We have a problem in psychiatry. The illness that is most likely to cause episodes of depression – bipolar disorder – is also the most likely to get worse on an antidepressant. But the news is not all bad, and today I’ll bring you the latest word with a little help from our friend and colleague Joe Goldberg.
Welcome to the Carlat Psychiatry Podcast, keeping psychiatry honest since 2003. I’m Chris Aiken, and my cohost Kellie Newsome is away in her homeland: Melbourne, Australia.
Last weekend I went to the 9th annual Mood Disorders Summit. I took some notes and will share some of the highlights, starting with an action-packed review of antidepressants in bipolar disorder from Joe Goldberg. Whether you are for or against antidepressants in bipolar, you’ll find something challenge your beliefs here, but first – to clarify – we are only talking about adding an antidepressant to an anti-manic med. Antidepressant monotherapy is not recommended in bipolar disorder.
Joe Goldberg worked hard to identify the type of bipolar patient who is a good candidate for antidepressants. Someone who
- Has a history of responding very well to them
- Has no history of getting worse on them
- Lacks recent mania, rapid cycling, or current mixed symptoms.
- Has no history of substance use disorders
- Is preferably bipolar II rather than bipolar I
The problem is, according to Joe, that these patients are very rare, particularly because of mixed symptoms. About half of bipolar patients in the real world STEP-BD study had mixed symptoms during their depression.
The DSM requires 3 of these symptoms to qualify for a mixed features diagnosis, but Joe expressed some concern about those who have just one or two. So watch out for depression with racing thoughts, irritability, hyperactivity, distractability, and impulsivity. Those manic symptoms look different when they overlap with depression. The racing thoughts are not going to be about exciting plans or pleasurable pursuits. Instead, they are anxiety-ridden, and the patient will complain “I can’t shut my mind off.” The hyperactivity is not likely to be goal-directed in mixed depression. When these patients where activity monitors they pace aimlessly, moving from room to room, restless and agitated. And the impulsivity is likely to have a destructive nature – breaking things, quitting jobs, ending relationships – rather than starting new pursuits.
Later in the conference Dr. Greg Mattingly warned that antidepressants can worsen mood when non-bipolar patients have mixed symptoms, only here I would emphasize that we should count all 3 symptoms before we start to worry, and I’m usually more concerned when these patients have other features of soft bipolarity such as early age of onset, family history of bipolar, and highly recurrent depression. Greg suggested atypical antipsychotics for these patients. In my experience lamotrigine and lithium are also helpful, but Greg is going by the evidence and indeed lurasidone has enough evidence in major depression with mixed features that it almost got FDA approved for this.
And although these antidepressant-responders are rare, you want to look out for them, because if a patient with bipolar disorder does have a really good response to an antidepressant they are likely to get worse when it is taken away. Joe drew that conclusion from a widely cited 2003 study by the late Lori Altshuler and colleauges, in which they followed 84 patients with bipolar depression who had full remission on an antidepressant. One year later, the ones who stayed on their antidepressant were about twice as likely to remain in remission compared to the ones who discontinued it.
Altshluer’s study was not randomized, but Joe drove the point home with another study from Altshluer’s group that brought more clarity. This time they randomized antidepressant responders with bipolar depression patients to either stay on their antidepressant or switch to a blinded placebo for a year. Again, the ones who responded very well – who had full remission on the antidepressant – did much better by staying on it, validating the former study. But this study added one thing more because they included patients who only responded part-way. For them, staying on the antidepressant didn’t do any good. The bottom line: If your patient has a full response to an antidepressant for bipolar depression, keep them on the med. But if their response was part-way or iffy – taper it off.
There is one exception to this rule – rapid cycling. Now that we have more long term, controlled trials of antidepressants in bipolar disorder we’ve become aware that rapid cycling is possibly an even bigger risk than acute mania on them. So you always have to keep in mind that that “great response” to the antidepressant is really just the upswing in a series of rapid cycles that are about to unfold. When antidepressants cause rapid cycling, they may just make the patient cycle in and out of depression. You may not see any hypomanias, or they may not tell you about them, but eventually – in my experience after 6-12 months – all those cycles are going to culminate in a mixed state where the patient is much worse off than before they started the antidepressant. So we’re looking for sustained recovery when deciding to keep the antidepressant going, not just a flash in the pan.
The final thing that can go wrong with antidepressants is mood switching – either into mania, hypomania, or a mixed state. Switches into full mania are rare - Joe estimates they happen about 12-15% of the time. Most antidepressants just aren’t powerful enough to flip people all the way. Instead, they just sprinkle mixed symptoms on top of the depression, in which case your patient is not going to call you and say “Doc, I’m in a mixed state.” For patients, mixed states just feel like worse depressions – with more anxiety. I tend to see more swings into full mania with the older antidepressants – the MAOIs and tricyclics.
But how do you know if the new mood state is due to the med or the natural course of the illness? As with a lot of things in psychiatry, you never know the cause, but you can know the classification. Mood switches that happen within 12 weeks of starting the antidepressant are officially classified as antidepressant-induced, according to the International Society for Bipolar Disorders.
Let’s recap. If you use an antidepressants in bipolar disorder, make sure to have an anti-manic mood stabilizer on board, like lithium, carbamazepine, valproate, or an atypical antipsychotic – preferably one that is FDA approved in bipolar mania because some of the non-approved ones like brexpiprazole (Rexulti) and paliperidone (Invega) did not past the randomized-controlled test in mania. Lamotrigine is less clear. This one prevents mania a little but does not treat it at all, and we don’t know if it’s strong enough to prevent mania on an antidepressant. It may be OK for a bipolar II patient who has all the favorable antidepressant signs that Joe mentioned, but I would not use lamotrigine alone with an antidepressant in bipolar I.
Joe did not advise to use antidepressants or to avoid them. Rather, he laid out what might happen if we do. Some patients with bipolar depression get better on antidepressants, and if they recover fully they should stay on them. Nassir Ghaemi estimates that about 20% have this very good response. Others respond part-way, and here it’s usually best to taper off to avoid two problems that antidepressants can bring: Acute mood switching and rapid cycling.
But in Joe’s view, mood switching and rapid cycling are not the biggest problems with antidepressants in bipolar. The most likely outcome he said is that the drug is just going to do nothing – no better and no worse than a placebo.
Joe’s talk prompted a lively discussion, including this very relevant question – what counts as an antidepressant? Monoamines was the answer – those that increase transmission and block reuptake of serotonin, dopamine, and norepinephrine. There are lots of medications that treat depression through other mechanisms – ketamine, antipsychotics, antiinflammatories, and the latest dextromethorphan – which just got FDA approved and is now on the shelves as a bupropion combo pill called Auvelity. But whether these novel agents cause mania or rapid cycling is unknown – one speaker mentioned that even the gold-standard antimanic lithium has a case series where it supposedly caused mania but that this is most likely a random finding. If enough patients take lithium for bipolar, sooner or later a clinician is going to see a few manias after starting it and write it up for publication.
The Mood Disorders Summit is a great conference and is going strong in its 9th year. Since COVID, it has been virtual, but next year they will bring back the old-fashioned in-person feel with their 10th conference in Scottsdale Arizona, September 2023. It will actually be a hybrid event so you can live stream if you prefer, and whether online or in person the conference is free to all.
I’ll be back in two weeks with more pearls from the conference. In the meantime, follow me on twitter or linkedin for a daily dose of research updates. Today’s study tested several mood stabilizers to see which one had the best cognitive profile in first-episode bipolar disorder. The winner? Lithium.