My name is Dr. Ron Winchell, moderator and my co-moderator, Dr. Kathy Krohn. Welcoming you to the first session of the last day of the annual meeting. Thank you for sticking it out. And we hope we see you the rest of the afternoon still today. Dr. Michael Gitlin is Distinguished Professor of Clinical Psychiatry at the University of California, Los Angeles School of Medicine. He is currently the Director of the Adult Division of the Department of Psychiatry and Director of the Mood Disorders Clinic. He is the author of many scientific articles and book chapters, as well as five books, including The Essential Guide to Lithium Treatment with Michael Bauer. Among his awards are Distinguished Educator Award in Teaching from the UCLA Department of Psychiatry and the Leonard Tow Humanism and Medicine Award. I'm going to add a personal note to this introduction. Those of you who may have been present or know about the turmoil in the 1970s when departments of psychiatry across the country were shifting from a primarily psychodynamic model to a psychobiologic model, remember it was a period of great turmoil. And this did not happen equally at the same time across the entire country. I was at the time a medical student planning to enter psychiatry at a major East Coast medical center that will remain unnamed at the moment, which still had not yet made that transition. In my second year, the Director of Residency Training decided to have a meeting for medical students interested in psychiatry. When talking about various programs, he referred to UCLA as anathema. The next day I signed up for a rotation at anathema, which I guess is the A in UCLA. When I arrived, I was taken under the wing by a senior resident who really shepherded me with warmth, training, patience, and education, and possibly one of the reasons, possibly to your regret, that I'm standing here today. His name was Dr. Michael Gitlin. The next time I encountered him was several months ago as we were coming up with names for people we would invite to give this presentation. I said, oh, that name sounds familiar. When we traded emails, we found indeed it was the same, Dr. Michael Gitlin. A more recent related event, several days ago, some of you may have been present at our presentation with Dr. Stahl on MAO inhibitors. It's our goal to get people to become familiar with the use of MAO inhibitors because we're at risk for losing them. Many people are not getting trained in use of MAO inhibitors. After the question-answer period, a somewhat younger woman, certainly younger than I, came up and asked a question. I said, oh, you're using MAO inhibitors. She said, yeah. I said, were you trained in how to use them in training? She said, yes. I said, what program were you at? She didn't mention the institution, but a big smile came across her face and said, Dr. Michael Gitlin taught us everything. So many, many years of being a distinguished teacher, I want to thank you personally, and I actually would ask the audience to join me in a round of applause for a wonderful educator for decades. Thank you. That was very kind. I even have two slides about MAO inhibitors for bipolar depression, not knowing that would come up. So what I'd like to do is spend the next hour providing an overview of bipolar depression and how to treat it. We're going to talk about, of course, mood stabilizers, second-generation antipsychotics, a large chunk on antidepressants, both about efficacy and switch rates, and then we'll have probably about 25 minutes for Q&A because the treatment of bipolar depression, for any of us who work with mood disorders, this is something we have to deal with pretty well every day, and the literature is just dreadful. I mean, it's wonderful to talk about evidence-based medicine, but to do that, you need evidence, and we don't have a whole lot of evidence. So first, even though, of course, mania or hypomania is the defining pole of bipolar disorder, it's what distinguishes it, of course, from major depression, the dominant pole is really depression. This is the five studies that have looked at it. These are both treated and non-treated people. The bottom line is the aggregate of the five studies is showing that the average bipolar patient spends three times as much time depressed as they do manic or hypomanic. This is even true about bipolar I's and conflicting data, but it may be even more true with bipolar II's. So depression is the dominant pole, even though there are far more studies on mania than there are on depression. Many of our colleagues obsess about the differences, the clinical differences between major depression, depression, and bipolar depression, but the fact is the similarities far outweigh the differences. Yes, there are some differences. Bipolar depressed people have more slowdown features, more hypersomnia, more fatigue, more apathy, psychomotor retardation, they're more likely to be psychotic, et cetera, but any feature you can see in bipolar depression, you can see in unipolar depression. So there are no pathognomonic features, and I'm not sure it's fabulously helpful to obsess about the differences when the similarities far outweigh them. When we treat regular old depression, major unipolar depression, all we really worry about, by and large, is does it work or does it not? You're not really worried about antidepressants making unipolar depressed people worse, but, of course, in bipolar disorder, bipolar depression, it's a little trickier. It's tricky because of the three factors on this slide. Efficacy, of course, as would be for any treatment for any disorder, but the second is the rate of switching, and this is obviously referring to antidepressants more than anything else. TEAS is the formal term in our field, treatment emergent affective switch. Everyone calls it switching, of course, meaning you give somebody an antidepressant and they overshoot the mark and they become hypomanic or manic, but the third or the second of the negative consequences of treating with potentially antidepressants, there are many different terms for it. It's classically talked about as triggering rapid cycling, which is a period of mood instability. It could be formal episodes with full time duration or it could be just mood instability. Later I'll show you some extraordinary data from one study about that. And again, there are many different terms for it. Rapid cycling, affective instability, or Terry Ketter, when he was involved in this area, called it the roughening of the course, which is, I think, absolutely the most poetic form of talking about affective instability. So what I want to do now is go over both acute treatment and preventive treatment of all the major options that are available to us. So first is acute treatment, treatment of acute bipolar depression with mood stabilizers. And it turns out, even though we've been doing this for well over half a century, again, we know way less than we should or even than many of us think we do. First, lithium, of course, is our oldest drug in bipolar disorder and there's some evidence that lithium has acute antidepressant efficacy in bipolar depression. But if you try to look for some RCTs, randomized controlled trials, you're going to be looking with not much success. It is startling how little data there are for lithium, our oldest drug, in acute bipolar depression. We don't have to worry about it making people worse, of course. We don't have enough good studies comparing lithium for acute bipolar depression to antidepressants for acute bipolar depression. Europeans and Australians and New Zealanders use it way more than Americans do. If you look at... There are all sorts of studies about that. They are way more lithium-centric than we are. And, by the way, they look down on Americans for that. It's like, you know, we're little children distracted by bright, shiny new things, new drugs that come out from pharmaceutical firms. That's really what they think of us. Carbamazepine, Tegretol, when it was non-generic, there were just a few trivial little studies, no great data. Valproate, there are data. There are four small RCTs, double-blind studies, each of which shows, and an aggregate show, that valproate is more effective than placebo in treating acute bipolar depression. Why don't you use it much, and why don't I use it much? Two reasons. One, increasingly we've learned, of course, that valproate is a very tricky medicine for women who are fertile. Between age whatever and about 50, because of the potential disastrous teratogenicity, it's a little tricky, and bipolar disorder is, of course, a disorder of relatively young people. The other reason is, by the time these four studies... Actually, I show you the four studies on the next slide. By the time these studies came out, divalproate, Depakote, was already on its way to losing its patent, and so Abbott walked away from it and never did a large-scale study. So all we have are these four small studies. Lamotrigine is one of the darlings of bipolar disorder. We know, as I'll show you in a few minutes, it's better at preventing lows than highs, which is unique among the mood stabilizers, but it does not have an indication for acute bipolar depression. There have been five double-blind studies, only one of which, and isn't it an amazing coincidence, that one was published, and the four negative studies weren't published. What a coincidence. Anyway, those four negative studies eventually came out in aggregate form, so one of five positive studies, that's why it didn't get an FDA indication. However... Oh, and gabapentin, there's nothing. Gabapentin's a good anxiolytic. It's not an antidepressant. And the same thing with Keppra, levatoracetam. Let me go to my Lamotrigine slides. There it is. Five studies, 1,000 patients. No four out of five were negative. However, in two different pooled analyses and meta-analyses, that's the John Geddes in the lower right and then Joe Calabrese, what they showed is if you add them together, which the FDA does not allow you to do, which is why that was never submitted, we find that there's a small evidence for efficacy. In bullet 2, a number to treat NNT of 11 is not clinically particularly relevant. It has to be 10 or under for it to be noticeable clinically. However, what Geddes did in his meta-analysis, or his pooled analysis, he then divided the group, the 1,000-plus patients, into those who were more depressed, Ham-D score over 24, versus less depressed. And as you can see on the slide, the patients who were more severely depressed had a drug-placebo difference that was pretty impressive. But if you look even more at it, this is again from the Geddes study, what you see is the difference is not efficacy. The difference is in placebo rates, response rates. So the reason the more severely ill people looked like they did better is they had lower placebo response rates, not that they had higher drug response rates. So do I think lamotrigine works for bipolar depression? I do. I actually use it pretty regularly, and I think it's effective with at least some people. But from a data point of view, it's really a bit shaky. There are other studies showing that lamotrigine does work. There are two that I'm going to show you. This is called the LAM-LIT study from Europe, in which you add lamotrigine to lithium and you find that people get more better. And then another study was another John Geddes study about adding lamotrigine to quetiapine, which we know has its own antidepressant effect, and it looks like lamotrigine augments the efficacy. So again, even though if you look just at randomized clinical trials, you'd say lamotrigine is a loser for acute bipolar depression, there's other evidence suggesting that it really does work, and I think it does work, and I'm sure many of you use it also. OK, second topic. Bipolar depression prevention, preventing depression with mood stabilizers specifically. And first, of course, is our old friend lithium. The last meta-analysis was done nine years ago by Emanuel Severus in Germany, and the reason there are only seven studies in this, and you'd think, well, there have to be more maintenance studies with lithium. There are, but some of the early studies with lithium had particular design flaws that you have to exclude if you're really going to do this right. What they did in the old days is they took people who are on lithium and then suddenly substituted placebo. And about 15 years later, we learned that when you do that, you have lithium discontinuation rebound manias. So the people who were switched to placebo had unusually high relapse rates. So those studies have to get excluded, and that was a bunch of the early studies. So there are seven good controlled studies, and there's no doubt that lithium is better than placebo at preventing any episode, but as you can see here, it's driven more by its ability to prevent mania than depression, as you can see on the slide. And that, by the way, is what our patients all tell us about lithium. They're sort of grateful because people are often ambivalent about their manias, but nobody's ambivalent about depression. Everybody with any taste hates depression. So they resent the fact that their manias were taken away and they're left still with recurrent depressions. And these data support what our patients tell us. Valproate, or in the form of divalprox, this was the first... This study from 23 years ago was the first study, the first modern non-lithium study, looking at mood stabilizers, and it was divalprox versus lithium versus placebo in patients who had all had a mania, and it turns out there was no difference between divalprox, lithium, and placebo, much to everyone's surprise. And the question is, why would that be? Most of us assume that divalprox actually does work. It does prevent mood episodes. And probably the reason this happened is in this study, first of all, the usual large dropout rate, but the second reason was to get into the study, this study, people had to be unusually stable, and therefore people did much better on placebo than you would have anticipated. High placebo response rates doom any active treatment in a controlled study. So most of us think valproate works, but from a data viewpoint, you can't... It doesn't... Those data don't exist. I was talking to a colleague of mine yesterday from Germany who said in Germany you can... In the EU, you can just about not... They're not allowed to prescribe divalprox for bipolar disorder as a maintenance treatment. It's just not what they do, but we do it certainly in the US. Third is lamotrigine, and I think most of you know lamotrigine has the FDA indication as a maintenance treatment for prevention of both episodes in general, but much better for depression. This is the two studies. You combine them. It's lithium versus lamotrigine versus placebo. And what you can see is the orange line is lamotrigine, and those people clearly did better than both placebo and somewhat better than lithium. As you can see, the p-value is unimpressive compared to lithium, but if you then look at a second study that came out of Denmark, it was, again, lamotrigine versus lithium, no placebo, and the two were identically effective at preventing agents, and in this study, just like in this study, lamotrigine was better at preventing lows, lithium was better at preventing highs. That's my clinical experience, and it's shocking and relieving when one's clinical experience actually agrees with the data in our field. It's rare, but every once in a while that happens. All right, so now I want to go to one of the more fun and controversial areas, which is the area of what is the place of antidepressants in treatment of bipolar depression. First, we're going to look at acute treatment, acute bipolar depression. There's this wonderful quote from Eduardo Vieira in Spain from an issue of bipolar disorders in 2008, and he said, we are still uncertain whether antidepressants are effective, ineffective, safe, or unsafe in bipolar depression. They might end up being the four things at the same time, depending on their class, particular pharmacodynamics, and concomitant therapy. An elegant sentence that says virtually nothing. I wondered whether Vieira was running for office when I saw this sentence, but in fact, he got it exactly right. That statement is about as perfect a one-sentence summary of what we know, really don't know, for antidepressants and bipolar depression as you can get. As my colleague Mark Fry says, the question should be, and I agree with this completely, shouldn't be do antidepressants work or not work. That's way too simplistic for clinical work and even for research. The question should be for which bipolar patients are antidepressants safe and effective, and for which patients should you avoid them, and we're going to talk about that. What plagues us in the bipolar depression antidepressant world is that there are very few RCTs. Again, embarrassing given how common this is as something that every clinician who treats mood disorders has to deal with on a regular rate. Shockingly, the placebo response rate in bipolar depression studies is unnervingly high. As you remember, we said a few minutes ago, high placebo response rates make it very difficult to then achieve drug placebo differences. And the third problem is underpowered studies. Most of these studies just have too small a number of subjects to really show efficacy if it's there. There have been a number, and I mean a number of meta-analyses, and I'm going to show you a little bit about them because it's important that you see that the meta-analyses disagree with each other, because, well, if there are only 5, 6, 7, 8 studies, how could the meta-analyses disagree with each other? And the answer is each one has slightly different inclusion and exclusion criteria, therefore they study slightly different numbers of studies, and they come up with disparate conclusions. Again, you know, as I say to my residents, if you really wanted certainty, you should have gone into pathology. This is the wrong field for that. So this is probably the most famous of the meta-analyses by Cedar McQueen from Canada, looking at six studies and showing, if you can look at the lower left, a p-value of 0.28. No evidence overall that antidepressants... This is just efficacy. This is response. Remember, response is a 50% improvement in whatever rating scale you use, the HAMD, the Madras, the QIDS, it doesn't matter. 50% improvement, no evidence of efficacy in these six studies. By the way, most of the subjects in these studies were on mood stabilizers. It turns out not all, but most were. They also analysed the same data for clinical remission. Remember, remission is that you're virtually all better, not just better. HAMD scores less than 7, etc. And here, too, you see no consistent evidence that antidepressants work. So you look at this, and the anti-antidepressant people, of which there are many in our field, look at this and say, see, antidepressants don't work. Or if you want to be more generous, you'd say the efficacy of antidepressants is unproven. I think that's a more generous interpretation. But there are other meta-analyses, as I alluded to a moment ago. Here's one from the group at McLean. Again, not looking at six studies, looking at whatever it is, nine or ten. And they find, again, because slightly different inclusion and exclusion criteria, that antidepressants do work some. This is, again, a meta-analysis not looking at individual studies. So, again, even the meta-analyses disagree with each other. And here's another one comparing the antidepressant effects for unipolar versus bipolar depression and finding the efficacy is pretty well the same. And the most recent good one is by McGirr in seven years ago, and he looked at 16 studies, 1,400 subjects, mostly bipolar 1s. All of these patients were on mood stabilizers, lithium anticonvulsants, second-generation antipsychotics. These are short-term studies, six to 26 weeks, a bunch of SSRIs, agomelatine, which is a melatonin agonist antidepressant available in Europe. It failed its registrational trials in the U.S., which is why it's never crossed over to the United States. And he, McGurr, like the Cedar and McQueen meta-analysis, found the same thing, no evidence for efficacy when you look at either response or remission. Okay, so that looks negative, but they did something very clever. They then dichotomized these, what, over 1,400 patients into those who were treated with mood stabilizers, meaning lithium or anticonvulsants, and those that were using antipsychotics, second-generation antipsychotics, as their mood stabilizer. And lo and behold, what they found is when antidepressants are added to second-generation antipsychotics, aripiprazole, quetiapine, olanzapine, et cetera, now you see that the antidepressants seem to work. As you see, the response remission rates, the NNT is seven for both, meaning that antidepressants work. Now, why would you see it if it adds to an antipsychotic and not to, let's say, lithium or an anticonvulsant? Well, obviously, one possibility is there's an additive antidepressant effect. We know that a number of the second-generation antipsychotics have their own antipsychotic efficacy, and maybe adding that efficacy to the antidepressant efficacy is what allows you then to see a signal, whereas lithium and the anticonvulsants may not work as well. Now, again, this is a post-hoc analysis, so nobody's done a random assignment study. There's been silence in the field after this was published, even though I think this is really intriguing results that we should be thinking about. So there's nothing new to add over these seven years. And I'm going to get back to the switch rate in a minute. And one of the most recent studies just wanted to, you know what, let me go on and I'll get back to this one. All right, so I want to, oh, no, let me talk about it now. So Nasir Ghami, who's been a leader in this field, has also talked about citalopram antidepressants. He does not think antidepressants are particularly effective. This is one of the most recent studies, so therefore it's not in the meta-analyses from about a year and a half ago on citalopram versus placebo. And you can see placebo is blue, citalopram is red, and technically there's no difference. But if you look at the left side of the slide, it sure looks like for the first six months or so, citalopram did better than placebo, but then that effect seems to wane. So for whatever reason, that's one of the most recent studies. And here are my two MAO inhibitor slides, not knowing that we would, that that would come in the introduction. So first, there were no classic randomized controlled trials on MAO inhibitors. There's always been this whisper in our field that MAOs have special efficacy for bipolar depression. Most of the collected case series are with tranylcypramine, PCP, or Parnate, for those of you who know. And it looks like it may work, although if you are going to be a hardcore data person, you would not be impressed by that, because we don't have any good RCTs. Most recently, about five, six months ago, Jay Amsterdam, who's done a huge amount of the work in bipolar depression at the University of Pennsylvania, he has this large mood clinic, and he compared the outcome of unipolar and bipolar depressions treated with MAO inhibitors. And he found, again, this is kind of just descriptive, looking at what he did in real life, it's not a controlled trial, but he found that MAOs were indeed more effective in bipolar depression than it was in unipolar depression. Again, fitting with the whispers of our age, even though, again, from hardcore RCTs, you can find it. And he also showed low switch rates. I'm going to get back to the switching issue in a little bit. Next, we have second generation antipsychotics. And I'm sure you all know that when, for those of us who are a little older and grew up in the first generation era of haloperidol and flufenazine and, you know, those guys, they were not, they had no particular antidepressant efficacy. All we did is give a lot of people with mood disorders tardive dyskinesia, unfortunately. But then the second generation came along, risperidone, quetiapine, olanzapine, aripiprazole, and suddenly we see that these medicines have both mood-stabilizing efficacy, adjunctive antidepressant efficacy in unipolar depression, and some of them have really good data on treating acute bipolar depression. So one of the first is this study. Some of you remember OFC or Symbiax, olanzapine-fluoxetine combination. I know it's shocking that both of those medicines were patented by the same company. What a coincidence. And they ended up in a combination pill. Anyway, this was their study where they compared olanzapine alone to placebo to olanzapine plus fluoxetine. The yellow line is the combination, OFC. And you can see that OFC clearly works, but you also see the pink bar, which is olanzapine alone, that olanzapine worked better than placebo. So that was the first one, the first study that we did. And we said, well, wait a minute, maybe there's something different about these second-generation antipsychotics. By the way, what's obviously missing in this design, where's the fluoxetine alone arm? How do you know that OFC didn't work just because it's fluoxetine? And our friends from Eli Lilly said the reason we didn't put in a fluoxetine alone arm is that it would be unethical to give antidepressant monotherapy to a group of bipolar ones. Remember that statement. I'm going to get back to it in a few minutes. So this showed that OFC works, whether it's driven by some unique combination effect or just the F. The fluoxetine is unclear. And then our friends from AstraZeneca really got into it, and they published five double-blind studies on quetiapine as an antidepressant, bipolar antidepressant. This is one of their classic studies, and what it showed, the red line is placebo. The two other, the yellow and green line, are 300 and 600 milligrams. What you can see is both doses are substantially and significantly more effective than placebo, and they don't differ from each other. All the quetiapine studies show exactly that, meaning that it is low-dose quetiapine that has the antidepressant effect, not high-dose. You should remember that because that's probably true about all the second-generation antipsychotics. At low doses is where you see the antidepressant effect. At higher doses is where you see antipsychotic effects, antimanic effects, et cetera. So, yes? Just for a moment, I'd like to say we are opening up the overflow room, but there are still a number of empty seats also here in the first two or three rows if people want to make themselves more comfortable. And please, go ahead. Thank you. And if you want to stand, that's okay, too. Anyway, so there was the quetiapine. Again, high-dose is not better than low-dose, which is a good thing because if you're going to use antipsychotics to treat bipolar depression, you don't need to go messing around to high-dose with obviously the chance of greater side effects, EPS, weight gain, whatever. This is another study looking at quetiapine. This one used paroxetine, Paxil, as a comparator. There was criticisms of this because the paroxetine dose was a fixed 20 milligrams. Paroxetine did not separate from placebo, but quetiapine was better than placebo here. By the way, look at the placebo response rates, 53 to 55%. Who are these subjects? I can't get 55% in my clinical patients using active drugs. So you wonder, when you have an unusual placebo rate, you should always wonder, who are the folks in these studies? And why are they not like the people I treat? Not that we take it personally. This is the Abilify, the aripiprazole study. There were two of them. Aripiprazole did not get the FDA indication for bipolar depression. Even though I do think it works, we all know it works well as an adjunct for unipolar depression. Why didn't it get the indication? Look at week six to eight in both of those studies. If they had designed a six-week study, Abilify would have had the FDA indication. And in week six through eight, they lost. You can see the drug placebo difference disappeared. When you submit data to the FDA, you have to tell them in advance how you're going to analyze the data. And unfortunately, Otsuka said, we're going to look at eight-week outcome. And that doomed them. The question is, why did the efficacy disappear between week six and eight? No one knows, but probably what many of us think the reason is, they used the wrong dose. They should have known, Otsuka, that in unipolar depression, we know the good doses of adjunctive aripiprazole are like 5 milligrams, 2.5 to 10. The mean dose of aripiprazole in these studies was 16 to 18 milligrams. I think they gave people too much. I think there was too much D2 blockade. And remember, aripiprazole has a three-day half-life, which means the level keeps going up. I suspect that by week six, there was just too much aripiprazole, too much D2 blockade, and they lost the efficacy, and they didn't get the indication. I do think it works at low dose. Litudo, lorazodone, also has the FDA indication, as most of you know. And here, too, we see the clear evidence that it's low dose. So if you look at the lorazodone doses, 20 to 60 versus 80 to 120, identical efficacy. So why would you use high dose when it's not better than low dose? You also get fewer side effects by using the low dose. And again, you see that in study after study, when you're going to use antipsychotics for bipolar depression, you should be playing in the low dose range, not the medium to high dose range. You save that for acute mania and psychosis. And again, we see the same thing with Vraylar, and our most recent friend in this area, Lumetepirone, Kaplita, has, again, the same data. So if we want to summarize where we are with second generation antipsychotics, for those of you interested, this is the best meta-analysis from about a year and a half ago. There are 18 RCTs, overwhelmingly more than any other medication class. And the reason for that must be obvious, because pharmaceutical firms had patented medicines, and they have the money to do these large scale studies. Lithium is unpatentable. How do you patent rocks in Chile and Bolivia? That's where Marlithium comes from. And so you end up with this large database on proprietary medicines, and pretty well, they work. Slightly different numbers to treat for each of them, but by and large, these are clearly effective medicines. If you look at others, there are a number of other medication classes. We don't have time to go over all of them. We'll do a few at the end. But one of the other well-researched groups of medicines are, as they're called, the vigils, modafinil and armodafinil, provigil and nuvigil. And remember, these are stimulants that work through orexin, not through dopamine, meaning they have a different cardiovascular profile, way less risk for abuse, et cetera. There are five RCTs on this, because again, when they were patented, they're now both generic. There was interest in whether these really would be effective bipolar antidepressants. The individual studies were negative, but if you add them up, as they've done here, you see evidence of mild efficacy. And then T's of 16 are nothing to write home about, but these are not switchergenic, so you should feel very comfortable if you needed to add something mildly stimulating to your bipolar patient. Let's say they were bipolar depressed and they really had trouble, as bipolar depressed people do, getting out of bed to get to work, go to school, do whatever they do. You know, a little modafinil in the morning will get you out of bed and get you to work so you don't get fired until hopefully you can treat them with other agents. And you can add modafinil and R-modafinil to anything, including MAO inhibitors, because it's not dopaminergic. There's no interaction whatsoever. So now we want to shift to switch data, because again, this is the much heat, little light in our field. So it's like a heavyweight match. On the left side, we have Fred Goodwin and Nasir Ghami, and on the right, we have Müller and Grunze from Germany. And again, everybody knows the same handful of studies, and yet we have passionate differences in how we interpret the data. So Goodwin and Ghami say, look, all antidepressants cause switching, so why would you use them? You know, you don't want to destabilize bipolars. They don't prevent suicide. And mood stabilizers like limotrigine, maybe lithium, work, so why would you take the chance of using antidepressants? And Müller and Grunze, looking at the exact same database, say, yeah, but wait a minute. Switch rates are lower with newer antidepressants, meaning everything from fluoxetine on, compared to tricyclics, let's say. And the switch rates are low enough that they are minimal to acceptable. We're going to go over that in a minute. We know that the proximate link to suicide is depression in 50% of coroner series. Therefore, if you treat the proximate cause, wouldn't that ultimately prevent suicide? And that there is evidence that antidepressants work. So again, we have this controversy in our field, which is always fun, given that there are only a handful of studies and everyone's read the same studies. In general, if you want to summarize a long literature and say, all right, well, who are the bipolar patients who are at most risk for switching? This slide shows you the accumulated wisdom of our field. Younger folks, bipolar 1s have about twice the switch rate of bipolar 2s, well studied. People who have unstable bipolar disorder, like rapid cyclers, tend to have much higher rates of switching. People who have mixed maniche features within their depressions have a higher rate of switching. There's some studies that show that a history of substance abuse, especially stimulant abuse, that may alter something in receptor physiology and make you more sensitive to switching when treated with antidepressants. And there's no doubt that tricyclics are far more switchogenic maybe than the vaccine. So if you're going to say which antidepressants are the least switchogenic, the answer is SSRIs and bupropion. And it is not correct. It is not true that there are data showing that Welbetrin has a lower switch rate than SSRIs. There are no data showing that. From what we know, those two, SSRIs and bupropion, have your lowest switch rate. But you know there are a lot of antidepressants that have never been studied. Cymbalta, Remeron, MSAM patches. We have no data on that for bipolar depression. Neither efficacy nor switching. As an example of the rapid cycling, this is again from one of Nassir Ghami's study, showing that the rapid cycling, basically what it shows is that the rapid cycling people have more switch rate, higher switch rates with antidepressants. All right. So remember I talked a minute ago about it's possible that Venlafaxine has a higher switch rate than SSRIs. Presumably because of its effect on norepinephrine. There are two studies showing it. This is the first one from our friend Vieta in Spain where he took a bunch of bipolar 1s and 2s on mood stabilizers and randomly assigned them to Paxil versus Effexor. As you can see, reasonable doses of both. Equal efficacy and a significantly higher switch rate. But if you look at the number of patients in this study, that's like 3 versus 1. Right? Or 4 versus 1. I mean when an N is 60, you know a switch rate of 3% means one patient. So that was intriguing but too small a study. Then there was a later study from a group called the Stanley Foundation found the exact same finding. That Venlafaxine had twice the switch rate of either an SSRI or Bupropion. So it's probably true that adding some norepinephrine into the mix in your antidepressant confers a somewhat higher switch rate. Remember when we were talking about some Symbiax OFC, I said that the design of the symbiax designers of that study said they couldn't use fluoxetine as monotherapy because it would be unethical. Well, our Paxil people didn't seem to find that because they did a monotherapy peroxetine study. So this is one of the quetiapine studies using Paxil as an active comparator. I showed you the efficacy data earlier. But these were bipolar 1s and 2s. Two-thirds bipolar 1s. And if you look at the very bottom in the middle, it says switch rate. It's a little too small. This is the switch rate during the study. And you can see the Seroquel quetiapine switch rate is 3%, very low. That makes sense, right? Because we know quetiapine would treat mania too and it should therefore prevent it. Peroxetine had 11% switch rate, but the placebo switch rate is 9%. So this means, and this is by the way the only study in our entire field on monotherapy with an antidepressant in bipolar depression. This is the entire world's literature. And the reason it's relevant is that there is no statistically significant difference between Paxil and placebo in switch rates in this short-term study, like an eight-week study. Now, am I saying you should be cavalier about giving antidepressant monotherapy to bipolar 1s? I'm not. But if you close your eyes and you think of the melody in our field, the melody is, oh my God, anytime you give an antidepressant to a bipolar 1, if they're not on a mood stabilizer, they're all going to switch. And this says that in an eight-week study, they're no more likely to switch than if you gave them placebo. So do I think that's a risk? I do. But it's not as much of a risk as we imagine. Another extraordinary study from Sweden, where they have the whole health system computerized, they looked here at 3,000 plus bipolar 1s who started antidepressants. And it's what's called a within-subject design, looking at other times for the same patients. And amazingly enough, in Sweden, a third of the bipolar 1s were treated at some point with an antidepressant monotherapy. That's a little surprising. And what they find, however, is that there is indeed a risk for increased mania. This is antidepressant monotherapy for BP1s, but it's only in the first three months of treatment. After that, antidepressant monotherapy is no more associated with switch than not taking an antidepressant. And the absolute risk is very low, with a number to harm of 112. They also found on the second bullet at the bottom of the slide, that people who are on mood stabilizers, whether it's lithium, Depakote, Lamictal, were the three they looked at, they did not have any increase in switching if you gave them an antidepressant. Meaning that mood stabilizers do indeed prevent switching. I don't quite understand the last of the sub-bullets. They even found that the patients who were on antidepressants along with the mood stabilizer had lower rates of mania after the first three months. Now why would antidepressants be associated with a lower rate of mania? One possibility is that clinical stability predicts clinical stability. It's like the jello is hardened. That people who are well tend to stay well. And people who are having a mood episode, depression, mania, tend to have more mood episodes. And it could be the fact that they were doing well predicted also fewer manias because the brain was quiet metaphorically. Back to the original meta-analysis by Cedar and McQueen. This is just looking at their switch rates. And you can see at the bottom the p-value is 0.89. that in a group, if you look at patients who are on mood stabilizers, predominantly, and you add antidepressants in the short term, six weeks, eight weeks, 12 weeks, no difference in switch rate, none at all. By the way, every meta-analysis finds that, that there is no difference in switch rates when the antidepressants are added compared to placebo, none. And again, this is from the group from Sweden, and there are now two studies, one that just came out two months ago, that's the study of the lower left, showing that the same thing is true with dopaminergic stimulants. They looked at bipolars who were given methylphenidate, Ritalin and its friends, and they found here, too, that if people are on mood stabilizers, you can add methylphenidate, I presume it's true with D-amphetamine, but we don't have the studies, safely. Because remember, there is a substantial number of comorbid bipolar disorder ADHD folks in our practice. I'm sure we all see them, and the question is, can we safely treat them with the treatments for ADHD, imagining that Adderall was actually available at any pharmacy? I imagine one day I'll tell our children that one day we were able to prescribe Adderall and it was available at pharmacies, as an example. Anyway, you can absolutely do this. You can add stimulants to mood stabilizers. I don't think I'd want to give dexedrine monotherapy to a BP-1 without a mood stabilizer. That would scare me, but if they're on mood stabilizers, it should be fine. And again, Ross Baldessarini has done just a review of all... This is a lovely paper. For those of you who want to read, probably the best review of treatments of bipolar depression. This is from three years ago. And he looked at all of the different drug groups, and if you look at the right side, you can see the relative risk of efficacy is basically the same for all the groups that antidepressants work, as well as virtually anything else we have. All right, so now let me switch. We have two other topics before we finish. One is, I want to talk specifically about bipolar 2 patients. In most studies, two-thirds to three-quarters of bipolar patients in the studies tend to be bipolar 1s, but clinically, we all have bunches of BP-2s. Should we think about them differently, and if so, are their data supporting doing it differently? And one of the people who's done the most work is J. Amsterdam at Penn. He has published a series of studies looking at antidepressants and bipolar depression, finding that it absolutely can be used, can be used safely and effectively. This is his Effexor versus lithium study for bipolar 2 depression, and he found that Effexor was more effective for 12 weeks, and there was no difference in switch rates. And my favorite of his studies is this one. It's now 13 years old. He took 81 bipolar 1s who were depressed and gave them Prozac monotherapy. Remember, these are twos, not ones. And the responders to Prozac, if you responded acutely, then you were put into a double-blind study. These are acute fluoxetine responders who get randomized to continued Prozac, continued lithium, switched to lithium, switched to placebo, double-blind. And what you see is that the people who stayed on Prozac did much better. The people who were switched to lithium and placebo, that's the orange line and the blue line, had substantially higher relapse rates. So that means Prozac was better at preventing depressive relapse among Prozac responders than lithium or placebo. And of course, he looked at switch rates, because remember, in this study, he's giving Prozac monotherapy for one year to a group of bipolar 2s. That qualifies as long-term treatment. And he found two things. One is, in formally defined hypomania, you know, four days of the usual symptoms, there's no evidence that the Prozac people switched any more than the lithium or placebo people. But, go two slides up, these are the actual young mania rating scale scores. The top line is Prozac, the middle graph is lithium, the lower one is placebo. You don't need to be a statistician to see there's a lot more noise on the fluoxetine graph. See that? What does that mean? That's what Keter is talking about, about roughening of the course or affective instability. The bipolar 2s on fluoxetine monotherapy had more affective changes, not enough to be formally diagnosed as hypomania, but more hypomanic days, more just mood instability. That's the price to pay for being on fluoxetine monotherapy. My guess is most patients would say, if this is more likely to prevent my depression, I'm willing to put up with that. But that's exactly what roughening of the course looks like. That's one of the rare studies where you can see it prospectively and it's so clear in the way they show it. A more recent study that we did at UCLA, Laurie Altshuler was the first author, was a 16-week study of lithium versus Zoloft versus lithium plus Zoloft. We found in a 16-week study, absolutely similar efficacy. Lithium did not add efficacy to the Zoloft and there was no difference in switch rate. So that's relevant. And again, I've told you that already. All right. So remember we talked about cycle acceleration. Does it cause, do antidepressants cause speeding up of the clock, more manias, more depressions, et cetera? There was an early study that Laurie Altshuler did when she was the NIH with Bob Post that found about a quarter of the bipolar showed cycle acceleration. But let us remember, these are all tricyclic MAO data. This is all pre-SSRIs, Welbitrin, modern antidepressants. And if you were going to see the speeding up of the clock from the antidepressant, you saw it within the first few months, not two years later. So if you're going to see it, you need to observe most carefully in the beginning of the antidepressant treatment. Then we have the classic study. This is also from NIMH from 40, 35 years ago, where they looked at, these were bipolar patients treated with tricyclics and what they looked at is cycle length, the amount, the number of days between one mood episode and the next one. And as you can see, these patients elegantly charted, you could see that when they were on the tricyclic, the clock sped up, the cycle length got shorter, right? And then when they went off it, it got longer. And then when they went on it, it got shorter again. Elegant data, except for four problems. Problem number one, these are NIMH patients. These are quaternary care. These are the hardest to treat, most stubbornly ill people around. Is that relevant to the everyday bipolars that we see? Number two, these were tricyclic treated patients, not SSRIs or bupropion treated patients. Number three, none of them were on mood stabilizers, which is not what you would do in today's world. That's just lousy care. And the fourth is, look at the large N on this slide. Anytime you see single digits, run, run quickly, right? So I absolutely believe these data, but these are unique patients treated in ways that are not the way you and I would treat them today. So even though this is true, I don't think it's relevant to what we do on a day-to-day basis. And one of the last topics to think of is, can we, should we? Is there a group of bipolar patients who should be on a maintenance regimen of mood stabilizers and antidepressants? And the answer is, I think yes. There are three studies that I show you on this slide that show that there's a group of patients on mood stabilizers and antidepressants, and anytime you take off their antidepressants, they get depressed. And then one day you wake up and you think, wow, I have a bunch of patients on mood stabilizers and antidepressants as their long-term regimen. And in fact, we all do. Now that doesn't mean that all of them should be, but there are a group who need to be on maintenance antidepressants with their mood stabilizers in order to not get depressed. Here's just what the graph looks like from one of the studies from 20 years ago. And again, the people who stayed on the antidepressants had half the relapse rates for depression as the ones that went off. So this is something that we do all the time in our mood clinic. And if you do it, I know you quietly won't tell people because you're embarrassed that you have bipolars on antidepressants, but it's okay. Let us not forget ECT. ECT is just as effective for bipolar depression as it is for unipolar depression. I show you the most recent meta-analysis showing that basically the efficacy is identical to unipolars. So you absolutely should consider using it. There's one interesting study from Germany comparing ECT to medication, and this is for treatment-resistant bipolar depression, and the ECT people did more better, and I absolutely believe that. I think ECT should always be considered for a stubborn-to-treat, moderate-to-severely-ill bipolar depression, as you would with unipolars. All right, so let's finish up with just a few, just some other ideas of things that have been talked about. Sleep deprivation, omega-3 fatty acids, N-acetylcysteine, sleep deprivation. By the way, sleep deprivation is there twice, obviously. I was sleep-deprived when I wrote it. Sleep deprivation is effective. The problem is it's only transiently effective. It works beautifully, and once you have a good night's sleep, the effect goes away. So who's going to spend all of their life not sleeping? That's hard to do. But I want to say a word about ketamine and TMS for obvious reasons. First is ketamine. There are three RCTs so far. It's used pretty regularly. Insurance companies are not thrilled they use the fact that the patient's bipolar frequently to not pay for the TMS, just because TMS is so expensive, but it's clear that it does work. So that's something you can do, and oh, I'm sorry, sorry, this was, I was talking about TMS. TMS certainly does work for bipolar depression, and I would urge you to prescribe it, and again, if you can get the insurance companies to pay for it, that would be great. And for ketamine, again, there are fewer studies, but again, it looks the same. It works just as well as it does with unipolar depression. One of the nice things about ketamine is you really can combine it with anything. There's a little concern about combining it with stimulants or MAO inhibitors, because ketamine causes transient increases in blood pressure, but by and large, you can add it to almost anything. So let me then summarize, and it's perfect when they will open for questions. As I said, my first sentence, bipolar depression is the dominant pole of the disorder, even though mania is the defining pole. Our database for the treatment of bipolar depression is embarrassingly weak. The fact is, we treat this stuff every day, and we do not have nearly as much guidance as to what to do as we should or would want. Among mood stabilizers, lamotrigine has the best data, and most of us feel it has the best efficacy both in treating acute bipolar depression and in preventing bipolar depression, which makes it the darling of the BP2 world. And for antidepressants, I think they do work. I think there's no doubt you can have switching. You remember the slide that shows, you know, what were the risk factors? Those are the patients that you might be a little careful about whether to use antidepressants, but for most bipolar patients, certainly if they're on a mood stabilizer, you can prescribe antidepressants safely. It may or may not work, but you're not likely to harm them. And as I showed you from some of Jay Amsterdam's work, there's reason to think that we can treat bipolar II patients with antidepressant monotherapy safely. Not everybody's going to agree with that statement, but I think there's enough data that's at least a consideration. I would not do that with bipolar ones. I don't, I don't, my stomach isn't strong enough to have somebody on antidepressant monotherapy who's been hospitalized for manic episodes. So with that in mind, let me stop and take some questions. Thank you. Hi, so a couple of things. We will be posting the slides on the app. And actually, for those of you who've been following the Clinical Update Series, we have posted many of the slides, but sometimes it gets posted a day or two after, but the app will function for weeks and months. Okay. So feel free to go back and use it. We now know why you win so many teaching awards. Thank you. And one of the greatest teaching techniques is to actually allow us 30 minutes for questions and answers. And apropos of other agenda we have during the Clinical Update Series this week, an extra dollar to anyone who mentions MAO inhibitors and you're going to get a cup of coffee after that and one more time we'll throw an extra provigil in. And by the way, when you ask your questions, if you're about to start your question with I have a patient, please reconsider and try to make it a generalized question as opposed to about a specific case. So but I am going to kick off with the first question. And that's about Valproate, two-part question. One, is it true that it has any greater efficacy for the rapid cycling or mixed state than lithium does? And two, I have not been able to find anything convincingly consistent about what target certain levels would be beyond getting below 40. Is there a good target? So two questions. Question one is about the atypical bipolar features like rapid cycling, mixed states, et cetera. So originally we thought that the anticonvulsants were better than lithium in treating those people. This is as maintenance treatments. And it really came from the early data showing that those features predicted a poor response to lithium. And someone had the radical notion of actually doing a comparator test of comparing lithium to divalpro-X for rapid cycling. And it turns out they both stunk. So it wasn't that divalpro-X has unique efficacy in these unusual bipolars. It was that those are hard patients to treat. And so lithium and divalpro-X are equally unlikely to be terrifically helpful. We do so much polypharmacy with rapid cycling because no single medicine seems to do the trick. In terms of valproate levels, the only study in our field that's, I think, helpful, and it really is, there's an acute mania study showing that if you have a valproate level of 94 or above, so make it 95 for round numbers, that you're much more likely to respond. That's an acute mania study, not a maintenance study. We don't have any great maintenance data, which is really too bad because we don't have anything to guide us. So is 95 the right answer for maintenance? I don't know. What I do is I tend to look for the ballpark of in the 70s and 80s. Obviously we know side effects do indeed correlate with blood levels. You don't want a valproate level over 120. People start getting really major tremors and nausea and diarrhea, and they just don't feel good. So for maintenance treatment, I target high 70s to 80s, but that's a guess based on zero studies. Oh, I'm sorry. Sorry, we just want to comment. We have a remote audience online, so we will be alternating questions between members who are here in the physical audience and those who are online. So please. Okay. So when bipolar depression is associated with high risk of suicide, does any drug besides lithium be used to treat both the depression and suicide risk? So yesterday we had another session about pharmacogenomics, and I gave the lithium talk there. So for years there's been a whole bunch of studies, not classic RCTs, but a lot of data showing that lithium prevents suicide, and it's the only one that we so far have seen it. And then three years ago or so in the American Journal was this large collaborative VA study where they added lithium to people with mood disorders, mostly unipolars, some bipolars, who had already had some suicidal events, so they were at higher risk, added lithium or placebo. And they found that lithium was resoundingly no more effective than placebo. That's this huge, it's the largest study in the field, and it was resoundingly negative. But the target blood level in that study was 0.6 to 0.8, which is probably correct, but the average blood level in the study was 0.5. And it's not clear whether patients didn't want to take higher dose, the doctors balked at prescribing higher doses. So all that study really showed is that low-dose lithium didn't seem to work. And the answer is that there are no other mood stabilizers that have been shown to prevent suicide. The only two in our field which are clear are the lithium, with the exception of that VA study, and clozapine and schizophrenia. Clearly shows fewer suicides. Could you emphasize the optimal timing for getting a lithium level? The optimal time for getting a lithium level is, and has always been, 12 hours. Make it 11 to 13 if you're not too obsessive. That's good enough. Now, it turns out, why is it 12 hours? You might want to ask that. And it's an artifact of early studies. So in Denmark, when they did the first studies on lithium, they looked at the pharmacokinetic curve, and regular lithium peaks at an hour and a half, and then gradually the level settles down. And somebody in Denmark, a pharmacologist named Amdi Anderson, Amdi, son of Amdi, decided that, you know, 12 hours is in the middle of that flat part. How about that as a good time to get lithium levels? This is how the 12-hour lithium level was invented. Nobody has ever studied whether that's the time frame that correlates with the most efficacy. Maybe it's the peak level. Maybe it's the trough level, which is 24 hours. Nonetheless, every study you have ever read, and the convention in the field is that 12 hours, 11 to 13 is fine. Now we finally get to your question. Thank you for your presentation. We all inherit patients sometimes on a regimen already, and sometimes we inherit patients with either a history of bipolar disorder already diagnosed, or a history suggestive of bipolar, that are only on an antidepressant. Let's say that patient is stable and not having any issues. Would you start prophylactically a mood stabilizer, or would you wait and see how the patient did and if they were symptomatic, and if so, which one would you choose? So let me get this right. This is a bipolar patient who you inherit who's on antidepressant monotherapy. Correct. Correct? Yes. Would you go for bipolar I or II? I've had both, but... Bipolar IIs, I'd probably leave them alone. I would say, look, there may be a slight risk. For bipolar I's, I would be somewhat nervous. The other thing I would look at, there would be two issues I'd look at in helping to decide what to do. One is, how long have they been stable? We're talking two weeks or five years. I mean, it really matters, right? Five years, you get a lot of cred for five years. The second issue is, is the diagnosis accurate? There are lots of people, especially people who have a history of substance abuse, has been well studied in the dual diagnosis world, who come in with a label of being bipolar, but really they have either drug-induced mood swings or personality disorder and mood swings, and somehow got labeled bipolar, which would also explain why they were doing fine on antidepressant monotherapy. The first thing I would do is a diagnostic review, and then I'd want to see really how long they'd been stable. Every once in a while, though, you do... We've inherited patients who come to our mood clinic who have been on antidepressant monotherapy for years, and they had a hospitalized psychotic mania. The diagnosis was not in doubt, but now they've been well for 10 or 15 years on, I don't know, Prozac monotherapy. Some of my colleagues would demand that they go on a mood stabilizer. I think the best history of the future is still the past. You put together 15 years of no symptoms, my goodness, I can't do that with most of my patients. So I'd probably leave them alone, but write a nice note in the chart about we documented the risks of antidepressant monotherapy. Thank you. How long do you need to keep the mood stabilizers, Depakote versus Lamotrigine, with or without antidepressants for a bipolar II patient for maintenance treatment. So what you're asking is, what is the proper length of what's called continuation treatment for bipolar depression? Now we know for unipolars, the right answer is in the ballpark of six months. If you actually look, there are only two controlled studies in the unipolar world. One said four months of euthymia was enough. The other said eight months. What's between four and eight? Right, six months. That's the answer. The same question has never been addressed in bipolar depression. So from a data viewpoint, I have no answer for you. Many of our colleagues are nervous about antidepressants for bipolar patients, obviously I'm not, and therefore will not do a six month continuation treatment. They'll do more like three to four, just because they want to get people off antidepressants quickly. I tend to, by and large, do the same. I borrow, I cheat, I borrow from the unipolar world. I do six months, and then I talk with the patient about, shall we try getting you off the antidepressant and have you just on your mood stabilizer, whether it's lamotrigine, divalproix, lithium, whatever. But the answer is probably three to six months. Again, with no data to support anything you do. Thank you very much for the excellent presentation. I usually keep my bipolar one disorder, depressed patients, one or two months on antidepressant. Which is also reasonable. I have no data to disagree with that. My question is, do you have any experience with bipolar patients on long-acting injectable, especially Abilify Mantenna? You know, the answer is very little. As you probably know, there are a few of them that actually have indications. Right, Risperdal-Constant was the first one to get a bipolar indication. And the only reason Constant was the one that was FDA approved is that oral resperatone was losing its patent. So Janssen tested the long-acting injectable. There's no doubt that the long-acting injectables work. Remember, that's just a different delivery system than putting it in your mouth. And of course, little better adherence because of the obvious factors. But there's never been a huge tradition of it in the mood world like there is in the schizophrenia world. So it's a perfectly reasonable thing to do. We have a handful of patients on long-acting injectables. But it's just, I don't know, it's just been a harder sell for bipolars than for schizophrenics. But it's perfectly reasonable to do. My experience is that bipolar patients on long-acting, on Abilifem antennae, for example, they do not become very depressed or not high. So their mood is level, and that keep them out of hospital. And basically, at times, I may add an antidepressant if they're depressed, or they do, even by themselves. Some patients with 30-year history of bipolar disorder, they treat them themselves. And they use antidepressant, for example, well, butyl, for a month or two. And then when they feel they're going high, they stop it. And they just report to me. You have very insightful patients. I'm really impressed. Oh. Yeah. So is there a known risk of rebound mania coming off of any mood stabilizer, or is that a phenomenon specific to lithium? So what's being referred to, there are, as I had mentioned, looking at those old lithium studies, if you stop lithium suddenly, and in the studies that I've looked at, and there aren't many, the magic number is two weeks. If you're gonna get people off lithium, or if your patient insists on going off, you should probably take a month to go off. Because if you go off suddenly, there's this clear phenomenon of lithium discontinuation rebound mania. That's been pretty well studied. That's true. Nobody's ever studied it with other mood stabilizers, like limotrigine, like divalprox, like carbamazepine, like antipsychotics. So nobody knows. The other question would be, clinically, frequently what happens is you're gonna take them off lithium because you're gonna switch them to another mood stabilizer, either because of lack of efficacy or side effects. So if you have another mood stabilizer on board, do you still need to taper the lithium? Or could you stop it suddenly because they're already on divalprox or whatever? And again, no one knows. Having seen this lithium discontinuation rebound mania on more than a few occasions, you know, when people go sour and you think it's because of what you do, it makes you more conservative. So I do the tapering, even if they're on another mood stabilizer. It may or may not be necessary, but that's at least what I do. Thank you. What about primiprexil for a resistant bipolar depression? So if we had had 66 minutes instead of 60, I would have added a few slides on primiprexil. That's what you're asking about, right? Yeah. So primiprexil, you may remember, Mirapex is a D2, D3 agonist used either for restless leg syndromes or in Parkinson's disease. And there are two small double-blind cells and two small double-blind studies, both of which showed efficacy of primiprexil. There is less controlled, but also some data on using it for unipolar depression. The usual doses are in the, you start at 0.25 and you gradually go up. Even though it's a D2, D3 agonist, it doesn't feel anything like classic stimulants. It can be sedating, a GI side effect. It doesn't feel like Adderall or Ritalin or Vyvanse or any of those. And I think it does work sometimes. So that would have, had we had longer, that would have been a slide to show because I agree with that. Thank you. Question about treatment options for elderly patients with bipolar disorder and major neurocognitive disorder combined. So that's a tricky one. So the big thing to know about geriatric psychopharmacology, certainly in mood disorders, is of course the start low, go slow. But even more importantly is there is this sodium lithium pump. And so it turns out as people get older, not only do they need lower blood, not only do they excrete lithium less well, so you have to give them less lithium to maintain a certain blood level, but they're more sensitive. And you can see an 80 year old get toxic on a lithium level of 0.6. So you wouldn't imagine that a level of 0.35 would be reasonable for a 23 year old bipolar one, but it would be very reasonable for an 80 year old with bipolar one. So I think what you'd want to be most careful about if somebody already had cognitive impairment is A, making sure that your lithium doses, no, yeah doses, but really your levels are low, and that you want to avoid things with anticholinergic effects because that obviously interferes with cognition and people who already have MCI or mild dementia already have almost no cognitive reserve and you don't want to mess with it. That's the key. Yeah, I had a related question or sort of related and it's about unipolar depression. When using lithium to augment treatment of unipolar depression, what kind of doses should we aim for? So the question is about doses or blood levels when you're using lithium as an adjunct for unipolars. And the answer is, so far as we know, you use the same levels that we now are targeting for maintenance, which is 0.6, 0.7. We used to target 0.8, 0.9, and then there's been this move, which I think is correct, that lower levels are A, better for your kidneys, better for your thyroid, better for your parathyroid. You have fewer side effects and you probably have no greater relapse risks. So 0.6, 0.7. You know, I don't even think in lithium doses. I only think in lithium levels because I can. Regarding lamotrigine, is there any data supporting what is an optimal dosage regimen and whether or not there should be blood levels? You would think, given how often we use lamotrigine, that there would be, I can give you a really good answer, but of course, this being psychiatry, I can't. In the maintenance studies that got at the indication, the dose range was one to 400 and they did some vague post hoc analyses and it wasn't obvious that higher dose was better or less good than lower dose. There have been a few stabs at lamotrigine levels. So I gather that our neurologist colleagues use lamotrigine levels to target a certain range, like two to 10, for epilepsy. There's a few little studies that suggest that lamotrigine levels for mood disorders should be, one was set above three, one set above five. But these are really soft data and it is not a routine part of practice. I don't think it's wrong, but it's not routine that we monitor lamotrigine levels, partly because lamotrigine is so insanely well tolerated. If medium dose doesn't work, so you go to three to 400. I know neurologists go higher. We inherited one patient from neurology in our clinic who was on 900 of lamotrigine for epilepsy. That got my attention. Can I add a point, which is, I don't routinely monitor lamotrigine levels either. However, because there's so many medications, particularly reproductive endocrinology, that is going to interfere with lamictal levels, that if I have a patient who's stabilized, and particularly if there's someone who's headed into the premenopausal phase, I will get a level when they're stable, not because I'm gonna change it, but I wanna know what I have to go back to when we introduce a drug that's gonna distort it perfectly right. So what's your experience in using Trazodone for your bipolar patients with insomnia? So Trazodone, of course, as we know is, it's only FDA indication is for depression, even though 99% of its use is for insomnia, as we all use it for. There are no studies on Trazodone for bipolar depression. It is a weak serotonin reuptake inhibitor. So you'd think, well, maybe that's not what I should be using as a sleeper in a bipolar patient. But in fact, its power in affecting the serotonin transporter is relatively weak, which is why you can actually combine Trazodone with an MAO inhibitor, and you should never prescribe an SSRI with a MAO inhibitor, which again is indicative of how weakly it hits the serotonin transporter. So I have no problem. I do use prescribed Trazodone for insomnia in bipolar patients who are either depressed or euthymic. Every once in a while, one of my colleagues prescribes Trazodone on the inpatient unit for acute mania to help people sleep. I wouldn't do that. That seems to be flirting with the switch gods. So I would not do it for acute mania, but in other states, I would. Since most of my, maybe all of my bipolar patients have at least one, maybe two or three comorbid diagnoses, I was curious in the studies with the add-on antidepressants how well they screened out other destabilizing diagnoses that respond to SSRIs like OCD, social anxiety, PTSD, that may account for the stability after the add-on of the SSRI. It's a very good question. I'm not, I don't know the answer. I mean, they screen out, some studies screen out, each one has a slightly different variation, but I'm not aware that there was any formal screen out for things like OCD, anxiety disorders. Usually what they screen out is acute substance disorders for obvious reasons, but otherwise, it's a mush. Yeah, those are huge confounding factors. It really seems risky to draw conclusions about somebody who gets better bipolar depression when you add an SSRI if they have PTSD. If you're saying that we need more and better data, you have my vote. There's a two-part question. They were asking about, according to your data, is it safe to say that sertraline is much safer than the other SSRIs in regards to the risk of switching? So there are few enough studies with SSRIs, but there are no convincing data that any one SSRI is more or less associated with switch. So I would consider them all the same. And the part two question is, which stimulants is safer? They're comparing modafinil or something like methylphenidate in augmentation for a patient with bipolar disorder. So remember I showed you that there were five RCTs on modafinil or armadafinil. So we can say with reasonable confidence that the vigils, provigil, nuvigil, really don't cause switching if you're on a mood stabilizer, at least during short-term studies. So I can say that safely. There are simply fewer studies with the dopaminergic stimulants, ritalin, dexedrine, Adderall, Vyvanse, et cetera. I think the data that we have indicates safety, but there's just fewer studies. So if you wanted to be conservative, you'd absolutely go with modafinil or armadafinil first, because you can be more sure that that's safe. Okay, would you please tell us more about most of patients with mood disorders in bipolar, including bipolar, they like to have a hit of cannabis. In my experience, I noticed that most of the time, that destabilize, especially the bipolar patients. What's your experience with that? Yeah, this is an interesting universal paradox. Patients all say it makes, that pot makes them feel better, right? All of our, well, not maybe not all, but lots of our patients say that, and the data in support of that is nonexistent. In fact, the data show that it's either neutral or makes people somewhat worse. The only exception might be GAD. That would be the only one, and there, maybe it's really CBD that's anxiolytic, not pot with a combination of THC, CBD. So by the way, I have many fewer concerns about pure CBD, A, because it's really completely different than THC, and there is this remarkable, it's a schizophrenia study, but it's about psychosis, showing that spectacularly high dose CBD, given double blind to schizophrenics on antipsychotics, made them better with no one getting worse. Boy, that's about as good a demonstrate, and those of you who know about CBD doses, so the gummies that you can buy in your pot store are like 10 milligrams, 25 milligrams. Schizophrenic patients in that study were getting 1,000 milligrams a day. You couldn't even afford that. I don't know how they did that. That's a lot of CBD, and it was utterly safe and even effective. So that tells you that THC is the destabilizing force within pot. Like you, I probably completely ineffectively urge them not to do it, and I get ignored about every day. Being a psychiatrist and having adolescent children, it's similar experience. I am seeing more and more that they really become more agitated and irritable and angry when they start. It also depends how much they do. You know, if somebody's doing a hit at night, that's one thing, but we all have these patients who basically smoke all day long. You say, well, how much do you smoke? And it's really every two to three hours they're having a half a joint. So those are the ones that really worry me, because that could clearly be destabilizing for sure. So we have a couple questions in similar vein. So it's about the concept of a bipolar spectrum disorder, or just the concept of bipolar spectrum. For patients with atypical recurrent depression, which doesn't fully respond to antidepressants or recurs too soon, is it still applicable to use mood stabilizers? Do you see patients which you conceptualize as bipolar spectrum patients? So the problem we have here is the DSM, our Bible, is a system of categorical categories. Categorical meaning we imagine that there are strict boundaries. So depression is five of nine criteria. If you have four criteria, you technically don't have major depression. But that's ridiculous. All of psychopathology is dimensional, not categorical. So we talk about people having bipolar one and bipolar two. But what if somebody has kind of really severe hypomanias? Can you call them bipolar 1.5? Or how about if their hypomanic episodes don't last four days, which is the minimum time threshold? How about if it lasts two to three? Are they bipolar 2.5? So again, almost assuredly, there are people on the bipolar spectrum where they have some features consistent with a very mild bipolar disorder, but not enough to meet the DSM criteria. Do I think those people exist? Yes. Do I think they really are on a bipolar spectrum? I absolutely do. The problem, of course, is, okay, how do you treat those people? Do you treat them like unipolars and see what happens? Or do you treat them like bipolars and use mood stabilizers and maybe avoid antidepressants? We have no data about that. Personally, I treat them as unipolar and because usually I'm not sure what they have. I mean, you know, these are subtle cases. And frankly, I'd rather be sure. And if I give them an antidepressant and they switch, now I'm moderately sure. So diagnostics certainly lets me sleep better at night and gives me a greater clarity as to what to do. But I have colleagues that I respect enormously who would look at that same patient and say, nope, mood stabilizers, and they would avoid antidepressants. So. What if you put them on an antidepressant and they're not getting better? So some of our colleagues think that one of the definitions of bipolar spectrum disorder is either the failure to respond to a number of antidepressants, also known as treatment-resistant depression, or recurrent tolerance to antidepressants. We have no data showing that that's true, but there are a number of our colleagues who believe that that should also be considered to be in the bipolar spectrum. Another area with much heat and little light. Hi, thank you for your talk. So I'm in my first year of practice after residency, and I'm in community psychiatry, so I'm getting a lot of new patients. Diagnostic clarity is a big thing, like you already mentioned, but also polypharmacy. And I'm finding, you know, I have these kind of messes of polypharmacy, which you alluded to earlier, with really difficult-to-treat patients, and I'm wondering if you could speak about maybe your approach to deprescribing, and also making sure that you're tracking, like if you, your approach to tracking, like efficacy and then deprescribing, so we're not just stacking things. So first of all, yes. Of all the disorders in the DSM, bipolar disorder is the most associated with polypharmacy, multiple medications. So we all see those patients, either we caused it, we prescribed them ourselves, or we inherited them. And I don't think it's always because, you know, the people who treated them, our colleagues who treated them before us are bozos. It's frequently out of desperation. You want people to get better, they're not getting better. And your hope is that A plus B is better than A or B. But it isn't, and then you add C, and then you add D, and then they're on six medicines, and it's not clear that any of them have really done a damn thing. That happens all the time. It happens to me. If it hasn't happened to you, it will. So that's just what happens. And you're right, once you realize that, then what you should probably do is, just as you say, you should sit down with the patient and say, you know, if you're not doing well and you're on these six medications, something's not right. I suspect the number of these are not helpful. Let's together figure out what order in which we'll try to taper and discontinue them. I always engage the patient in that, because patients develop these deep convictions about efficacy, even if they're incorrect. And I always do it one at a time, because otherwise I have no clue what's actually happening. So you're right. You'll get used to it after a few decades, really. Well, you've earned your cup of coffee, and I will throw in the provisional, even though you didn't mention the male inhibitors again. We want to thank all of you for coming today to this wonderful, wonderful presentation.

Dr. Ron Winchell

Dr. Kathy Krohn

Dr. Michael Gitlin

UCLA

Mood Disorders Clinic

bipolar depression

psychobiological models

treatment emergent affective switch

mood stabilizers

second-generation antipsychotics

lithium

tailored treatment