 Well, thank you very much. Great pleasure to be back here and have a chance to share some ideas about where we're going in the application of genomics into medical practice. I understand this is a CME offering, so here are my disclosures. Take the perspective, first of all, that a lot has happened, as you'll see, and certainly, as you know, in genomics in the past now, decade and a half, more or less. But one of the fundamental questions is this one, when will physicians actually routinely use genomics in their day-to-day practice? You may have seen this so-called hype curve. I pulled my version of it from Wikipedia, as you see. The idea that something impressive happens, such as the sequencing of the human genome, there's a lot of high expectations. Things don't turn out to be quite as easy as they seemed. I'd like to think we're on this slope of enlightenment and maybe even have approached this productivity plateau for some things. And what I'll try to show in this next hour or so are some examples of things that are happening now in the application of genetics and genomics in today-to-day practice, and then make some comments about where things may be going, realizing that this is such a fast-moving area, predictions are really very difficult to do. I'm going to use a kind of thought experiment as a way to go through this, and that is to imagine a family with this family tree. We're going to focus on Laura, and we're going to follow her from birth until maybe her 60s, and ask the question, where does genetics and genomics contribute to life decisions or medical decisions in this individual as she sort of passes through different epochs of her life? Now, there are a few things you'll just have to sort of accept as not particularly real world here. The first is I am not going to try to guess what genomics is going to look like in 60 years. So if the clock started today, I'm certainly not any better than anybody else at figuring out where it's all going that far into the future. So therefore, at every point in her life, we're going to be thinking about how genetics and genomics as it is applied more or less today, maybe plus a few years in a few cases, but fundamentally where it is now. Therefore, as she ages, in a way, the clock is frozen. The second point you'll see in a minute is that this is, I guess, more or less a sort of white middle class family, and it's not an attempt to make any statement about access of individuals of different backgrounds and socioeconomic groups to genetics and genomics. I realize that could be a significant challenge. It just didn't make sense for her to change her ethnicity every time she grew a few years. So you'll have to accept that. And by the way, all the pictures here are just stock photos. They're not real patients. So we're going to consider these different epics. Look at first newborn screening soon after she's born, diagnostic testing, in this case for a sibling, preconceptional screening, what happens when she's planning to start a family, prenatal diagnosis, pre-symptomatic testing, and finally predispositional testing. So the story begins just after she's born. She has some blood drawn from a heel stick, which is true for all babies born in the U.S. and throughout much of the world. And her parents really never are the wiser, actually. They may question why the band-aid is there. But beyond that, this is just done, and they never hear again about it. And that's a very good example of no news is good news, because if all of the screening tests are negative, then that's typically the last the family will hear. So newborn screening now is just over 50 years old, and it dates to the ability to detect the levels of phenylalanine in the blood with a very simple and expensive assay. And the reason why was that if you detected PKU at birth and instituted a diet restricted in phenylalanine, it made the difference between a child who would have severe intellectual disability and epilepsy and having a fundamentally normal developmental future. And so this was a huge public health advance, probably the most significant one of the last century in terms of genetics and genomics as it applies to day-to-day practice. It has evolved quite a bit from the bacterial inhibition assays that got this off the ground in the 1960s now to tandem mass spectrometry. Don't need to go into the technical details. The blood is drawn on the same filter of cards as it always has been. It goes into the mass spec at one end, and you read a spectrum out the other, which in principle can diagnose dozens of different inborn errors of metabolism more than you actually know the natural history of. And there's been a very productive effort actually led by the American College of Medical Genetics and Genomics to standardize using an evidence-based approach what are the specific analytes that diagnose things for which you have clear understanding of natural history and interventions that are proven to work. Since it doesn't make sense to screen for things where either they don't have major medical significance or there's nothing you can do about them. Although it's determined on a state-by-state basis exactly what to screen for, there is now a kind of baseline of screening that in the U.S. at least pretty much all the states do. Although some will add some things on top of that baseline. Well as you probably know, actually NHGRI is funding a set of pilot projects to look at the question should sequencing be the basis of newborn screening instead of the analyte testing. And there are a few others like testing for deafness. Obviously if one did sequencing in principle any genetic disorder could be detected so it vastly increases the scope of what's possible to test for. However, it comes at a price, never mind the cost of doing it, but also all the various things you may find that are not so clearly medically actionable and about which perhaps the natural history is not known. And I think that's really the crux of the currently funded projects to try to explore what the utility of this approach might be, the acceptance of it and compare that to the standard which has been in place for a very long time. All right, so let's move on into diagnosis. Laura is now three and she has an older brother who has been diagnosed as having autism spectrum disorder. One of the biggest maybe arguably one of the first advances in genomics that saw the light of day in the clinic was the ability to do cytogenomics as it's now called microarray testing. I guess some would say that genomics really from a medical point of view had its birth in the 1950s when it became possible to look at the full set of chromosomes and detect extra or potentially missing material. But it had to be fairly, well not fairly, it had to be very dramatic to see it back in those days. A whole extra chromosome as in trisomy 21 was detectable. But much more subtle changes back in those days were not. I guess it was sort of like taking a picture of the earth from about the level of the moon and trying to detect subtle differences. Well you could see volcanoes maybe but you're not going to see things happening on a kind of street level. And that resolution has gradually been increased to the point now where copy number changes down to a few thousand bases can be detected. And this has increased the sensitivity of genetic detection of changes that are related to intellectual disability, autism spectrum disorder and developmental problems. Probably in the range of what it was was perhaps 5% with cytogenetics to 15% or more with the genomic microarray testing. Meanwhile any certainly medical geneticist is familiar with this so-called diagnostic odyssey. Wherein an individual has a clinical problem, you make your best guess as to what might be going on and then perhaps order a genetic test that reflects your theory about what the etiology is. Get sent various places whether to academic or commercial lab. It could cost up to a thousand or even several thousand dollars depending on what the test is. And often it turns out that's not the answer and people cycle around this sort of loop sometimes for years causing huge frustration for the family, for the physician or physicians oftentimes who are involved in racking up bills that could be in the thousands of dollars realizing each turn of the cycle involves lots of medical tests. So the ability now to do sequencing, exome sequencing or increasingly whole genome sequencing has really changed the landscape here quite dramatically. Here's an example of a child seen in our clinic where genome sequencing produced an answer and in some ways kind of a maybe unexpected set of answers. It was a child with severe seizures that began soon after birth, intellectual and developmental disability and a MRI which is shown on the left with significant cerebellar and also brainstem atrophy. He had gone through multiple rounds of standard testing by the way including microarray, no answer resulted from any of that. And ultimately we did exome sequencing together with our colleagues at Hudson Alpha and looked at both this child and both parents. And what was found actually was two de novo variants in the child that is two variants that were not found in either parent, SP-10-1 and the set D-5. It turned out the SP-10-1 variant which is a duplication has been described before that exact same variant has in other individuals and I took a screenshot of a paper where that was described and the clinical history and the MRI are remarkably similar to our patient also cerebellar and brainstem atrophy, not as severe actually as in our patient but qualitatively very similar. So we felt very comfortable that this is very likely to be the pathogenic variant. It's been reported actually in several cases aside from this one publication. It encodes a spectrum cytoskeletal protein that is expressed in brain and the condition is referred to as infantile epileptic encephalopathy type 5. Set D-5 is a gene that has been associated with intellectual disability mainly because it's within an area of deletion in a copy number change that's been found in several individuals with intellectual disability. We feel pretty strongly that the SP-10 mutation is the one that's most relevant to the child's problems. This one we don't know how much it's contributing since this child would have been fairly profoundly intellectually disabled just on the basis of this one variant. And you look at what you learn from sequencing and, you know, this is true for us but I think it's true for most groups who are doing it. You find a handful of different kinds of things. Occasionally you find a pathogenic variant that's very well described in a gene that is quite well understood in terms of its pathogenic relevance and sort of kick yourself afterwards. How come I didn't think of this? We had a family of three SIBs with intellectual disability and kind of a degeneration of bones especially the hip joint who had gone through years and years of evaluations by lots of groups. Nobody had any idea what it was. They were being treated with anti-inflammatory drugs. It turned out after sequencing the diagnosis was mucolipidosis which is a pretty well known inborn error of metabolism. Lysosome will enzymes don't get trafficked to the lysosome. And so it wasn't such an obscure diagnosis and we kind of felt a little foolish that nobody had thought of it. And the only consolation was that we, I was at the long line of a number of clinicians who hadn't figured it out. And I think this is what happens when you're dealing with really rare obscure presentations that sometimes nobody puts everything together and it just popped out from the sequencing. So that's one thing that happens. The second is this kind of result. I think it's fair to say certainly for me and I bet it would be true for most of my colleagues that we would never have thought of testing this gene. In fact if we had I'm not sure we'd have found a lab that actually does the testing because it's so vanishingly obscure. It's just not on the differential diagnosis of this kind of condition certainly hadn't come out in any of the lists that we had generated either from our own experience or from looking at databases. And I think we find a lot of examples of that exceedingly rare obscure things that you can find a case here and there that convince you that it's real but still exceedingly obscure. Then the third thing of course are variants of uncertain significance and we certainly have our share of those as anybody who's doing sequencing does that we just can't really know if this is truly pathogenic or not. There often is a convincing case that you can make based on either segregation in the family or maybe de novo or maybe that you can infer a change in the function of the protein by computer modeling. Even go so far sometimes as to make animal models but you know when is the threshold crossed that you're confident that it's really pathogenic is a tough area and perhaps one of the things that may help in this is data sharing among groups which I think offers at least the possibility of changing an N of one to an N of many which may convince everybody that some of these truly are or for that matter truly are not pathogenic but however you look at it this has really changed the landscape in terms of diagnostics. It's not always so easy these days to get insurance coverage for the cost of the sequencing even though you can make an argument that whatever you're paying these days it's usually in the sort of six or $7,000 range when you package the sequencing and the analysis depending on who's doing it and various sort of insurance deals but getting coverage is not always so easy and I don't think the insurance industry is so much doubting that it can be helpful. Very few of these disorders or at least not enough of them when you say are treatable because we now know the answer this family benefited because among other things they learned it was a de novo event. Yes, two de novo events but still de novo and therefore unlikely to recur or germline mosaicism so there was some value. It short circuits that diagnostic odyssey they aren't gonna be going through lots more diagnostic tests because we know what is causing this problem. So I think there's been tangible benefit and it will probably save in terms of additional studies far more than it would have cost to have done this. So I think they get that. I think what is the area of concern includes the variants of unknown significance that I mentioned earlier and how much of a chase do you go on trying to validate those and then the issue of incidental findings. This is a screenshot from the paper in genetics and medicine now a couple of years ago for a committee that I was part of trying to make recommendations for what would the threshold be for looking at incidental findings and reporting them. The recommendations in the original paper were the top three things. A list of mutations I'll show you in just a moment should be reported to the referring doctor regardless of the age of the patient. They should seek and report mutations on this list and the ordering clinician then is responsible for counseling both pre and post test. Part that wasn't stated in the original paper was number four that patients could opt out of learning about incidental findings. And I think the rationale in the committee's mind and say this having been part of it was, one of the things concern that people could blithely opt out not really understanding what it was they were opting out of. And the metaphor that gets used a lot which I know is imperfect but probably as close as we're gonna come in daily practice is the person who gets a chest x-ray because they've let's say bruised their ribs in an accident. And the radiologist notices a shadow on the lungs that has nothing to do with the bruise but might indicate early lung cancer. And you probably know that first of all if they didn't report that and it turns out someday it is a lung cancer and you go back and look at that x-ray and realize it was there all along there could be liability for having not called attention to that. And I think that's a fairly well established issue in radiology and nobody as far as I know ever can sense a person who's about to have a chest x-ray. Do you only wanna know about your rib or do you wanna know about all the other things that might be found? So using that sort of metaphor which again is not a perfect one the committee didn't offer an opt out it created a firestorm sort of be careful what you wish for you like to get attention for the things that you publish and we certainly got attention but there were a lot of people who were offended at the idea that there wasn't an option for a patient to not want to know. I think that now it's been added a year later it was added into the recommendations. Talking to folks doing this very few people choose to opt out but at least now the possibility is there. This is the list I won't read it to you but the concept here were there was a very high bar that a gene had to cross so to speak in order to be listed it meant that this had to be something where there was really pretty much fireproof evidence of pathogenicity for that variant. It had to be medically actionable in other words it had to make a difference to the care of that individual to have determined this prior to onset of signs or symptoms. Of course you wouldn't know if they had signs or symptoms at the time the test was sent. So it had to be a manageable condition where you could alter a person's outcome based on the finding natural history had to be well understood and the penetrance had to be well understood. So lots of things that many people in the committee and now there's a community effort to add or for that matter remove appropriate genes from this list but the committee has a fairly stringent test and it won't put things on where there's any ambiguity about what the significance might be. A lot of these are cancer predisposition syndromes these are associated with aortic aneurysms and dissection cardiomyopathies that can occur insidiously arrhythmias hypercholesterolemia and malignant hyperthermia so it's a fairly short list of types of conditions and fairly modest list of genes. Some groups have added on their own to this list so this is still a very much work in progress and by the way it was recommended that a child who might be tested let's imagine for intellectual disability and incidentally might be found let's say to have a BRCA1 mutation that is of known pathogenicity that we would report that assuming the patient has knocked it out of any incidental findings that would be reported to the referring doctor and presumably in turn to the family even though we don't generally recommend testing children for adult onset disorders and the rationale here first of all was that who knows if this family has a history of in this case breast and ovarian cancer sometimes families are small and that diagnosis hasn't surfaced in anybody that anybody knows about in recent generations now you know this child has it and what would be the rationale for waiting for cancer to occur which could be the outcome if you don't disclose it so we wouldn't have offered testing to a child with a known family history because then the child when he or she grows up could make a decision about testing but if we have no information about family history and you discover it incidentally it could have significance to the child and incidentally also to a parent who probably is a carrier and in a way the child benefits by the diagnosis of this risk in the parent who can then be offered surveillance and risk assessment and risk mitigation by the way this is some of the stuff that kind of ironically in a way scares insurance companies I think ironically because you'd think it would be a good thing to figure out that somebody's gonna get let's say breast and ovarian cancer in advance of actually having that diagnosis since from a purely financial point of view it should be less expensive to prevent the cancer than to treat the cancer nevermind the toll on the individual and I don't think they doubt that really but for one thing there's this concern about opening Pandora's box on lots of testing for incidental findings that may or may not really lead anywhere but probably will cost a lot of money and general concern about doing all these things when realizing people pass through insurance companies pretty quickly so the investment for your future health in our system is not necessarily the first thought on the mind of all the companies who may only be covering that individual for a short time so this is still a very dynamic area and I guess I'm optimistic that the case will be made that this will become more and more often covered as time passes. So moving on to preconceptional screening now Laura is married and her husband considering starting a family in there of Northern European ancestry offered carrier testing for cystic fibrosis perhaps among other things there has been for a long time in the US a sort of opportunity to offer preconceptional testing to couples oftentimes guided by ancestry this is not a full list but here are a few examples of tests that are commonly offered based on ancestry and fundamentally based on a particular set of variants that are found more commonly in individuals with a specific ancestry versus others and some cases the carrier frequencies can be quite appreciable that information then can be used as a guide in terms of family planning including possibilities of prenatal testing this is a moving target too I pulled a screenshot from one company with which by the way I have no involvement and you can't read it you can just be impressed by the number of conditions that are offered for screening and I think this is increasingly becoming the case some of those conditions are vanishingly rare and so therefore very few couples will turn out to be both carriers for a pathogenic variant I guess the philosophy though is the incremental cost of adding a variant to the list now is dwindling to almost zero and as a consequence there's an argument well maybe this is vanishingly rare but somebody is going to be affected by this condition and if you could figure that out in advance why wouldn't you and especially if you do know the natural history of the condition these are not all actionable conditions in fact probably most of them or at least the high proportion aren't which is why they're on a preconceptional screen so the question is beginning to be asked well what about sequencing here would that be a better way of picking up carrier tests and well the answer there is it's sort of a complicated territory I think one of the issues is that there are plenty of sort of ambiguously and maybe even erroneously annotated variants in the literature so you can find a variant and think it's pathogenic and it will sometimes turn out the evidence base for that wasn't as good as was perhaps imagined and so you could actually at least potentially give misinformation with wholesale sequencing until we have a better handle on what the pathogenicity of variants actually might be another issue is just what the carrier burden is so I pulled a screenshot from this paper quoted at the upper left they didn't do whole sequencing actually it was really looking at I think it was 450 genes but you notice that only a small proportion of people just with that list of genes were not a carrier for anything almost everybody was a carrier for some things and a small number of people for many things and I think it's a fair statement that if you sequence the genome of everybody in the room we are all carriers for something that if it were homozygous would cause some kind of important medical condition chances are partners are similarly carriers unless we're from the same ethnicity or even in other ways related maybe low however the carrier frequency in the population is likely collectively across the genome to be quite high which means that somebody's gonna have to be there to provide counseling to individuals based on this test and argument could be that the number of counselors out there is just not scalable to the potential need which I think argues on the one hand for training more counselors but on the other hand for coming up with new paradigms of how counseling can be done that might not require the kind of traditional one person, one patient in a room with one counselor talking one gene at a time it just may not really work as you start looking at a genome's worth of data. All right so what about prenatal testing? Actually they are found to be CF carriers but then you like to have a prenatal test and the fetus is not found to be a carrier. There is a long history as I'm sure you're aware of prenatal diagnosis beginning with amniocentesis now the possibility of chorionic villa sampling and even pre-implantation testing biopsy of a blastomere testing it for a variant that the fetus or the embryo is at risk for and then implanting embryos back into the mother demonstrated not to be carriers of that particular mutation. What about looking at non-invasive approaches? This is a screenshot from a paper a few years ago looking at what has now come to be called non-invasive prenatal screening or some would say prenatal testing wherein a sample of mother's blood is obtained we all have DNA circulating in our blood from cells that degenerate included in that if a woman is pregnant is a small proportion of fetal DNA by doing sequencing and very sophisticated bioinformatic analysis it's possible to essentially count the numbers of copies of chromosomes 13, 18, and 21 the three associated with live-born trisomy syndromes and have a reasonable ascertainment then of pregnancies where a fetus with trisomy is being carried. In addition now this is being looked at as ways of picking up other copy number changes and this is likely to be a moving target over time you're probably aware that the same technology is now beginning to be used in a very different context so-called liquid biopsy because not only can you pick up these fetal changes but if an individual has a cancer which includes cells that are dying and releasing their DNA into the bloodstream you can also pick up cancer specific rearrangements in fact I think there have been a few examples of pregnant women who incidentally happened to have undetected previously undetected cancer who have had it detected because there were a large number of copy number changes well beyond 13, 18, or 21 indicative of genomic instability in a tumor that up until that point they didn't know they had. What about sequencing the fetal genome? Is that a feasible option? And well again a screenshot from a paper a few years ago now where this was done as proof of principle that same DNA could be sequenced and if you know the mothers and fathers sequence together with what you can see in the DNA for mothers blood you can infer the fetal sequence it's not a mainstream test. Guess a year or two ago a colleague who's a bioethicist asked if I would work with them on a paper sort of a editorial that ultimately was published in the New England Journal should parents who want to have their fetal sequence done be allowed to do that and the premise was that yeah if that's what they wanted to do that there wasn't really a sort of ethical grounding to say no you can't have that realizing all the decisions that parents make on behalf of their children. Honestly I expected having had the experience with the ACMG guidelines that this would generate a lot of discussion and it really didn't and certainly it's not being done on a routine basis nor was the argument in this paper that it should be it was more a question of would you say no if somebody wanted it then that you would recommend it I don't think we were I mean recommending it be done but not saying that it shouldn't be done if somebody was willing to do it and I guess implicitly willing to pay for it. Okay let's move on to pre-symptomatic testing so Laura is now 45 and she's just learned that her older sister has been diagnosed with breast cancer and now she's concerned about her own risks and there in fact are others in the family now this is not meant to be their family tree it's actually a kind of modified pedigree of a family we did see in our own clinic this was the person who sought counseling she had a sister with breast cancer a cousin and a grandmother and then also a aunt with ovarian cancer and an uncle with prostate cancer and actually there was cancer on the mother's side of the family too so based on that family history this was a few years ago testing for BRCA1 and 2 were offered however it was offered first not to the pro-band but to her sister and her sister indeed was found to carry a known pathogenic variant in the BRCA1 gene and that allowed us then to test the pro-band and it turned out she was negative and reason for doing it that way it's not intuitively obvious that you would test a relative before you test your own patient is that knowing that her sister actually carried a mutation it's a fair guess that it accounts for the family history of breast and ovarian cancer on the father's side of the family so that the negative test has great significance because it more or less erases that concern whereas if we had only tested her and had nobody in this side of the family who had had cancer be tested there would always be a bit of doubt as to whether this particular pair of genes accounts for the family history and if it didn't we might have not tested for the right thing in our patient and we wouldn't know for sure what her risk really was so it was very powerful where possible to test an affected family member because then you have a high confidence in the ability of that variant to be predictive of the risk in your patient you know there's a discussion now since the tradition well tradition is probably the wrong word but the sort of standard has been when a person is concerned about risk of cancer on a genetic basis usually it's based on family history sometimes it's based on being relatively young at the time when the cancer occurs or it may be that a person has had more than one primary cancer those are the red flags that tend to point towards a genetic predisposition and so we usually see individuals I can tell you in our clinic we're seeing people almost every day now where that kind of history occurs and you make an assessment of risk and then send whatever tests seem to be appropriate given the spectrum of cancers that has been identified in the family but there is an increasing sort of discussion about whether BRCA testing particularly should be offered even on a non-family history basis but to all women potentially of certain ancestries or even all women period and the rationale is that there are some examples of people that have very small families that just nobody happened to have cancer in at least recent enough generations anybody knows about it going back more than a few generations is usually hard to do and so you will miss some individuals I can tell you for sure in our clinic we have had a few examples of people that ended up being tested that I probably at the time when it was sent wasn't so sure this really made sense because there really wasn't much family history and then you get surprised when a positive result comes back and realize that it would have been easy to miss this it's a debatable point because there are issues of cost and potentially variants of unknown significance that get found it's being tested actually my colleagues at Hudson Alpha are doing this in Huntsville now as a sort of pilot program there have been others so it's a point that I think is likely to be discussed for a while and hopefully there'll be data that will give us a better sense of what the clinical utility of that approach is you may remember going back now I guess about 20 years more or less when the genes were first identified there was a debate also like why would you want to know that you had a genetic variant that predicted a high risk of breast or ovarian cancer or same could be said for colon cancer if you can't do anything about it now you've just learned the name of the thing that you may get but maybe won't and is it just gonna sort of obsess you now with this worry without having anything you can do about it and in the early days though there were some things that made sense to do there wasn't a lot of data but there is now this is just one of many many papers showing a significant benefit to South Pingo opherectomy compared with surveillance in terms of prevention of breast cancer even though it's not the breasts in this case that were surgically removed so between surgery mastectomy and opherectomy and South Pingo opherectomy various other approaches to surveillance including MRI and chemo prevention there's a lot of options now that can be offered and so this is really I think proved itself to be extremely clinically useful well sequencing not only may contribute to identification of individuals at risk of cancer based on family history but in addition to selecting the appropriate treatment for an individual actually diagnosed with cancer and increasingly now there is either a tendency to compare normal and tumor or sometimes because of costs just look at tumor which can even be done from paraffin embedded samples and then identify a set of genes that are believed to be driving that tumor and use that information as a basis for selecting appropriate medications that target specific mechanisms and it first of all logically is an extremely compelling argument so much so that I don't know what it's like here I can tell you in Birmingham you can listen to the radio and hear about why you should if you have cancer go to some clinic or still sequence your genome and there even billboards about it so it's become a I would argue a fairly hyped area actually how much good does it do there are fair number of examples of individuals where it was possible to target a therapy to a gene that you might not have previously imagined would be relevant and it has made a big difference at least in the short run problem is number one there aren't drugs for every possible target you might identify now so sometimes you get information but it doesn't necessarily guide your management at least that's my perspective as a non-oncologist but the other point is sure you're all aware cancer is a dynamic entity if you didn't believe in evolution this will convince you because it is evolution in real time happening so you give a drug and it works for a while and then the cancer sort of mutates its way around that drug and the interesting in well I don't know interesting is the right word the challenging thing is that you can have multiple metastases and there's no guarantee the genomes for those individual metastases even will be the same anymore so it may not be any one drug that you have to choose but could actually be a whole family of drugs so it's a it's a very tough area I think in terms of evolving but over time whether it really makes sense to sequence your cancer genome today and expect that it's going to make a long lasting benefits probably true in some cases and perhaps debatable in others and there's tendency now to look at panels of genes that are targeting things that we know we have drugs to treat but over time since cancer is fundamentally a dynamic genetic disorder it certainly will help inform us about the cast of characters and the pathways that are relevant as time goes on and we hope also as we proceed into better approaches to treatment wouldn't exclude by the way treatments of rare single gene disorders I think for a while as genomics was sort of evolving there was this perception that you know the where this really was going to make a difference was common disease and rare disease was kind of 20th century stuff and now we're in the 21st century and what I think has happened between what I mentioned earlier on about breaking the diagnostic odyssey but also now even in therapeutics is a realization that rare disease offers huge opportunities to improve outcomes this is one that has resulted in the FDA approved drug tuberous sclerosis complex which is an area of clinical interest of mine where among other things individuals get these sub appendable giant cell astrocytomas or multiple renal angiomyopomas and a particular drug everolimus has been approved by the FDA it inhibits mTOR which happens to be ultimately regulated by the genes that are involved in tuberous sclerosis complex there too only one will be mutated in any one individual it disrupts the complex it results in excuse me high levels of mTOR signaling in this drug among others inhibits mTOR I think this will probably remain as a kind of unique case history in one way which is mTOR stands for mammalian target of rapamycin and it was discovered I believe in yeast as a factor that was regulated by an antibiotic rapamycin and once it was realized that the hammered and tuberous complex ultimately when disrupted leads to increased mTOR signaling it was like the name of the drug that you should test was written on the label of the pathway and sure enough rapamycin originally did show benefit and subsequently a rapamycin like drug was as you heard approved by the FDA and so now I see patients with these lesions and it's been a while long while actually since I've sent any of the Sega patients to a neurosurgeon or the renal patients used to go to radiology for embolization of the lesion these days most of the time right a prescription similarly in cystic fibrosis which is a mutated chloride channel uh... there's been effort to develop drugs that either improve the function of the mutated channel or reduce its degradation uh... when there's abnormal protein folding this is a screenshot from one paper for one particular drug I have a caftor uh... the lung function was dramatically improved in those treated versus untreated think there are now two FDA approved drugs in this setting uh... so it's an example of drugs that I think it's fair to say wouldn't have been on the list treat or even didn't exist until the genetic mechanisms were known and then subsequently informed an approach to development of treatment and uh... that's certainly been the area of interest for myself in neurofibromatosis as you've heard uh... where we have a clinical trials consortium doing clinical trials with drugs that only are now on the list to use because we have some understanding of the underlying pathogenesis which followed from the discovery of the genes involved in these uh... this family of disorders well finally let me say some words about predispositional testing so Laura's now sixty been well and now the question about whether genomic testing should be partake in this based on internet available options and few years back as the ability to do the genotyping came down in cost and as mostly GWAS studies had shown associations with common disease a set of companies entered this sort of marketplace with direct to consumer testing two of them are no longer in that business and this one which is still in the business has had a significant encounter with the FDA i did it by the way at my own cost or my own expense uh... two years ago before the FDA got in the picture uh... this is the kind of data you could get then you wouldn't get now because they no longer are focusing on these common disorders uh... as they did at the time but uh... this is what my risk of type two diabetes was based on whatever genotype they identified you can see it's not too different from the population average and in a nutshell that's kind of what i found in general i was at average risk for most everything uh... you heard of which i didn't think was so bad but i also i think a fair ability to take this with a grain of salt and realize that you're only chipping away at the edges of risk with this and that not all the risk of type two diabetes is accounted for by genes and even the number of genes that might be relevant only a subset of them were being tested for here the fear was well nor one how valid really is this i think is what concerned the FDA secondly what would a person do with this information if your risk was found to be low would it lead you to conclude that you're immune to type two diabetes and hence ignore the advice which by the way is the advice for most every common disease on the list they tested for lose weight and exercise more and if you decided well now that's not relevant to you because your genotype puts you at low risk you might actually put yourself at high risk the opposite could also be true maybe it's motivational to learn that your risk is increased and you will finally follow that advice that you might have previously ignored it debatable point as it stands now the concerns have been sort of addressed i think in terms of realizing the evidence base for common disease prediction is less robust than it needs to be to use it in this kind of context so that's not really the focus anymore you do by the way get this list of pharmacogenetic variants this is mine i'm at increased sensitivity to warfarin so a standard dose would put me at increased risk of hemorrhage printed this out and gave it to my primary care doctor and got the expected now what am i supposed to do with this put in a pile i don't know if it ever got scanned into our electronic health record even if it did i can tell you it's buried there the day should ever come when i need this drug it's unlikely to be realized unless either they've heard this lecture since i've given it enough times or unless they met unless i'm conscious enough i guess to uh... tell them but that truth is that the issues on clinical utility my judgment anyway a pharmacogenetics hinge as much on systems issues as they do on actual clinical utility debatable whether the cost of testing actually is sort of justified given the kind of societal costs that you save in terms of poor outcomes and whether there are alternative ways of making these decisions that might be less costly but i think that equation may change if one does the testing at a single point in time for multiple variants where again the incremental cost of putting one on a panel becomes really negligible uh... then the cost part of the equation really falls away and i don't think doctors are ever going to consciously say before i put a person on this drug i'd better get a pharmacogenetic test they will for some things perhaps but mostly i think this sees the kind of mainstream is going to do so through living in the background in the electronic health record and being a sort of pop-up that says you know this is somebody who either might not want to put on this drug if they're at risk of some severe adverse outcome that by the way happens now in parts of southeast asian china were a specific hl a variant predicts the stevens johnson syndrome based on exposure to drugs like carbon isapine everybody gets tested in some areas they carry a card that's what informs their physician that they should or shouldn't put on this drug if there's a clinical reason to use that drug that kind of approach over time i think is likely to be the way this will be used if you kind of take a big picture look now at uh... where's all of this you could ask two questions you know if should everybody be sequenced first of all yes but also then if you did when would you do it and where would the data live so when could you do it well i mentioned you could in principle do it prenatally raises lots of questions how much do you want to know about a fetus in terms of what might be untreatable unpreventable adult onset disorders we could perhaps do it in the newborn screen as you've heard somewhat similar set of questions yes you can diagnose some things that maybe you could treat if you knew that they were going to happen but a lot of things you'll figure out you might not be able to do much about but how does it change the nurturing of a child to know about these things do it in childhood just a little older so all of them raise those same set of questions so maybe you should do it if you're going to do it all in adulthood only challenge there yes you can give good information and obtain informed consent but maybe you've already lost the opportunity to intervene on some things that would happen earlier in life and then if you were to do it where would you keep the information well in the electronic health record you might say couple of challenges there any of you who are clinicians and whatever electronic health record you're using are probably aware that the ability of current day electronic health records to mobilize genomic information is still some ways off as a matter of fact the ability to mobilize even routine health information in some cases isn't as robust as we wish it would be I think many people would be concerned that the electronic health record maybe it'll be the long-term solution is not ready for accommodating large-scale genomic information right now and even if it were pretty much nobody I think these days are very few maybe not nobody very few live their whole lives in association with one particular health person grows up they go from a pediatrician to an internist maybe people move therefore they're not necessarily getting their care today from where their genome sequence might have been done two years ago will it be available and even if perchance there's a way of dealing with that if they're traveling and they need access to the sequence is it gonna be there if it's locked in an electronic health record somewhere so I don't know that it's the most robust way to do this could be stored in the cloud which I think it already is for many people good news is that you should be able to access it pretty much anywhere anytime you do have to have confidence in the security I think that's an answerable issue could live on a personal device that already does for some people which is fine as long as you have it with you that day or don't put it in the washing machine or whatever else can happen so uh... it would be a viable alternative and finally my favorite is the most efficient place to store genetic information seems to be the cell nucleus and nobody ever forgets to bring those to the doctor and it is possible to imagine that the cost of sequencing could go to a point where it's cheaper to re-sequence when you need it then to bother storing it then just re-analyze it for whatever the question of the day turns out to be I have no idea really how this will evolve anyway I think all these possibilities are on the table back to the original question I asked when will doctors use genomics on a routine basis in their day-to-day practice well arguably some are especially I think medical geneticists have very much embraced cytogenomics genome sequencing I think you'll find those that do preconceptional and prenatal testing have embraced it but what about sort of mainstream day-to-day practice and here's my answer so this is you can't read it I know but it's a screenshot of iTunes so the reason I show it is if you think back I don't know how long it is now it must be close to twenty years fifteen plus years maybe it was possible to store music on an electronic device in the late nineteen nineties and if you try to do it you can either copy a CD you owned or may remember you would go online and they were illegal it turns out music sharing sites that you could go to and download seemed like a great thing everything was free and all these things you never thought you don't you could own uh... then it became clear it wasn't a legal approach well alright so then the uh... companies came into existence and I remember using one one time and it was so complicated I gave up quickly I couldn't really make it work and it was when this approach came into existence and now many others that are out there it became easy to use sort of fun not expensive and ended up becoming mainstream so my answer to when genomics will be used day-to-day is when somebody invents the itunes of the medical record uh... something that's readable easy fun works seamlessly I think that information access is when physicians are likely to use it I'll end with a point about education and about the deployment of this so a lot of people talk about the genome as the book of life and I like to think of it not so much as the book of life but as the library of life because you could argue each gene is sort of a story unto itself or by this metaphor book the genome is the whole collection of genes so the library is the collection of books look at this from the perspective of a practitioner who wasn't educated in genetics and genomics you could say alright so genomes a library full of books and what do I remember about genetics well one thing I may remember is that the genetic code is a three-letter code so you can say alright what's a book that relies on three-letter words and this may come to mind and you could come to conclude well how hard could this be and my answer is that if you read the genome as a book and you want to look for a literary metaphor a better one is this that you could teach a first grader I think to read Ulysses if by that you mean pronounce the words even that might not be so easy maybe but should be feasible but the chances they'd understand it as a matter of fact the chances a college student would understand it not so easy right because it's full of obscure illusions and you have to read between the lines so that there are books about how to read this book just as there are books about how to read the genome and I think that's fair statement of what reading the genome is like and I can tell you when we have our conferences as we do now regularly with our colleagues at Hudson Alpha looking at genome sequencing data from patients that we've been seeing you sort of feel like you know a discoverer in a new world uncovering things sometimes no one's ever seen before or been seen very infrequently at best and there are debates about is this or isn't this pathogenic and it requires a pretty high level of sophistication to figure that out into stretch of literary metaphor doesn't take long to pass through the looking glass and the rules you thought you knew don't always apply quite as you thought you know there are examples variants that do not change the amino acid so we would have previously overlooked them but then they actually turn out to alter splicing and they are pathogenic so a lot of examples now of things that don't seem to behave quite the way we would have originally predicted so it's not necessarily an argument that the only ones who should ever do genetic or genomic testing are bonafide medical geneticists or certified genetic counselors because if that's the position we take I think it's going to be a very difficult way to scale this to really becoming mainstream in use I do think there are going to be futures for people training in medical genetics and genetic counseling for a long time to come because someone's going to have to help interpret this complex Ulysses like genome and you know figure out when you pass through the looking glass what's on the other side have a very strong bright future for people who are interested in that training but I also think we have to work very hard to provide point of care education to physicians and other health professionals who are really going to be at the front lines of using this and increasingly innovative tools to help explain genetic and genomic information to an increasingly large population who stand to benefit from it as you all know here especially stand to learn a huge amount in the coming years as the precision medicine initiative occurs so that I think there will be a much better annotation both of the genome and its relevance to both health and disease uh... that will strongly inform this effort uh... but i'll end with what's almost now becoming a cliche it's been used a lot but i think it really is a a exceedingly powerful statement coined by ellen kaye who as I understand it is viewed as the father of the laptop computer which i think would be a really great thing to have on your cv if you could but anyway he said the best way to predict the future is to invent it that's certainly what goes on here it's what's going on now around the world and i think it's what makes this era the one that's by far the most exciting time in the history probably of the of humanity to be involved in practice of medicine and for that matter specially genetics and genomics so i have no idea the future is going to look like in sixty years other than to say it will look nothing like the world that we're in today and i think genetics and genomics will be a huge driver helping it to get there guess there's time for questions yes the question is how hard did i scrutinize the twenty three me results to that validate that they were based on accurate risk estimates to be honest the answer is not that much because i think i went into it with pretty high level of skepticism from the start i didn't really think this was i thought it was interesting i suppose you could say musing a lot of people view this as recreational genomics and it kind of was that from my point of view so i didn't take it seriously enough bothered you know delvin t you know the number of variants was too many for me to put the time into it and i guess i didn't think there was a high enough likelihood that i'd learn anything that would be that meaningful so i never really bothered my sense was also that whatever risks they were looking at were only you know part of the true risk equation for anything that you might be at risk for i will say parenthetically uh... i did learn that i was a carrier for something i didn't know before wouldn't have changed anything in terms of family planning i don't suppose but it actually could have implications for the future there's not a whole lot i can do about it and so i don't you know i've looked at that a little bit it kind of is what it is so the answer is no i just never really felt like the common disease risk was going to be as powerful as maybe it was originally hoped for a kind of you common disease risk at this moment kind of in the same category as weather prediction maybe it's more like climate prediction and i kind of viewed this is about the equivalent of looking at the farmers almanac to figure out what you should wear tomorrow and feeling was kind of interesting but not could really change anything