Morning everyone. Can you guys hear me okay at the back? Perfect. Thank you. Well, morning everyone. Thank you everyone for coming. The topic of the symposium today is Emerging Potential Biomarkers to Inform Bipolar Clinical Practice. My name is Dr. Balwinder Singh. I'm a mood psychiatrist at Mayo Clinic in Rochester, Minnesota. I'm an assistant professor and I'll be chairing this session in the presence of three very esteemed bipolar disorder clinicians. Our first speaker is Dr. Michael Gitlin. He's a distinguished professor of clinical psychiatry at Geffen School of Medicine at UCLA where he directs the adult division of psychiatry and has directed the mood disorder clinics for over 40 years. Good morning. I'm not sure what the attraction is to that side of the room but hopefully we'll balance it out. So good morning. I'm gonna start off by talking about lithium and it'll be a combination of topics in our 15 to 20 minutes. Brief FDA indications, then clinical indications, some pearls, and then some issues about genetic markers, such as they are, and we'll finish up with some new uses for lithium, kind of looking towards the future. So first let's remember, let's go over, oh, for the FDA indications this is old stuff, nothing new here. Monotherapy for bipolar 1 disorder and the treatment of course of acute mania. The real issue is how do we best use lithium now? And the convention of a 12-hour level is still what you do, not that that's the right parameter to use. There are actually no data showing that 12-hour levels correlate better with efficacy than, let's say, 4-hour levels or trough levels, 24-hour levels, but it's the convention. Every study uses it. You should too. Oh, it seems to move by itself. There's this question. We know that if you change your lithium regimen in your patients from once daily to BID, TID, or vice versa, the level will actually change based on differences in lithium, renal excretion of lithium. So the question is, how do you do that since there's good reason to think that patients should be on once daily lithium? It's probably better for your kidneys. And so what should you do? Because all the studies are with BID and TID regimens. And the answer is the ISBD, International Society for Bipolar Disorder, this keeps moving by itself, says since we don't have any good data to help us with that, ignore it. Whether your patient is taking once daily lithium or BID, just assume it's a 12-hour level and treat that number the way it is, and you want that number now to be between 0.6 and 0.8. In the old days, we used to think that we should use higher levels. There's more and more data, mostly coming out of Europe, that lower levels are A, just as effective, and B, better for your patient's kidneys. You should be nice to your patient's kidneys. We know that for older folks, even given the fact that they excrete lithium less well, that they not only tolerate lithium less well, but they will respond to lower levels. So for older folks, levels of 0.5, even 0.4, may be sufficient. And again, we don't have any data suggesting that any one indication requires a different level than another. I want to do a brief review on lithium and renal effects, since that's been a hot topic with ongoing data that, of course, given that this is psychiatry, disagrees with each other. So let's remember that the kidney is composed of two parts for our purposes. I'm sure nephrologists would see it as more complex than that, and that's the tubular effects, which has to do with salt and water balance, and the glomerular effects, which is filtration. We know that lithium affects the tubular system far more in a negative way than the glomerular system. Initially, it's functional, kind of like when you drink alcohol and you urinate more because it interferes with ADH. So lithium does the same. The difference is, over time, that functional change turns into structural change, and you start seeing destructions of the tubulars and the interstitium, and you end up with what's called interstitial nephritis and diabetes insipidus, which is more than 3,000 milliliters a day of urine. The glomerular effects from lithium are less consistent. Certainly, you do see a decrease in EGFR, creatinine clearance, but it's not necessarily as common. I'm going to show you a little more data about that, too. The question, and we don't have time to go over the details, if you have what Dr. Jefferson used to call creeping creatinine, you're following your patient's creatinine and it goes up slowly, slowly, slowly over years, decades, when should you lose your nerve and stop the lithium? The field is unclear. There's one study recently that showed that if the EGFR goes below 32, there's almost like an irreversible decline in that point in renal function. So if you're going to stop the lithium, you should probably do it in the EGFR range in the 40s or maybe the 50s. 60 or above is fine, and we'll get back to that, too. So the definition in the renal world of CKD, chronic kidney disease 3, is an EGFR less than 60. If it's over 60, don't worry about it. If it's below 60, you go from screening to monitoring, meaning you start getting EGFRs and serum. Creatinine is a little more frequently, maybe like every three to six months instead of once a year, and end-stage renal disease is less than 15 cc's per minute. It does correlate with how long you've been, your patient has been on lithium. Comorbid disorders, hypertension, diabetes, clearly help do that. Oops, it went away again. And I think there's enough good evidence to say that yes, lithium is associated with renal damage. If you look at bullet 3, the rate of CKD 3, the percentage of people is 30% higher than a non-lithium treated population, but this is a slow effect. The mean time to get down there is over 20 years. So you don't need to be worried in the first year or two, and if you're not sure what to do, you don't need to stop the lithium suddenly. This is really slow stuff. And again, whether you compare it to a controlled population or those patients treated with other mood stabilizers, you're fine. And this is one of the more recent studies that looked at it. Look towards the right where you see the part that says greater than 10 years lithium exposure, that's the group where you see a greater acceleration in EGFR. If you want to see what it looks like graphically, it's that. Look to the right part of the slide, and that's where that's the key number. The people who have been on lithium for more than 10 years have an acceleration of the decrease in their EGFR, which by the way goes down for all of us as we get older. By the time we are 60, a third of our glomeruli have gone to glomeruli heaven, I guess, glomerulus heaven. So we have to compare it to that. The Scandinavians have far and away the best data on this. They have their whole country computerized in Los Angeles. We can't do anything for more than three blocks, but in Sweden they seem to be able to do this remarkably. And again, they show that there is indeed risk for that. This is not the majority of people on lithium. This is a relatively small percentage, but it's true. So take the renal effects seriously. Okay, so what are the predictors since that's our focus for today? The predictors of lithium response and maintenance treatment, and here they are. This is the best meta-analytic review on this. It's the MDI sequence, meaning if your patient has a mood episode, it's mania followed by depression followed by the well interval. The less common pattern is the DMI pattern, depression followed by mania. So the MDI pattern predicts a better response to lithium. Rapid cycling predicts a poorer response. More severe illness, like more psychosis, predicts a more severe illness. Family history of lithium response indeed is an inherited phenomenon. And again, the less severely ill you are, and the more you meet the classic criteria to discreet episodes of mania followed by depression with euthymic intervals, with grandiose euphoric manias, that cartoon picture predicts the best response to lithium. And this is a study from Germany, now 13 years old, looking at atypical features, schizoaffective disorder, rapid cycling, mixed features within episodes, etc. And as you can see, the more mixed features you have, the more atypical features you have, the less well you do on lithium. So the cartoon picture predicts the good response to lithium. Everything else predicts a less good response. All right, so now let's go over the genetic predictors of lithium response. And let me give you the end of the story, which is that it's not ready for prime time. And part of the problem we have is there is still unbelievably little useful knowledge as to how is it that lithium works. And until we know more about that, it's really kind of difficult to find the right candidate genes. So the first reasonable study was that of Roy Perlis from whatever that is, 14 years ago. He found a few SNPs, a few places where it looked as if there was some link, and that was great as a first order. This just shows you the same thing. I will move on. And then there was there are some other studies looking at telomere length. And some of you may know that telomeres have to do with the size of areas at the end of chromosomes, and it's a good thing. And so you want lengthening telomeres, shortening telomeres are correlated with cell death or cell destruction and death. And if you look at the second bullet where it says lithium in telomere length, you can see there's increased length in bipolar patients treated with lithium compared to healthy controls. And when I show you a little bit of the data on lithium as being neuroprotective, meaning it may actually decrease the evolution to dementia, maybe this is why. Unclear. The telomere data and the lithium preventing dementia are two separate literatures. Whether one is the mechanism by which the other happens is an intriguing idea, still unclear. And the most recent study is called the Conley-Jen study, sponsored by the international group, the Ixley group. And again, here too, there are some SNPs that are found to be relevant to lithium's efficacy, but it's not good enough to be able to use in clinical practice yet. And there was one other. There it is. So this is the most recent study that came out six, eight months ago. And again, they were looking at whether a polygenic risk score could correlate with lithium efficacy. And it turns out the biological data did not enhance the clinical predictors that we've talked about. So the question of, do we have biomarkers for lithium response? There is a lot of active work going on, like this study here. There'll be more going on, but for right now, this is absolutely not ready for prime time. Now I want to talk to you a little bit about the future, which is always amusing with lithium, which is just about our oldest drug. I don't think any of us prescribe barbiturates anymore, and chloral hydrate has fallen out of favor. So lithium may be now our oldest drug. And then there's basic science data and animal data suggesting that lithium is neuroprotective. Now this is ironic, since your patients all tell you, and my patients all tell me, that lithium is correlated with cognitive impairment, right? Our patients all say to us, at higher doses and blood levels, it's exquisitely blood level related. I'm not as sharp. I don't think as well. Monskow, who did the first studies on lithium in Denmark in the 50s and 60s, found it then. He actually took lithium himself to see what it was like, and he describes perfectly that a level of 1.0, basically he felt like a bozo in faculty meetings, and at 0.5 he felt fine. Not clear how academics can tell the difference, but all right. Anyway, so we've known this for a long time, so it seems rather paradoxical that a medicine that causes acute cognitive impairment could be neuroprotective. But I think that's exactly what's going on with lithium. And there are two relevant studies. The first is the Chen study, which is eight, nine months old, that showed that lithium lowers the risk of development of dementia of the Alzheimer's type, DAT, and vascular dementia. So people on lithium were less likely to progress from MCI to dementia if they were on lithium. That's an extraordinary finding. Now, by the way, one of the other things we're going to have to figure out as a field is what blood level of lithium might we need for this efficacy? Some of you may know that there's, and I don't have slides on it today, there are data on lithium levels in the drinking water, which are really non-detectable, but even teeny, teeny, teeny little lithium levels, that more lithium in your drinking water correlates with less violence in the population that drinks that water. So do we really need a lithium level of 0.6 for this neuroprotective effect, or could you use teeny little amounts of lithium? That to me is incredibly exciting because we can only, if we use lower doses, that means we don't have to worry about renal effects, thyroid effects, parathyroid effects. That would be fabulous. And again, this is from a second study from a few years ago showing that people who took lithium had a lower rate of an evolution to cognitive impairment. So I think that's exciting. The last thing I wanted to show you is a recent study that was, that flies in the face of what we typically think. So for years we've, one of the other issues about lithium is that it would prevent suicidality. There are multiple studies, few of them are random assignment, you know, classic RCTs, but a lot of other kinds of data. And there were a few studies showing that lithium did indeed prevent suicide, excuse me, even in non-bipolars. This is the largest study and it's a resoundingly negative study. Let me first show you the next slide. You take, this is a collaborative VA study, you take patients who, mood patients, most of whom are not bipolar, 85% major depression, and they're at high risk for suicidal events because they've already had one, and you add it as treatment, to treatment as usual versus treatment as usual. And what you can see is there's no difference. In fact, there was so no difference that the the monitoring board stopped the study in midstream because there was no evidence that lithium was going to help. However, how do we reconcile that with all these prior studies showing that lithium does prevent it? And the answer is in the sub-bullets you see here. So the goal was to get, in this study, was to get people to lithium levels of 0.6 to 0.8, just what everybody else does. Right. But in fact, the actual mean lithium level was not that. And it's not clear whether patients balked at taking higher doses because of side effects, or the doctors in the study were wary about raising the doses because of whatever factors in their own thinking. But the fact is, the mean level in this study may have just been too low. So it's not clear whether this is a negative study or really a failed study based on the fact that patients didn't get to the right blood level to show the anti-suicide efficacy. So be careful over over-interpreting this study. So with that, I know we have numbers of things to discuss today. So a reminder that even though lithium is an old drug, there are new thoughts about it. So with the suicide data, especially the neuroprotective dementia early data, I think is just terrific. Renal problems are real, but you can manage them, and you shouldn't be in a panic about it if the creatinine goes up a little bit and the EGFR goes down a little bit. And again, genetic markers, there's going to be a lot more work on it. One of these days, I suspect, this will help us clinically, but not today and not tomorrow. Okay, with that, I will stop. And Dr. Singh, thank you. I forgot to mention, we'll take questions at the end of the session. All right. So I'll be talking about lamotrigine. These are my disclosures. All right. So indications of lamotrigine. So lamotrigine is FDA-approved in adults for the maintenance therapy for bipolar 1 disorder. And the way maintenance therapy is defined is to delay the time to occurrence of mood episodes in patients treated for an acute mood episode with standard therapy. The lamotrigine is also FDA-approved as an adjunctive therapy for generalized seizures of Lennox-Just-Start syndrome and partial seizures and tonic-clonic epileptic seizures. I won't go too much into details regarding those indications. All right. So this slide is presenting the data on the basis of the lamotrigine was FDA-approved. There were two randomized controlled trials. And in this paper published by Dr. Goodwin et al. in 2004, they combined the data from both the studies. So they looked at patients with bipolar 1 disorder who are either in a manic or depressed phase. And then in the open label phase, they are over a time period of 8 to 16 weeks. They started stable patients on lamotrigine. And patients who responded to lamotrigine, then they went in the double-blind phase. And they monitored those patients for 76 months or sorry 76 weeks or 18 months. Patients who did not respond to lamotrigine, they did not go into the double-blind phase. In the double-blind phase, patients either received lamotrigine 100 to 400 milligram, lithium where the target level was 0.8 to 1.1 or placebo. Now this slide is looking at the time to intervention for a depressive episode. So there are two primary outcome looking at time to a depressive episode or time to a manic episode. So when we look at the data comparing lamotrigine to placebo, it was significantly superior in dealing the time to a depressive episode. Lithium was not superior to placebo, and lamotrigine was not superior to lithium when we look at time for a depressive episode. Now, when we look at time to intervention for a manic or a hypomanic episode, lamotrigine was superior to placebo, lithium was superior to placebo, and lamotrigine versus lithium, when we look at that, we saw a bit more with lithium, they had a lower risk of relapse. On the x-axis, you're looking at time in months, on the y-axis, you're looking at percentage of patients who are intervention-free. This was another study looking at lamotrigine as an adjunct treatment to lithium. Here we had about 124 bipolar-depressed patients receiving lithium who were randomized to lamotrigine or placebo, and after eight weeks, paroxetine with an SSRI was added to non-responders for another eight weeks, and then the responders, they continued the medication and were followed up for up to 68 weeks, or until a relapse or recurrence of a depressive or a manic episode. Now, after eight weeks, what they saw is that the addition of lamotrigine was significantly more efficacious as compared to the placebo group, while after addition of paroxetine in non-responder, both groups, they saw similar improvement. But the important thing was with the lamotrigine, the median duration or the time to relapse or recurrence was much longer for the lamotrigine group as compared to the placebo group. So their main conclusion of this study was that in patients with bipolar depression, despite continued use of lithium, addition of lamotrigine, there was superior benefit as compared to placebo. This was a meta-analysis looking at lamotrigine for acute bipolar depression, and when they combined data from these six studies, they showed that lamotrigine was superior as compared to placebo, and the risk ratio was 1.26. And when the data from the additional study was added, it was not published, so from the Vanderloo study, the hazard ratio or the risk ratio increased further from 1.26 to 1.47. Now what's the evidence base for lamotrigine for acute unipolar depression? So this is a meta-analysis from about 18 studies published in 2016, where they looked at studies where they used lamotrigine for both unipolar or bipolar depression. In unipolar depression, there were four studies, so the number of studies was much lower. The sample size was much smaller, about 187. One study compared lamotrigine versus lithium. Second study used lamotrigine as an augmentation agent of antidepressant. For bipolar disorder, there were 14 studies, and looking at comparing as monotherapy versus placebo, monotherapy versus lithium, or olanzapine plus fluoxetine, that's the Symbiox, and the augmentation of antidepressant versus placebo, and augmentation of mood stabilizer. So when we looked at the results, lamotrigine sufficacy for depressive symptoms did not differ significantly in monotherapy versus augmentation, or in unipolar versus bipolar. So at least when you look at this meta-analysis, we saw a response even for unipolar depression, although the number of studies are much smaller here. It was superior to placebo, with a relative risk of about 1.42. And lamotrigine did not differ regarding efficacy on depressive symptom response or remission from lithium, Symbiox, or olanzapine fluoxetine, citalopram or inositol. Now what's the evidence base for lamotrigine in other psychiatric disorder other than bipolar disorder? So there's some data for borderline personality disorder. So the initial two randomized controlled trials were positive. It showed that lamotrigine may help in reducing anger, affective instability, and impulsivity. There's a recent randomized controlled trial from England with a much larger sample size, and lamotrigine was not superior to placebo in that. We did a meta-analysis, and we combined the data. Lamotrigine was not superior to placebo for a patient with borderline personality disorder. What about borderline personality disorder with bipolar disorder? So about 10% to 15% of patients can have borderline personality disorder as a comorbid condition. So this was a retrospective study, and there was some evidence that dimension of borderline personality disorder improved significantly with treatment and corresponded with the response of bipolar symptoms. Whether it's the improvement in bipolar disorder that is correlating with improvement in some of the borderline personality disorder. So it's a weak evidence for lamotrigine in borderline personality disorder. Treatment-resistant schizophrenia, treat with clozapine. So there's a meta-analysis about five studies, small sample size, 161, and lamotrigine was superior to placebo in both the primary, looking at PANS and BPRS, and the secondary measures for both PANS positive, so both positive and negative symptoms. So there's some evidence there in combination with clozapine. What sort of things we should think about when we are considering lamotrigine? So we should think about lamotrigine inducers and medication which could reduce or increase the lamotrigine serum level. Because most of the studies, when you look at, they do not account for that. When we see a lot of the retrospective study, you'll see patients who are on lamotrigine, they're either on oral contraceptives in females, or they are on carbamazepine or valproate. So carbamazepine, phenetine, or estrogen-containing contraceptives can reduce the efficacy, not the efficacy, but reduce the serum level of lamotrigine, so they induce lamotrigine metabolism. So if you have a patient or a young female who is on oral contraceptive or who is stable on lamotrigine, and you or the patient decide to start an estrogen-containing or oral contraceptive, you want to monitor the serum level, they may need a higher dosage. So be aware of that. So a lot of people are not aware of that in particular. Valproate can reduce lamotrigine metabolism, so that's the group where we see a little bit of higher risk of rash with lamotrigine because it can raise the serum level of lamotrigine. So when we look at the dose titration with lamotrigine, so this is the most common one we see, 25 milligram orally for two weeks, then 50 milligram for two weeks, then 100 milligram for a week, and then 200 milligram. So most of the studies, when we look at lamotrigine, dose is 100 to 200 milligram, but you can go as high as 400 milligram. Now what if you have a patient who is on Valproate? Can we use lamotrigine? The answer is yes, but we have to make some dose modifications. So if you have a patient who's on Valproate, but they're not having manic symptoms, but they're having more depressive symptoms, and we're thinking of adding lamotrigine, so we want to slow the titration here, 25 milligram orally every other day for two weeks, then 25 milligram for two weeks, 50, and then the maintenance dose of 100 milligram would be very similar to about 200 milligram of someone if they're not on Valproate. Now if you have a patient who is on carbamazepine, so we kind of have to be a little bit more aggressive with the dosing here because carbamazepine is an inducer, so it will reduce lamotrigine's blood level. So we do 50 milligram orally for two weeks, then 100 milligram, and then 200, and the target dose here is 400 milligram. Now there's a paper just got published in Journal of Clinical Psychopharmacology last year where they compared patients who titrated, where they titrated lamotrigine faster as compared to slower. The problem with this paper was, you know, the odds ratio was pretty significant when you look at the fast titration group increased the risk of development of rash as compared to slow titration where they showed it decreased the risk, but there are only nine patients in the faster titration, and they went pretty quickly from 100 to 300. There was no rationale why in that particular group whether there was a particular reason they went from 100 to 300, but it's a very small sample size, and for this odds ratio, their confidence interval was very broad. But overall, when we look at that data, it seems like if you do a faster titration, those are the patients that may have a higher risk of rash. However, none of the patients had Steven Johnson's rash or 10, but we worry about. So it was the rash, but not the Steven Johnson's rash. All right. So when we look at mood stabilizers, I think pretty much all the mood stabilizers have at least one black box warning, and with lamotrigine, the warning is for severe rash, more Steven Johnson's syndrome. It's not very common, so the prevalence is very low. It's about 0.03 to 0.08%. There is a test we could do, HLA-B1502, so if you are worried and you are thinking about, you know, starting lamotrigine and you are extremely concerned about the risk of rash, this is a blood test, or it can be done via the swab testing. Most of the pharmacogenomic tests these days, they do this testing, and if that test is negative, it cannot completely rule out, but it can give some additional information if it is positive and you want to avoid lamotrigine. There are a couple of new warnings now that it can increase risk of hemophagocytosis, lymphohistiocytosis. It's an immune reaction of what we see with lamotrigine. It's not very common, but in April of 2018, FDA added that, that it can cause a rare but serious reaction where there's a severe inflammation throughout the body, so if you're a patient and you start lamotrigine and they start having fever and feeling sick, you want to get assessed and get the blood workup done. There's a second warning now that it can increase risk of arrhythmias in patients with heart disease. There's a dose correlation there, but those are the two newer indications, so just to be aware of, you are considering lamotrigine if you have a patient with a history of arrhythmia or heart disease, they can increase risk of arrhythmias. Now, so Dr. Gertlund presented data regarding lithium, and so I get asked this question all the time, well, is there any serum level of lamotrigine? So we did this systematic review published in 2021, and we ended up with seven studies, and all of those were observational studies, so most of the observational studies, they have a higher risk of bias as compared to a randomized controlled trial, so the data for lamotrigine serum level is much weaker than what we see in lithium, but we saw a constant theme in most of the studies that a minimum lamotrigine threshold level of three microgram per milliliter in patients with mood disorder was, most patients responded they had at least that level. The other thing was they did not look only at bipolar disorder, so some of the studies had patients with treatment-resistant depression, whether they had unipolar or bipolar disorder, so they kind of looked at overall. One of the studies showed that the plasma concentration of greater than 3.25 microliters at eight weeks did correlate with early improvement in patients with treatment-resistant depression. There was another study looking at patients with bipolar disorder with rapid cycling, and they showed if the serum level was greater than five microgram per ml, they had a better mood improvement. So data was inconsistent regarding the therapeutic range, so if you have ordered lamotrigine level, you'll see there's a big range from three to 13 to 15, depending on your lab, so that range is usually for patients with seizure disorder, epilepsy, but if we want to extrapolate data from the systematic review, if you have a patient who's on lamotrigine and they're not responding at least to a blood level if it's less than three, you could optimize the dose higher if they are, let's say, 200 milligram and they're not on other medication which could reduce the efficacy and their level is greater than three, then maybe you could think about switching to a different medication. There's lack of double-blind randomized controlled trial assessing the efficacy of lamotrigine in clinical improvement in mood disorders. So I'll talk a little bit about some of the biomarkers. So there's not much data with lamotrigine as compared to lithium and acetyl aspartate, so that's one of the neuronal metabolite markers of mitochondrial activity, and we measure this in brain with magnetic resonance spectroscopy. It's a source of acetate for myelin synthesis and is a precursor of this N-acetyl aspartyl glutamate, so a truncated, right? It's synthesized in neuronal mitochondria from L-aspartate and acetyl coenzyme, and it's linked through the glutamine and TCA cycle, so it's involved there. So this was one of the studies published by Dr. Fry in 2007. So what we are looking at on the left side is a section of brain, and this is a voxel here looking at the anterior cingulate cortex, and the right side here we are looking at the N-acetyl aspartate. So the question was, can NAA stage disease, and what we saw here is that there was a positive correlation in the basal ganglia NAA creatine ratio with the phosphorylation for mania, so patients who had a lower level of basal ganglia NAA, they had a higher or larger number of phosphorylation as compared to patients who had a higher NAA, which was a significant R was 0.57. This was another study done by my colleague, Dr. Paul Kowarkian, published in 2015, looking at patients' NAA level at baseline, and we looked at patients with bipolar disorder. They had a lower level of NAA, and then 12-week post-lamotrigine treatment, and NAA was normalized after lamotrigine treatment. This is some additional data from our bipolar biobank. It was established in 2009 by Dr. Fry and Dr. Joanna Pernaka at Mayo Clinic in collaboration with the Linner Center of Hope, so Dr. Sue McElroy, she'll be presenting some of this data in detail because we did this study looking at some of the biomarkers for mood stabilizer group, and we have other colleagues from Chile, University of Minnesota, University of Mississippi, and from Mexico as well. So in this study, we had 199 patients with a diagnosis of bipolar disorder who were enrolled in the Mayo Clinic Bipolar Biobank. Forty-three percent were males, and the mean age at the time of interview was 42, and 72 or 73 percent had bipolar disorder type 1. Forty-three percent were on valproate, 38 percent were on lamotrigine, and the other 7.5 percent were on valproate and lamotrigine. The other mood stabilizer combinations were not as common. So I won't go too much into detail because there's some overlap with my slide and Dr. McElroy's, so she'll be talking more in detail about some of these findings, but we did see that there are three particular SNP-level findings, the THSD7A, ABCC1, and DISP1, which were significantly associated with mood stabilizer response in bipolar disorder patient. In particular, the THSD7A, when we look at drug stratified analysis, it associated with both lamotrigine and valproate response, but ABCC1 and DISP1, they associate mostly with better valproate response. So in conclusion, lamotrigine is an effective therapy for bipolar depression. We see more evidence for depression than for mania or hypomania. It has a very low risk of Steven Johnson syndrome, so most patients will have rash, but not necessarily the Steven Johnson rash, and the risk is lower with slower titration. There is some weak evidence of lamotrigine serum level in mood disorders. There are significant interactions with valproate, carbamazepine, and oral contraceptives and probable biological markers here. I'd like to invite my colleague, Dr. Sue McElroy, next. She'll be talking about valproate, so thank you. Okay, so good morning, everyone. I'm going to give you a brief overview of some genetic, some clinical pearls and potential markers of response to valproate in bipolar disorder. These are my disclosures. So valproate in bipolar disorder. Valproate's an incredibly useful tool for people with bipolar disorder. You don't use it in everyone, but it can be really useful in certain patients. This drug only has one psychiatric indication. It is indicated for the treatment of acute manic episodes associated with bipolar 1 disorder, and that's been indicated for acute mania since 1995. It's important to remember that the therapeutic levels with valproate are higher. We hear about how we can get away with lower amounts of lithium. That's not so for valproate. In fact, I think one of the most, to this day, one of the most common mistakes made with valproate in the treatment of bipolar individuals is not using a sufficient enough dose. So we've got good data that, you know, initially we thought the therapeutic level was around, well, sometimes 45, then 50, but it's clear it's 80 and above. The other really important thing to remember about valproate is, you know, unlike, certainly unlike limotrigine and pretty much unlike lithium, you can load it, which can be very helpful for treatment of both inpatient and even outpatient emergencies. So you can begin the drug anywhere between 15 to 30 milligrams per kilogram per day and achieve therapeutic levels the next day. And I'll show you some slides on that showing that valproate loading, not only can you do it and people tolerate it, but you do get more rapid onset of efficacy. It's a really important thing to remember. We do have data. There's never been a large double-blind placebo-controlled study of valproate in bipolar depression. There are several small studies, however. And a meta-analysis of those studies suggest the drug may also have acute antidepressant effects in bipolar disorder. And we also think the drug may have some prophylactic effects. I have not gone into the whole story of the prophylactic study of valproate in bipolar disorder. If we have questions, we can talk about that. It was one of the first studies to be done and it was a true prophylactic study. So no matter what you responded to, you got randomized to valproate or lithium or placebo. And what we learned from that study is we really need to do enrichment studies. We need to treat somebody with the drug we want to test. If they respond, then they get randomized to stay on the drug or go on placebo. So what that valproate prophylactic study taught us, because the findings were equivocal, it was not overly positive, what it basically told us was it taught us how to do more appropriate maintenance studies with lamotrigine and with many of the atypical antipsychotics. Other uses of valproate, epilepsy, a number of different types of epilepsy. I'll use it in people with clozapine to help prevent seizures. Migraine prevention, which is an important thing to remember because migraine is such an important comorbidity of bipolar disorder. And there's some emerging data suggesting the compound may have some anti-cancer properties. Okay, the FDA approval of valproate for acute mania was based on two pivotal studies. One, our group did in McLean ages ago. This is the second one done with the lead investigator of Dr. Charles Bowden. And these were bipolar I patients with mania that got randomized to divalproate, or lithium, or placebo. And you can see, you know, boing, it's right there. Valproate and lithium were both superior to placebo in the treatment of acute mania and they were basically similarly effective. Now in this study, valproate was begun with a titration, 250 TID, okay. Now what we've learned is that, like I said, we can load this drug. And here is a pooled analysis of a number of studies comparing different dosing strategies of divalproate with other anti-manic agents. And you can see, so if you look at the middle two lines, the green and pink lines where divalproate and lithium are being compared, neither the divalproate nor the lithium are loaded there. And then if you go down to the bottom two lines where valproate is loaded, and this was I think around 30 milligrams per kilogram per day, you see basically equal efficacy with an atypical antipsychotic for the treatment of acute mania. I just can't stress what a nice strategy this has because I do this in outpatients too. Now with outpatients, I tend to load at 15 milligrams per kilogram per day. But if you have somebody, but sometimes you can maybe even avert a hospitalization. Now I don't think we have any clear-cut, definitively proven clinical markers for valproate response in bipolar disorder. We have a bunch of hypothesized clinical biomarkers. One of the leading clinical features, it was initially thought that both mixed states and rapid cycling might be associated with valproate response. I think now that's because the people in the old days who got valproate were people who failed lithium. And so the people with mixed states and lithium tend to fail lithium. And then some would go on to respond to valproate. So it's really, I don't think it's clear whether or not mixed features in rapid cycling really are predictors of good response. In fact, I don't think they are. They're severe variants of the illness that tend to predict less response to our available treatments. There's also some emerging, there's old data that so-called secondary mania, or mania after head trauma, post-stroke mania, or mania associated with EEG abnormalities might be associated with valproate. Now again, these are really more like case reports. There's really no study that proves this. But we'll circle back to this issue. There is, however, one very nice double-blind placebo-controlled study in people with bipolar disorder and alcoholism. And valproate did, in fact, it didn't just reduce bipolar symptoms, it reduced drinking. So that's really something to always, if you have a very severely ill patient who's bipolar and actively drinking, you might want to think about valproate. Now we know valproate is approved for migraine prevention at much lower doses than you use in bipolar disorder. I am unaware of a study that took people with bipolar disorder and migraine and randomized them to valproate or placebo. So again, I don't know if migraine is an actual biomarker of response for valproate. But it's another good thing to remember just because the comorbidity of bipolar disorder with migraine is so large. These are some of the hypothesized clinical biomarkers for valproate response. Valproate, like the other mood stabilizers, has some serious side effects. I guess the one I'm going to talk about is hepatotoxicity and hyperammonemia. Because we may have biomarkers for those potentially severe side effects. So there's two types of hepatotoxicity with valproate. There's just a regular elevation in liver enzymes, which is usually no big deal. It often just goes away with time. But then there's this idiopathic, very, very severe hepatotoxicity where you get hepatic failure and people can die. Well we now know, before we didn't really know what that was due to. But now we have found there's a clear association in people who have mitochondrial disorders caused by Polg mutations. Polg codes for mitochondrial DNA polymerase. So it's basically contraindicated in people with a mitochondrial disorder due to a Polg mutation. It's also, you want to avoid it in children who are two years or less in age suspected of having a mitochondrial disorder. And I've listed for you that Alpers-Huttenlocher syndrome is an example of one of these disorders. And you can see epilepsy is part of it. And then hyperammonemic encephalopathy. We now feel valproate is contraindicated in patients with known urea cycle abnormalities. This is a group of uncommon genetic disorders. I guess a more prominent example being ornithine transglycerin deficiency. So we do have some biomarkers of severe side effects. What about pharmacogenetics of AED mood stabilizer response? Dr. Singh began talking about this study. We, a bunch of us here, conducted GWAS to identify genetic variations that were associated with AED response and bipolar disorder. Again, this is, Dr. Singh showed this slide. Most of the patients were receiving valproate or lamotrigine. And then here is the Manhattan plot showing that multiple SNPs in THSD7A were significantly associated with anticonvulsant mood stabilizer response in patients. And this applied to both valproate and lamotrigine. Okay, so what is THSD7A? This is what I was able to look up. THSD7A encodes thrombospondin type 1 domain containing 7A. Interestingly, it is expressed in glutamatergic and GABAergic neurons, which we know are targeted by AED mood stabilizers. I think it's important. A good way to understand it is that it's a potent neuroangiogenic factor that coordinates growth of nerves and blood vessels. And it has been associated with both bipolar disorder and schizophrenia. And now I'm going to just come back to this whole notion of valproate and angiogenesis and valproate possibly being helpful for secondary mania. Valproate enhances endothelial sprouting in vitro and there's an animal model that shows that chronic valproate treatment can promote post ischemic angiogenesis and functional recovery in a rodent model of ischemic stroke. Conversely, and don't ask me to explain it, valproate actually reduces angiogenesis in cancer cells. So another time our group really thinks about Depakote is when somebody has either, you know, there's a clear-cut central nervous system insult that precedes a manic episode. Now there were two other genes that were significantly associated with AED mood stabilizer response. As Dr. Singh said, they were both associated with valproate response. And this was ABCC1 and DISP1. Okay, what are these? ABCC1 encodes multi-drug resistance associated protein 1. It's a key protein linked to multi-drug chemotherapeutic resistance. It plays a role in cellular efflux of drugs and so maybe it regulates AED mood stabilizer effects in bipolar disorder. Now DISP1 encodes a protein essential for normal hedgehog signaling. It is important for normal cellular proliferation and differentiation. The role of this gene in brain and mood disorders is unknown. However, a snip close to DISP1 to the DISP1 promoter region has been associated with response of OCD to SRIs. Okay, but as Dr. Gitlin said, unfortunately these genetic findings aren't ready for prime time. They're not clinically useful at this point. Hopefully they will lead to clinical useful modalities, but not right now. It's too premature. So to conclude, Valproid remains an important part of the therapeutic armamentarium for bipolar disorder. We think it might, you know, it can work in mixed states, rapid cycling, secondary mania, including when lithium doesn't. There's a one nice study showing that it reduces alcohol use in people with bipolar disorder and alcohol dependence. It also works in migraine. And it's important to remember that it might be an important treatment for that subgroup of bipolar patients who have comorbid migraine. You want to avoid this compound in people with POLD mutations and urea cycle disorders because those people can have the serious adverse effects. And we think there may be potential genetic markers for Valproid response in bipolar disorder. Maybe these genetic markers will become targets to help development of other compounds for bipolar disorder. But as Dr. Gitlin said, it's not ready for prime time. And thank you. Good morning. I'm Mark Fry and I'm the caboose of this symposium. And I had asked Dr. Singh if we could talk about antidepressants and bipolar disorder because we have very little data. And you saw these excellent presentations on clinical predictors of response with very well studied treatments in bipolar disorder. That will not happen in this talk. We have very little evidence base that these drugs work. And yet they are extensively prescribed in ways that we need to try to better understand and perhaps find that group of individuals who for antidepressants are in fact effective mood stabilizers. And we'll kind of walk you through that here. I'm showing you my disclosures. I want to really point out that we've had investigator initiated study support from AssureRx, Myriad Genetics, and I will not be presenting any of that data as part of this presentation today. I will be presenting pharmacogenomic data of antidepressant response from very separate studies and grants. So, Balwinder had asked, can we first review indications of the drugs that are part of this symposium? And you see the indications for antidepressants and really want to underscore how expansive our clinical research portfolio is for antidepressants in major depressive disorder, seasonal affective disorder, premenstrual dysphoric disorder. We have excellent regulatory studies that have gotten those indications in children and adolescents. Midsection. I'm trying to see if I've got an arrow here. We have that middle section is all the anxiety disorders, GAD, SAD, panic disorder, obsessive compulsive disorder, post-traumatic stress disorder. There has been a profound investment in really looking at these drugs for major mood disorders and anxiety disorders with the exception of bipolar illness. We further have indications in bulimia, fibromyalgia. Recall that adamoxetine is a neurogenergic reuptake inhibitor, FDA approved for attention deficit disorder. Not available in this country, but roboxetine is a neurogenergic reuptake inhibitor, FDA approved for major depressive disorder. We have antidepressants that are FDA approved for smoking cessation, for obesity. And yet there has been virtually no interest, no investment in looking at these medicines and their efficacy profile or safety profile in bipolar depression with one notable exception. And you see that here. Fluoxetine in combination with the lantopine is FDA approved for bipolar depression and that's it. We have 25, 30 regulatory approved antidepressants in this country, underscoring the bottom part of that slide. They are off label, meaning there's no FDA approval. So let's take a second and really remind ourselves, what does that FDA approval mean? You may say that's just a lot of bureaucracy. The label tells clinicians, patients, there are controlled studies showing that this drug works. There are controlled studies to show this drug is safe. And if there are concerns of a safety profile or a black box warning, there's information to really make informed decisions as to whether these should be treatments getting started. And by and large, while it's a very separate process from the regulatory approval, there tends to be a pretty substantial preclinical research database of these drugs that start to give us an indication mechanistically as to how they might be working. That is all absent in the context of antidepressants in bipolar depression. So lack of FDA approval, how extensive are these antidepressants in our practice? Profoundly so. What I'm showing you here is a 20-year trend of pharmacotherapy standards in bipolar disorder. This is from a number of insurance actuarial data sets associated with several institutions in the Upper East part of our country. The left part of the slide, what you can clearly see, is that our any sort of mood — sorry, let me rephrase. The takeaway from the left part of the slide, that blue line, atypical antipsychotic mood stabilizers are increasing substantially. When you look at the percentage of office visits, including one of these medicines, that now is over 50%. That really shouldn't come as a surprise. There's been a 15-year investment in multiple atypical antipsychotics, showing the regulatory level data of efficacy and safety, that does correspond. But at the same time, what we see is that green line of lithium use decreasing over that 20-year period of time, and mood-stabilizing anticonvulsants in yellow doing the same. So then if we look at the right part of the slide, again, these are percentage of office visits that include a specific pharmacotherapy. More than 50% of office visits include an antidepressant. And yet, with the exception of fluoxetine, we have no FDA approval for these compounds in this illness. So let's take a little bit of a closer look at that. And I really chose kind of different types of studies and the best examples of that. So the first one is the EMBOLDEN2 study that was led by Dr. McElroy, largest study of an antidepressant monotherapy in comparison to placebo and quetiapine. In that study, paroxetine was no better than placebo. Quetiapine was very effective and was really one of the studies that led to its FDA approval for the depressive phase of illness. A number of meta-analyses have been done. I'm simply showing you one in that second blue bullet. Six trials over 1,000 patients. When you look at an antidepressant as adjunct to a mood stabilizer versus placebo, there is no significant benefit of the antidepressant from a response outcome, typically a 50% reduction or achieving remission. Now let's drop to the bottom bullet. This is actually where one of the meta-analyses I think starts to become interesting. And maybe it guides clinicians in some way. This meta-analysis similarly identified no overall benefit of antidepressant versus placebo. But when they started to look at the type of mood stabilizer, a signal came through. Now remember, the majority of these studies would require that an individual be on lithium valprate atypical antipsychotic. Then we study the antidepressant versus placebo. In that secondary analysis, there was in fact differential efficacy. There was a better outcome of that antidepressant when prescribed with an atypical antipsychotic versus when prescribed with lithium or an anticonvulsant. Think about that for a second. These atypical antipsychotics, many of them have FDA approvals and treatment resistant depression. And it may be that that's the sort of signal that we're seeing here. So if we are needing to think about differential anti-manic mood stabilization prior to considering an antidepressant, there might be a clinical pearl for us. But I would argue pretty small and pretty great. Where is the best evidence for an antidepressant in bipolar disorder? The last decade, we've learned a lot. It's a pretty clear cut, I would argue, narrow group of bipolar patients. But they tend to be individuals with bipolar type 2 disorder, depressive symptoms that are not associated with psychosis. In my opinion, I still think the rapid cycling piece is under review. This data would suggest that bipolar 2s who are depressed do very well with antidepressant monotherapy. Jay Amsterdam's group has looked at venlafaxine compared to lithium. The last bullet was a study many of us did with the late Lori Altshuler, where in fact sertraline monotherapy was just as good as lithium. The combination gave no additional benefit, but certainly was associated with higher dropouts because of side effect burden. So we have a project at Mayo that's being led by Manuel Fuentes, and we're trying to develop a conversation tool that individuals with lived experience and providers can use to guide how we might think about choosing a treatment for the depressive phase of bipolar disorder. We thought this was going to be a simple task, much like antidepressants and major depressive disorder, and we've come to realize it's anything but simple. But Manuel and his team with Mete Ercis and Ashley Zhang have tried to really consolidate the clinical evidence base in a way that's really simple to understand. So if you as a clinician are wanting to consider an antidepressant for bipolar disorder, the data that is the best, but again pales in comparison to FDA-approved treatments would be the ones that you see here, bupropion, escitalopram, fluoxetine, sertraline, venlafaxine. But to be very clear, if there is a modest evidence base, it doesn't mean that a particular patient will respond to that, and that's where I think the individualizing or precision medicine will be critical in this space. So this is a study that I continue to cogitate about some 20 years later for the following reasons. This is a very small sample, but it's a critical observation that there are a minority of bipolar patients for whom antidepressants really are maintenance mood stabilizers. So this was a study that we had with our Stanley Foundation Bipolar Network. We simply were following naturalistically individuals treated with an antidepressant. Bipolar, depressed, on a mood stabilizer, still depressed, got on an antidepressant, didn't have side effects, it worked, it didn't get manic. We watched and identified participants who were in remission and on that drug and followed them for a year. The left part of the slide, what we're showing you is in this cohort, if you stopped that antidepressant before six months, you relapsed, not into mania, back into depression. The right part of the slide is trying to stratify that a little bit more to better understand where does there seem to be that risk. So that red line are bipolar, depressed patients, again, on a mood stabilizer, got on an antidepressant, responded, didn't get manic, didn't have side effects, but stopped that medicine within six months. They are relapsing into depression far faster than the green group who's on that antidepressant six to 12 months, and they are relapsing far faster than those individuals who have stayed on that antidepressant for a year. Now this phenomena represents about 15% of our Stanley Foundation cohort, so it is not common, but our field got stuck for about a decade with these arguments that antidepressants work in bipolar, they should be prescribed, antidepressants don't, it's malpractice to do that. Those types of unsuccessful arguments need to stop, and I would say the better question is for whom are antidepressants really good mood stabilizers, and how do we identify that either by clinical prediction or by a biomarker. But as I had stated, there are no clinical predictions as to how these medicines are helpful. So I'm going to shift and really talk about some biomarker data that I would say is not quite ready for prime time, but I do think it's setting a roadmap as to how we can think about biological markers, clinical predictors, and how we actually need to study them as an interaction. So we've been very keen on trying to understand the neurobiology of the adverse event when an antidepressant is started, and with a matter of weeks a patient becomes manic. This slide is simply showing you that those incidence rates differ by the type of study that you have. Retrospective chart review, 40%, our Mayo Clinic Bipolar Biobank required a physician to review electronic health records and confirm antidepressants started within eight weeks and induction of mania or hypomania. I think the most relevant point, if we start thinking about clinical variables and how we might then study them in the context of biological markers, that bottom study by Karolinska Institute colleagues in Sweden is really remarkable. They found that there was indeed an antidepressant-induced mania phenomenon in a number of patients, but there was a window of vulnerability. And if you were on that antidepressant and did not get manic within the first three months, the likelihood that you would further was negligible. What is that neurobiology of a transient but profound risk of an antidepressant in the context of the illness of bipolar disorder? And what's happening where that neurobiology risk or vulnerability fades away after three months? That's a question we need to answer. So we've been thinking about this phenomena and really underscoring we need to think of those clinical factors, we need to think about biological markers, and we need to think of these as interactions and not study them in silos. So the way we're kind of thinking about this as it relates to that risk factor as a switch, are there gender and age risks? There are. Women, adolescents. Is the phenotype conferred differential risk for antidepressant-induced mania? You bet. Bipolar 1s far more so than bipolar 2s. This is really peculiar because the bipolar 1s are on these anti-manic mood stabilizers and yet they have these higher rates of starting an antidepressant and switching. We know the number of prior episodes of depression contributes to this risk. Mixed depression, that little bar graph on the right is a study of ours from a number of years ago showing even very soft symptoms of mania in depression are associated with a subsequent treatment emergent mania when that antidepressant gets prescribed. The comorbidities, the drug therapies, and we've been very interested in starting to look at genetic markers. I'm showing you on the left a meta-analysis of the serotonin transporter in antidepressant-induced mania and bipolar disorder and it is nominally significant. My statistician geneticist hates this study because she thinks it's really not meaningful and she's right. But it's meaningful for an audience like this. This is suggesting there may be a signal with this serotonin transporter in antidepressant-induced mania. But I would argue to the right part of this slide we have not even touched norepinephrine transporters, dopamine transporters. We haven't done any GWAS analyses. Our genomic analyses need to go to a different level. For example, we can use polygenic risk scores. This is a study that's showing the polygenic risk configuration of someone with major depressive disorder who responds to an antidepressant actually looks very similar to a bipolar patient who gets manic when exposed to those antidepressants. This gets to a really important point. We have to remember the biology of the illness when we start thinking about clinical outcomes or genomic outcomes related to treatment. Sumacro showed you a signal in only 200 patients where THSD7A is associated with responding to lamotrigine and valproic acid. But THSD7A is also a risk gene for bipolar disorder. We need to start thinking about the biology or the genes of the disease and then how on top of that the omic markers of treatment response may impact that. Here's another example. Let me just close with this last study. We've been very focused on looking at mitochondrial function. If you look at the upper left, that orange little candy bar is complex one. That's an area of research where anandriaza is clearly identifying electronic transport chain dysregulation and bipolar disorder. Complex four is where Su Tai at Mayo Clinic developed an animal model of mania that's impacting that. Complex five on the far right is where Hillary Bloomberg's been studying energy dysregulation and bipolar illness. One of our predoctoral students at Mayo reviewed all the preclinical literature of psychotropic drugs and how they impact mitochondrial energetics. That's what you see in figure B and in the bottom left. We simply took that preclinical data and said, does mitochondrial energetic profiling and how that changes with antidepressants confer differential risk for antidepressant induced mania. This data was just published in Molecular Psychiatry. It shows that antidepressants that increase mitochondrial energetics in preclinical settings, and that's primarily peroxetine, venlafaxine, and nortriptyline, they have much higher rates of treatment emergent mania associated with antidepressant therapy in comparison to antidepressants that do not increase mitochondrial energetics. First data that is starting to look at mitochondrial profiling, regardless of serotonin reuptake inhibition to see if that can advance some of our thinking. Evidence-based treatments for antidepressants and bipolar disorder, profoundly limited. The extensive use of these drugs does not align with the clinical evidence base. We actually think biomarkers with very narrow phenotypes and large sample size that take into account clinical variables and the biological marker as interactions are really the future of our research. And if we can encourage you to come to our symposium tomorrow afternoon, we'll show you how a learning health system may be the very roadmap or footprint to start studying that. This work has been funded by a number of different organizations with wonderful colleagues and friends that are listed here. Thank you. We are up for questions now, so we have some questions online, and so I'll try to take one question from the audience and one question from the online. Is this working? Hi, everyone. My name is Jan Ellison-Bizuki, and with my husband, David Bizuki, who's the CEO and founder of Roblox, we have a family foundation and a nonprofit called Metabolic Mind, which advances metabolic therapies for psychiatric illness, in particular bipolar disorder. And Mark and I work together on the BD2 Initiative, which is a funding collaborative for bipolar disorder. If you want to read about metabolic therapies, I actually have an op-ed published today in the San Francisco Chronicle that tells our family's story, so I won't go into that now. But my question is, we're talking a lot about mood-stabilizing drugs that are used in epilepsy and their efficacy for bipolar disorder and other psychiatric illness. The single most effective intervention for pediatric epilepsy is a ketogenic intervention, nutritional ketosis, otherwise known as a ketogenic diet. So I'm wondering if anyone can speak to the use of ketogenic diets in psychiatric illness. We do have five studies underway, but I'm particularly interested in mechanism and how we might compare those to the mood-stabilizing drugs. So for me, the ketogenic story lines up with a historical component, really recognizing the benefit of ketosis for pediatric epilepsy. And some of those studies were actually published in 1921 in the Mayo Clinic Proceedings, really showing that this patient population, when in ketosis, had a powerful anticonvulsant response and when not having that nutritional intervention, seizure is returned. And if you think about how some aspects of epilepsy, electroconvulsive therapy, mood-stabilizing anticonvulsants may actually have similar neuronal stabilization, this gets to be very interesting indeed. I don't want to suggest that that nutritional intervention is going to function like valproate or lamotrigine, but there's a common base where we can start to think about that and study mechanistically. I do know that a number of early studies for therapeutic ketosis and bipolar disorder will be presented at the International Society for Bipolar Disorder meeting in Chicago in several weeks' time. And I think that will be critical data to look at, both for feasibility and treatment response and hopefully setting the stage for larger collaborative work. Thank you, Mark. So I have an online question for Dr. Gitlin. And the question is, in your experience, is there a difference in outcome in lithium's efficacy in BID versus TID dosing? The second question is, there's a patient who is on lithium for maintenance treatment for a long time, their creatinine level is 2.0, they do not want to go off lithium because that has worked the best. What sort of advice or suggestion would you give? So two questions in one. So the answer to the first question is, there are no great studies comparing efficacy of lithium with the different regimens. Nobody is doing TID, because nobody can remember to take anything TID. But even comparing BID to 1-Stale, I start all of my new lithium patients, those I can convince at this point in time, to do it 1-Stale. I didn't show the data because we didn't have time. But there's good evidence that 1-Stale lithium, if you start it in the beginning, prevents or slows down the renal problems. So yeah, there's no evidence there. The second one was? High creatinine level of 2.0 with lithium. So a creatinine level of 2.0 probably translates to an EGFR in the 40s, is the guesstimate. that's flirting with problems. Yes, you can have informed consent and document and all that stuff. That would make me, it has made me profoundly nervous having run the Mood Clinic for 45 years. I have young people I put on lithium who are now old people on lithium. How did that happen? And you watch their renal function go down and I've already had a few patients end up with transplantation or dialysis where lithium was the only known risk factor. Because of that, when I see a creatinine go up, an EGFR go down to a certain place, I lose my nerve and I say to the patient, I can't do it. Can I just ask a very relevant question because you showed data that after about 10 years of lithium, your kidney function really starts to decline. I think all of us have had clinical scenarios where we're reaching that ninth or 10th year on lithium. The course of illness has been pristine. Do you think we should consider sort of pulling off at a decade to minimize that acceleration or just further watch? Again, it has to do with, first of all, we should get nephrologists involved in this so they'll frankly know more about lithium in kidneys than they do. They know more about kidneys, hopefully. Low, well, you can lower the dose, but hopefully you've done that earlier. It depends what the EGFR is. So if it's in the 50s, we can just watch and monitor carefully. Once it gets down into the 40s, again, I just, having caused this a handful of times, I lose my nerve. So, all right. There were really no other factors, like no hypertension, no? Sometimes there are and sometimes there aren't. When you do renal biopsies in patients on lithium, in patients whose renal function has decreased, they all look the same. It's not glomerular, it's all interstitial nephritis, which is relative preservation of the glomeruli with destruction of the tubules and lots of fibrosis in the interstitial. I think Dr. Singh should present a counterpoint. Yeah, I think we'll need a whole new session for that question. I'll just. How about just a little pearl? Yeah. These are pearls. Yeah, so my observation, so I'm interested in lithium and chronic kidney disease. So it's not a drug for everyone, just like any other medication, but it's a drug for some people for whom it works really well and they do not respond to any other interventions. So I've published a few case series, observational studies looking at patients who are in long-term lithium who are at a higher risk of developing chronic kidney disease and diabetes mellitus and patients who are on high dose of benzodiazepine, which is more as a surrogate marker probably for hypertension and shown to be increased risk of chronic kidney disease. Prevalence rate is about 26%. 50% of those patients discontinued lithium, 50% continued. The risk of relapse was eight times higher. So once you discontinue lithium, your patient 99% of the time, they are going to relapse and they may not respond to any other medication. And one third group may not even respond to reinitiating lithium. So I'm very cautious with discontinuing lithium. I think we'll have, we should think about another symposium. That's another symposium. Yeah, I'll just mention there's a drug, a class of drug called as SGLT2 inhibitor. A lot of clinicians are not aware of that. This is a newer drug, which is FDA approved to reduce the progression of chronic kidney disease. So when I talk to nephrologists, they are okay if you have a patient who has lithium induced chronic kidney disease, have that discussion, see if they're a candidate for that drug. That may help slow down the progression. And those drugs also have, they reduce heart disease as well. That is correct. Okay, thank you. Next question from the audience. Okay, yeah. I have a question about patients. I'm the medical director at a clinic, substance abuse treatment clinic. We get a lot of patients discharged to us from the inpatient service who are bipolar and they come out of the hospital and they're still manic. They're not stable. They're not stable. And they're on long acting, they put them on long acting injectables. And a lot of times they put them on long acting injectables of first generation antipsychotics, prolixin, Haldol, and they come to me and I have to stabilize them. So I have a couple of questions. So they're on prolixin, Haldol. I have two questions. One is, what do you think about putting a patient on Haldol, 100 milligrams injectable a month, on lithium? Because it's controversial and I wanna hear your opinion about that because it's very effective but it's controversial. And the second thing is, if they're on a first generation antipsychotic, what do you think about using a second generation antipsychotic like Latuda or whatever you wanna choose, adding that to the prolixin or the Haldol to stabilize them? So those are my two questions. So circa 2023, which is really important to reference because these drugs have been around since the 1950s and 60s. I think all of us would agree that the first generation antipsychotics are very effective anti-manic agents. And it may just, and your observations coming out of the hospital like that, I think are shared everywhere. The data are very clear. They are not effective for the depressive phase of illness. There is more post-manic depressive symptom severity when using medicines like Haloperidol. There's more, there's faster and more relapse into recurrent episodes of depression over time. So there's just no data to suggest that these are helpful medicines from the depressive phase of illness. Moreover, these are probably the drugs that are really gonna increase rates of tardive dyskinesia in this patient population. So I think I would really try to work with hospital administrators and insurance groups and just say, it's 2023. We have to do better for our patients in this regard. If you are suggesting that you shift from the first generation to the second generation. No, adding. I'm saying adding. So. Well, whatever you suggest. You have to come to me. Our hospital will kind of get, we get a check saying, why is this person on two antipsychotics? I think in this case, I would say, I would like to try to transition them off of the first generation. And I'm gonna do it carefully. But the second generations are where the evidence base is, right? So you would get rid of the Haldol? I absolutely would do so. Okay. So you wouldn't supplement it, you would get rid of it? I would. And the Prolixin? Absolutely. Okay. All right. So now I'll take an online question. Sorry, we are above the time, but we are here to answer your question. So this is a question for Dr. McElroy. What are your thoughts about Volproate and polycystic ovarian syndrome and weight gain, especially in young females? And the second question to piggyback on that, would you recommend to screen for P-O-L-G mutation before starting Volproate? Okay. Well, the second one, the second question's easy. No. If somebody has a P-O-L-G mutation, it's usually pretty evident. It's a mitochondrial disorder. Those things are evident very early in life. I guess, unless you're a pediatric bipolar specialist. Now, okay. I was hoping maybe this would not get brought up because I don't have a good answer. But I have to admit, it has affected my practice. Okay, Volproate absolutely causes weight gain. There's no doubt about it. It probably causes, of the major mood stabilizers, Volproate definitely causes more weight gain than lithium, than carb, and then lamotrigine, which we know is actually associated with a little bit of weight loss. Now, the whole polycystic ovary situation, I have to admit, I think it's still, I don't think we really know. But because there's a possibility and because you get weight gain and because the drug is a teratogen, I tend not to use Volproate in fertile females, especially young women. Because I'm much more, I'm gonna use lithium and an atypical unless they fail everything and we have to go to Depakote. And then if I'm gonna have to do that in a young fertile woman, I really, I like to talk with her GYN, make sure she's on birth control. Does she have PCO or not? It's an important issue. I wish we had it better resolved, but I'll be honest with you, it's affected my practice. I use, I'll use Depakote in men and post-menopausal women with no qualms. But with fertile women, I'm very careful. And then, but the whole thing with, then we could talk about PCO because that also is associated with psychiatric disorders, including obesity and binge eating. So you just, if you've got a, if you have a young woman who's overweight, the last thing you wanna do is put them on a mood stabilizer that causes weight gain. Discontinuation of treatment for weight gain is an enormous, it's much more common than people realize, so. Can I go ahead? Sure. Hi, Lee, or? Are we switching mics? Sure, go ahead. Yeah, we'll take both the questions. All right. So I'm a PGY-2, I'm from Cleveland. So when I started learning, I mean, when I got into residency, it seemed like I was really enthusiastic to learn about STALs and like the neuropathways, the specific neuropathways, so the serotoninergic pathway, the dopaminergic pathway, and how it translates to clinical practice. The more and more I realized, even going through the text, it seems like, it doesn't seem like we have a clear answer as to how each mechanism works independent of the other. They're all interfering with each other's mechanism. Similarly, we're now learning about TAR1 and different molecular mechanisms of treating these conditions. And clinically, I've learned that CLAWS, you know, the most effective disease medications are probably the dirtiest ones. So CLAWSipine for schizophrenia and Serocol for bipolar depression. So my question is, overall, in your assessment, is there a utility for like targeted treatment in these complex conditions? And if not, where are we in that stage? Because we have a lot of the genes coming out. We go through the textbook. It may give us a sense that we're far, way far ahead than we actually are. So is that a, it's more of a question as to whether that's a fair assessment or understanding of the situation right now? Yeah, so great question. I don't think there is any sort of a biomarker right now, per se, but kind of looking at the disease state, you know, if you have a patient with a, let's say, depressive predominant polarity or a manic predominant polarity, so you could look, think about, you know, if you're thinking about an atypical and a psychotic. Seroquel is one of those, or Cotiopine is one of the medication which has, you know, data for bipolar depression, bipolar mania, or mixed symptoms. So, you know, usually when I'm teaching resident, if you don't know anything and someone asks you a question about someone with bipolar disorder, pick Cotiopine. It will target all of those. The question is tricky because there are other comorbid conditions. So if someone has obesity or someone has, you know, someone, let's say they have depression but high anxiety, Cotiopine is a good option. But if someone already has a higher BMI, Cotiopine has a higher weight gain potential as compared to if someone has a depression predominant polarity to their illness, then maybe Lorazadone is more weight neutral. Keroprazine is a little bit more weight neutral. Lumetapron is a little bit weight neutral. Versus if you have someone who's a treatment resistant bipolar depression, I think that's, or bipolar with mania or agitation, I think that's the group where you think about maybe, you know, if they're not responding to Cotiopine and all, then you think about Clozapine. So we use it, but we use it not as commonly because of the higher side effect burden. It's a good drug. So it's a, from a biomarker standpoint, I don't think there's any biomarker in particular which could predict the response to the medication per se. The pharmacogenomic testing can tell us how people metabolize the medication if they're poor metabolized or ultra rapid. So I think there's a role there, but I don't think those biomarkers can predict who'll respond. So that's how the companies are selling those. And I would only add, your comments are completely valid. The problem is, is we've started thinking about a drug, it's a serotonin drug, let's look at serotonergic pathways and see how this works. There's still probably a place for that, but I think the types of collaborative studies that are going forward are turning that completely upside down. And it's identifying a disease risk gene or a neural network that we think is fairly disease specific and then starting there and kind of reverse engineering that or flipping it upside down. For example, so we have some genetic variants that are either associated with the disease or treatment response. We're now actually studying those variants and looking at them functionally in different types of models. That's probably gonna be a more successful way to say, here's a gene, it acts like this, it responds to this type of treatment. Who are the sorts of participants that would respond to that and how do we get this drug to them far faster than starting with a drug and figuring out who it works with? That's the problem with trying to glean pearls from antidepressant trials and bipolar disorders, they started from the wrong point. Okay, thank you. Thank you. Hey, I have actually taken lithium for 20 years and I have kidney disease. And I've got three clients in my practice who have kidney disease take lithium. Lithium is very important to their recovery, we watch it. Dr. McElroy pointed out a few minutes ago, well, was anything else going on? Well, in my own case, actually six different things caused my kidney disease. And one of them is stratospheric copeptin. I mean, who even tests copeptin? I do, and I've been watching it and the trend for like the past five years and I finally found an endocrinologist who wants to work on it because my nephrologist couldn't have cared less. So testing is really important, like investigative, exciting people to work with is really important. There's a disparity in that that's not available to everyone which is like super sad, but I really advocate for that for my peeps. There's one thing that I've never been clear on that I wanna know your opinion on, which is I've heard some people say, use the CR form of lithium because you're not gonna get the peaks and that's gonna be easier on the kidney. And then I've heard other people say, use the immediate release form of lithium because the kidney will get a little break during the day. What do you guys advocate? So your first point is of course fascinating and yes, any other comorbid disorder that might predispose to renal problems, whatever it is, whether it's hypertension, diabetes or others, certainly those need to be looked at. The answer to the second question, there are some early studies showing that what the kidneys need to stay okay is a low trough level. That's really more the issue. How do you get the lowest trough level? The answer is immediate release lithium once daily. 23 and a half hours later, you have a pretty low lithium level. If you take TID lithium of sustained release, you have a nice even curve, but there's always a bunch of lithium coming through the kidneys. So probably once daily of regular lithium capsules is the safest way. Okay. Can I ask a second question or do you want me to wait? Us? Go ahead. Yeah, I think we're about time. I refuse to give antipsychotics, new ones, old ones, to anybody for longer than three to six months because of not only the 20% risk for tardive dyskinesia, but also the very well established trend of cortical thinning. We have so much more disability now in our bipolar and schizophrenia populations than we did 20 years ago before literally 60% of people with bipolar disorder were on these drugs. So how do I manage keeping people well, and I have a good track record, without using these medicines? I use nemodipine, varapamil, allopurinol, aptium, why isn't this being discussed? So the purpose of this symposium was to really review. Well, not here, but I mean in our, I'm sorry, let me qualify that. Why isn't this being discussed at any meeting? I attend like, I don't know how many meetings, like three to five per year. I never hear this stuff. So the calcium channel blocker story, very interesting. The challenge is when those early study, and mechanistically they make sense, but those larger scale studies did not separate from placebo. Now you can, and so what, and so another example would be modafinil. Our group was very interested in looking at that compound. Placebo-controlled trial, very positive with three sites. Larger initiative with 35 sites, it doesn't separate. Company says we can't afford to go farther, we're stopping this program. That happened with many of the calcium channel blockers. You could argue maybe we need to go back to those studies and figure out for whom were they helpful, and figure out why. I don't think the allopurinol ever really got to that level, but I think what you're emphasizing is where the field needs to go. Even if their failed studies are negative, if there are individuals who've done very well, isn't it our responsibility to try to understand why and how that might be, and can that be scaled up for a particular type of patient population? We could look at the CACNA1C polymorphism, for example. Just one quick addition. One of the reasons those drugs are not being investigated is that they're all generic, and the NIMH is not interested in clinically relevant thing, they're interested in hypothesis-driven issues. So that's issue one. And if one would use calcium channel blockers, as you said, verapamil, which is the most commonly used and the most investigated, may not cross the blood-brain barrier very well. So it should be nemotipine or aziratopine as the two calcium channel blockers that we know cross into the brain. It's too bad we don't have more studies on those. I'll say we tried to launch a study looking at this with Bob Pope, and it was gonna be $1.8 million to get the drug, and so we couldn't afford to do it. Very unfortunate. I'll take the last question, and then we'll end the session. We are here to answer your questions. So this is the best question, one of the best. So what is the best maintenance treatment for bipolar disorder? What is the best treatment? Best maintenance treatment for bipolar disorder. I think the most senior clinicians should answer that. Do you wanna take my social security payments? Only about 20% of people of bipolars have this light bulb response on lithium, where it is just astoundingly effective. For those 20%, lithium is clearly the drug of choice. If I had bipolar II disorder, I would wanna go on lamotrigine. As was discussed, it's better against depressions than manias, and of course, depressions are the dominant feature, especially in bipolar II disorder. So I don't think there's one drug. I am, again, I've been doing this for a long enough time. I am so grateful that we have so many options. So if our patient doesn't respond to A, I have B and C and D and E to think about. That to me, I started when lithium was the only mood stabilizer around, and it is so much better now to have options, and I'm thrilled about that. Do you wanna add anything to that? Okay. Thank you, we'll move on. Thank you.

Bipolar Disorder

Biomarkers

Lithium

Renal Effects

Genetic Markers

Lamotrigine

Unipolar Depression

Drug Interactions

Valproate

Acute Mania

Antidepressants

Personalized Medicine

Clinical Practice

Treatment Response