PARIS – Existing evidence on the best ways to treat bipolar depression is limited, and must balance against what works in the real world based on empirical experience. When treating bipolar depression "we are in the land of the uncertain, where [psychiatrists] must improvise and do their best with what’s available," Dr. Andrew A. Nierenberg said at the annual congress of the European College of Neuropsychopharmacology.
The trials that get drugs approved by the Food and Drug Administration often do not inform routine care. Patients in trials often differ from "real" patients, generally have milder disease, and are less susceptible to the worst outcomes. The published evidence base for treating bipolar depression should not be viewed as synonymous with best practice. Treatment is not a cookbook, and evidence-based medicine is very difficult to actually implement, said Dr. Nierenberg, medical director of the bipolar clinic and research program at Massachusetts General Hospital, and professor of psychiatry at Harvard Medical School, both in Boston.
Dr. Andrew A. Nierenberg
Treatment of patients with bipolar disorder in actual practice requires personalization, including prescribing agents that work and expecting patients to continue to take them despite possible adverse effects. Published evidence leaves it unclear how best to treat any specific patient, he said. That requires not only evidence but also clinical experience and knowledge of the results obtained by others, followed by feedback and outcomes assessment to guide future changes in treatment.
Perhaps the biggest question in the treatment of bipolar depression today is whether antidepressants work. A 2004 meta-analysis said they did (Am. J. Psychiatry 2004;161:1537-47), but this was "a terrible" meta-analysis because the overall numbers of patients were small, and the results from a single study heavily influenced the overall meta-analysis results, Dr. Nierenberg said. In addition, results from the STEP-BD (Systematic Treatment Enhancement Program for Bipolar Disorder) trial showed that adding an antidepressant to the mood stabilizer lithium gave no significant added benefit (N. Engl. J. Med. 2007;356:1711-22).
Taken together, the literature provides insufficient evidence to document a beneficial role for antidepressants in the treatment of bipolar depression, along with some evidence against any efficacy, yet Dr. Nierenberg said he prescribes an antidepressant as monotherapy or as an add-on treatment to many patients, as do many other psychiatrists. Psychiatrists generally feel that the data against adding an antidepressant are unconvincing, because they have seen antidepressants work on many prior patients. "The data are inconsistent with their clinical experience," he noted.
The roles of several other drugs or drug combinations also feature conflicts between the evidence base and their use in everyday practice.
The combination of olanzapine and fluoxetine showed efficacy in a highly influential 2003 study (Arch. Gen. Psychiatry 2003;60:1079-88). But the olanzapine and fluoxetine pairing is seldom used today because of concerns that it causes weight gain and tends to trigger metabolic syndrome. It’s ironic that olanzapine and fluoxetine is an FDA-approved combination that few patients actually receive, he said. Patients don’t like to gain weight, and when they do, they often stop taking the drugs.
Lamotrigine showed efficacy for treating bipolar depression in a 1999 study (J. Clin. Psychiatry 1999;60:79-88), but failed to show substantial efficacy in five other studies. In addition, lamotrigine failed to receive FDA approval for treating acute depression. Despite that, U.S. physicians appear to have embraced lamotrigine for treating depression in patients with bipolar disorder, "a discrepancy between what the literature says and what people actually do," Dr. Nierenberg noted. Results from a 2009 study showed that adding lamotrigine to lithium led to an improved decrease in depression, compared with lithium plus placebo (J. Clin. Psychiatry 2009;70:223-31), but physicians also prescribe lamotrigine monotherapy, even though published findings do not support that approach.
Perhaps the drug with the best clinical performance on paper is quetiapine, which has produced the largest effect size seen in depression (Am. J. Psychiatry 2005;162:1351-60), both monopolar and bipolar, and has had efficacy for mania, anxiety, insomnia, and psychosis. Based on this evidence, Dr. Nierenberg said that he would expect the extensive prescribing of quetiapine, but in reality, few patients receive it. Again, it’s a drug with significant side effects, including sedation and weight gain.
Other approaches to treating bipolar depression that are often used as first-line therapy, despite a poor evidence base, include optimization of a regimen of lithium, valproate, or carbamazepine that the patient might already take. Psychotherapy also can help, but less so once a patient has had 10-20 depressive or manic episodes. And electroconvulsive therapy is an important option for treatment-refractory bipolar depression. But the right treatment in many other clinical situations might involve unclear options, such as how to handle patients who are already on antimania treatment, and how to continue an antidepressant that produced partial improvement. Common complications of bipolar depression that can influence treatment decisions include attention-deficit/hyperactivity disorder, substance abuse, a chaotic life, and poor drug compliance. These and other factors make it impossible to rely just on evidence when choosing a patient’s treatment, Dr. Nierenberg said.