 Thank you very much Scott for the very generous welcome and thanks Julia for the invitation to speak. The work I want to talk about is an area that I've been thinking of moving into for a number of years and it's interesting that around the world when we started this I became aware that there's a small number of groups doing exactly the same sort of work that have just begun around the same period as ourselves and I think this comes out of both my intrigue with the mechanisms of illness but I'm also a clinician and the awareness that at the moment we have no way of identifying early individuals who at high risk of developing bipolar unlike the schizophrenia programs and I'll use that as the analogy we can only intervene when people develop clear and robust illness so really the translational aspect of this area of research is that hopefully this work being undertaken around the world will inform early identification and intervention programs so the question really I suppose just put it in other words is what factors determine which children of those with bipolar will go on to develop the condition and who'll be resilient to this and while we're looking here at a very focused and enriched sample a genetically defined sample this also has the potential to influence broader practice about early identification of bipolar disorder and as I was saying before that elucidation of such factors would inform because at the moment it's it's very much guesstimate and clinical intuition about who you intervene early in a seriously considering has bipolar I want to at the beginning go over a number of the aspects of where the genetics is up to and some of the points have already been made by Brian just so we can put in context this very specific high risk strategy that we'll be looking at for those of you that aren't aware like schizophrenia bipolar disorder very strongly genetic disorder most studies show that the heritability is in a range of 80 to 90 percent the those sort of delightful Scandinavian data sets that we all salivate over have has looked at heritability cross populations and shown that first degree relatives of individuals hospitalized with bipolar disorder have 10 to 14 times the population risk of developing the condition so all the epidemiological evidence confirms that this is a very strongly genetic disorder the question in recent years has been what's the mechanism as Brian saying as for schizophrenia the predominant means of inheritance appears to be these common genetic variations you get a critical mass or aggregation of these common variants and you get the illness at the moment we don't know how many as with schizophrenia that appears this is the main mechanism Brian was talking about copy number variants or structural variants the moment it's not clear whether that plays a role in bipolar if it does it's certainly not as common a mechanism as in bipolar you saw this before and I'll just point out the relevance to our story here this is looking at replication of the large international schizophrenia consortium data set and you can see here that Brian pointed out the replication and other schizophrenia data sets including his own but what's striking is the fact that this also replicates in two independent bipolar disorder data set so schizophrenia common gene variant pattern replicating in bipolar but not in other physical disorders like coronary artery disease Crohn's disease hypertension etc so there appears to be from several strands of evidence major sharing between bipolar and schizophrenia the other strand of evidence that's really clarified that story comes again from Scandinavia where they look at population data and the essential story here is that they looked at individuals with a number of hospital diagnoses of bipolar looked at the population data sets how many their relatives had bipolar and as you'd expect higher than population rates but what was striking is when they looked at those relatives they found higher than population rates of schizophrenia they did the opposite they looked at people hospitalized with schizophrenia looked at the first degree relatives as well as fine finding higher rates of schizophrenia there were higher rates of bipolar disorder so this is a graphic from that paper and you can see here these are shared genetic effects as the paper of Lichtenstein that was published in the Lancet about a year and a half ago so that there are there is major sharing in terms of the genetics between the two disorders it's not saying they're the same conditions as some had interpreted in some some of the media discussion you can see there are also unique genetic effects so what we see is that there are shared genetic origins between these two conditions but distinct genetic origins which account for the clinical differences what genes are involved the story here is a little bit behind schizophrenia but catching up very quickly it's interesting that the schizophrenia research community were much better at collaborating than the bipolar disorder community but I think the story is forcing people to come together this was the most recent data presented at the World Congress on psychiatric genetics and Athens late last year numbers are now just staggering I mean these are the biggest studies ever in psychiatric research so 12,000 cases with bipolar predominantly bipolar 152,000 controls and the amongst the loci coming up the two that appear to be the strongest one the calcium channel subunits gene coding for that and also for one of the the genes involved in sodium channel regulation so ion channels are looking an interesting story so it's starting to become apparent what some of the specific genes are as well as this being strongly genetic what about some of our own work and this is also a crowd lot of interest we were invited to contribute to a European consortium the study was only published four weeks ago by Stephen Kitchon from Germany from the Mannheim group and this study demonstrated and associations as a large association study about 9,000 individuals with bipolar about 12 or 13,000 controls and you can see that this gene is really interesting gene for bipolar neurocan is a gene that the protein is a glycoprotein that's involved in cell adhesion and cell migration so very intriguing prospect in terms of abnormalities in neurodevelopment in bipolar disorder and there's a lot of interest in seeing whether this is turning up in schizophrenia as well this was just a methodology there was an initial large German case control study which showed a number of interesting possibilities there was a European replication sample and this held up after the European replication sample the journal asked for further replication so as well as a number of American and English samples were invited to contribute to this and you can see here that once once all these groups are combined that there's still a significant effect here it's 1.2 and as Brian was saying these are small effect genes these are not sort of powerful robust genes in and of themselves we're looking at interplay between these and what also makes this intriguing is that animal work has indicated that it's only expressed in brain which makes it a substantial interest and it's expressed and this is with some of the animal studies involved in this paper indicating it's involved in areas of interest to us in psychiatry for example near cortex near cortex and basal ganglia areas so an interesting one to see with really strong evidence for that as Brian was saying in schizophrenia that despite the growing knowledge that we can only account for about two or three percent of the ideological variance at the moment so the big challenge of the field is what's the nature of the missing heritability as it's being turned the missing or hidden heritability one of the possibilities is that you can't demonstrate the effect when you look at individual genes in isolation but you can demonstrate it when you look at interplays between genes and I'll come later on because one of the other issues is is the part of missing heritability the interaction with the environment and that's actually a much tougher question to answer in many ways so this was a paper we published in biological psychiatry last year the specifics aren't important but the methodology is that now we've got the statistical capacity to look at interplays between genes is very demanding statistically and what we found was that we looked at individual genes on their own that were associated with illness and then we looked at the interplay so this is the chromosomes like a clock the the black bars here are individual genes that came up associated with illness and most of the ones that we demonstrated were actually had been reported by other groups which gave us some confidence there was a validity to the areas that we demonstrated but the grey lines are the significant interactions after statistically correcting for multiple gene interactions so it gets very tough in terms of the level of statistical significance required and you'll see here the number of these are between individual genes that had demonstrable individual effects but you can also see that some of them are with areas that on their own didn't come up that weren't sort of on their own strong enough to be demonstrably associated with illness so we see here that and if you understand this is probably biochemical pathways if you have two pathways hit then the illness becomes apparent at that point so we see here that some genes won't become obvious until you start looking interactions so various levels of complexity but the story is starting to become more obvious in terms of some of these stories so let's get to the main game so we're saying before that I become intrigued about can we demonstrate factors features be they clinical or biological that might alert us that this is an individual who is at particularly high risk of developing bipolar disorder it's a methodology in medicine if we can demonstrate risk factors that has shown its benefit if we look at public health studies for example with cardiovascular disease have been major reductions in in mortality in morbidity with stroke and myocardial infarction by identification of of significant risk factors and early intervention programs and so there's been a major impact when death rates due to stroke have reduced substantially within our community by this methodology so how do we identify what our predictors and risk factors for mental illness with schizophrenia schizophrenia has the the advantage and benefit of having a reasonably well-defined prodrome in bipolar though people talk of prodrome there's really no clear prodrome that's been demonstrated now look at some of the studies so that we're stumbling around in the dark a bit more in terms of current knowledge well what do we know about the psychiatric illness amongst the offspring of people with bipolar disorder I'd give a few of the seminal studies there was a meta-analysis published in the late 1990s looking at rates of illness in the offspring of individuals with bipolar and in those studies they showed that the there were rates amongst the various studies of bipolar sort of ranging from 4% to 15% now most studies show population risk of about just over 1% so we're looking at 4 to 15 times population risk compared to what you'd expect in the control population of 0 to 2% individuals are also almost three times more likely to develop any mental illness and four times more likely to develop an affective disorder and that's bipolar or depression it was a large study published just two years ago by Burma Ha and that was the biggest study that's been undertaken of 388 offspring of individuals with bipolar compared to the offspring of 251 controls and this showed and really confirmed very strongly the increased risk of bipolar disorder in these relatives I found an odds ratio of over 13 mostly bipolar 1 and bipolar NOS has actually been debate about some of the diagnosis of bipolar in that as to how liberal that was even though it appeared to be more DSM based what about following through because to me the there are two essential questions here one is when you get a group of offspring compare them to controls can you demonstrate differences at that point are there any baseline differences so that's the first question for the second question in many ways is probably more interesting one if you follow people over time can you demonstrate or tease out predictors of those who convert to bipolar disorder so let's have a look at the literature that's looked at clinical predictors of outcome because to this point in time there've been no studies that have involved biological markers one of the larger studies was undertaken Stony Brook the Stony Brook high-risk high-risk project in New York and they followed 134 kids over time and found it compared to controls there were higher rates of a range of disorders and I think this is an interesting story we'll come back to when we look at some of our own data that they had more behavioral attention or problems and more psychopathology in general however the question that came out of this is how specific are these findings to families with bipolar disorders when they looked at the offspring of those with other disorders like schizophrenia and depression there were no major differences between the groups there may be some non-specific effect of growing up in a family with illness another study that looked at neuropsychological predictors study of Maya and followed up a relatively small sample but over a long period of time and found that those with abnormalities in executive function the Wisconsin card sort test predicted they were the group that were at much higher risk of developing bipolar disorder Janice Aiglin some of you would have come across her name Janice was the first author in the first molecular linkage study of bipolar back in the late 1980s it was interesting as it was that study that really inspired me to get involved in genetics research didn't realize the story would have gone on this long since then since the late 1980s but she looked in the Amish population and followed up children of the Amish who developed bipolar and found a sort of broad pattern of abnormalities that were more common than in those with the control families that they had more mood, lability, low energy, anxiety, a number of attentional problems and school difficulties they didn't follow through to conversion and that study like most didn't have an illness control group most studies don't I think that's a limitation that's there for a number of these. So where's the field up to at the moment? We started our study about 18 months ago when I got involved in this was discussing with some US colleagues who said well look there's a similar study that's just started across the US and we've eventually agreed to collaborate so now we're sort of inactive collaboration and share much of our data set with that sample. So that's the study being led by John Nernberg of Indiana John very long track record in genetics of bipolar and other disorders. There's our group we're looking at 12 to 30 year olds. There's a Canadian group that's published in clinical work and Duffy. There's a Scottish group and I'll I'll show some preliminary findings from the Scottish group because I'll talk about some of our imaging results and also talk about some very recent results coming out of Andrew Macintosh's group from Edinburgh. Michael Bauer's group in Germany are looking at cognitive therapy as a means of preventing the development of bipolar. So rather than looking at prospective features they're going straight into a randomized prevention trial. So we're collaborating with the US group between us we're sharing our data set for the 12 to 21 year olds. One of the big issues here is numbers particularly when you're looking at predictive capacity for future illness development. So in the joint sample there are now 290 at risk and 170 controls. So this is now a substantial group. We share completely our clinical data set and we got blood for genotyping in these samples. We've contributed about 60 of the 12 to 21 year olds a major contribution to that. But I'll focus on the work we're doing here in Sydney. We've engaged the graphic artists who come up I think with this very catchy logo catch them before they fall and I think that really has captured the essence of what we're on about that really the whole point of this research is to pick up bipolar before we get into the devastation and some of the work that we've seen through I think some Michael Burke's work indicated that the worst outcome in people with more illness. So if we all have this hope and belief I think it's not perhaps as well substantial as we'd like that if you can get in early you can improve outcome. So we went in probably a bit ambitious after 500 but we're aiming for 200. I think that'll be very reasonable by the end of this year as well as control. So we call it the bipolar kids and Sibs study if you see something about that in a minute. So I said the same before there's two components. This one to look at factors at baseline that are different between the two and then to follow up over time to look at what are the factors that predict who converts to bipolar. So very quickly just take you through the method and look at some of the results because I'm aware that time's moving on. We're doing a very detailed assessment. So people come in for a whole day and any of you that are interested the families are very keen on this. I wondered how this would go. But the families the bipolar families really very worried about their kids. So we've flown people into Sydney from all over the east coast of Australia. It's been one of the most expensive studies per capita I've ever done. But our recruitment really has been going very nicely. They come in for a whole day. They have a very detailed clinical assessment. They have a very detailed imaging protocol with structural functional fMRI. I'll talk about those in a minute. We've done DTI diffuse intense imaging and also resting state. We're doing genetics. We had a benefactor approach as recently who's given us money to do epigenetic studies. One of my hypotheses has been that perhaps those at risk of those who have more childhood stress and epigenetics may be a way of getting a handle on the biochemical results of that and also a battery of neuropsychological testing. The moment we've got 105 kids. We've done a complete assessment on 85 controls and we've also brought in a bipolar arm. So there's an illness control arm. These are people with bipolar in this same age group. And at the moment we've followed up over 20 and we've had clinician diagnosis of three. We're chasing those up that appeared to have already converted to bipolar, which is encouraging that we got a sample which is going to convert at an acceptable rate. To demonstrate predictors you've got to have enough converting to the illness. Just to give you a feel for the cohort. So while this isn't earth-shattering it's given us some confidence validity of the sample. So you can see here and this focuses on the 19 to 30 year olds because we've got the best control group for this. That there's a much higher rate of illness. This is lifetime illness and this is not just definite but it's also probable. So that's why the rates might seem a little bit inflated. So you've got to be aware we've used the NIH best estimate criteria and these are the definite and probable. And you can see here that those at risk have six times greater likelihood of having any psychiatric diagnosis as childhood or adulthood illness. This is 19 to 30 year olds. When you look at the rates of affective disorder the odds ratio is 6.9 for those who are offspring. So this is a group with high morbidity and it's very striking when you actually see this happening sort of in your own group. I won't go through this in too much detail. We haven't had anyone as yet I was saying we've had a few clinically diagnosed but at the point of interview had bipolar NOS but the major affective condition you see is depression in these samples much higher rates, strikingly high rates of anxiety disorders and also behavioural disorders. It's a very broad picture of psychopathology. I'll just take you through some very quickly some of the results that are starting to come out. This is work coming out of the US sites on one of the co-authors in this group in this paper and you can see as you would expect that those in the at risk families have an earlier and greater rate of affective disorder particularly depression. What's interesting and starts to give us some hints about clinical presentation is that those and this is only in the bipolar families not in the control families that in those who are developing major affected disorders at the moment that's mainly depression. There are high rates of anxiety disorders so maybe anxiety is a prodrome in these particular families. I want to give you some very brief results from the imaging. We've got here the results for our 19 to 30-year-olds because we've got good comparison group and also this avoids some of the developmental problems when you're looking at adolescents and combining different samples. This work has been mainly undertaken by my post doc Gloria Roberts and Michael Brakespear has been going through the statistics and confirming that. We hypothesised that there were abnormalities or dysfunction in emotional regulation in these families. We thought that was a reasonable working hypothesis or something that might be predisposing to illness. So we used a facial emotion go no go task that had been used in people with established illness. This is a task that activates the inferior frontal gyrus and the anterior singlet cortex in some studies. The particular task where we found abnormalities was inhibition of response to fearful faces and a very striking result when Gloria first told me about this I said there must be an artefact or problem here. Michael's gone over it and convinced that really this is a very solid finding and that on the whole brain analysis there was a reduced activation with inhibition to fearful faces in the inferior frontal gyrus, an area which clearly is of interest in terms of mood disorders. This is a whole brain analysis and survived multiple correction. So a very strong finding. This is consistent. There was a recent meta-analysis in bipolar disorders and the summary statement of that was that there were two areas that came up in the functional imaging tasks in people with established bipolar in the main areas with inferior frontal gyrus and ventral lateral prefrontal. So it looks as though that we're picking up the same area that's been picked up in a number of the studies with people with established illness. The only, so ours is the only functional study that's been undertaken in a high risk population because it's so few of those and we're the only group actually doing that. There's a paper looking at diffuse intense imaging that was published just three weeks ago and it's in press in bipolar disorders by the Scottish group and then demonstrated abnormality in with diffuse intense imaging in white tracts in the brain of people with bipolar disorder at risk for bipolar disorder. Like our sample, none of these have established bipolar disorder at the moment. So both at risk samples. So both studies, ours functional which we're writing and preparing for publication at the moment submission for publication and the structural study of Andrew McIntosh suggests that you can actually demonstrate in an at risk sample imaging abnormalities. I think this is really a new way of looking at the illness and I was not sure whether we'd be able to demonstrate this. So whether you're looking at structural or functional, there appear to be demonstrable differences between an at risk group and controls. I think I probably made most of this points. We're looking at correlations with genetics to see if there are any obvious genetic determinants of these functional findings but it's suggestive that there is a predisposing or pre-existing difference in how people at risk to bipolar actually deal with emotional challenges. So where does this take us? I'll just finish off with acknowledgments to colleagues. I doubt that it's going to be one factor that is going to be clinically applicable. I suspect we're going to be looking at an algorithm of clinical imaging, perhaps genetics, maybe neuropsychs. I think this is a very interesting story and I think looking at our translational focus that these sort of studies really have the capacity to influence future service development. So thanks very much, Mr. Jay. Good.