 Well, good afternoon, everyone. Thank you for coming. I'm Eric Green, Director of the National Human Genome Research Institute. I'd like to welcome all of you to the sixth lecture in the Genomics and Health Disparities lecture series by way of reminder or background. This series is cosponsored by four NIH institutes, the National Institute of Diabetes and Digestive and Kidney Diseases, the National Heart, Lung and Blood Institute, the National Institute on Minority Health and Health Disparities, the National Human Genome Research Institute, and then in addition to those four components of NIH, it is also sponsored by the Office of Minority Health at the Food and Drug Administration. These five organizations get together and choose the speakers who we ask to come in and present their research on the ability of genomics to improve health equity, or at least to help us understand the issues around the use of genomics for such purposes. The speakers in the series are deliberately picked to approach the problem from different areas of genomics research, including basic research, population genomics, translational and clinical research, and so forth. And the lecture topics have a focus across many disciplines by design, and today's speaker is especially notable for his ability to use genomics to advance public health and precision medicine. So that's the series. This is lecture number six. More will follow, especially in the fall and beyond. And so to introduce today's speaker, I'm glad to invite to the stage one of our co-sponsors, the director of the National Institute of Minority Health and Health Disparities, and a good friend, Dr. Elisayo Praz-Stablig. Elisayo. Thank you. Thank you, Eric, and thanks all for coming on a critical day with a lot happening across campus and in the city. So it's a real pleasure to introduce Mark Collin. I'm trying to remember the year, but I think we met in 1989 in Aspen at the annual meeting of the Henry J. Kaiser Faculty Scholars in General Internal Medicine group. I was the first-year awardee, and he had just completed his time. And we got to hang out and talk and become friends, and over the years these kinds of relationships end up coming back, as I'm discovering being here in Washington with a lot of people. Mark is extremely accomplished. He spent most of his career at Yale, but in 2008, right, 2009, he was lured to the left coast in part accompanying Michelle Barry, his wife, who is an accomplished scientist in global health. And he became chief of the Division of General Internal Medicine as well as brought his research program there. Mark's legacy is that he and others reinvented the discipline of occupational health as far as a clinical discipline is concerned during 20 years of research and really was a definitive reference point in that area. He subsequently then reinvented himself or moved into looking at big data, trying to understand all the different components that lead to health outcomes. As someone who was very focused on the environment, he understood the importance of the physical and cultural environment, not just toxins, the individual behavior and demographics, and the biology. And I think this has led to his current role as director of the Stanford Center for Population Health Sciences, which he's been for the last two years at Stanford, which I think is a very exciting enterprise to pursue precisely what NIMHD has proposed is our research model of looking at the different components that lead to different health outcomes. And I'm sure you can weave the health care system into it in your role as the co-PI of the Stanford CTSA. So Mark is also an elected member of the National Academy of Medicine. He's currently the IOM, and he was a student, has an MD from Yale, where he spent most of his career before moving to Stanford. So it's a real pleasure to have him come and tell us about this exciting work. Mark? Thank you. Well, thank you for the generous introduction. Thank you, Eric. Thank you, Eloseo. As I said, Eloseo and I go way back. And thank you all for coming, frankly, with the director's council meeting across campus issues as senior associate dean at Stanford that are near and dear to my heart. And with whatever it is that's going on down at the Capitol today, I'm shocked that any of you are here, and very grateful. Thank you. Thank you. This is one that you clearly won't be able to read about online tonight. I want to start in the following place. So I was one of the, because in my more administrative capacity, one of the early recipients of the PAR that emerged from NIMHD, suggesting the development of disparity centers whose central conceit would be the application of the tools of genomics in addressing health disparities. And like many of you who have been engaged in this, I greeted this with both remarkable interest but also a very healthy amount of skepticism regarding whether or not this was an attempt to make something that was problematic politically correct or whether there was really underneath it all some potentially enormous scientific opportunity, which those of us who'd worked in various ways in health disparities, whether it was in the workforce or in other of the important life stages. So this was a big open question. And what I'm going to do today is walk you through a little of that angst that we went through in thinking about whether or not this was a good idea, starting you with reasons why it was a terrible idea, moving you to think a little bit about why it might not be such a bad idea. And then I'm going to finish by showing you just a brief window into the projects which we have been able to launch as part of our new center, which graciously funded by NIMHD, and in which I think there at least are some opportunities that hopefully we will begin to explore. As a spoiler alert, I don't want anyone hanging on the edge of your seat waiting for the answer to this question. We are not quite there yet. In fact, I don't really know yet whether I can view myself in the role I'm now playing as an intrepid explorer into a really great new world or someone acting out of remarkable hubris and trying to solve a problem that is effectively insoluble. But in any event, here goes. So let's start with the reasons we should not go there, because I said the first 10 discussions that I had with my own team and with colleagues were in this space. And the most important of them I call first things first. And I don't think anything I'm about to show you is new. I'm treading over territory you've seen before. I'm going to show you work in my own hands just so you see how I personally got to where I worried about this. But it's all laid out on this single map. You've seen many versions of this map. This is actually a rather old one. But at the time we did the study that I'm going to show you, we didn't have full access to the decade later data, although I'm going to show you a little of that a bit. And this just shows the probability for white men in this particular slide of making it to the age of 70, not exactly old age. But we use survival to 70 because we wanted to be able to look individually at all of the counties, at least all of those with reasonable mortality in terms of size in the entire country. And I think you can see you don't need to be a rocket scientist. There's no big surprises on here. You can see that there are some counties out and about heavily concentrated in the South, but by no means uniquely in the South in which the probability of a white man making it to a social security check, assuming they decide to hang out for the real thing, is barely a little over 50%. And on the other hand there are places like the area that I'm happy to be living in as we speak today in which that chance is well over 80%. This is an incredible geographic disparity, not surprising. What we decided to do was see to what extent social factors that we could easily measure at every county level could explain those. And so here are a bunch of pretty predictable predictors that we could get from the decennial census. And here are a few more of those, and oops, and then a few more that could be easily gleaned from other available reference points, including looking at climate, looking at at least one indicator of available healthcare, an indicator of food supply and air quality, given my long standing interest in physical environment. And then a few demographic risk factors. And it will not surprise any of you in the slightest that a handful of factors for each of the four populations, we were unable to, in this analysis, look at populations other than whites and blacks because of the numbers in the various counties. So we looked at white men and black men and white women and black women. And you can see certain factors just show themselves as being enormously significant in explaining those health disparities. Again, I don't think there are any surprises on this slide. But what was a little surprising to us was this slide. So this is just looking at the residuals, which is to say, if you take those 18 or 20 variables and really only six or seven that add much weight to the regression, how much do they explain? And if you look at the four curves, you see they explain basically everything. So 80 plus percent of the variation is explained by those variables, all social variables. Again, this is ecological. We're looking at average health effects. We're not looking at individual effects. There's not a single big outlier. There's more variation in men than women, and I'll show you why it is set. And there's more variation in African Americans than there is in whites. But frankly, there's no surprises here. It doesn't even leave a hell of a lot of room to look for another explanation, let alone a genetic explanation. This is just how they distribute. And I think that again, the shocking information contained in counties in which African American men don't do very well. It's just horrific to look at a place where barely 40% of the population is going to ever see a social security check. This was again in year 2000. It has improved for African Americans, as you know, in the subsequent decade. I'll show you a little of that data in a bit. You can see that white men and black women are pretty much more or less the same distribution, and then the lucky white women over on the right. One question we wanted to ask ourselves is just doing a thought experiment. Just imagining that we could do the big project that unfortunately in the social climate we cannot do, which is level the playing field and give the inhabitants of all of these counties values not their own. And what we did first was a black white flip using a Blinder-Wahaka decomposition strategy. So again, we just redid the regressions. But now, for the black men, we substituted the white coefficients. And lo and behold, there ends up virtually no county in the country in which there's a big difference between the whites and these sort of make-believe blacks. And the same was true for women, although there was a little bit of a disparity that remained after you adjusted for it. One of the things we could not do incidentally using the same strategies to explain the difference between male survival and female survival. I'm going to give you some insight into that in another slide in a few minutes. Now, as we were doing this, we were also looking at possible, even though the study itself doesn't beg for alternative explanations. But we were curious about various things. And one of the ways we looked at the possibility that some of the variation, incidentally, I should mention that geography fell out completely from this analysis. So living in the south, which you might imagine is a remarkably bad thing for your overall health. And indeed, most counties in the south have dreadful health statistics by comparison. But it turns out when you adjust for things like education and occupation and wealth and the like, most of that disappears from sight. But in this particular slide, we took the data from the health and retirement study, which now has genomic information available. And we took the state of birth of all of the people in the HRS, which includes thankfully every one of the 50 states. And we looked for those states, using again the available data, at the distribution of social factors in the states in the decade that the HRS subject was born. And we looked at their genetic risk scores using three of the currently available polygenic risk scores. And I think you can see what's pretty obvious, which is there's no variation at the ecologic level in these risk scores. Basically from a genetic point of view, and again, unfortunately it's state level, not county. But from a genetic point of view, these things distribute pretty evenly. On a social point of view, however, there's four standard deviations in wealth, education, and the like, and it's true of any variable you look at. So just from, if you're looking for an opportunity to explain health disparities, you probably shouldn't spend too much of your time on the left side of that slide. Just again, this is remember, why we shouldn't go there. Now, we even went further with the health and retirement study, because we were particularly interested in seeing to what extent those early childhood predictors, either from a social point of view on the right, or from a genetic point of view on the left, actually, again ecologically predicted outcome. And on the left, the fact of the matter is they didn't, not very much. Again, not individually. This is not individual genomics, but the state they were born in. Whereas the state social factors 70 years later had predicted a substantial amount of the chronic disease burden that this population had experienced. In a model in which we took into account every other observable individual factor, smoking, non-smoking, and the like. It's quite stunning that this much survives of our early childhood. It's itself a huge line of interesting research, and does suggest that there may be an opportunity to try and study early in life what's going on that embeds many of these things into our overall health opportunity. Now, I would love to say that for those of us in the healthcare world, especially those of us like myself who are clinicians, that has nothing to do with us. It's all about social factors. One did notice a very small but real impact of the single measure we did in that study of health quality. But I think the slide outlines four bullets. I'm not going to go in detail. I assume that there's no one sitting in the room that is unaware of these issues, which contribute obviously to health disparities and which, in theory, could be addressed. So once again, why would we go looking at the genome when we've got this heavy lifting to do staring us in the face? Now, a second reason, and I'm not going to spend as much time on the last two reasons, because again, I think that this is probably that you are probably very well acquainted with most of what I'm about to show. But we have substantially less available information to study the genetic influence on health, certainly in minority populations. And there are two reasons for it. One illustrated by this slide, which is that we haven't done it. We just haven't assembled those populations. And many of us, as I've been discussing throughout the day, are waiting with bated breath for all of us program to all of us research program to mature so that we can begin to look at populations with both genomic and social information as well as clinical phenotype information covering the real spectrum of the US population. But in this slide, you can see that we've made apparent progress over the past seven years in the amount of GWAS studies that have been done in non-white populations. But you can see almost all that progress is in Asia. And we still have done virtually nothing to look at minority populations other than the Asian population that are major parts of the US population. Now, the other problem, again, for those of you who are in the genomics world, this will not come as any news. But there's also enormous differences in the ability to study some populations. And not only have we not studied African-Americans, but to study them is going to take a long time. Because there is, as you well know, an enormous amount of variation. This is from some work that's being done by Esteban Bouchard and his lab in which they have been trying to assemble some African-American populations and other minority populations for very important work they're doing in respiratory disease. I'm going to show you some more in a sec. And you can see that the sheer number of variants in the African-American population literally dwarfs what's seen in here, the Puerto Rican population, but which is at least largely white. In fact, it's not even entirely white and includes some overlap here. And then Native Americans, who have a much smaller amount of variation, another area that I will come back to before we're done. And then, of course, there is the part of this discussion that's actually most difficult for many of us to meaningfully talk about. Which is that the real concern for many of us around the genomic revolution and its applications in finding new drug targets and fantastic new therapeutic agents, which I think we can certainly say is rapidly emerging in the cancer field and is likely to emerge in a wide range of other therapeutic areas. That unless we substantially change the social organization of our healthcare system, these advances are likely to widen, not narrow gaps in that the part of the population that has access to the testing, has access to the knowledge, has access to be able to afford the therapeutic interventions. If indeed, Coumadin won't work, but you need a thrombotin inhibitor. It's unfortunately an area in which the parts of our population most in need of investment is the part of the population least likely to be able to exploit that. I just want to add one additional point before moving on to the glimmers of hope, that there may be some potential value in the direction that we are headed. And that is that even if, even if we could level the playing field on access and cost, even if I didn't continually hear things, we have various focus groups going now in some of the disparities work that we're doing, which I'll come to in a bit. And one of the things that we hear most often is we assemble focus groups to talk about what do they understand about the genome? What do they understand about gene testing? What do they understand about how knowing about your gene might influence your health? We are repeatedly assaulted with the following basically now predictable comment that comes out at least once in every focus group, which is, yes, my doctor said that I should go get this test only then to discover that even the test was not paid for by insurance, let alone any evidence that they could benefit from the result of that test. And we're talking about people that are, we're talking about 40-year-olds with breast cancer in whom their insurance would not pay for them to be tested despite a family full of women, that kind of thing. We really are, I mean, the last thing I'm going to say of this, but we really are in a deplorable state in terms of our ability to bring the genomic revolution to the bedside to deal with disparities. So I want to now turn the corner a little bit. I feel like I've gotten a little bit preachy here to talk about where potentially some of the opportunities may lie. And again, I'm not trying to in any way cover this in a comprehensive way. It's really more conceptually to sort of get people to think a little bit about where some of the work may go. When, for example, some of the worst of these problems are addressed, or at least where we have some confidence that they might be. So the first, and I think it's not a huge leap, and many others who you've heard from and who are publishing very actively, we're early to recognize that it's not the genome itself that is likely to be the explanatory factor that's going to help us really address deep disparities. And the reason for that's pretty clear. It's because the genome's not the cause of those disparities and probably contributes only modestly if at all frankly in some populations to what we see. It's very, very likely that environmental factors are contributing quite substantially. And so the question now lurks, given that we have the technology to explore it, is to what extent might understanding the relationship between those environmental factors and the genetic alleles of interest? To what extent might that relationship itself actually offer us the beginning of some clues as to a pathway for application? So again, here this is not to push the genome off the page, but only to recognize that it's not the genome per se that could be the target of our research and certainly not our applied research. But in fact, it's potential interactions with the environment. Now, usually when people talk about the subject, they show what I would basically call pretty lame examples of little findings. There have been some interesting ones, don't get me wrong. They're scientifically very interesting. But we don't really yet have a very, very broad pallet of places in which showing that social disadvantage as an environmental exposure or any of its components easily interacts with an allele or even something larger in the genomic world that would help us exploit it from an interventionist point of view. So I actually want to show you one that you probably don't think about in terms of a gene environment interaction. But for a variety of reasons, I think it's the right way to think about what I'm about to show you. So this is plots from... It's exactly the same thing I just showed you before from the S70 study. This is now with the newer data. And this is the black data, which I didn't show you before. Again, I don't think it takes much imagination to see that men and women are rather different here in terms of where they are on the plot. Women have an enormous advantage. And then over on the right-hand panel, you can see what the whole population, all the counties look like, all glommed together. And I think that the social gradient is pretty obvious. I didn't draw a line there because I wanted to draw the line here. So this is again the black data. And it may not be completely obvious looking at the curves how remarkably different the slopes of those two curves are. But you can see it a little more transparently when you look at the county by county ratio of male survival over female. It's a number always less than one, except in a few remaining post-conflict countries around the world. It's true in every single society. Women enjoy an advantage. But you can see that that advantage is not fixed. And in fact, here I've just done the quantification for you. You can see that the slope for men is about twice as great. The gradient for men is about twice as great as the gradient for women, suggesting that something in the biology of females is resilient relative to the biology in the same social climate. And the x-axis here could be substituted for any bad predictor. I mean, income could be there or proportion with high school education or whatever you want. I mean, it really doesn't matter which big social predictor you use. The curves look the same. If you doubt, if this doesn't look that compelling to you for a variety of reasons, because there is some truncation on the left side, because, frankly, there aren't too many counties in which the proportion of Africans-Americans are living outside of poverty. But when you look at whites, you can see that better as things get closer to zero. And you can see how remarkably divergent those curves are. So whatever is conferred by that second X chromosome is an unbelievably valuable factor. As I said, there's not a population in the world, post-transition population, where women don't live longer. But it is hardly a fixed biology. And what's stunning about this, at least to my eye, and I hope you would agree, is how remarkably different the response of men and women is to their social environment. In fact, I'm not going to go much further here with this particular line, because it's not the central focus of my talk. But frankly, when people say, talk about the biggest health disparities in the United States, the biggest health disparity is men. They enjoy remarkably poorer health and longevity than women over the long haul. And it's an area I know we don't typically talk about that as a disparities population. But you can see that it is the men who are the recipients or the victims, I should say, of much of the social variation that we have observed. And again, it doesn't matter what you use for an x-axis. So there are other reasons. Another reason for at least thinking in this space is that not only do we see rather huge differences in overall mortality, which I think from a data point of view is a very secure place to work, rather than sort of trying to dissect it by individual causes. But there are some diseases in which there really is a substantial difference in both the prevalence of the disease, and perhaps shockingly and more importantly, the mortality, especially for a disease like asthma, which is intrinsically treatable, and for which we in the biomedical community, even if we can't ourselves, prevent asthma, because we don't control the environmental factors yet or don't understand the environmental factors that contribute, we do control the treatment of patients who develop asthmatic symptoms. So the question here is whether or not this is an example or any of these other diseases that are hugely varying between or among our various populations, and therefore single themselves out as extremely relevant for disparities-focused research. And I was glad to see that some of the institutes that focused on particular diseases, cardiovascular disease diabetes, are active participants in this series, because I think this is one of the areas in which there may be some real potential opportunity here. So the second thing is whether or not diseases that we do end up not entirely living out in that social universe of disparities, but in fact come into our midst, they become our patients, we take care of them, we have opportunities directly to substantially modify the natural history. And in fact, it's particularly important for those diseases that occur relatively early in life, in which we can make real differences in health value over the life course, not just cosmetic differences at the very end of life. And this is one again, this is from Bashar's lab up the road from us at UCSF. And one of the observations they've made, as I said, they've been assembling some non-white populations and doing some fabulously interesting work, looking at GWASs within those populations and also looking at particularly looking at markers that may have direct physiologic or clinical relevance. And one of the observations is a substantial, at the ecological level, substantial difference within the populations in their responsiveness to standard bronchodilator challenge and albuterol challenge. And as you can see, the Mexican Americans have, on average, the kind of response that is actually fairly typical in the white population, whereas the Puerto Ricans and the African Americans that we recruited actually have substantially lower average levels of responsiveness, leading at least to the idea of here's a gene environment interaction, not that it'll help us make one particular treatment decision, you know, use this drug before this drug, but which might fundamentally change the way we treat if this entirely pans out and this is relatively new work. If this pans out in further trials, might introduce entirely novel ways to treat very important diseases that disproportionately affect minority and underserved populations. And finally, an area that has really blossomed in my own institution in which, you know, I recognize this now becoming widespread is to extend substantially our notion of what genomics is really looking at and what the real focus of genomics should be into all of the more downstream products of what the Human Genome Project has uncovered and the things that we are now learning. And it may be, and I'm just gonna show you a very, very simple example in a second, just to illustrate the point, it may be that again, even if in the gene environment interaction, we can't immediately see a reality that allows us to intervene as we fill in the missing pieces and try and understand mechanistically what's happening between the cellular level and the ambient environment or the social environment, we may find that there are very exciting opportunities for at least understanding and possibly intervening. So here's an example, this is from Steve Kohl's lab. Steve, if you don't know him, is an oncologist at UCLA but has been doing some extremely interesting work looking at gene environment interaction. His particular interest has been in the bottom of three steps leading to the genome. The reason for showing the slide is you don't ever see the genome at the bottom. I mean, that's just not the normal way, you know, normally it's the genome first and in fact, a lot of the inferences from gene-wide associations and others have been the primacy of the genome in the biologic pathway and quite pointedly, Steve's group has developed a series of nice illustrations like this in which they point out that it's actually not the genome that starts this process necessarily but what the genome does. Obviously, huge portion of the behavior of the genome, some understood, some not yet fully understood, is biologically determined and over which we don't, you know, that regulatory process, the target of an enormous amount of research not aimed primarily at disparities but basic mechanistic understanding. But what Steve's lab has been trying to do is sort of understand how it is that the social environment might be getting under the skin and in this fabulous drawing and this sort of paradigm, the genome now sits as the target of the environment in an odd way, again, not the genome itself, we don't imagine the genome is changing, thank God, at least not for this purpose but the factors around it, maybe there's methylation going on, maybe there are other pathways by which transcription is being rapidly accelerated, turned off and so forth. And so one can think about this pathway and say this is interesting, let's try and better understand how these pathways work under a variety of what he calls social processes, I would translate that into environmental circumstances, the world in which we actually live. But the really more interesting question is to try and understand these two issues, which is not what's happening time after time after time after time as sort of a kind of a system that's auto-regulating as many of us who went through learning basic biology, learned about the way these systems auto-regulate, but how it might turn out to be that environmental factors or social factors might lead to a more persistent embedded change in the signaling process. And this is just one slide, this is again from an animal study that he did because he's looking for examples in the laboratory, I'm sorry, this is not animal study, these are from human leukocytes rather in vitreous study in which he is looking at cells drawn from different phenotypes. And in this one just showing that persistently individuals that they classify again, this is the psychologists in their group classify as lonely or isolated. And the relative difference between the inflammatory pathways and their obviously more protective pathways and these turn out to be constituent. I mean this is persistent in these populations and presumably ends up having something to do with why these, how this is getting under the skin and why it is, if you think back at my first slide, one of the far and away biggest predictors of survival to age 70 is the fraction of the population that has partners, the adult population that has partners. So it's probably a pretty important example. So he shows this nice example of an animal, this is now from animal work, of an animal raised under two conditions. So in the experiment they initially raised animals on what they consider to be very favorable social conditions in a peer group and then basically throw them out of the family at the age of six months. And what the slide illustrates is I just want you to look at the black dots which illustrate an interesting phenotype that emerged and turns out to correlate with a genotype in some new work that they're doing which turns out to be a particularly volatile one. So most animals in fact after some decline in the more productive part of their gene expression will revert back to a probably reasonably stable, healthy mean. Certain animals about 5% of the total population they studied actually warped off into orbit after they were thrown out of the family and ended up doing very badly and most of them died early in the experiment. But this is an example of a single phenotype and again corresponding to a genotype that they identified that appeared quite resilient. So an enormously exaggerated response to being kicked out but then what appears to be a sort of a palliative or resilient overall response so that by the time that they were two they rejoined the biologic pattern of their peers. So maybe we've got examples of that too and maybe understanding what biology of resilience is as understanding from my point of view the biology of being a woman will turn out to be that protects women from some of the worst scourges of social adversity. Maybe this will turn out to be advantageous as we begin to try and explore other forms of disparities dealing with populations that appear to be quite environmentally sensitive. So in that vein we decided Huber is not withstanding to go ahead and apply for the opportunity of putting together a center and this is our sort of logo and we're just gonna show you some of the projects that we have now embarked in. Again, I warned you at the outset that I wasn't gonna have you leaving here saying wow, look at what they can do because we haven't done it yet but we've made a really lovely start. So the first of our three projects called Bracelet is a project looking at the relationship between genomic determinants in the American Indian population and we identified a set of partners among the Lakotas in South Dakota on a couple of reservations. This is our team and their team at the first of what have now been five meetings. This is there, you can't see it very well but this beautiful marker that marks the entry to the reservation. And frankly what we're trying to do is something that will address one of the big gaps which is the willingness or potential willingness of our underserved populations to embrace what we learn in the genetic revolution. Will people even seriously consider the information? And for the Native Americans and I think many of you probably know some of the struggles that have gone on with engaging Native Americans in this kind of research, the biggest problem has been in control and ownership of information. And one of the things, there had been a failed attempt by folks in our genetics department some years ago to work with the Lakota around a very high prevalence of autoimmune diseases and it was clear from a national point of view and the local tribe as well that this research couldn't go on because the data was being handled too much by us. And so what we're gonna do, very simply you can look at the aims but the goal here is to help them build a biobank, help them learn how to do genetic testing themselves, learn how to interpret the information, to educate the tribes on what's important and what's not important in the hope over the period of what we imagine we're just now starting to build the physical biobank, I mean like the freezers, teaching them how to process material but more importantly beginning an educational process across the reservations to engage people in the potential value that knowing more about their own biology may be critical to their long-term health. Now this is a very, very, very steep uphill struggle. These are reservations in which between you and me the biggest problem is an enormously high adolescent substance abuse problem and suicide rate. That's what they're dealing with on a day-to-day basis and we do only with a certain amount of trepidation go on to the reservation and talk to them about underlying biology but we do think that at a minimum we're gonna learn something about what it is we are gonna have to do as a scientific community to help develop this level of expertise and knowledge in the populations who are potentially most at need for the information. Now a second project epitomized by this, oh I didn't get the pretty slide, oh we have a lovely slide of the kids, sorry. This is a project of Tom Robinson that actually has been going on for a couple of years. It's a very, very intense long-term intervention strategy to deal with the obesity epidemic in Mexican-American adolescents. And the focus is, yeah, pre-adolescence. So the focus is on identifying vastly overweight or clinically obese children between seven and 11. They've been put through and you're gonna see what we're adding to the study. They already are about two years into an intervention that involves very intense home intervention. So there's a family component, there is a school component and there's an after-school component and I had a nice picture of them playing out on the sports field so you can see kind of how this is going. But the results have been very mixed which is to say that some fraction of the families thankfully it appears over 50% are doing amazingly well and kids are having a really good result from this and it appears over a couple of years that this may be sustainable. Of course you never know whether it's sustainable until you go there. But we decided, and this was sort of the conceit that we added as part of our sphere center, we decided to use the omics idea and Mike Snyder who's the chair of genetics at Stanford has popularized the notion of iPop which is integrated personalized omics profiling in which he not only does these studies but does them repeatedly on population sometimes as frequently as every two weeks in a study he just finished looking at pregnancy which is gonna be amazing when we start seeing the results. But in this case what we're doing is taking advantage of the material that Tom's group thoughtfully had managed to biobank. So we have a nice trove of the periodic specimens that have been obtained during the several years of the trial. We did have to re-consent everybody because it wasn't clear at the beginning of this that we were ever gonna do this kind of work. We've been very gratified that somewhere in the 80 plus percent range have been willing to consent for the omics testing which has been great. I should point out incidentally that this trial has had a 95% adherence rate over two and a half years. So it's a testament to the principal investigator and his community engagement team that we've done so well. What we'll learn from this I'm not sure but the hope is is that by looking at the entire pathway between on the one hand the underlying genetic alleles but on the other hand what the expression of those alleles has been we'll be looking at the methylation pattern we'll obviously be looking at the expression patterns over this period of time as well as all of the intermediate products of that. So at least at a minimum we are sort of it's preemptive of what will be going on in the all of a study it incorporates many of the same many of the same concepts in this population. Now with the superimposed clinical trial which will hopefully give us some insight somehow into what it is that appears to be working in those who respond and what's working or what's not working in those who have failed to. So this is what the aims of this piece of the trial are. The good news is virtually all the materials in hand everyone's been reconsented who's willing to do it and we are well underway so that the next time you see anything coming out of our center at least some of the descriptive statistics will be available and it is just as chance would have it kind of a nice opportunity to look at an interesting cross-section of the Mexican-American population of that age. And finally a study that's aimed again at the people. And this is a study that V.J. Paracoil one of my junior colleagues who's a geriatrician and a palliative care physician based on her quite substantial experience primarily with cancer patients has anecdotally observed a both relatively poor performance on the part of physicians who identify enormous opportunities either on the therapeutic or on the diagnostic side for genetic testing in cancer patients but also an extraordinary reluctance on the part of many of the ethnic groups that we deal with in Santa Clara County which has one of the most diverse populations virtually any county in the United States 40% of them were not born in the United States and 100 languages being spoken. I'm trying to get some insight into what it is that's happening in the communication between the healthcare providers and the patient and the patient's family that is impeding the general willingness to use genetic information. So this is a population generally they're all Medicare patients that's the focus of her study is looking at the elderly cancer patients so they're all insured and that's actually part of the purpose is to look at people who are being given various kinds of advice about diagnostic tests about the availability of unique treatment strategies given the surface markers or the somatic genome of their tumors in which we have good prior reason to believe that information is being very ineffectively incorporated ultimately into care trying to identify where the lesions lie in the hope that we can develop some strategies for ameliorating them over time. So with that, I will close as I said hardly a solution to the problem that confronts us all but we think we've at least identified some areas in which as greater access to populations accrue as our experience with being able to effectively interact with those populations improves we can meaningfully start to exploit the extraordinary opportunities created by the omics revolution. So I'll stop there, thank you. And I think we've got time for at least a couple of questions I unfortunately have a short leash and have to get out to Della soon but I'm more than happy to entertain a question. There's looks like one here and I see another one over there. Courage people though to please come to microphone since we're videotaping and videocasting or the videotaping it, I think. So there's a microphone over in that. I was welcome. Maybe you started to talk a little bit about it but like for the three studies you talked about at the end of the Dakota, what exactly is the conversation around any genomic data and the plans for what's gonna happen with that data? Right, so the conversation is all about genomics data. I mean, we're trying to slip in a little health information while we're at it but the focus is, and again some of this goes back to the history in which very early on a senior member of the tribe had actually himself developed I forget some complex form of RA and noticed that lots, there was lots of arthritis on the reservation affecting young people and it actually through a series of calls contacted the geneticists at Stanford and they'd actually started, they'd come out and drawn blood on some dozens of members of the tribe and it started to do some work when in 2014 the national outcry from Native Americans began and basically the leadership of the tribe said you can't do this. So we actually had some specimens sitting around in freezers in Palo Alto and it stopped and so the good news, if you can call it good news is that the awareness of the excessive amount of rheumatic disease was very clear to members of the reservation. So there was some interest in trying to understand why so many of their kids were getting disease before the age of 15 or 20 and so there was interest in that and so the whole focus has really been on we don't know what the cause is, we don't know whether there's an environmental cause or whether there's an underlying biologic cause but because you're a relatively isolated subgroup who've been intermarrying over generations there's some chance that this actually lies in alleles so that's actually been the direction of the education. And at least up till now, again, what we haven't done yet is go out to draw more blood. What we have done is recruit people for focus groups to come to large meetings which have been extremely well attended. We have identified technically skilled people to help physically build a biobank. They're making some investment, I mean a lot of the investment's coming from us but they're actually investing in some of the hardware to do it. So we'll see, it's a terrific question, Eric and I'm just not sure where it's gonna go but we do think that at a minimum it's changing the dialogue that we're saying you're gonna learn how to do these tests, you're gonna learn how to read them, nothing's coming to Palo Alto, you're gonna have the information, we'll sit there and talk to you about what we think it means, you're the ones that'll tell each other, you will learn HIPAA, you will learn what it means to do all that and that's what we're teaching them. So we'll see. Hi, good afternoon. Just a question about the last slide that you had. You talked about assessing confidence of clinicians, communicating genetic test results and risk information to the breast cancer patients. So the question is, is it really a question of assessing the confidence of the clinicians or maybe the knowledge base of the clinicians around genetic testing? So are there genetic counselors involved as well? It's a great question and unfortunately, VJ is not here and I would, if she were, defer it to her. This was her choice of word and we actually had substantial discussion. Do you think they're not competent or do you think they lack the confidence to talk to these patients? And she said that at least from an anecdotal point of view as someone who was frequently called in as a palliative care consultant to see the same patients and so she interacted both with the clinician, the primary clinician and the patient. She said very often it was that this was a conversation these physicians were having every day. If you treat breast cancer, you're talking about genomics every day and yet they were seemingly not doing it. And so again, she felt that the problem was that they felt too nervous because they'd gotten pushed back. But I agree, it's a great question. Between you and me, I'm guessing both are true. Question, given the disparity as it relates to African-American males, I was just curious why you didn't have an initial project or initiative trying to take a look at what's happening in those populations. Well, if we were in Birmingham or in Nashville or even in Oakland, we surely would have done it. This was not for lack of interest. Well, you have East Palo Alto. We have East Palo Alto, which once upon a time, so you're probably dating yourself here, once upon a time was 80 or 90% African-American, it's about 10% African-American. Wow. So to give you a little reassurance, virtually all of the African-Americans in the peninsula have moved north and in fact the same is happening now in San Francisco and the African-Americans are moving north and most of them are settling in and around Vallejo on the North Bay. So we have started now a big county level project with Solano County, which is the county they now live in and we also have opened up a little office in Oakland where we're going head to head with our partners at UCSF to get access to the area. But the problem, frankly, was just lack of access. Only 3% of an otherwise amazingly diverse Santa Clara County is African-American. Thank you. Hello, my question is specific for your program, your institute. I wanted to know if there have been any considerations for research for recently arrived immigrants or has they acculturated into the US system? So the answer is we talk about little else and in fact in the two pilot projects that we're now able to fund about three quarters of the applications were in this space for a wide variety of reasons. It's not that it's hard to get access to some of individuals. What's very problematic is getting organized access, which is to say they don't have electronic medical records. What little care, what fragmentary care they get has unfortunately been shifted from federally qualified health centers to the safety net health system that the county itself runs so that it is probably next in queue for something to study. The other problem is that we're dealing now with, it's not an it, it's a they. So we're dealing with very, very different immigrant streams. I said there are 100 languages spoken. So we have, you know, we have Khmer, we have a large Vietnam, huge Vietnamese population in San Jose, but again, getting access to them and getting them closer to us is gonna require substantial improvement and enhancement of our community engagement core, which we're actively working on. And we view that frankly as the first step, which is really making some inroads in those communities before we feel comfortable in trying to attract them for research. Thank you, I understand that there might be many social factors or political barriers in order to have access to those populations, but I'm happy to know that there are initiatives at least. No, and thankfully we're in California. Thank you. These are the kind of studies that you couldn't do in many parts of the country, but we can do them. I think most Californians feel as safe as you can feel in today's America. I really appreciate how you structured your talk from don't even go there to the potential of scientific knowledge that could be gained. So my question is a little bit more along the don't even go there side, but I wonder if you would sort of reflect on what is the likelihood that this omics revolution will should change our strategies to reduce health disparities in the short term, five, 10, even 15 years? Well, of course it's a great question. And one I would happily try to sidestep if I were forced to do. I think the possibility for benefit in the next five to 10 years will most sharply focus on advances like the kind of work that Esteban Bouchard's lab is doing at UCSF, which is looking at very common, very important, very earliest in life, like as unfortunately increasingly type two diabetes is becoming, disorders in which there is a reasonably high likelihood that understanding the pathway could at a minimum change the way we treat such people much as there's been a substantial revolution in the way we treat hypertension so that it's one of the few things in which the basic algorithms for how we treat whites and blacks has changed. And although it wasn't initially based on a genetic observation, it was based on population level, you know, old fashioned heritability in which it was pretty clear that the physiologic pathways were gonna be very different and therefore your first best guess as to where to start would be different in the two populations. That's the kind of thing that I think offers relatively short term promise. I think in terms of will we develop tools that will substantially change the way we deal with disparities, I'm a little less optimistic, but I think it is incumbent on us to learn enough to make those probability confidence intervals narrower, which is, you know, I don't know, but right now the confidence intervals are bigger than the point estimate. I mean, I'm sorry, that's not a great answer, but it's probably the best I can do. Anyway, I have time for one more and then they're gonna swallow me up. Hello, Dr. Cohen. First off, thank you for a really informative talk. I'm just a summer intern here, so I apologize in advance. Good place to come for the second. I apologize in advance, if my question is not as impressive as some of the others, but I noticed there was a huge disparity which you point out between men and women in the probability of reaching the age of 70. And I was wondering if, in your opinion, this disparity is more based on biological differences, for instance, menstruation, which releases toxins out of the bloodstream, versus genetic differences by which females have the extra X chromosome and so can kind of read off the other chromosome if there's a problem of one. So it's the subject of a whole nother talk and so forth, as to how that sex difference emerges. So my own belief is that it has mostly to do with the way women live and that women are socially quite different in terms of their ability to support each other so that women don't always need a partner, as most men do, to be socially connected. And I think that, of course, derives from some underlying biology. So there's biology, and so we talk about that as a social biologic difference. So it's probably not estrogen, and it's probably not the singular advantage of menstruation or childbearing or any of the other more obvious biologic functions, but probably something that over evolution has come to protect the female of the species for very good, we think very good teleologic reasons in terms of the preservation of the species. But we don't know. The striking observation is the one that it's not a fixed difference. It's not like women live five years longer on average than men. It depends a hell of a lot on the social conditions. Thank you. Anyway, thank you all. Thank you.