July 2011
Reviewer: Jeff C. Huffman, MD
Combining medications to enhance depression outcomes (CO-MED): acute and long-term outcomes of a single-blind randomized study
Rush AJ, Trivedi MH, Stewart JW, et al
Am J Psychiatry 2011; 168(7):689-701
Background: Response and remission rates for major depression are limited when depression is treated with antidepressant monotherapy, even when the medication is dosed appropriately and aggressively. There has been some evidence from smaller trials that co-treatment with two antidepressants may be more effective than montherapy but this key question had not been studied in a large trial.
Methods: Subjects were recruited from 15 sites (6 primary care, 9 psychiatric). Subjects were required to have either chronic (>2 years) or recurrent major depression, diagnosed via the DSM-IV, and to have a score of 16 or more on the Hamilton Depression Rating Scale. Subjects with psychotic symptoms or bipolar disorder were excluded, but the study otherwise had broad inclusion criteria. Subjects were randomized to one of three arms (escitalopram-placebo, escitalopram-bupropion, venlafaxine-mirtazapine). Medications were flexibly dosed (up to 20mg/day escitalopram, 300mg day venlafaxine, 45mg/day mirtazapine, 400mg/day bupropion) over the course of 7 months, based on standardized assessments by study clinicians at multiple timepoints. The primary outcome measure was response and remission on the QIDS-SR at 12 weeks, though outcomes were measured throughout the 7 month period. Response was defined as 50% reduction of symptoms, and remission defined as consecutive measurements with scores less than 6 and 8.
Results: A total of 665 subjects were enrolled; in general subjects in the three groups were similar on baseline measures. Eighty-six percent of subjects completed at least 4 weeks of treatment, and 72% completed at least 12 weeks. Response rates at 12 weeks were 51.6%-52.4%, and remission rates 37.7%-38.9%, with no significant differences between groups. Similarly, there were no between-group differences in response/remission at 7 months. Of note, by 12 weeks, in the escitalopram-placebo arm, the mean dose of escitalopram was near the maximum of 20mg/day, but in the escitalopram-bupropion group was 12.5mg/day. At 12 weeks, the mean doses of venlafaxine and mirtazapine were 200mg/day and 20mg/day, respectively. Subjects in the venlafaxine-mirtazapine group had a greater side effect frequency, intensity, and burden than the escitalopram-placebo group at 12 weeks and 7 months.
Discussion: Though this is not strictly a psychosomatic medicine topic/trial, it is a large, important trial in the treatment of depression, and seems an important study for all psychiatric practitioners to consider. Overall, the results were in some ways disappointing — combination treatment with antidepressants did not seem to be a method to increase the somewhat limited rates of depression response and remission seen in real-world patients. The study was well-designed, allowed enrollment of patients with many comorbidities using broad inclusion criteria, and utilized a variety of sites. The choices of agents were reasonable, as bupropion added to an SSRI is the most commonly used combination of agents, and the investigators also looked to explore a second combination of agents that had greater effects on the noradrenergic system.
The major issue with the trial is whether the doses of the agents used in combination were sufficient. As noted, when used in combination, the doses of antidepressants were relatively low, and among the dual-action agents, did not reach doses at which their noradrenergic effects would be fully recruited. This issue may explain why these findings are in some contrast to prior, smaller trials that found combination therapy to be superior. The main question, of course, is whether it is possible to get most patients to tolerate full doses of two antidepressants; for example, the venlafaxine-mirtazapine group had higher rates of side effect burden even at these relatively lower doses.
The take-home message is, perhaps, not that combination therapy is ineffective, but that combination therapy at these doses is ineffective, and if one is going to treat patients with combination therapy for depression — which still may be a good idea — it is vital to get patients to full, good doses of both agents, and if that is not possible, one may be at least as well off with a full, high dose of one agent than more modest doses of two.