dr. alan guttmacher:i have to tell you, for a couple of weeks now i have really been quite proud of myselfbecause when we got together to talk about this workshop, francis said, "alan, why don'tyou take the last talk." and i thought, that's fantastic, what better than to have your bosssay he wants you to bat cleanup. i thought, you know, i i just looked at the agenda, andcounting the two times that francis had himself bat, which is a different issue, it turnsout i'm not the cleanup batter, i'm in the ninth position in the batting order. [laughter] so that's probably more correctly. and i hopeby the end of the talk maybe you won't realize
that. so what i'm really talking about ishow wonderful it is that we have these gwas studies, but, in fact, how do we translatethose into something to improve diagnostics, therapeutics and particularly, in some ways,prevention. so there are sort of two major points here.the first is the use of genetic information regarding common disease to individualizedproviders' approach to patients, and to also change patients' behaviors in ways that leadto improved health, so called "personalized medicine" in many ways.the second is the use of genetic information regarding common disease to understand thebiology of human disease to lead to improved diagnostic, therapeutic and preventive approaches.i'm going to spend most of my time talking
about that second one, because the whole personalizedfor several reasons, including that personalized medicine is probably a three-day seminar ratherthan a 20 minute talk, but i will at times bring in a little bit of this. now, of course, we spent the whole day talkingabout gwas and most of us who spent time in genetics, either practicing medical geneticsor trying to write about it have spent time thinking about single gene disorders and nowwhat is this about genetics and common disease. and it's really been the theme for the day.we're moving from thinking about the larger effects of single genes and the rarer singlegene, so called single gene diseases, and i put those in parentheses because there'sprobably no disease that only one gene really
has, has an effect on. to the smaller effectsof multiple genes and common, complex diseases. now, one way to sort of, just sort of makea quick, you know, distinction in your mind between these two kinds of situations i thoughtwas to talk about a typical single gene disorder, and several have already been mentioned today,but since, you know, i come from the nih i thought i should be erudite and quote theliterature, reference the literature for you so i have referenced my favorite i hate topick favorites here but i have referenced my favorite journal here. and those of youwho buy your own groceries undoubtedly, at least, read this journal. i'd like if anyof you write for this journal sometimes i'd love to meet you. this is from the weeklyworld news. i take particular pride in this
because i was the only medical geneticistin vermont at the time that this appeared, "creature captured live in vermont, bat withthe human face. â€˜he's smart as a whip,â€™ says stunned scientists." i should note, iwas the only medical geneticist and though i'm often stunned, i'm not the stunned scientistwho was quoted here. though it would be great to have this on my cv if it had of been me.but this was the kind of the old model of a single gene disorder. some strange conglomerationof physical findings, et cetera, that led to a diagnosis usually carrying some tripleeponymic [spelled phonetically] name, you know, with three people in the world whoâ€™vehad it. now, obviously, these single gene disorders,they're much more common than three in the
world and some of them are quite importantfor health, but in general, single gene disorders are rare enough that they rarely approacha kind of public health concern. and that's why when we start thinking about the impactof gwas studies on health, there really is a somewhat different impact than even thetotal impact of single gene disorders, because when we start talking about gwas, we're reallytalking about all these disorders. you will notice, i think, the first page ofthe handout from my talk starts with a variant of this slide. i tried to put my time to gooduse during the break this morning and actually updated the slide that you have from 2002to 2004. you'll notice sort of interestingly, the only thing that changes between thosetwo years is that diabetes and alzheimer's
have flipped places in terms of excuse me,alzheimer's and pneumonia and influenza have flipped places in terms of the order of things,both as alzheimer's becomes a more common cause of death and pneumonia and influenzabecome slightly less common. but essentially, if you do it any of the last number of yearsyou'd have this same list of ten disorders. what they all have in common, of course, isthat by the very definition, of course, they're common disorders, and that in terms of whatwe have known about the genetics, all we've really known until quite recently is, gee,if you have a family history of one of these, you should be concerned you have an increasedrisk of developing this and perhaps dying from it, et cetera. but we have not been able,prior to the gwas kind of era, to identify
for most of these, cancer being the one exceptionwhere we really have been able to identify numbers, specific genes involved. becauseof the very nature of all cancers being genetic and not all hereditary, of course, it's madeit easier to find genes involving cancers. the one that's listed that gets a questionmark is injury. many people would argue, gee, injury, that is by its nature an accidentas it's often called incorrectly, and so genes must not play a role there. of course, there is a genetic variant thatwe have known for some time. you don't need a gwas study to find this genetic variantthat makes one much more likely to die of injury than other individuals. and as an experiencedclinician, i can tell you, just looking around
the room, that i can spot a number of peoplethat have this genetic variant, so without appropriate form consent i'm just going totell you a number of you that you are more likely to die of injury than others in thisroom and that genetic variant, of course, is the y chromosome. if you happen to havethe copy of the y chromosome you're much more likely to die of injury than if you lack it.but that may not be truly biologic, which is why that gets a question mark. but it'san important thing to realize, and one of the real reasons for showing this slide isregardless of whether genes play a role in the likelihood of your developing an injury,though, in fact, they probably do, they clearly do play a role in what happens once you getthe injury. and that's not just peculiar to
injury, it's also true of all the disordersin this slide. we often think about genes in causing disease, but clearly, particularlywhen we're talking about or specifically when we're talking about common diseases, geneticvariants have much to do with how the individual host handles the disease as well as the therapieswe use for the disease once they get it. so if you look at several different people withthe same injury, who ends up developing acute respiratory stress syndrome in the icu andwho doesn't. well, lots of factors can contribute to that, but we're beginning to be able toidentify genes, in fact, that play a role in those kinds of things. so if, you know, we're in the era of genomicmedicine and using the whole genome to understand
health, gee, can't we just have a good gwasstudy today and have a new drug tomorrow? after all, you know, last week genes involvingdiabetes, how come there's not a new drug for you to write about now, it's been severaldays, after all. well, the problem is that there are all these steps. and basically whati'd like to do is walk you through this diagram, i'll go through it some of it in a littlemore detail than others, but basically, showing the flow from gwas down to both, well, prevention,diagnostics and therapeutics, all three of them, just sort of talk with you a bit abouteach of the steps involved here and why there are both problems here. but while there areproblems, there are reasons to have optimism about these various pathways and that we actuallywill be able to get through them.
so the first one, replicate and validate.well, people have already talked about that, several speakers today so i really don't needto go into that one anymore. the next step i'm going to try to keep highlightingthese in yellow, identify genes and gene products. once you've replicated and validated the gwasstudy, you clearly need to not just, as we have talked about again, not just have theregion in mind, you've got to identify the genes or gene products that are responsiblefor this for the ideology of the disorder, to be able to really get anywhere. a goodexample of that is the story that many of you will know, now a couple of years old,using a gwas approach, both originally in this study and then in a couple of othersafterwards, which among them have been able
to identify three different genes which accountfor approximately 74 percent of the attributable risk of that age related macular degeneration,which depending upon how you define things, as either the leading or the second causeof significant vision loss in the elderly in the u.s. and that's interesting, becausebefore that study, of course, no one would have referred or thought of amd particularlyas a genetic disorder. it was a disorder that was pretty common. we thought, gee, if youhad family members that had it you were at somewhat increased risk for it, but it wasn'ta dramatically increased risk. now, once you have identified the gene orgene product, we'll go down the center aisle here talking about developing a diagnostictest. and, of course, the test methodology
is going to vary incredibly depending uponexactly what the gene or protein product, perhaps, you're looking at. so, you know,you can use your imagination, you can run wild in terms of the various technologiesand methodologies one might do to do that. now, it's important to realize that howeverwhatever the technology is, methodology you're going to use to develop the diagnostic test,that's only going to happen if someone financially supports development of the test. sometimesthat comes from the nih through research dollars. sometimes it comes from private industry,particularly; of course, private sector support tends to depend upon the, at least the perceptionthat there's a market for the test. but this is a real issue, how you go firstdeveloping the test and then moving it from
the research to the clinical arena, and particularlythat move from the research to clinical arena because even if you have a well developedmethodology, et cetera, et cetera, there are very few clinical labs that perform testsjust because they can. they tend to perform tests because they're going to get paymentfor doing so. so there are all kinds of issues about health, economics and other kinds ofthings mixed up in here. so there are multiple steps really to developing the diagnostictest and it's not just a question of technology or scientific approach. in terms of once one has the test, you actuallyneed to show that the test i'll go through all these terms with you. that the test reallyis valid, first of all, the analytic validity
that the test does what you really expectit to do. but even if you have tests with good analytic validity that does not necessarilymean it's of any use in healthcare. so you need to show both the clinical validity, andin some ways, most difficultly, i suppose, but perhaps most importantly, the clinicalutility of the test. that's not to say that there aren't lots of genetic tests out there,and more broadly, lots of other medical tests that aren't genetic at all for which clinicalutility has never been demonstrated and which, in fact, if you did a vigorous study you probablywould not be able to demonstrate true clinical utility. they are used and sometimes they'reuseful and sometimes they're not. it's interesting as we get to sort of genetic testing; thisis a standard that in general is being utilized
to think about new genetic tests, et cetera.we haven't very much retrospectively demanded all three of these for existing medical teststhat have nothing to do with genetics itself, but it's a good sort of rubric, i think, tothink about as you think about testing for not just genetic disease but any disease,whether it be gene based or not. and then, of course, even if you go throughall those steps you really have to obtain third-party payer coverage for the test. becauseif you can't cover the lab cost, then not very many healthcare providers are going toorder the test. and even if a healthcare provider might suggest to a patient, very few patients,though some would, very few patients are going to utilize the test unless their third-partypayer covers the cost of the test. another
part of this is actually covering the healthcareprovider time involved in the testing, which can be significant, particularly with newergenetic tests that haven't been used routinely in medicine, take a lot of explanation offor discussion between the healthcare provider and the patients and, of course, we have ahealthcare system that tends not to reward providers for time spent simply talking toa patient. so there are a number of financial hurdles here. sometimes you can get aroundthe need to cover health professional time, specifically in terms of a test but you certainlyneed to be able to pay the lab in some manner if you're going to have the test reach wideuse. so going down the other arm here that hasto do with developing therapeutics, you really
need to define the function of the gene orthe gene product if you're going to get very far with that. so here nick wade, i don'tthink is here today, but here's his story from last week about the diabetes genes findings.and part of the story says, "the importance of the new genes is that they point to previouslyunknown pathways involved in diabetes. several of the new variant genes make the pancreaticbeta cells produce less insulin, dr. altshuler said. that suggests that diabetes may startas a disease of too little insulin production, even though patients turn up in the doctor'soffice making too much insulin to which their tissues have become resistant." well, of course, speculating upon the functionof a gene in the "new york times" does not
necessarily make it true, though lots of readersprobably think so, and i'm sure you're aware of the power of your pens when you write thesethings. but clearly, that's an important point, that already david altshuler and others involvedin this are trying to figure out what is the function of the genes implicated because it'sby understanding the function that you're really going to get somewhere. so even if you understand the function, whatdo you need to do next? well, based upon your understanding of the function, you need toidentify a drug target. and this gets a lot of attention in the biotech press, amongstother places. that's an example of it. of course, you can use more or less conventionalstrategies once you have found a new drug
target to develop a new drug. again, you needto have the idea that there's going to be a market there. most drug development is donein the private sector and so there needs to be a feeling among some part of the privatesector anyway that there is a market for it. one thing to talk about, which i'm just goingto sort of throw this in as a parenthetical, interesting, and i think important fact. wetalk about whether it be genome wide association, some other things are happening in genomics.it's important to stress that a lot of the new drugs we're talking about aren't simplyas old; new drugs tend to be tweaking something so, you know, you add a hydroxyl group hereor you take away a methyl group to get better, you know, coverage or better transmissionacross the blood brain barriers. we're talking
about completely new categories of drugs fora number of these diseases. the interesting, i think, to think about isthat if we say that we humans have some place roughly around 20,000 genes, well, maybe notall of them are going to turn out to be good drug targets. they may not all be "drugable."if one guesses, and it's still a guess at this point, that perhaps half the human genomepresents drugable targets, the genes or the proteins they produce, probably about 10,000genes. currently, if we look at all the drugs in the pharmacopoeia, they target about 500genes in their products. so that would argue that 95 percent of potential sort of drugspace is unoccupied at present. so that by understanding, again, the way that genes workin disease it may be to understand mechanisms
of disease in ways differently than we havebefore. so that for instance, if we go back to the story about adult onset macular degeneration,the genes that have been shown to have this large role in attributable risk, are involvedin inflammation, it appears. so that has led to very interesting thoughts about, gee, maybewe should try using drugs that attack the inflammatory pathway. now, they could be olddrugs that we have had to do with inflammation or one would hope eventually of understandingthe precise mechanisms here, perhaps some newly designed drugs to be very specific interms of their effects. but a lot of this, again, is understanding the basic biologythrough our knowledge of genetics. so how would you design a candidate drug onceyou have identified a drug target? well, there
are lots of ways to do them. one of the thingsin terms of a genomics approach to particularly try to occupy this other 95 percent of thedrug space out there, is the idea, the use of so called chemical genomics. the nih hasa chemical genomic center, which is shown there. it's part of a larger initiative, themolecular libraries network, which has been started by the nih which has multiple componentsto it, including lots of external labs, ways of something called pub cam where the resultsare shared, et cetera, et cetera. the idea is to use high throughput screens to understandboth the action of genes that are newly discovered but also very importantly, to make a hugedifference in terms of developing some new candidate drugs.
now, of course, even if you do that, you'vegot to do clinical trials to show that the drug is safe and that's its advocatious [spelledphonetically] and you've got to get an fda approval. so this is in your handout, it'sjust a diagram, courtesy of chris austin, who runs the nih chemical genomic center showingbasically with some approximate time horizons for you there. the stages of drug developmentwith basic biomedical research, which has a very indefinite timeframe to it. and thenthis molecular libraries initiative that i mentioned goes a little further than thatand there are other efforts in the public sector to do that but this is probably thelargest one to move a little closer, a little farther down the field in terms of achievingthe ultimate goal of actually having a new
drug. but there's still many steps beyondthat. so if you're talking about gwas leading to a completely new drug, then you have multiplesteps that need to be taken and it will take some years. and it's important for both you,obviously, and your readers to understand that. now, there may be some shortcuts. for instance,if we go back to the amd story, if it turns out the drugs are already out there, widelyprescribed, perhaps, as anti-inflammatories, if someone does a study to show they're helpfulin amd then you skip all this other steps of drug development. you can use an old drugthat's already been clinically approved by the fda, et cetera, et cetera. you just needto expand the indications and the labeling
from the fda, which is a lot easier to dothan to go through all these steps and much more likely to both shorter timeframe andmuch more likely to lead to success. but if you're talking about starting de novo, really,a completely new drug, then you can shortcut some of these steps but it's still a matterof some years until you're going to see a completely new drug. so looking at our other pathway, the preventionpathway, you can see here you might well come from diagnostic tests. once you have a diagnostictest you could offer that to lots of people and think about prevention strategies. sothat, for instance, at an onset of macular degeneration you could think about developinga diagnostic test based upon these genes that
we have shown or implicate in the diseaseetiology and you could, you could diagnose people in the population as being at a significantlyincreased risk for this. and then, for instance, you could simply work counseling with thosefolks about smoking behavior, for instance. it turns out that the attributable risk ifyou add smoking in much increases your risk if you have one of these at risk alleles.so those kinds of prevention strategies based upon the diagnostic test are one way to go.but in fact, for some of these you don't even have to develop the diagnostic test. you cansimply, if you can replicate and validate the gwas study, et cetera, et cetera, youcan occasionally go without the test but simply be able to devise prevention and non-drugstrategies that would be of use even without
a diagnostic test leading into it. now, this is an interesting article from jama,not very long ago, which, again, talks about amd. i'm going to concentrate on the partin yellow but i think the part in white is actually interesting just in terms of well-donescience, in terms of the folks writing the article saying wait a minute here. "we believeit is premature at this time to consider genotyping individuals with various stages of amd. screeningshould consider, one, the genotyping of about 30 individuals with drusen/pigment changes",that's some sort of sign of amd, "would be required to identify one individual who washomozygous for the risk allele for both genes. and two, the observation that many but notall individuals of those genotypes will develop
the disease." so they're saying, wait a minute,before you develop some kind of test and try to market it broadly here, let's look a littlebit basically, and they're not using the terms, they're talking about let's look at the clinicalvalidity and clinical utility before we just start doing this test because we can. "however,in the future, a risk profile that includes genetic and environmental factors, such asthe one calculated herein, may ultimately lead to targeted screening and closer monitoringof individuals who are at higher risk of visual loss due to amd progression." and that's clearlysomething that many people in the public health community are beginning to think about, thatonce we show genes that are implicated in disease causation, to think about findingindividuals who have variants that make them
more likely to develop disease and developtargeted preventive strategies to try to help those, those individuals. which, of course,are all of us for one disease or another, or for many of us for multiple diseases. but even if you have done that, you have tovalidate these things via outcome studies. and the idea there is that, you know, youmight have a test that looks good, et cetera, et cetera, but unless we really do rigorousoutcome studies to show that making the behavior changes, et cetera, have a positive impacton eventual health status, it's sort of so what. now, again, lots of things in medicinewe have done over the years without showing that they really have a positive impact onhealth status, but the idea here is we shouldn't
just do this blindly. it's going to take sometime if we think about all the steps here. even if we had this prevention strategy inplace today, it would take some time to do a good study that shows by counseling youabout your individual risk, let's say, for amd and then counseling you about the particularlyinjurious effects of smoking for you. does that really change i mean, there are lotsof people today who smoke who already know that smoking is not such a great idea forthem. by adding in their personal risk for amd, is that really going to change theirbehavior or not? we don't know. and we don't know how you might do the counseling in orderto be able to achieve that. so those are the kinds of things that still clearly need tobe figured out.
but even if we do all that, would we reallymake an impact on health? so i think to do to make this work, it's going to require bothinformed interested providers and informed interested public, that is us. and we thankyou for your help in accomplishing both of those things actually. so there are many ways to talk about thisand one particular sort of resource, i guess, and tool to mention to you, if you're notalready aware of it, in terms of reaching healthcare providers or something which hasnow been around for over a decade called the national coalition for health professionaleducation in genetics. the name may be the only thing that's more unwieldy, really, thanthe acronym of nchpeg. but nchpeg is an umbrella
organization that includes literally dozens,scores actually, of health professional organizations, everything from hospital chaplains to physicianassistants to lots of nursing organizations, a lot of medical specialty groups, et cetera,that's in many ways sort of serving the clearinghouse and the catalyst for health professional educationin genetics, and particularly some of these newer things.and then, of course, informed interested public, there are multiple ways to try to achievethat. you know, we tend to look at you as perhaps the major way to do that. we knowthat you often do not feel that your job is to do our job for us, to educate the public.but we like to think, good days i guess, of it as a partnership that we can do that togetherand even if it's not the prime objective of
any of us, somehow we'll still get about doingit. a specific other tool for this that many ofyou will know about but some of you probably won't is something called the u.s. surgeongeneral's family history initiative which was started several years ago by then surgeongeneral carmona, continues even after his leaving that job, and it's a multiple agencyin the federal government effort and lots of private nonfederal partners at this pointto try to encourage use of family history in healthcare with the idea that way beforewe get to sophisticated genetic tests, et cetera, family history in some ways is thecheapest and most accessible genetic test that we've got. precise? not very. but cheapand easy to access, yes. so this is an attempt
to try to make family history more usefulin healthcare in various kinds of ways that i'd be happy to talk about. so again, just to make the point that whilemany of us of today's standard medical practice have never been proven to improve health,though it makes sense economically, we should use rigorous outcome and cost benefits todecide which genomic medicine practices to utilize. i'm sure we'll be using some as westudy them obviously, but that's what we seek to do to do this in a kind of rational, effectiveway. so i realized that i needed to come up withan executive summary. one of the things i have learned in my eight years within thecapital beltway is that nothing is worth knowing
unless you can reduce it to one page or less.something that i know you think you get a few column inches, we do, too, sometimes.and so the question is, how could you reduce this to one page or less? so i figured thequestion is easy, i'll just look at the title of my talk, "will gwas actually lead to prevention,diagnostic and therapeutics for common disease"? the wonderful thing about at the nih, lotsof very bright collegial folks are running around. can anybody give me a short answerto this? i don't know whether it's the short part that threw them, but for some reasoncouldn't answer that. so that, you know, i thought some more. gee, there must be somepeople around here and then i realized this is an executive summary. i am a member ofthe federal government so i have an executive,
i have a chief executive. so i thought, aha, executives must know aboutexecutive summaries, perhaps the chief executive would have a nice short answer to this. isee some of you look skeptical. i can't believe that. well, sure enough, president bush onapril the 10th, 2002, in a saturday morning address, actually, when he was saying somethingabout the need for genetic nondiscrimination legislation, he mentioned "our age may beknown to history as the age of genetic medicine, a time when many of the most feared illnesseswere overcome." well, i thought that's of interest, that's short, it's an executivesummary. i have unpublished data that suggests that the reason why executive summaries areso important within the capital beltway, because
of variants and genes having to do with attentiondeficit disorder, being much more common around here. but anyway, that will remain unaccomplished.i thought, that's wonderful. and then i realized, well, you know, this is a crowd who may notnecessarily believe that everything george bush says is true. i know you have no it'sjust that i'm a federal bureaucrat, you're a member of the press so you have no morepersonal, political beliefs than i do, clearly. so i thought that might not work. but still balance, i know that you love balancein your stories, right? you like to have both sides. ideally you like to have at least twosides but you like to have balanced coverage. so i thought, aha, when i first came to d.c.,there was another chief executive. he came
from the other party. what could be more balancedthan that? but what's the chance of two chief executives in a row talking about, you know,this kind of stuff? and then i remembered that, in fact, there was a ceremony at thewhite house in which francis was a lead participant at which then-president clinton announcedthat the draft human genome had been achieved, human genome sequence. and at that ceremony,sure enough, here's what clinton said. he said, "because of the draft sequence now beingavailable, it is now conceivable that our children's children will know the term canceronly as a constellation of stars." now, that shows a couple of things. one shows he hada heck of a good speech writer. now exactly. now, it also some people wouldsay, well, boy, that's gene hype. and there's
clearly been a lot of gene hype out there.and if truth be known, of course, will our children's children and cancer still existin a couple of generations? of course it will still exist. however, i think there's actuallysome gene hope there more than gene hype. and that is, if we look at the family of disordersthat we call cancer and we compare those to, say, infectious disease because many people,i think, more or less correctly say that our knowledge of genetics and genomics have thekind of impact on health that our knowledge of infectious disease did in the last century.so we compare the family of disorders that we call cancer with, say, tuberculosis. ithink we'll see a similar kind of evolution in this century for cancers as we did fortb in the last century. that is, cancer will
become as tb has become, no longer somethingthat if you did a rigorous three generation family history, you would find multiple individualsin the family who, in fact, had had it, several of whom have probably died from it. like tb, cancer would become something thatpeople no longer have a legitimate personal fear that, gee, i've got a good chance ofdeveloping that during my lifetime. like tuberculosis, for those relatively rarer individuals whodo develop cancer, it will become something which in many instances is curable and inothers becomes treatable so that it becomes a chronic disease rather than a cause of death.now, like tuberculosis, there will still be huge issues about healthcare access, socialeconomics status and other kinds of things
that even in the age of genomic medicine willstill have a huge impact on healthcare. even in the age of genomic medicine, the bad newsis we will all maybe it's the good news actually, we will all still die but presumably we willlive longer and healthier, but i do want to, you know, i want to both say that there'sa real, if not pot of gold at the end of this rainbow, there are some real positive impactsthat we can discern gwas studies and some of our work will lead to. but also at thesame time say, it will take some time to get there. there are multiple steps beginning,you know, it is a revolutionary era and all that, but there are logical steps that needto be completed before we truly see the health benefits. for some diseases, those steps willhappen very quickly. for others they will
be frustratingly slow, but it will be an interestingstory for all of us. so i'm going to stop there. male speaker:questions? anybody have any questions? use the microphone, please. sharon begley:sharon begley from "newsweek.â€ a question about the part of your slide where you showin the flowchart how some of this work might lead to a diagnostic. will it be importantfor any such diagnostic test, i guess including those that already exist, to achieve clinicalvalidity and utility for it to determine the, the expression status, the epigenetic statusof whatever gene allele you're looking at?
dr. alan guttmacher:it's not going to be important for every test to be able to do that, but there will be somefor which simply knowing the gene sequence isn't going to be enough. there will be somefor, you know, knowing that won't tell you much of anything. there are others that willtell you a lot, but not the whole story. to know but of course, often in medicine youdon't need to know the whole story to have a positive impact. ideally, you want to knowthe whole story and be able to take all of that into consideration. so there will besome situations i don't think we really know yet, but there will be some situation in whichepigenetic phenomena, like all kinds of other things, are very important. there will besome situations where genetic status alone
may not tell you much but it's going to bethe interaction of that specific genetic variant with some very specific environmental intermediary,that makes a difference in health. so that even knowing, you know, the genetic statusin some situations may not tell you much but if you can, you know, connect that to specific,as we learn these specifics, something about the specific genetic environmental interactions,it may be knowing that extra piece of information that going to be important. male speaker:bob langreth, "forbes.â€ while people, geneticists debate whether like that paper, whether it'spremature to have macular degeneration, other genes available as a test, isn't someone insilicon valley just going to put this stuff
up on the web, you'll get your own alumina300k app or whatever, $500 and you'll be able to find out all the things they want you know,all the things you want and whether or not it's "premature" or not? dr. alan guttmacher:absolutely. i look very much forward to your story illuminating that dynamic. and explaining to people how they might beable to discern what's, you know, sort of i mean, this is a wonderful you know, if weall wanted to go into snake oil sales together, this would be a wonderful 21st century opportunityto do it, there's no question. and on the other hand, if we wanted to do something that'sreally wonderful and a real benefit to humanity,
this is also a wonderful opportunity. so therewill be i think, there's clearly, because this is complicated stuff. as you all know,the question you know, the discussion we had during lunchtime, how do you explain thisto people? well, that's not just of interest to you all, it's very much of interest ina clinical setting, for instance, how you explain this to people. so it's a situationwhere, you know, people are kind of ripe for the picking, perhaps, for people who eitherinnocently, they may think they're doing something really worthwhile by bringing these teststo everybody, or perhaps not so innocently, just thinking, gee, there's a way to do this.and you can go on the web and find some pretty bizarre kinds of offerings. there's lot ofclaims in terms of nutrigenomics and is nutrigenomics
a real thing? yes, absolutely. is there somethingabout genomic makeup that interacts with our nutrition to make a difference in who we are?absolutely. do we understand that very well yet? absolutely not. but there are peoplewho are willing to do you know, take a swab and then, you know, based upon that do a geneticprofile, tell you exactly what kinds of diet you should be on, but particularly what kindof minerals, vitamins, other things you might need to take. and nicely enough, just to makea one stop shopping, they're willing to sell those to you actually, just to make life easierfor you, i'm sure. so there's it's a whole mixed bag out there. dr. alan guttmacher:i would be interested in asking you all who
write about this, as we can anticipate, ithink, during the course of this calendar year, the discovery and publication of dozensof gene variants that are clearly correct in terms of showing association with commondisease, at what point do you get to the threshold where at least some segment of the generalpopulation says, i want to know. i want to see my own report card. there are certainlypeople out there who take preventive medicine very seriously and are looking for whateverinformation they can find to try to individualize their own plans, at what point does our cautiousapproach of, well, you know, we really ought to test this out in a research study and makesure we know exactly what the quantitative risk is and prove that there's clinical utility,not just validity. and at what point do we
get run over by an avalanche of demand ofpeople who simply want to know now. and if we can't tell them that the information isbogus anymore, because it's actually turning out, i think, that we've got real data here,are we in a position to tell this [unintelligible]? what do you think? [unintelligible]. female speaker:i just think it depends on how much it would cost. because i mean, i'm thinking about,you know, the different genetic tests for cancers that are out there and people aregetting them. dr. alan guttmacher:and that's an interesting point. apparently the cost of, for instance, the brca1 and brca2tests, which remains in the thousands of dollars,
in part because that's exclusively licensedso there's no competition. there's no reason that testing for dozens of snps in a givendna sample should cost more than pennies. i mean, we indicated that the carton costof a snp is about an eighth of a cent so the actual, actual laboratory costs are goingto be quite small. the problem, of course, is there are going to be a big markup by thosewho are wanting to put this out there as a for profit enterprise but if there's somekind of competition in that, that could drive it down. so i don't think we should counton cost being a huge barrier for this kind of broader application. so what's going tohappen? female speaker:i think people also want to know, i wouldn't
have a test unless i knew there would be somethingi could do about the problem. i mean, if it's hopeless, why know? i mean, maybe you do yourwill or something but dr. alan guttmacher:and there are, you know, there are whole social science studies about the people who are informationseekers and the people who are not. you know, we've known for now, gosh, ten years, whatthe major common risk factor for alzheimer's disease is, whether you have the apoe4, spellingor not and yet there's been very little uptake of that information because there's nothingyou can do and it's a pretty horrible idea that you would have this cloud following youaround from then on after you have had your genetic test. a lot of the kinds of things,though, that come out in the next few years
or this year won't be quite that dark in theirimplications, but they won't either be connected with a certainty that you can interfere. ifyou're at higher risk for diabetes or for a heart attack, well, you know, we do thingsto try to reduce those like watching your diet and exercising. would that be useful?even if you couldn't prove right now that your particular genetic predisposition wasgoing to be helped by those factors. female speaker:i do think it could be added incentive. i mean, as you said, at most it may increaseyour risk 20 percent but, okay, as a chronic dieter that i am, you know, if i knew i hadalleles that, you know, would increase my risk of diabetes, hypertension, heart attack,that might, you know, be a good boost.
dr. alan guttmacher:so larry, you should tell them about this multiplex project which is just getting underwaywhich is attempting to assess this in the real world, by offering real people the chanceto have this information and see what they do with it. larry thompson:so alan simply outlined almost a decade's worth of research bringing these tests totheir final stages, and in collaboration with the social institute, colleen mcbride [spelledphonetically] and i decided we should start down this path. even though we picked geneslast year we're not going to deal with the flood that's coming out this year. and wewill be offering testing for five different
chronic diseases, heart disease, a coupleof different kinds of cancer, osteoporosis, diabetes to individuals that are covered bythe henry ford health system, so they're in a covered hmo-like organization. and thisis a really early stage project. we will be offering this multiplex test where they canfind out what their risk is for these complex diseases, some of which were found by oldgenome association studies. and our major outcomes and measures that are going to be,first of all, do they understand the test? because this is hard stuff to package. asyou guys are struggling how you write about this, this is hard stuff to package for thegeneral public. and whether or not they find the test useful, what's their attitudes aboutit?
and most of the early indications will be,and my behavioral biology people will hit me for talking about this, but this reallyis somewhat marketing issues, not testing the market, but see how people deal with this.ultimately we will be able to follow them up long term, although it's a relatively smallsample, to see whether or not people who have some of the worst combinations of mutationsactually lose weight or go on an exercise program. more immediately we'll be able to see, basedon the report card, and these folks are going to get a genetic report card. it's going tocome in the mail, a little folder that will tell them how they scored. we'll be able tofollow them and see what they avail themselves
of in the programs that are available as partof the health system. so there's wellness programs, smoking cessation programs. so ouroutcome will be, does your genotype make you more likely to go and enroll in one of theseprograms or at least inquire to one of these programs? male speaker:joe. male speaker:i just i've always been male speaker:this one is available [unintelligible]. male speaker:i mean, he's been using mine. male speaker:all right. all right.
male speaker:very articulate. male speaker:all right, all right, all right. no, the thing that i still have a little bit of troublewith in all of this is what makes genetics so special? i mean, you can listen to wtopradio and there's a place up in rockville that will test you for every known heart problemthere is, and some people are running up there to get it and some people aren't. i mean,is there something special about the kind of information or the increased risk factorsthat the genetic studies are going to give or is it just because it's less painful andyou can just do it with a needle stick or a buccal swab or something like that? that'sone thing. i mean, i have thought about this
all along. i mean, there are obviously familialimplications, but you're not talking about that now necessarily, you're talking aboutindividuals who are going to change their health behavior, right? dr. alan guttmacher:i can answer before francis does because his will be the short answer is that we don'tknow whether or not there will be something special about genetics. we think there mightbe. and part of it builds upon this foundation where, for better or for worse, we convincethe public genetics is very important and deterministic. now, we have to back off alittle bit and say, well, it's not as deterministic as we were telling you before. there are variantsthat occur that have little slight increases
in risk. but i think the real answer is, i'mnot sure we know whether genetics will be different or not. male speaker:and i'm going to also answer before francis, and say absolutely, we don't know but it'salso the breadth of the information. there's almost no piece of this that's unique, butthe breadth of the information as well, the ability once we have enough of this to beable to tell, you know, in their genetic report card. not just talk about one disease in termsof your risk but to talk about multiple, multiple diseases. now you could say, gee, that's likea whole body scan or something. well, most of the whole body scans aren't found by science.we would hope that this would be.
dr. francis collins:yeah, i guess the other thing is that it has the potential, at least, of making these predictionsbefore there is even the tiniest indication of actual pathology. you don't have to havean abnormal body scan outcome before the data may give you some information. so becausegenetic information is a permanent, it's there, it's there as soon as you want to look atit. whereas most of the other things we do to sort of make predictions about future risk,you already are having a high blood glucose or a high cholesterol or, you know, you'vegot degeneration of your spine or you've got a coronary artery that maybe isn't quite doingit's thing because you just had a thallium stress test. this has the potential of movingthe timetable back to an earlier point so
that you could begin to practice preventionbefore you're already sort of half in the grave. male speaker:but that sounds like, you know, the definition of a healthy person is someone who [unintelligible]. dr. francis collins:we used to say that on rounds, actually. yeah, well, it does have these semantic consequencesof exactly who's normal anymore, especially if we are all recognizing that we have thatsort of dna equivalent of original sin, we're all flawed. dr. alan guttmacher:but it's also based on the idea that one could
perform interventions to make a differencein terms of outcome. not simply saying, gee, you're not healthy anymore. but in fact, talkabout individualized prevention and other kinds. there may be treatment modalities thatwould stop you from becoming diseased. male speaker:you've got to be a little discouraged by looking at people who know absolutely for sure, positively,as you said, that smoking is probably not a good thing. dr. alan guttmacher:absolutely. male speaker:and so is dieting and so is eating more vegetables. and i mean, in a way you keep coming backto all the things that you already know that
your mother told you you should do anyway. female speaker:and also dr. alan guttmacher:but we do have to be a little discouraged but not totally discouraged because for onething, we don't know, and this is a big, huge question. if instead of hearing, gee, smokingin general is dangerous for you kind of thing, if you understood something about particularindividual susceptibility and also, if you understood about something, and smoking isprobably not going to be a good example, in fact, but something you could do that wouldmake a difference in terms of outcome, then, then it might be more valuable.
male speaker:[unintelligible] that you should find the allele that prevents you from getting sickfrom smoking. dr. alan guttmacher:that might tell you something about the biology for which smoking is a disease. male speaker:well, but i mean, i should see about r.j. reynolds as a partner instead of pfizer ornovartis. female speaker:and don't, don't you have to also maybe be a little concerned because you're talkingabout, you know, looking at alleles that increase your risk of heart disease 20 percent, andpeople who have a family history of heart
disease, they don't have the allele, thenthey may think they can get away with not living a healthy lifestyle. dr. alan guttmacher:sure. and part of this, again, i purposely spent much less time talking about the personalizedmedicine model than i did about the other stuff. and, again, i would argue that by understandingmore genes involved heart disease, we're going to repeat this story that we got from familialhypercholesterolemia. we'll learn other things about the basic biology that goes to heartattack. i mean, obviously, it can be multiple different biological pathways that all endup the common end, too often, of heart attack. but we will learn much more about the biologyof the disease that will come up with other
kinds of things that will make a difference.so it's not personalized medicine is absolutely an important part of this, but it's not thewhole story. female speaker:going back to the whole personalized medicine and report card thing. the thing that i thinki would find discouraging, and i have talked to some other people, is this idea that moreand more genes are now being implicated. so, i mean, there could turn out to be 100 genesor, you know, genomic regions that are associated with diabetes or obesity, in which case wewould all have one or several of them. then what do you do with it? dr. alan guttmacher:but it might sure. but it would be important,
again, not to know what your total risk isof that. it may be important to understand the variants that you have that confer increasedrisk, how do they translate from a variant to disease? and maybe there are particularthings in the steps that lead to diabetes, depending upon those risk alleles, that requirea different kind of thing. so maybe, you know, for one diet is going to make a difference.for another one diet has absolutely no difference whatsoever. but, in fact, for instance, let'ssay that, you know, your increased risk for diabetes is some receptor in some place orother is over-expressed, then maybe some new small molecular drug that was specificallydesigned to sit on that receptor and block it. maybe that's what you should be takinginstead of worrying so much about exactly
how many donuts you're having after lunch. male speaker:[unintelligible]. male speaker:yeah, just speaking personally as someone who confronts this as potential news, i thinkthere's sort of an analogous question to what consumers will make of this; and that is,what reporters will make of it. i was struck in one of these diabetes papers last week,this sentence, "while the eighth type 2 diabetes variants discussed in this report, each conveysa substantial population attributable risk, five to 27 percent at each locus. each contributesvery modestly to overall variants in diabetes risk, 0.04 to 0.5 percent, total of 2.3 combinedacross all eight snps." so, you know, i think
this is actually, you know, was an issue lastweek about what, what does one say about this? you know, it's obviously an incredible scientificaccomplishment, but what does it mean? and of course, there's always the big push tosort of, you know, speculate on what kind of very utilitarian outcome will be. but ithink that both the reporters and the consumers, what they're going to make of it really dependson how much magical thinking we're going to invest in, you know, single digit changesor expressions of our risk for, for diseases. and that's why i think this is going to bea hard story to sell even as this tsunami, you know, breaks over us. female speaker:maybe i could comment on that a bit. we didn't
have time really to talk about populationattributable risk and what those measures mean and they are, you know, fairly complicatedconcepts, but they do depend a fair amount on both the size of the risk associated andthe prevalence of the variant. and so if you have a variant that is not very common, you'regoing to have a low attributable risk. familial hypercholesterolemia, that variant is very,very rare and yet it opened the doorway to lipid lowering and we're hoping to find othergenes very much like that. so, you know, we shouldn't, you know, throw up our hands andsay, gosh, it's only .5 percent of the variants of diabetes in this population or other populationsbecause really what we're looking for is clues to pathophysiology.
dr. alan guttmacher:and i think part of the reaction that many of you are having is the idea that when you'rewriting a story for somebody, of course you realize that, as all of us do, that most peopleare narcissistic. i mean, you write the story. they want to know, gee, i've got diabetes,it's in my family. they just found these new genes, so what does it mean for me? and it'sa longer story to talk about it in terms of public health kind of sense and understandthe biology of disease is different from saying, you know, this is the particular personalizedmedicine kind of impact of it.