Hello, thanks for coming to our talk. I'm Dr. Hines. I'm the chair here, which means my only real job is to introduce smart guys in the room. In as much, I'm going to introduce everybody now, and then we're just going to flow through the talk, and we'll take questions at the end. For the guys in the virtual audience, I do have questions here. So just put your questions in, and we'll also answer those at the end. First up is Matthew Rusling. He completed his undergraduate studies at the College of Bard with an emphasis in informatics and microbiome analysis. He joined the Army through the HPSP scholarship program and went to medical school at Des Moines University. He's now completing his psych residency at Walter Reed and the National Capital Consortium Residency. He's received several awards for teaching and contributions to military research. Dr. Rusling also serves as the research chair for the Society of Uniformed Service Psychiatrists military chapter at the APA. And Dr. Rusling's long-term interests are in seeking opportunities for research or industry upon completion of his military service. Next will be Dr. Zach Monahan. Dr. Monahan is a combined MD-PhD from the Uniformed Services University of Health Sciences. His doctoral work focused on protein biochemistry and protein fused in sarcoma and its relationship to the pathogenesis of amyotrophic lateral sclerosis. After serving a tour as a battalion surgeon in the 2nd Maintenance Battalion, 2nd Marines Logistics Group at Camp Lejeune, he returned to his psychiatry residency and is currently completing his third year of residency and has research interests in the pathogenesis of neurodegenerative diseases. And our last speaker will be Dr. Landon Sorensen. He enlisted in the Army after high school and served seven years prior to attending medical school at East Tennessee State University Quillen College of Medicine. He's currently enrolled at the National Capital Consortium Residency at Walter Reed. He's also a teaching fellow at the Uniformed Services University of Health Sciences. And his interests are focused on clinical applications of research. And he's planning to do a child and adolescent fellowship. With that, further ado, Dr. Ruslan. Hey, thanks, Dr. Hines. And thank you, everyone, for coming. To get started, because we are wearing these uniforms, everything that we discuss here is our views only and does not represent that of the Department of Defense or the US government. So just getting that out of the way for all of the rest of us. So learning objectives for my part of the talk is focusing on the basic role, limitations, scope of pharmacogenomic testing using PGX as the acronym, and when we should be considering it, and when it will likely not have significant benefit for our patients. And then building some insight into the base rate for pharmacogenomic mutations and understanding whether this is a common or uncommon phenomenon. And then finally, looking at the relevance of pharmacogenomics in considering substance dependence. So we're going to start out with a brief case. So we have this 45-year-old male with history of PTSD. He comes into our office and says, hey, doc, I have been on every SSRI, every SNRI that you've heard of, and none of them have worked. I've also gone through CPT, and it's been a really good therapeutic relationship, but nothing is really working. And we inherit him on 200 milligrams of sertraline for the past three months, and he continues with Persistent Criterion B persistent symptoms and a PCL-5 of 64. And he's having no side effects, good adherence, and no benefit. So we have a couple of different options in managing his care. We can increase his dose above the FDA-recommended max. We can continue at the same dose. We can decrease the dose or abandon medication trial and try something new or explore augmentation strategies. So when we look into how do we make this clinical decision with four very oppositional treatment pathways, we should be asking if there are additional data points that we can seek out to better inform our therapeutic decisions. And my argument is that pharmacogenomics do have a role in a small subset of our patients. But first, starting off, going back to MS1 in medical school, we're going to remind ourselves of what a substrate is. So a substrate is what a reaction starts with. An enzyme catalyzes that reaction. A metabolite is what that reaction produces. And an inducer is something that increases the rate of metabolism. And an inhibitor decreases the rate of metabolism. And the CYPs is a group of enzymes that makes up the P450 system and does something like 90% of the metabolism of the medications that we prescribe. We can think of this kind of like a sink, where if we turn the faucet to the left, which drains the basin, it drains the substrate. Or to the right, where we slow down the flow of the substrate out of the basin, that is kind of like inhibiting the reaction. So really focusing on substrate concentrations is a good way to conceptualize induction and inhibition of the CYP system. So giving an example that we encounter very frequently in psychiatry is the 2D6 role of fluoxetine being converted into its active metabolite. So in patients, let's say we are thinking about, how can we increase the serum concentrations of fluoxetine? So we give them some bupropion. And this 2D6 inhibition increases the serum concentrations of the prodrug. When we look at the actual metabolic profile of what's happening inside the cells, we knock out the functional CYPs, which decreases the production of the active metabolites and increases the serum concentrations of the prodrug and the inactive metabolites. So when we're thinking about our pharmacologic decision making, we know that interactions between drugs are really important as we make our treatment decisions. And we can break down drug metabolism into modifiable and non-modifiable components. The modifiable being medications that they're taking and foods they're eating. And non-modifiable is their genetic profile. So does this non-modifiable component matter as much as what we're already thinking about when we are already making drug treatment plans? And how frequently is a patient's CYP metabolism mutated? When I was preparing for the talk, it was this last point that surprised me the most at how high these mutation base rates really were. So there was this excellent paper. And all the references are at the end. We'll provide slides so you can look at these individual papers if you'd like. So this was looking at a meta-analysis of European studies at the CYP2C19 allele, which was a non-functional mutation. And the prevalence ranged from 11% all the way up to 20.8%. So this means that there was a loss of function in between 1 in 10 and 1 in 5 patients in Europe. And then when we look at the 2C1917 allele, which is a rapid metabolizer, the rate of rapid metabolism was between 1 in 5 to 1 in 4 patients, depending on the country of origin. If we were to look at Poland, for example, and we were to say, what's the probability based on these two alleles that someone would be a normal metabolizer, there's only a 56% chance on two alleles alone that an individual from Poland would have a normal CYP2C19 genotype. And again, considering only two alleles. So in pharmacogenomic studies, there's a huge component of ethnicity and a wide range of ethnic difference based on country of origin in the pharmacogenomic mutations. This study looked at a collection of all pharmacogenomic mutations across all known studies. So this is a global sample size, which found that there was only a 41% of individuals having a normal 2C19 genotype, which I found pretty shocking. I would have anticipated that normal would be like 80% to 90%, and these rapid and ultra-rapids would be much less frequent. But with the rapid and ultra-rapid, off of the 17 mutation alone, giving us about 30%, and then 3% being poor metabolizers. So to put this a different way, one in three of our patients could benefit from either higher maintenance dosing or BID scheduling, and one in 30 of our patients may require a lower maintenance dose and possibly alternative dosing strategies. So if we kind of think back to what medications are metabolized by 2C19, we see a lot of the medications that we prescribe every single day. And with it in mind that only 40% of our patients are going to have a predictable response, we should be very cautious in individuals who have had abnormal drug reactions to multiple medications on this list. And then calling into the importance of one of the non-psychiatric medications on this list, there is a strong FDA recommendation that in patients with a poor metabolizer profile of 2C19 to avoid clopidogrel and also to minimize use of 2C19 inhibitors in patients who are taking clopidogrel. This is a good reminder for me that this is a red flag drug that when I see it on a med list, I now very carefully look through to make sure that there is no 2C19 inhibition of the medications that I'm prescribing as a medical legal issue. So looking at 2D6, which is the other most common CYP enzyme that metabolizes the drugs that we prescribe, in a American, so United States origin data set, 80% had a normal metabolic profile, poor metabolizers 5%, and ultra-rapid metabolizers 2%. So the base rate of a abnormal 2D6 genotype is much lower than that of the 2C19. So we've already discussed the base rate of two CYP enzymes and identified that the base rate of a normal genotype is 41% for 19 and 81% for 2D6. So using averages cited so far, the probability that any one patient, assuming independence between 2C19 and 2D6, that they have a normal metabolic profile is about 1 in 3. So the next question is, does this matter clinically? Because all of this is all well and good, and maybe it does or does not actually impact patient care. This is a phenomenal study that was done in Norway that looked at a couple of thousand patients who were prescribed sertraline. And they genotyped them, they drew their blood, and harmonized their sertraline dosing to 100 milligrams because everyone is on a different dose of sertraline. So this allowed them to compare different doses in a normalized way. And what they found was in the poor metabolizer group that the blood levels of sertraline were 2.7 times higher than those of the normal metabolizers. Looked at a different way, 1 in 30 of our patients are taking 300 milligrams functionally of sertraline instead of the 100 milligrams that we think we're prescribing them. And then within the poor metabolizers, 1 in 4 patients are taking 150 milligrams instead of 100 milligrams. So I strongly believe that some of this information does help form individualized dosing plans, as the studies that have been done do somewhat definitively show that there is a strong relationship between actual biologic outcomes in terms of the drug level and the genotypes. So there was a big study that was done, published in 2022 in JAMA by Oslin et al, which looked at how much of a difference does it make if we take a huge sample size across a lot of different clinics and hospitals, and we give 50% of the providers genotype results, and then for the other 50% we withhold them. And we look at, does this change prescribing practices, and does it change outcomes? And what we found was that giving genotyping results does reduce the prescription of medications which have a genotype medication interaction, and that remission was achieved a little bit earlier. And for the first couple of weeks of the studies, the genotyped group did have improvement, but at the 24-week endpoint, there was no difference. There were, I think, a couple of important limitations on this study, the big one being that these were simple cases, so no augmentation or antipsychotic use was allowed. It involved both psychiatry specialists and primary care, so not necessarily representative of a sick patient who's failed multiple trials, who comes to us after going through primary care, multiple trials with them. And also, most clinicians only treated under five patients, which means that many of them were likely unfamiliar with the use of pharmacogenomics, and it's unclear if the clinicians were waiting on the pharmacogenomic results before making prescription decisions, what kind of educational materials, and that in the testing that was used, it provides an almost oversimplified interpretation to guide your clinical decisions. So my takeaway was that when clinicians who have infrequent use of pharmacogenomics with limited training in its use in a kind of straightforward patient population, that there's likely limited benefit. So if pharmacogenomics is something that you intend to use in your practice, I think that it's worth using judiciously and being selective towards a population that has, in fact, failed multiple drug trials. I don't think this is a panacea, but I think it does provide a lot of benefit to the right patient. So let's look at a sample of PGX, and should we only prescribe from the recommended list or avoid these medications? Not really. So when we get pharmacogenomic testing, no matter from what commercial supplier we're getting it from, they're going to give you an enormous amount of data, like so many things. And this is maybe 50% of what you would see from a GenoMind or other sample. So this is really overwhelming. There are some rapid metabolizers, some intermediate, some ultra-rapid, some normals. I have no idea what to do with this information. Good thing is that there is an expert group called the CPIC that is the kind of framework of how we should be using pharmacogenomics if we are to approach them in the most conservative way possible. So as of April 2023, there were only sufficient evidence for profiling three things, 2D6 to guide the dosing of paroxetine, fluvox, venlafaxine, and vortioxetine, 2C19 to guide acetalepram, acetalepram, and sertraline, and then 2B6 to further dose sertraline in individuals who had a poor 2C19 metabolic profile. So the takeaway is for a lot of hospitals, the Roche AmpliChip 2D6 2C19 is one of the in-house standards. And that is actually evidence-based as enough based on what we know about pharmacogenomics and its limitations. But there is emerging evidence. There is so much research going on in this field. So the anticipation is that we will likely gain more and more insight as we move forward into the future. So going back to this panel, we can cross out everything that is less evidence-based, really focus on the 2C19 and 2D6. And instead of saying, oh my gosh, sertraline, for example, is metabolized by 2C19, 2C19 is abnormal. I should just avoid the use of sertraline because their genotype falls into the abnormal range. Now we can say that we can actually use this information to create a dosing plan for our patients that's individualized, have a normal starting dose, and likely a higher maintenance dose. So going back to this first case that we started with, where we have this 45-year-old male, history of PTSD, multiple failed drug trials. Now we ask, what should we do with this gentleman now that we have identified due to his multiple drug failures that he would be a good individual for? And my timer just went off. I'm so sorry. There we go. So now that we've identified that because of this guy's multiple drug failures that he's a good candidate for pharmacogenomic testing, we now have new evidence to further individualize his care. And we can increase his sertraline dose beyond the FDA max with something that is guided by more than intuitional alone. When it comes to using genotyping in the treatment of substance dependence, there is, like a lot of things, less evidence for this than other domains of medicine. However, when it comes to methadone and buprenorphine in particular, there seems to be some possible benefit in the genotyping of the 2b6 and 3a4 genotypes. There are a number of case reports in particular of individuals who clinicians were concerned about non-adherence or drug seeking. And then when some studies will genotype these individuals, they will find that these drug seekers are in fact rapid metabolizers, that these genotypes and biologically are receiving a lower dose that is going away faster, so their withdrawal symptoms are going to be much more severe. So would I genotype every patient that I'm prescribing buprenorphine to? Probably not, but in the individuals who I have a confidence that they're being adherent, but they are reporting severe withdrawal symptoms or requiring higher doses, I would have, you know, identified these patients as likely benefiting from pharmacogenomic testing before I start advancing dosing above recommended levels. And then with methamphetamine, there are some upcoming targets which are really exciting, especially the BDNF-VAL66 met mutation that could alter COMT metabolism, but that is very, very far in the future. So the takeaway is pharmacogenomic testing can have significant benefit for our patients. It is unclear if it's going to be helpful when used infrequently and will be most beneficial for patients with multiple treatment failures, and does have a growing role in clinical psychiatry, especially when we use it cautiously. So if you're interested in collaboration, I have experience in micropalm data analysis. I have a big interest in PTSD and human resilience. My email is up there. I would love to hear from you for collaborations and also have some relationships with other labs nearby, so look forward to hearing from you. I have a bunch of references, and these will be available on the slides that we'll present at the end. Thank you so much. All right, thanks Matt. So I'm going to be presenting an overview of some recent literature, highlights from literature about inflammatory biomarkers, specifically in psychiatric disorders. And as Dr. Hines said, we're third-year residents at Walter Reed, and as Matt mentioned, for all of us, I have no disclosures. The views I'm presenting today are my own. I'd like to take a moment to kind of establish the state of the problem that we're hoping to address with my part of the talk, at least, which is that reliability is a foundational concept in medical diagnostics. And that's true for psychiatry as well, but psychiatry has historically been at a disadvantage compared to other fields of medicine that haven't had, we haven't had the same kind of objective lab tests, blood tests, that we can rely on to improve reliability among diagnostics and to make sure that our diagnoses are accurate. One of the emerging tools that we can use that's becoming more mainstream are tissue biomarkers, and that can be an adjunct now to structured diagnostic interviews, and also to brain imaging, which has been historically some of the things that we've had to rely on in the absence of tissue biomarkers. So my objectives for today's talk would be to, first, to appreciate an experimental model that we can use to characterize potentially relevant biomarkers in psychiatric disorders, to summarize key suspected biomarkers that exist now in the literature, to articulate some examples of the potential clinical utility of inflammation-associated biomarkers, and then to discuss the potential for how a detailed study of these biomarkers may actually help us understand the underlying pathophysiology. So just very briefly, inflammation is obviously a foundational concept in pathology and medicine. Back to antiquity, the cardinal signs of inflammation have been appreciated, redness, swelling, pain, warmth, and we have come to appreciate that this is a product of molecular mediators, cellular participants, and a dynamic relationship between them in tissues that are exposed to an exogenous insult. So this is a image of some skin that's being exposed to an injury, and you can see kind of as the population of cells and molecular mediators that drive those changes over time and space correspond with structural changes that hopefully lead to a resolution of the damage that was brought onto the tissue, but can also lead to damage in its own right. And this general model of inflammation can be applied to neuroinflammation specifically. So the actual, in this particular example, the accumulation of beta amyloid aggregates in nervous tissue leads to upregulation of cytokines, activation of microglia, which are resident nervous immune cells, the peripheral invasion of leukocytes, and then in this case, the damage to nervous tissue and to neurons themselves. That's a very brief model of inflammation and neuroinflammation specifically, and the question that I want to answer today first is, are inflammatory pathways at all, is there evidence that inflammatory pathways are involved in psychiatric disorders and specifically in affective disorders? That's the example that I'll use. So there is evidence that that is the case, that there are inflammatory pathways at least involved in affective disorders, and some of this data I cite here includes the administration of LPS in both human and animal models that can induce the symptoms of mood disorders. And somewhere between 25 and 40 percent of patients with hepatitis C who are treated with interferon alpha will go on to develop a major depressive episode. This study from 2022, I'm going to walk through for a couple of reasons. These authors, I think, lay out experimental evidence of a potential role for neuroinflammation in the development of a depressive and anxious phenotype. But secondly, I also think that it lays out some general practices for how to effectively interrogate suspected inflammation-associated biomarkers that may come to clinical or scientific attention. So just very briefly, these authors have a mouse model where they subject the mice to chronic social defeat stress, which is effectively having a bigger mouse beat up a smaller mouse. And then they monitor the individual protein players in a sequence that culminates in what's called a theraptosis or a therapeutic pathway analogous to apoptosis, programmed cell death. And ultimately, with the actual development of a depressive or anxious phenotype in the mouse. They use an anti-inflammatory agent, Adaravone, which is one of the FDA-approved ALS medications, Ravacata, approved in 2017. And it's an ROS, a reactive oxygen species scavenger. And they use this to manipulate both the individual protein players as well as the phenotype in question. This is their experimental layout. So they have a 10-day period where the mice are subjected to chronic social defeat stress, which is the daily physical defeat of a weaker mouse at the hands of a stronger mouse. And then those mice that are stressed, as well as control mice, are treated either with an intraperitoneal injection of a blank vehicle or with the anti-inflammatory. And then subject to a series of behavioral tests. In this case, I have a few here. Sucrose preference test, open field test, tail suspension test, and forced swim test. And they show that the phenotype that develops is consistent with results from these tests that are generally considered to be depressive or anxious. So the degree of preference for sucrose feed is diminished in the stressed mice compared to control mice in the far left there. That change in the behavioral phenotype is attenuated by a treatment with edaravone. So that anti-inflammatory can reduce the depressive and anxious phenotype. And that carries forth for all of these remaining behavioral tests. The open field test, tail suspension, and forced swim test. And these are really measures of, for example, with the tail suspension test, how long does a mouse stay immobile when suspended from its tail? The forced swim test, how much time is a mouse immobile when it's forced to try to prevent itself from drowning? So the conclusions from that portion would say that anti-inflammatory agents, in this case the edaravone, are potentially capable of mitigating depressive behaviors in an animal model. So they go a step further and they look at the hippocampal tissue lysate. So on the left-hand side here, this western blot, is hippocampal tissue that's been homogenized. And then a series of those individual protein players in that ferruptotic pathway are analyzed. So sirtuin, sirtuin 1, Nrf2, Ho1, and GPX4 are all players in that ferruptotic pathway. But I'll draw your attention to TLR4 and phosphorylated NF-kappa B, which are widely appreciated inflammatory biomarkers. They're involved in an inflammatory pathway that leads to the transcription of IL-1, IL-6, and TNF. And you can see that in mice that are unstressed on the far left and untreated with an anti-inflammatory, you can get a kind of a base level of how much these proteins are expressed in the hippocampus. When you stress the mice in the middle two lanes and you do not treat them with an anti-inflammatory, the level of phosphorylated NF-kappa B, which is activated, producing the downstream inflammatory mediators, is increased substantially. And if you treat with the daravone, the anti-inflammatory, that level returns to more approximate the baseline. These graphs on the right are simply the quantification of the western blotting. So mitigation of depressive behaviors by anti-inflammatories would be correlated with changes in the inflammatory mediators at the tissue level. Then the authors do a complementary assay, which I think is a nice way of solidifying their conclusions. So in this experiment, they take the mice with the 10 days of chronic social defeat stress, and they treat them either with a daravone, a vehicle, a blank vehicle, or a daravone, and EX-527. And EX-527 is a sort of two-in-one blocker. So it blocks this pathway, this pheroptotic pathway, that is being implicated in this experiment. So you can see in the graphs here, this again are behavioral tests. And there's control mice on the left on each graph, and they have some baseline behavioral pattern. And then when you stress the mice, their behavior changes in a way that's consistent with a depressive or an anxious phenotype. And then you treat them with the adaravone, the EDA, and you have an attenuation of those. But if you block this two-in-one, the initial pathway target of this pheroptotic pathway, this inflammatory pathway, you can actually attenuate the consequences of treatment with the anti-inflammatory adaravone. So blocking specific targets of anti-inflammatory compounds can attenuate the behavioral and the molecular consequences of anti-inflammatory treatment. So then these series of experiments help us do two important things. First, they establish an association between the potential tissue biomarkers and a behavioral phenotype. But they also establish that this relationship is manipulatable. So at an individual level of proteins, you can get down to that level of resolution to change the outcome by blocking specific mediators. But the most fundamental thing is that these data suggest that inflammatory pathways are involved in the development of a depressive or an anxious phenotype. So the next question would be what potential pathways and associated biomarkers exist in the literature that reflect this potential, that reflect the association of inflammatory pathways and psychiatric symptoms of interest. In 2007, the recognition of NMDA receptor encephalitis and the subsequent 2016 publication of diagnostic criteria, I think, is when a lot of the inflammatory biomarkers, the utility of inflammatory biomarkers in diagnosis, became more mainstream. And in the years since then, a number of potential inflammation associated biomarkers have emerged. I list a series of them here. This is an abbreviated list. And what's happening now, and I have a citation at the bottom of this slide and in the references slide, what's happening now is what these authors have done, which I think is really going to be the next immediate steps in the development of this work, which will be to link specific mediators to specific neurotransmitters and to specific psychiatric symptoms. So for example, to link an IL-6 or CRP to the serotonin pathway and to suicidal ideation. The beginnings of that began to emerge as these papers were published in 2021 that make use of IL-6 and CRP, but IL-6 especially is a central node in the inflammatory pathway. It governs a series of other downstream effects of inflammation. And these authors in 2021 showed, in the case of Kaplan et al., they showed that IL-6 potentially has a causal relationship. So tissue levels of IL-6 have potentially a causal relationship with the subsequent development of suicidality. And they studied that using Mendelian randomization, which is a genetic technique to help establish the causality of an association. Additional work showed IL-6 and CRP specifically associated with neurovegetative symptoms of depression in a Dutch biobank, and that was the UK biobank using about 3,000 individual samples. Sorry, 3,000 from the Dutch biobank and then 150,000 from the UK biobank. And one specific inflammatory-associated biomarker that's become most prominent recently is kynurenic acid. And in this paper from 2003, Brown and colleagues, they obtained brain tissue samples from the NIH neurobiobank and completed mass spec to show that there were reduced levels of kynurenic acid in the anterior cingulate gyrus of women who died by suicide and who had been diagnosed with major depression, compared to women who had been diagnosed with major depression but did not die by suicide or controls. Kynurenic acid is an NMDA receptor antagonist, and it's produced in this tryptophan pathway on the right that I lay out. And it potentially offers a means for us to understand how inflammation may lead to downstream effects that affect the development of psychiatric phenotypes. And so tryptophan is committed to this pathway, be the two enzymes IDO1 and TDO2. IDO1 is upregulated by IL-6 and IL-18. And the tryptophan is then sent down either the kynurenic acid pathway or to quinolinic acid. Kynurenic acid is an NMDA receptor antagonist, and quinolinic acid is an NMDA receptor agonist. There is some evidence that the ratio of those two things, the higher the quinolinic acid ratio to the kynurenic acid ratio, can raise your risk for suicide. And that would be borne out in these data from the neurobiobrain brain bank. What was interesting is that kynurenic acid is preferentially produced by astrocytes, and quinolinic acid is preferentially produced by microglia. And so activated microglia, which would be a central component of neuroinflammation, may lead to an increased relative concentration of quinolinic acid relative to kynurenic acid. And because quinolinic acid is an NMDA receptor agonist and kynurenic acid is an NMDA receptor antagonist, there would be reason to think that that would have a reduced potential for neuroprotection. So kynurenic acid via its NMDA receptor antagonism is thought to be neuroprotective, the same way that some of our medications like memantine are neuroprotective. So this gets us to a point where we can say activated microglia might be a means to explain why a baseline increase in neuroinflammation can subsequently lead to the development of psychiatric phenotypes. And I'll close with a comment about the potential utility of biomarkers. So this study from 2020 took 88 patients with major depressive disorder and 88 healthy controls and found 20 metabolites with significant concentration differences. They also found two kynurenic acid and kynurenine that had potential predictive benefits. So we can take biomarkers and kind of break them into two categories, diagnostic and predictive. Diagnostic biomarkers obviously come up with diagnosis. Predictive biomarkers can help us predict the response to treatment. Of the 20, only two were predictive and only one was both. So kynurenic acid was the only biomarker they identified that had both diagnostic utility and predictive utility. And they found that lower levels of kynurenic acid were associated with a diagnosis of major depression and lower levels were also associated with an improved response to treatment with acetaleprime over six weeks. That predictive potential for biomarkers I think is really important and maybe is the next step to go beyond simple diagnosis. And so in 2022, a publication came out, Targum et al., that showed data that showed that the peripheral platelets from patients with major depression could be used to generate both diagnostic and predictive biomarkers. Before I get into the details of that, I want to just quickly review the GS-alpha signaling pathway. So GS-alpha signaling pathway in the context of major depression is, this kind of lays out how it's thought to interact. So there's a G-protein coupled receptor, that's a lipid-based signal transduction protein that takes an extracellular signal and then transduces it into an intracellular mechanism to generally lead to increased transcription. So this G-protein coupled receptor here on the far left is anchored in the membrane and G-alpha S, which is this small purple subunit just below it, when an exterior, an external ligand binds the G-protein coupled receptor, that G-alpha S then goes to the adenolyl cyclase to the right, increases the activity of adenolyl cyclase, which turns ATP into cyclic AMP, and that cyclic AMP activates transcription factors that go into the nucleus and lead to, in this case, increased transcription of brain-derived neurotrophic factor. G-alpha S is thought to be ensconced in lipid rafts at a higher rate in patients with major depression and also to be anchored more tightly to microtubules in those patients, which inhibits the ability for G-alpha S to get to the adenolyl cyclase to turn it up. As a result, you'd have lower levels of BDNF. These authors took platelets from peripheral blood samples from patients with major depression and controls and found that in an in vitro assay using platelets from that sample, that G-S-alpha was, the adenolyl cyclase was less responsive to PGE1 stimulation via that G-S-alpha pathway in patients with major depression, and that treatment with acetalepram over six weeks, again, was shown to have an increased response to PGE1 stimulation, and that could be used as a monitor for the treatment response itself. And my last point I'm going to close with is a paper from from last year that I think does a really interesting job. It takes a series of patients who have dementia, uses the neuropsychiatric inventory questionnaire that's given to caregivers for those patients, and associates the neuropsychiatric inventory questionnaire results with underlying tissue biomarkers. So in this study, the caregivers were given the neuropsychiatric inventory questionnaire. And the severity, the overall severity of the NPIQ score was then plotted alongside results from a PET scan that monitored a protein called TSPO. So TSPO is a translocator protein. It's 14 kilodaltons. It exists in the outer mitochondrial membrane. It's been used for decades as a marker for microglial activation. Microglia seem to upregulate this protein. And you can use radio ligands and PET scans to monitor microglial activation by looking at TSPO. So the more active microglia are is kind of a marker on the x-axis. So the x-axis is showing us how active the microglia are. And on the y-axis is your NPIQ severity score. You can see a pretty tight correlation that the more active are your microglia, the worse the caregiver reported NPIQ severity score is. But they go a step further, and they take the individual components of the NPIQ severity score and see how that when you leave one out, a leave one out analysis, see how that changes the tightness of the correlation. So if you remove irritability from the NPIQ severity score, the correlation on the graph on the left drops by about 23%. So that is the greatest contributor individually to the tightness of that correlation in the graph on the left, suggesting that irritability potentially has a direct relationship with the degree of microglial activation. So what these data would suggest is there's a potential to use an understanding of biomarkers, tissue-level biomarkers, to understand the underlying pathogenesis of individual psychiatric symptoms, which I think is really an exciting potential utility for this. And it goes beyond simple diagnosis. It goes beyond even predictive use for response to treatment. So we discussed an experimental model that talks about how we might interrogate potential biomarkers at the tissue level, and also how to establish that inflammation does have data to suggest it's involved in the development of affective disorders. We summarized some key inflammation-associated biomarkers. We articulated some potential clinical utility, both diagnostically and predictively. And then we discussed how we might even use biomarkers to understand the underlying pathophysiology. And with that, I'll turn it over to Dr. Sorenson. Thank you. Good afternoon, everyone. Thank you for joining us. My name is Landon Sorenson. I'm a psychiatrist resident at Walter Reed as well. And for those of you standing in the back, there are some chairs up front if you'd like to join me. I plan to teach some things today, and I think you'd be more comfortable seating. But overall, I have two real goals, summarize known and emerging diagnostic biomarkers, and discuss predictive biomarkers. And I have nothing to disclose on the last slide. So the clinical interview is always going to be one of our most important ways to screen for substance use and abuse. And not all patients report accurate alcohol use. And similar patterns are seen with substances that are both illegal as well as misuse of prescription substances. Clinical biomarkers can supplement our screening as well as accurately monitor levels of consumption over time in specific cases. So these biomarkers are continuing to be approved or improved, while many are already in use today, and several we will discuss. So as a physician, when I was in medical school, I was told to treat the patient, not the labs. But I also like to use labs to improve my patient care. Can anyone guess which labs are from a patient of mine who is not yet ready to disclose his heavy alcohol use? I'm sure we all have some guesses and we can interpret these. These are three different conditions, of course, almost textbook. But only the one in the middle demonstrates the de Rittis ratio, one of the oldest biomarkers for addiction substance use. In 1957, it was published, noting that the AST to ALT ratio of 1.5 to 2 over the other is sensitive for alcoholic hepatitis, in the sense that 92% of patients with alcoholic hepatitis demonstrate this ratio. The mnemonic I was given was awful scotch and tonic is elevated in alcohol use. And of these three patients, while they all could, this one in particular was later tested with additional biomarkers. And thankfully, we don't have to wait for the severe hepatic impairment that occurs to demonstrate this ratio. Because as Harris et al. in 2021 demonstrated, and this is an excellent article that I encourage individuals to read, there are a plethora of well-documented biomarkers that are more sensitive and more specific than the awful scotch and tonic ratio. So first and foremost, what I'd like to highlight is phosphatidyl ethanol, also known as PEF. It is the most sensitive and specific related to heavy alcohol use in patients currently. In fact, a significant proportion of patients evaluated with PEF values elevated above the cutoff for heavy drinking still maintained a normal carbohydrate-deficient transfer in, or CDT level, which is also a very sensitive and specific test I've used before. Further, while CDT remains a valid biomarker for alcohol use, it also appears that CDT will increase with age and has been found to be higher in males compared to females, while no sex differences appear to exist with PEF. Unfortunately, for at least now, both those tests take a little bit of time to go to the lab, get sent off, and come back in my facility. So I frequently use and have seen GGT be used, because it can come back the same day. It's something we're ordering for other conditions as well as an initial screener in certain cases. Unfortunately, GDT is exceptionally nonspecific as it's demonstrating changes to the liver in particular, and you have to have certain levels of damage before it turns positive. As an interesting note, the patient I highlighted on the slide before was one of my patients, and his BGGT was positive. He then told me that it was specifically related to his keto diet that he had just started and that he had not returned to alcohol use. He then enrolled in a partial hospitalization program, where upon screening with PEF, he was positive, and then later felt comfortable to disclose to the treatment team there that he had returned to alcohol use. While we do not have nearly as sensitive or specific biomarkers for the majority of the rest of the things I'll cover today, there is obviously ongoing need, and there's a lot of research being driven by those need. For example, with trends across the nation for legalization of cannabis, there is need for a so-called driving under the influence of cannabis test, and I've presented some details here as far as different cutoffs and detection. One of the more inconsistent things we have, though, is when we look at synthetic compounds, synthetic cannabinoids, synthetic other things as well, because frankly, it's difficult to detect. Some of them have different alterations that make it less than ideal for us to detect from a baseline test, and sometimes require mass spectroscopy or other things to detect them. So one study I'd like to highlight to bring that up is an interesting study where they looked at patients admitted for first-break psychosis into the inpatient psychiatric ward with concern for substance use and a negative UDS. They then invested the money and the time to find specific metabolites from the synthetic cannabinoid they were taking and just trended that with time, monitoring their symptoms of psychosis as well as positive values. The interesting case I wanted to bring up was one of those patients was psychotic and positive for 24 days during his stay on that inpatient unit, which is something that I want to highlight the importance of this need for more rapid testing as we involve and improve this science, because the reality is many of us may have seen patients with brief psychotic episodes that were actually synthetic cannabinoid-induced psychosis. Moving on to the category of stimulants, one of the less expected outcomes of this research was developing something that Dr. Rusling was talking about, but a marker for CYPA activity. In fact, so 3-hydroxycotonine to cotinine ratio is related to nicotine use. It is how we could detect if someone was using nicotine, say a clozapine patient who suddenly has a clinical change and we wanted to check to see if they're smoking again, or perhaps an adolescent who's in an at-risk population. But also, this ratio is in a non-invasive probe for CYP2A6 activity and something that may be used further. For most stimulants, there are several challenges in developing both accurate diagnostic and predictive biomarkers, largely related to shorter half-life, shorter detection windows, different variants. But as Dr. Manahan briefly mentioned with neuroimaging, there is interesting trends that neuroimaging may be a biomarker of the future, specifically, for example, listed here with cocaine. FMRI studies investigating the ROID system, especially the striatum, are starting to notice changes in patterns consistent with these substances. If we focus on the need for monitoring patients receiving S-ketamine, while the specific dosing protocols and methods of providing S-ketamine in clinic make it significantly less likely for patients to lead to neurocognitive impairment or side effects of use, a patient may elect to use ketamine separately, either while undergoing this or separately in our clinics. And as such, they may risk the neurodegeneration that can occur at higher doses. Interestingly, if neurodegeneration begins occurring, you may find higher levels of neurofilament light chain, which likely represent neuroaxonal injury and disruption. And this may be a peripheral biomarker we could use to screen the difference between separate personal use and treatment. As far as kratom, GHP, rohypnol, these are all undergoing research in ways we can more rapidly detect and monitor treatment. One of the new ones is surface-enhanced Raman spectroscopy in particular for these substances, and this is all obviously ongoing and needed. An interesting biomarker for MDMA that I discovered when I was preparing for this talk is the variations in serotonin levels in particular. And similar to ketamine, this likely is reflective of neuronal damage. Unlike ketamine, though, the specific changes that I'm including are only detectable in the CSF for now. Whereas ayahuasca and psilocybin have been studied in treatments of affective disorders, no pertinent biomarkers for today, but more evidence for what Dr. Monaghan was just talking about as far as inflammatory markers, both of which were being tested as treatment for affective disorders, and so far have only been shown to decrease inflammatory markers such as CRP. And then LSD as well is also at this point currently difficult to detect, but there are lots of interesting studies looking at fMRI, whole brain MRI, and using AI to monitor these different neurocognitive networks more precisely and quickly, finding interesting patterns in disruption between connectivity of certain circuits. Moving on to prescription substances in particular, clinical overview and physical exam are going to remain one of the most sensitive ways to detect benzodiazepine use. The most consistent effects observed include psychotic peak philosophy and visual analog scores of alertness, and common limitations also exist on our traditional UDS. When screening for benzodiazepines using traditional immunoassay tests, it is notable that nordiazepam and oxazepam, a common metabolite, are given an antibody antigen response leading to a presentation on your lab study of UDS positive, UDS negative. From there, other benzodiazepine compounds are then cross-tested for reactivity. In other words, low cross-reactivity of other drugs can result in false negatives for other benzodiazepines. And detection is especially important in the next section we're going to discuss, opioid use disorder in particular, as there's been a lot of research looking at patients and increased risk of relapse if undergoing methadone maintenance therapy and also taking benzodiazepines, either for one reason or another, I mean prescribed or illicitly. And the other study looked at veterans, noted that nearly half who died from opioid overdose were co-prescribed and taking benzodiazepines. And last, before we discuss opioids, I wanted to highlight anabolic steroids, because again, one of these need-based innovations is looking at the ways of doping and sports performance is now leading to microRNAs being potential ways to measure both to detect as well as trend these substances. So let's talk a little more about opioids in particular. There were 11 million Americans age 12 plus that reported using opioids in 2017, and of those about 2 million met the diagnostic criteria of opioid use disorder. After the opioid epidemic, when I was in medical school, there was a lot of research going through looking at ways to better detect and trend and ways to improve outcomes for these patients. One of the interesting changes that seems to come from that research is several changes to the inflammatory pathway markers, especially in special populations, which I'll discuss on the next slide. For example, as listed here, increased IL-10, which may be indicating the body's attempt to circumvent the detrimental effects of opiates. Up-regulation of pre-pro-dinotropin MRA and pro-dinotropin peptides in peripheral blood lymphocytes have also been appreciated, and notably, this only occurs in patients who have been prescribed or exposed to exogenous opioids. With respect to the labs, because once again, I brought my lab box back, there is a trend in that this picture that you're being seen is consistent with opioid use at a prescribed or perhaps if not prescribed use way for us to potentially screen for some of these things. There are significantly higher white blood cells, lymphocyte count, red blood cell distribution width, and in those with opioid use relative to those with no history of opioid use. In contrast, the percentage of monocytes is significantly lower, as well as monocyte-to-lymphocyte ratio and platelet-to-lymphocyte ratio, also with a reduced in patients with opioid use disorder. When I said special populations, this is where things are really starting to get interesting now, as far as the clinical standpoint, is there is a couple studies looking at patients with opioid use disorder and other things, and then measuring changes in biomarkers. An interesting study looked at early life stress or adverse child experiences and opioid use and compared. There were specific biomarkers just to some populations where patients with early life stress without exposure to opioids had significantly higher TNF. Patients with opioid use and early life stress in particular were found to have changes in lower levels of cortisol, and whereas only patients exposed to opioids without deemed early life stress had the laboratory values I was just disclosing. So something is happening, leading to us not seeing these lab changes. An additional interesting population is opioid use disorder in HIV patients. Patients with opioid use disorder have a much higher rate of being infected with HIV. Persons living with HIV and opioid use disorder also have these unique biomarkers listed here. Decreased CD68 has been found in frontal white matter and in the thalamus in patients with both of these opioid use disorder and HIV. And opioid use was also found to affect CD4 count in patients with HIV. Individuals with suppressed HIV-1 RNA at baseline had increased CD4 when they had consistent negative opioid toxicology. Or another way to say it is when they begin to no longer take opioids, their CD4 count increased. As far as predicting treatment response, buprenorphine also seems to have an interesting action as it is itself an opioid in long acting. It decreases monocyte chemotaxis. Recall that earlier in the laboratory studies I presented, I lowered a monocyte to lymphocyte ratio as a known biomarker of opioid use. There are also specific changes to IL-6, but they do not appear consistent across populations and may or may not predict treatment response. One of the interesting things when I start looking at inflammatory markers and how they predict treatment response is that it does appear that when patients have a higher level of inflammation in particular before starting methadone maintenance therapy or other specific opioid use therapies, they do tend to stay in treatment longer. As far as predicting initial opioid use disorder events, this is one of the more interesting sections of things that I didn't know before starting this talk. There is a lot of correlative data between vitamin D and initial opioid use disorder or dependence after surgery. It's been found that there is lower vitamin D in patients in, sorry, I apologize. There have been lower vitamin D, it has been found that lower vitamin D appears to lead to post-operative opioid use disorder more often, and the same patients required higher levels of opioids during their surgery. Separately, another study found a correlation that there's a higher prevalence of vitamin D deficiency in patients diagnosed with opioid use disorder. From there, they then started doing animal models like what Dr. Manahan was talking about. They produced vitamin D deficient rats. They also ended up having vitamin D deficient knockout mice and controlled their diet and found that when they were vitamin D deficient, they had a much higher response to exogenous opioids and compared to the other populations. And interestingly, when they corrected the vitamin D deficiency, they then lost that over-amplified response to opioid use. So given that vitamin D supplementation is such a low-risk supplement that we could easily be prescribing, perhaps we should be slightly more aggressive with our vitamin D supplementation in certain patients. And also, as we might recall, that vitamin D supplementation has a synergistic effect with SSRIs in treatment for major depressive disorder, and this interesting correlation will continue to trend for the next several years. Moving back to alcohol use in particular, remission prediction of alcohol use disorder was perhaps more complicated than I originally expected given all those peripheral biomarkers and things I was discussing at the start of my talk. Interestingly, it looks like MRIs and EEG monitoring right now are the most promising as potential ways for us to predict, but perhaps that's just demonstrating what we already know clinically because one of the best ways we can project whether or not someone will respond to treatment, and several of the treatments that Dr. Ruslan was talking about at the end of his talk, is their amount of cravings, different subjective effects that they disclose, and perhaps what we're seeing on MRI and EEG is substantiating what we're already doing. There is also a study that I'm going to follow for the next little while as it goes because it is a longitudinal study looking at single nucleotide polymorphisms in a large GWAS study of patients who have a mother and a father with alcohol use disorder, and perhaps that will lead to some interesting ways for us to predict in the future. As far as predicting treatment response to medications in particular, so far the pen alcohol craving scale is the scale I use, but whatever scale you prefer to look at how much patients are craving before you initiate treatment, higher levels of craving in particular does lead almost across the board to improved outcomes and improved opportunities for treatment from these medications. That, plus for acamprosate in particular, while several biomarkers originally came forward, it looks like that and aspartate are both pretty good at predicting whether or not they'll respond to acamprosate. From there, gabapentin, no biomarkers identified yet. However, similar cravings appear to improve the chances of them being able to respond to abstinence-promoting medications. And then as far as naltrexone response, naltrexone relative to placebo is known to significantly reduce alcohol Q-elicit activation in the right ventral striatum, more of those neuroimaging things I was just discussing, and the medial prefrontal cortex and the orbital frontal cortex, and reduce heavier drinking over 16 weeks. Overall, one of the things that I might have said a few years ago was I might have talked about OPRM1, but thankfully there is an expert panel of guidelines who are looking and trending this data because while we predicted that that might be a biomarker, they are currently recommending against. And I am including that link again just for those who didn't write it down in the future because the reality is these are areas of ongoing research that are going to continue to change. So overall, in closing for today, we went through several different clinical biomarkers with the goal of being able to supplement our treatment and detection of substance use disorders, of psychiatric conditions, and practice of psychiatry. This is in short, by definition, precision medicine and precision psychiatry to be able to use these specific things to improve outcomes for our patients. And I look forward to learning as we get more of these, such as the neurofilament light chain that just came out for ketamine, as well as all the different ones that have now been developed for alcohol use disorder. At this point, we'll move forward with any questions. Here are all of my different resources. And this is the QR code that allows you to keep these slides for future reference. Thank you. So the question about ordering the pharmacogenomic testing is a really good one. There are so many different companies, and I don't want to endorse any one of them while I'm in uniform. But if you just go to Google and you're like, hey, I want to do some pharmacogenomics, there you go. Some hospitals will also offer the Roche AmpliChip 2D6-2C19 as in-house sequencing because it's applicable to so many different organizations. That one is also drawn by LabCorp and other outside organizations. The other ones are going to be consumer-ordered or health care provider-ordered. You just go to the websites, enroll as a provider, and then you can order them. Also, you might have patients who self-enroll, do it themselves, and then they'll come in with this big stack of papers and be like, hey, doc, can you please interpret this for me? That's always fun when that happens. Yeah, some of them are also covered by Medicare and Medicaid. So if you just look at the websites for each of them, they'll give you some guidelines of when it is and is not covered by the insurance companies. But that is a whole can of worms in itself. So yeah, that is how to order pharmacogenomic testing. But I guess the other piece of that, my initial question, of course, was are you able to get this genomic testing done and paid for by insurance? And I guess you sort of answered that, at least on the getting it done part. The whole paid for insurance seems like it would be very complicated. On the other side of that, you also mentioned that you would, in certain circumstances, be able to prescribe above the dose based on someone being a rapid metabolizer or something like that. And I just wondered if you, and again, you may not be able to answer this, but is there pushback on that from payers when you are trying to prescribe beyond labeling when you have genomic reason to do that? I'm not sure if payers are there yet is what I'm trying to get at. So has that been problematic? And I know that you're in a military hospital and it might be very different than what you would find in commercial insurance, but I was just kind of wondering what that looked like. Yes, that's a really good question. And we are really lucky at the Walter Reed Program to be able to rotate with a lot of civilian programs to get a little bit more awareness of all of the insurance challenges that our non-TriCare providers face. Luckily, when it comes to dosing, there's not been pushback. It's more so for the approval of the medication itself if we're going for non-formulary medications. And the pharmacogenomics is helpful more so for pharmacolegal purposes so that I can say I have a rationale and here's a study of why I'm doing this thing that isn't recommended for everybody. So even if there is an adverse outcome, that I have more coverage for my team than just saying, well, I just kind of felt like doing it, which when you're going above FDA-recommended max doses, you really need to have very good reasons for why you're doing that. But that was a great question, ma'am. Thank you. Good afternoon. Biomarkers are sort of a hobby of mine. I mean, I'm a general psychiatrist, but due to a family situation, I'm trying to help somebody with autism. And there are a lot of, I don't know if you want to call them, non-conventional approaches that use certain biomarkers. And I'm just going to throw some of them out there. They're brain folate antibody receptor antibodies that some people measure. Have any of you heard of those? Yeah. There's a thing called central folate deficiency, and they found a lot of correlation with kids and people with Asperger, autism, and central folate deficiency. So you can get antibody tests that tell you if they're either binding or blocking proteins. And then if there are, you use high doses of folinic acid, which somehow bypass the regular folate receptor. That's one of them. And then there's this thing called the mitoswab. Have any of you heard of that? It's a way to check mitochondrial functioning based on a swab, as opposed to doing a muscle biopsy. And I'll just throw out a last set, which is sort of controversial and comes from the so-called functional medicine field, urine organic acids. In the promotional or the blurb, quinolinic acid was mentioned, and that's one of them that comes up a lot. There's a whole plethora of things that suggest or allegedly talk about brain inflammation and mitochondrial function. Just if you could comment on some of those things in general. I'm just noticing more and more biomarkers being commercially available. Thank you. Yeah. So I'm going to share something really quick, because this actually just came up in a patient case. So MTHFR genotyping can be helpful to identify methylfolate deficiencies. The evidence is kind of mixed for heterozygote intermediate metabolizers. But if you have a homozygous intermediate or poor metabolizer of the MTHFR gene, really helpful to supplement with methylfolate. When it comes to central deficiencies, I'm not too familiar with that. But in the patients that I do take care of who are homozygous for an intermediate or poor metabolizer of MTHFR, foot stomp that if their B12 is not replaced or is not within normal limits, when you supplement with folinic acid, they're going to develop a hyperfolate state, which can lead to an irreversible neuropathy. So if you're going to be going down that road of high doses of folinic acid, you need to also make sure that their B12 levels are normal to avoid permanent neuropathies. And that was my only TED Talk point. Can you guys hear me now? Yes. OK, good. Before we go, I promised I would alternate with the online group. So the question is, were the medications mentioned in the first section of the study the only psychiatric medications with sufficient data around adjusting based on CYP genes for recommendations, or were those just the antidepressants? And if so, are there other types of psychiatric drugs? The answer is yes. There's people working on all groups of the various psychiatric drugs. And as I think you hopefully got from the talk, we obviously have a lot of knowledge gaps still. But the ultimate goal is to be able to say, we know if this person has a hypoactive salience network combined with a chironic acid and whatever else, then we can treat this person this way and it'll work. But that's obviously some years away. But hopefully we're getting there. You want to add anything there? Yeah, I just want to add that. So the CPIC, which takes the absolute most conservative stance that is absolutely imaginable, they only recommend that small handful. But if you take a rational look at some of the smaller meta-analyses published and just a knowledge of the metabolism of these drugs, you can come to your own conclusions that I think are also perfectly valid. It's just the CPIC is unimaginably conservative. So their recommendations I have a lot of faith in, but there's still a lot of good other evidence out there. Hi, good afternoon. My name is Morgan Freese, I'm a PharmD. My question is regarding, you were talking about the PRIME study, and I know you mentioned the abstract that talked about the non-persistent effects, but I'm curious, on the primary endpoints for that study, what were they looking at for their primary endpoints, and did the study meet those primary endpoints or not? Yeah, I'm so sorry to be hogging this. But yeah, so for the PRIME study, the primary endpoints were, was the remission rate of depression superior between the pharmacogenomics and non-pharmacogenomics group? The second primary endpoint was, were fewer highly interactive medications prescribed in the pharmacogenomic versus non-group? And the first endpoint, there was no difference. Maybe they got a little bit better earlier on, but at 24 weeks, no difference. And then for the interactions, fewer interactive meds were prescribed in the pharmacogenomics group, but they didn't comment on things like adverse outcomes or side effects or tolerability or dosing and other important things like that. Gotcha. I was under the impression from reading the paper that they actually did show over 24 weeks that there was a statistical difference. So there was a significant difference at weeks 8 and 16, but at 24 weeks, there is not a significant difference. I'm going to triple check that right now. I'm talking about over the 24-week period of that program. So up until 24 weeks, yes, there was a difference. At 24 weeks, that small difference did disappear. Gotcha. And then I have one more question, and it's just in regard to, what are your thoughts on, I know there were several meta-analyses that were published in 2022 and 2023. There was the PREPARE study that was also published in Europe about pharmacogenomic testing and adverse reactions. So what is your opinion on all this new data coming about the clinical utility of pharmacogenomics? Kind of like I said, that it's helpful in patients who have multiple treatment failures. I think if we apply it grossly over an entire population, the individual level benefit is kind of washed out. So I think we need to be careful in who we're using it on in a cost-conscious model, and I think that way it's really helpful. But should we be giving it to everyone? I think the data is pretty compelling that now. Thank you very much. I would add that there's definitely a medical economics component here, but for sure, in the next 10 years, those genetic tests are going to get less expensive, right? And as that becomes the case, there will be a better case to be made that the juice is worth the squeeze, so to speak. So hopefully we'll get there soon. Thank you. All right, so I'll have a question for Dr. Sorensen so that Dr. Rusling doesn't have to keep on answering them. So the question is about how do we bring this into the clinic? And so any specific conversations that you have clinically with clients about pharmacogenetics? And so one example that I've come up with, I'm sure you have others, is CDT. So if I have a normal AST, normal ALT, and a question about alcohol use, I tell my client, hey, there's a lab that can tell you and me how well your body's doing at keeping up with your current level of drinking. Are you interested in knowing that? And so I've actually had people decide to start weaning off of alcohol because their CDT came back twice the upper limit than normal. Are there any other conversations that you've been able to initiate just using pharmacogenetics in the clinical sense? Yes. I'll keep it brief for the sake of time. But the same case from today, I saw him last week before coming here, and he was asking what his results were, because we've discussed in the past about how those results will monitor. One of the things that's important that I'd like to bring up now is I am a member of two special populations with respect to substance use disorder, if I were to develop a substance use disorder. I'm both in the military, and I'm both a physician. Now when you look at special populations, the reason why that matters is, can anyone guess which population in particular tends to attain recovery the easiest and have the best results? Professionals, like all of us, physicians, et cetera, because we have potentially, the one colloquial way to think is we have the most gain and the most to lose. So patients who are interested and aware of these concepts, such as patients who are physicians, telling them that I have this very sensitive and very specific test, and we're going to, with your consent, test you every two weeks, can be a pretty interesting way to assist in motivation during their process of recovery while they're still trying to schedule AA meetings and all of the above. This is a very wide-ranging question. From a transgender session, is there any effort to try to search for a biomarker for the underlying true gender before someone goes into a surgery, which is irreversible? A big decision has to be made. Is there anything to guide that? Yeah, that's probably too big of a topic for us at this point. There's a lot of specialists in transgender care, both psychiatrically and medically. I would defer that to them. Anyone else? I'll take the liberty of asking another question. Somebody on the panel works in the microbiome area. If you could say a few words about that. Really, are things like probiotics useful to treat psychiatric illnesses? Great claims are being made along those lines, so please. You don't know what box you just opened, but we'll let Matt talk for two minutes. Is the box Pandora's box? When it comes to probiotics, I have a lot of concerns, which is why I do not use them in any of my patients. The way that they are manufactured, there are not steps that are taken to actually confirm that compositionally they are what they say they are. If they were to do this, they would use 16S sequencing, which tells you the names of things that are in the pot, but not their actual genetic function. There are some ... I think if I were to use a microbiome-derived intervention, it would be with prebiotics. Things like non-digestible fibers and fermented foods, but specific probiotics. I'm very leery of organizations that sell very, very expensive things without a lot of validation steps that are involved, and not a lot of RCTs that show efficacy and outcomes. I think that the big ... If I were to put my money down, it would be that tryptophan metabolism and the production of ketogenic and ketogenic acid within the gut, leaching into the blood and then crossing the blood-brain barrier, is really where the money's at. If I were to develop a probiotic, it would be in manipulating tryptophan metabolism to increase ketogenic acid production and minimize ketogenic acid production. Regrettably, that's not how probiotics are being developed, but really should be on targeting specific metabolic pathways, because I care less the name of the things that are in your gut, and a lot more on functionally what's actually happening. I would also just add that, of course, the most important thing is what you eat. If you eat fermented foods every day, your microbiome will be good. If you stay away from ultra-processed foods and seed oils, your microbiome will be better. That's a whole separate talk that we do, but you probably do better by getting it the natural way, so to speak. I have two questions, one for Dr. Rusling and one for Dr. Monahan. First, for Dr. Monahan, since you just got done talking, I guess. The question I have is, we know that there's a chicken-and-the-egg phenomenon with inflammation and depression and vice versa, right? The question I have really centers around treatment. Is there anything that we can do to help address some of those inflammatory markers that we could decrease, potentially, to help? Yeah, I think that's the next step, right? Right now, the first step is to get a clear understanding at the individual level of proteins and mediators, what's going on. What came out of my research for this was a move toward understanding causality as opposed to association, which I think will be a really important step to inform that kind of a thing. I mentioned Mendelian randomization in part of my talk, which I think is a really powerful tool to look at associating, understanding what the nature is between a biomarker and the outcome. Just to be really brief with it, I think a good example that we can understand MOR with is smoking. If you take two populations of people, one that's smoking, one that's non-smokers, and you can identify genes that are associated with heavier smoking among smokers. If you take those two populations and you look at a gene that is associated with heavier smoking and you find that that gene is associated with a higher all-cause mortality in smokers, but not in non-smokers, that suggests a causal relationship between smoking and death. Using that kind of a tool with inflammatory biomarkers, I think, will help us understand the causality, and then we can take the next step into actually manipulating them for therapeutic purposes. Makes sense. Thanks. For Dr. Rusling, you kind of touched on phenoconversion a little bit, but I guess my question for you is, in the clinic, how do you assess that phenoconversion if you have pharmacogenomic information available? Can you clarify a little bit on what you mean? You talked about drug-drug interactions and how that can change how somebody metabolizes the medication. How do you deal with that in the clinic when you have that, say, pharmacogenomics? You know somebody, say, a normal metabolizer for 2D6, or an intermediate metabolizer for 2D6, and you're giving them bupropion. How are you identifying that those two things are going on? Do you use EMR? Do you have a pharmacist that reviews the medications? That kind of thing. Yeah, absolutely. The EMR question is a really good one, and oh my gosh, is it a pain, because pharmacogenomic results aren't necessarily stored within the EMR in a way that's easy. Like in our EMR, they're typically stored as scanned PDF documents. So when I have that result that a patient has, I'll print it out and give it to them, and kind of put the responsibility on the patient and say, it's important that your docs know about this. Please bring this with you to follow-on appointments, especially for psychiatry, because I agree it is a challenge. Hi, question for Dr. Monaghan. Great talk to all of you, by the way. The Adderavone is mostly for ALS patients, but have they ever looked at behavioral changes in ALS patients after administration of Adderavone? That's a really good question, not that I saw. But it did come on the market in 2017 as the, I think it was the second drug that was FDA-approved for ALS after, oh gosh, Rilisone, yeah, right. And so I haven't seen that, but that would be a really interesting study to look at. Thank you.

psychiatry

biomarkers

pharmacogenomics

psychiatric medication

therapeutic outcomes

inflammatory biomarkers

affective disorders

kynurenic acid pathway

substance use disorders

personalized treatment

clinical practice

biomarker research