 Thank you very much. I'd like to thank Julia For inviting me to come to this meeting. I think Translational research is at the heart of success In terms of solid mental disorders, and that's what many of us are looking at in addition to neurological disorders It's a challenge and what I'd like to present today is the relationship between translational research and What is now being seen as personalized medicine is looking particularly at a person and coming up with an appropriate therapeutic Treatment for that disorder, whatever it may be And I think we have to push ourselves to move in this direction and personalized medicine has been Coming up slowly But I think it's now here and we have to see if we can make it work In translational research. I think I worked at the IH for a number of years And I think we first started talking about it in the late 80s and early 90s And we said people had to do more trade translational research and the answer from the field was what is translational research? and I think I have some definitions that I'll offer up For both translational research and for personalized medicine as we go through this I think I have some thoughts about ways that we may be able to accelerate some of our progress as we move To understand the brain and the disorders. I apologize for the vagueness in this cycle of definition of translational medicine research, but It also is a little vague how you actually do this and the hours should be going in both directions and In particular at this meeting we've heard an awful lot about what I would say good translational research But it's always coming bottom up Into the clinical area. I don't know how much goes back down into the basic area from the clinicians It would be good to think about that But overall the primary goal of translational research have we have we as we have heard is benched the bedside and I think it is as I just said that Direction we need to get more going back to the bench so that basic research scientists can think about enhancing the clinical research efforts And I think integral to all of this our collaboration and data sharing and integration and coming up with standards I think we are as a group of people Unbelievably creative and we all like to do our own thing But I I really believe that if we're going to solve the problems that we're working on we have to start doing more standardization and Keep doing individual efforts of research, but also combine our efforts together That personalized medicine is a medical model That's emphasized in general the customization of health care And we're looking to do this because a lot of our health care is not good enough and we believe that diseases are Variable enough across a disease pattern that we have to look at the individual patient and make our Prevented efforts and treatment unique And I think translational research is core for success in this area now whose personalized medicine critical for Well, obviously at the top of the list it's the patient consumers, but it's also critical for clinicians Diagnostic development drug regulatory system health care delivery and pharmaceutical discovery stakeholders Why why do I feel it's urgent to do this? Why do many people feel it's urgent to have personalized medicine especially for mental illnesses? One of the problems is that the illnesses are increasing The diagnostic methods that we currently use are subjective and Treatments are often based on trial and error So we try one drug and if that doesn't work we try another one if you look at data from the United States for 2008 this represents the number of people suffering from mental illnesses in Millions in in one particular year and the numbers are enormous and if we take this across the world They're staggering the number of people that are impacted by this It's not just individual suffering, but it's suffering within the families and impact on overall life All these diagnostic categories that coming out of DSM and I find it fascinating that the depth of DSM knowledge Grows and grows and the criteria that we use get broader and broader to describe what disorders One of the big problems with DSM is I believe is that we move from More of a Pathological condition into a medical model so we describe symptoms and clusters of symptoms and it's not biologically based And I think we need to move to a biological basis for mental disorders if we're going to try to relate as we've heard here today Our illnesses to genes and biological processes I don't think we can relate them to the medical model. We need another way to look at diagnosis So the problems in diagnosing mental disorders are that it is subjective There's heterogeneity in the categories There's extensive comorbidity in mental illnesses. It's it's interesting today in these two days We've heard about many illnesses being discussed and They all are presented as one illness, but are they really one illness? Are they multiple illnesses? And I think we have to look at this with great clarity There's over-specification of illnesses and Therefore, there's a lot of not otherwise specified categories And the biggest problem is that the bottom line which is following the end is that these diagnostic systems drive our research clinical trials journal publications and FDA approval and I and many others don't believe that they're doing an adequate job This gives you an idea of comorbidity It is enormous across all these disorders and the leader is bipolar one They have the most comorbidity of all illnesses, but when we do studies we talk about one disorder not many The other issue that comes up in treating mental illnesses is who does the treatment? Well, ideally you'd like to seek psychiatrist treating the mentally ill, but in reality More people go to a general medical service to treatment And then you run into problems because they don't do a great job in terms of the treatment and the people who work the Practitioners are who work in the general medical field while they're good They don't have enough time to deal with the subtleties of mental disorders and don't give a thorough diagnosis So we have perfect drugs we think or at least we hope we've been developing perfect drugs And then why does the treatment fail? Well, these are a couple of reasons one is as I mentioned a few times diagnostic accuracy and comorbidities Treatment effects. So you start with one drug add another one maybe without stopping the previous one. So there's polypharmacy Dosage length of treatment are all factors and then there's the environmental interactors Does the patient comply with taking the medications as prescribed? Does he adhere to all the things that are supposed to be going on? And then the fact is we don't know about external factors like trauma. So What is personalized medicine offer for you? Can we be clear what would happen? hopefully With understanding a genomic regulation Disease processes will be able to provide accurate diagnostics appropriate treatment predict risk or onset of illness predict response to therapy Support remission and cure of disorder and define etiology all of these things. We hope we'll come out of looking at the genes that regulate these illnesses But then you have to step back and say the genes really play a role in this and I think we would all agree that genes do play a role, but it's it's two different degrees in the illnesses So major depressive disorders I think there's enough data that says the heritability is about 37 percent schizophrenia 46 and bipolar disorder 40 Although at this meeting we heard a much higher figure talked about but we do have to identify the genes that put us at risk and We need to have objective tests to do this So I like to ask the question about how important is our DNA and obviously our DNA is extremely important in terms Of the way we express ourselves and when we express diseases But I wonder how meaningful the whole genome studies are for the diseases that we're interested in and so far They really haven't had a high yield And I think that part of that reason is that we have the complexity of diagnostics and co-mobility and What may be critical to measure is not whole genome but gene expression in the brain in particular areas at a particular time in the person's development Also, I think that as everyone else does in this room that all mental illnesses are multi-genetic and impacted by epigenetic effects making it even more complex, but We still have to come up with phenotypes. They don't have a choice and They need to be objective accurate and reliable They need to be such that they can be done on a large number of subjects And provide direct evidence of brain gene expression for an individual and Useful to predict treatment response as well as other aspects of disease One of the major problems we have is not only genetics complicated, but I think our brains are exceedingly complex So if you think about your own brain It has a map of the body. It has a map of the outside world and has a map of all of your experiences That is a lot of information that you're integrating using constantly In addition, the brain has a hundred billion nerve cells a million billion connections two million miles of wire And it's very tiny but highly connected and the way it's connected is extremely important to our function And this is what we're trying to understand how it works in a meaningful fashion and how it goes awry in disease And that complexity is huge and We don't really know the connectivity of all the nerve cells. We're starting to try to collect that information We don't know where genes are expressed in the brain at what times again was starting to collect that information so how can we move forward and Solve all the diseases we would like to solve without having all this information We're doing a pretty good job of trying to do that But we need to work harder at it and I think there are some ways we can do that This is a slide I put together a number years ago that reflects humans or animals birth to death cycle that we have and the ways that neuroscience or Anyone doing research on the brain tries to look at brain function And the question that I have for you is you could see here we go from single neurons the whole brain studies as well as from Genetics up to imaging and these all give you different levels of analysis But the question is how can we take all of these level of analysis and take the information from these different levels and Bring it into clinical studies. How many of these methods can we use in clinical studies? Not too many, but we could take the information that we lose use it Learn from these basic levels and move them into clinical types of questions that we could answer With greater efficiency and information. I think we're at a point where we have to think about doing a paradigm shift in the way studies are done and this I'm talking about clinical studies and On the left side, I have the way current studies are done on the right side I have what we what I think we need to do so currently most of the studies that are done are Small to medium one to two sites. We have to do multi-site large global studies We have to be sampling a bigger percentage of patients at different sites We all do our own unique protocols and this is good because it provides new information But to prove a particular disorder at its basis, we need to do standardized protocols. We need to do things the same way We all do our own assays and unique methods that we'd like to have answers to but we need to try to do some studies with standardized assays and assessment methods We have to move from subjective assessments to objective assessments in particular I think the diagnostics that we do is are too subjective and they need to be much more objective For brain disorders, we have to I believe even though we've heard some interesting blood measures these two days We need to go to functional measures of of the brain. So we need to have brain markers to understand what the brain is doing And also we need to move to a system where data data is available for all to look at and re-examine So you have database and we need to be able to do data integration across systems to understand what the interactions are No small task all of this So if we can define the phenotypes in this new paradigm paradigm, we would have objective standardized diagnostics and Objective standardized clinical signs and symptoms as Well as the environmental data All this is a very big task For functional brain markers. I can't include too many but right now what I would include in this is brain imaging Neurophysiological measures neuropsychological measures and as we can develop more measures I think we need to add them to this list So it was really pleasing to see how many people here reporting on measures of cognitive function This is a nice measure direct measure of brain activity and this is just a restatement of those in more detail and What I'd like to do now is talk about some recent activities that I and others have been involved in in terms of personalized medicine for the brain and Also translational research so in 2009 held a meeting in Washington To look at and ask the question is do we do we have enough information? From genomics to apply it to mental disorders and we brought together people that represented all of the Aspects of care in Terms of working towards personalized medicine. So we had basic research as clinical researchers health care companies we spent the day talking about this and Basically, we came up with a number of recommendations from the meeting which Say we have to integrate the knowledge that we have we need standardization of methodology Studies need to be done with real-world populations to meet real-world needs and we need large studies So harness the power of numbers The real issue is to take all the specialization and boil it down and focus it and integrate the information We have to come up with what a disorder is and how to treat it best There are some other activities that are going on to look at translational neuroscience and psychiatry One is from the NIMH 2008 strategic plan They are moving to develop research diagnostic criteria and this is looking more at biology Another one is an international study to predict optimized treatment for depression and ADH And I'll give you a little bit more detail on that On both of them. So the strategic plan that NIMH is developing is looking at biology And they want to develop research diagnostic criteria for each of disorders looking at direct measures of brain function That are observable and quantifiable and Hopefully if you do each of these measures across all disorders, we will come up with a way to compare disorders and and Put them into categories That are based on biology and that are not based on signs and symptoms So it's looking at the translation of basic functional dimensions Reflecting circuit level activities and behavioral activities and they're not limiting what measures can be used other than they have to be Generate they have to be different units of analysis that can be independent variables generating Classifications from basic behavioral neuroscience Rather than starting with an illness definition and seeking its neurobiological independence We really want to start with the neurobiology And this is a mock-up of what the research domain criteria would look like And it's a construct of looking at behaviors which are listed on the left going across genes to Circuit cells molecules self-reports and it as this gets filled out one is looking at a Documentation of disorders that can then be looked at more significantly with genes and treated better In the icebox study, they've taken an approach of standardized methodologies and protocols and the study includes a Slightly over 2,000 people at 20 sites around the world in a treatment study with three different drugs for depression and they're sampling genomics and doing magnetic resonance imaging functional magnetic resonance imaging and DTI and Looking at EEG and evoked potentials and autonomic arousal Measuring cognition and other psychological tests and personal history and the study is also incorporating the Regular measures of diagnostics, so you can look at DSM criteria ICD criteria But what we want to see is more like what our doctor is looking at What's the basic biology of these disorders and can you come up with biological? tests that Segregate together to say that an individual will do best on this type of drug Recently did an edited book on integrating neuroscience and personalized medicine And we have about 15 20 chapters in there where we've asked people to focus on what are the genomic and physiological and psychological behaviors that are associated with any disorder and Well, it was Educational to do this book It's also controversial in the sense that some people look strictly at genomics Measures other people try to combine lots of measures to look across disorders But nothing right now is really ready for personalized medicine, but we have to move it in that direction And most of the problems that you see and that are reported here are small studies non reproducibility And we have to get to the point where we can test All of the leads that are there where there are many but they need to be replicated So if we want to move to personalized medicine, there are other issues other than basic understanding of an individual disease We need to establish Electronic health records that are functional and informative to the treating physician We need clinical decision support systems So this means how does the clinician interpret the record of the person's history and new information that we learned about Classifying disorders and what's the best treatment and how do they use my biomarker information? there's other issues that go across the reimbursement and People are suggesting to have perhaps have financial methods that include paying for performance outcomes Not just clinical service delivery. So to summarize I'm a strong proponent of saying that new studies need to include functional brain measures and These could be seen as surrogate gene expression markets And we need real-world samples of patients including those with comorbidities We need standardized objective methods and protocols Databases and data sharing we need more than one therapy to be tested in every situation There has to be a replication design within the within the study a Statistical and analysis plan and we need to report sensitivity and specificity of measures The one would think that all these things are always done But if you look at many publications, they're not there obviously the patient will benefit from all of this and Will inform us about impending disease will have accurate diagnosis Effective treatments treatment responsiveness and also prediction of side effects and we could actually say when we have a disease remission or cure Those are the ultimate goals So is personalized medicine all good? I think there's a lot of issues out there that we have to be careful about and concerned about as we move into this area somebody we know very well who's here made this statement back in 2004 that covers both the area of personalized medicine and translation research Julia put in one of his editorials and molecular psychiatry as the immense progress has been made in basic research The challenge is now translate fundamental discoveries from the area of pure science the reality of health care Leading to better preventive approaches more refined and successful treatments and improved health care I mean he's really saying their Translation of research will lead to personalized medicine Those are not the words that he used So when do we do this? One would have to say that we need to be doing it now. We are starting to do it now and As this graph demonstrates if we do it We can collect the information that we will save an awful lot of Costs in terms of our efforts to treat and cure diseases Interestingly enough this Francis college at the NIH and I think I don't know for sure But when I said that we started talking about translational neuroscience. It was in the late 80s and 90s. We did this at NIH And now this is being Revisited again there where they are going to have a national center for advancing translational sciences He talks about establishing a center for translational medicine So we are revisiting it even though it's been claimed. We should be doing it all along and Hopefully this will add Enough money somehow us 650 million for this effort doesn't seem to be enough, but it's a start But it's a revitalization of looking at translational sciences and lastly I think that you know the couple of books that I published in the past few years deal with Databasing and And personalized medicine and I think the two really into date for interdigitate very well And we do need a database all our information and leave it open for others to look at and That's my message. I hope that I can convince some of you to think about it and that you would Let me know where I'm thinking about it the wrong way and better ways to improve it But how can we come up with solutions to treat people who are Have illnesses of the brain and do it in a better or more efficient way, and I thank you for your attention