My name is Dr. Balwinder Singh. I'm a consultant psychiatrist at Mayo Clinic in Rochester so I welcome you all for coming to the session. I am co-chairing this session with my colleague Dr. Gustavo Medeiros and we have three speakers and a discussant. So Dr. Medeiros, he's a Mood Fellow at Johns Hopkins School of Medicine. Dr. Jennifer Vandervoort, she's my colleague at Mayo Clinic and she's an Associate Professor of Psychiatry and the Director of the Child and Adolescent Clinic there. And Dr. Gisele Skaiene, she's an instructor at the UT Health Houston, Texas, and she's a discussant for our talk. And this is our agenda for today's talk. So we'll take questions at the end of the session except for Dr. Medeiros. He has to run, he's at the APA Junior Colloquium, so thank you for coming. But after his talk, feel free to ask questions to him and then the rest of the speakers will take questions at the end. So Dr. Medeiros will talk about blood-based biomarkers of antidepressant response to ketamine and s-ketamine. I will discuss about some metabolomic biomarkers of racemic ketamine. And Dr. Vandervoort will discuss about anti-anhedonic effect of ketamine and in vivo increase in mTOR protein expression. And then we'll have Dr. Skaiene as a discussant. So I'd like to invite Dr. Medeiros for his talk. Thank you. So very happy to be here. I think that this is a pointer maybe. No, okay, not a problem. We go there. So just a little bit about my background. I'm originally from Brazil. I did medical school and psychiatry residency at the University of Sao Paolo. Willing to have more research opportunities, I came to the United States eight years ago. I redid my residency here. I finished at University of Maryland and now I'm doing a fellowship at Hopkins. I finished my fellowship next month in July. I will be starting back to University of Maryland to help them build the depression center. Okay, so our talk today is about blood-based biomarkers of antidepressant response to ketamine and esketamine. I have no disclosures. Sorry. Okay, perfect. Perfect. I have no disclosures as a good fellow. So good. Yeah, we start by there. So we have some variability, right? Response to ketamine and esketamine varies depending of the study, but mostly somewhere between 40 to 60 percent of patients with TRD, they respond. So approximately half of the patients, but we are not very good in predicting that. So our goal during our fellowship at Hopkins and now as assistant professor is try to better understand which are the patients that respond to ketamine and esketamine. We decided to start taking a step back. So yes, we want to get the patients, we want to kind of study blood-based biomarkers, brain-based biomarkers, clinical variables, but what do we study? Which are the most promising biomarkers? So we took a step back and we said let's do first systematic reviews and meta-analysis to try to see which ones are the most promising, and then we can try to build from there. So this is a work that we divided in three parts. So first we did a meta-analysis on blood-based biomarkers, which is the one that we're going to present today. The second one is one in brain-based biomarkers, which was accepted this month, and the third one in clinical variables, which probably we're going to submit by the end of the year. So here we are focusing on blood-based biomarkers of response to ketamine and esketamine. This is a 15 to 20 minute talk, so there are a lot of things that we're not going to be able to discuss. If you're interested, I strongly encourage you to see the publication. Okay, so blood-based, we all are in, in, we have some sort of clinical contact, and we understand that's much more acceptable and much easier to get blood from the patients than to doing, to do other modalities. So as a result, it's the one that is most investigated. We see that in the literature, numerous biomarkers were reported, but there were a lot of inconsistencies. Samples usually were small, so we decided to conduct this meta-analysis. First, we need to understand the difference between two types of biomarkers. So we have the baseline blood-based, this is actually any kind of biomarkers or predictor, right? We have the pre-treatment predictors, which are the baseline. So before I give ketamine, before I give esketamine, what are the characteristics that predict a good response? Another type is longitudinal. So after I give the medication, what are the changes in blood levels, in brain-based biomarkers, clinical markers that will predict? So in our talk, we are going to first start with pre-treatment and then go to the longitudinal. Something that we noticed in the three different meta-analyses that we did is it's much easier to find longitudinal biomarkers than baseline biomarkers, at least it's what we see. So baseline biomarkers, you are comparing one patient to a lot of different patients, right? So the variability is much larger. While longitudinal, you compare what happens to you after, so you somehow are in your own control. But what we really want are the baseline biomarkers, because before treatment, we want to know who's going to respond or not. So challenging field. Okay, so we think we did a pretty good job in terms of methodological aspects. We pre-registered the protocol at Prospero. We followed the PRISMA guidelines. We searched in five databases, also did some manual searches. So this is an important point. The searches were conducted in 2021, August 2021. So what we are going to present today is not the whole literature, and we did meta-analytical calculations if there were at least three studies in the same biomarker. Okay, so we found 56 manuscripts that were included in the systematic review. In the meta-analytical part, we tried to get the most data that we were able to. We got almost all the papers that we wanted. I would say that 95-98% we were able to get, but we just conducted meta-analysis with biomarkers that were investigated at least in three independent samples. So 26 studies were included in the meta-analysis. Although the title is Predictors of Response to Ketamine and Esketamine, all the 56 studies, they gave ketamine. Two studies also gave esketamine. So in theory, what we are going to talk here is related to racemic ketamine. In total, more than 460 biomarkers were examined. The most frequently investigated were neurotrophic factors. The most frequent one there was BDNF. Levels of ketamine and ketamine metabolites. The most frequent metabolite was norketamine and inflammatory markers as well. So basically, the talk will focus on these three types of biomarkers. Neurotrophic, inflammatory, and then levels. Okay, so we start with the baseline. So let's try to understand a little bit how these tables work, right? So we have the biomarker there on the left, and then we have... We need to show the table, right? Okay, so we have the biomarker, then we have the number of studies, then we have the sample size, the responders and non-responders, and then we have a standardized mean difference. So let's go to BDNF, for example. So 11 studies, about 330 people. So the SMD there, if the value is negative, that means that people with lower BDNF would respond better. The opposite, actually. That responders would have lower levels of BDNF. So what we see is that both BDNF and VEGF, the SMD very close to zero, so not statistically significant at all. So baseline levels of neurotrophic factors, they did not predict response to ketamine. And then we have the inflammatory markers there. So these studies were varied from three to six studies. We see, again, that responders tended to have a lower level of inflammatory markers. So we see there that all are negative, but they don't reach statistical significance. So this was a little bit surprising, because our hypothesis in the beginning was people with ketamine have some anti-inflammatory properties, so maybe people with higher level would respond better, and you have one or two papers showing that. What happens is that one or two papers, they are very highly cited, and then you have a lot of papers that are not statistically significant and that are not really cited, and you have a small that even not get to be published. When you put everything together, it seems that there is a trend that if you have lower inflammation, you respond better, just like SSRIs or SNRIs. But again, did not reach statistical significance. When we see also even two, so we did meta-analysis for studies, for biomarkers that were investigated in at least three studies. When we see the systematic review, so those papers that, the biomarkers that just have two studies, we did not find a clear consistency there neither. So baseline, the take-home message is not very promising. Potentially lower level of some inflammatory markers, but we need a larger sample. So longitudinal, right? So here we have two time points of assessment. People measured at baseline, they received the medication, and then we measure after the medication. Good. So here is BBNF. So how we see this is on top we have the responders, on bottom we have the non-responders, and then there on the graph we see that if it's towards the right, that means that's higher after the treatment. If it's towards the left, it's lower towards the end of the treatment. So what we see, we see increase in BBNF level in responders, but in responders we see that pretty much is the same level. But it's very important for us to see some details, right? So it's a statistically significant increase, but the first thing is there is a lot of variability. When you see the studies, you see that it's kind of huge standard deviation there. And the second thing is that although it's statistically significant, the effect size is 0.26. So it's very small. We do find that there is a dose response, dose relationship there. We wanted to see if studies that just give one dose, they saw increase, and then if there was a progressive increase with more doses. We were not able to do that because of the number of the studies, but we see when we divide single effusion versus multiple effusions. So there is a trend that over time this BDNF increases. Again, it's 0.35 if you see there. So it's a small to moderate effect and probably at that point you're going to see clinically if they are responding or not anyway. Okay, now we go to longitudinal, the inflammatory markers. So again, we have each type of inflammatory marker there. Interleukin 6, TNF alpha, and here we have responders and non-responders. So interleukin 6, we see that's pretty much in the middle. Pretty much all of them we see that not statistically significant, not in responders or non-responders. So pretty much we don't see a clear trend of change in inflammatory markers. Again, it's the same thing that we see. So if we go there to interleukin 6, there is this paper Yang 2015, a very nice paper, but was the first one that was positive. That paper is highly cited, was published in biological psychiatry. And then you have several papers that do not see anything or even that paper of Dr. Zarate's group that sees an increase. So there is a little bit of publication bias there. There's at least a little bit of attention that we give also to the papers, especially because of what kind of journal they are published. So those were the ones that we were able, that we had more studies. So we had six in each of them, but the other inflammatory biomarkers we did not see neither. Finally, we go to ketamine in the nor-ketamine levels. So we don't have two time points. So although it's longitudinal, that means after you give the ketamine, we are comparing basically the concentration between responders and non-responders. So what we see there is that if it's toward the right, it's higher in responders, towards the left is lower in responders. We see that's pretty much close to zero. It's not statistically significant. Again, this is mean concentration, right? There are other questions that we could ask here. Is there a minimal concentration that's effective? Is there a range? We are not touching that. Basically, we are just talking about average concentration. Average concentration. So it was, even when we decided to publish, we did not have a major positive finding in this meta-analysis, right? So mostly we did not see any baseline predictor. When we see kind of longitudinal, we see BDNF, but the effect size is small. And then we started the paper, I think, that turned into a broader discussion about how to examine blood-based biomarkers. It's very complicated. So we are trying to predict something in the brain. The brain is kind of just like the most well-protected organ that we have. We have the blood-based barrier, and we have so many peripheral confounders that's extremely difficult. We did a quality assessment of the papers. This was rated by two independent raters. We see that a lot of studies had high risk of bias. So we see there, we have each study in a line. The last circle is the overall risk of bias, we see that most studies they had higher bias. That means that there were not many descriptions of were the patients fasting, not fasting, what was the time, did the statistical analysis control for gender, control for age, do we know about concurrent medications, do we know about comorbidities, and then there are so many factors that, and then we put in the supplemental material for each of the biomarkers, we see there for example the brain, BDNF is the first biomarker, and then we see which are the factors that are associated with higher levels and lower levels. It's so many factors, if you measure what time of the year you measure, if they smoke or if they don't smoke, if they did exercise in that morning or not, so it's extremely complicated and just like our job, not only to try to investigate larger samples, but try to capture that in a more standardized way, because replicability is a big problem in psychiatry, replicability of blood-based biomarkers is another problem as well. So take-home messages here, okay, we have more and more literature, but at this moment very little evidence that they are associated, they are predictors of response to ketamine, and no current evidence of clinical utility. We see that responders had a statistically significant but small effect increase in levels of BDNF. The more treatments you give, the higher the effect size, and we need to conduct larger trials with more comprehensive consideration of the confounders. So this is the team that is helping us with the systematic reviews and meta-analysis. Those are our mentors, Dr. Goyes from Hopkins, Dr. Gold from Maryland, Dr. Zarate has been amazing with us as well, and thank you so much. So this time if you have any questions for Dr. Medeiros, please feel free to ask. Would you mind using the mic, please? Hey Gustavo, great talk. My name is Mani, I'm with the Zarate group, actually. I was gonna ask you, do you think with improvements in testing, do you think it's still worthwhile to pursue biomarker, blood-based biomarkers for ketamine response? I think that it's worthy. We need better control, you know. I think that we do blood-based studies as psychiatrists. I've seen that so, so many times in residency and in other discussions. Oh, I'm studying ketamine, okay, take some blood and let's run some tests. Which tests? No, let's run pretty much everything that you have there. So I think that we need to be much, much more strict. I think that some genetic markers that we can, we can try to understand this. So the family history of alcohol use disorder is something fascinating. So we see that people with a positive history of alcohol use disorder, they tend to respond better, right? So, so far we collected from four studies, independent samples, and we see that the odds ratio for that is about three to four. So it's a moderate effect size. So I'm very intrigued to see what's behind that. So another point is that, I think that my here, that the symposium is talking about opportunity and challenges. So I think that my systematic review, our systematic review was much more about the challenges, right? But they're going to present as well potential metabolomics targets, potential biomarkers related to the mTOR. So I think that the idea is we should not abandon that, but we should do better research, and I include myself in that. We should try to analyze larger samples, which we should try to be more standardized. We should better think before we do the study what are the biomarkers, and at the same time assess the confounders for those biomarkers. I think that there we are in a better route. If I had to put my money, I would put more maybe towards brain-based modality. So we are trying to focus more on EEG, those increasing gamma power, because such a rapid acting antidepressant that we can see we have a good temporal perception with EEG. I don't think we should abandon. I think that we should be more strict and more demanding with our research. That's very helpful. Do you think the subtyping of using the biomarkers will give you clues to the clinical subtyping as well? That might be effective. Can you say that again? Like for example, like if you look at biomarkers that might predict metabolic profiles, do you think that would guide maybe, like you mentioned the alcohol history, maybe there's something in the metabolics, right? Yeah, there are, Bob Winder will talk about that a little bit more. And the outbreak idea, I think that's developed multivariate models. So those univariate analysis, they are very limited. If you see the clinical utility, the prediction power, and we are talking again about larger samples. And when you move to real-world samples, there are so many scatamine clinics right now. We have so many catamine clinics as well that collaboration becomes something very, very crucial. So all of us, the four of us, we are in the NNDC, National Networks of Depression Center. So that's just like one thing that helps is collaboration between different institutions as well. But I think that the ultimate goal is multivariate factors. You have a patient, you put their family history of alcohol use disorder, second variable. You think about predictions of increasing gamma power, third one, BMI, fourth one, and then you have a very beautiful model that says, okay, the chance of responding to catamine, the scatamine is very low, maybe should pursue TMS, ECT, something different. Thank you. Thank you. Any other questions? If not, I think I'll start my talk. All right, so my talk is focused on metabolomic signatures of catamine response. These are my disclosures. But I'll start a little bit with neuroimaging since I like that question, you know, about the blood bias biomarker and the challenges with those. And when you look at the depression itself, it's such a heterogeneous disease, and then we don't have a biomarker for depression, but there are studies looking at the neuroimaging field where we, you know, see some inkling there. So for this study, we start with neuroimaging, but it's hard to use neuroimaging in clinical practice. It's easier to do a blood draw. You have a patient with depression, you can get blood draw, and you can, you know, if you can find a biomarker, you can predict who will respond to catamine or other treatment intervention, but visual thinking at this point at least. So I'll talk about neuroimaging just in brief. Magnetic resonance spectroscopy, it's a unique non-invasive methodology to investigate in vivo brain biochemistry. So if you are interested in glutamate GABA, GLX, so looking at one of the hypotheses with ketamine that being an NMDA antagonist, it leads to increase in glutamate, or if there's a relationship between that GABA and glutamate, that MRS can be a nice technique to kind of measure those metabolites. And anterior cingulate cortex, so that's one of the emotion regulation center has shown to be associated with depression, so it's a perfect target. So on the figure you're looking at this, that's nice small voxel right in front of cingulate cortex, that's where we see the anterior cingulate cortex. In patients with major depression, there's data suggesting that they have, they can have low glutamate, GLX, GABA in the anterior cingulate cortex. Increase in GABA has shown to be a potential mechanism, and we look at some of the prior studies with antidepressant, TMS, and even ECT, that increase in GABA could be a predictor for response. It's ketamine, do we see that in ketamine too? Now the other thing with the conventional MRS scanning is we look at limited time points. It's hard to do it throughout the infusion, so most studies look at the baseline, or they look at end point, or every 13 minutes. So there's sort of a lack of continuous measure, so we can see what's happening with the GABA or GLX level pre-infusion, 13 minutes you have to stop the MRI, and then kind of adjust the scanner again, or at the end, but we don't know what's happening during the infusion. So this is an ongoing open-label, non-randomized feasibility trial, so high risk of bias, you know, I don't claim that we are, you know, predicting everything, but it's more of a hypothesis generation. It's a pretty tricky trial. We include adult patients with treatment-resistant depression, 18 to 65 years age group, with at least moderate to severe, moderately severe, or severe symptoms of depression. Now what we do, that once we identify the patients who are depressed, have very minimal anxiety, who can stay fasting for at least six hours, and who can stay still within the MRI with no claustrophobia, no contraindication to MRI for an hour, so you need to find those patients. Good luck finding those, it's really hard to find those, it's very hard. So what we do is that patient will receive a single IV infusion in the scanner, so it's a 40-minute infusion, but before we start the infusion, first 13 minutes, we'll do an anatomical, check for anatomical abnormalities, everything looks good, and go ahead with the scanner. Patients will be monitored for additional 60 minutes after the infusion, and then we measure metabolites at baseline, at the end of infusion, 60 minutes, 100 minutes, and 24 hours. So this is, this data is from, the imaging data is for just the seven subjects, and this is where the challenge with the neuroimaging studies, because it's really hard to sit, to stay still in the MRI scanner once you get the infusion, because the moment you move, the voxel is gone. So we started, and we learned it the hard way, we started with 12 patients, and we lost, initially we lost a data for about five patients, two pump failure during the infusion or software update, and two patients that moved during the infusion, so go figure. But we had some interesting findings, so I thought it's worth presenting that. So the imaging data I'm presenting today is from just the seven patients, but the blood-based data we have is for more than seven patients. So mean age was 45 years, BMI of 29, 86% had comorbid anxiety. So we developed this novel methodology where we could measure the GABA, glutamate, and GLX measured throughout the scan, throughout the infusion, rather than at every 13 minutes. So I work with Dr. John Port, he's a neuroradiologist, and he developed this technique by which we can measure GABA or the metabolites level at every minute, rather than every 13 minutes. So what we are looking here in that figure on the right, on the x- axis is time in minutes, on the y-axis we are looking at the percentage change in GABA during the infusion, and this paper was published in Psychiatry Research. So the red line, what we are seeing, those red lines are patients who remitted with ketamine at 24 hours, based on Madras score, and the blue patients who did not remit at 24 hours. And what we start to see is that there's a sort of an increase in GABA level during the infusion, because there are prior studies where they showed that there's some increase in GABA and even glutamate, but they were only measuring at baseline, 13 minutes, 26 minutes, 39 minutes, but here we could see that clearly it's increasing, and it started to come down, but when at the end of the infusion it stayed pretty high. The next thing we saw is that the peak anterior cingulate GABA during the infusion was associated with next-day remission, and also correlated with the degree of clinical improvement. Then we looked into, it's a small sample size, but we explored why are we seeing that increase in GABA, because prior studies suggested that should be the glutamate. We did not see any significant finding in glutamate, whether it's a small sample size. So we divided the patients into remitters and non-remitters, and at baseline, which you're looking on the left, is that remitters had a GABA deficit to begin with as compared to non-remitters, and after, when we look at, you know, they received the ketamine infusion, there was no difference between their GABA concentration between remitters and non-remitters. So maybe the patients who are improving with ketamine, they have GABA deficit to begin with. Can ketamine is maybe just attenuating that deficit and normalizing that GABA deficit. Now let's go to the blood biomarker. What's happening there? So metabolomic studies in depression, so that in metabolomics, what we do, we use a panel of, a panel where we look at different metabolites. Some of the studies are targeted, where we target one panel, whether it's acyl carnitines, but majority of the panels, they are, we call it known targeted metabolites, where they look at different classes of metabolites, almost 500, 600 metabolites, looking at sphingolipid, lipid, ceramic, and then we look at the changes. What we have seen in the studies in major depression, that there are changes in mitochondrial beta-oxidation, lipid metabolism, and neurotransmission. In one of our earlier studies, we mapped global changes related to use of SSRIs and defined pathways implicated in response with that, but it took eight weeks. So what we are seeing on the right side is a paper published in Journal of Affective Disorder, but what we did here, rather than looking at depression as a one entity major depression, we kind of divided depression to different phenotypes based on symptoms. So when you're looking at CD is core depression, NVSM, neurovegetative symptoms of melancholia and anxiety, and we saw a different profile for each phenotype. So we saw some very interesting phenotype, and what we're looking at is just focus on the acyl carnitine. So from C0 to C5 is the short chain, and C5 to C10 is the medium, and then C6 into 18 is the long chain. So very different sort of profile based on the phenotype. Now the validation of peripheral markers of ketamine response would potentially transform the TRD practice with a greater precision, so that's the hope with the neuroimaging and combining neuroimaging with the blood marker. So connecting the peripheral and central with the use of ketamine, either to metabolomics is one of the tools, not the only tool, and using imaging data can provide greater insight that inform about central changes. Now this is a paper published by my metabolomics mentor Dr. Rima Khaduradow, their group in collaboration with GNJ, and in this paper they showed that they use just the metabolomic platforms and with both ketamine and s-ketamine, and they showed that both drugs alter metabolites related to tryptophan metabolism, urea cycle, at two-hour post infusion. They also noticed changes in blood glutamate and circulating phospholipids were associated with decrease in depression severity. So with this study we aim to investigate changes in baseline metabolites after a 40-minute infusion with the ketamine treatment in patients of the treatment-resistant depression. So we utilized two approaches. First we used a non-targeted metabolomics platform where biocritics about 630 metabolites, looking at 26 biochemical classes including lipids, triglycerides, amino acid, biogenic amine, so it's a very standard panel we used. MedDress was used to measure depression symptoms. And so we define changes in metabolites as early and late. Early means change in metabolites from baseline to 40 minutes or at the end of infusion. And late, looking at from 100 minutes to 24-hour post-infusion, what's happening there. Standard statistical technique was used to measure the metabolites. Tongue twister, right? So what we are looking here is early and late metabolomic signature of intravenous ketamine. So on the top, we are looking at baseline to 40 minutes. And so when you look at the red dotted line at the top, and those orange sort of circles at the top, they are reflecting as metabolites. They had at least 50% increase. And then you look at the lower dotted red line, they are showing the decrease. And so what we saw is that there were significant changes in multiple metabolites or multiple platform. About 10 acylcarnitines, especially the short-chain acylcarnitine, they decreased within 40 minutes. And some of those, they bounced back when we look at the lower part of that figure, 100 minutes to 24 hours. So it's almost like some of the short-chain acylcarnitines, they are going down and they are bouncing back. It's almost like they are rebounding within 24 hours. Now we saw something similar with SSRIs, but at eight weeks with different phenotypes. And here we are seeing changes within 24 hours. This is a figure just looking at the late changes. So acylcarnitine amino acids, bile acid, biogenic amine, ceramide. So a lot of lipid and energy pathway metabolites and hormones. So the hormone listed here is the cortisol. We saw an increase in cortisol within 40 minutes with ketamine and then sort of normalize after that. Now, then we looked at, specifically looking at the lipids. Very similar, we saw a significant decrease in 93 triacylglycerol. And in the recovery phase, significant increase in 27 swing lipids. So within 24 hours or even less than that, we are seeing significant changes in these metabolites, which are involved in the energy pathways. Then we looked at, okay, what happens? Do they correlate these metabolites, the change in blood metabolites, do they correlate with central metabolites at all? Do they correlate with the central GABA? So on the top, we are looking at correlation of 40 minutes metabolite change and phenotype, so with Madras. And then at the bottom, we'll look at the GABA brain. This is, we are seeing a lot of the triglycerides. So the TG stands for triglycerides. A lot of the triglycerides, we are seeing a positive correlation with GABA. And there's some negative correlation with some other metabolites as well. Again, our sample size is small, but we are still seeing some signal here. And as I said, these are all hypothesis-generating findings. We saw early change in some triglyceride and correlation with the central and tiered singlet GABA levels. This is one of those, the TG18.3. Early changes are highly correlated with change in Madras as well as GABA. So the main finding from this analysis was the rapid utilization of triglycerides with ketamine treatment, followed by recovery within 24 hours, highlighting a possible role of lipid metabolism. Lipid peroxidation plays a vital role in CNS. Homeostasis and probably a lipid-based pathway is involved in depression and with ketamine response as well. And we noticed some correlation with the brain imaging finding. Then we looked at, is there any correlation with the GABA blood and brain level? So we saw that the blood GABA level increased from 40 minutes to 24-hour post-exposure. There was a positive correlation. It was weak. Our sample size is small, so the p-value was 0.08. It's a small positive correlation between the blood and the brain GABA level, but it did not reach the statistical significance. How about the correlation between glutamate and remission status? So as compared to the earlier studies published by Rotoroff et al., in our study, we did not see that glutamate did not change significantly in our study. However, we saw an inverse relationship between baseline glutamate with MedRAS, and remitter seems to have, they had a higher drop in the glutamate after 24 hours. And then we looked at the targeted acylcarnitine platforms. I was particularly interested in this because in the SSRI cohort, we focused on the acylcarnitine, and we wanted to see, what is ketamine doing? Are we seeing a similar signal here? So this paper is published in Psychiatry Research, and what we saw, very similar to SSRI, is that with ketamine, there's a change in the short-chain acylcarnitine. They are decreasing, except C2. C2 is the acetylcarnitine, which is involved in the TCA cycle. That's sort of moving in the opposite direction. We saw something similar even with the SSRI cohort. So probably the mitochondrial, it's involved with the mitochondrial fatty acid beta-oxidation, and two metabolites in particular, C3 and C5, they increased at 24 hours. So it's almost like there's a late-stage accumulation of C3 and C5 acylcarnitines, likely as a compensatory effect of amino acid consumption because these are derived from the branching amino acids there. Okay. So this is just looking at, we focused on those C3 and C5. We saw that remitters had a higher drop in C3 compared to non-remitters at 40 minutes, and there was no correlation with the brain GABA here, and drop in C3 within 40 minutes, it correlated with change in depression symptoms. Pretty similar finding with C5. So there's something happening, especially with the small-chain acylcarnitine. Limitations, these are very small sample size, hypothesis generation, lack of placebo, the role of prescribed medication cannot be ruled out. All the patients were fasting for a minimum of six hours, but they continued their antidepressant, so we cannot rule out the impact of other medications there. Now, then we did this systematic, we'll quickly go over this. It's not published yet, we just presented some of these findings at SOBP, but we did a very focused systematic review on just the metabolomics biomarker, not the BDNF, not the inflammation marker. I'm more interested in what's happening with the energy pathways on all the metabolites. So we identified five studies, and the problem with the studies is they're all different. There's no mention, or none of the studies, they follow the same protocol. They have a different technique, or different time points, or the fasting blood levels or data is not provided. So these are the five studies looking at patients with depression and changing metabolites. You know, some of the main findings, as we saw with the C2, but it's not consistent. So when we look at the Modell study published in 2022, there's a two-hour post-infusion, there's a decrease, whereas when we look at our study, there was an increase, at least in the early stages. So there's no consistent finding with those metabolites, but we are seeing some of these metabolites which are involved in the energy pathways, they are getting involved within 40 minutes to 24 hours. So ketamine is leading to alteration, and several metabolites. Studies suggest involvement of energy metabolism, and in healthy control, acetylcarnitine, so C2, decreased post-infusion, whereas inconsistent finding were observed. So we need larger and more rigorous studies to better understand the role of these energy pathways and metabolites. And I think these are some of the references, and this is my team on the left, the Mayo team, my primary mentor, Dr. Mark Fry, John Porte is my, Dr. Porte is my neuroradiology mentor, Dr. Vandewoordt, my colleague who's here, and then other colleagues in the metabolomics team, my Dr. Kaduradaw, my metabolomics mentor, so thank you. I do want to highlight my disclosures. Obviously I'm an employee of Mayo. I did have a research study with Asharax Health, it has nothing to do with today's talk. But I do want to highlight that I am talking about IV ketamine that is not FDA approved for any type of psychiatric condition. So I am using ketamine off-label in today's talk. So by way of introduction, let's start with the concept of anhedonia. Anhedonia is defined as a diminished interest or pleasure in nearly all activities. We know that this is a core symptom of depression. It is a driver of morbidity and a predictor of suicidal ideation independent of other depressive symptoms. Anhedonia in and of itself as a symptom is a barrier to patients seeking care and engaging in care. And even when patients do seek care and engage in care, anhedonia is still a barrier to functional recovery and a predictor of poor long-term outcomes. Anhedonia also has a notable negative impact to the economy, particularly when you're looking at things like absenteeism and reduced productivity. Unfortunately, the effect of our antidepressants on anhedonia is understudied and really more research is needed. There was a systematic review published in 2019 suggesting that our available conventional antidepressants really have a very beneficial effect when particularly looking at anhedonia. Acetalopram for example in this study was inferior to CBT when targeting anhedonia. There's also other literature out there that suggests that SSRIs, SNRIs can further blunt the reward as well as the emotional system, furthering the anhedonia's effect. So having said all of that, there's a growing interest in specifically targeting anhedonia as a symptom of depression. Understanding these underlying kind of biological pathways is an active area of study. And one of the pathways that I want to highlight today is the immune system. So immune dysfunction in pro-inflammatory states have been associated with both depression as well as anhedonia. These impairments impact metabolic homeostasis, neurotrophic signaling, as well as synaptic plasticity. Now we have preclinical as well as clinical data that suggests ketamine modulates the immune and neurotrophic signaling and increases synaptic plasticity. There's also data that very early came out of the NIH showing that ketamine reduces anhedonia. However the mechanism of the anti-anhedonic effects is still unknown. And before we go further, I do want to just make sure we're all on the same page with the mechanism of ketamine. So here you see ketamine blocking an NMDA receptor that stops GABAergic inhibition, allowing for glutamate release and increased glutamatergic transmission through the AMPA receptor seen here in yellow. Through downstream effects we see BDNF release and then activation of mTOR. From there we see increased protein synthesis and increased synaptogenesis. So mTOR is a critical molecular mechanism that regulates both inflammatory as neurotrophic signaling cascades. So it starts to beg the question when we put these pieces together, can mTOR be a biomarker for anti-anhedonic effect? And so that was the question that really started one of our studies. Now our study has two objectives to it. There's a clinical objective and a biomarker objective. I'll separate these objectives as I go through the methodology and the results. So the clinical objective is to investigate the correlation between change in anhedonia and change in depression after ketamine infusions. This was a two-site study with Emory University as well as Mayo Clinic. The biomarker objective was to conduct a post hoc analysis at the Mayo Clinic site only to evaluate the association between anti-anhedonic effects of ketamine and a change in peripheral immune cell mTOR protein expression. In terms of our methods, again, two cohorts, Emory University had the larger sample size at 44, Mayo Clinic at this point was a bit smaller with 12 subjects. All were adult patients, all had treatment-resistant depression. We define that as at least two failed trials of an antidepressant at adequate dose and duration. We also included failed TMS and failed ECT as one of the two trials if it was applicable. Both sites used IV racemic ketamine at 0.5 milligrams per kilogram given either two times or three times a week for a total of up to six IV infusions. Now at Emory, they used the PHQ-9 as the primary outcome measure. And we looked at anhedonia as a single item on the PHQ-9. This is reflected in Question 1. We also looked at anhedonia phenotypes in the literature. And this is based on research domain criteria, RDoC criteria. And the two phenotypes were core depression and interest in activity. So for the core depression phenotype, we're not only looking at anhedonia, but then we included down and depressed mood reflected in Question 2. And then for the interest in activity phenotype, we included anhedonia, tired, fatigue, energy levels reflected in Question 4, and concentration reflected in Question 7. At Mayo we utilized the Madras, a similar concept. We looked at anhedonia as a single item, Madras Question 8. And then again looked at the core depression and interest in activity. For the Madras, it's referred to Question 1 and 2 as apparent sadness as well as reported sadness along with anhedonia. And then for the interest in activity, we looked at concentration, lassitude or the difficulty initiating activities as well as anhedonia. For the biomarker component of the study, again this was at the Mayo Clinic site only, blood samples were collected immediately before and immediately after the very first ketamine infusion. We sent these samples to our collaborator, Dr. Sue Tai at the University of Queensland. She isolated the PBMCs and then looked for changes in total mTOR and changes in phosphorylated mTOR levels using Western blot. For the analysis of the clinical objective, we used linear regression to investigate the relationship between the change in anhedonia, core depression, and interest in activity phenotype scores and percent change in modified total depression scores. Now the modified total depression score was calculated by excluding the anhedonia items that I shared with you on the prior slides from the total depression rating scale scores, whether it be the PHQ-9 or the Madras. For the biomarker analysis, the linear regression was used to investigate the relationship between the change in anhedonia and the anhedonia phenotypes with percent change in PBMC mTOR expression. We also did a secondary analysis to investigate the relationship between non-anhedonic features of depression and change in mTOR. So the non-anhedonic features that we investigated include sadness in and of itself, negative thoughts, this included pessimistic thoughts as well as suicidal thoughts, and neurovegetative symptoms. This included things like inattention, sleep, and appetite. For the demographics of the Emory cohort, again 44 patients here, median age was 52.5 years. A little over half of the sample was women. Mean total PHQ-9 score was 17.5. The median anhedonia score was quite high at 3. 63% of the overall subjects had some sort of reduction in anhedonia scores. And there was a 39% reduction in those scores when utilizing the single item on the PHQ-9 And that median difference was a reduction of one point on the PHQ-9 scale. That was statistically significant from baseline to endpoint after ketamine was given. For the Mayo cohort, smaller sample of 12 patients, mTOR data was only available for 10 on the biomarker objective. So our small sample got smaller, which is unfortunate. The median age was a little bit younger than Emory, 47.9 years. Unfortunately, 100% of our sample was women. So it's hard to extrapolate some of the results to men. The median total Moderisk score here was 28. And the median anhedonia score was 3. In the Mayo cohort, the median change in anhedonia score using the Moderisk Just Question Number 8 in and of itself was a reduction of one and a half points, showing a significant improvement in anhedonia from baseline to endpoint after ketamine was given. Now for our clinical objective, the results showed a correlation between percent change in anhedonia, core depression and interest in activity scores, and modified total depression scores was high in both cohorts. So the Pearson correlation coefficient here on the slide ranged anywhere from .66 to .84. You can see the Mayo Clinic cohort here on the top, Emory was here on the bottom. This is just looking at that single item on the depression rating scale. This is when sadness was added in with the anhedonia, and this is when energy and concentration were added in with the anhedonia. For the biomarker objective, there was a significant correlation between change in anhedonia, looking at just that single item, question number 8 on the Moderisk, and an increase in mTOR protein expression in the PBMCs. So the greater the reduction in anhedonia, that correlated with a greater change in mTOR protein expression. The p-value here is .026. The correlation between the change in the core depression and the interest in activity phenotype scores and the change in mTOR protein expression did not reach statistical significance. So when we added in that sadness factor, the p-value here became .08. When we added in the concentration and energy components, that p-value is .07. So missing that cutoff by just a bit. Interestingly, there was no significant correlations between non-anhedonic depression phenotypes and change in mTOR. So again, when we're looking at just sadness, or when we're looking at those pessimistic thoughts and those suicidal thoughts, or the inner tension, sleep and appetite, we're not seeing these factors correlate with change in mTOR. So there's something about that anhedonia factor and change in mTOR that we were finding. So in terms of our conclusions with regards to the clinical objective, I think it's fair to say that we replicated prior findings of ketamines anti-anhedonic effects in a heterogeneous group of patients with treatment-resistant depression at two different sites. And these anti-anhedonic effects are independent of general depressive symptoms. We also showed an in vivo increase in peripheral immune cell mTOR protein expression correlating with the anti-anhedonic effect of acute ketamine response. So ketamine-induced mTOR engagement within the peripheral immune cells may be a key moderator to anti-anhedonic effects and could maybe serve as a predictive biomarker for those anti-anhedonic effects. But really, we do need this to be replicated in a larger sample. We have collected that larger sample through the National Networks of Depression Centers. We have a sample of 74 that we've collected, but we're currently in the analysis phase and Dr. Tai is doing the assays for these biomarkers to see if we can replicate this finding or not. And I do want to acknowledge our team. Again, Dr. Tai is the brains behind the biomarkers at the University of Queensland in affiliation with Mayo Clinic. I want to thank Patricia Riva-Pace, who was able to gather that larger set of subjects for the clinical objective. And then, of course, my colleague, Dr. Balwinder Singh. He spearheads a lot of the ketamine research in our clinic. Vanessa Pasernik is our statistician. And of course, Dr. Mark Frye, who's been a mentor to all of us over the years. And I want to thank you all for your time this afternoon. Thank you. Thank you. So, good afternoon, everyone. First of all, I want to thank you guys, our co-chairs for organizing this panel so we can have the possibility of discussing the use of ketamine biomarkers for ketamine response or not and how we can get better for the future. And just to start, no disclosures. And we know that ketamine works, but we still have patients that don't respond and we still have patients that getting even worse. So we do have a need to identify biomarkers, blood-based biomarkers, neuroimaging-based biomarkers, and my best guess will be a panel of biomarkers where we can put all of these different factors together and use machine learning and very sophisticated analysis, which are gonna help us to identify these patients. And we do have this strong need to not only identify how the patient's gonna respond to treatment or not, but we need to identify the basis of the disease. Why is the patient do not respond to previous treatment? That is gonna guide us to better understand why that patient could respond to ketamine or not. What is happening at the cell at the beginning, way back when we start before thinking about treatment, that is gonna help us to stratify the patients. And when we are able to stratify patients in subgroups, this group has more like low-grade inflammation who are gonna respond better to these options of treatment. These patients, they have a mitochondrial dysfunction baseline which will respond better to this treatment. And when we have that defined, we're gonna be able to have better clinical trials to identify response and non-response because we are stratifying the patients in a better way. This is just to give a little bit more regarding to the mechanism of action of ketamine. And we do have the main mechanism of action is inhibition of the NMDA receptors in GABA-ERG interneurons. But nowadays we have the unified model of ketamine action that includes not only the inhibition of the NMDA in GABA-ERG interneurons, but also in glutamatergic, sorry, paramedial neurons. And when we have these two working together, we're gonna increase glutamate and all of that gonna lead to the change that Jennifer just talked to us. But the main thing is all these pathways that are going downstream to the increase of glutamate levels, they are gonna help us to differentiate what is the fast-acting antidepressant effect and what is the long-term antidepressant effect. When we're talking to fast-act, we are talking more to BDNF acting, increasing protein in a fast-acting kind of looking for the homeostasis, the synapses homeostasis. When we're talking about long-term effects, we are going way downstream that is that mechanism. We are talking about phosphorylating MECP2, which that are gonna increase synaptogenesis and that are gonna lead to a more long-term antidepressant effects. So what do we need to think about when we are talking about biomarkers for identify if the patient is gonna respond or not, we need to start thinking not just only at the ones that are at the beginning as Gustavo presented to us like BDNF, Emitor, but we need to start thinking about the downstream effects that that drug is gonna change in our cells. And that is where we can start talking about metabolomics and metabolism. If you think in a very simple way, everything that you're gonna give to our brain is gonna need what to work? ATP. If the cells doesn't have ATP, if the cells doesn't have energy, the ketamine that we just gave to that patient we're not gonna have any effect. So bear with me, if you have a patient that has a mitochondrial baseline disorder, a disease which is very common, most of our patients, they could have mitochondrial diseases which are genetic diseases. 7% of these patients, they just are diagnosed with mitochondrial disease like 10 years after they show psychiatric symptoms. So would that patient respond to ketamine in the same way that other patients? No, because they have an inability to produce ATP in a normal way. So when you put all this data together, and trying to put this, look to this big picture, create a puzzle and put the puzzle together, we need to remember that we are not talking about one drug that has one specific effect. That drug changes a very large kind of like cell signaling pathway and has several downstream mechanisms. At the beginning of this process, what could happen is, and studies have shown that is, ketamine act as an anabolic factor, is gonna try to use the energy that the cell, the body has to provide all the continuous that that cell ask. For example, it start burning fatty acids, it start breaking proteins to have energy to continue the BDNF signaling, the synapses and all of that. When that is done, and we get to the long term, long antidepressant effect, the cells start to get a feedback the cells start regulating and start catabolism. And it starts like producing, regulating the baseline, start to change the beta-oxidation that was, as we can see, some acylcarnitines, they are more decreased, the short ones are decreased, but 24 hours they are increased. That is all adjustments because of the energy demand of the cell, and how the cell we're gonna respond to that ketamine shock that we can kind of say. And most of these alterations, we do see in MDD patients, in TRD patients. Some of the patients, they do have mitochondrial dysfunction, we have a strong evidence of that. We have evidence of inflammation. Not all the trials, they gave us the same results, but we do have some data there. We have data showing that patients with an increase in the median chain acylcarnitines, they have a high risk for depression. And we have like reduction of BDNF. So what do we do have regarding biomarkers? What are the problems? Lot of studies, lot of data, but they are mixed results. We cannot reproduce them, and basically we get to the point that we cannot use them. And we do have, Gustavo showed us, that BDNF has a small effect size, could be a beginning, but what we need to do is work better on how we start our clinical trials. We need to define our inclusion criteria based on stratifying patients in subgroups. How is this patient, how is the, it's the turning off. How is the patient can be stratified? The other thing that we need to kind of get more organized is basically creating a standard procedure where all the research groups we're gonna follow, so we can put that data together and get a large sample size. So we also need to think about depression scales. What is the depression scales that I'm using? I'm using Madras, I'm using PHQ-9. All of that give us different responses. I'm using just a factor of the, a dimension of the scale, like a Hedonia. That is gonna give us different responses. Time points, 40 minutes. We have studies with 40 minutes, 100 minutes, 24 minutes. We have 13 minutes, 60 minutes, 24, 48 minutes. How are we gonna standardize all of that and put together and try to get one biomarker? Very hard. Other things that we need to pay attention is lifestyle. Most of the things that we see here is, and if you're talking about metabolomics, that is metabolism, is if you exercise, you're gonna change your metabolism. If you're fasting more than eight hours, that are gonna change your metabolism. If you're fasting only six hours, that are gonna change your metabolism. You need to have a very strict protocol so you can put the data together. Medical comorbidities, most of our patients, they have cardiovascular conditions. That is gonna change all the metabolomic panel because that changes all the energy process. The medications that these patients are taking, they change the process. They change how the cell we're gonna produce and use the substrates that the cell has. Even OCT drugs that these patients are taking. Very like, not harmful, it's just like some supplements. Magnesium, is that a problem? Yes, it is. Magnesium is a cofactor of several enzymes in our cells that could change the way that the ketamine were gonna be metabolized and the result that the patient are gonna show to us and to our trial. The other thing is small sample size, big issue. We need to increase the sample size. We need to do longitudinal studies and not just evaluate the response and non-response, but also the efficacy and how secure is to use ketamine in like four years, five years. We still don't have that data. And we do need like lack of placebo group. We, most of the studies are just pre and post. We don't have groups, placebo groups, or we don't have like, we are investigating one specific marker, but we don't know if that marker is changing in, that that marker is associated with the disease per se, with disorder per se. So we need to know before, is this, so we need to know before, is this marker different between TRD patients and MEDD patients and health controls? All back steps that we need to go and fill out. And the only way to do that is further large trials with more comprehensive considerations of confounders. As much, as more, you can patronize your clinical trial. What is gonna be your inclusion criteria? What is gonna be the information that you're gonna get to that patient? It needs to be a very deep interview. What is the diet of that patient? What is the activity of that patient? What is the drugs that patient's taking? Is this patient doing therapy? If not, all of that we're gonna change the final result of our trial. And to be able to put all of that together and get one biomarker, I would say we have a long, long, long way to go. But if we keep doing what we're doing here and discussing and trying to align all the thoughts, I think we can get there. So thank you, and I wanna also thank my group at UT Health, NIH, the Dunn Foundation, and the Linda Gale Research Fund, which is fund my research. Thank you so much. All right, we're up for questions for any of the speakers. Hi, I'm Dr. Dennis Henkel from Philadelphia. Very nice to talk for all of you. You mentioned right towards the end very briefly just about magnesium with ketamine. At my institution, we infuse magnesium with ketamine for headache treatments. So I was wondering if you think that's a good way to do it, or if you think it's not a good way to do it. I mean, I think it's a good way to do it. I mean, I think it's a good way to do it. So I was wondering if you think an infusion of magnesium would make any difference with ketamine response? Yeah, I suppose to tell you, yes. So I think the question is. Yeah, it's about like the supplements. I didn't question about magnesium plus ketamine for headaches because I do have a lot of headaches. I didn't, actually, I didn't read anything about ketamine for headaches and associated with magnesium. That was just one of the examples that nowadays a lot of people are start taking magnesium supplement. But thinking mechanistic way, it does, it could make sense because magnesium, it is a cofactor for different enzymes that we're gonna help with the metabolism, not only at the periphery, but also with the metabolism of the brain. And it's gonna help some neurotransmissions and all of that, but specifically, I don't know how to answer that. You've gone to a lot of trouble to tell us about how ketamine works. We have now a couple of new drugs, for us psychiatrists to use. Let me ask a stupid question. But does the ketamine work in any way similarly to the albedity, the new dextromethorphan and Welbitrin? Maybe you can comment on those mechanisms of action for people that are kind of curious about these new drugs. So I can take that one. All right, thank you. So the question is, does ketamine work similar to ovality? So the newer drug, ovality is a combination of dextromethorphan and bupropion. So theoretically, if we believe that ketamine only works by NMDA antagonistic action, then the answer is yes. Because dextromethorphan, the primary mechanism is the NMDA antagonist. And the rationale behind adding bupropion is when you take or ingest dextromethorphan, it breaks down pretty quickly. So the role of bupropion is to just delay these. So it's a 2D6 antagonist, sorry, inhibitor. So it just slows down the breakage so the dextromethorphan stays longer. So theoretically, mechanistically, yes, it is similar. And we saw, when we look at the studies, a pretty quick response within first two weeks. 14 days we saw improvement in depression symptoms. But the question with ketamine, is it only the NMDA antagonistic action which is leading to antidepressant effects? There are other pathways which have been involved with the opiate receptor pathways, the GABAergic acetylcholine. So maybe yes, but I don't think we know the exact answer if that's the only pathway involved. But very similar, yes. Great talk, and again, good to see you after SOBP again. Thank you. Towards the metabolites, the metabolomics that you were doing, I would say our group, we've been looking at the metabolites of ketamine, hydroxynurketamine, and we recently looked at CSF levels of the metabolites. And the HNK, I mean, that's what Carlos is very passionate about. And it seems like it lasts longer, and it transfers over. So we're kind of wondering if it's, ketamine's always been like a pro-drug, and maybe it's the other pathways, the AMPA receptor, maybe that's a little bit more involved. And I was going to ask you, do you guys also, with the metabolomics, also look at, I guess, the metabolites of ketamine in your studies? Is that something that you guys have been looking at? Yeah, it's a great question. So we did look at the S-ketamine, and R-ketamine, and N-ketamine. Again, I didn't present all of that data because it's very small sample size. We got nine patients, and we did almost more than, I don't know, 600 analysis. So all those P values, those fancy P values of less than 0.05, they may not hold up any of the bond for any correction. We did see some signal with the N-ketamine, more with the R-N-ketamine as compared to the S. From a mechanism standpoint, I think that the dihydroxy norketamine would have a slightly longer mechanism, or sorry, longer duration of action. How that would impact the metabolomics, I don't think I know the answer for that. I think the NIMH, they're doing some trials there, so hopefully we'll know more. Yeah, Rune is doing the analysis, so maybe next month or two, we'll figure it out. And towards the Nudexa question, why is this on Nudexa? We did a Nudexa case series after our ketamine patients, and we didn't find that ketamine response really predicts outcomes with that either, so just to the dextromethorphan question. Yeah, no, thank you. I mean, the CSF data is fascinating because what we're all checking in the blood is, you know, we don't know how accurate that is, but the CSF, looking at those metabolites, that's a different deal, yeah. That's the same challenge with the cytokine analysis. Correct, yeah, yeah, yeah. So I think that's where the standardizing, at least the fasting level, and maybe the drugs, so when we do future trials, that can at least help reduce some of the confounding covariates there, yeah. Thank you. Thank you. I'm Sunil Katragada from Atlanta, Georgia. I'm a psychiatrist and use sketamine and ketamine in my daily practice in a hospital-based and outpatient-based practice. I've been using for four years or so. As you know, even for ketamine, even though responses are good, the response is not the same, and it highly depends on the patient population and selection and the type of response, even the degree of response is not the same. We're still trying to kind of figure out the clinical biomarkers, you know, who responds and who doesn't respond. I don't think we were able to standardize, in my knowledge, and who are a typical patient, 100% sure, I mean, we have hunches about, oh, this patient has chronic and this has no response to ECT, this has fewer trials, this has no psychosis, or this patient has more melancholy, so we all have a lot of ideas in figuring out clinical biomarkers to kind of predict the response. I was wondering, in your knowledge, was there any correlation between clinical biomarkers, or we have knowledge about, with any of the blood-based or nootropic or metabolic biomarkers? If we can reach to one biomarker, it was wondering if we can reach a system where clinical biomarkers tie to blood-based or nootropic biomarkers where we can predict better. Yeah, so, great question. Is there a clinical biomarker which correlates with ketamine response and which correlates with the response or with any of the biomarkers? So, the short answer is no, but when we look at, you know, I think what Dr. Vandewoerde presented from her talk is if we break down depression into these phenotypes, so maybe that anhedonia, we saw a signal with anhedonia and mTOR, it needs to be replicated, we are seeing, and these are small sample size studies, these need replication, but so far, there's no clear sort of a clinical indicator. As I said, you know, depression itself is a very heterogeneous condition. There are, you know, different types of depression, but we haven't seen a clear indicator. There's some initial data with the family history of alcohol use disorder, and there was a higher response in ketamine. We don't know the reason why, but there are, you know, some of the initial studies looked at the BMI, higher BMI correlated with the response, but then it was replicated in a couple of studies, not in others. We are doing some additional analysis, so there's a poster, I think it's on Monday, looking at sleep phenotypes, and we saw a signal with hypersomnia. Patients who have hypersomnia at baseline, they have a higher response to ketamine as compared to patients who do not have hypersomnia, and then we add the hyperphagia piece to it. We are seeing maybe atypical depression. That's a subgroup where we're seeing a slightly higher response. Again, it's not, you know, I'm very cautious when we look at the observational study or small sample size studies without kind of accounting for some of those variables. So, so far, at least to my knowledge, there's not a consistent finding with any particular covariate by which we can predict who is a good responder to ketamine. There are some data, you know, with at least what we see in a clinical practice with, you know, if there's a comorbid personality, sort of that is driving the depression. We see those patients may not do as well, maybe in the acute phase, and early on, they may respond better, but as they receive multiple treatments, we see the response start to fade away. So I hope that was helpful. Hello, thank you for the great talk. My name is Juliana, and I'm applying for residency this year, so my question is probably gonna be like more on the basic side. There is indication that we do have a mitochondrial dysfunction in major depressive disorder and TRD. And it seems like the ketamine bypasses this mitochondrial dysfunction, because we also seen like increase in lactate and lactate to pyruvate ratios. So my question is, do you have any idea on how the ketamine bypasses this issue with the mitochondria and just stimulates the mitochondria to produce energy again? I'll let you take that one. It's working, yeah. So my hypothesis is that not all the patients are gonna have a mitochondrial dysfunction that will be so impaired that we're not gonna be able to produce ATP. The data that we are talking about TRD, our TRD patients, and I know because our data from our group, they are TRD patients that they are like the ones that do not respond to anything. So they are like a very depressed patients with, they are starting the DBS trial, which means they tried everything, even ketamine, some of them. And in that case, yes, the mitochondria, looks like the ketamine could bypass the mitochondrial dysfunction. And some of them even didn't respond to DBS. We do have some data, but it's very previous, but it's like we have patients, and we have data showing that patients with, with TRD patients with higher lactate to pyruvate levels, which is a biomarker, is a marker for mitochondrial dysfunction using, using mitochondrial disease clinics. These patients with higher lactate to pyruvate at the baseline, they didn't respond to DBS one year after. So it seems like if you don't have, if the mitochondria is not working properly, it doesn't matter what you're gonna give to that patient, the response will be very low. We need to maybe try to improve or add something that are gonna help the mitochondria to get better. Ketamine has effect on the mitochondria, not just like, not stimulating the mitochondria to produce ATP, but ketamine can, through Amytor, can increase biogenesis, mitochondrial biogenesis. So ketamine will increase the number of mitochondria that the cell has, and that will gonna be a possibility to increase the energy that is being produced, and they can sustain all the new, the synaptogenesis, synapses, and maintain this long antidepressant effect. But again, my main point was, we need to know what is the baseline of the patient. Otherwise, maybe the treatment not gonna work. Thank you so much. You're welcome. Hi, a quick question. Thank you for your talk. I'm from Mexico. I've been doing ketamine for about seven years, and I almost have 200 patients right now, and I have not seen clear results about the clinical predictor of dissociation with our patients, but not only dissociation, but what do we do with that dissociation? I mean, there are a few papers that says that if we do psychological therapy during dissociation, maybe the result will be different. So what are your thoughts about dissociation as a clinical predictor with ketamine? But I think with esketamine, the evidence says that there's no difference, but with ketamine, I don't know. So what are your thoughts about that? Thank you. I can take it. So great question. Does dissociation predict response to ketamine? So I think some of the earlier studies of the earlier data, so that's the problem. When we start for the smaller sample size study, dissociation showed that higher dissociation can predict response. But as more and more data start to come out, we do not see dissociation as a predictor. And the question is, what do we do with dissociation during the infusion? We try to manage the best we can. And the problem is there are some centers where they will keep pushing the dose of ketamine till people get or start to have prominent dissociation based on that earlier data that you need to have in a pretty good amount of dissociation to have a response. So dissociation by itself is not a predictor. So stick with what we do in our, at least in our ketamine clinic, we stick with the research dose, which is half a milligram per kg body weight, regardless if we have dissociation or not. Dissociation can be problematic in some cases, especially if someone has severe PTSD because dissociation can, next thing you know, there's a flashback coming. So we have to be cautious with that, but we don't look for dissociation as a predictor. If someone has it, we try to manage it the best we can. Thank you. Hi, I'm from Mercer Island. I'm Tina Dimopoulos and Dr. Dunner, Dave Dunner and I have a Spravato Clinic. I just find it's really fascinating speaking about the PTSD piece. So I've been tracking people with PCL-5s relative to their ketamine treatment, just to look at kind of response and sometimes dissociation absolutely can trigger people. But I wonder if, in terms of modality, anyone's doing something on the line of psilocybin where people process post-ketamine to facilitate response in terms of the more complex, so we deal end of the algorithm, so really complex people like comorbidity of PTSD, depression or bipolar to depression. That's one thing. The other is, is anyone also looking inter-clinically, and I see this too with a lot of PTSD folks, that they'll also have a lot of comorbid, autoimmune and inflammatory diseases. And looking at CRP as a marker in some of these folks post-ketamine or eskimo. So, I think in terms of CRP, just clinically, I think that's... It's not very specific. Right, it's not specific, clinically, anecdotally, I would say that those patients do less well in our clinic than patients with less comorbidities, but we, in our clinic, I would say we treat primary treatment-resistant depression. But yes, a lot of our patients come with comorbid anxiety or PTSD or those sorts of things. There's some literature out there saying anxious depression may respond a little bit better, but I'll be honest, in our clinic, when chronic pain is such a big factor, we probably don't think that they're necessarily the best candidate. We just haven't had a great... I would say we don't have a lot of good track records with a lot of the chronic pain patients with treatment-resistant depression. So we're very thoughtful about which one comes first in terms of our clinic. Yeah, I mean, even on the pain piece, like other associated autoimmune illnesses too. You ever see in any... Like if people have a host of autoimmune skin disorders or autoimmune, other things, no? No, I don't think we have seen necessarily that autoimmune disorder, that's a predictor or to response per se. I think you raised a question regarding the psilocybin in psychotherapy and ketamine in psychotherapy. So it's a very tricky area because there's no randomized controlled trial especially with ketamine with assisted psychotherapy. There are some open-label studies suggesting that yes, maybe the ketamine-assisted psychotherapy can be helpful to the question is what time before post-infusion, during infusion. So it's a... Manualized. Yeah, I don't think there's a clear cut data at least from a randomized controlled trial. This is rest, once we have the case series or open-label data, that's usually... There's a high risk of bias there. At least for ketamine, there is very minimal data. There's no randomized controlled trial showing ketamine-assisted psychotherapy as compared to just ketamine. Anybody use like acetylcarnitine in conjunction with ketamine or lithium? So again, great point. So theoretically makes sense. With lithium at the mTOR activation, we see it can have synergistic effect. There was one trial where they looked at lithium. It was a negative trial from James Morrow's group, but there were question whether the timing of lithium induction or lithium use with ketamine could have been different. The acetylcarnitine piece, I'm not aware of any randomized controlled trial where they use acetylcarnitine, which is C2 along with ketamine. Theoretically makes sense. There's data suggesting treatment-resistant depression have a deficit or lower level of C2 and you replace with L-acetylcarnitine that that would help, but there's no randomized controlled trial kind of showing that effect. Thank you. Thank you. All right, our time is up, but feel free to ask. We'll be here if you have any questions. Thank you so much for all of your attention. Thanks.

anhedonia

depression

conventional antidepressants

ketamine

treatment-resistant depression

mTOR pathway

synaptic plasticity

dysfunctional brain circuits

immune cells

ketamine infusions

anti-anhedonic effects

immune system

mechanisms of action