TCPR: Dr. Baldassano, could you please share with us your definition of bipolar depression, and why we hear so much about it these days?
Dr. Baldassano: Bipolar depression is simply the term used when someone has a depressive episode in the context of having bipolar disorder type I or type II. We hear a lot about bipolar depression for two reasons. First, our bipolar patients suffer many more days with depression than they do with mania or hypomania. Many studies have found the ratio to be as much as three to one (Kuka RW et al, Bipolar Disord 2007;9(5):531–535). And second, it is very difficult to treat. We just don’t have as many effective treatments for bipolar depression, and bipolar depression seems to be more resistant to treatment than unipolar depression. It seems to be an entity distinct from unipolar depression.
TCPR: Historically, the definition of bipolar depression is simple: a patient who has had a manic or hypomanic episode in the past presents with a depressive episode. However some literature lately is talking about bipolar depression in patients who present with depression without a clear history of mania or hypomania.
Dr. Baldassano: This is true and it is where this diagnosis becomes more difficult. Patients often enter into treatment with depression. Upon first presentation, unipolar and bipolar depression look the same, so you can’t really distinguish the illness based on the depressive symptoms. Our job as psychiatrists is to try to understand what came before—what the longitudinal history is for the patient. This is especially challenging because often patients lack insight and it is not always easy to diagnose previous episodes of mood elevation, especially in type II bipolar where the elevations are hypomania rather than mania and often go unnoticed.
TCPR: If the presentations of bipolar and unipolar depression are clinically similar, if not identical, what is it about these depressions that is different?
Dr. Baldassano: A good illustration of this is the BRIDGE study (Bowden CL et al, Arch Gen Psychiatry 2011;68:791–799). The BRIDGE study was a multicenter, multinational study that looked at both prevalence and characteristics of more than 5000 people who presented with major depressive disorder. They found that many of them were actually undiagnosed bipolar disorder patients. They did this not through a systematic interview using DSM criteria, but with a set of “bipolar specifiers” they were able to diagnose a fairly large percent of this MDD cohort—47%, to be exact—as bipolar according to these “specifier” criteria, which look at factors like longitudinal history, family history, age of onset, and number of previous episodes. By comparison, when the researchers strictly applied just DSM criteria, they found fewer (about 16%) who were diagnosed with bipolar disorder. By the way, we don’t know whether the 47% of patients with “bipolar-specifier” features would respond better to a mood stabilizer than to an antidepressant. In other words, we don’t know whether their depression is more like bipolar or unipolar depression.
TCPR: Can you briefly explain how the results of that study guide your practice or how they should affect our practices?
Dr. Baldassano: What I always say to my residents is that when patients come in they give you a snapshot. But what a good clinician needs to do is create the photo album. And what is in the typical photo album of a patient who has bipolar disorder? I tend to look at five different dimensions: age of onset, course of illness, family history, treatment response, and co-morbidities. We know our bipolar patients tend to present earlier with their first active episode; in fact, in the STEP-BD program, the mean age of onset was about 17 years of age (Perlis RH et al, Biol Psychiatry 2004;55(9):875–881). We also know that bipolar patients tend to have multiple recurrences of their illness and inter-episode recovery. Bipolar patients often have a loaded family history, including suicide, schizophrenia, substance abuse, anxiety, and unipolar and bipolar disorder (Serretti A et al, Eur Arch Psychiatry Clin Neurosci 2012;May 9:online ahead of print). And in terms of treatment response, our bipolar patients tend to have more treatment resistance, more failed antidepressant trials. I think it is very important to look at the longitudinal course of a person’s illness, because it is essential to help you to parse out what is depression in the context of bipolar versus depression in the context of unipolar.
TCPR: So in a patient with depressive symptoms, even if they have never had a manic or hypomanic episode, if you still feel they may be bipolar, how do you approach treatment?
Dr. Baldassano: If I have a strong suspicion that a patient is bipolar, or if they have depression with one of the bipolar features I described earlier, then I am more likely to go ahead and treat with a mood stabilizer, for which we have more evidence. There is not an antidepressant that has shown in any large-scale study to be effective in the treatment of bipolar depression. (See for example, Sachs GS et al, N Engl J Med 2007;356(17):1711–1722. Editor’s note: in this study, all patients were taking a mood stabilizer and the study medications were added.) So in an attempt to practice evidence-based medicine I am more likely to utilize mood stabilizers, which we have more evidence for. My second concern is switch rate: in placebo-controlled trials the switch rate tends to be low associated with the more modern antidepressants such as the SSRIs, SNRIs, and Wellbutrin (Sachs GS ibid). However, if you look at observational data, which mirrors the real world, instead of clinical trials, you see higher switch rates. One study found self-reported switch rates in STEP-BD participants were as high as 44%. (Truman CJ et al, J Clin Psychiatry 2007;68(10):1472–1479).
TCPR: In those rare circumstances when you give an antidepressant to a person with bipolar depression, do you always give some sort of mood stabilizer prophylaxis, too? And if so, which do you use?
Dr. Baldassano: Yes, personally I do, because bipolar disorder is a cyclical disorder. I prefer that patients are on mood stabilization to help reduce this cycle. There are two agents that are approved for acute bipolar depression: quetiapine (Seroquel) and the olanzapine/fluoxetine combination (sold under the brand name Symbyax). Also, there is a plethora of data supporting lithium in both acute treatment of bipolar depression and maintenance treatment. I also use lamotrigine (Lamictal), which has fairly robust data in prevention of bipolar depression, and is FDA-approved for this indication.
TCPR: Is there anything that makes you less or more likely to try an antidepressant in bipolar depression?
Dr. Baldassano: I am much less likely to add an antidepressant if someone has had previous manic switches with other antidepressants. In the Truman et al study I mentioned earlier, the odds ratio for a manic switch was as high as 7 or 8 if the patient had had a switch with previous antidepressants. I’m also hesitant to utilize an antidepressant in someone who has a history of rapid cycling, because research shows antidepressants in patients with bipolar rapid cycling can increase the cycles or cause more cyclic acceleration (Ghaemi SN et al, Bipolar Disord 2003;5(6):421–433). And I am less likely to use an antidepressant in postmania depression.
TCPR: Should we approach depressive episodes differently in bipolar II vs bipolar I?
Dr. Baldassano: I am equally hesitant to prescribe an antidepressant in bipolar 1 and bipolar 2 if there is any history of rapid cycling. My decision about using an antidepressant has more to do with proximity to their mood elevations and history of multiple mood elevations rather than type I or type II. This hasn’t been well studied in the research.
TCPR: We’re taught that one of the reasons a depressed patient may fail a trial of an antidepressants is because he or she is actually bipolar. While this is often true, is there a risk of doctors undertreating MDD because they are led to believe that their patient actually has bipolar depression?
Dr. Baldassano: Not responding to an antidepressant or having multiple antidepressant failures can be a sign of bipolar disorder. But just assuming that a patient who has failed an antidepressant trial is bipolar is a bit simplistic. If I have a patient who does not respond to a good therapeutic antidepressant trial, it may prompt me to take a step back and take a better longitudinal history or review their longitudinal history, or to get collaborative data from a family member. And I would look again at what I think of as the soft signs of the bipolar features: the course of illness, age of onset, and co-morbidities.
TCPR: Thank you, Dr. Baldassano.