My name is Nitin Gokte, I'm the Chief of Research and Deputy Medical Director at the American Psychiatric Association, and welcome everybody for this morning's Gralnick Award Lecture. It's my honor to introduce Dr. Dost Angur, with this audience probably he does not need any introduction, but he's currently the Chief of Division of Psychotic Disorders, responsible for three inpatient units, residential facility, a PAC team, two specialty outpatient clinics, and a community support program, along with maintaining an active research life. In addition to his clinical work, he receives, you cannot hear? Is this better? Yeah. Sorry about that. So Dr. Angur is the Chief of Division of Psychotic Disorders, responsible for three inpatient units, a residential facility, a PAC team, two specialty outpatient clinics, and a community support program. In addition to his clinical work, he receives funding from the National Institutes of Mental Health and other sources for his research on the clinical manifestations and neurobiology of schizophrenia and bipolar disorders. Dr. Angur is currently the William P. and Henry B. Test Professor of Psychiatry at Harvard Medical School, and has authored more than 250 articles on the research into the neurobiology of bipolar disorder and schizophrenia. The Schizophrenia and Bipolar Disorder Research Program led by Dr. Angur was founded in 2006 and has grown greatly in scope and depth since then. The long-term goal is to develop new treatments for patients in the Division of Psychotic Disorders, including medicines, computer programs, and early intervention strategies. He has won awards from McLean Hospital, Partners Healthcare, which is now Magendal-Brigham, and Harvard Medical School for his teaching of medical students, residents, and his mentoring. He also serves as the editor of JAMA Psychiatry, a premier journal in the field of psychiatry. It's my pleasure and honor to welcome Dr. Angur. Thank you, Dr. Agakde, for that introduction. Everybody can hear me okay? Great. And I also want to thank the award committee for selecting me for this honor, and also, especially, Dr. Brittany Gauss for nominating me in the first place. I'm really grateful to be here with all of you this morning. I also just want to start out by remarking about Dr. Alexander Goralnik, whose name this award carries. Dr. Goralnik was a clinical leader and director of a hospital in New York State for a good part of the 20th century, caring for people with severe mental illness. And as you heard, I also cut my teeth on the inpatient units at McLean Hospital. So I'm very familiar with the challenges, but also the opportunities for improving the lives of our patients and their families, our patients with severe mental illness. So I think it's very apt that I'm here, you know, for a talk in Dr. Goralnik's name. This is my funding and disclosure slide. I want to make sure you can see the slides. I don't think you can see my cursor that I can see here. And I do have one interaction with a for-profit entity in the last 12 months at the bottom of the slide. The outline of my talk today, I'm going to start by giving you some of the philosophy behind early intervention, why we're here talking about early intervention in psychotic disorders this morning, and then really start at the very beginning with the emergence of psychosis and at-risk states in psychotic disorders, then go into some of the current research avenues and controversies in early psychosis treatment, presenting some of our own work in this area as well, and then early psychosis neurobiology, and leave you with some conclusions. So speaking of early intervention, I'm going to go back almost 30 years now, more than 25 years, to this very influential notion that schizophrenia as a disorder, of course it's a heterogeneous disorder and not with a well-understood neurobiology, but nonetheless, when you think of groups of people that could meet criteria for the diagnosis of schizophrenia, that following the onset of illness, in the middle of the slide, there's actually a deterioration, that there's a downward progression. People with schizophrenia lose community functioning, their impairment progresses, they accumulate insults, they also are more likely to develop medical conditions that end up claiming their lives earlier than they should, and so that there is this long-term trajectory that is an adverse trajectory. This has been a very influential idea. It's not always true, but for many people it certainly is, and the question is, when you're at the beginning of this journey on the left-hand side of the slide, how should we be thinking about somebody with schizophrenia? I'll tell you, when I started my training almost exactly 25 years ago now, I talked to the mother of a patient whose son had become sick in the 1970s and had been hospitalized, and she told me that she met with the treatment team in the 1970s, her son's treatment team, and the doctor told her, you know, you shouldn't come visit every day. You're giving him hope. You're going to make him think that things are going to just sort of go back to the way they were, that he's going to recover, but given the illness that he has, that's not going to happen, and so you really need to prepare him for that. This is a first-person account. This is what she told me she lived through when her son first got sick. Of course, we really live in a different world now, thankfully, and this is really exactly the challenge, isn't it? When you have a young person just getting sick with psychosis, how do we think about their life trajectory, what lies ahead for them? While the early intervention philosophy, I'll give you the punchline right up front, and I'll keep doing this. I'll tell you the punchline and then go into some of the data, is that we should view the onset of a psychotic disorder as a critical window of opportunity for intervention to prevent long-term disability. It's not the start of a long-term irreversible decline. We want to focus on improving long-term outcomes, and we can't wait for a breakthrough, miracle breakthroughs in research. We have to do it now. We always have debates about the role of discovery and the role of new treatments versus the role of implementing treatments that we already have available and doing it properly and doing it right, and there's always a tension between those two things, and that's fine, and of course we want new treatments, but we can't wait for some new breakthrough that's going to reverse the long-term course for our patients. We have to be thinking about what can we do now that will improve the long-term trajectory for our patients. Of course, this doesn't mean we are going to be Pollyannas about it and that all of our patients with schizophrenia, if only we said the right words or have them go to the right meetings that they're going to improve and they're going to do great in their lives, but if the overall orientation of the clinical team is that we're looking for better outcomes, we're looking for improvement, then of course we're going to be more likely to find it than if the overall orientation is that this is a long-term life sentence and we shouldn't be giving our patients hope. Just a couple of papers over the last decade or so from Australia on the top of the slide and from the UK at the bottom of the slide, but there's certainly similar data in the United States as well. Long-term outcomes following a first episode of psychosis are actually surprisingly positive. In Australia on the top, you see a study which looked at over 600 individuals and 57 percent of them over subsequent follow-up reported some paid employment and symptomatic remission was observed in more than a third and perhaps as much as half, social vocational recovery in 31 percent. These numbers are actually surprisingly good. We often have the clinician's illusion in the field of psychotic disorders. If you put up a shingle and you say, if you have a problem, come see me, the people who are going to come see you are the people with the problems. We don't always see the people who are doing well, who have better long-term outcomes to the same extent. As a result, clinicians who work with people with schizophrenia often think that schizophrenia is associated uniformly with poor outcomes. Of course, it is plenty of times, but when you actually follow a cohort systematically like this, you find that there's some people out there, probably a third or more, who can also do surprisingly well. Similarly, at the bottom from the United Kingdom, this is Paula Dazon's work. It's a little smaller study, but when you look at people who, based on some systematic criteria, were classified as not recovered versus recovered, there's actually 63 out of about 208 who are classified as recovered among the non-affective psychosis group. The numbers are, of course, even better for the mania and depression with psychosis groups. What are the principles in today's world for early intervention? Another influential paper from now about two decades, almost two decades ago, from Pat McGorry, who is one of the leaders in this field from Australia. The way to think about this is not to think about reinventing the wheel or coming up with some magical new formula that's going to help a 19-year-old, 20-year-old who's just coming into the clinic and being diagnosed with schizophrenia. It's practical preventive strategies, mainstream healthcare. What kind of clinic can we start that will take care of young people with psychosis, that can actually deliver the care we know already how to deliver? A population-based approach, improving access, pathways to care are a big deal in this field. Many young people come to care only when they're more severely ill, or only when they have an advocate. There's a huge difference between seeing, as we do at McLean, a child of two college professors who are staying up late nights surfing the internet looking for every and any opportunity, any treatment option, any new research study, and so on, versus the child of a single mother who has to, you know, who can't take time off from work and has to take two buses to get to our clinic. Those two kids are in very different situations. So we have to adopt an approach that's going to take all of those kids and improve outcomes. Systematic early delivery of existing and incremental care. So this is, again, not about, you know, waiting for some new medication to be discovered or some specialized neuromodulatory treatment to be discovered. We can already deliver systematically good care. And then treatment proceeds in stepwise fashion, depending on safety, response, and progression. So the intervention should be commensurate with the magnitude of the problem. So coming back to the slide that I already showed you, the question really is, what does that red line look like? How do we deliver early intervention that will actually prevent at least some of that deterioration for a majority of our patients? This is the philosophy for early intervention. This is overall where the field has been coming from and where it's going. Of course, we don't at the moment have good evidence that systematic early intervention improves long-term functional outcomes. We have some, but it's controversial. And most of what we know suggests that the moment you stop the early intervention program, outcomes actually start reverting back to the long-term trajectory as if they were never in the early intervention program. So it's more controversial than you might think. But of course, the hope is that there is going to be that red line there. And with better-designed programs, we can develop the research evidence base for showing that come into our program, and you're going to be more likely to be working 10 years from now or more likely to be happy with your life 10 years from now, as opposed to if you don't come into this early intervention program. So as I said, I'm going to go back to the beginning now, sort of march through the at-risk states and then the first episode, the emergence of severe mental illness, because this is, of course, where the early intervention is going to happen, and then also talk a little bit about the neurobiology. So I'm not going to dwell on the at-risk states. I'm going to show a few slides from work in the field, really just to highlight the complexity of the problem. So one of the issues is that at-risk states are heterogeneous. There is no sort of metaphorically railroad track that you get on, and the train takes you to schizophrenia, versus another track that you get on, and the train takes you to bipolar disorder or severe depression or substance use disorder, personality disorder, OCD, whatever it is, eating disorder. Instead, there is probably a pluripotent stage. On the top of the slide, you see these dots of different colors that are sort of floating together in the same pool of patients. There's psychopathology that's emerging over time, but it's really contingent on what happens to people, how much cannabis they're smoking, for example, or what kind of adverse life experiences they're going through. Their genetic loading is in there, of course, and so on. But how psychopathology unfolds and manifests itself actually changes over time, and ultimately, some people from that heterogeneous pluripotent pool on the top of the slide will end up being diagnosed with schizophrenia at the bottom left, but also some with bipolar disorder and so on, as you can see in the slide. The notion of going from this pluripotent group to specific DSM-IV, DSM-V, DSM-VI diagnoses is really going through different stages, and that's what you see on the left side of the slide. There is this notion that there is a complex group of patients that are not predetermined what kind of disorder they might have, and there's a lot of research interest in this. Also, the emergence of psychotic disorders, specifically if you're going to end up being diagnosed with schizophrenia, what happens to you in the years preceding the diagnosis of schizophrenia is also complex. On the right-hand side of the slide, you see this growing black blob. That's the early psychosis that's following a diagnosis. But to the left of the slide, you see people in these different colored groupings with different abbreviations. I'm not going to go into detail about all of these things, but APS is attenuated positive symptoms. COGDIS is cognitive dissonance. There are UHR criteria, the ultra-high-risk criteria. There's different ways of looking at this. But not every patient that ends up being diagnosed with schizophrenia takes the same path to that diagnosis. Some people start with unusual sensations as captured in this cognitive dissonance notion that they don't belong in the world, or time flows differently for them compared to what it does for other people, or that there's something going on that everybody else knows about that they don't understand, that the principles of life, the principles of society, they have not understood. Cognitive dissonance, that's the idea. For others, it's attenuated positive symptoms. So there are people who start to feel like they are getting meaningful looks, that there is a conspiracy that people are talking about them. But in the attenuated version of this, they can still tell themselves, well, that doesn't make sense. Everybody's going about their business. This is not about me. So they actually retain some insight. That's why it's attenuated positive symptoms, but eventually they blossom into full-blown positive symptoms, and so on. So there's all these different pathways to getting to a schizophrenia diagnosis. So the complexities continue, even for people who end up with schizophrenia. And then this is our own work from the early days of our first episode psychosis clinic. Even after receiving a diagnosis, there is actually very significant variability in what happens to people in the subsequent two years. So this was our first 100 patients, a chart review. And these are clinician diagnoses, so these aren't systematic structured interview diagnoses. But you can see that about half of all patients retain the same diagnosis from when they came into the clinic to when they were assessed two years later. But actually, the other half have changed. So even the first episode diagnosis is not really the final word. So there's a lot of variability in how people evolve. At-risk states are being more and more recognized, and there is a lot of intervention programs that are being developed. There's a lot of things you can do in the at-risk state. So these are young people who don't yet have a diagnosis of schizophrenia, schizoaffective disorder, bipolar disorder, psychosis, or anything else. But you can improve detection of these individuals in the community so that you can actually intervene early and hopefully prevent the progression. Reserve clinical need for help-seeking individuals. Many people who are in these at-risk states are actually anxious, they're depressed, they're withdrawn, they might be getting into problematic substance use patterns. And all of those things you can intervene upon and reduce the burden of psychopathology in the population. Preventing the onset of a psychotic disorder, that's sort of the holy grail. And there's not a lot of evidence for that at the moment, but hopefully it will be forthcoming in the future. Shorten the duration of untreated psychosis. So people also come to these clinics, billed as having an at-risk state, but when you assess them in the clinic, you find that they already have frank psychosis, and they're already in a first episode. And of course, this is a great way to get people into care who have a first episode. And reduce symptom burden and disability generally. There's good evidence, including meta-analytic evidence from a decade ago, but there's even more now, demonstrating that outcomes are improved for people in these high-risk syndromes with early intervention and specialized at-risk programs. But I said there is more to come on this. And a big part of that is that the NIMH, the National Institute of Mental Health, has launched a major program. On the top of the slide, you see Accelerating Medicine's Partnership for Schizophrenia, or AMPSKITS. It's called AMP-SCZ. That has committed well over $100 million to the research of these early at-risk syndromes, and also treatment interventions, studying treatment interventions. The U.S.-based version of this, funded by the NIMH, is called PRONET. There's a second parallel effort, also funded by NIMH, as well as others in Australia, called PRESCENT. That's on the right side of the slide. And the first papers are starting to appear, describing what this work is going to be all about. But there are literally dozens of sites around the globe that are recruiting and assessing people in these early at-risk states, learning about their characteristics, collecting translational data, including brain imaging, genetics, and others. And then eventually, the goal and the hope is to convert this into a clinical trials network of a kind, where early intervention studies can actually be done. So stay tuned. A lot more to come, probably, in these earliest phases of illness, before people have received a formal diagnosis. Going on to early psychosis treatment now. Once again, I'm going to give you the punchline. This is a relatively mature field, especially compared to the at-risk syndromes. The standard of care for early intervention now, in the U.S. and around the world, is Coordinated Specialty Care, or CSC. CSC, again, is not a magical formula. There's no special medicine. There's no potion that you deliver. It's actually team-based and multidisciplinary care, intended to be comprehensive, intended to be intensive, similar to assertive community treatment, you know, those PAC teams and ACT teams that some of you are probably familiar with. Person-centered, focused on engagement, recovery-oriented, prioritizing return to work or school, over control of symptoms. Psycho-education is important, including family psycho-ed. Phase-specific treatment, I mentioned this already, you know, that the intervention should be commensurate with the problem. What this really, one of the biggest principles that comes up in this notion of phase-specific treatment is not being heavy-handed with antipsychotic medications. Young people coming in for the first time into the mental health care system, we have made often the mistake of being heavy-handed with medications that both creates a very significant side effect and adverse event burden, but also drive people away from care. Young people feel like, this is not meeting my needs, this is not what I'm really looking for in terms of improvement, and that creates a tension. So actually, start low, go slow for the earliest phases of antipsychotic treatment is a big part of this. And finally, time-limited. CSC is not something you do for life. CSC is what you're supposed to do for maybe the two years, I'll get back to this, there's some debate about this, probably the longer the better, but it's time-limited. It's a program with a beginning, a middle, and an end, and you're trying to accomplish certain things while the patient is in treatment. So you can see, this is, you know, very different from having a patient who comes to the emergency room or gets hospitalized with the first psychotic episode, and when they're ready to leave your hospital, you say to them, go see this doctor, you know, Tuesday morning, 9 o'clock, in the community somewhere, in some town near you. Or here's the number of a clinic, call them, and they'll give you an intake appointment, and maybe you'll meet with somebody, you know, two weeks or a month from now, and then they'll tell you what to do next. That is ineffective, overmatched, compared to the urgency, the intensity of the problem that that young person and their family are really dealing with. You know, the 18-year-old, 20-year-old, just developing a psychotic illness that's been severe enough to come to a higher level of care. What they really need is a team that's going to come meet with them, and they're going to be talking about school and work, substance use, social relationships, where are you going to live, how are you feeling across the board, not focused on, you know, how bad are the voices, are the hallucinations getting better, and so on and so forth. And then it's going to be intensive. It's going to be assertive. It's going to say, we're going to come visit you in your house, if that's what it takes. We're going to see you next week or tomorrow. You've got to come to the clinic. And it's got to be youth-oriented. It's got to meet the needs of an 18-year-old whose number one agenda item is not to get proper treatment for their medical condition, but rather their number one agenda item is to get back to their friends, to get back to their family, to get back to school, to not be asked, hey, where have you been? You've been missing for weeks, right? They want to get back as quickly as they can. And we have to understand all of that. In fact, the Australians, Pat McGorry sometimes says, what you really need is young staff. These clinics need to be staffed by people who understand the priorities of the young people that we're serving. That's probably open for debate, but you get the point. So this is the standard of care. This is what we need to be doing. And how do we know that? Well, the story goes back, again, 20 years. One of the first randomized control trials of integrated versus standard treatment for patients with the first episode of psychosis was conducted in Denmark, Marita Nordentoft's group. And they were able to show that this integrated team-based, youth-oriented approach is associated with better outcomes. And not just better outcomes clinically, for example, in psychotic symptoms, negative symptoms, and lowered substance use patterns, and so on, but also economically. You actually, if you intervene early, you save money in terms of the number of dollars that you're spending on taking care of these young people over the subsequent few years. So this is not new. That literature has been out there for a long time. In the United States, the most influential study, of course, was the RAISE study. Many of you might have heard about this one. It's recovery after an initial schizophrenia episode. This was funded by the NIMH and, again, conducted about a decade ago. So we were a little late to the party in the US compared to the Europeans, the Australians, and even Hong Kong and a few other places in the world had been conducting these studies. But the US-based study was a well-done, large, definitive clinical trial that provided clear evidence that team-based care and coordinated specialty care, or CSC, is actually superior to routine care. I'm just going to show you one slide from that famous paper from 2015. John Kane and colleagues published this paper. And the orange dots are the NAVIGATE, which is the CSC, the standard of care that I just have been talking about, versus the green dots, which were the community care. And on the left, you see greater improvements in symptom burden in people who are receiving team-based care. And on the right, you see – sorry, on the left is greater improvements in quality of life – quality of life, not symptoms. And on the right, you see greater improvements in symptom burden in both cases in people who are receiving the standard CSC treatment. So there is now overwhelming evidence that when you design a team like this that takes care of young people with these problems, outcomes are going to be better. And there is meta-analyses. Plenty of studies have been done. This is from 2018, and there is a few additional studies since then. But in pretty much domain after domain, whether it's treatment discontinuation, hospitalizations on the top, remission, recovery, involvement in work or school, and then symptom severity, and then care utilization, psychiatric hospitalizations, and at the bottom, quality of life and global functioning. In every domain, you're better off if you're in one of these coordinated specialty care programs. So this is really a well-established literature now. This is what we need to be doing. And of course, there's a lot to talk about. The question, as many of you might be thinking, well, in my care environment, what am I supposed to do? How do you actually deliver this kind of care? And it's team-based care. So this is not something you can just decide to organize for yourself and try and implement. Of course, it's great if you can. But oftentimes, public policy needs to get involved. And it has. In many countries around the world and in the United States, we now have government support, both in the U.S. federal government and state government support, and networks of treatment centers that are actually learning from one another. In the U.S., the example for this is EPINET. And EPINET stands for Early Psychosis Intervention Network. It's funded also by the NIMH. It brings together, as you see, over 100 early psychosis clinics across 17 states. And there's a national data coordinating center. We're part of EPINET in Massachusetts at my hospital, at McLean Hospital. I'll go into that a little bit more in a minute. But the notion here was to create a network of these early psychosis intervention centers funded in part by the federal government. The Substance Abuse and Mental Health Services Administration, SAMHSA, actually earmarks federal dollars by law to be put towards early psychosis intervention. And then the states administer those dollars to create these treatment centers around the country. So at this point, you know, Bob Heinsohn, who's one of the NIMH leaders in this area, has a great slide showing that by 2022, 49 states in the union actually have an early psychosis strategy and a network of centers. Now at the national level, bringing them together in EPINET, the goal is to create a learning healthcare network. So to be collecting standardized data as a matter of course, routinely when people come into the clinic, certain kinds of things get recorded and they all get sent to, with the proper agreements and permissions, to a national data center where at this point we have data from over 4,000 young people all around the country who are being treated in these coordinated specialty care clinics, including again ours in Massachusetts, to see how things are going, what's working, what's not working. Learning healthcare, the notion is that, again, you're collecting standardized data as a matter of course, but you're also studying the patterns. You're noticing things that you can actually intervene upon. So if you're seeing that people in rural areas are dropping out of CSC treatment because they can't access care, you might do something about that, but you're going to see that in your data set, right? That kind of thing is happening already. So early intervention has gone national and global now and networks of clinics are delivering this kind of care. So our involvement in this comes from the LEAP Center, Laboratory for Early Psychosis. It's funded via an IMH grant, an Alacrity Center mechanism. It launched in 2019 but renewed in 2023 and I'm the corresponding PI with my colleagues, John Hsu, who's a health economist and internist at Mass General, and Miguel Hernan, who's an epidemiologist at the Harvard School of Public Health. So we've brought in people into this early psychosis field in our center who are not in the mental health care field themselves. So we're hoping to take advantage of some of their knowledge and expertise and experience that they otherwise wouldn't be involved in early psychosis care. So the LEAP Center coordinates 14 early psychosis clinics, first episode psychosis clinics within Massachusetts, accessing statewide, nationwide databases, some of these things that I was talking about, and we're part of EpiNet. We've been sharing some of our earliest experiences, also collaborating with Dr. Keshavan, Macheri Keshavan, and the Massachusetts Department of Mental Health folks. We've been collecting harmonized data collection and publishing on the results of our findings. This paper from 2021 was one of the early ones that showed improvement across the board in the first six months of coming into these clinics. So we're simply replicating what many others have found, that these clinics actually are associated with good clinical outcomes. The LEAP Center has also collaborated with Dr. Keshavan's group, MapNet, to create a statewide strategic plan for early psychosis care in Massachusetts. We've been bringing together stakeholders, clinicians, researchers, payers, and policymakers to develop what's called the MASS STEP, so the Massachusetts Strategic Plan for Early Psychosis. We hold annual meetings. This is the last one that took place in November of 2023. The plan is written now, and I'm going to talk about that in a minute. So the meetings now are really mostly about reviewing progress and identifying priority areas for the coming period. If you want to read more about that, this is the paper that we wrote about strategic roadmap for Massachusetts. So what's in the strategic plan? Well, I already talked about access to care, expanding capacity. I'll show you in a minute, but we already know from our own data in the LEAP Center that the capacity that we have for taking care of young people with psychosis is woefully inadequate. We probably are capturing about 10% of everyone who has the need. Serving underserved communities, diversity in clinical teams, this is a big deal because our patients, if we want to keep them in treatment, we want the treatment teams to look like the patients and there to be mutual trust, and that's not always easy in our settings. Role of peer providers, this is something that was not in the, if you remember the definition of CSC with all those bullet points, there was nothing about people with lived experience. Certified peer specialists who are now getting employed as clinical workers along with the rest of our CSC teams, so that's part of the Massachusetts strategic plan. Reducing staff burnout and turnover, this has become a real problem, especially since the pandemic. It's actually interesting to talk to colleagues like in Denmark, I already mentioned Marita Nordentoft's group. In their context, getting a job in one of the CSC clinics is considered a plum job and people, once they get that job, they don't want to leave because they think it's good work and they feel good about their work conditions and so on. Whereas in the US, turnover is actually at breakneck speed. It's actually very challenging to keep people in these jobs. I talked to a colleague in California who told me they have to renew the basic training every year because the entire clinic turns over every year. There's nobody left from a year ago in the clinic and so they have to train everybody again in the principles of early psychosis intervention. That's how bad things can be. There's a lot that goes into this, of course. There's pay, there's work conditions, there's large caseloads, we all know that mental health care is under-resourced across the board and this area is no exception. Finally, an idea that came from actually the mother of a young man who had developed a psychotic illness, been hospitalized and put in a general psyche ward with people who were much older than him. Many of them had chronic mental illness, were impaired, hadn't worked or lived independently in decades. There was also people with opiate use disorders, there was people with personality disorders. You take an 18-year-old and you put them in this kind of environment and tell them, this is what we're doing to help you. It backfires. This mother said to us, can't you guys have some kind of service like 8 beds or 10 beds just for the young people with these problems? That's specialized. My son didn't have to go into that hospital because now it's taken us a year to convince him that he should go back to get some care because he didn't want to do it after that. We don't have that unit yet in Massachusetts but that's under active discussion and it's because of the strategic plan process that the LEAP Center was able to help implement. This is the paper that I mentioned. I'm not going to go into the details of this. This is identifying diagnoses of schizophrenia spectrum disorders in large data sets, looking at a data set from Massachusetts, the All-Payer Claims data set. These are insurance claims. Anybody in Massachusetts who receives any kind of care that gets billed to any insurance provider, the state has mandated that they have to report those data to a single pool called the All-Payer Claims Database, APCD. We have access to that database at the LEAP Center and when we do an analysis with very rigorous criteria to identify first episode schizophrenia, we find that every year there's about 1,400 young people being diagnosed with schizophrenia in our state. Massachusetts is about 6 million people. It's one of the medium-sized states in terms of population. 1,400 people. I know from collecting the data from all of our 14 clinics that exist in Massachusetts, the CSC clinics, that we're doing about 100, maybe 110 intakes every year. It's not even 10% of the need. This is why I say access is woefully inadequate. When we look in these data sets, what is happening to the rest of the 1,300 people who are not getting CSC? It's just the worst case scenario. They're bouncing around. They're in and out of emergency rooms. They're in hospitals. Some of them then get lost to the healthcare system, which probably means they're incarcerated. We don't have that data in this data set because this is insurance claims, but lots of bad things are happening that are entirely preventable if we can actually set up a public healthcare system that meets the need. Some of the projects we're actively working on now, I just mentioned this first bullet, what happens to people who don't make it in the clinic door, who are not getting CSC. Another one that's a perennial favorite in this field is impact of cannabis use on first episode psychosis trajectories. I already mentioned this. I say it's a perennial favorite because there's good evidence that cannabis use in teenagers is a risk factor for the subsequent development of psychosis. The problem is that there's not a gold standard randomized control trial or other experimental evidence. The way to really prove this would be to take a large number of adolescents, randomize them, and force half of them to smoke weed daily, every day, and prevent the other half from smoking any weed ever. Do that for maybe 10 years and come back when they're 25 and see who has been diagnosed with schizophrenia and who has not. You can imagine that's never going to be done for lots of reasons, ethical, practical, and so on. Short of that, everything we know is that if you're smoking a lot of weed when you're 15, 16, in your middle teenage years, you're asking for trouble. The odds of being diagnosed with psychosis is going to be three to fourfold higher subsequently. We're doing some work on that as well, looking at our Massachusetts data. Another area is first episode affect of psychosis, people with bipolar disorder with psychotic features. This is a condition that's just as common as people with schizophrenia, and yet the research database for this field is actually much farther behind compared to where it should be. We also have data on first episode affect of psychosis that we're studying. I just want to say one word about our signature project. The way the NIH mechanism is written, we needed to pick an intervention that we wanted to test. What we decided to do is test an intervention that we're calling CSC 2.0. In this case, let me start with the logic, which is really in the bottom of the slide, the last bullet point. The primary outcome for our intervention is retention in the clinic. The idea here is that we know these interventions work. There's already really good clinical trial evidence that if you come to this clinic, you're going to do better than if you don't come to this clinic. The problem we face now is retaining young people in these clinics. When we look at our Massachusetts data, among the people where we have baseline information, by 12 months, only 50% of them are still coming to the clinic. We're losing half of our audience within the first 12 months. The logic here is very similar to the famous Cady study, again, from 20 years ago, some of you will be familiar with. We know antipsychotic medications will reduce symptom burden. The question is, will people take the antipsychotic medications? The Cady study decided the primary outcome measure they wanted to look at was adherence to the medication. It's the same thing here. Can we keep people in the clinics? We're comparing routine CSC versus CSC plus these care elements that are in the middle of the slide here. Peer support that I already mentioned. Digital outreach, a smartphone app that retains the young person's connection with the clinic. Care coordination. This is specifically referring to when patients... One thing that we hear from CSC clinicians all the time is that when patients get hospitalized or they're in the emergency room, no one calls us. We don't even know that the patient was hospitalized. True story. We've had patients in our clinics in Massachusetts who get hospitalized, who have a problem. They are exacerbating. They go to the hospital. They're in the hospital for a week or 10 days. They get discharged. They come to the clinic a month later for their routine appointment. The clinic has no idea about anything that's happened. In fact, sometimes the patient forgets to mention or to them it's not a big deal. The patient doesn't tell you and you still don't know because of the fragmented United States healthcare system that we're all laboring under. Care coordination is actually really important because sometimes you have young patients whose medication regimen has changed or they've been told something in terms of psychoeducation that actually runs counter to what the CSC team was trying to do, but nobody actually knows that. So we're actually hiring staff in the clinics to specifically work on care coordination to stay in touch with patients' families and also inpatient and ED teams to make sure that everybody's on the same page. Cognitive remediation. This is for the subgroup of people with psychotic disorders who have cognitive impairment. Either objectively measured using cognitive testing or self-report, self-requested, self-nominated shall we say. People who say I want help with my thinking and my memory and so on. So we're actually providing computerized software-based interventions for improving cognition. This is, if you're familiar with Sofia Vinogradova's work, we're using the same software package from Posit Science that she has used in her studies. And finally, multifamily group therapy. The logic there was that if you can keep the families engaged in the clinic, that will improve the chances that the patients will also remain engaged in the clinic. So this package of five different interventions added to routine CSC. The question is, can we do better than what we're doing already? We know CSC works and it's a perfectly good intervention, but maybe we can improve upon it and call it CSC 2.0. So we're doing a randomized controlled clinical trial, again funded by the NIH in our LEAP Center, really turning the Massachusetts clinics collectively into a laboratory for research. Everybody's delivering care, but we're also testing out these new strategies. A few other things in the interest of time I'll just mention briefly. Of course, we studied the COVID impacts on patients and providers, especially telehealth. We're looking at the role of peer providers in the clinics that already had them, expanding capacity, some of the same things that we already talked about that the LEAP Center is working on. So this is sort of, I've been dancing around this issue a little bit. Although things are much better than they were in early intervention, clinics are overstretched. They lack resources. There's a lack of a long-term financing plan. Right now in the US, but also around the world, actually, including Australia, UK, and elsewhere, governments are providing the funding for early intervention as a special case. Early intervention is not incorporated into the healthcare system. So it's actually not like a cardiology clinic or an oncology clinic where you get some specialty care. It's more like there's a separate source of funds and you can open that clinic. But of course, what that means is if those funds were to stop tomorrow, there is no clinic anymore. So we actually have a lack of a long-term financing plan. This is being addressed. There are some good things happening, actually, in this area. The Center for Medicare and Medicaid Services recently approved these new so-called HCPCS codes. It's not my specialty to know what HCPCS stands for, but it's these billing codes that will actually allow for monthly flat rate reimbursement of clinics as part of the healthcare system. So at some point in the future, hopefully, if you're sitting in a CSC clinic, you'll be able to bill Medicare and Medicaid. And then, of course, usually when CMS passes these rules, private insurers follow, you'll be able to bill insurance for everything you're doing for that patient and receive one monthly lump sum payment. And this will be similar to what's done, for example, in other settings, people with long-term disabilities, children with intellectual disabilities in similar situations. There's actually monthly payment systems that keep those teams running. So there's some hope that early intervention will actually become much more institutionalized in that way, but we're not really there yet. So a couple of things I'm going to focus on next in the treatment of early intervention. One big one, of course, is always antipsychotic medications. I have colleagues who say to me, we're so frustrated with basically serving as a medication surveillance system. Are you taking your meds? Make sure you're taking your meds. Don't forget your meds. That's what the psychiatrist's role gets boiled down to. And of course, it's very frustrating with the treaters. It creates tension. It's an adversarial relationship with the patients. So there is this burning question about should patients with first episode psychosis remain on antipsychotic medication long-term? Treatment guidelines say yes. Existing evidence is controversial. On balance, of course, patients should stay on antipsychotic medications. They improve the clinical condition significantly. But there are things like this. This is a much-discussed paper from the Netherlands, again, a decade ago. I'm not going to walk you through the slide, but what you're looking at here is a clinical trial where people with first episode psychosis all went on medication. And after one year, they were randomized to stay on the full dose of the medication or to start dose reduction and, if possible, discontinuation. And what happened in the first couple of years is what you would expect, that people who were going down on their medication doses crumped. There was many more relapses, hospitalizations in the first couple of years. But after the first couple of years, the first three years, the curves actually started crossing over. So people who were on lower doses of medication on average tended to do a little bit better. And they were more likely to be functional in the community. Not their positive symptoms, but their community functioning was a little bit better. This paper is not the final word, and nobody here should rush back to take their patients off of their medications, right? And we haven't certainly done that at McLean. But it raises these questions about there are probably patients in our patient population who could go on lower doses of medication, and it would be better for them. What we don't know is who those people are at the front door and what to do about it. It's certainly extremely risky to try this for everyone. You know, let's taper everyone's medications down and we'll find out because 60% or 70% of them will relapse and then we're going to know they can't do it and the remaining 30% won't relapse and we're going to know they can do it. You know, that's extremely risky and it's unethical. So we have this conundrum of knowing there are probably some people, especially in a more mixed first episode population, this paper was not specifically about schizophrenia, this was actually a broad psychosis phenotype. There are people out there who can probably reduce their medications and it would be better for them. Well, these studies are extremely difficult to do. Again, our colleagues in Denmark have tried this and were actually not able to conduct a study. So they wrote this paper about lessons learned in the terminated clinical trial they were trying to do. One of the challenges that this group had was when you say we're doing a research study on discontinuing medications versus staying on your medications, do you want to join? Who actually signs up for that study? It's the people who want to stop their medications. So they actually had trouble with the continuation arm. They couldn't recruit enough people who were willing to stay on the continuation arm. So the sociology of the relationships and the patient's priorities, the family's priorities, the clinician's priorities all play into this. It's very complicated. Well, this is a tough issue. It's not going to be easy to do good clinical trials that give us definitive answers about how to handle medications. Everybody goes on the medications with a first episode psychosis, but what to do after a period of stabilization for the long term is an open question. One thing we're doing in the Leap Center is something called a target trial emulation. So when a randomized control trial is impossible, impractical, unethical, can you actually do something with observational data that gives you an answer that can actually be useful? The way this works, I'm going to briefly describe a target trial emulation, is if you have a good enough observational data set, meaning a cohort of patients that you've recruited and you're measuring how they're doing and what treatments they're on with some frequency, and there's some treatment variability in that cohort. Some people are going on medication A, some are going on medication B, some are stopping their medication, and so on. Can you go into that data set and basically select a data, a group of data, that is basically like what you would have done if you'd done a clinical trial? So go to the beginning of when a patient goes on that medication and start following them. First of all, everybody with the right inclusion-exclusion criteria, like you would have for a clinical trial, and then you go to where in your data set they started taking the medication, and then if the clinical trial would have done weekly visits measuring their positive symptoms, then you look in your data set for weekly visits measuring positive symptoms. So you're emulating a clinical trial without actually having done a clinical trial. Of course, you have to have a data set that's good enough to do this. You have to have those weekly visits with the right kind of data and so on. But if you have that, you can actually start to do an emulation. Surprisingly, this kind of approach of basically looking for a trial buried in your observational data set, your cohort data set, can account for a lot of the flaws that are often seen in observational studies, but not all. Of course, there's no randomization. So there's always a reason why one patient goes on medication A and another patient goes on medication B. And if they are not randomized, there's going to be systematic biases. So target trial emulation can't take care of everything. It's not the same thing as a clinical trial. And in fact, I'm starting to see people talk about target trial emulation as a magic bullet. You know, you don't have to do a clinical trial anymore, and we have to be careful about that. That's simply not true. But anyway, so this is an innovative way of looking at data sets that might give us better answers than what we have so far. So my colleague, Miguel Hernan, who I already mentioned, has done this work in statins and in cancer. And what they are able to show is that if you have an observational data set, again, that's good enough. You should do it this way, emulate a target trial, because what you will do then is address a lot of avoidable flaws, and your conclusions will be much more reliable. And we have started working together on a couple of specific questions. One thing we did was use data from the STAR-D randomized trial to model dynamic treatment strategies for major depressive disorder. This was really just a proof of principle. This is not our own data set, but the idea was, can we start to implement this target trial emulation idea in psychiatric data sets? And, you know, this paper actually has some interesting points in it about STAR-D results. But the things we're moving into now— shoot, sorry, I should have updated this. This paper has now appeared in the American Journal of Epidemiology. It says in press, but it's now in AJE. We're calling our own approach to this FEPCausal. We've been collecting an international consortium of observational cohorts with this explicit purpose of emulating target trials of antipsychotics. The first thing we've done is benchmarked results to estimates from UFES, which was a randomized control trial conducted in the 2000s. So the idea here is, okay, this is an innovative new approach, and maybe we can actually learn something from observational data that's going to be similar to what you learn from clinical trial data. But first, what we should do is select a clinical trial and benchmark ourselves. If we do everything right in our data set, do we see the same things as they saw in that clinical trial? Because if you do, that gives you confidence that the emulated target trial idea is working in this kind of data set. So FEPCausal had over 1,000 patients when we did this first analysis that benchmarks to UFES, and we saw similar results for discontinuation and absolute risks of hospitalization in our data set, emulating a target trial as they did in UFES. So, of course, this doesn't prove that everything in our data set is going to be as good as what's in a clinical trial, but, again, it seems to be working approximately to remove much of the confounding. So this is an approach that I'm actually quite excited about because I think we can provide some additional context and information for the field that has not been easy to collect. We have a second project that's now emulating ULAST, which is another European-based clinical trial of long-acting injectables. Over 1,400 individuals now. FEPCausal continues to grow. And we also replicated, basically, what ULAST has found with long-acting injectables, meaning there is no benefit to long-acting injectables in first episode of psychosis. This might sound a little counterintuitive because long-acting injectable medications are often associated with better outcomes in at-risk patients and unstable patients, but in clinical trials you typically don't see that because who signs up for a clinical trial? It's somebody who is taking care of themselves, is interested in research, will come back and follow up with you, and so on. And those are not the patients who need the long-acting injectable, right? So the selection bias for a clinical trial actually has limited the efficacy of long-acting injectables in clinical trials, and we see the same thing in our dataset. But actually now there is an example of extending findings to a new subpopulation. ULAST could not look at people who had a first episode, went on the medication, and then relapsed again with a second episode. But in our dataset we have that because we're collecting data clinically, so some people actually relapse, and we see a strong protective effect of long-acting injectable after a first relapse. So this is an example of how you can both benchmark yourself with these observational studies if you do them right to clinical trials, but then maybe also be able to develop some new observations that are not available in the clinical trials. One of the things we're very interested in is what happens to people when they discontinue their medications. So, for example, can we emulate that Dutch clinical trial that I showed you that after the first three years or so the curves crossed over in terms of stopping medications? We don't know, but we're very interested in that, so more to come. Stay tuned. Okay, finally, I'm going to tell you a little bit about neurobiology of early psychosis and schizophrenia. Once again, I'm going to give you the punchline up front. Schizophrenia is at least partly a neurodevelopmental disorder. What happens at age 18 or 20 or 25 when a young person comes to the doctor and gets diagnosed with schizophrenia is not the start of the process. It's a culmination of the process that starts with genetics, in utero brain development, birth events, early childhood brain development, and adolescent brain development. If you think of psychosis this way, it really changes priorities because it brings up the notion of early intervention. Of course, this is exactly what happened in cancer. Screening in the population, early detection and intervention can change the game and outcomes for cancer. It's happening also in neurodegenerative disorders. We now know that many processes that ultimately lead to a diagnosis of a neurodegenerative disorder started a decade or more prior to the diagnosis, and that's where we need to intervene to prevent that long-term trajectory. The exact same thing here. This is an influential paper from 14 years ago from Tom Insel. I show it not because the specific content really matters for us, but primarily to show this notion of how, like in the bottom panel, you see these curves. The solid lines are the healthy curve, and then the dotted lines are what the best evidence indicates happens in schizophrenia, that there's a deviation from healthy brain development. Over time, people with schizophrenia actually no longer have the average healthy brain development that you would want. Again, a neurodevelopmental perspective argues for primary prevention and secondary prevention. If we know what the exposures are, what the offending agents are, one way to improve outcomes in schizophrenia is to intervene early so that you can actually improve those outcomes. If schizophrenia is the culmination of something that takes 20 years to unfold, maybe if you can intervene in those first 20 years, you'll actually have ultimately, of course, either fewer cases of schizophrenia or at least less severe outcomes for people who are being diagnosed. Things like poor prenatal care, adverse childhood experiences. One interesting one is urbanicity. There's a lot of new evidence coming out. These effects are more subtle compared to some of the bigger things like neglect and abuse in childhood. But still, there's actually a connection between air pollution, for example, and emergence of mental illness, and also cannabis, which I already mentioned and I'm going to get back to in just a minute, as well as secondary prevention via early detection and intervention, which is what we've been talking about. So the neurobiology is taking us in this direction. This cannabis issue I already mentioned, but I'm a believer of what Sir Robin Murray likes to talk about. The causal relationship is going to be very difficult to prove, but everything we know says that there is a relationship. No one has ever found evidence that says smoking weed in your teenage years does not increase the risk of psychosis. Anyone who's looked at this finds the same relationship. So Sir Robin, along with Marta De Forti, wrote this article a few years ago now. You know, a little bit of an exasperation saying, what level of proof do we require? What more do you want us to do to send this message out there that cannabis is risky in teenagers? And my colleague Eden Evans and her group at Mass General have written an article where they say, the time is right for the United States to follow Canada's lead and invest in public health education and messaging regarding cannabis as legal but not safe for younger users, particularly high-potency products. So the point here is that it's not about legalization or criminalization. It's about public health messaging. I mean, it's really quite amazing the way the pendulum has swung in the United States and in the Western world. You stop the average person on the street and you ask them about tobacco, and they'll tell you lung cancer. You ask them about alcohol, and they'll say, don't drink and drive. You ask them about weed, and they say it's totally safe. I mean, I don't know, maybe you're hearing something differently, but this is certainly what I'm hearing. It's totally safe. And it's true that cannabis doesn't cause lung cancer. It doesn't even cause tooth decay, which was something, apparently, I didn't know anything about this, but in the literature there was this concern that it was going to be associated with tooth decay. It doesn't even cause that. But it's associated with psychosis in teenagers. So the message isn't, oh, cannabis should be illegal, it's a gateway drug, it's going to be terrible, you'll lose your mind, blah, blah, blah. The message is probably everybody in this room can smoke all the cannabis you want, and you're not going to have any adverse effects from it. And I have been quoted out of context on that topic. A journalist who interviewed me, Massachusetts had a referendum in 2018 about legalizing cannabis, and they interviewed me, and I talked for 45 minutes about the psychosis risk and the public messaging and so on, and the only thing she quoted me on was, you know, I tell my trainees, anybody after the age of 25, if you smoke all the weed you want, you're not going to have any problems. But the average person on the street needs to know, oh, it's not good for kids. Teenagers need to be careful because it's associated with schizophrenia. But that's not what's happening at all right now. And who's smoking weed in the Western world? I mean, of course, there's the occasional person who after a hard day's work goes home and smokes a joint to relax or something, but the majority of cannabis is consumed by teenagers and young adults, and they are the ones who are at risk. So we really need to just change this debate. Similar story actually with stimulants, psychostimulants. This is my colleague Lauren Moran's work that appeared in 2019. The risk of a psychotic episode after starting psychostimulants was twice as high if you're taking Adderall, which is known on the streets as Speed, right, compared to if you're taking methylphenidate. So, of course, the message is not that Adderall will cause psychosis. The absolute risk is only 0.2% in the 60 days after starting amphetamines. But the risk is higher for amphetamines. If there's no reason to use amphetamines, start with the safer drugs. That's really the message from this. So developing brains are susceptible to the emergence of psychosis with specific interventions, and the primary prevention can actually reduce the morbidity coming from psychotic disorders if we can identify these kinds of exposures and mitigate them. So the other things I want to say about the neurobiology of schizophrenia is that the emergence of illness is not a linear progression. I think many people, including lots of researchers in the field, have this mental template that there is a lesion, and the lesion manifests itself. So first, you have a small problem, and then that problem progresses, and then you have a bigger problem. But actually, it's a very dynamic situation, and there's lots of risk but also lots of resilience, and the brain is changing all the time. We know a lot of that, actually, from the work of Dr. Gokde here and many others as well, and that every new step triggers a new physiologic situation. So outcomes may be determined not by the initial conditions but by brain response to perturbation. So we really have to get smarter about understanding somebody who has a first episode of psychosis, what has changed in the brain, and what's going to get better and what's going to get worse. And then, of course, for many people with psychosis, there's an active progression of illness. So the tension between neurodevelopment and neurodegeneration, these are sort of two camps in the field of psychosis research, arguing about what those processes are. I mean, I think there's no doubt that neurodevelopmental trajectories are involved, but I actually, just from clinical experience, you know, probably many of you will know this picture that I'm about to talk about. I think there's actually a neurodegenerative component because we see young people who come in with a first episode of psychosis, and many of them kind of, you know, regain their footing, their equilibrium. The meds help, their families swing into action, they catch their breath, they metabolize a little bit what happened, they apologize to the people they've offended, or whatever it is, you know, they try to get back to life. But then there's actually a subgroup of people where that first episode triggers a really toxic progressive process. And no matter what we do, things continue to get worse. Those are the people who end up on clozapine, they get ECT, they end up on first-generation antipsychotics, even though the side effect burden is more severe, and so on. But, you know, whatever we do, there's a few months improvement, and you think, oh, maybe finally this is getting under control, but then it continues to deteriorate. I've personally seen a couple of people die by suicide in that kind of profile. So, in those people, the first episode is the same, but the response to the first episode is different. So it's not the same thing, right? Having a psychotic disorder doesn't mean everybody has the same process. Everybody has the first episode, but then some people respond to that first episode in a toxic way. The way I imagine this, it's almost like what we learned about COVID, you know, where it's not really the COVID infection that killed a lot of people, it was that inflammatory reaction, the ARDS, the acute respiratory distress syndrome, that killed a lot of people. It was a homogenous reaction to the infection. So, just like that, the brain reacts, in some patients, in this degenerative manner. We know there's progressive cortical volume loss, so once the diagnosis has been made, the next few years are really important. But actually, at least in some studies, this volume loss can be reversed by, in this case, cognitive enhancement therapy. This is Machere Keshevan and his group. So, losing brain volume is not destiny. If you intervene early, you can actually arrest or even reverse that loss. As you can see, for example, in the top of the slide here, the amygdala and fusiform gyrus, the solid line is treatment as usual, the dotted line is cognitive enhancement, and with the solid line, that's trending down, whereas the dotted line is trending up in the amygdala and the fusiform gyrus, and even in the hippocampus. So, you can actually retain tissue if you intervene early enough, at least in this one study. I'm going to mention a couple of things from the work that we do in this early psychosis field in terms of brain imaging, and that work has to do with brain energy metabolism. So, this is an old slide from a very influential review article about the importance of energy metabolism in the brain. So, here you're looking at a synapse. On the top, you have the presynaptic neuron, and on the bottom, you have the postsynaptic neuron, the two yellow profiles, they come together. The red spheres are glutamate, and they are being released into the synaptic cleft, but then the blue profile is a glial cell that's picking up the glutamate and then detoxifying it. But I'm going to point out the red arrows with ATP and the pink profiles, both in neurons and glial cells that you can see on the slide. So, those pink profiles are mitochondria, and mitochondria are little organelles in all of our cells, and mitochondria generate ATP. ATP is adenosine triphosphate, it's the energy currency in all of our cells. That's what your cells need to do work, whether you're running down the road so your muscles are burning ATP, or you're doing synaptic transmission, your neurons are talking to one another, and your neurons are burning ATP. Energy is really crucial for normal brain function. And there's been a kind of a slow burn, simmering literature on energy metabolism abnormalities in psychiatric disorders for a few decades now, but it's really starting to pick up steam now. It turns out you can measure brain ATP levels, as well as some other interesting things, using an MRI imaging technology called Phosphorus MRS. I'm not going to go into the details here. I'm just showing you a slide here from my colleague, Faye Dew at McLean, who does this work. And at the bottom, you see these peaks and valleys, these squiggly lines, and you see there's three ATPs, gamma ATP, alpha ATP, and beta ATP, on the right-hand side of the slide. Those are actually the peaks of the three phosphates. It's adenosine triphosphate, so there's three signals, and these are the three signals. The bigger those peaks, the more ATP in your brain. And this is an MRI scan, just like a fMRI or a structural MRI for research, or if you hurt your elbow or your knee and you go get an MRI. There's no radioactivity, there's no needles, it's just measuring brain ATP levels noninvasively in vivo. Doing this work, we have found that there's actually a significant decrement in ATP synthesis in the brain in schizophrenia. This is chronic patients on medication, stable, and you're seeing two curves here. The solid curve is healthy controls, and the dotted line is schizophrenia. And in this case, the farther down the line goes, the more ATP is being produced, for reasons that I'm not going to get into, but it's sort of, you go down, and that means more ATP, and you can see that patients with schizophrenia are not producing as much ATP. And the x-axis here is seconds, so over even a few seconds, you start to see this divergence, that people with schizophrenia are not making ATP in their brains. This calculates out to about a 22% reduction in the rate of ATP synthesis. This is a, you know, in all my years, this study is from 2014, when I saw these data, I thought to myself, I've looked at a lot of functional MRI, DTI, structural MRI data, some of you may be familiar with these modalities, and we're happy when we see a 0.5% difference between patients and controls. And there's always extensive overlap between the patient group and the control group. Here, you're looking at a 22% reduction in ATP synthesis. Energy, the very basic thing that your neurons need to function. So I thought this is really something important and we should pursue it, so we've been doing that. We spent the last decade looking at all aspects of energy metabolism and I'll show you what we've been finding. In the interest of time, I'm not going to go into the details of this, but one thing you can do with phosphorus MRS, which is this modality, is redox biology. So you can look at oxidative stress and reductive stress, and there's big-time abnormalities in the redox ratio in schizophrenia. So on the left, you see the chronic schizophrenia, where you see the average ratio in healthy people is almost 5, if you look at the y-axis there. And in people with schizophrenia, it's 3. Here, it's a 40% abnormality. And then first episode, schizophrenia is even lower. It's almost down to 2. So very significant redox imbalance with oxidative and reductive stress. The details get complicated, but it's there. We've also been looking at the impact of these abnormalities on brain connectivity. Some of you may be familiar with this kind of image showing brain networks, different brain regions that are all connected to one another. And the warm colors here, the reds and oranges, is the default network, the default mode network, and the blues and greens are task-related networks. And these two networks are distinct from one another. We all have them in our brains, and when you're doing internally-oriented processing, like thinking about where you're going to have lunch today or thinking about something that happened yesterday, the default network comes into action. And when you're engaging externally, like solving a crossword puzzle, the task-related network comes into action. So these are both happening all the time in our brains. It turns out that energy metabolism actually controls these networks. So again, without going into the details of this complex slide, on the x-axis here on the bottom is flux. That flux is the ATP synthesis that I just showed you. So we know it's abnormal in our patient population, but ATP synthesis also seems to control brain connectivity, which makes perfect sense, because you need energy to connect your neurons. We already talked about that. So our findings are very clear that this energy problem... This is, by the way, in early psychosis, first episode schizophrenia and first episode bipolar patients. So the energy metabolism problems are leading to brain connectivity problems as well. We also have looked at insulin sensitivity, which is in the rest of the body. So in this study, we asked the question, OK, so if in the brain, glucose metabolism and energy production is poor, what about in the rest of the body? And you can measure that using the glucose tolerance test, which pregnant women undergo to test for gestational diabetes. So many of you may be familiar with that. But you can do that in other patient populations as well. And as you can see on the left side of the slide, healthy controls have an elevated insulin sensitivity, which is the normal way it's supposed to be. Patients with first episode psychosis have half the insulin sensitivity. And remarkably, their siblings, who are healthy, lean young people... In many cases, we had about 18 or 20 people in that sample, in the siblings. These are well above average kids. They're in, you know, high-powered universities or have gotten their first jobs and are doing very well. And yet they carry this trace of a glucose metabolism abnormality. Their bodies don't respond to insulin the way they should. They're not prediabetic, they're not overweight, they don't have diabetes and so on. So it's actually this vulnerability that seems to track with what their siblings, who are the patients, are experiencing. And also insulin sensitivity in this data set is related to the cognitive testing scores, the matrix composite score, which is a test of global cognition. So it's not just a brain problem either, it's also a body problem. The two are related to one another. So energy metabolism is really a key component of these abnormalities. This is work from the United Kingdom, where they were looking at fasting insulin levels and body mass index, so insulin on top, insulin sensitivity, and BMI on the bottom. And when you take developing kids, you start measuring them early on in life, so the x-axis here is age, you see different classes, different groupings of kids. And, well, guess what? The kids with the high fasting insulin, class 3 that I circled on the top panel, are kids that were at multiple-fold higher risk of being diagnosed with schizophrenia once they made it to young adulthood. So the insulin sensitivity that I just showed you in our own data set is there also in these population-based data sets from the United Kingdom. And then the bottom one, the body mass index, there's a class 4 there, which is actually the kids who had normal BMI in early childhood, and sometime around age 10 or 11, they start to increase BMI, and those kids actually end up having a higher risk of depression. So body metabolism actually is very closely linked to psychopathology. So the implications of this, of course, are that we see major abnormalities in brain bioenergetics and psychotic disorders. It's the machinery underlying neurotransmission and synaptic plasticity, and it's compromised. And there's a suggestion of abnormalities in the molecules, the proteins that actually generate the ATP as well. That's not our work, but from others. So I have a couple more slides, but in the interest of time, I'm not going to go into them. But these are dynamic emerging abnormalities. Again, I'm not going to march you through each of these curves. Suffice it to say that some of these things are also happening in the prodrome. And in fact, just now we have a new grant where we're starting to collect data in the prodrome to measure energy metabolism as well as circuit-level activity. And the early data that we have already is showing that these things are already happening in high-risk individuals even before they've developed psychosis. And then I'm going to skip this slide because it'll take too long. A few things that often come up. We see no evidence of medication effects. Because these are biochemical abnormalities, it's reasonable to say, well, what about what... You know, these people are all taking antipsychotic medications and a variety of other things. Could that be causing the energy problems? So a few answers to that. One is that we always look at the medication load, how much medication patients are taking, and we never see a relationship between that and the energy measures. We have a small group of patients who volunteered for our studies who are not taking any medications, and their results are right in the middle of the rest of the patient group, so they're not different somehow, suggesting that people who are not taking medication don't have a different pattern. And we just completed a study of olanzapine in healthy people. Our ethics review board allowed us to give 5 mg of olanzapine a day for two weeks to healthy people. They wouldn't allow higher doses or higher time periods, which is reasonable because olanzapine can be such a toxic medication and cause weight gain and lots of other problems. So this is the best we can do. And then we did these energy brain scans before people took the olanzapine and at the end of the two-week period, and olanzapine did not cause any changes in these energy measures. So at least 5 mg of olanzapine doesn't actually hurt these energy measures. Just another piece of evidence suggesting that these energy metabolism problems are not an epiphenomenon of taking medication, but are probably associated with the illness themselves. And of course there's the sibling study that I just showed you. You know, the first degree relatives of our patients also show some of these abnormalities. Do I think that this is really what schizophrenia is all about? No. There are similar abnormalities reported in epilepsy, dementias, autism. It's likely to represent a general vulnerability for brain function. So I wouldn't say that schizophrenia is a metabolic problem, an energy problem. And if you have brain energy problems, you're going to get schizophrenia. That's really too simplistic. I think these probably are underlying vulnerability factors. What they mean is that you might have a brain, you know, complex brain disorder like epilepsy, like autism, or like psychotic disorders. The exciting thing about these energy studies is that there's lots of interventions that we know from diabetes and other fields about how to improve people's energy metabolism. There's ketogenic therapies in the middle of the slide. The ketogenic diet or the keto diet may be familiar to some of you. It actually improves brain energy metabolism. And there's some evidence already emerging that it might help people with psychiatric disorders, especially with psychosis, schizophrenia, and related conditions. So we just started a clinical trial of a ketogenic diet for first episode psychosis. And we're doing those energy scans, the MRI scans, before and after to see if ATP synthesis is improving after the keto diet. There's ketone salts, NAD supplements, and even intranasal insulin can be used. On the bottom right is intranasal insulin. It turns out that if you spray insulin into your nose, it doesn't get into the rest of your body and drop your blood sugar. It actually gets absorbed. Remember the cribriform plate and the olfactory bulb. So it actually gets absorbed into the brain directly. And that improves brain energy without causing any changes in the rest of your body. So we just completed an intranasal insulin study as well. So the results are still pending. Stay tuned for that. Okay, I just have one conclusion slide to just remind you of. It was a long talk. I had a little extra time, so I went into a lot of detail. But, you know, the point is that early phase of illness is a critical period for intervention to improve outcomes. It's not the beginning of a downhill slide. It's a time, a critical time, when we can do good things for our patients. But, of course, we have to learn a lot more about this. We talked about many of the controversies and the challenges. And it has to be founded on the clinical and the neurobiological reality. And that's where this notion that the early illness is actually a fluid and malleable time comes from. So there is much more to learn. But I think it's really an exciting time in early psychosis. And I just will end by thanking my entire team and all of our patients and their families who volunteer for our research studies as well. Thanks for your attention. Hi. Thanks for a great talk, first of all. I really appreciated it. I'm Howard Margulies from Montreal in Quebec, Canada, where we have an extensive first-episode network of clinics. And, you know, I started my clinic many, many years ago. And I started my clinic when I was 18. And, you know, I started my clinic many, many years before the government got involved. But it was really only when government got involved that we had the network to be able to provide the care to most of the population. So do you think the problem in the United States is one of more organization rather than willpower or, you know, either, like, the will, the desire, the expertise is there, but is it an organizational problem and a funding problem more than anything? Yeah. I mean, you know, we ask frequently, what's the problem with the United States, right? There's so many riches and so much expertise in the health care system in so many ways. And yet the outcomes are not what they should be. So, yeah, I think it's both. First of all, you know, the population is large. And the resource per capita that's going to early psychosis is actually still low. So it's under-resourced. You know, we should acknowledge that even though the absolute magnitude is large, it's actually still under-resourced. So there's a need for more resources. But even when you do have the resource, it's not being applied well and evenly. You know, you think of a state like California with 50 million plus residents, and they have the county system. And, you know, each county is doing it a little differently. And they don't have the ability in these massive counties like Los Angeles County to really monitor what the clinics are doing and make sure everybody is delivering the same high-quality care and so on. So, yeah, I mean, it's a combination of the resourcing but also the organizational problem. And it's just, you know, the original sin of the American health care system is that it's fragmented and it's poorly regulated. A follow-up question is are you involving people with lived experience in all phases of your program, including the research aspect, including like developing the research questions and that aspect? Yes, we are. So, you know, we were not the innovators in this field. We're latecomers. But we have now a very robust group of certified peer specialists, as they're called in Massachusetts, involved in care delivery, both in the early psychosis program but also in the inpatient units and all levels of care. And they are involved, you know, we have a research advisory board formed of people with lived experience and family members. That's where we go present ideas before the grants are written, before the projects are started. So we're doing a better job with that now. But, again, you know, I wouldn't take credit at McLean for having figured this out. You know, this is obviously happening elsewhere as well. Hi, Fran Cornos, Columbia University. Thank you for a wonderful talk. I run a program for people with schizophrenia. I retired from it. And I've also worked at the interface of HIV infection and schizophrenia. And one of the things we do in HIV is we look at people, the subgroup of people who don't progress despite having HIV. And I was involved in a project for the group for the advancement of psychiatry where people wrote in about being told they had schizophrenia and throwing away all the medicine and all the care and getting better. And it just makes me wonder if we might be doing in schizophrenia what we do with HIV, which is to try to understand what subgroup of people get better on their own and what makes them get better on their own. I think that's an excellent question. Couldn't agree with you more. I don't know why it doesn't get done. You may have been involved with Ellen Sachs, and she's been trying to develop a network of people around similar ideas. I'll tell you one anecdote from our first episode clinic where we do all this research that I showed you as well of a poster child patient, somebody who came in with a psychosis diagnosis, non-affective, looked like schizophrenia spectrum, went on the medication, did much better. A year later, was back in full-time college at a high-powered college and doing well. So the psychiatrist, thinking that they're very enlightened and progressive, said, you know, you're doing so well. If you want to talk about lowering the medication dose so that you become even more functional, I'm happy to do that. Just let me know when you want to start talking about that. And the patient goes, I'm going to tell you something. I haven't been taking the meds the whole time. So, you know, right, these are the ones that we know. There's plenty more out there. What that says to me is not that the meds are junk and we shouldn't be taking meds, but rather that outcomes are determined not by the meds but by patient-specific factors. There are some people out there, just like you said, who are non-progressors, and they do very well. But we, you know, we have our blinders on, so we think, you know, diagnosis, treatment, problem solved. And so we're just giving them what we know how to do and then not actually being curious about what else is going on out there. And I just don't know why it's not being done more, but there are absolutely these people out there, and we really should be learning from them. Thank you. Hi, I'm Allison Swigert. I'm a psychiatrist at a first-episode psychosis CSE in Harrisburg, PA. I've been doing that about four years. I really appreciate your talk and the overview. My question is sort of maybe any tips about sort of language for how to talk to patients and families about some of the most difficult questions that we get are from family members, which is sort of like, when will my kid go back to normal? And then the questions around how long to take medication for. And I think my struggle is around, you know, how to sort of balance conveying the seriousness of this and the importance of preventing a second episode or a relapse while also maintaining hope and, you know, the possibility that things will improve. So just wondering if you have any sort of like tips on language or types of explanations you use for patients or families around that. Well, I think you pretty much said it right there, really. So one thought is, you know, the Navigate handouts actually have some of this. You know, metabolizing the first episode, there is quite a bit about this tension between, you know, the goal is recovery, and we're hopeful that everybody who walks in this door can actually recover and do well. But, of course, we know that's not going to happen to everyone, and this is a serious problem, serious illness, and we have to respect it and make sure we're doing everything right. So that's the kind of language that they use there. And this is what I was referring to when I said you can't be a Pollyanna about this. You know, this is a serious problem. The idea is not that, oh, hey, everybody's going to do better. You know, it's going to be fine. The idea is if we don't project hope and if we don't try, then the outcomes will uniformly be bad. If you're trying, you're giving people a chance for the ones that might actually improve. So I think that's the kind of orientation and the language that we talk about. Hi, thank you for the great talk. I'm Laura Safar from Massachusetts. I have many questions, but I'm going to select four. So bear with me. So one is about the slide where you have the comparison between the community centers and the special programs for early psychosis. The community centers, it seems they still did okay. So my question is for the rest of us out there who are not in an early psychosis program, what are the key elements that you think we should pay attention to? So medication, taking the patient as a whole, paying attention to family, being part of the treatment. So the things we do as psychiatrists kind of thing. I don't know if you have any comments on that. So this is what I was emphasizing at the very beginning. There's no special potion. There is nothing magical. There is no new treatment. This is team-based care. This is remaining engaged and getting the patient and the family educated and delivering the highest quality available. That's really what it's about. So no, I actually think what you just said is exactly right. And then there are two questions. So one is imaging, right? When do we do brain imaging? It's an all-time question, right, in early psychosis. The other one, I am particularly poor at helping patients who are using cannabis and have first episodes of psychosis. As the previous colleague asked, any tips of how you talk about this? Yeah. Imaging, yeah, I'm glad you asked. I probably should have brought that up. You know, in the early days, going back 20 years, the recommendation was that every first episode patient should actually get brain imaging, an MRI scan to establish a baseline and also to potentially identify any reversible causes of psychosis. The reversible causes of psychosis are relatively rare. There is still a little bit of controversy in the literature, but by and large, the recommendation for imaging has been removed. So we are not doing imaging for every patient that comes into our first episode clinic. You never find anything. It's just not going to happen. Or what you find are these nonspecific white matter abnormalities, and you don't know what to do with it, and there is no treatment implication, so we all move on. And imaging is still costly. It was costlier, but it's still costly. So right now, the general agreement is no imaging, although there are a few centers that think differently. And then in terms of cannabis, yeah, that's a tough, tough problem. I had a patient once say to me, I will never give up my cannabis. I mean, this is just such a reinforcing, addictive thing. It's not about the instantaneous pleasure effect the way it is with cocaine and opiates, right? It's actually this long-term behavioral addiction in its own way. And so people will sometimes be like dedicated to the cannabis at the expense of anything else. They are prepared to give up anything else except for cannabis. It's pretty wild. So the things we try and, of course, don't always succeed are, you know, you were using cannabis. Things were bad. You came in the hospital. You stopped using cannabis. Now, you know, three weeks later, four weeks later, things are better. How about a trial period? Let's give it a few weeks. Let's see how you're doing. If you say to a 19-year-old, you will never do this, the first thing they will do is go prove you're wrong, right? So developing that kind of adversarial relationship will never help. So looking for those kinds of openings, agreeing to trial periods, harm reduction. Even if they are using, can you prevent it from getting out of hand? You know, parents getting involved, you know, one joint on a weekend may be okay in the long term. You don't want any, but if it's going to happen, try to reduce the harm. So those are the things we're trying. I'll tell you, they don't work very well, and the ones with poor outcomes are the ones who don't give up cannabis. So there's a really clear relationship to outcomes. That's probably not causal. You know, like the ones with cognitive impairment and poor adherence and poor insight are using cannabis, and they also have the poor outcomes, but it's clearly marking out people in our experience. Thank you. And if I may, one more. Any data on the anti-psychotics? Is there any preference in terms of selecting? We always select based on, say, the profile, et cetera, et cetera. Any data on selecting one versus the other? Nope, not at the moment. Thank you very much. Among those second-generation anti-psychotics, yeah. I appreciate your comment about the keto diet, but that leads to another set of issues that families have brought up, and that's actually in the popular media frequently, are other forms of megavitamin or nutritional supplementation, omega-3s, large doses of various nutritional supplements, acetylcysteine. Do those play any role in an adjunctive treatment for this group? So the evidence base for all of these things, including the keto diet, right now is thin. There aren't any well-done, large, controlled clinical trials that say that the keto diet will lead to better outcomes. There is just mostly case series and a couple of small clinical trials showing some improvement, and then one long-term problem with the keto diet is it raises your cholesterol, and if you're a 22-year-old looking at 30, 40, 50 years of living with this illness, it may or may not be a good thing. And then the only, I mean, omega-3s, exciting early study, definitive negative later studies, so we're not using omega-3s anymore. It's really not associated with better outcomes. Acetylcysteine, probably better than omega-3s. There are definitely some well-done studies showing some improvement, but you have to take high doses for very long time periods. I mean, talking about six months or a year before you start to see the two groups actually diverge from one another, so it's unlikely that that's going to be a game changer. And I'll tell you, at our clinic, some of our clinicians prescribe acetylcysteine, and I've never met a patient who was doing poorly, went on an AC, and did better. It's always like, well, you know, are you sure? Maybe it's a little better. I'm not sure, so it's not a strong effect. And then these, you know, megavitamins, the B12s and some of these other things, the evidence base is even weaker. It's just not, you know, so there's not much we can say about that. We're not doing any of that. Okay, I think we're over time. Well, thank you for your attention and for staying the whole time. Thank you.

psychotic disorders

early intervention

schizophrenia

bipolar disorder

neurobiology

Coordinated Specialty Care

RAISE project

brain energy metabolism

Phosphorus MRS

ketogenic diets

Dr. Dost Angur

National Institutes of Mental Health

patient outcomes