“What Has Changed?”
Antidepressants should be used “only if there is a history of a positive response,” whereas maintenance therapy should be considered if a patient relapses into a depressive episode after stopping the drugs, the report notes.
Pacchiarotti noted that since the recommendations were published nothing has changed, noting that antidepressant efficacy in bipolar depression “remains unproven.”
The issue is not whether antidepressants are effective in bipolar depression but rather are there subpopulations where these medications are helpful or harmful, she added.
The key to understanding the heterogeneity of responses to antidepressants, she said, is the concept of a bipolar spectrum and a dimensional approach to distinguishing between bipolar disorder and unipolar depression.
In addition, the definition of a mixed episode in the DSM-IV-TR differs from that of an episode with mixed characteristics in the DSM-5, which Pacchiarotti said offers a better understanding of the phenomenon while seemingly disposing with the idea of mixed depression.
Based on previous research, there is some suggestion that a depressive state exists between major depressive disorder and bipolar I disorder with mixed features, and hypomania state between bipolar II and I disorder, also with mixed features.
Pacchiarotti said the role of antidepressants in the treatment of bipolar depression remains “controversial” and there is a need for both short- and long-term studies of their use in both bipolar I and bipolar II disorder with real-world inclusion criteria.
The concept of a bipolar spectrum needs to be considered a more “dimensional approach” to depression, with mixed features seen as a “transversal” contraindication for antidepressant use, she concluded.
In Favor — With Caveats
Taking the opposite position and arguing in favor of antidepressant use, albeit cautiously, Goodwin said previous work has shown that stable patients with bipolar disorder experience depression of variable severity about 50% of the time.
The truth is that patients do not have a depressive episode for extended periods but instead have depressive symptoms, he said. “So how we manage and treat depression really matters.”
In an analysis, Goodwin and his colleagues estimated that the cost of bipolar disorder is approximately £12,600 ($16,000) per patient per year, of which only 30.6% is attributable to healthcare costs and 68.1% to indirect costs. This means the impact on the patient is also felt by society.
He agreed with Pacchiarotti’s assertion of a bipolar spectrum and the need for a dimensional approach.
“All the patients along the spectrum have the symptoms of depression and they differ in the extent to which they show symptoms of mania, which will include irritability,” he added.
Goodwin argued that there is no evidence to suggest that the depression experienced at one end of the scale is any different from that at the other. However, safety issues around antidepressant use “really relate to the additional symptoms you see with increasing evidence of bipolarity.”
In addition, the whole discussion is confounded by comorbidity, “with symptoms that sometimes coalesce into our concept of borderline personality disorder” or attention deficit hyperactivity disorder, he said.
Goodwin said there is “very little doubt” that antidepressants have an effect vs placebo. “The argument is over whether the effect is large and whether we should regard it as clinically significant.”
He noted that previous studies have shown a range of effect sizes with antidepressants, but the “massive” confidence intervals mean that “one is free to believe pretty much what one likes.”
The only antidepressant medication that is statistically significantly different from placebo is fluoxetine combined with olanzapine. However, that conclusion is based on “little data,” said Goodwin.
In terms of long-term management, there is “extremely little” randomized data for maintenance treatment with antidepressants in bipolar disorder. “So this does not support” long-term use, he added.
Still, although choice of antidepressant remains a guess, there is “just about support” for using them, Goodwin noted.
He urged clinicians not to dismiss antidepressant use, but to use them only where there is a clinical need and for as little time as possible. Patients with bipolar disorder should continue to take antidepressants if they relapse after they come off these medications.
However, all of that sits “in contrast” to how they’re currently used in clinical practice, Goodwin said.