 Thank you very much for that introduction. It's a great pleasure to be here, and I particularly want to thank Julia for the opportunity to come and talk to you about the work that we're doing, which really dovetails beautifully with the notion of translational psychiatry. And what I'm going to try and do in the course of this lecture is cover three areas. The first is some of the background evidence for the fact that there is a progressive process of neuroprogression in bipolar disorder, and for that matter, many other disorders. Secondly, touching on what we think are some of the key biomarkers underpinning these changes. And thirdly, for us most excitingly, what new therapies might these point to? So for a long time, we've struggled with this black box of pathophysiology, not really fully grasping what lies inside the black box. We know of a whole lot of risk factors that go into the development of psychopathology, and out-pop a variety of syndromes, but we really are struggling with understanding what lives inside this box. In terms of bipolar disorder, and in terms of the progressive process that exists within bipolar disorder, a critical message is to understand that people with bipolar disorder really do start off their course of their disorders firing better than their peers. There are a number of studies that show this, and this is one that was published in the British Journal last year. And what you can see is that kids who develop bipolar disorder and have a standardized grade average, in this case looking at more than at the top end of academic performance, they have an odds ratio of 3.3 for being in the brightest group. Quite unlike the scenario with schizophrenia, these guys start off firing ahead of their peers, but that changes. And there's a consistent picture of evidence that points to the fact that there's a progressive neuropathological change in the disorder. The first comes from as early as Emil Kraplan, who showed that with each episode of illness, the gap between the episode that you've got and the next one that's coming gets shorter and shorter and shorter. So there's this accelerating cascade where illness seems to beget a greater vulnerability to subsequent episodes. The next clue that we had, that there's a progressive process, comes from work of Swan and others, just focus on the white curve. And what you can see is that lithium response runs fairly flat in the early phases of illness. But once you get over about 10 or 12 episodes, the likelihood of you responding to lithium goes off the cliff. So after about 10 to 12 episodes, you cease to respond to lithium as effectively as you did in the early phases of the disorder. And this has been replicated, another study which shows that if you catch bipolar disorder in the first five years, the likelihood of you responding to treatment is very much greater than if you catch it, for example, after 11 to 20 years, or greater than 21 years. Even in psychotherapy literature, there's evidence that stage of illness matters. So this is a well-known paper by Jan Scott, which essentially showed that CBT didn't work for bipolar disorder, but buried inside this data for us was a little treasure. And this was it, that if you look at treatment response by a number of episodes, focus in on the left-hand side of the graph. And you can see that in those people who had between one and six episodes, you had a clear separation with CBT and treatment as usual. CBT really helped. But in people who had more than 30 episodes, they clearly deteriorated with CBT. So stage of illness not only predicts response to pharmacotherapy, but seems to predict response to psychotherapy as well. And we just published this paper looking at atypical antipsychotics. And again, we looked at maintenance treatment of bipolar disorder. We looked at people who were given a lanzipine. And what you can show is that the people who had one to five episodes consistently did better than people in the other groups. And in the relapse to depression, if you look at the bottom part of the figure, people with six to 10 episodes did better than those who had more than 10 episodes. So again, showing evidence that treatment response just declines with the chronicity of the process. So this is the Kaplan-Meier curve from the same study, again showing that those people in the solid line who had one to five episodes clearly did better than those who had greater numbers of episodes. And we've now got new data. This is now unpublished from the STEP-BD study. Now the STEP-BD study, for those of you who might not be familiar with it, is a very large, non-randomized study following up people with bipolar disorder. And this is now, because it's an NIH-funded study, the diet is in the public domain. And so we were able to divide up people using the same broad methodology into those who had one to four in the yellow, five to 10 in the green, and more than 10 episodes in the blue. And what you can clearly see on Madras in the top left, the people had the fewest episodes, had the lowest depression symptoms. On the top right, you can see they had the lowest manic symptoms. And on quality of life and functioning, you saw the exact same pattern, that the guys who had the fewest number of episodes were doing best. And we also had the intriguing opportunity to look at antidepressant response in the study. Again, one of the very few papers in psychiatry that gets published in the New England Journal of Medicine was the antidepressant study in STEP-BD, where they looked at, did antidepressants work for bipolar disorder? And the answer from that study was no. But in the data mining that we've done, we've come out with an intriguing clue. So if you look at those people who've had one to four episodes, look at the top of it. People who are in the early stages of the disorder seem to have a transient response to antidepressants. So they have a response to antidepressants that is evident between about weeks four to eight and nine to 12, and then disappears. So it's early. It only occurs in early illness. And it's transient. It doesn't last. Together with that, you get an increase in manic symptoms. But once you get into five to 10 episodes and more than 10 episodes, antidepressants seem to be of no value. And again, this is the kind of data I like because it really accords with my clinical impressions. So Jin showed you some neuroimaging data. And again, I don't want to go into the details of it. But again, the picture is fairly clear that there appears to be a degenerative process that people in the first episode seem to have relatively preserved ventricular volumes. But with progression of the illness, one gets ventricular enlargement and loss of brain tissue. So again, we've got clinical evidence. We've got neuroimaging evidence of a progressive process. This is the slide that Jin showed. I didn't know he was going to show it, so I won't show it again. But I do want to show you this neuropsychology slide, which is a truly elegant little study coming out of Columbia. What they did is they looked at first episode patients, second episode, and third episode patients. And if you just focus in on the left-hand side of the figure, where you can see it quite clearly, in fact, I suppose you can see it fairly clearly across the figure, you can see that the first episode patients in the blue are doing better than the second episode patients in the red, who are doing better than the third episode patients in the green. So you see this progressive decline in cognitive functions with number of episodes. So there's again this beautiful, consistent picture of a progressive process. There is hope in the sense that there's evidence that lithium-treated patients, lithium-responsive patients, may not show this cognitive decline. So this is data by Yanis Rubikowski, showing that excellent lithium responders seem to have relatively preserved cognitive functions. So there is a suggestion that we might be able to, in some ways, arrest this neuro-progressive cascade. And there's evidence also from other data. This is from the dementia literature, showing, at least in community-based studies, that people who take lithium, if you look at the first column on your left, have a reduced odds ratio for the development of dementia, which is not shared by anti-convulsants, anti-depressants, or anti-psychotics. Again, intriguing data. So let me come to the second bit of this talk. And that is very briefly what might be the processes underpinning this neurotoxic cascade, this process of neuro-progression. Well, one of the key components we think is oxidative stress. And I'm showing you just one for illustration, because one can't show all the data. But this is a meta-analysis of lipid peroxidation, measured using an assay called T-bars. And you can clearly see a very consistent picture that people with bipolar disorder have elevated levels of lipid peroxidation compared to controls. And this is now a very well-replicated finding. There are many other measures of oxidative stress, which for the reasons of time, I can't show you. There's also post-mortem evidence that individuals who have bipolar disorder and for that matter schizophrenia have elevated levels of lipid peroxidation. So this is post-mortem data looking at lipid peroxidation in the singulate. And you can see significantly elevated levels of lipid peroxidation. So where the hell is this coming from? At least one of the sources whereby lipid peroxidation might be occurring is abnormal mitochondrial energy generation. And many of us are now beginning to believe that at the core of bipolar disorder lies aberrant mitochondrial energy generation. And this is one slide by collaborator Tina Andriesa who showed reductions in mitochondrial complex one activity in people with bipolar disorder. It's not yet been replicated, so this is just an intriguing new clue. And we know that chronic mild stress decreases mitochondrial function. So if you take an animal and put them in a chronic stress paradigm, not only do you see reduced neurogenesis, as explained beautifully by our previous speaker, but you can see increases that stress decreases mitochondrial oxygen generation. Lithium seems to up-regulate mitochondrial complex one activity, particularly complex one activity, which is the complex which is predominantly down-regulated as a part of the disorder. So again, this, of course, does not prove anything, but it's a really interesting correlation. We've been interested in glutathione. And glutathione is the brain's primary free radical scavenger. And there's data that it's significantly reduced across psychiatric disorders, so in schizophrenia, major depression, and bipolar disorder. Glutathione is the body's predominant defense against free radical damage. And not only is this relevant to psychiatric disorders, but it's clearly relevant to other disorders such as Parkinson's disease. And we've been interested in NS-style cysteine. Jen showed you, in his presentation, a schema showing that NS-style cysteine is a precursor for glutathione. And there's now evidence that NS-style cysteine has an impact on neurogenesis and on neurite extension. So this is a study from the plastic surgery literature showing that NS-style cysteine enhances neurite extensions. And what these authors also showed was that NS-style cysteine increases the survival and regeneration of sutured peripheral nerves. So if you just eyeball the two blocks in D and G, you can see a significant increase in regeneration of nerves with NS-style cysteine, showing that it might be potentially neuroprotective. So on the basis of this, we went ahead and did a couple of large-scale clinical trials. The first one that we did was in schizophrenia. So we took 140 individuals, treated them for six months, and gave them NS-style cysteine or placebo, in addition to treatment as usual. And what you can see here is a significant effect of NS-style cysteine on many of the core symptoms of schizophrenia. So we saw an effect size of 0.43 on CGI. We saw an effect size of 0.52 on negative symptoms. This was an effect size of 0.57 on PAN's total symptoms. We also saw a reduction in extra-paramidal side effects. And there's fairly good evidence now that there's substantially increased levels of oxidative stress in people who have tartar dyskinesia. Please start to let NAC might have antidepressant effects. So this is preclinical evidence showing that NS-style cysteine reduces immobility in the forced swim test. The forced swim test is the industry standard preclinical paradigm for determining antidepressant efficacy. And you can see at the dose of 50, NS-style cysteine has comparable levels of antidepressant efficacy in this model to imipramine. So we went on and did a clinical trial in bipolar disorder. This had a sample size of about 74 individuals. Again, a six-month study. Again, it was placebo-controlled. Again, it was add-on to treatment as usual. And we were surprised to find even larger effect sizes. So we saw on the Madras, which is our primary outcome, we saw an effect size of 1.4 on bipolar depression rating scale. The other symptom of depression, again, large effect sizes. And on measures of functioning and quality of life, we also saw large effect sizes. So we saw 1.1 on the Rift, 0.93 on the Slice Life. And on quality of life, we saw an effect size of 0.95. So again, this for us highlighted the utility of the translational paradigm in starting with a biomarker of interest and seeing if we can try and modify this to come up with potentially useful treatments. We've now just completed a maintenance study. And I'm afraid it doesn't look quite as good, largely because I think we were underpowered and because of the placebo response. So let me just talk you through what we found. So this is a study we did together with Jen Marley as our primary collaborator. And we gave everybody open-label NAC for a period of eight weeks. And you can see a robust decrease in depressive symptoms in the open-label treatment. The problem is, having got down to very low levels of depression, nobody got ill again. So people in both the placebo and active groups stayed well. So the groups really did not separate. And if we look at survival to depressive episode, the visual, the trend towards superiority of NAC was not statistically significant. And again, we just didn't get have enough events. So even though we had 154 individuals in the study, only about 14 in each individuals had a relapse. So we were really underpowered to show an effect. Let me change gears to the other biomarker of interest. And that's cytokines and other inflammatory markers. Now, there's now a wealth of information showing that people with most major psychiatric disorders, but particularly mood disorders, have elevated levels of inflammatory cytokines. And this is now a well-understood area, and people like Bernard Bowen, amongst others, have contributed to this literature. In bipolar disorder, you see a very similar pattern to what we see in depression, that both in the manic and depressed phases of the illness, you've got significantly elevated levels of cytokines. What's interesting is that this seems to predate the development of illness. So this is data from our epidemiology study, where we followed up 1,000 women for 10 years. We measured cytokines at baseline, and we saw, did cytokines predict who becomes depressed? And here you can see that those people who had the highest levels of cytokines in the green curve had the greatest risks of developing a daynova episode of depression. So nobody had ever before been depressed. And you can see that the people with the lowest levels of cytokines had the lowest risks of developing a first episode of depression, and the people with the highest levels of cytokines had the highest risks of developing a daynova episode of depression. So this motivated us to try and look, well, is there anything you can do about this? Before I get to the treatment, I wanna just show you that these biomarkers really do additionally differentiate stage of illness. And to my knowledge, this is the first data that supports the staging model that Ian Hickey and Pat McGarry have developed. So what you can see is that BDNF is normal in early stage disease, but BDNF starts to drop in late stage disease. So you start to lose your BDNF as the illness progresses. You can see that interleukin-6 and TNF-alpha, the inflammatory cytokines, are elevated both in early and late disease, but the anti-inflammatory cytokine, IL-10, is elevated in early stage disease and falls off in late stage disease. And the hypothesis is that you get loss of your anti-inflammatory compensatory mechanisms as the illness progresses. And you can also see clear changes in the glutathione system with progression of the illness. So let's get back into translational gear. What can you do about this? So, Neri and colleagues looked at the cox inhibitor, selicoxib, as a treatment for bipolar depression. It's a small study, it's underpowered, and it was significant at week one, but not at week six. But it's still an intriguing clue that if one gives agents that are active at inflammatory pathways, you might be able to do something interesting in terms of the management of these disorders. There's an intriguing study came out last year on aspirin. So this was only published about a few months ago, showing that if you give aspirin to people with established schizophrenia, you get a significant reduction in symptoms. So this is not only in PAN's total symptoms in the top left, but particularly in PAN's negative symptoms in the bottom left and PAN's general in the bottom right. So again, who would ever have imagined that aspirin might have utility in psychiatric disorders? But now we're looking at this possibility. The other anti-inflammatory therapy that we've been really interested in is statins. Now, apart from lowering cholesterol, statins have really robust effects in lowering cytokines. So this is relatively old data showing that pretty much all statins lower CRP and other inflammatory cytokines. So we were curious to know, do statins have any effects on risk for psychopathology? And the answer appears to be yes. So again, in our epidemiological study, we looked at people who were or were not on statins at baseline and looked at their risk of developing de novo depression over a 10-year period. And you can see a substantial reduction in the risk for development of de novo depression in people taking statins. The odds ratio in this study was 0.2, indicating an 80% risk reduction. We were really surprised by these data and we wanted to replicate them and we did. So we had another study that we'd completed a while back and realized that we had the data for this. So this was a prospector study following up people who had had a cardiac event. And some of these people were or were not given statins and we looked at, if you got statins, did this influence your risk of developing depression? So at three months, if you took statins, you had a reduction in risk of 69%. And at nine months, you had a reduction in risk of 79%. And there were a whole lot of other factors that were covariate in this model and this held up. I need to emphasize, of course, that this data, as intriguing as it is, does not equate to randomized prospect of clinical trial data. We need to have that kind of data to show that there might be clinical utility. But I think it's an intriguing clue and I think it talks to the validity of the translational model. And there are a whole host of other biomarkers that we're interested in. So this is data looking at leptin and again, using the same design, your levels of leptin seem to predict whether you're gonna be at risk of developing de novo depression. I'll pass on that one. So to summarize, we have a host of potential neuroprotective agents. Some of these are relatively old agents, like lithium, but there are a number of agents acting on a range of novel pathways that might be neuroprotective, that might address biomarkers of interest in these disorders and might lead to the development of novel therapies. So to finish off, we know now that we, there's pretty good evidence that we have got a process of neuroprogression where you've got a accelerating and progressive disease process. There are a number of biological process that seem to underpin these, including dopamine and glutamate, mitochondrial energy generation, oxidative and inflammatory processes in neurotrophins. And we now understand that these are involved in the activity of conventional agents, but these open the door for the development of novel therapies. So I'm gonna finish at this point. Thank you for your attention.