Tuesday, May 26, 2009

Antidepressant Combinations for Resistant Depression

In many cases, a single form of therapy does not relieve symptoms of depression or anxiety. For every person, I recommend a range of healthy lifestyle changes, and I usually would encourage psychotherapy. For many, I encourage trials of antidepressant medication as well. For a significant number of people, a single medication does not relieve symptoms. In these cases of treatment-resistance, it can help to search for an effective combination of several medications. In this post, I will discuss the evidence-based foundation for combining antidepressants.

I find that several different antidepressant combos are worth trying, including an SSRI + mirtazapine, venlafaxine + mirtazapine, or an SSRI + bupropion. Adding trazodone to another antidepressant can be helpful for sleep, although I am less convinced of this combination substantially improving depressive symptoms. Combining newer antidepressants with older tricyclics is another area of interest, which I think we should be open-minded about, but I don't tend to prescribe tricyclics very often, mainly due to concerns about safety and side-effects. There are far fewer careful studies of antidepressant combinations compared to studies looking at single medication treatments, so the level of evidence in this area is not very strong yet. But here are a few references to existing articles from the research literature (I will try to expand my list over time):

This is a 2009 article from a Montreal group (Blier et al.) showing that a combination of mirtazapine + SSRI (paroxetine) led to significantly more improvement in depressive symptoms over 6 weeks, compared to groups taking either medication alone. The combination was well-tolerated for the most part-- in particular there was little difference in side effects experienced by the combination group compared to the group taking mirtazapine alone. Unfortunately, there are conflict-of-interest problems here: the study was funded by Organon, the mirtazapine manufacturer. And several authors of the study were "consultants" paid by the drug manufacturers for speaking engagements. The continued behaviour of psychiatrists accepting corporate money to give "educational lectures" is unfortunate. Usually large sums of money are involved. Please see my posts on industry-sponsored research: http://garthkroeker.blogspot.com/2008/11/biases-associated-with-industry-funded.html and http://garthkroeker.blogspot.com/2009/05/my-experiences-with-industry.html

Here's another recent article (2010) by Blier et al, again making a strong case for using combination antidepressants right from treatment initiation: http://www.ncbi.nlm.nih.gov/pubmed/20008946

This is an open study looking at combining escitalopram (Cipralex) with bupropion in chronic depression. There was no placebo or comparison group, so the study has a weak design. But it shows the combination was quite well-tolerated, and led to 50-60% of patients experiencing a remission of symptoms after 12 weeks, which in general is a better result than most single-agent trials, especially in chronic depression. The study was funded by NIMH, which reduces the likelihood of bias.

This is a 2006 review from Biological Psychiatry looking at combinations of bupropion with SSRIs. It is a good paper, but really most of the evidence presented is of mediocre quality. It does support the idea of using bupropion-SSRI combos, to improve treatment response in depression, and also to reduce SSRI-induced sexual side-effects such as delayed or inhibited orgasm.

This interesting 2004 paper is looking at the treatment of patients with depression plus chronic headache, using citalopram alone, or amitriptyline (a tricyclic antidepressant) alone, for 16 weeks. Then, a combination of the two medications was given to non-responders, for a further 16 weeks. The combination group showed substantial improvement in both headache and depression.


This is an important 2004 study from Biological Psychiatry showing that 6 weeks of a fluoxetine (SSRI) + desipramine (tricyclic) combination was more effective than either drug alone, in treating depressed patients admitted to hospital. While similar proportions (about 35-50%) of patients failed to respond to any of the three treatments, those patients who did respond improved much more with the combination. That is, the combination treatment led to a significantly higher chance of total remission compared to either single antidepressant treatment. The study was funded by NIMH.


This 2002 study looked at several options to treat non-responding depressed patients taking fluoxetine: the first group took a higher dose of fluoxetine (40-60 mg/d), the second group took a combination of regular-dose fluoxetine + desipramine, and the third group took fluoxetine + lithium. The results favoured the first strategy, of maximizing the dose of fluoxetine, instead of combining with something else. However, this result is not very useful, since normally we would maximize the dose of a single agent first anyway, before resorting to a combination. Yet the study does show that combining is not necessarily the best first step in addressing treatment-resistance. It was funded by NIMH.

This is one of the few studies looking at trazodone combinations: in this case, trazodone+fluoxetine showed some promise in treating resistant depression.


A weak old study from 1992 showing that trazodone can sometimes help as a combination with an SSRI (in this case, fluoxetine). 3 out of 8 depressed patients in this study improved with the combination (of course, this means that 5 out of 8 did not improve).

This is a 2009 review article on combining antidepressants. It is in German, which makes it hard for me to read!


Dr Shock said...

Nice overview, overall not much evidence that combinations can be more efficacious. The danger of interactions and side-effects increases with more psychopharmacological drugs.
Kind regards Dr Shock

GK said...

Resistant depression is extremely tenacious and debilitating, and I think we have to keep an open mind about exploring options including drug combinations.

We need more unbiased research data, looking at larger numbers of patients for longer periods of time, to help guide treatment decisions better.

Meanwhile, the existing evidence--limited as it is--does not discourage careful trials of combinations, provided patients are followed very closely for interactions and side-effects.

Anonymous said...

My Dr is pretty much someone who prescribes Seroquel 300 mg XR to all his bipolar patients. He actually told me this, not as bluntly, but still. Also, as a rule he is against prescribing more than one medication per patient.

As you know, not many people with difficult to treat bipolar disorder are to be helped by that kind of general solution.

I'm on Seroquel and Zoloft. My biggest fear is that Zoloft will stop being effective as it has many times before.

When/if that happens, could Mirtazapine as an add-on be an idea worth trying in your opinion? Or is a combination of Seroquel+Zoloft+Mirtazapine impossible? I take Seroquel before I go to sleep and Zoloft in the morning.

I'm bipolar 2 and chronic depression is my biggest problem. The manic side is not an issue anymore thanks to Seroquel.

An answer would be much appreciated!

Kind regards
/Elin, Sweden

GK said...


I can't give extremely specific personal medical advice on a blog, but I can comment in general:

A seroquel + zoloft + mirtazapine combination is not contraindicated. As with any therapeutic trial, it would be important to weigh possible positives and negatives:
1) Positives: could improve depressive symptoms. Might reduce SSRI-induced side effects or reduce need for as high a dose of SSRI.
2) Negatives: could provoke mood cycling; more side effects such as sedation, weight gain, or rare but more immediately serious side effects such as liver problems, severe agitation, etc. Would cost more money. Does not have an extremely clear evidence base in research, as yet, aside, perhaps, from individual case studies.

There may be more items in both the positive and negative column, that could be specific to individual situations or histories. Finally, it would be up to the individual and caregivers to weigh these factors and decide one way or another.

Best wishes,