The Next Big Breakthrough

What if mental health care worked more like cancer treatment—tailored to the individual, informed by biology, and driven by data? Charles Marmar, MD, Chair of Psychiatry at NYU Grossman School of Medicine, takes us through the latest advances in precision psychiatry. From brain imaging to digital phenotyping, Dr. Marmar outlines the tools shaping a future where treatment is fully customized. He also shares stories from the front lines: a patient whose depression was treated with the help of a brain biopsy, why PTSD and depression aren’t one-size-fits-all, and how quantum computing could radically accelerate psychiatric research.

🔍 Topics Covered

00:00 Introduction
02:24 Personalized Treatment Approaches
05:13 Challenges in Precision Psychiatry
11:12 The Role of Genomics and Brain Imaging
23:35 Digital Psychiatry and AI
25:57 Quantum Computing in Psychiatry
33:04 Future of Precision Psychiatry
37:22 Conclusion

📚 Related Resources

Charles Marmar, MD – NYU Faculty Profile
NYU Langone Psychiatry
Adriana Heguy, PhD – NYU Faculty Profile
Javad Shabani, PhD – NYU Faculty Profile

🙌 Support & Subscribe

If this episode challenged your thinking, like, share, and subscribe to help bring smart, science-forward conversations to a wider audience.

Visit our website for more Insights on Psychiatry.
Watch this episode on YouTube
Executive Producer: Jon Earle

Transcript

DR. THEA GALLAGHER: Welcome to the Insights on Psychiatry podcast. I'm excited to introduce our guest today, Dr. Charlie Marmar, who is the chair of our psychiatry department here at NYU Grossman School of Medicine and NYU Langone Health. Dr. Marmar, thank you so much for being with us today. 

DR. CHARLES MARMAR: Dr. Gallagher, it's a great pleasure to be with you. And Thea, it's wonderful to have you as our leader in these communication efforts in our department. I especially have fun working with you on our shared innovative research projects to advance precision psychiatry. So perhaps we can talk about that too. 

DR. THEA GALLAGHER: Yeah, absolutely. And what's really interesting is that we were able to talk two years ago about precision medicine, especially in psychiatry. Can you fill us in? What have been the advances in the last two years in precision psychiatry and are you happy with the pace at which things are moving?

DR. CHARLES MARMAR: I would say the major advance, separate from specific studies which aim to try to personalize treatment rather than to apply a one-size-fits-all, or I would say to go beyond the traditional way in which we've tried to provide treatment guidelines, so we have over the last 30 years or longer been evolving specific guidelines for the evidence-based practice for the treatment of stress, anxiety, depression, psychosis, addiction, et cetera. And if you look at the guidelines that have been published by the American Psychiatric Association, the American Psychological Association, the World Health Organization, and many others, they will generally say it is recommended for a patient with a new onset of depression to start with the first line treatments. If they're successful, great. If not, go to the second line, then go to the third line and then if the patients are still very difficult and refractory to treatment, use specialized approaches and special assessments and consultations for the very complex cases. That has been helpful to some extent. It's certainly an advance over one-size-fits-all treatment, but it's not precision medicine in the way that, let's say, cancer is treated now. 

And the major advance, I think, Thea has been a growing recognition in the field of psychiatry of the urgent need to develop precision medicine so that if I refer you a patient with depression, you'll be thinking about that patient not in terms of first and second and third-line treatments, but who is the person I'm working with today? What are their unique biological features? Age, gender, ethnicity, other features. What are their unique psychological features, their conflicts? What are their unique social and cultural features? Are they highly educated? From stable families? Are they from vulnerable backgrounds? Did they grow up in poverty, in violent neighborhoods, and really were disadvantaged in many ways, didn't have good health care growing up, and maybe not a great education, they may all present. A younger woman, let's say, South Asian woman from a highly educated, prosperous family, presents with sadness, loss of interest, difficulty sleeping, concentration problems, feelings of despair and hopelessness. An older Caucasian male who's had a history of head trauma and comes from a violent neighborhood and from a poor background, hasn't had great education or the privilege of excellent healthcare, and may also have the same symptoms, but clearly they're very, very different.

We talk, Thea, about the bio-psycho-social model, but we're now beginning to ask the question, how are the unique biological, psychological, social, and cultural characteristics of my patient that I'm working with today and how do I adjust the treatment? So there is much more buy-in to that. There's progress in some studies. We're beginning to find certain clinical features, cognitive features, and even genomic and neurocircuitry features will help us say which treatment for which person at what time. But the biggest revolution is the understanding that we have to move in psychiatry to where cancer has moved so successfully. 

DR. THEA GALLAGHER: And almost moving away from trial and error and moving more toward gathering all the data early on and then having a more precise treatment plan for that patient moving forward and not just, hey, does this work for you or not? And then you have this pile or this grouping of treatment refractory patients who haven't benefited and I think precision medicine is saying, we want to make sure that we're not having this group of treatment refractory patients that we don't know what to do with just because they didn't benefit from the first, second, third line treatment. 

DR. CHARLES MARMAR: Exactly. And I would say if we go back 30, 40, 50 years into the history of the treatment of psychiatric disorders, psychological disorders, not only was there a trial and error approach, there was something even more problematic, which there was a school-based approach. So I remember early on as an educator, asking a group of residents when I was maybe junior faculty, and I had a group of wonderful residents at the time. I was an assistant professor at the University of California, San Francisco, UCSF, one of the great medical schools. And I asked the residents I was teaching, I presented a case to them and it was a case of a patient, let's say, with a third occurrence of a major depressive illness. And I asked them each to present what they thought would be the best care plan for that patient. And the response actually was not to start with one common treatment, it was completely school-based. One resident was very interested in psychoanalytic psychotherapy, and they started with a psychoanalytic formulation. Another was very interested in neuropharmacology and psychopharmacology, they started with a medication approach. Another was a behaviorally interested person, they started with a cognitive behavior therapy, and a fourth was becoming interested in some other approach. And fundamentally, the treatment that they would have given that patient if referred to them was actually not based on the specific needs of the patient, it was based on the interest, passion, and bias, so to speak, of the clinician depending on the school. 

So in practice, if a patient with depression at that time was referred with depression to a psychoanalyst, they would get psychoanalysis whether they had bipolar depression or not. If they were referred to a psychopharmacologist, they would get psychopharmacology whether or not the origins of their problems was in childhood trauma. If they were referred to someone with another point of view, they would get that approach. And I was thinking to myself as an assistant professor at the time, imagine if I was referring a patient with atrial fibrillation to a cardiologist, and the treatment for that arrhythmia was determined by the cardiologist's preferred school of training as opposed to the basic needs of that patient. 

DR. THEA GALLAGHER: I think what we were saying is, the more we can understand kind of all of the factors, then we can have a more precise treatment plan and I think this is also really important. You think of how many people have comorbid conditions. Then you add another layer of maybe that trial and error because we're not looking at treating conditions maybe in concert. 

DR. CHARLES MARMAR: We have to look at what are the presenting problems. We have to look at all of the complexity of the medical illnesses, neurological illnesses. But we have to look also at education, culture. We have to look at life experience. We have to look at also biology, including sex. For example, there's some developing data to suggest that if a man in their mid-30s and a woman in their mid-30s both present with relatively similar depression symptoms and they are suitable for initial treatment with medication, that on average, the woman will do better with a selective serotonin agent anytime from menarche to menopause and the man maybe do better with some mixed receptor agents, drugs like Wellbutrin, Bupropion, venlafaxine, Pristiq, and others because of the differences of biology. And interestingly, after menopause, a woman who may have done very well with a selective serotonin drug may do less well and may need a drug more like the 35-year-old male may need, unless they're also on estrogen replacement therapy after menopause. So forget about all the great complexity of culture and genetics and family background and brain imaging findings, just the simple age and sex may be profoundly important. 

DR. THEA GALLAGHER: And it makes so much sense. I guess the bigger question is, how are we getting closer to that? And what are you excited about that's actually happening, getting us closer to taking all this data and understanding the psychiatric complexities in an individual?

DR. CHARLES MARMAR: Two things have happened which are very, very favorable. One is there's a growing consensus among clinicians, students, and researchers that this is essential to move our field together. The second thing is that there's a growing consensus about the agencies that provide the funding for the research to advance, the NIH, DOD, VA, DARPA, and other agencies that fund the majority of psychiatric and psychological research. They are also becoming very, very interested and initiating requests for applications for precision psychiatry proposals. And the final thing is, which is very interesting, is our field is maturing to give us the tools we need. Leaving aside simpler things like age and sex and education and cultural background which we've been able to assess for a long time, of course, we now have sophisticated human genomics to study the genome in psychiatric illness. So more and more clinical trials now include collecting DNA and RNA so that we can study the genomic contributions. We also have much more sophisticated, non-invasive brain imaging. And we have a whole series of other fascinating ways to try to understand the neurobiology of psychiatric illness.

For example, there are brain banks of people who've been visionary and have said at the time they pass away they'll contribute their brain because they've struggled a lifelong history of bipolar disorder or PTSD or schizophrenia or some other alcoholism. And those studies, those postmortem studies allow, of course, for a very detailed interrogation of brain structure and functioning. Those become published and part of public databases, and then researchers such as myself and others interested in precision measurement can use those as reference samples to say, oh, so these specific genomic pathways may be disturbed differently in these disorders. Now I have the DNA or RNA of my patients in my trials. I can be informed by the postmortem reference samples to know where to look in the genome in my patients. There are many other ideas. We go beyond blood. We're starting to learn about exosomes, which are actually coming directly from the brain and have genomic cargo on them. We are doing studies, we're planning to do some collaboration at NYU with colleagues at the Mayo Clinic in which we're harvesting live cells in our clinical trials and those cells will be blood cells, but they can be programmed backwards to stem cells and then forward to neurons. And we can even grow brain organoids, sort of brains in a dish and we can then study those brains, we can expose those brains to alcohol, we can expose them to drug treatments. So we're beginning to triangulate all of those things with animal models, and I believe we'll be able to advance our field and become precise. 

The goal is to do in psychiatry what is now done in breast cancer. You remember how breast cancer was treated in the '50s? Do you have a sense of how, if someone presented with breast cancer, how it was treated then? 

DR. CHARLES MARMAR: No.

DR. CHARLES MARMAR: One-size-fits-all. Everyone got some combination of surgery, radiation, and chemotherapy. It was not specific to the person or their genomic tumor, and it was often associated with great side effects and often not great outcomes. But we didn't know beyond the fact, we knew that radiation could help some women with breast cancer. We knew that chemotherapy could help, and we knew we had to remove the tumors with surgery where possible. But we didn't know in precisely for whom. If a patient has breast cancer now, obviously, we biopsy the tumor, we find its genomic features, and we begin with treatments that are highly specific to that form of cancer. As a simple example, estrogen receptors are studied, progesterone receptors are studied, and a set of receptors called HER2 receptors are studied and if the patient is estrogen negative, progesterone negative, and HER2 positive, they're likely to be started with the drug Herceptin, which is a monoclonal antibody specific to their tumor, does not attack other parts of their body, does not attack healthy cells in their gut or their brain or elsewhere and think of the enormous advance. This is a treatment specific to their form of cancer, which targets the vulnerability of those cancerous cells and does not attack the entire body of living cells. 

DR. THEA GALLAGHER: And it seems like psychiatry has not been able to have those advances in part because the animal models hasn't been a possibility. But it seems like with technology and all these opportunities to gather more data in these larger systems, or as you said, even after their death, it might help us to kind of make more of these precision medicine models. 

DR. CHARLES MARMAR: It's a wonderful observation. I think there are two major reasons that have limited—three—that have limited the advance of precision psychiatry. The first has been a failure of imagination to understand that we could do that in our field. That's changing. The second is we have animal models, but many of them are not very good representations of the illness. There are, so to speak, animal models for psychosis, but rodents don't have delusions, so they're very, very difficult. You have to do an animal model of, let's say, a circuit disturbance in the hippocampus, but not the illness itself. We have several diseases that are quite amenable to animal disease modeling. The best are trauma because the same circuits in rodents are altered when they're exposed to predator threat as are humans when they're exposed to combat. So that's very helpful and perhaps the very best animal models are alcohol and drug abuse because, for example, rodents will become addicted to cocaine. They will become addicted to alcohol and to other stimulant drugs and there you can then understand the animal models are very close to what we need to do in humans. 

But the final thing that has held us back up till now is in precision medicine, in all other areas of medicine we biopsy the organ of disease—breast cancer, prostate cancer, we have access to arrhythmias, we have access to the tissues of disease. In the case of human neuropsychiatric illness, except under unusual circumstances, we don't biopsy the living brain of the patient and what we're trying to do is approximate what is happening in the brain by looking very far downstream. You know, Thea, it reminds me of that story about the person who lost their keys at night, and they're looking under the lamppost. But they didn't lose their keys. So they were asked, why are you looking under the lamppost? Well, that's where the light is. But the keys are not there. So the problem for us is when we study DNA and RNA in peripheral blood, the expression of DNA in peripheral blood cells is not necessarily closely related at all to single-cell gene expression in specific neurons, in specific areas of the brain, in a given specific psychiatric disease. So the question is, can we find a proxy for the brain biopsy? We call them just for ease of communication, a liquid biopsy. Is there something in the blood that actually reflects what's going on in the brain? We're very interested in exosomes for that reason, because they're in peripheral blood. They can be isolated. They are vesicles that carry a genomic cargo and they have some genomic features on them or markers that would tell us more or less where they're from. Are they from the brain or the liver or the kidney? If they're from the brain, where in the brain? Are they from the hippocampus, the amygdala, or the basal ganglia, or the frontal cortex, PFC? And maybe even more specifically, among those that come from the amygdala, are they from the central nucleus of the amygdala or another nucleus in the amygdala? 

So that's interesting because in a way, those features are then much closer to what may be dysregulated in the brain. We also like the idea of these postmortem reference samples because they give us clues of where to look because we actually have the genomic features from specific brain cells and brain regions and we like the brain organoids also because they replicate actual brain structures. There may be other strategies, but when you ask me what's held us back, a lack of understanding that this is critical, that's changing, a failure of animal models to closely replicate human brain diseases in psychiatry, particularly diseases with cognitive disturbances such as in psychosis, and then a lack of ability to actually biopsy the organ of illness. We are the only field, psychology and psychiatry, are the only field in medicine that does not routinely biopsy the organ of illness. 

DR. THEA GALLAGHER: It sounds like what you're saying is you know brain scans are not the answer in precision medicine necessarily. We need scalable interventions or scalable ways to assess looking at, like you said, these records or gathering data in these large-scale medical records.

DR. CHARLES MARMAR: I would say brain scans are becoming more sophisticated. PET scans, MRI scans, other kinds of scans, SPECT scans. We have a lot. We're making progress. They provide useful information. I would say studying DNA and RNA and peripheral blood is useful. It provides information, but it's not going to be at the level of specificity of actually…for example, you and I have a shared deep interest in post-traumatic stress disorder. We have some pretty good ideas from animal models which circuits are involved. We know the amygdala is involved, we know the hippocampus is involved, we know the insula is involved, and we know the prefrontal cortex is involved, among other areas. But when we have—you and I are collaborating together in the clinic on a patient who's not responding well to treatment for complex trauma—we do not biopsy their amygdala to find out what's different about this patient. That's the main limitation we have. That's why we are much less advanced than the treatment of prostate and breast cancer. 

DR. THEA GALLAGHER: And the hope is that we find ways to make it more accessible and precise by looking at some of these records as well. 

DR. CHARLES MARMAR: Yeah, alternative models, more accurate, sensitive animal models, post-mortem reference samples, exosomes, brain organoids, and other things we haven't thought of yet. But the point for our listeners is very straightforward. Until we can accurately reconstruct what the neurotransmitter and neuro circuit disturbances are in these illness in a patient-specific way, we will be somewhat limited in precision medicine.

DR. THEA GALLAGHER: And ultimately what do you hope? That we are able to get that information or do you hope that in combination with other factors? I know you're into facial action coding and voice markers, using other pieces of data to put it all together into a larger model to get more precise if we can't do these biopsies?

DR. CHARLES MARMAR: Well, so great point. So in addition to advances in genomic psychiatry and imaging psychiatry, you and I and others are working together to advance digital psychiatry. The advantage of digital psychiatry, of studying voice quality, of studying the structure of language with large language models using AI, studying facial expression, using facial action codings, I was very influenced by this, interested in this because in my years at UCSF, I got to know and had the privilege of knowing and becoming friendly with and asking for research advice. One of my senior mentors, Paul Ekman, actually developed the Facial Action Coding System. Paul is a remarkable person. He actually came to the conclusion that there was a limited number of unique human facial expressions. He mapped out the muscles that were involved in creating those facial expressions. And then he went to multiple different cultures, including South Pacific Island cultures, Asian cultures, many different cultures, and found that even though cultural differences were great, even though language was different, that actually the same range or palette of human emotions, anger, sadness, shame, etc., were expressed in all these cultures in the same muscle groups, which is a fantastic discovery. So Paul really inspired me to do this work. 

But we can also use cell phones, which give an enormous amount of information. We know that when people are depressed, their geolocation is different. They do less things. They stay home more. They retreat into bed. They avoid social situations. All of that's very interesting and valuable. And the final area, not the final but another major area which is going to advance precision psychiatry is computational modeling. The ability to use advanced AI-informed machine learning and very fast computers. I'm beginning a collaboration now, which I'm hoping will be radically interesting. I have a colleague, a renowned quantum physicist, Dr. Javad Shabani at NYU, who is one of the world's leaders in developing quantum computing. His laboratory at NYU is developing next-generation qubits, which are like the chips for quantum computing. Quantum computing, even in the kind of rough and inaccurate proof-of-concept models can handle vast amount of information and process it in lightning speed time. So we're going to collaborate together, Dr. Shibani and I, on a project in precision psychiatry. In this particular study, it happens to be a clinical trial that was completed by NIAAA, 346 subjects. We have DNA and RNA on every one of them before treatment. They were treated with gabapentin, an anticonvulsant drug, which reduces alcohol craving and we also collected RNA after treatment because it changes with treatment. Okay. 

So I will ask you a question to be playful. How many unique variables do you think we have on each of these 346 participants, patients, that we could model to try to understand why some people respond well and others don't respond well to gabapentin for AUD, for alcohol use disorder? Take a guess how many uniquely different variables are in the database on each of these patients?

DR. THEA GALLAGHER: Two hundred.

DR. CHARLES MARMAR: Okay. How about a little higher? 

DR. THEA GALLAGHER: Okay. 2,000. 

DR. CHARLES MARMAR: How about a little higher? 

DR. THEA GALLAGHER: 5,000. 

DR. CHARLES MARMAR: Twenty-two million. We have 22 million unique features because in designing this study, I had identified, I and my team, other colleagues, had identified 1,300 genes in the human genome that we thought might be relevant for understanding how people respond or fail to respond or get side effects or don't get side effects to gabapentin for the treatment of alcohol use disorder. And I asked the director of our wonderful NYU Human Genomic Center, Dr. Adriana Heguy, who's a great colleague of ours and part of the study to estimate what the cost would be to sequence those specific genes and their transcripts. And she gave me a price. It was pretty high and she said, you know, Charlie, we can do something much more interesting because we have now the sequencing sophistication to sequence the entire human genome, 22 million features, for less money than the targeted sequencing of those 1,300 genes. And of course, you'll get those genes in their transcripts, but you'll get all 22,000 coding genes in their transcripts. And you will also get all of the so-called dark matter in the human genome, which plays an important regulatory role, even though those genes don't code for proteins, and it directly affects cell function. They have a role in managing how the 22,000 genes express themselves, okay? Chromatin remodeling, transcription factor regulation, post-transcriptional changes of the structure of transcripts of proteins, etc. So we have it. And we have also, of course, clinical features. We have other things, but the big data is the genome itself, right? 

So now we have 346 people with more than 22 million features. Now, what were you taught in your statistics courses about how many subjects you needed for the number of variables? In other words, in traditional statistics, you're limited by sample size, and you need a certain number of subjects, the variable ratio, right? And do you remember what you were taught about that in conventional statistics? It's about, you know, you need maybe a 10 to 1 ratio or something. In the world of AI, there's not actually a limit on the number of features you can have per subject. There's a limit on the computational resources needed to manage that information and make sense of it. And we think that quantum computing will help solve that problem. So that's the next thing. But even using NYU Langone Medical Center's high-capacity imaging, we can do pretty well. But right now, when we run a machine learning model to say which genetic features or which imaging features or which clinical features or cognitive features predict who will and will not be a likely responder to a treatment, let's say gabapentin for alcohol use disorder, it can take weeks to run those models, and these quantum computers can run them in a fraction of a second.

DR. THEA GALLAGHER: And can make sense of, like you said, twenty-two million….

DR. CHARLES MARMAR: And the other thing about quantum computing is it probably has the capacity to create a much more sophisticated map of the interrelationship of these factors than conventional computing. So it's not just light-years faster, it's also going to be more precise.

DR. THEA GALLAGHER: And is the goal you think to get even down to go to the complex to get more simple because it's so many pieces of information and it might be over fitting the model or the data, is the goal then to kind of know what is actually most important and then we can kind of pull that out? 

DR. CHARLES MARMAR: The goal is to do some targeted hypothesis testing with a smaller number of features, then do a very broad exploratory model with all the features and then reduce those to the most important features. Part of the output from AI models is a variable importance plot which will read out among these 10,000…100,000…million…10 million…20 million features or more, which five are the ones carrying the signal? And then we can study those and we can ask which medications or psychotherapy or TMS, neuromodulation is likely to move those features. 

DR. THEA GALLAGHER: So going from the complex to the more simple. And in our final question here, looking ahead, what are you most excited to see happen in precision medicine in the next decade? 

DR. CHARLES MARMAR: I think we're actually going to do two dramatically interesting things. The first is we're going to take psychiatric illnesses which are not unitary, they're heterogeneous. For example, we already know that there's six or seven or more forms of depression. There's bipolar and unipolar, there's depression which involves psychotic symptoms or does not involve psychotic symptoms, there's typical and atypical depression. There's others as well. Okay. So we know that depression has multiple subtypes. We suspect that trauma probably does. We know there's PTSD that has prominent cognitive impairment. We know there's PTSD with severe depression. We know there's PTSD which is complex rather than simple. Those who have suffered enormous trauma exposure in early development have a different set of problems than a high-functioning person who survives a traffic accident. We know that there is PTSD with prominent association symptoms and not. So there's not one PTSD, there's probably a half a dozen major types of PTSD. I think we'll be able to confirm what they are and confirm them with biological and psychosocial features. And the second thing is, we'll be able to do these precision medicine treatment models that say who's likely to respond to which treatment for which subtype. 

DR. THEA GALLAGHER: Which will totally change the way that we're looking at psychiatric illness and I think even starting with the diagnostics, we're already starting to realize, like you said, there's a spectrum and there's all these different types. So the better that we can diagnose and then treat patients will lead to better outcomes. 

DR. CHARLES MARMAR: So a woman with a complex PTSD with associative features, who is 60 years-old, is not going to be treated the same as a man with a more simple PTSD and a man who came from a very difficult background and had three tours of duty in the military or a woman who was in the military and has a very complex PTSD, will not be treated the same as someone who was walking down the street as one of my patients was and tragically, as they were walking down the street in San Francisco at the time, I was doing the study, someone jumped to their death from a building from high above, and they witnessed this horrible event and the person had gruesome injuries, of course, and it was completely horrifying to them. But they were very high functioning before. They went through a period of nightmares, flashbacks, and startle reactions, but they responded to a short course of treatment and were very well. Obviously, that's a different kind of PTSD than someone who was sexually assaulted and abused during their childhood and neglected and has suffered enormously during their lifetime. 

DR. THEA GALLAGHER: So I think it's going to give a lot of hope to patients. And also, I think to doctors and clinicians being able to have more confidence in the treatments that they're actually disseminating and implementing with their patients.

DR. CHARLES MARMAR: If a patient has a solitary nodule of breast cancer, which has not escaped the duct, with certain genetic features, they're going to be treated…they are currently treated very differently, obviously, than someone who has a very aggressive metastatic breast cancer with other genomic features. And so the illnesses are treated as precisely as possible for their stage, their aggression, their genetic features, and other factors. We want to be able to treat depression that way. 

If we have time, I'd like to conclude with a very interesting and dramatic example of precision psychiatry in depression. Around 2015, our neurology psychiatry program which we have at NYU where we have experts trained in both neurology and psychiatry, they are board-certified neurologists and psychiatrists, were asked to consult—and this was in the Department of Neurology initially—on a patient who had had roughly 20 years of completely failed treatment for depression. So let's say for the majority of their adult life, they were depressed, miserable, very difficult to work, very difficult to maintain love relations and family relations, and enormous personal suffering. The patient had been treated with all of the evidence-based treatments, psychotherapies for PTSD, psychodynamic, cognitive, behavioral, others. That patient had been treated with virtually every type of medication treatment for depression, SSRIs, non-SSRIs, other antidepressants, mood stabilizing drugs, neuroleptic drugs and others, completely failed. I don’t remember right now if they had neuromodulation treatments, but they definitely had ECT and a complete failure.

The patient was thought to possibly have a more rare neurological disorder, and it was justified to perform a brain biopsy to rule that out. So they had a small needle brain biopsy done under neurosurgical conditions. A needle was presented through the orbit and into a silent part of the prefrontal cortex, relatively safe thing to do. The main complication is infection, but if it's done under neurosurgical conditions, it's safe to do. And then the tissue was taken and sent to a neuropathologist. And the neuropathology report came back massive invasion of inflammatory cells. Quite unexpected. And the patient was treated with a novel anti-inflammatory drug which crosses the blood-brain barrier, and for the first time in 20 years, had significant relief from their depression. A single biopsy with a single confirmation. We can't do that routinely, but I mentioned that case as a proof-of-concept, as an illustration of what could be done if we knew what was actually occurring in the brain. 

DR. THEA GALLAGHER: And I think you're saying, even with that example, is kind of the theme of this, that going from the complex to the simple and maybe we can understand more about the brain when we kind of understand all of it. Maybe it does come down to one single biopsy, or in what you're talking about, it might come down to just understanding a few different factors and how they integrate together to present for this patient. 

DR. CHARLES MARMAR: Correct. And if we knew that it was invasion of inflammatory cells in the brain that were causing the depression, we might then be able to look in the blood or somewhere else or in brain imaging or in the genetic background to find a marker that was associated with that, even though we weren't routinely biopsying. So let me give you another example. We talked about it on an earlier broadcast today. In the 1800s, a common reason or perhaps the commonest reason why patients had long-term hospitalization in state mental hospitals, psychiatric hospitals in America, including in New York, was it turned out to be the neurological complications of syphilis, neurosyphilis. And so that was a more common reason to end up spending years, decades, or your life in a state mental hospital than was, let's say, schizophrenia. When the infectious basis of neurosyphilis was discovered with the spirochete and treatments were developed, of course, now that disorder is an infectious disorder, it's no longer considered a psychiatric disorder and it's treated and we don't have patients with neurosyphilis in psychiatric institutes. If we really understood and that's an example of going from the complex and the completely treatment unresponsive condition requiring long-term psychiatric hospitalization to a clear, simple idea of what is disturbed, a specific infection of the brain by a specific organism with a specific treatment. Nothing could be more clear than that, right? So there is going from the complex and the enormous pain and suffering and a sense of hopelessness to an understanding, a treatment, and a cure. 

DR. THEA GALLAGHER: Well, it's exciting for, I think, patients, for clinicians, and for researchers, and I think it'll be great to see where it continues to grow. 

DR. CHARLES MARMAR: Thank you. 

DR. THEA GALLAGHER: Thank you so much for talking with me. 

DR. CHARLES MARMAR: It was a pleasure talking to you, and it's an even greater pleasure working and collaborating with you. 

DR. THEA GALLAGHER: Wonderful. Well, thank you to all of our viewers and listeners and if you liked this episode, please rate and subscribe to our podcast where you listen or watch your podcasts. Thank you so much from all of us here at NYU Langone Health and the NYU Grossman School of Medicine.