Antidepressants in Bipolar I Disorder: Never as Monotherapy

Bipolar depression is difficult to treat and prevent. Were this not so, the controversy about the use of antidepressants in patients with bipolar disorder would have faded away long ago. Physicians want the best for their patients, but they are often faced with difficult decisions related the benefit-risk profile of a treatment. And increasingly, patients have their own views on what they expect from treatment and which risks are worth taking and which not. At the end of the day, prescribing a treatment is an individualized decision based on knowledge, experience, empathy, and common sense. In an era of sophisticated genetic studies and dazzling neuroimaging tools, what makes a good psychiatrist is still sound clinical judgment. Last year, the Journal published an international consensus document on the use of antidepressants in bipolar disorder (1) that summarized the current evidence on their controversial efficacy and safety in bipolar disorder and provided a number of recommendations based on the state of the art. Recommendation 4 was “Antidepressant monotherapy should be avoided in bipolar I disorder” (1).

The article by Viktorin et al. in this issue (2) confirms that antidepressant monotherapy can induce mania in the absence of a concurrent mood stabilizer. The authors used Swedish national registries to identify 3,240 people with bipolar disorder who started treatment with antidepressants, and they used a powerful within-individual design to control for confounding by disorder severity and genetic and early environmental factors. The study results clearly indicate that the risk of antidepressant-induced mania can be reduced by the prescription of an adjunctive mood stabilizer (lithium, valproate, or lamotrigine). The study does not, however, resolve the question of the efficacy of antidepressants in bipolar depression, which is, at best, arguable (1, 3, 4). While clinicians, and perhaps patients as well, tend to believe that antidepressants are effective in the treatment of bipolar depression—and even beyond, judging from the high rate of prescriptions in this and other studies (2, 5–7)—the evidence base is very weak, with the strongest findings coming from one trial that found the combination of fluoxetine and olanzapine to be superior to olanzapine alone (8). The other question that is not resolved in the Viktorin et al. study is the risk of switch in bipolar II disorder, given that data on hypomania were not captured. Thus, we are still left with the precise recommendation of not using antidepressant monotherapy in bipolar I disorder only (1).

Questions and Answers

Why were up to 35% of the patients with bipolar disorder in the Swedish registry treated with antidepressant monotherapy? Why are 15% of manic patients maintained on antidepressants despite the emergence of manic symptoms (6)? Why are serotonin-norepinephrine reuptake inhibitors so widely used (7) despite their reportedly having a greater risk of inducing mania than selective serotonin reuptake inhibitors (9, 10)? Why are so many patients with mixed states treated with antidepressants (11)? There may be several answers to these questions. First, it may simply be that antidepressants actually work. In fact, the evidence that they work is as weak as the evidence that they do not. The absence of financial incentives to seek a bipolar disorder indication (given that marketed antidepressants all have an indication in major depression, and the diagnostic cross-sectional criteria are the same for unipolar and bipolar depression), as opposed to the case for antipsychotics or anticonvulsants, has hindered research on these particular compounds for bipolar disorder. Second, the extensive use of antidepressants may be the result of limited treatment options, safety concerns about overuse of antipsychotics, extrapolation of antidepressant efficacy in other indications, and perhaps pressure from patients themselves asking for rapid relief of the unbearable pain of depression.

Facts and Recommendations

The Viktorin et al. Swedish registry study quite strikingly confirms early findings that antidepressant monotherapy is potentially harmful in bipolar disorder (12). Patients should be advised that without the addition of a mood stabilizer, there is a serious risk of manic switch. The study does not resolve the question of whether antidepressants should be used at all. Research progress has yielded several alternatives to antidepressants, with better evidence of efficacy, albeit far from ideal tolerability (13). In the current situation, it is truer than ever that every patient is a unique case deserving an individualized approach. Every patient should have the right to receive a detailed summary of reasonable treatment choices, psychoeducation on the illness, and objective information on the benefits and risks of each strategy. For many patients, antidepressants will not be the first-line recommendation, but perhaps those with a depressive-predominant polarity, with bipolar II disorder, or with a previous response to antidepressants may be candidates for an adjunctive antidepressant (1, 14–16). For patients with a history of mixed states and switching, with rapid cycling, or with a mania-predominant polarity, we may want to avoid prescribing antidepressants at all costs (1). We may want to look at comorbidities, especially those that may respond to antidepressant therapy, before we decide what to prescribe. Entering the era of personalized medicine, which in psychiatry at this point means stratifying patients according to clinical profile (17, 18), we want better evidence for what to do in patient subgroups, rather than simple registration trials that speak about the drug but not about the patients. As mentioned earlier, if excellent treatments were available for bipolar disorder, this editorial would not make sense, and therefore any effort to develop and market novel drugs must be applauded. Quetiapine, lurasidone, the combination of olanzapine and fluoxetine (an antidepressant!), and ECT are evidence-based treatment options (13), but it is obvious that many patients need something else. That “something else,” at present, is what their physician or treatment team may figure out, after a careful investigation of the individual patient’s needs and profile, from a number of available options, including adjunctive antidepressants. And the recipe is still the same: scientific knowledge, experience, empathy, and above all, common sense.