Dr Jonathan Meyer (University of California, San Diego): I want to welcome you all to our second CME session, entitled Troubleshooting the Tough Bipolar Case: The Who, When, and How of Mixed Features and Anxious Distress. This session is supported by an educational grant from Intra-Cellular Therapies. And I'm pleased to introduce our presenter, which I think some of you may know very well, Dr Joe Goldberg, clinical professor of psychiatry at the Icahn School of Medicine at Mount Sinai in New York, New York.

[00:00:30]

So if we go to the next slide, we can see his disclosures.

[00:00:36]

And if we advance it, we can get to learning objectives.

And I'll pass it along to my friend and colleague, Dr Goldberg. Take it away, Joe. Thank you, Jonathan.

Dr Joseph Goldberg (Icahn School of Medicine at Mount Sinai): All right, everybody's relaxed? Because I'm here to make you tense. So there's good news and there's bad news. Which do you want first? Bad news? We have a lot of material to cover. And the good news? We have a lot of material to cover. So depending on your cognitive orientation and your framework, this will be an enriching experience.

We're going to cover a lot of territory in the world of bipolar disorder, and here are our humble objectives. We'll talk about recognizing the presence of bipolar depression by diagnostic criteria, and talk about the DSM-5 specifiers.

How many of you actually use the DSM specifiers? Okay. How many of you know about the DSM, that you just don't use them, or you think about them, but you don't really confess to, because nobody else looks at. Okay, well, they're important, actually, because they really give you description and make more vivid the characteristics of the patient.

So whether you write them down or not, at least by the time we're done, you'll think about them more. There's like 9 of those. There's mixed features specifier, there's the anxious distress specifier. We're going to talk a lot about those. There's others. There's seasonal onset, there's peripartum onset, rapid cycling, so get to know them, get to know them. We're going to talk about two in particular today, and their implications for course outcome, morbidity, mortality.

We'll talk about the evidence for treating bipolar depression, particularly when either or both of these course specifiers is present. So moving us beyond just how do you treat the diagnosis, but actually how might you try to personalize and individualize the treatment for a given patient, course specifiers are 1 way to help you think about that and do that.

And so then to develop an individualized treatment plan for a given patient with taking into consideration things like these course specifiers.

[00:02:43]

So let's start with some voting.

[00:02:44]

How many people with bipolar disorder do you provide care for in a typical week?

1. 1-2;
2. 3-5;
3. 6-10;
4. 11-15;
5. More than 15; or
6. I'm just here for the beach.

Okay, I think we nailed that 1. All right, so 12 of you, I will see you outside with your wristbands after this.

[00:03:28]

Question 1

Now, let's get into some information about diagnostics. So which of the following should not be counted toward making the DSM-5-TR mixed features specifier designation during a depressive episode because of the no double counting rule? Is that:

1. Pressured speech;
2. Grandiosity;
3. Irritability; or
4. Increased goal-directed activity.

Okay, all right, we've got some spread here. A little more than half of you identifying irritability is, eh, not sure that that's diagnostically descriptively distinct. So let's see. We'll go on with that.

[00:04:30]

Question 2

Next question. Which statement best reflects the evidence for lamotrigine monotherapy in acute bipolar depression?

1. It shows robust, consistent efficacy across registration trials;
2. It failed to separate from placebo in most FDA registration trials, but meta-analytic signals are modest, so it's not a first-line acute treatment; or
3. It's contraindicated because it consistently worsens mania; or
4. It's effective only when mixed features are present.

And now, let's see. Oh, okay. Well, we have top-down and bottom-up. So is it robust, or does it fail to separate? You can't be both robust, and it doesn't work. So we'll reconcile that conflict. We go into mental health because we like conflict. So there's a conflict, if ever there was one.

[00:05:40]

Question 3

All right. Third and final question. For somebody with bipolar depression and mixed features, which regimen is evidence-based and appropriately dosed as monotherapy? Would that be:

1. Quetiapine, 50 mg at bedtime;
2. Lumateperone 42 mg/day;
3. Risperidone 2-4 mg/day;
4. Aripiprazole 15 mg/day.

The answer is 1 of these 4. That's a hint.

Okay. What do you all think? We have conflict. We're so conflictual. All right. So I went into psychiatry because I wanted to help people who are conflicted. So we'll see about that in the next 50 minutes or so, if we can clear some of the shrubbery away and get to a resolution.

[00:06:42]

So let's start with some basic phenomenologic things. So on the 1 hand, you could think about bipolar disorder and major depressive disorder as categorically distinct phenomena.

These characteristics that are shown on the left. Let's see. Does my thing work? No, this doesn't work. Let's go backwards. We're going to use my electric pointer.

So these characteristics on the left would be more defining of bipolar illness. You have manias and hypomanias in their episodes. They're syndromal. And you can have syndromal highs and have unipolar disorder, kind of by definition.

Antidepressants, the biggest concern with traditional monoaminergic antidepressants in the bipolar disorder world is no one's ever shown that they work. There's not a single placebo-controlled trial with any traditional antidepressants in bipolar I disorder, at least, saying that it's better than a placebo. If you're with a patient, you might want to know that in terms of informed consent. Whereas in unipolar disorder, antidepressants work just fine, right? Well, better than in bipolar depression, let's say.

We see trait impulsivity and aggression a little more so in bipolar disorder. And maybe some would say there's more hypersomnia, hyperphagia in bipolar than unipolar depression.

In the middle is stuff that isn't going to help you any. You have major depressive episodes in both. You can have psychosis in both. Suicide rates are higher in either condition than the general population. This notion of mixed features, that means elements of the opposite pole, coincident with whatever the other pole is. So low-grade high symptoms with a major depression or low-grade depression symptoms with a major high is mixed features. And DSM-5 now lets us span both poles with this notion of mixed features. And irritability doesn't get to double count for mixed features because it can happen in both. I'll show you some data on that.

So on the 1 hand, we could say they're categorical opposite ends of a spectrum. On the other hand, we could say there's more fluidity in this sense. And to say in almost like a more of a Kreplinian kind of way, rather than is this bipolar, yes or no, we could maybe talk about some of the elements that speak to a bipolar diathesis in someone who may not have had a formal episode of mania or hypomania. So, DSM-5 now gives us major depression with mixed features, for instance.

That means I've got a full-blown depression. And I'm allowed to have a few mania symptoms, but not enough to equal a full syndrome or episode of mania. And the chance of that progressing into categorical bipolar disorder is about 1 in 5. We'll see that as we go along.

So, takeaway from here: dualism, it's a dialectic. It's opposing forces. On the 1 hand, yes, there are instances of categorical dichotomy, but there's also fluidity. Very current concept.

[00:09:40]

So, irritability is really not diagnostic in the STAR*D study. Like, 40% of well-screened unipolar patients had anger attacks and prominent irritability. So that doesn't really help you diagnostically. That's like saying, "I have a runny nose. It must be a cold. It can't be the flu. It can't be pneumonia. It can't be allergic rhinitis. The symptom is just not diagnostic."

So, when you're asked things like, does it count toward the specifier, it's a bit of a tricky question. It can't count in the sense that it clinches things any more than if you have psychosis. I can't say, "Well, then it must be schizophrenia," because you can have psychosis from other things. So it doesn't not count to talk about irritability. It's just diagnostically not specific. So that's going to make your lives harder because you can't just say, "Oh, she's irritable. Must be bipolar." Doesn't work that way.

On the other hand, you can certainly build a story. And another reason I wanted to become a psychiatrist is because I like stories and narratives, often about people. And so you can build a story. You can talk about someone's age at onset, their family history, their trajectory of episodes, their treatment responsiveness. You can put all kinds of characteristics together, and some of them are going to point you in a direction, and some maybe not.

So irritability, duly noted, but it's not going to clinch anything.

[00:11:04]

You can screen diagnostically for bipolar disorder. Heck, we screen for cancer. We screen for diabetes. We screen for all the things we think we can treat. Someday, there'll be National Borderline Personality Disorder Screening Day. It'll probably be on a Friday afternoon at 17.00. And the reason we don't have that yet, you agree, the reason we don't have that yet is because we don't really have fantastic treatments for that condition.

Here's a little bit of bias. We screen for things that we can treat, and we sort of go like this when we, well, maybe it's something else. But here's the thing about—I didn't mean for that to happen. Here's the thing about screens. Screens are meant to not miss anybody. Screens are meant to cast a wide net. So think, if you're screening for cancer, you would much rather have false positives than false negatives.

Like, you know, I think you all are at risk if we're doing mammograms or pap smears or PSAs or CEAs or you pick your favorite screen. I don't want to get missed. I'd rather be told, "Oh, good news, we were wrong, you're okay," rather than the opposite, which is, "We were wrong, you're not okay."

So when you screen positive on a screener, like the Mood Disorders Questionnaire or the Rapid Mood Screener, it's no more of a diagnosis than saying, "Wow, your SED rate is up. You have lupus." No, no, your SED rate is up. Now we have to work you up for a bunch of things, including lupus. So screens are really helpful to get your foot in the door and also to rule things out. If you score below the threshold on 1 of these screeners, it means a very low likelihood that you have casehood. And now you are free to move on to 1 of the other 264 diagnoses in the DSM-5 before you're done with your diagnostic interview. So, I like screens as entry points.

[00:12:58]

I said a moment ago, I like to build stories.

So here are some of the things that go into the story of identifying bipolar illness, which if you have the full syndrome and you meet the DSM duration criteria and sufficient number of symptoms, is about 1%-2% of the population. It is not anywhere and everywhere. It's the rarer form of mood disorders, major depression having a lifetime prevalence of more like 17% and a past-year prevalence of about 8%-12%.

But you don't want to miss that 1%-2%. It'd be like saying, well, most respiratory tract infections are viral. Fewer are bacterial, and fewer still are fungal. But you don't want to miss the fungal 1 because the outward symptoms may be similar to the other kinds, but the treatment's different, the course is different, so you don't want to miss it.

So here's some clues and tips. Early onset. Early onset means adolescence. Most people with bipolar disorder will have symptoms of some kind before they're 18. And if you happen to have a depressive episode, which is the most common polarity of onset, before puberty, that's no slam dunk for the diagnosis, but you're at high risk.

This depressed, syndromally depressed 12-year-old has like a 20%-40% chance of evolving into having a manic or hypomanic episode over the next 10-15 years. So that's not the majority, but it's a pretty substantial minority. It's another building block in the story.

Family history. Genetics and mental health are not Mendelian traits. It's not like eye color or monogenetic illnesses like cystic fibrosis or muscular dystrophy. You have the gene Huntington's. You have the gene, yes or no? You have the illness, yes or no? We call it complex traits or non-Mendelian genetics, which means there isn't a gene. There's likely many genes that exert small effects and have cumulative risks, and they have to play off the environment because the environment can change how your genome expresses itself. That's called epigenetics.

So even 2 identical twins who have the same DNA don't have 100% concordance for expressing whatever these vulnerability genes are for bipolar disorder any more than they probably don't have 100% concordance for breast cancer or for lung cancer because this one smokes cigarettes and this one lays in the sun. And this one uses substances and alcohol, and is sleep deprived, and poorly handles stress and maybe takes antidepressants when they're feeling more agitated. And so you may actually change the expressivity of whatever the underlying genetics may be.

Episode durations tend to be short but frequent. So the average episode in bipolar disorder is about 13 weeks, according to the National Institute of Mental Health Collaborative Depression Study. That means if you get admitted for a mania on Friday night, you're not going to be all better on Monday. If you are, check the tox screen and see what cleared from your system. But it takes a few weeks to really see resolution through the natural course of things.

That also has math implications. If an episode lasts about 3 months, you can't have 50 episodes per year or per day, for that matter. So you can have mood variability, and this also gets into the notion of mixed features, where I've got changing elements of high and low within an episode that would count toward mixed features. But distinct episodes like rapid cycling, 4 or more episodes per year. An episode has a beginning, a middle, and an end, and it's got to have these other features that we talked about like speedy thinking and not sleeping. So count episodes from beginning to end and think about them as discrete departures from someone's usual baseline state.

Core features are listed here. And then response to treatment is helpful, but it's not diagnostic. You know, if I give you Tylenol when you have a fever, it doesn't mean you had a bacterial infection, or if your headache goes away, it doesn't mean you don't have a brain tumor. It's consistent, maybe, with a presumed diagnosis. If I give you a stimulant and you're distractible and you're more focused, it doesn't clinch that you have ADHD, but it's consistent with it.

And as I said a little earlier, there's never been a study with antidepressants showing that they were better than a placebo to treat bipolar depression. So if Mr. Highly Recurrent Depressed Patient is not getting better with antidepressants, that doesn't mean, wow, this must be bipolar, so much as it means, you know, I should probably pull out my MDQ and do a little more screening and look at family history because people that don't get better with antidepressants after many, many exposures certainly could have treatment-resistant depression, could have psychotic depression, could have poor adherence to their antidepressants, could have a TSH of 100, could be using substances. So there's a differential diagnosis for hard-to-treat depression, but in that differential diagnosis of many antidepressants haven't worked is might this be bipolar disorder?

[00:18:08]

Anyway, these are some clues.

Everybody probably knows this mnemonic or if you don't, now you do. DIGFAST are the core elements that define a manic or hypomanic episode. And just by way of review, it's a matter of magnitude. So if you're DIGFAST-ing and that lasts for at least a week and has psychosis and causes trouble and functioning, that's mania, it's bipolar I.

If you have bipolar II, when you DIGFAST, it need not last as long as a week. I mean, it often will, but it's got to be at least 4 days, and it doesn't involve psychosis. There's no such thing as psychotic hypomania. That's not a thing. And it doesn't cause trouble with functioning for the most part. In fact, it can look enhanced. So manics get hospitalized and hypomanics get put on committees or volunteer to be on committees and then over-volunteer and then they make off-color comments to the chairman of the committee, and then somebody gets fired, and then eyebrows get raised, and they cause a little chaos, and they stir things up, but it doesn't lead to the kind of psychosocial devastation that we see in bipolar I.

It's the depressions that really can be, well, lethal and the proximal source of morbidity. True, you can wipe out your life savings in the blink of an eye if you're psychotically manic. You can professionally embarrass yourself. You could make enormously life-changing decisions that you'll later regret, but the time spent with depression-to-mania is remarkable. Look down at the bottom.

This is from that NIMH collaborative study I mentioned. In the bipolar I world, if you line up time spent with mood symptoms over long-term follow-up, the time spent in the southern hemisphere, depression compared to the up, is 3:1. And in bipolar II disorder, 39:1. Blink, and you'll miss a hypomania. Hypomanias don't last. They're transient. They're like isotopes. So who wouldn't want to be, you know, hypomanic because it doesn't last, and then it goes into depression, and people are concerned they're just going to get depressed. So that's what makes this condition so challenging.

[00:20:15]

Here's the DSM criteria we think about for mania symptoms. Once again, here's your DIGFAST. And if we're going to talk about the mixed features specifier, that essentially means that you've got to have at least 3 of these things, but you can't double count.

[00:20:28]

So I think maybe you were asked a question about this a little while ago.

DIIS, distractibility, irritability, indecisiveness and sleep disruption certainly can go with mania or hypomania or depression. So it's not that it doesn't count. It's that it's not specific, right? So if I'm evaluating autoimmune diseases, I'm not going to say, "Well, fever doesn't count because it has many etiologies." What I'm going to say is it's not discriminating. So, that's an important distinction.

When you're thinking about is the mixed features element there. I get calls from colleagues; I want to say to this patient, she's kind of mixed. What's kind of mixed? She's kind of septic? She's kind of pregnant? She's kind of dead? I mean, what does kind of mean?

Kind of should impart that there's the presence of the opposite pole, but the kind of part is, I may not be sure if these DIIS things owe to the high vs low. And that's fine. I'm happy to try to figure that out along with you and vice versa because there's ambiguity, right?

I mean, these things can co-occur. If you're really talking about mixed features, the stuff that's not on this list, like fast thoughts, talking fast, excess energy, and activity that's goal-directed, now we're talking. Because mania is more than anything, some would say an energy disorder, more than a mood disorder. I'll show you a slide of that in just a bit.

[00:21:54]

So in the NIMH STEP-BD study, we showed that most depressed bipolar patients are not mania-free. It was 30-something percent, way on the left. I'm ignoring you, I'm sorry. It was, sorry. It was about 30% of bipolar patients whose depressions involved no elements of mania at all. A full mixed episode in the DSM-3R or 4 meant syndromal mania with syndromal depression. DSM-5 said, back away from that. Just elements of the opposite pole, but you can't double-count this, okay?

But look what we found in STEP. About half of bipolar depressed patients had subsyndromal mania. I think this was 1 of the studies, along with another couple of studies from Europe, that influenced DSM-5 to sort of soften the concept of mixedness. So, as to say, 1 reason why antidepressants may not work so well in bipolar depression is when you have mixed features, it could make things worse compared to having no mixed features.

[00:22:58]

Let's do a polling question. Serena is a 26-year-old woman with current major depressive symptoms. Six years after a brief first depressant, recently began escitalopram 20 mg/day, went up to 40 mg/day after a month. Now she's restless, hard to sleep, irritable, in her tension and agitated. What's the best course of action for Serena? Do you want to:

1. Rediagnose her based on this information as having bipolar disorder with mixed features and start an FDA-approved treatment for bipolar depression; or
2. Do you want to say, "Serena, adios the escitalopram, and let's see if this sleep, irritability, tension stuff gets better?"
3. Start a beta blocker because maybe she's apathetic from all that escitalopram, how about that?
4. Give her lamotrigine as a mood stabilizer.

What do you want to do for Serena?

So, yeah, most of you would say, "Look, you know, if I'm seeing an adverse reaction, the first thing we do in medicine in general is stop the potential offending agent and see if it resolves."

And in fact, that's what DSM-5 would tell you to do. DSM-5 would say, if you suspect emerging mania symptoms from an antidepressant, newsflash, take away the antidepressant. And if those emerging mania symptoms continue in the absence of the antidepressant after 5 half-lives, you have a good case to diagnose bipolar disorder.

But what if those mania symptoms go away? So there's a fancy name for that, it's called a side effect, okay? Just like if I give you an antihistamine and you get really drowsy and cognitively impaired, I don't say, "Oh my gosh, I've unmasked your dementia and your narcolepsy. Quick, let's get a behavioral neurologist in the room."

[00:22:04]

Well, if you're still fuzzy-headed and lethargic a month after you stop the diphenhydramine, sure, maybe I've unmasked a diathesis. But think about this factor in terms of the revelation of timing.

Now, you could make an argument to say, what if she is apathetic? And this is kind of like a little curveball here. It's possible. This is a real clinical decision. Is she clinically better from the depression, but she is restless, in which case you might say, I'm not sure I want to ditch the antidepressant. If she's apathetic, I'd probably lower the dose.

You could make an argument for a beta-blocker, and whoever the snarky person who wrote this question is, well, that's me, you could say, well, if it gets better with a beta-blocker, then I have some indirect evidence that this was akathisia. Okay, you could make an argument for that. I mean, if I was going to argue with myself, which I did when I wrote this, yeah, I'd probably lower the dose and see what happens.

Last point, in real life, it's sometimes hard to say, "Well, let's just pull away the offending agent and see what happens." Managed care doesn't like that in the hospital. "Hi, I've got a psychotic, manic patient who just began an SSRI last week. She needs to be in the hospital." "What's your treatment plan?" "I'm going to stop her meds." "Okay, and then what?" "I'm going to see what happens." "Okay, and the reason she has to be in the hospital for that?" "Oh, well, anything could happen." "Okay, and you want us to pay for that, why?" "Well, because I'm following the DSM-5 criteria. Hello? Hello? Hello?"

But that is technically what one would do. In real life, it's not that easy. So you may intervene, but do stop the presumed offending agent if for no other reason than to see what happens.

[00:26:41]

Can SSRIs cause akathisia? Yeah, I mean, they're not like dropping out of the sky frequency. There's mostly case reports. It can happen, so it's in the differential, but, you know, let's not get too carried away.

[00:26:56]

Activation, what's, oh, my patient got activated. What does that mean? Well, let's go back to DSM-5. You can't diagnose mania or hypomania just based on a couple of symptoms alone. So Serena is irritable, edgy, and agitated.

This is DSM-5 verbatim. Can't be taken a sufficient basis to diagnose a hypomanic episode or even indicate a bipolar diathesis. So thank you, DSM-5, for calling out what is sort of an unfortunate truth is, this is not quite enough.

Just like when I was growing up, if you sneezed 4 times, my mother said, "You're getting a cold." You were allowed 3 sneezes in our house. Four, you're getting a cold. To this day, when my kid sneezes, 1, 2, uh-oh, 4 sneezes.

This is called folklore. This is called maybe index of suspicion, but it's not in the ICD-10. So think about syndromes and constellations.

[00:27:54]

And think about energy. I told you a bit earlier that increased energy is the essence of mania or hypomania. You cannot say, "Oh, manic right now." I mean, unless you're under the influence of a substance. But by definition, nor can you, you know, be talking a mile a minute and gesticulating and say, "I'm so depressed," and fail to notice the mismatch between the signs and the symptoms. So the hyperkinetic, hyper motoric aspects of things.

Oh, I'm going to steal a minute from myself. There's this guy named Marty Teicher at McLean Hospital, real innovative psychiatrist. So he was doing actigraphy back in the 90s. And I visited his lab once, and they'd sit around, you know, actigraphy, you're just tracking movements over the course of time. And they'd look at these actigraphs of their patients and say, "Yep, yep, he's manic."

They didn't want any clinical history. They wanted to know how much movement was going on. And for them, the diagnosis went much more centrally to increased motor activity and psychomotoric acceleration and speediness. So I don't know if you can dispense with the other symptoms, but these later analyses that came along in the last 10 years say you really need to have elevated energy and goal-directed activity. That's now a necessary criterion to make the diagnosis of mania or hypomania.

[00:29:16]

All right, back to Serena. Well, for whatever reason, akathisia or not, bye-bye escitalopram. Let's give her lamotrigine, titrate her to the manufacturer's recommended dose for maintenance, which is 200 mg/day, and call her my favorite diagnosis, bipolar disorder unspecified, ICD-10F39. That means, I don't know, most common diagnosis I make, at least for third-party purposes. Her sense of feeling anxious and wired continues, unchanged. Her mood remains depressed. She's more irritable, minus escitalopram. She's ruminative, she's preoccupied, but she denies fast thoughts or overactivity of energy. Ooh, she read that paper I just showed you about high energy. Serena's making our lives difficult, and it's almost lunchtime. What are we going to do?

1. Well, do you want to just keep going higher than 200 on lamotrigine?
2. Do you want to add a benzo?
3. You want to do pharmacogenetic testing? When in doubt.
4. You want to give her some quetiapine? Or
5. Something else?

What do you think?

Yeah, I know what I wouldn't do. I'm not sure I know what I would do. Okay, quetiapine's certainly reasonable. Do you want to cross your fingers and hope maybe I'll get a little more juice out of lamotrigine for whatever the heck this thing is? When in doubt, here's a benzo. I'm not sure if pharmacogenetic, why not? Let's try a benzo, come back in a couple of days. Let's see if it works. I bought some time. It's fair.

[00:31:06]

So actually, you could do any number of things. I'm not sure pharmacogenetics is going to tell you very much. I mean, I guess you could make an argument, if she was a poor metabolizer for the cytochrome P450s that metabolize escitalopram, she may have gotten like a buildup, but now you've stopped it. So how's that going to help you? You're not going to put her back on another drug that goes to the same substrate maybe, but that's not necessarily going to help.

I love answer E. We'll think about what that might be.

[00:31:31]

Let's delve a little deeper into what, meaning something else, always like something else.

So where does lamotrigine fit into the pharmacopeia? Well, this is going to take a little bit of time. So in 1999, the very first study was done with lamotrigine in bipolar depression. And it's sort of worked on 1 of the measures of depression that the research team was looking at, but not on the main measure. So it was technically a negative study. It didn't work on the outcome of interest.

And then the manufacturer did several successive studies using the rating scale that worked in the first study, but wasn't the 1 that was chosen. When you do a study, you have to, in advance, declare, this is my outcome measure. If it doesn't work, I can't say, actually, this is not my outcome measure. So that's cheating. But for the subsequent studies, they used that other outcome measure called the Montgomery-Asperger depression scale, hoping it was going to work, and it didn't. None of these studies separated from placebo, and people kind of hung their heads and said, "How can this be? None of it works."

And then along came a meta-analysis where you increase the statistical power of finding a signal, and it turned out, on an overall basis, there was a signal better than placebo for lamotrigine to treat acute bipolar depression, but the size of the effect was small relative to placebo. And so it never got its FDA blessing as a treatment for acute bipolar depression. It's got its blessing to prevent relapses. And it never has ever been shown to prevent manias or hypomanias, only depressions.

So it doesn't really have a catchy name. I mean, you could say it's a mood stabilizer, but more accurately, you could say it's a compound that can help you achieve euthymia by virtue of preventing future depressions. And you're kind of leaving somebody uncovered on the high side of things. And if you're going to use it acutely, that's off-label, and that doesn't mean it's bad. It just means there may be other things that would be more potent. Well, we need a catchier name than that.

[00:33:27]

So back to Serena, what if she has anxious distress, right? You know, we talked about, this is another course specifier. It involves a couple of features of generalized anxiety disorder and a few features of panic disorder. I'm tense, keyed up, restless, can't concentrate. That's right out of the GAD page. And impending doom, and I think I'm going to lose control, is kind of out of the panic page.

But unlike having those as comorbid diagnoses that are there, these are just features that are present during the time of the current mood episode. And they use the word distress. I always wished DSM had included some psychosocial context. Gee, I miss axis 5, axis 4. Isn't axis 4 psychosocial stressors? We don't talk about them anymore.

So what am I distressed about? You know, if I'm getting audited or my significant other turns to me and says, "We have to talk," or worse, my 16-year-old turns to me and says, "We have to talk," I will have anxious distress. It's not necessarily a syndrome, but it's going to be a course specifier if I'm depressed to begin with. So I don't like that DSM sort of leaves out axis 4, distressed about something.

[00:34:42]

But anyway, it's a thing, and it's got some prognostic importance. And it's prevalent, and it can cut across both unipolar and bipolar patients. And it's common.

[00:34:52]

This is some work that Roger McIntyre did. So now we're toggling back to mixed features. So mixed features, where you've got symptoms of high and low at the same time. Anxious distress is when your mood episode includes 3 GAD symptoms and 2 panic symptoms only for the time frame in which you are having your mood episode. Got that?

[00:35:14]

And then there's more morbidity with these course specifiers. Mixed features may have a little more risk for cardiovascular disease, earlier onset.

Let's skip ahead with some of these things.

[00:35:25]

Maybe a poor response to treatment.

[00:35:28]

This is a slide that started with people with unipolar depression who had low-grade mania symptoms that we would now probably say comes close to the DSM-5 course specifier, and says over the course of 5 years, about 20% will polarity convert over the course of 5 years. This was a study done long before the DSM-5 even came along, but it was a useful study by the Collaborative Depression Group because it said a couple things.

One, don't ignore low-grade mania symptoms. You don't have to be syndromal for it to have relevance. And number 2, don't assume it's always going to convert onto mania. So this is sort of the part of the hard part.

Are you unspecified? I mean, before DSM-5, yeah, you kind of were. You don't have enough mania symptoms to qualify for the syndrome of hypomania or mania, but you got something. So it's a little bit more real-world in that sense that low-grade features count. Longer time till recovery, poor response, and maybe a 20% chance of polarity conversion. So you don't assume that that's the case, but you just have to think these things in your head and hold them in mind.

[00:36:37]

Poll 4

Which of the following would you say is the most significant challenge if you're going to apply these course specifiers?

1. Distinguishing mixed features from rapid cycling;
2. Distinguishing mixed features from anxious distress
3. Knowing whether irritability is an expression of mixed features vs a high or a low; or
4. Managing treatment resistance in bipolar depression with mixed features?

And if you want to say all of the above, just raise your hand.

So, okay, I mean, yeah, there's something here for everybody to think about. Mixed features is 1 episode. Rapid cycling is many episodes. They're different, right? Anxious distress, well, there is an element of autonomic hyperarousal in both when you think about it. You know, you're sort of on edge and you're a little more hypervigilant and I bet you your cortisol levels are up. But technically, mania is goal-directed energy, and anxiousness isn't necessarily goal-directed. It's a little more just fight or flight in some respects. What to do with irritability? I have no idea. And differentiation from, well, don't even ask that question. I mean, these things just co-occur.

So let's talk about treatment, right?

[00:38:16]

So thank you, Sunovion Pharmaceuticals, for going to the trouble of doing the 1 largest multi-site randomized trial of the DSM-5 entity called Major Depressive Disorder with Mixed Features, MDDMF.

These are hard people to diagnose because they have to have a major depression but never have had a mania or hypomania. Now they've got to have enough mania or hypomania to not be a full syndrome but enough to be present, and it can't be DIIS. Yeah, this was a fun study for people to enroll patients in. But they did it, and what they found was using lurasidone monotherapy actually worked better than placebo in reducing overall depression symptoms, overall clinical improvement with a pretty substantial magnitude of effect. 

The effect size of 0.8 is actually a pretty large one. So if you are able to say, you know, I'm not encouraging using unspecified F39.0 all the time, but if you are saying, you know, there's some elements of a high that are present and you have a major depression and I'm a little on the edge, I'm a little on the fence about whether or not there's enough here for a syndrome, you could certainly invoke these data and say, lurasidone does treat elements of both poles in unipolar patients.

That's an off-label use, by the way. The FDA never approved anything for MDDMF. So it's evidence-based but off-label.

[00:39:38]

Antidepressants, I said earlier, they've never been shown to work. That's their biggest drawback. Response rates are really statistically no better than a mood stabilizer alone, and mood stabilizer alone aren’t great.

In the STEP-BD, we found a 24% response rate for depression with lithium or valproate. It wasn't much different adding an antidepressant. So this is 1 of the biggest risks with antidepressants, not that they're going to throw you into mania so much as that they're just not going to work. Lack of efficacy is a side effect, if you like.

[00:40:11]

Here on the left in STEP-BD, we showed, well, if you have syndromal bipolar depression with any mania symptoms, giving someone an antidepressant will exacerbate those mania symptoms. We like to say mania symptoms are like gasoline fumes in the air, and the antidepressant is striking the match and the mania symptoms get worse.

If you're in that 31% of syndromal bipolar depressed patients with no mania symptoms, good news, the antidepressant won't make you worse, it just won't make you better. So we're still not thrilled with antidepressants. The study on the right comes from the Stanley Foundation, which also showed even low-grade mania symptoms during depression when exposed to an antidepressant is gasoline fumes to a fire.

So screen for these things, and even if someone's got low-grade, talking fast, thinking fast. I'll even cut you some slack on the DIIS thing because if you're thinking clinically, you're not sitting there enrolling patients in a clinical trial, some of you might be, but if you're not, thinking this irritability, this activation, certainly could be in the mixed domain. Let's step back from antidepressants and maybe think about things that are more efficacious when high and low are coexistent.

[00:41:30]

The International Society of Bipolar Disorders says in bipolar I disorder patients, if you're going to use an antidepressant, have a ceiling, have an anti-manic drug in the picture and don't use them if you've got, they say, 2 mania symptoms out of your DIGFAST list, in part based on what I just showed you.

[00:41:49]

Here's a handy-dandy little treatment algorithm decision tree. Memorize it, and we'll move on. Well, what does it say in essence?

I mean, you can read about it more, but it basically says most depressed bipolar patients just don't benefit from antidepressants. If you're looking for the needle in the haystack to say, is it a decent idea to try an antidepressant, let it be somebody who has no mixed features, no rapid cycling, no substance use comorbidity, they have to have bipolar II, not bipolar I disorder, and hopefully a previous excellent response to an antidepressant. Now we're talking maybe 1 out of 10 people.

The NNT is big; it's 39. You've got to treat, 29. Got to treat 29 people with an antidepressant before someone's going to get better. So you have to make an argument to use this obscure treatment that doesn't work very well in the first place.

[00:42:42]

So here's what's FDA approved for bipolar depression, and most of these compounds also have their indications for mixed features, present or absent. The companies that make these medicines did their homework, read the DSM-5, come into your office and say, "Do you see patients with mixed features?" And you say, "Yes, I do. Go, go away. What'd you bring me?"

But mixed features, like I said, is 1 of the things that would really disincentivize you to think of antidepressants, and that's like two-thirds, 70% of depressed bipolar patients. So no surprise antidepressants are not spectacular.

I give you here the dosing with these agents, response rates. I like the NNTs. The National Institute for Clinical Excellence, NICE, says an NNT less than 10 or 10 or less is considered a good thing. Meaning you don't have to expose somebody – you don't have to expose a lot of people before you're going to see a benefit.

I think of anesthesia, right? So we'd like an NNT of 1. Hi, Dr Goldberg, I'm your anesthesiologist. You know, for every 10 people I put this mask on, 1 falls asleep. Let's see if you're that 1 in 10. We'd like an NNT of 1.

I don't think we have any NNTs of 1 anywhere in medicine, including anesthesia. There's always the oddball person who doesn't fall asleep. But so single-digit NNTs are way better than the 29 with antidepressants, are better than the 13 with lamotrigine and feel free to explore off-label options, but these are the 5 FDA-approved. And anything under 10 sort of counts as being useful.

[00:44:14]

So real quick, cariprazine, dopamine D3/D2 partial agonist, works with or without manic symptoms present.

[00:44:24]

lurasidone in bipolar depression with or without mixed features, manic symptoms are present, so it's robust in that way.

[00:44:32]

Lumateperone recently got its FDA indication as an adjunctive therapy in unipolar disorder, joining its data in bipolar depression with or without mixed features. So that's almost like a descriptor of the illness that is being taken for granted. With or without mixed features.

You could go through all the criteria of all the course specifiers. Does it work with or without rapid cycling? Does it work with or without peripartum onset? Does it work with or without seasonality? I mean, if you incorporate the course specifiers, you're now really fleshing out the features of the illness in much more detail.

[00:45:11]

Here's some of the newer data that were just published a few months ago in the Journal of Clinical Psychopharmacology with lumateperone for mixed features, not just with bipolar depression, but also in MDD-MF.

So this is now the second study database that speaks to a molecule that has efficacy for major depression with mixed features.

[00:45:33]

More of the same.

[00:45:34]

Asenapine actually does have data for bipolar depression with mixed features. So it does work better than a placebo. It's even maybe a shade better than olanzapine. So it's not often widely thought of as having efficacy in that domain, but here's the data.

[00:45:49]

So let's not talk about individualization of treatment. Let's dig a little deeper into Serena's experience.

Serena makes an impulsive suicide attempt. She overdoses; she's hospitalized. And there, she's agitated, pressured, loud, labile, tearful, despondent, hopeless. She lost her job and her boyfriend, anxious distress. She's apprehensive, she's keyed up, she's overwhelmed. Her PHQ score is way up high. And her Young Mania Rating Scale score looks like this.

Who knows about the Young Mania Rating Scale? Good, now you shall.

[00:46:23]

Bob Young, when he was a resident at Washington University, St. Louis, for his residency project, came up with a scale to measure mania symptoms. Lucky him, this has been used in every bipolar study ever. Unlucky Bob, who was a supervisor of mine, he didn't get it copyrighted. If you leave here today with nothing else, if you make a scale, get it copyrighted.

Poor Bob, it's been like 50 years since his residency, he sees no dividends. Anyway, he did the field a great service because he took what are the DSM still, to this day, criteria, mood, elevation, energy, hypersexuality, sleep, irritability, talking fast –

[00:47:04]

– disorganization, distractibility, grandiose content, aggressivity, you kind of look disheveled, you know, the lipstick on your forehead, and fire engine red nail polish, and this variable, lack of insight. "I'm fine, what's wrong with you?" It goes with this.

So if we score this thing, on a good day, you're going to have as low a score as possible to get into a clinical trial. Normally you need, like, a 14 or 15 or more to get into a clinical study. If you're looking at full-on mania, we're talking more like in the 20s to 30s.

So Serena here, Serena's got elevated speech, irritability, and racing thoughts. So you scratch your head, gee, when you rate her on this thing, doesn't really come out as symptomatic as you might think, which in all fairness, may as much reflect the imperfections of the Young Mania Rating Scale as whether or not her symptoms are severe enough, because that ER description sounded pretty severe.

[00:48:08]

Poll 5

So now what do you want to do with Serena? You want to write a letter to Bob Young and say, I think your scale's not sensitive enough, because I saw this case in Miami, anyway.

1. She's bipolar I manic with rapid cycling;
2. She's bipolar I with both mixed features and anxious distress;
3. She's still unipolar with mixed features and anxious distress; or
4. Bipolar II with mixed features.

So, nobody said you're going to write Bob Young a letter and say your scale isn't sensitive enough for someone like Serena. So I'm intrigued that some of you thought this was bipolar II and not bipolar I because of the level of dysfunction and impairment. Unless you want to make the argument that her impairment comes more from the depression side than the psychomotor activation side, but I bet others of you would say, who's to tell? I mean, she made a suicide attempt.

Mixed features patients are much more likely to impulsively act on their despair than pure anergic motorically retarded patients. So these folks might say, no, this is a mixed presentation, and it's functionally impairing. You'd have to really minimize the extent of activation and elevation energy to say it's not impairing. You kind of see the dilemma with that? She doesn't have rapid cycling because she hasn't had 4 episodes in the last year, so good for that. And she had a major depression with mixed features, a third of you thought that.

I mean, you know, she's loud, she's pressured, she fulfills the non-DIIS issue. So she, unfortunately, or fortunately, either way, doesn't get a membership card in the bipolar club at this point, Serena.

[00:50:34]

Poll 6

What do you want to do with her?

1. You want to give her quetiapine 50 mg?
2. You want to give her lurasidone 20-120 mg?
3. You want to give her risperidone 2-4 mg?
4. You want to give her lumateperone 42 mg?

What do you want to do? There can be more than 1 answer.

Survey says, okay, sort of nice. Yeah, lumateperone certainly can treat bipolar depression with mixed features or major depression mixed features. Boy, they cover both bases, kind of sidestepping that. lurasidone's entirely appropriate.

Risperidone, I kind of agree with you. It hasn't really been looked at for bipolar depression with mixed features, and the 1 study that Janssen did with it in bipolar depression, it didn't work. Not only typical antipsychotics treat bipolar depression.

And 50 mg of quetiapine, I'm going to duck when Serena comes lunging at me in the ER and I tell her she's being involuntarily hospitalized. You're going to give her 50 mg of quetiapine and that's going to keep things chill? Okay, all right, all right.

[00:51:36]

How about individualizing treatment and healthcare provider goals? Well, so what are we treating? What are the target symptoms? Who's doing the treatment? Serena, at this point, takes the village. This is probably more than any 1 individual is going to be able to accommodate.

We're talking about immediate management, the interpersonal context of things, safety. Should she be in the hospital? I mean, maybe the question isn't should she be in the hospital, but is she going to get discharged the next morning because the CL psychiatrist person says, "Well, she made an overdose, and she's agitated. Off you go, involuntary admission." And the next morning, the attending comes in and says, "She looks fine to me."

So how do you make a plan and a trajectory? What's going to be the vertical care plan when she does leave? Is somebody going to put her back on an SSRI? I hope not. How do you deal with the interpersonal fallout? I mean, she's got a lot of moving parts going on and so it's just important to consolidate what the treatment plan is going to be.

[00:52:34]

So in kind of wrapping up here, when we talk about personalizing treatment, I like to invoke this language of identifying characteristics about a given patient that have informative value for the likelihood of a response or non-response.

Those are called moderators, patient characteristics that are shown on the left where you might do better with 1 treatment than another. For instance, euphoric manias do much better with lithium than dysphoric or mixed features presentations. Lithium's not as successful.

In mixed features when you're manic, valproate is. If I have mixed features, I'm probably not going to use an antidepressant after all the stuff that we talked about. We didn't say much about the rapid cycling course specifier today. Maybe next year we'll talk about that. But if you've had 4 or more episodes in the last year, adding an antidepressant has been shown to buy you more episodes in the coming years. So that would not be a criterion for saying you're an antidepressant candidate.

Psychosis, I probably want to include an antipsychotic. I might even think about ECT. Anybody want to give Serena ECT? It's lunchtime, okay. You can go through all of these things. Episode numbers, some medicines work better sooner than later.

Jonathan's written a wonderful book on lithium, and I think it's still true to say that lithium will do its best job if begun early in the course of illness rather than after multiple, multiple episodes have gone by in time. Is that still a fair thing? It's not a fair thing. Jonathan will tell you why that's no longer a fair thing. But the earlier studies said your best shot at getting success with lithium is before things like kindling and multiple episodes have set in. So maybe we'll hear more about the evolution of that.

Mediators are things that happen after the treatment has begun. Nonadherence, side effects, maybe she gets pregnant, there's a drug-drug interaction, and then you can sort of sort through your different treatment options.

[00:54:30]

So now let's wrap up.

[00:54:32]

You've seen these questions before. This is a short-term memory test. Which of the following should not be counted toward the DSM-5-TR mixed feature specifier during a depressive episode because of the no-double-counting rule?

1. Pressured speech;
2. Grandiosity;
3. Irritability; or
4. Increased goal-directed activity.

All right, I need to see 5% of you after this. You're just yanking my chain.

[00:55:12]

Okay, we talked about that. Let's move on.

[00:55:14]

Next. Which statement best reflects the evidence for lamotrigine monotherapy in acute bipolar depression?

1. It showed robust, consistent efficacy across registration trials;
2. It failed to separate from placebo in most FDA registrations
3. It's contraindicated because it's consistently worse since mania; or
4. It is effective.

Survey says, okay. Wow, 99.3%. So there's 1 or 2 of you. I'll see you after class.

[00:56:02]

So yeah, it has been shown to work for that purpose.

So let's move on because it's lunchtime coming up.

[00:56:08]

For a patient with bipolar depression and mixed features, which regimen is evidence-based and appropriately dosed as monotherapy?

1. Quetiapine at 50 mg/day;
2. Lumateperone 42 mg/day;
3. Risperidone 2-4 mg/day; or
4. Aripiprazole 15 mg/day.

And, all right. Yeah, lumateperone does the job here, and the other one's not so much. Okay, and with that, with that.

Dr Meyer: Thank you very much, Dr Goldberg. So we have 3 minutes and 53 seconds. The 1 thing I'll say about lithium, which you've referenced, is that more recent analyses have seemed to indicate that if people go off their lithium and go back on it, it should respond the same way.

So the idea that you only start it early, it may not work later, is not necessarily true. Some of these people are just sicker if you start it later, but it's not necessarily a lack of effect of lithium. So we think in general it's still probably the place to go for a lot of people.

Dr Goldberg: Important point. There was a study by Bob Post in 1992, which said if you stop lithium, it may not work well the next time around. A lot of debate about that. The 1 compelling argument may be if your illness gets worse. So you started out with simple mania and now, years later, off lithium. You've got mixed features and rapid – da, da, da – then, it may not work for that.

Dr Meyer: And I think that's the point, is that some of these people were sick, but when they're fine, they're stable. But once they destabilize, they go south, and it takes sometimes heroic efforts to get them back. It's not a lithium failure per se. It's a failure of their illness to respond simply because they are not the same as they were, you know, before their 10th manic episode.

So there's been a couple of questions about anxiety disorder comorbidity, with the idea, if you had a patient with OCD or panic, for example, and they clearly have a bipolar disorder diagnosis. Do you dare stick your foot into the waters of SSRI?

Dr Goldberg: So you can certainly stick your foot into the waters with the caveat that no one's ever studied SSRIs to treat anxiety and bipolar anxiety, and we made the assumption that SSRIs do treat bipolar depression, and we were wrong about that. So until someone is a study, the answer is we don't know.

The second piece to this is anxiety comorbidity is common. 45% of bipolar patients have an anxiety disorder. It's remarkable to me that the field has not studied this as a primary outcome unto itself. If you look at the Canadian practice guidelines, CANMAT, they acknowledge this, and they will say things like, well, firstly, some of the atypical antipsychotics will reduce anxiety symptoms. So go with that. Number two, some anticonvulsants have some antianxiety properties. There's 1 study with lamotrigine. There's 1 study with valproate, a study we did. It's got some anti-anxiety properties. I probably wouldn't use a stimulant for anxiety. I probably would not favor anything that's activating for the patient. I'd steer clear of drugs that can aggravate akathisia.

I might think about behavioral approaches. Here's where CBT can sometimes be helpful, especially for certain kinds of anxiety like social anxiety. OCD exposure therapy with response prevention, but it's very empirical. Lastly, the Canadian guidelines do call out a potential role for gabapentin, pregabalin for anxiety. Not well studied but theoretically of interest for anxiety.

Dr Meyer: Would you hypothesize that this anxiety, especially when it's sort of a generalized somatic presentation, is simply the manifestation of inadequate mood stability?

Dr Goldberg: I mean, it's certainly a possibility, and there are some people who will say, "Well, just treat the Dickens out of their mood, and maybe their comorbidity will go away." And I hope for that tool. It's just I think if you have true comorbidity, if I have diabetes and hypertension, and they can co-occur very often, treating 1 isn't going to make the other 1 go away necessarily. So I very much would say let's optimize the treatment for 1, but let's not assume that 1 will automatically resolve. Same is true with substance use disorder, if I stabilize your mood, you'll stop using heroin, right? Right?

Dr Meyer: Seemingly not. Okay, so we didn't get into older patients, but let's say you have a 75 who's bipolar I. Would you prefer, in this individual, lithium or an atypical like lumateperone, for example, in that person aged 75?

Dr Goldberg: So the good news is we've got increasing databases. Clinical trials now enrolling patients up to 75. Used to be up to 65. So we've got a database with drugs like lurasidone, and Martha Sajatovic did some studies in older adult patients, and the atypical antipsychotics that work in younger patients seem to do just fine in older patients. And so that's good news as a moderator is concerned. You wouldn't say, "Oh, sorry, you're over 65. It isn't going to work for you." You know, there is data with lithium and valproate in older adults, and it does work. So I don't think you have to be as concerned about age as much as things like, how's your GFR? How's your liver function? What other drugs are you taking? If you're on a thiazide, diuretic, if you've got cardiovascular disease, that may steer me away from or towards something else. So I wouldn't make it by the age as much as by you.

Dr Meyer: And I'll leave 1 last thought. When you give older bipolar I patients lithium, they stay on it for a few years, you reduce their future risk of dementia by –

Dr Goldberg: Oh, yeah. Oh, good point.

Dr Meyer: – 49%. And that's the reason you want to get good, assuming their GFR is adequate, and it's often the medical comorbidities as much as anything, which do in the kidney function.

So it looks like we're out of time. I want to thank Joe again for an entertaining and educational lecture.

[END OF TRANSCRIPT]