Dr. Sussman is editor of Primary Psychiatry as well as Associate Dean for Post-Graduate Programs and professor of psychiatry at the New York University School of Medicine in New York City.

Dr. Sussman reports no affiliation with or financial interest in any organization that may pose a conflict of interest.

Email questions or comments to ns@mblcommunications.com


Selecting the best strategy to optimize antidepressant response is a major ongoing clinical challenge. The need for more effective approaches for producing remission has been made clear by recent evidence that confirms treatment with any single antidepressant drug produces remission in only ~33% of patients, and that when antidepressants do work, they are of most benefit to those with more severe depressive symptoms. Patients with moderate levels of depression who seek care because they are either distressed or impaired by their symptoms may, paradoxically, be more difficult to bring into remission than those with a more pronounced disorder. An abundance of studies have shown that numerous augmentation or switching strategies may be effective for some patients, but no body of evidence demonstrates consistent superiority of any. In summarizing lessons learned from the Sequenced Treatment Alternatives to Relieve Depression (STAR*D) trial, the most ambitious attempt to date to address the question of antidepressant enhancement options, Rush wrote that “the gap between what we do in practice and what we know is very large.”1 In the near future, another study should be published that looks at whether it is better to start treatment with a combination of antidepressants rather than monotherapy. In anticipation of that article, I want to review the STAR*D trial and its major finding and describe the Combining Medication to Enhance Depression Outcomes (CO-MED), which looks at whether it is better to start treatment with two drugs instead of one.



STAR*D focused on non-psychotic major depressive disorder (MDD) in adult outpatients. The primary purpose of this research study was to determine which treatments work best if the first treatment with medication does not produce an acceptable response. All participants at first received the selective serotonin reuptake inhibitor (SSRI) citalopram (open label). If symptoms remained after 8–12 weeks of treatment, up to four other levels of treatment were offered, including cognitive therapy and other medications (Table 1). There was no placebo treatment option.

The study involved a highly representative clinical sample of depressed outpatients. In that regard, the STAR*D population was different than the highly selective cohort usually enrolled in industry-sponsored clinical trials.

At each level change, participants were asked to indicate the unacceptability of the potential treatment strategies (eg, to augment or to switch medications). Participants were then be eligible for random assignment the next treatment options.

Level 2

Participants who either did not have an adequate response to or could not tolerate citalopram are eligible for Level 2. The Level 2 treatment strategies were:


i) Medication and Psychotherapy Switch: switch to sertraline, venlafaxine extended release (XR), bupropion sustained release (SR), or cognitive therapy (CT).
ii) Medication and Psychotherapy Augmentation: add to citalopram either a) buspirone, b) bupropion SR, or c) CT.
iii) Medication Only Switch or Medication Only Augmentation: options were available for participants for whom CT is
iv) Psychotherapy Only Switch or Psychotherapy Only Augmentation: options were available for participants for whom additional medication is unacceptable at this point in the study (participants must be willing to continue citalopram)


Level 2A

Participants without a satisfactory response to their Level 2 treatment were eligible for random assignment to additional treatment at Level 2A. Level 2A was included so that all participants entering Level 3 had an opportunity to respond to at least two medications. Level 2A consisted of medication switch to one of two antidepressants (venlafaxine XR or bupropion SR).


Level 3

Participants without satisfactory response to Level 2 and, if appropriate Level 2A, were eligible for random assignment to one of the following treatments:
i) Medication Switch to: a) mirtazapine or b) nortriptyline, a tricyclic antidepressant.
ii) Medication Augmentation: add (to current Level 2 or Level 2A medication) either: a) lithium or b) thyroid hormone (T3).


Level 4

Participants without an adequate response to Level 3 were eligible for random assignment to Level 4 treatment. Level 4 includes two medication switch options: to tranylcypromine (a monoamine oxidase inhibitor), or to mirtazapine plus venlafaxine XR.



The study, which cost $35 million, failed to find any significant differences between drugs at each step in terms of the primary outcome measure, which was remission of depression. While there have been other significant findings from the study, the data did inform practitioners about “next step” management of patients with who fail on initial therapy.

CO-MED, like the STAR*D trial, was National Institute of Mental Health sponsored.2 Completed in late 2009, the results of this research have not been published. In contrast to STAR*D, this study compared whether a combination of antidepressants is better than one antidepressant alone when administered as initial treatment for people with chronic or recurrent MDD. CO-MED was designed to test whether two different medications when given in combination as the first treatment step, compared to one medication, enhances remission rates, increases speed of remission, is well tolerated, and provides better sustained benefits in the longer term. There were two arms to the study.

In Arm A, one of the following were given along with placebo: bupropion, escitalopram, mirtazapine, or venlafaxine.

In Arm B, two of the following drugs were given in combination: bupropion, escitalopram, mirtazapine or venlafaxine (Table 2).

Hopefully, the results of CO-MED will prove less nihilistic than those of the STAR*D trial with respect to answering the major question: Which treatment works best? One of the more disappointing aspects of both studies is that neither one looked at the use of a second-generation antipsychotic as an add-on treatment. Given that we now have increasing use of these drugs to augment antidepressants, it would be reassuring to have comparative data on both the safety and effectiveness of this approach.



Perhaps the answers we are looking for in terms of predicting treatment response may not come from large comparative clinical trials, but from molecular psychiatry research.

I want to thank David A. Mrazek, MD, FRCPsych, for serving as guest editor for this issue. Pharmacogenomic testing, he observes, may soon become standard practice based on the patient-specific evidence base that already exists. The four articles in this issue that address progress in individualized molecular psychiatry hint at the possibility, he notes, that we will be able to abandon our traditional trial and error approach to medication selection and begin providing our patients safer and more effective individualized psychopharmacologic treatments.

I also direct your attention to an article by Racha Nazir, MD, and colleagues regarding advice on starting the outpatient clinic. It touches on practical issues as office design, charting, knowledge of pharmacotherapy and psychotherapy, and individualization of patient care.

This issue marks the introduction of a new bi-monthly column titled “Clinical Updates in Child & Adolescent Psychiatry” by Margaret D. Weiss, PhD. Dr. Weiss is the head of the Provincial ADHD Program and clinical professor at the University of British Columbia Children’s & Women’s Health Centre in Vancouver, Canada. The column will highlight the clinical approaches, science, and new developments in child psychiatry. I look forward to her contributions and hope our reader-practitioners find the information useful.  PP



1. Rush AJ. STAR*D: what have we learned? Am J Psychiatry. 2007;164(2):201-204
2. Combining Medications to Enhance Depression Outcomes (CO-MED). Available at: http://clinicaltrials.gov/ct2/show/NCT00590863. Accessed April 19, 2010.