access to -- accessed medical claims data. I took advantage of the fact that I'm a researcher and used that with those, posed toal by ter ram, how many times folks with a certain genotype had evidence of an exacerbation. The -- so what we actually proposed, we went through long process, was that we would invite eligible patients to participate in the study. We would actually mail a swab to them. They would swab their cheek, mail this back. We would do the genetic analysis and be able to track asthma outcomes and thousands of patients very sufficiently. I thought asthma was relevant because a lot are concerned because asthma treatment is expensive. So I shopped the study around to a handful of medco's most forward looking clients. I got to tell you I was turned down by everybody. It was not that they didn't agree that the science was compelling and not that they had to pay anything to participate. They didn't. They primarily said no because their perception of the problems, the ethical, legal and policy problems with inviting their members to participate. Since I was a passionate supporter and remain a passion in the supporter I decided to pursue formal training to see if these concerns were women founded and if so what could be done to develop practical policies to address these concerns while advancing the science. Hope play in provide as little context for my remarks today. A couple of the folks today said pharmacogenomic testing reflected a paradigm shift. I don't believe it's a paradigm shift. We have lots of tools and experience available to us as well as ethical rationnals. The idea of stratifying patients is not U.N. we know people with elevated cholesterols, blood pressure is at increased risk of cardiovascular disease than those hop don't. I actually think that some of the excitement about pharmacogenomics is due to the fact it's the first technology to come out of an enormous public and private project. I think some of the concerns and the idea that we need a novel framework to deal with these ethical, legal and policy issues, firstly is genetic testing. Clearly, with the sad history of you genics in the United States and people's concerns, that's one reason why genetic testing is a sensitive special. That DNA is special, the idea of DNA determinism as this floats through all of these discussions. I think pharmacogenomics challenges for rare disorders because people are thinking that we're going to have to do pharmacogenomic testing in primary care settings and people aren't sure we can put the same models into the primary care settings. Drug exposure is very common. About 70% to 80% of people who have access to drug benefits fill at least one prescription a year. Another issue is managed care. It's characterized by the cost containment focus. I think that's why people don't trust them. Also public payers like cms. They're going to be a big player. With the cost containment focus, the traditional approaches of managed care like creating restricted foremew Larrys or using therapeutic substitutions run counter to it. So people are concerned that these may be barriers for pharmacogenomics in the most appropriate way. Typally, we have the pharmaceutical industry. They have a rather poor public image. People don't trust them because of their concerns they haven't been transparent about the safety issue of some of their drugs, that there may be concerns over the high prices being charged for drugs. What we're not sure about is whether they chan be trusted to do the right thing with pharmacogenomics or are they going to cherry pick certain aspects to address their pipeline and profitability problems. What I would like to do for you today is to really break my talk into three areas. The last onele' spend very little time on. I think there's a number of ethical, legal policy issues on the research front either with new drugs or existing drugs much the whole series of issues and finally post marketing surveillance, performance of of the test as well as the drugs that are associated with those tests. I'm not going to go into a lot of detail because I think the current system would require a large redesign and investment. So what are the concerns in clinical research? What I tried to do today is provide a comprehensive list of what the issues are but I will only go into a couple in detail for illustration. One, I'll talk about informed consent in the era of DNA banking. Informed consent is a way in which we protect people. Many argue that we have to modify T there's a whole series of private confidentiality concerns. The degree varies with the degree of anonymousation. Coded versus economically anonized. It's very common to use trusted intermediaries that essentially are the gatekeepers of the supply of information from patients and ultimate lit researchers and information that's coded and the potential for discrimination. Folks have described that maybe pharmacogenetic testing would reveal a group of patients that would not respond to a drug. If that was the only drug to treat a condition that could be problem mat tick because a lot of people would be could be certained that it would be more expensive. Harms to families, harms to individual families or groups. Whatever you do in pharmacogenetic tests you don't just learn about that, but you learn about disease susceptibility. When you test, it dpifs information about how -- it gives information about how someone will respond to statin therapy and it can also give information about susceptibility to Alzheimer's disease. Another category, the race related issue. Byle has been linked to the field of pharmacogenomics. The whole idea of stratifying individuals, particularly with pharmacogenetic tests have made people concerned that we would create new orphan drugs. That is unique to the field. We have heard that one of the benefits of pharmacogenomics that you can do smaller, faster, clinical trials if you select people for trials on the basis of their pharmacogenetic profile. That may result in having less safety data. By the time the product comes to market we know the doctor doesn't always prescribe according to label. Then finally, a big, big topic, and I won't go into really today, do we have the right incentive structure. Clearly intellectual property issues are critical. People are concerned about patent bottle necks. That's due entities holding patents, there by driving up the costs of having to obtained molt Mel licenses and translating into tests that are quite expensive. Then the focus by the pharmaceutical industry I would argue predominantly on new drugs, not necessarily to study marketed drugs, whether they're branded or generic. More than 50% of prescriptions written in the United States are for generic drugs. Those companies have no resources to do pharmacogenetic research. From a public health perspective what can we do to alter it. I said I would spend a little time on viral repos stories. We know these are being done on a mass scale. They're different. The folks or collecting the sample may not be doing the research. You're not asking for informed consent from a single study. You can't specify -- you don't know what the study is in the future. A number of different groups would try to take advantage of the viral repos tore rise. That takes it away from the traditional emphasis to protecting subjects from informational harm. What facilitates this research would be blanket research. Yes, you can use my specimen for any future use. From an ethical perspective it may not be sufficient. That may be too broad. There has to be balance while asking people to consent to some studies while recognizing it's difficult to to go back and ask them to consent. Documents today talk about risks and benefits to the individual. I think that's are a brair ry. Practically speaking we should be talking about risks and harm to groups, which can lead to the potential for group harm. For example, if you found out for a serious discease native-Americans were not responsive to the only drug that treated that disease, that there could be a potential for group harm that would be stigmatizing to that group. There's clearly a lot of debate. The participants have to have some measure of control. What's done is folks are asked to give a tier consent where they're asked what type of study they're willing to consent, any type of study, cancer study or just a breast cancer study. There's a lot of discussion that the studies can go on for many, many years. Do the investigators have a duty to contact participants years after a study is complete if the study reveals important results that could impact a person's ability to use certain drugs. The results of the studies are not validated. You're dog more harm than good by giving them information. But we would have a duty to contact participants. What's done now is to separate the informed consent for pharmacogenetic testing. You can say no to one, yes to the other. That's done for practical reasons because people are concerned because of ethical concerns over the DNA testing and the biobanking procedures and from an ethical stand point of I think what we're trying to do is strive towards the roast balance between Fostering pharmacogenetic research. We heard how the pharmacogenetic research network is trying to address this issue. I'm aware of the national cancer institute having a workshop next week talking about processes, and I think that will be the key. Will we be able to Harmonize the approaches. Let's spend a little time on the concept of race. There's no precise definition, sort of the prevailing thinking from a social perspective is that race is really a social construct. It's not biologically defined. We know from research that certain variants are done in some eght Nick and racial groups. There have been published studies in demonstrating differences to treatments. A lot of people debate scientific validity because they say self-identified race is an imprecise way and you can get a lot of noise when people say they are African-American. There's an advisory board today and there will be the first the -- eght Nick drugs. There's no information that explains why African-American appear to do better with bile. I think that pharmacogenomics can resolve some of these problems. They would say it's better to genotype than to ask people what the race would be. The potential harm is that we're going to be reinforcing notions that racial differences have a genetic basis. Emare concerned about that and statements about how a drug works in a particular population. We've heard that there are important differences in the distribution depending on where the study is done. I think from a practical standpoint it could be marketed to a particular racial group. You can give -- you can say that all African-Americans would benefit from it or this particular drug is more effective than other nonracially dephened medicines and we know that's not true. There's a theoretical concern. That group could be stigmatized. I think ultimately people think that physicians will take a shortcut and use race rather than genotype. I said I would talk about other types. You can find out that a particular drug is safe for a particular genotype. These people are a difficult to treat subgroup that we don't classify. Or it might resthreel a disease formerly thought of as large, is really composed of subgroups, individuals with the disease and no one of those subgroups is large now of attract commercial investments. That is the potential concern, is that drugs will not be developed for the subgroups. Firstly it's not attracted to a large pharmaceutical company because of their size and scale and their commitments to wall street. Might be very attractive to a small start-up company where they don't need to make billions of dollars. I think the ethical concerns arise when there's no other safe treatment for the disease. It's unlikely it would be so small that -- clearly we must be in the context where we're dealing with serious deseases. I think if those conditions are met then we would have ethical con srns. People have talked about modifying the orphan drug law. To hear some of the issues in clinical practice, we heard this all morning, I'm concerned that pharmacogenomics is coming into the marketplace without adequate validation. Professionals, such as pharmacists and physicians have a huge knowledge gap about genetics and the difficult about determining probable information. When I would talk to payers people would be excited if they could have a scientific rationnal. I think the nuances of where the cut point should be, where the threshold for saying I'm justified in denying access. That's where it's difficult. When are physicians obligated to offer a pharmacogenetic test. When are they actually obligated to follow these test results. They come back. You have a 30% response. Is that too low? Then I think, a lot of folks say the field will and vans if we focus on lie bit. It's not just liability for physicians but pharmacists. It derives from negligence theories because they didn't offer a reasonable standard of care. Pharmaceutical companies would be liable because they didn't disclose a potentially knowable information. I think it's important top understand that that is a real possibility, but I think it requires that pharmacogenetic testing be reviewed as the standard of car. Folks are saying informed consent. So we'll be thinking of this more like a cholesterol test or should we have it viewed as we do for Alzheimer's disease. I think those are two extremes. At least initially, we'll probably be somewhere in the middle where we'll talk about how to treat, use this information to guide therapy. Because the test is linked to an FDA approved drug and the doctor has already made the decision to prescribe a treatment, I actually think the pharmacogenetic testing will not be that controversial. Inappropriate use of the pharmacogenetic testing, that's all direct marketing. I know you covered that yesterday. That might be controversial. It might be appropriate for consumers to do their own pharmacogenetic testing without going through a physician. The other one, psychosocial harms. There's also the concern not that you learn other bad things about the individuals but you learn band things about their family members that they're more difficult to treat have a certain disease risk or the disease might be a proper progressive form. Discriminatory uses. I know everyone's in support of the nondiscriminatory legislation. I think folks felt like that sort of legislation is necessary to help people feel comfortable about getting genetic testing. I' concerned about higher drug costs. We heard that Herceptin was over a billion dollars. Well I've done a lot of cost effective analyses. One reason it could be over a billion dollars, not because it's very expensive. Pharmaceutical companies say even though they may develop it faster and cheaply, I don't think they will pass the savings on to the consumer. I think the higher drug cost are likely what we see. Then we talked about this, that there is a real problem of rapid and unmanaged introdiewx of genetic testing in the marketplace. I would just say here that predictive values of pharmacogenetic tests are too low to be clinical useful. On all the genetic testings that have been done, in retrospective, what's the genotype, you have to do prospective studies to determine what is the positive and negative predictive value of these studies, so we're going to get all excited and shift it from a more effective way. I think we talked about the other points. So payers have a lot of insight. These are the hopes that they have on how pharmacogenomics may be used in the real world. We're hoping there will be decreased costs, for all the reasons listed here. There's also the concern in reality like any new technology, it will actually be cost increasing. It will be cost effective but not cost savings. That's for a number of reasons. I've given the reason for higher drug prices. It will cost more for safeguards. There's concern that patents could be extended if you combine the drug and toast together in a specific use. Right now we're not paying for many of these tests today. If we do broad population screening, those will add up over time. This is just a little bit how they might think about pharmacogenomic testing. You know this. Whether this becomes an important element of clinical practice depends on how and whether and how to is reimbursed. I think we really need to think about pharmacogenetics. So today we're having foremew layer rise. We're asking consumers to pay more out of pocket. We have prior authorization. Seems like pharmacogenomics is clearly on par. It does tailor the drug to the individual. It's desirable not to give swlawn drug that you have evidence that would show it's unsafe and effective. It's also ethical at the group level. There's an obligation. And I think that that's really difficult to operationalize because of the problem. I would argue that the cut points will change depending on the disease, depending pont severity of the side effect and predominantly because of cost. Payers are very worried that that will break the bank and pharmacogenomic testing will put people into receiving it or not receiving it. So that's the longstanding new technology that has existed between what's rationnal at the policy level versus what's rashal at the individual level. I might say I want everything that will benefit me. I think all of this is predicated assuming the tests are reliable and predictable. I thought I might be provocative to say a blanket statement. They should never do testing to consumers. You have to have the unique starnsd. You need to convey the results, but a lot of the start-up companies say they know that. They know for people to buy their product -- bus -- because they do cost hundreds of thousands of dollars. Dietary things will be affected. How can we say you can have access to a drug over the counter but not access to the test. So, for example, if we found out -- let's aassume there could be a test to say who's at risk for cardiovast czar or ansaids, when people have insurance coverage that's an appropriate setting. When individuals are concerned about discrimination so they want to go around the system because they're afraid thir employer or insurer would get access to the result. So you would need a separate framework. Pharmacogenomics comes down to this slide. You can tell my bias, but I think it's important that I share w-with you the reasons why people have said yes. The DNA is uniquely identified. We all know he that from CSI and trials. Niece thinks can -- these things can live in banks for years and years. It's uniquely predicted. There's also the view that science is very complex, so we have to treat it differently. Then there's issues about race, stigmaization because of the likelihood of generic variability. I think we should really think of pharmacogenomics as a prescribing tool, helping physicians decide the best intervention. You're not going to give out a microbut something that's more die jepsable. I think it's really important for us to acknowledge, if you don't you're reinforcing the bad ideas. I think ultimately it will make patients less willing to be tested. So far we don't have strong evidence of genetic discrimination. I would argue that it's less likely for pharmacogenomics. So in conclusion, pharmacogenomics highlights the needs of the long standing problems. There's lack of information across a number of areas. We heard about that today. I think we need to think about how much political will we have and -- to support changes. One thing I didn't talk about, it's clear that the information technology that will be necessary to support this is going to be huge. People are moving to standardization. There's been a lot of investment but that's clearly an enAbeling piece -- enabling piece. My experience, payers still decide on price. We don't necessarily understand cost effectiveness information and we haven't made explicit the values that have to be built into any cost e texttive analysis. So let me end there. Thanks [ APPLAUSE ]
Thanks very much. I would like to go to Q&A. Are there any pressing questions?
I have one question. It was a very interesting presentation. There is a whole kind of literature lain line of thinking which has a lot to do with privacy and right to know and all of that. In a consent document, unless it's clearly specified that the person wants to be contacted in the future, he's not contacted. You're talking about something that may or may not a and some people may not want to know. In the case of pharmacogenetics I see it differently. You're talking about the drug that the person may be erks posed -- exposed to. So if that person did not allow to be recontacted and you know the person has a vearnt of a gene that could cause adverse reactions to a drug, they don't respond, they could die. Is it ethical not to give the person information when there is no charity about that or even when the person says I don't want to know about my gene in general but you know something bad is going to happen, how ethical or unethical is that. Do you use the same standard of ethics for genetic testing or should the standing different?
I think it's important to always allow folks the option not to be recontacted. I know that's common practice with some genetic testing for disease sus eptable. I -- sus sentable. I'm trying to imagine a scenario where you would have information that would affect their outcome that would not be, either would be no other treatment for a serious condition like cancer. I think you have to respect their decision. In most cases people don't really have a means of recontacting folks. Either the samples are permanently anonmust and there's not a mechanism to do that. So from an eght cam standpoint I -- ethical standpoint I would follow their wishes and consent
Thank you for your presentation. I thought it was good. At one point, you talked about the higher cost of implementing some of the privacy standards. I'm not aware of any data that shows that. I wonder if you can talk about that a little bit more
Folks have talked about the cost of implementing HIPPA. That that graphic is a hypothetical for the increased costs as well as the potential costs that would decrease healthcare costs, so I'm not aware of specific studies that talk about the cost of protecting genetic information. It's logical to me to think if we're treating that information differently, that this will have a cost associated with that. Kevin?
I know you were trying to go back and forth and balance yourself between is it a paradigmic shift, so how do you see the way forward for a development of this technology and the emphasis on the importance of this technology while at the same time the things that cash out from this naturally in people's minds when they hear about the power in this technology
In addition to what I already said we have a framework for evaluating new technologies. It last lot of deficiencies. We're not well served by putting this in a special bucket. I just lost my train of thought. Sorry. Can you say your question again?
It seems to be, from not just empirical evidence but the way one looks at the various frameworks, if you push this or just talk about the potential for this, it's going to be interpreted, absorbed or seen by many people as furthering a genetic essentialism
One major step is the vocabulary. I think people have talked about not using the word genetics when we talk about test profiles. We when we say we would like to talk about a test that would guide you. It has a completely different connotation to see if you're at risk for a disease in the future. One of the big things we can do is pay attention to the vocabulary. That's sort of my remarks. In the research setting our ethical position is to db some do contain the potential risk for discrimination and we need to disclose that and allow them to make an informed decision about that.
I had a couple FDA oriented questions, so I'll put them out here on the floor land let which everyou guys from FDA want to respond. We heard a comment this morning from the folks who are enveloped in developing laboratory developed tests, that they would like to see some recognition from FDA that those tests have some status in terms of if the biomarker is validated, that a test developed could still be used in pharmacogenetics, why or why not. Currently it seems from the comments we mered this morning on tpmt and the white paper that there's really no formal recognition or utilization of that mechanism by FDA as a way to provide pharmacogenetic services.
If you're talking about the biomarker as described in the guidance document and talking an lit cam only and there's no clinical validation, you can get an answer, but that won't tell you what the possibility is of being resonsive to the drug or developing a toxic reaction. If you go ahead and develop the test you can get it marketed. That's a simple answer
Let's take TPM3. FDA did fall short, while they said tests are available, they didn't really acknowledge the only way to test today is through laboratory developed tests. Is there a requirement that we move to an IBD before we have something that's formally recognized in labeling as a pharmacogenetic test
Other than a biomarker, yes. If you want something beyond that, then you have to go through the regular approval process because --
The ability to make a utility claim
It's not an FDA approved one
A test result produced by a laboratory is a research product?
The test itself is researched. It's not an FDA approved test. It's not clinical validity. Did that help?
It raises concerns.
okay. The test is not FDA approved. The only way to get the approval is to go through the process.
that, I understand, but in the practice of medicine, is that we can't say that a test for this entity be performed. Seems that for GLEVAC --
You want to try that one?
I think there are two separate issues here. One is a combination product or codeveloped product where a test is required for a drug to be used. Those tests need to be FDA approvaled. Beyond that, I mean, in many, many drug labels, probably 100 or more we point to pharmacogenomic information particularly in the area of me tabism. -- me tap live. If you're looking at the polyMoore fix in drugs for depression, that the drug is influenced, but it's not a severe toxicity. The recommendation is not there yet. A lot of this information has come forward over the past few years and the drug is a lot older, so we don't see it in the label, but the development in recommending that the test is being done is definitely going to be part of the label and there's no problem in putting that in the label, even the absence of an FDA approved test.
Okay. Other questions for this group of speakers? All right. Well we're going to -- thank you very much for your presentations. We'll take a 10-minute break and resume promptly at 3:15. &%F0 &%C1 &%F0 ?5A70?
I guess what I'd like to do is fig see if we can figure ?9972? out if there are some familiar airreas, well, two or three ?9B72? things that I think we should work on.?9D72? Do knee we need to try and ask staff to put together a part 3 ?A572?6-16-050233 to this program, we had part one this morn ging, part two ?A772? this afternoon.?A972? Do we need another half day or so of information ga?tther?A972? gathering and edgeucation.?AB72? The other is, can we try and bend some of these things ?AC72? into different airreas you know, are there research ishsues?AE72? issues, consent ishsues into some kind of lodgegical groups ?B272? that we could somehow tackle into makeing a summary ?B472? report of where they are and then some specific reck men?B672? recommendations for what this committee would like to ?B772? see happen.?B872?
Let me just take the chairman of the day prerogativeto try to frame this the way we dealt with largepopulation studies yesterday which is to get thecommittee to focus on what kind of direction can it giveto the task force so that the task force onpharmacogenomics can make best use of its time betweennow and the October meeting.The issue as I was listening today is for thecommittee to decide, are there still issues and gapswhere we feel none of these existing groups are tacklingthem and/or do we complootly lack information.It's going to take some discipline to keep ourdiscussions along that track.There are many interesting questions aboutpharmacogenomics but some of them may well bedecide areunder control and being well attended to by existinggroups in which case we don't have much to do except topay attention to that and monitor that as time
So people still have their hands up.So we'll go Kevin Agnes, Cynthia, did you?Okay.So we've got four people in the queue here.
Okay.As a member of the task force, I just -- a couple ofother I'd like to be able to see to get input.I think one of the things that I'd like to pursue alittle bit is I'm not sure that we had people set to I'd like to get more perhaps the financialside of the industry as to what their parameters are onsome of these issues, in familiar as we heard the desirefor partnerships with academia, with government and thatsort of thing. just want to get a better Zinns sense of how thatwould flesh out, that partnership, and also, I'mwondering where the judiciary is even in the geneticdiscrimination sort of thing.How do they see this cashing out.
mean, are they waiting for the lawsuits to come?
I'm just wondering what's their perspective on allthis.What do they see as the red flags and things like thatthat we're just hearing.I don't know.I haven't heard any of that yet so I'm just wondering ifit's possible to get somebody in October to speak to uson that.
Okay.On the financial aspects we also really didn't hear frominsurers
Right.
Is there some interest in trying to hear frominsurers as well? I think we had to have that -- I don't know if would have to be somebody necessarily from eachindustry but somebody that has that information orstudies that information.
Right.Agnes?
I think this was Sam had brought this up earlierabout the electronic health infrastructure, and I thinkthat would be something that we would need to hear alittle bit more on both for the area of pharmacogeneticsbut I'm sure it's going to have impact on the whole areaof personal genetic information that we should be moreup to date on.And the secondary ya that I just have a question on wasthat for the task force for the large population studiesthat where -- is there an overlap with what we'relooking at in the pharmacogenetics studies inpopulations, possibly large populations with the largepopulation study that we're -- that we're examining forour group? You want to just take that up?
I mean, there certainly are some questions that willbe in common to those two groups.And there's also substantial overlap, I think, betweenthose two task forces so I think we all need to bemindful of that as we go forward but it's a good point.
Okay.Cindy?
Because I work with Congress, I tend to have tooversimplify things so maybe this is too simple for thisgroup, but I was listening to everything that peoplewere saying and I divided the remarks into kind of aflowchart.Over here was research, the pharmacogenetics, theresearch needs.And then once you get the research going and you've gotsome conclusions and all that, then the question was howdo you integrate that into practice. are sort of two main issues.Leaving aside the integrating clinical practiceseems to me that there are big, big gaps in the researchthat is being done or that is yet to be done and so Idivided that further.Research with regard to existing drugs, drugs that havealready been approved.They received FDA approval so do you do there, whodoes that research.Is it the pharmaceutical companies, do goback and do some research on their own products that'sgoing to be improved, is it academia, is it governmentand how do you coordinate those?I think we heard a little bit from -- about that earlierThere's got to be some mechanism to coordinate thosethings.Is there a systematic way of conducting research,pharmacogenetics research on drugs?In other words, that it's not ad HOC it's not that someguy decides hm I'm going to go look at this and onepharmaceutical company says we'll go back and look atour drug -- there's got to be a systematic way to do itso how do you coordinate that and then the other box isof course pipeline drugs and in that case to me burden would fall on company itself because ones that are inventing the product.I mean, nobody else has access to that.I mean, they are the ones, so if it's a pharmaceuticalcompany, how do you get them to do that kind -- thatlevel of research?Do you have a man date?Does FDA require it or is it more an incentive basedsystem?Seems to me that a lot of different questions andsubquestions and ethical questions that we can put undereach one of those but that was my attempt at kind ofsimplifying what we heard today, the things that we'regoing to be faced with.And so I don't know who else we need to hear from as faras that goes.I think we got a good base of it but I'd like for us asa group to contemplate what can we advise the secretary do so that we can really encourage this kind ofresearch, both for -- in drugs and then inpipeline drugs and who's the best entity or industry orsector to do that?
And I would add even under approved drugs there's twobends.One is where you know the bio marker and one where youdon't know the bio marker but you know there's some sortof adverse event that you'd like to know the bio marker and I think that's two different bends as wellwithin that group so I think the task force coulddefinitely consider trying to make a flowchart andcoming up with some tentative outline of who might bebest suited to do that to throw out on the table fordiscussion at the next meeting.Debra, did you have some more commentary?
Yeah, about what we'd like nor information on andthis kind of ties into the framework that Cindy justpresented which was very nice.I -- I do believe that Japan has mandated that allexisting drugs be evaluated for pharmacogenetic impacton the Japanese population, and -- and maybe it would beuseful to hear how they are doing that and how it'sfunded and what they're actually looking at. don't know a lot of details about it.I believe knack mur ra is one of the major researchersinvolved in the Japanese FDA equivalent.I don't know what that's called.But like with the bio banks that we heard from otherpeople doing this, it might be interesting and I don'tknow if there are other ethnic groups or -- orpopulations where this sort of thing is being done, butat in Japan, it is.And then the second thing is with the FDA presentationthere was information that several submissions ofpharmacogenetic information have been done.Are you willing to share what the FDA is learning fromthat process and -- and when?Because one of the things is, with drugs anddevelopments Cindy, you were saying is there an FDArequirement for the pharmacogenetics, I think that'swhere FDA is moving and so can you give us an idea ofwhat you're learning and what your time line is to bethinking about making this part of the FDA approval rather than a friendly submission ofinformation.And I don't know that you have to do it now, but maybe something that could be done in the future.
I -- I'd be happy to, you know, present you all theseanswers actually.I've just put a presentation together for that veryreason because it's now one year since we started to getthese submissions and we have learned quite a bit.We're certainly not at the point wherever you're going move it into a required type of submission simplybecause the data is too complex and we need to make surethat we create the appropriate policies and guidelines that, but we are moving in that direction, that's nodoubt and I'm happy to share at any point what we havelearned and what we're doing with that information asyou deem it appropriate. Maybe that would be useful to hear about next time sothat -- I mean, maybe drugs and development, reallythere's a process in place that will move in the rightdirection for drugs and development through the FDA.I mean, we may be able to say, you know, move it alongfaster or get more resources if you need more resources whatever, but I think one of the major issues is theexisting drugs with the book that was shown by dick, youknow, it's not a small task for the existing -- existingdrugs.
Right.Yeah, I personally am still struggling with what do you have to do to get something in a drug label.So I'll probably keep asking you guys that question,because it's not really clear to me still.
It's not clear to me either and I think that's a veryimportant thing to be clarified.If death doesn't do it, I'm not sure what does.
Okay.Tim?
One quick addition.You might also want to talk about the personalizedmedicine coalition and get some of their perspective asit relates to
Okay.And now, one thing that was brought up to me during thebreak was that there apparently are differeringstandards for informed consent and what you're allowedto with banks, if you're a government agency versusif you're a private entity trying to do basicallyexactly the same research, but under a different hat.Is there someone we can get from the human protectiongroup that can clarify that for us what's going on, whythere's a double standard if there is a double standardand --
Can you clarify where did you hear this that there'sthis double standard?Did somebody say that today?
Who said that?
Yeah, okay.So you're volunteering to -- you want to come up and your comment to the committee and express yourconcern?
Yeah.Hi.The government agencies which are going to actually havea workshop on bio banks next week participate under adifferent set of regulations.45 CFR part 46.Industry has to operate under a different set 21 CFRparts 50 and 56.Where trusted third parties are used to hold the keys totrace back to source documents, that system is allowedin a government.What's in the industry is a part of FDA calledthe bio research monitoring group has said this is notallowed because they reserve the right to go back to thesource documents and without having to go through atrusted party.This has been an issue for quite some time and we thinksince we're trying to do public and private consortiumsworking together on pharmacogenomics we can't have twostandards.
Well, I thank you for clarifying that.I now understand what you're talking about.I thought you were talking about a different standardfor government agencies but what you're referring to isthe different regulation -- set of regulation thatgovernHHS funded research.And it's true that the common rule and FDA regulationsdo have a different approach to research involving humantissues and even the definition of the human subject isdifferent, the definition -- or the allowance for waiverof consent is different, and actually NIH, through itsprogram the clinical research policy analysis andcoordination program, an initiative of the NIH road mapis very interested in this problem and we've talked withFDA, the Joe hack it's colleagues in his center are, Ithink, develop -- certainly looking at this issue.And I don't know if Joe can speak to it any further, butI think there is a consciousness at FDA that -- thefact that they have a different approach is an issue andit's certainly a concern for NIH.And -- and if you're referring to the work sho*p thatNCI is sponsoring I'm sure that will be an issue.I know there's also a group primmer has a tissue workinggroup that's very concerned about this too and alsomaybe making some recommendations about it as well.
Thank you.
Certainly seems to me that if we're going to talkabout doing public private partnerships that we have tobe able to operate under one set of ground rules whereall agencies are accepting of a set of ground rules thatworks for everyone.So I -- I would like to see us talk about that a littlebit more and see if in our role as an advisor to thesecretary there's not anything that can be done tomediate normalization of things between agencies withinHHS.Other comments and concerns?Kevin?
Just one other thing.To bring this up and we can talk about it again in thetask force, but something that kept coming up andsomewhat tangentially during the various presentationsis this idea of benefit and the therapeutic things thatare going to be done, the clinical usefulness, that sortof stuff and at the end one of the reasons I asked thequestion of the ethics presentation and her answer wasyou know, you've got to get good language, of the thing we face today even in se phase Iclinical trials where you have wonderful informedconsent forms and yet the patients still walk awaycertain that this is gling to benefit them in sometherapeutic way in spite of the fact that it's a phase IIt's called therapeutic misconception.My fear is there's going to be huge therapeutic surrounding this sort of technology andit's going to be difficult to get a really goodunderstanding out in the public and some people who arevery good at that sort of thing are some of thesociologists who have been starting to study this thingabout risk awareness and different ways of conceptchulizing risk and all that sort of thing so that mightbe an area we'd want to look at. So you're talking about the public perceptions?
It's a little more complicated than just public persipgss.Different groups have different -- under differentfilters, different ways they interpret the very samewords and data and material and how does one thenaddress that sort of situation?It's something that I'm -- you know, the geneticcounselors see this all the time when they come in butit's something a lot of sociologists have begun to lookat in a more systematic way.
Okay.Agnes?
This comment relates not so much to a gap butsomething to keep in mind that as we're going to be together a document together or resolutionstogether that we include a section about the educationfor health professionals in this area.That was brought it many, many times for physicians,pharmacologists, nurses, other health care providers so think it would be something the task force has to makenote of.
Yeah, I actually made note of that in a largercontext, because I think we heard from several peoplethat education is not sufficient to create clinicalimplementation and -- and I would like to really explore going on with the clinical implementation pieceboth for things that already exist, what is reallyhappening that's keeping that from working as well as isthere some mechanism that we could propose going forwardfor sort of best practices?Here's, you know, when you get through the point whereyou have all the evidence, how do you turn evidence intoimplementation for better health care?And what are the -- the things, you know, sort of thesteps you have to go through on that implementation sideso I think most of the work that's been done to date hasfocused how you get to the evidence.And -- and we've seen a couple of examples even withevidence we're not seeing full uptake.I think eric's little chart where he compared her 2 andher accept tun testing with 2D 6 testing all of whichare quote valid bio markers where we know what they meanand we're still seeing this variation in upstake and weneed to understand that a little better.
Several points.Two quick ones and then a question, I think, for Tim.We heard several times also today about gene patents and impact that this is going to have on restricting thedevelopment of, you know, broader pharmacogenetic and I know we're dealing with gene patentsseparately but maybe we can hear this as we hear thereport of the NAS task force that's going to have areport coming out this July, that hopefully we will getbefore our next meeting.One point --
Can I just say something on that?Sarah whoever's going to be organizing this, since going to have some kind of report on that report II sum at that meeting can we ask whoever's doing that totalk about it both in the general as well as in thepharmacogeneticsGo ahead with your other point.
That's okay.The second point.One statement kind of struck me which is that whenthere's FDA approval then CMS should pay.And we just finished a coverage and reimbursementdocument and I don't know that that's in there anywhere,but it did seem like a logical connection between the I don't know whether it and Idon't know at this point -- don't worry, we're not goingto go changing the coverage and reimbursement document it was something to think about, I think, in thecontext of coverage and reimbursement andpharmacogenetics.And my third question is really in the model of the NCIcancer, they're not core facilities but they'rebasically resource facilities that are set up to helpwith certain types of cancer anal cease that are doneacross many different kinds of research.What would it take to have the same sort of research todevelop pharmacogenetic analysis from patients ofclinical trials in a more centralized way and it couldcome out of the pharmacogenetics research network and infact, dick said that they had applied for this and itwasn't funded but it seems like that would be somethingsince they already have data analysis and -- andstatistical analysis and many resources within thatnetwork, if there could be this -- a type oflaboratory created and I don't know what mechanismswould be needed, but could you speak a little bit tothat, Tim?
not sure I can speak very specifically to that.I mean, I -- we provide a lot of sort of the basicresources for genomics resource and that we dointramurally in our institute as well as just -- thepower of the con vie nor on these kinds of things andhaving workshops to try to provide the basic kind ofinformation for people so they can better understandthese types of things but I think it would require aproposal of someone to present to our institute as tohow we would, you know, they think we should proposeproviding those resources.I think it's something we would definitely consider but don't think I know the best mechanism at this point.
Just to clarify, there are such cores that are outthere.NHLBI supports major sequencing cores which werementioned in ro shell's talk where people can submitprojects for gene resequencing and pharmacogeneticscertainly falls under that.To me gene typing is dirt cheap and can be done in athousand plus cores and facilities around the country.So don't think it's an access to technology that'sholding up any of these studies it's a concept chulblock to pulling together the large studies at thetranslational end but the -- getting the data out oflabs I don't think is a major roadblock.
Sandra?
I disagree.
Oh, I'm sorry.
Yeah, on the point that you had made earlier, I thinkyou might want to hear from CMS themselves about theeffect of FDA approval on their reimbursement policies.As you know, they have responsibility for the elderlyand disabled population and there's recently been a drug added.You might want to hear from them about how thesetechnologies may then impact their responsibilitiestoward these populations and also the responsibilitiesfor -- in the area of cost containment because they dohave some responsibilities in that area and you know,they don't address the totality of the population, but Iknow that insurers and other -- the peers in generalkind of look to them to see what decisions they've madeabout that and the populations that they address butthey also have the other program medicaid in partnershipwith the states.They may not -- they don't make coverage determinationsthe same way, but certainly these technologies are goingto impact upon those populations and so you might wantto hear from them as well on that.
Okay.Deb, did you have a followup to your previous comment orsomething new?
Well, I disagree.Because I think that a general sequencing facility orgene typing facility isn't going to have thepharmacogenetic and form *p pharmacologic -- to asthmadrugs or antidepressants or whatever. want to help with these or designing what typeof gene typing or resequencing you would choose to do I think many of these products may come out ofclinicians who may not have the statistical information,the bio information and so to have a more focusedpharmacogenetics type of core rather than the genericsequencing kind of core might facilitate this research.
Yes.Then we're -- we're disagreeing only on what to call itbecause to me then it ceases to be a core because ifyou're wanting to be driven intellectually and conceptchully by the core where physicians and cliniciansaround the country might be able to offer cohorts apatients and from that would derive from geneticconclusions and data.So to me that's different from a quote core so whateverwe call it then I might agree there's a need for sucha -- think a lot of the farm GKB labs actually had acomponent where they both collected clinical samplesthat were well characterized as well as had to provide amechanism for resequencing or gene typing needed to bedone on those.So I think within the individual awardEEs of thosegrants there is expertise and it's a mixed expertisewhere you've got clinicians and the high throughput genetyping and sequencing support team.
But in talking with dick afterwards, he was sayingthat he had made a proposal for this type of thing that integrate and would be ongoing so that you couldevaluate the specimens pharmacojet cli and use theresources within the pharmacogenetics network and thatwould not fund it.
I'm going to let JULIO talk.
thing is that what you're referring and Idon't want -- the network put together and was notfunded, it was part of a road map for translationalcenters and the whole RSA was cancelled.So not that it was not funded as a specificproject, the whole initiative kind of disappeared.But I actually just very recently a couple of weeks ago editorial about this because the point you'rebringing up is very important and maybe we aulth toconsider now or in future meetings, I think having in this for a while, if you look at it carefully,there are some people who do outstanding work on bothsides I'm not talking about those but where you seethe biggest deficiencies are the people who work on thegenetic side and have the more genetic background.They call the clinical materials they just call samples years back, I was asked to consultant in order to doa calibration with a company, and they, you know, askedme to calculate like a Jr. Karm cogenetics trial thatwould result in blood samples and they said oh, the costper sample is too high.Where as with genetics research I can go out there andget a thousand schizophrenic patients. just go to the large state hospitals and you can a thousand people in a day.But you cannot for pharmacogenetics you have to screenthe people and then treat them and observe the resultsof treatment in a controlled way, which is extremelyexpensive and the people who drew the genetics side theydon't understand the clinical issues they don'tunderstand the clinical issues and they don't accept thecost that's extremely high. you see this all the time.You see very sophisticated genetics daunt on very simplequestionable value.So from my own, from my own GKB study, I had to screen2,111 people because if you're study pharmacogenetics ofa drug ideally that person should have that disease andnothing else.So if you're studying pharmacogenetics of Andy hepressosants you don't want a diabetic who's alsodepressed because if they change you don't know what'schanging and out in the real world it's very rare to a person who has that disease, only that disease,nothing else and is really to take that one drug andnothing else and does not have back pain, does not have,you know, is not taking a ton of natural supplements,not taking this and that thing.So the geneticists, they fail on that side.And the clinicians, they fail on the side that some ofthem have a more clinical background they collect verygood samples and very good trials and they don't knowthe first thing about the genetics and that may be wherethis thing could be helpful and they ask for a fewpolymorphisms here and there.They do a lot of tests.So what I see often are clinical people coming from theclinical side, the pharmacologist side without theknowledge of genetics and people from the genetics the knowledge of the farm cottics so maybe, youknow, some type of interface between -- and thepharmacogenetics network is wonderful, but it isrelatively, you know circumscribed tho those people thatare in the network.But it doesn't really at this point and I know it's agoal for the future, doesn't reach to the clinician outthere or to the clinical researcher out there and a lotof so it should be important pab from thispanel to try to bring those two communities togetherthrough a core facility, through, you know, some sort ofmechanism because that's where the divorce happens, youknow?
I think that's a really great idea and we'll see ifwe can try and figure out if we can make some sort oftask force recommendation.
One point on that and then another one to followingup on pat's topic.I think Dr. Davis this morning made, I think, a veryrational and impassioned plea to figure out how to dotranslational pharmacogenomics that is linked somehow tohealth outcome.That is as he points out that's a vrp different kind ofscience than the people who are trying to do the basicscience in the laboratory and it may be that thesenetworks which are valuable for one area of sciencedon't necessarily completely bridge that gap and thetask force may want to look more closely at themechanisms that specifically lead to addressingnot the basic science but assume the basic science isthere, how do you then take those discoveries and thatknowledge base and push that through with a series ofthings in the clinical ano*al sis but the and financing et cetera because there'sa whole number of different avenues that would need to into play in order for there to be success inadoption of this orny oh technological advance in thepractice of medicine.The other two things that I jotted down that the task might want to look at which I'm not sure we orother groups have taken up at least fully, one was theissue of genetic exceptionalism again and this we dealtwith years ago, I believe, but it comes back upspecifically in this context that I think is veryrelevant as she presented to the issue ofpharmacogenomics.Is it something, I mean, is this really a truly new that is going to have to figure out a wayto deal with it or is there a way to slip this intoexisting paradigms, regulatory or other woo*isz and thatseems to me as a reasonable task force question.The other is racing genomics.There may be other examples as well.There certainly been other examples coming down the pikeand to address that from the standpoint of are theregaps in knowledge and what would the secretary need toknow about those issues where we might be able to be ofsome help.
Do you think it would be useful to hear a shortsynopsis of what actually happened today?Clfr way it goes?
That probably depends on what actually happenedtoday.
Well, I mean, whether it was approved or notapproved, is there something -- a lesson to be learnedthere that, you know, I mean, as a potential topic for October update?
Let the task force do what the task force will do.I think it depenteds on what happened today and whatother kinds of examples that may need to come along. sure there will be plenty of opinions on what theydid.
I just wanted to rejoin the discussion that Debra andsome others were having. to sort of state the obvious, the example ofpharmacogenomics in this area of interdisciplinaryresearch is a very edifying one but far from a uniqueone.It really crystallizes I think the challenge to the NIHbut not just NIH but to academia to private new industryet cetera to think about how we do research and whennobody has the degree of knowledge in enough areas to beable to do the research anymore, and I think the GKBnetwork is a wonderful example of how to move in thatarea.And certainly NIH continues to deal with this, realizesit's a very fluid area, needs to come up with new modelsfor doing it but it's not just the funders that need todo it.It's all the funders.It's not just the funders that also challenges academicinstitutions and many are obviously trying to do this,how you come up with ways of trying to do this and it'sa farther challenge but perhaps an opportunity wherethis gets to translation and research where there areprivate industries that are interested in the knowledgegained here and how one creates interfaces with privateindustry as well as obvious interest in this kind ofinformation.You know, there are no, I think, easy answers to this,but everyone, I think involved represent -- recognizesthe fact that they don't have the answer as of yet soany advice I think the committee could offer as tostructures, I would about just look at the funders, Iwould look at them but I'd look at other ways in whichwe approach these things.
So I think part of our focus on funders might have todo with our charge to deal with HHS and not stray toomuch from our man date to be a sgrup that makesrecommendations to the secretary but we certainly couldtalk about how HHS agencies could do outreach and joinwith nonHHS agencies whether they're public or privateto move forward. commentary?I think the task force has plenty of meat.We'll do to put together a program that'sorganized.Sarah has some comments.
Actually it's more of a question.Does want to talk or give any furtherguidance to the task force about the long range goalhere?It sounds like you're not ready to begin writing anykind of report. still exploring and needing to put together anadditional presentations and fact finding for theOctober meeting, but not ready to think about theproduct that will come out of all of this yet.
Yeah, well I'm hoping we'll come out with somerecommendations but I'm not sure if we'll come out witha big book like coverage and reimbursement that withinit has imbedded recommendations or whether the workproduct will be more like our letters to the secretaryon education and discrimination that just points outsome specific things.I think this subject is so complex in many ways that youmay have to have some white paper at least thatbrightens the issue and talks about the specificrecommendations.
Well, would the committee like to give the task force latitude to think about what -- what form the -- Iguess that's a -- inherent in this, but think it wouldbe good for the task force to think about that early onand --
Yeah.So I mean, is there anybody that has any objection tosort of an open thought process at this point for how wemight convey whatever recommendations?Okay.Good.I'm seeing everybody in agreement that we can have somelatitude.
When you mentioned the white paper, one of thespeakers and I can't remember which one, had mentionedthat there were four white papers that were published inthis area.
Ro shell long
It would be very helpful if it was made availablewhen they come as they come.I want to thank everybody who participated in thesession from the speaking side and -- and all the peopleon the task force who participated in getting us thisfar, in familiar faith who did all the work oforganizing everybody to actually here and put theprogram together.I certainly appreciate having everybody's help and the shaker saying of many hands make lightwork and really does make a difference to have a lotof people participating.We thank all of you for your participation and lookforward to additional input and discussion.Did you have one more thing for the task force before weclose it
Actually not.We can talk about this later.I was more responding to Debra and the translationalresearch centers RSA or PA that was cancelled they had ameeting a couple weeks ago to think about what to doinstead of that, I think, and so we could hear fromthem.That could be something else you might want to -- andmaybe the NIH road map in general might be somethingthat might be of use to hear from -- if only for thetask force or the full committee maybe.
AllI'm turning it back over to hunts for the last closingnext steps and closing remarks.
Thank you to Emily and the task force.That was a terrific all be it exhausting day and mythanks to the speakers as well.I think we never fail to learn something and todayactually learned an enormous amount and I thank you allfor that.It falls on me simply to announce our next meeting isOctober 19th and 20th.And at least currently is scheduled to be held hereagain according to my notes and meeting dates for nextyear are in your table folders for those who like toplan your long range calendars.I think it -- all of us wanted to both recognize andthank and say good-bye to Barbara and Joan.This being your last meeting.Ed totally forgotten he was ever on this committeeby now.His 12 hours have passed but you've been terrificparticipants and we will misyou, but wish you well inyour retirement.Any other business?
Are the meeting dates set for going out?I don't find that
were supposed to be but we're having to work outa -- we haven't found sites meetings yet sowe're holding off on setting them in stone yet.But we hope to do it very soon because we know yourcalendars will fill up soon.
Could you send out at least tentative dates that wecould hold or --
Can we do that?Yeah.We can certainly do that.
March, June and October.
Don't put anything on those months.
A theme that has been going through the hall work andcertainly few days is access and I'm bringingit separately from the pharmacogenomics because itreally is underlying everything we've been talkingabout.And I know a fair amount of studies have been fundedthrough the LC regarding access.What we don't know is if they have solid evidence toreport about that, but that's something I would like usto think about for our future meeting and have ourcolleagues do the homework to know whether they're at a to want to be presentsing that but I think that'scriticaledly an important underlying work that we'redoing and if the science is moving along in that area itwould be hoov us to know what the science is.
Thank you for that.We do need to keep coming back to it so I appreciatethat.Agnes?
Not that I want any more work but just the beautifulchart that we put up regarding the time line of all thepriority areas, that is there anything else that we haveto address besides the pharmacogenomics for the nextmeeting?
Large population studies is the other major one.So with that and seeing no other red lights, thank youto everyone, both on the committee and in the audienceand those who are still hanging in at home and withthat, this meeting will be adjourned. you