Chapters Transcript Keynote Address: Updates on Treating Patients with Bipolar Disorder Course: 2025 Annual NYU Langone Health Annual Psychopharmacology Review Thanks so much, Lynn. Uh, thanks, Charlie. Um, thanks for inviting me to come back here. Um, I was a resident here from, uh, 1980 to 1984, and that completely shaped everything that I did. We dealt with really, really ill people and, and, um, fundamentally. And this would be related to the talk. What I started to become interested in and still interested in, and I hope you're interested in, is epistemology. How do you know what you know? And what is the Evidence that you have in your head that you use to make the best decisions that you can for your patients. And that has been really a a lifelong. Quest that I don't know if I'll ever be finished with, but what, what's remarkable and what I wanna also mention before we go into some of the details here is that, uh. Charlie has been willing to take a risk. Um, along with, uh, with Danny Oeffecu, and we are conspiring. Literally conspiring to try to change the system of care to radical collaboration across multiple health care systems and we're doing something called the Learning Health Network. It's actually called the bipolar Action Network and we are willing to actually share our outcomes. And in talking to Charlie, and it's also relevant for all of you. Do you know the outcomes of your patients? Not only do you know the outcomes of your patients, but what do you try to do to get better outcomes, and that is at the heart of what the bipolar Action Network is trying to do is can we get better outcomes? And I'll talk about that a little bit later. I have a lot of disclosures over time. So this is, this is the, the, the heart of one of the things that we face all the time. And that the major problem for our patients in terms of the burden, in terms of the duration, in terms of how much it really interferes with their life, is bipolar depression. And, and that's really what I'll try to focus on more than anything else. Um I don't know about you all, but I, I have 3 groups of patients. I have one group of patients, they're like, they just respond to stuff. They're OK. And then I have another group of patients who don't respond to anything. I mean, nothing, right? And you, you keep throwing things at them and somehow, uh, you know, things don't work. And then you have another group of patients who kind of wax and wane over time, in between, uh. And what, what, what I wanna get across to you is the epistemology of this is a lot of times we just don't know what to do. And that most patients are taking multiple treatments, and they have multiple problems. And there's, it's not like you can just look up and chat GPT and it'll tell you what to do. Right, it's just unclear. Um, given that, you know, of course we have some information, and we do the best that we can with the information that we have. And this is an oversimplified. Slide of the course of patience. Uh, from Ellen Frank and David Kupfer from quite a while ago. And I say it's oversimplified because it lacks a dimension of anxiety. And also the patients who have mania have depressive symptoms, the patients who have bipolar depression have manic symptoms. And it's not, they're not mutually exclusive. And the other challenge I have for you. Is to think about how do you actually assess your patients in your actual clinical practice? What do you have time to do? What can you do systematically? Do you take a narrative? Do you use any sort of instruments at all to be systematic? And in talking to a lot of people, we generally don't do that, right? We generally try to take a history, we have some sense of what the criteria are. And we try to follow people over time. And while that can work, it may not be optimal. And again, that's what we're trying to do, uh, in the bipolar Action Network is figure out how can we do this in a way that's reliable, valid, and practical, right? It, it, it's not easy to do that. And really the point of this slide is that if you do not assess the symptoms of mania in people who are depressed, you're gonna miss something. And if you don't measure the symptoms of depression in people who are manic, you're gonna miss something. Right, that that's really the point. And the other, the other dimension of anxiety is if you miss that, if you don't really treat people's anxiety and follow it, then they're gonna suffer in ways that have nothing to do with the mania or the depression, but actually the anxiety itself. And so, uh, this is from Eduard Vieta and colleagues, right, bipolar one, you can have this period of mixed features. But part of the point here, and, and also I, I won't talk about the math of this, is you cannot predict the next episode. It follows what's called nonlinear dynamics, and you cannot predict it. You will not know when the next one is going to come, um, but what you can do is, of course, give treatments that will reduce the probability, or at least extend it out as best as you can. With bipolar 2, there's a misconception that it's not as severe as bipolar 1. That's only for the mania. It turns out that depression is much worse. It's, it's harder to treat, it lasts longer, it's more of a proportion of what people have. And then if you look at, at the total time ill and the total time depressed, so when the red is the total time ill that people have. And the blue is the percentage of time depressed. That means. That's 70 to 80% of the time that people are ill, they're depressed. Right, but so that's really the great, the great burden of it. The other thing, if you look naturalistically, untreated, the depression will last longer than the manic phase, right? Overall. The other thing is you would think, you would hope that over the past 20 years or so, that, you know, we've gotten better at treating bipolar disorder and you'd hope that the results were better and the outcomes were better. It turns out they're not. And it's still highly recurrent in, um, and this is buried in the literature in the European neuropsychopharmacology and what Smedler and colleagues did. Was they looked at the course of Recurrence in bipolar disorder based on the number of prior episodes. So in, in Blue is one prior episode, orange 5, and so forth. So what you can see immediately is if you have more prior episodes, you get ill sooner. But what's really remarkable about this is that the X axis is logarithmic. Right, the X axis is logarithmic here. So, at 5 years, let's see, does this show on here? No, it doesn't. At, at 5 years, um, If you look at that on the left, what happens is that. Um, You have, oh, there it is. How do you like that? Did that work? Oh, so, so look, here's 5 years, right? You go up here and go just up here where it's 1 prior episode, that's about maybe 20% of people who are remaining well, well. That means 80% have already had some episode within 5 years. If you've had a lot of prior episodes, you've got about 2% who are doing well, which means 98% had another episode. This is 2019. It's not that, you know, even though it's before the pandemic and seems like it's 5000 years ago, it wasn't that that long ago, right? So we still have the problem of, of that. The other thing that, that's, that's curious and um I was lucky to be part of the study that uh uh Greg Reed did, um looked at how, what are we doing? What, what are we actually prescribing for our patients over time. And this is from 97, 98 until about uh 2015, 2016. And it won't surprise you, but here's the rise of the 2nd generation antipsychotics for bipolar disorder. And here is the drop of lithium from about 30% of the patients. To about 15%. We looked at our data at Mass General for the system. 13% of patients with a diagnosis of bipolar disorder are taking lithium. And um the other, the other thing is that if you look at any mood stabilizer, right? So the mood stabilizers, it's a funny category, right? But lithium. Valproate Carbamazepine, oxcarbazepine, right? I mean, that, that's sort of what's in that category. Um, we could argue what's a mood stabilizer or not, but look at that decrease, right, from 60% down to 30%. So the real question. With the rise of 2nd generation antipsychotics in our actual practice. Are we getting better outcomes? Right? And what are we trading off? Now, one weird thing at MGH Is that the residents have no idea how to use lithium. I, I mean, they don't prescribe lithium and, and the, the myth of lithium, it's dangerous, it's bad, it's gonna make your kidneys, uh, uh, collapse, and that the risk-benefit ratio doesn't favor it, except lithium is the only drug for bipolar disorder that's associated with a decrease in all-cause mortality. Right, so, so you're kind of, you have to weigh these things, right? It's a pain in the neck, it's a hassle, you have, you have to watch the, the levels, of course, you know all of that. But is it better for patients that they're getting the SGAs. And not getting lithium Is it better over time, right? And what is the real risk of getting kidney failure from lithium, right? And, and how can you manage that? But, but that's sort of the decisions that we have to make all the time. The other thing, which, which is um. Personally, I'll tell you why it's a it's a problem for me. Any antidepressant, 45%, 97, 98%, over 60%, 2015, 2016. Right, an antidepressant for. Bipolar disorder. Is that good? Is it bad? There is Virtually no evidence that a standard antidepressant works for bipolar depression. But nobody believes it. And everybody prescribes it. And if you give it with an antipsychotic, is that OK? And I'll, I'll show you a little bit later, uh, uh, some things on this. But we don't actually know. Right, and, and maybe for some subpopulations, it's worse because you'll make them go into rapid cycling, maybe not, maybe mixed states, who knows, right? And, and it goes back to the conspiracy with, with Charlie and Dan, right? About maybe we need to find out. Maybe we need to actually look at the outcomes of our patients, uh, put in a system to do that, share the data radically and transparently, and learn from each other and learn from what we're actually doing because we don't know. Now, the reason I, I take this personally is because in 2007, um, we published a study from. Um, this huge NIMH study, the systematic treatment enhancement program for bipolar disorder, and we publish it in the New England Journal, and we, we have, um, placebo added to a mood stabilizer versus paroxetine or bupropion added to a mood stabilizer and no matter how we tortured the data and we tortured it, believe me, we squeezed it, we put it on the rack, we stretched it out no matter what we did. Placebo did a little better, numerically. Which, which was like wild. We, we, we couldn't really believe it, but that's what it kind of showed. It also showed in those people who had treatment emergent mania prior, if they had a history of it, you kick them into mania, right, with it. So, so it's like, wait, wait a minute, we did all that work, and look what happened. Nothing, right? It had no effect on practice, which has made me a little cynical about what we do with the evidence that we have and also it's why we're doing the bipolar action network. Can we get better outcomes from actually doing stuff rather than just publishing papers. Now this was published in 2023 by uh my friend Lakshmi Yatham. And If you actually look at this, what they did was they treated people with antidepressants plus mood stabilizers. They got them better, and then they wanted to see what happened if they removed the antidepressant. If you dig into this, you'll, and if you read the, the abstract, it'll say there was no difference. Right? But Lakshmi says, actually, there is a difference, and people did better if they stayed on an antidepressant. It's just a statistical fluke that the New England Journal made them do. It, it had to do with like, when do you start the time, when do you count, um, and they counted before randomization, which was kind of weird in, in some ways. Anyway, The real point is we don't know. Right? The real point is that we have evidence against it. This is some evidence for it, for some people, and you know, you'll hear from Charlie later about, you know, precision psychiatry. Well, we don't, we don't have a way to do that with bipolar disorder. You cannot match treatment to patients based on biology. We are not there. We may get there, and, and Charlie's optimism may help drive us to get there, but we ain't there, right? We, we, we just gotta do what we can do in the best way we can do it with the, you know, what we got. So The, the simple story is, you know, OK. Maybe we should just know what the FDA approved treatments are, and you all know these, but I'm gonna try to go through some details about it that you might find interesting and maybe didn't actually know. But that's, that's the sum total of, uh, of the treatments for bipolar depression. That's it. That's, that's all we got. Um, and which is really remarkable if it's the burden of the disorder, right? Is that, is that this is all we really have. Um, so one of the things I want you to, I want you to notice. Two things in the data. That I'll show you. One thing is that everything works in about 50% of patients. It, it's all, it all hovers right around 50%. You've got some differences in side effects, you'll know that it'll be familiar to you in many ways. Um. But that means that if any one treatment is working 50% of the time, What do you do for the other 50%? Like, like, what do you do? What do you do if one thing doesn't work, uh, what do you choose next? Right, and, and there is a giant hole. We actually don't know that. So Dan and I are conspiring on a study that's called SMART BD which is a comparative effectiveness study for competing treatments for bipolar disorder. I'll tell you about that later. So this was the first, right, the first one that was approved, uh, Mauricio Tone, um, currently chair of New Mexico, formerly of Lilly, formerly of McLean Hospital. This was the first treatment to really show there was a difference, um, and that something worked. So, so, you know, here you've got your, your 50% or so. What you may be surprised about is that olanzapine didn't do that bad. Olanzapine alone for bipolar depression. It wasn't that bad. But none of us like olanzapine because it causes metabolic syndrome, right? We don't like that. But now we have GLP-1 agonists, other things to, to manage it, Sammy Dorfin, when, when it's added to it. So who knows what, how that'll change over time. Um. The other weird thing about this particular study is that the OFC group was much smaller than the other two groups, and I think it was just an artifact of what the heck they did there. The other thing I want you to notice across all the treatments that I'll tell you about, is that they have complex pharmacodynamic profiles. They hit a lot of targets. And it may be that you have to hit a lot of targets to get efficacy, and we don't know why. But I think what is clear. Is that you can't hit one target. One target is not gonna do it. And, and Dan and I have also put in a lot of grants over time to NIH and we were criticized and didn't get the grants. Because they said it hit too many targets. And they were interested in funding things that hit one target. And we tried to make the argument, you have to hit a lot of targets, but it didn't work. We tried, Dan, right? Uh The, the other thing here that, that, that many of you know, because of the metabolic syndrome, it's probably caused by the histaminergic and muscarinic effects more than anything else. That's what it's approved for. You all know that. The other thing that I want you to look at. And think about These studies are kind of weird. And, and they're kind of weird because You have different dropout rates, and one of the, one of the ways to look at the dropout rates is look at how many people dropped out of placebo. That's one thing to do. Uh, and, and even though I'm gonna show you in aggregate all of these studies for the FDA treatments, you can't quite really compare them directly. Um, and you'll see this here, right? It's, it's over 61% discontinued. From placebo. Just keep that in mind, OK? Just keep that number in mind. And this, this is probably. Why this happened. Right? It's like, well, if it's good enough for OFC must be good enough for any antidepressant and any antipsychotic. Can you generalize? And that's what we do, right? We, we do that. But uh you'll see if you dig into the literature, you won't find data that it works. But again, it doesn't mean it doesn't work. It probably means there's a subpopulation where it works quite well, but you don't know who that is. That's the problem, right? That's, that's the epistemological problem. You don't know. You absolutely don't know, which is why As you treat people, you may frequently be faced with uncertainty. Anybody faced with uncertainty when, when they're like, you know, with your patients, right, all the time, cause you don't quite know, right? You don't know. One of the other things which is aspirational for the bipolar Action Network. You see a patient, you put it in the system, and up comes 5000 patients just like yours with treatments and outcomes, right? That would be extraordinary if we can build that and we're working on it. Quetiapine 50%, whether you give a little or a lot. Right, about 50%. What is I think very interesting in all the years since this came out in 2005 by Joe Calabrese. They didn't study any dose below 300. So you don't know if a dose below 300 will work. Right, you just don't know. Um, You do know that. The typing will work for mania. Depression, psychosis, anxiety, insights of anything that bothers you, you can take a typing. Now you would think then that if that was actually true. There would be a large proportion of patients who would take quetiapine monotherapy. It's not true. If you look at the data, you actually look at what people are being treated with, tiny, tiny proportion take quetiapine monotherapy. So I wanna ask you all, how many of you have at least 10% of your patients on quetiapine monotherapy? Anybody? No, It's cause it doesn't work well enough. And remember, it works half the time. So if it works half the time, a lot of times you add something to it. Right? And, and sometimes you do things that you're not supposed to do, but it works. And sometimes you add one antipsychotic on top of another. That is like forbidden. You're not allowed to do that, but we do it anyway because sometimes you gotta do it because they're not responding and there's no data to do. What do you do if they're not responding? You don't know, right? Complex pharmacodynamic profile, you'll notice it, it's dopamine and these 5HT2A and so forth, and alpha, um, adrenergic, and again, because of its effects on histaminergic muscarinic receptors, you got the problem of also the metabolic syndrome. Um, you all know the, the side effects. But here I want you to look at the placebo discontinuation rate. Remember what it was before OFC? Does anybody remember? 60%, right? Here it's 40%. OK, that's still a lot of people dropping out. And you'll see this get dramatically lower with the other ones. Then you have lorazadone, and here too. About 50% Less is not necessarily worse than more. More is not necessarily better than less. Um, what do we do when people aren't responding to less, we give more. Does this show that more is not better than less? No, because it doesn't answer the question, if someone's not responding to less, do you have to give more? That study has not been done. So you actually don't know. Here again, people forget to tell people to eat with it, so it can get absorbed. If it doesn't get absorbed, it doesn't work. It's like if people don't take it, it doesn't work. If it doesn't get absorbed, it doesn't work. If you forget your patients to take it with food, it's not gonna get absorbed as well as it could. And here the side effects are, are different than the other ones, uh, with more akathisia, um, you know, it's, it's sort of like 30% of people get activated, 30% get. Get sedated and you can't tell who and then in the middle, you know, people tolerate it pretty well. but look at that discontinuation rate of placebo, right? So about 60% for OFC 40% for quetiapine, 6%. For placebo, that's interesting. For whatever reason they were able to keep people in the study. Then you have cariprazine, a little bit more, 50%. Some studies even suggest that 1.5% is better than 3.0. Who knows? What do we do when somebody doesn't respond to 1.5? We raise it up to 3.0. Does it work? Does anybody think it works? You don't know, because it's also complicated by time and time of exposure. So, so, it's like, how long do you keep somebody on 1.5? And then say it's not working, you make some determination, either the trajectory is not getting better or or something like that, and then you decide, well, maybe they need more. Right, um, again, we just don't really know. Again, it's, you start to see there's a similar pharmacodynamic profile, dopamine 5HD 1A 2A, but here, of course, no affinity. For those receptors associated with metabolic syndrome, so it's really not that much of a problem with lobrazodone. Uh, and it's interesting if you talk to people who have a lot of experience with lorazodone, it feels more like an antidepressant. Right. And, and then you've got stories for what they're worth, right? It's part of what we know of people get kicked into mania with lorazodone, which you think shouldn't happen. Have you ever seen that, Dan? Have you seen anybody kicked in? Yeah, low doses, right? Low dose of, right? It can kick him into mania. Um, it's the partial antagonism, right? Right, and aripiprazole also. Um, Yeah, so, um, Again, you know, again, this restless is a catesia, somnolence, kind of this mixture of things that people get or not. And here again, now you have 2.5% placebo, right? So, so the point is it's hard to compare this to OFC that has 60% dropout from placebo, right? So it's, it's hard to balance those things overall. And lumeperone. Fixed dose. Here they decided 42 is it. And if you read, you know, you say, oh, you can slam people with 42, and then you do it, and then people can't stand it. So you start at 21, so they can get used to it, and then you can get up to, to 42, which is prac it's just practical, right? Uh, uh, um, you can try to get away with slamming people with 42, but you're gonna lose some people from side effects from that. And again, your pharmacodynamic profile, um. And, and this one is really interesting is, is, is it starts to hit glutamate indirectly, um, through dopamine, right? And, and that's pretty interesting. You got usual side effects here also, but no weight gain, no metabolic syndrome. So What we, what we really don't know, right, if, if you give lumeperone, will you have better outcomes? Than if you give cariprazine, lorazadone, quetiapine, or OFC. Olanzapine fluoxetine combination. Do you, do you have some sense that one is stronger than the others? Do you have some sense, you know, if, if they don't do OK with this, when I go to that, they do better. Right? And, and here again is the big hole in our, in our epistemology is, I couldn't tell you what the outcomes of my patients are. I have no idea. I can give you some sense, but if, you know, if, if I had to go into the electronic medical record and find out like, well, how my patients, I would have no idea how to do it. And again, that's one of the things we're trying to fix, uh, because I, I think it's, it's, it's a huge problem in all of medicine. So, so this is the overall response rate, right? 50, it hovers right around, right around 50%, you know, more or less, but you can't make anything about that difference of more or less. You can't say one is better than the other cause they haven't been directly compared. If you look at weight gain, not surprising, OFC has the biggest proportion of people who have greater than 7% increase in weight. As compared to the other ones. I thought cataipine would be higher, but again, you know, this is an indirect comparison, they're not direct. So, see, so take this with a grain of salt. In other words, don't believe everything I say. Maybe not be a good idea. This I was really surprised about is that I would have thought OFC was more sedating than quetiapine. Again, this is indirect, but the proportion of patients who have whopping sedation seems higher, at least in this, although, you know, again, you just don't know. And here's one of the most interesting drugs. It's interesting for a lot of reasons, but Even though it's not approved for the acute treatment of bipolar depression, and there've been five trials, of which one is positive, 1999, Joe Calabrese was, was clearly positive, and 4 other, it couldn't separate out from placebo. But you've got a methodologic problem here of when do you start the clock, right? So, in order to avoid Stevens-Johnson syndrome and toxic epidermal necrosis, you start low and you go slow. So when do you start the clock, right? If, if you're comparing it to placebo and it takes you 8 weeks to get up to 200 mg or 25 mg a week, if you're really conservative, right? Then, then, like, when do you start the clock? Um, so it, it becomes a big methodologic problem. If you look at the meta-analysis by John Geddis, what it suggests is that in those 5 trials, you might separate out from placebo for the more severely depressed patients. There, there may be a little bit of a more thing, which is why some clinicians really like to use this acutely. And then, um, and again, it has a complex pharmacodynamic profile, but, but here, you know, it inhibits glutamate release and it decreases glutamate transmission in the dentate gyrus. So Yeah, interesting stuff, right? And, and have you, has anybody seen a Stevens-Johnson syndrome in their patients? Anybody seen it? Yeah. So it's, um, it doesn't happen often, but it does happen, right? You have to watch for the rash that's very characteristic with, what's the word bulla, B U L A E, right? It's, it's like a pustule that people can have with a red target around it, um, also in their mucous membranes, that's what you wanna watch out for. But if you start really low and you go really slow, it's not gonna happen that often, but people freak out if there's a general rash, right? And, and then you have to send him to the dermatologist and worry about that. And this is buried in the literature from 2009. Um, this is Van der Luo and also Eduard Vieta. Uh, they did an interesting little study where it was lithium plus placebo versus lithium plus lamotrigine. And the combination actually worked, you know, it, it's separated out. Maybe that's what we're doing clinically, maybe that's what we've discovered can help. And there's also the, the odd. Indication for lamotrigine, which is you can edit at any time, because it can decrease the probability of having future events. Although also if you read the literature carefully, it suggests that lamotrigine is more protective for future depression than future mania, but even that's weird, because if you look at the studies, The proportion who got mania on placebo was very small, so you couldn't detect a difference cause they may not have had it. So who knows? Maybe, maybe that's true. So, the antidepressants, right? It's used widely. We don't know if it hurts, we don't know if it helps. You, you know, who has any idea. But, but, um, one of the things that we do, and it's, this is in all of medicine. If you do something and something happens, you assume that something happens because you did it. Right, you did something before, something came after it happened, there's cause and effect. You give an antidepressant, somebody gets better, you'll be convinced it's because you gave them the antidepressant. But people fluctuate with with bipolar depression, right? The, the, the, the if the episode is gonna end anyway, and you do something before that, you'll assume it's because you gave it. This is just a fundamental. Reality of, of, again, what we're doing here. And as I said earlier, the antidepressants persist. But here's a recent meta-analysis that I found, uh, deep in psychiatry research 2022 by Hug and colleagues. And basically, what this meta-analysis says is this really, if you look at all the literature on an antidepressant versus placebo for bipolar depression, for people taking other stuff that's making so they don't get manic, there's not a big difference. But as, as Dan and I think, I think would agree, they just seem to seem some some people who need it. You know, you just, you just can't tell overall. And that gets back to Lakshmi Yatham's, uh, interesting study where if they do respond, and if you look at the data carefully, it looks like they do better if you keep them on it long term, right, which is at odds of what we're saying otherwise. What about lithium for bipolar depression? Well, Sue McIlroy did probably the most modern study of this in Embolden One, which was actually looking at quetiapine. And they were using lithium as a comparator. Well, lithium didn't separate out from placebo at all in, in this, and this was a large study. This was not small, um, which is kind of interesting. You, you think, maybe not, um. The, the other thing I just wanna mention is, I have not been a big fan of Valproate for a long time. Um, it can cause, um, polycystic ovarian syndrome. Um, there's recent data from Europe that it can also cause other reproductive problems in men, um, and that the, the side effects can be pretty bad. Um, so, um, we just published an editorial basically saying you have to really rethink. Or the value of valproate. Also, there's a study in the literature, uh, called Balance that was done by John Geddis, and there they gave people lithium plus valproate. Got them better Randomize them to stay on the combination. Go on lithium alone or valproate alone. Valprate alone looked the worst. Um, the valproate plus lithium looked the same as lithium. It wasn't any better, but the valproate did not look good in that particular comparison. But I recently looked in the literature. You know, there's not, there's not unreasonable data that Valprate might work for bipolar depression. Uh, which I was surprised about. Um, so maybe it's something to, for me to rethink and reevaluate, but also for, you know, for us to look at the real-world data of, like, you know, what's actually out there. So these are other things that, that, uh, you know, we can turn to for people who are just not responding the way we would like them to, um. And I can go over a couple of these things, and then I'll tell you some strange things that if you're really desperate, there's a little bit of evidence that maybe you can think about these things. So Carlos Eroti, NIH, um, sort of the first study to look at, at ketamine for bipolar depression. This is one infusion, right? So people get better and, and then it can last actually for days, uh, probably because of norketamine, um. Uh, ketamine and S-ketamine for bipolar depression. So Santucci colleagues just published this last year with a response, response and remission. Um, 39%, not that bad, 13%, you know, it, it's. It's important that we don't overestimate how well people can do with ketamine, right? There are people who just don't do well with ketamine, right, just, just happens. And then there's this funny French study. And the funny French study was they were looking at remission from suicidal ideation. And they were looking at um 0.5 mg per kilogram added to the current treatment, 2 infusions, 24 hours apart, and um they were looking at remission at day three. Overall, right, placebo versus ketamine, ketamine did better. Except all of the effect was in bipolar. All of the effect was here and not here. Which is kind of strange. So who knows? This is a really important study from over a decade ago, where ECT was better than the best guideline pharmacotherapy you can get. Almost twice as better in terms of, of ECT. The problem, as you all know with ECT is what do you do if somebody responds to ECT? Do you give them maintenance CCT? Do you switch them over to something else? What do you do? Right? And, and this is what I find so interesting and what we have to face because we face these decisional dilemmas all the time, right? You gotta figure out what the heck to do. But if you know somebody's desperate, ECT. Now, do you give ECT or ketamine for bipolar depression? That study hasn't been done, but it has been done for unipolar depression, right? Where they look about the same and maybe ECT is a little bit better if I remember correctly. Repetitive transcranial magnetic stimulation for bipolar depression. First of all, it's not approved, so it's hard to get coverage for it. But the data say basically it works and my friends who are real experts in TMS thinks it works even better in bipolar depression than regular depression. And then there's this weird stuff, right? The weird list of, of things you might be able to turn to, right? Pramipexole, which is a D2, D3 agonist that we steal from the neurologists for Parkinson's disease. Um, the late Jan Fawcett, who really favored using monoamine oxidase inhibitors, um, Before he died, switched over to pramipexole for his patients because he thought it worked better. The, the pramipexole, has anybody played with pramipexole at all, right? It's, it's, um, you have to watch out for nausea, so you have to start really low, like a 0.25 of a milligram. Try to get to a target of anywhere between 3 to 5 mg. Um, it can cause somnolence, and the weirdest side effect, the strangest side effect, which I have seen with pramipexole is gambling. And it changes your, your reward system, and it makes people who never gambled before, have a gambling problem. It's just something to watch, right? It doesn't happen that often, but, but it, it can happen. Um, Has anybody given anybody pioglitazone at all? A pioglitazone is an anti-diabetic drug that affects the um paroxysome proliferator-activated receptor system, which I won't torture you with. But there are 3 studies in the literature that suggest that it might work for bipolar depression. There's a 4th 1, which I don't believe because it comes from a non-reputable group in Iran. Right, I mean, these, these guys in Iran, they'll publish studies and the curves look the same in every single study, like, gee, what's going on there? Yeah, ADHD too, right? Yeah, it's an interesting group. Um, the, the, uh, the study that Dan and I didn't get was in minocycline, uh, which is an interesting antibiotic, uh, that also has interesting effects on glutamate, um. And we couldn't get funding for that. There is one study that was done subsequently that didn't show that it worked, that came out of England, um, and another study said maybe it works. And acetylcysteine, has anybody given people or recommended they take a acetylcysteine NA? Yeah, NA is NA is really interesting, um. NC can, um, it's not only anti-inflammatory, but it, it affects glutamate in a funny way. So there's an anti-porter, right? It's, it's like a revolving door, and basically, uh, you put two cysteines together, you remove the sulfhydrol groups, it causes cystine. Cystine will transport in and it'll push out glutamate extra synaptically. And that then dampens down through negative feedback. The problem with acetylcysteine is that it's hard to get it into your brain. And, you know, the penetrance is 16, 17%. And there's a trick, and the trick is to add probenecid, and probenecid is an ancient drug to get antibiotics across the blood-brain barrier. And there are Two studies in the literature where they were able to get an acetylcysteine into the kids who had traumatic brain injury with probenecid, and it actually can treat traumatic brain injury if you give it soon enough. Um, so if you wanna give it, give it with probenecid, it's like 500 BID. Pimvivanserin is really interesting, right? It's, it's an antipsychotic approved only for the psychosis associated with Parkinson's, end-stage Parkinson's, and it doesn't touch dopamine. Right? Doesn't touch it. And there's one case series in the literature from um uh Al Melik, um, who suggests maybe it can work. But again, there's no data on hard to get hold of it. Epsilon is not on the market, but it may be a lithium analog, and again, we don't know if bupropion dextromethorphan actually works. So I hope I gave you a couple of basics, maybe some new stuff, uh, went over the FDA approved treatments. Mixed stuff about antidepressants and other treatments, and I hope I was able to. Provoke you a little bit with the reality of the limits of our epistemology, with the fact that as clinicians, we all face uncertainty, uh, we do the best that we can, uh, that, uh, Dan and I are trying to, trying to fix that with a Smart BD, um, uh, which is, um. Uh, the sequence multiple assignment randomized trial for bipolar depression, and what we're gonna do in this study is compare. The combination of escitalopram and aripiprazole, cause that's what people are doing, versus quetiapine, lorazadone. And cariprazine. Treat people for 6 weeks, they could be taking anything else, we don't care. It's a pragmatic study. Pragmatic studies like real world, and if they're not doing well, we re-randomize them. Now that causes all sorts of interesting things to happen, especially for the people get re-randomized. There are 12 pathways. It's 4 x 3. Um, so it's really complex statistics to actually compare all of that. We'll follow people up to 52 weeks, see what happens to them naturalistically, um, funded by the Patient Centered Outcomes Research Institute, which was funded under the American, the ACA and is actually funded by a tax on the insurance companies. We'll see if it survives this interesting time that we live in. Uh, we'll see what happens. Uh, and, and again, I, I just wanna emphasize, um, Charlie's courage and in in in supporting the bipolar Action Network, uh, because we really aim to have this be a model for how collectively we can improve the outcomes for people with serious mental illness, um, and, and expand it out so we can make it so that every clinical encounter adds to our knowledge and that we collaborate radically. To figure out what's causing people to have better outcomes and, and what do we do for the people we don't know what to do? How can we study it? How can we look at it, and how can we aggregate those data together. So thank you very much for, for your attention and Charlie for inviting me back. Published March 1, 2025 Created by
CME Knowledge Hub