According to a study just out in the American Journal of Psychiatry, starting depressed people on two antidepressants leads to much better results than starting them on just one - Combination of Antidepressant Medications From Treatment Initiation for Major Depressive Disorder. But how reliable is it?Currently accepted practice is to prescribe one antidepressant to begin with, and if the patient doesn't feel better after about 6 weeks, to either change to a different antidepressant (switching) or add a second drug while continuing the first (augmentation).
But in clinical trials and also in "real life", the proportion of depressed people who achieve "remission", meaning that they're fully or almost fully recovered, with their first antidepressant is rarely more than 1 in 3. Some antidepressants may be slightly better than others as first-line treatments, but any such differences are small.
Do two mediocre drugs combined add up to one good treatment? In this study, Blier et al. took 105 depressed people and gave them either one antidepressant or two. The one antidepressant was fluoxetine (Prozac) 20mg, and the two was mirtazapine 30mg and either fluoxetine 20mg, venlafaxine 225mg, or buproprion 150mg. The study was double-blind; patients didn't know which drug(s) they were on. There was no placebo group, however.
Mirtazapine (Remeron) is an antidepressant which is commonly used as an add-on treatment in depression, because it can be safely combined with most other drugs. So it makes sense to use mirtazapine in research like this, but take note: this study was "supported by Organon Pharmaceuticals", who make... mirtazapine.
What happened? All three combinations of two antidepressants were equally effective, and all three were considerably better than just Prozac alone, in the initial 6 week phase of the trial. The difference was massive by the standards of antidepressants - about 5 Hamilton scale points, considerably larger than the average benefit of an antidepressant over placebo.There was also a 6 month follow-up phase to the study in which everyone who had been taking two antidepressants had one of them replaced by placebos, so everyone ended up only taking one drug (either fluoxetine or mirtazapine). Discontinuing one antidepressant seemed to cause relapse in about 40-50% of the people who were taking two, as opposed to a 25% relapse rate in the people who started on just fluoxetine and kept taking it. If you believe it, this is further evidence that two drugs are better than one, although the total sample size was just 66 for this bit, and I'm not sure I do.
What are we to make of all this? This study joins a previous one finding that mirtazapine plus paroxetine is better than either drug alone as a starting treatment. But that paper was also by Blier et al and it was "fully funded by Organon Pharmaceuticals" although apparently "The sponsor had no role in the study design, in the collection and interpretation of the data, in the preparation of this report, and in the decision to publish this manuscript".
Personally, I'm not so much troubled by the industry sponsorship in these studies as I am by the nature of the add-on treatment, mirtazapine. Mirtazapine is an unusual drug, with a pharmacological profile very different to that of most antidepressants. Notably, it's a powerful hypnotic - it makes you sleep - and it increases appetite. Patients on mirtazapine in the present study put on over 2kg in 6 weeks.
Why does this matter? Because the two scales used to rate depression in this study, the Hamilton Scale and the Montgomery-Asberg Scale, both count reduced appetite and sleeplessness as symptoms of depression. If you're on mirtazapine, you're unlikely to have either problem - you'll be more worried about the exact opposite, insatiable hunger and drowsiness. So mirtazapine could reduce your total score on these scales even if it didn't change your mood. I have no idea to what extent this is a factor in these results, but it could be important.
So, are two drugs better than one? Should antidepressants come with a side-order of mirtazapine as standard? Maybe. But it's far from proven.
Blier, P., Ward, H., Tremblay, P., Laberge, L., Hebert, C., & Bergeron, R. (2009). Combination of Antidepressant Medications From Treatment Initiation for Major Depressive Disorder: A Double-Blind Randomized Study American Journal of Psychiatry DOI: 10.1176/appi.ajp.2009.09020186
13 comments:
Interesting. The general lore that I hear from psychiatrist on the lecture circuit is that there is no real evidence that polypharmacy is more effective than monotherapy, regardless of disorder.
On the other side of the coin, the combination of drugs and psychotherapy does appear to be superior to either one alone. The caveat being,however, none of these studies have subjects taking two placebos.
I would assumed that the "knowledge" that one is taking two "treatments" will lead to a superior performance over just one placebo.
We know the average difference between one tx and placebo is quite small (usually 10%). I wonder how large/small the difference would be with two placebos?
A dumb question from someone outside the field --
The way I read this, it seems like one group is taking a larger amt. of drugs total compared to a group taking strictly less. Group A was taking 20 mg of fluoxetine compared to group B taking 20 mg of fluoxetine plus 30 mg of mirtazapine. Wouldn't a more appropriate comparison involve group A taking 20 mg + 30 mg = 50 mg of fluoxetine? Or have studies shown that 50 mg of fluoxetine is no more effective than 20?
Anonymous: You're right that the two drugs group were taking more drugs than the fluoxetine only group.
you can't really "convert" 30 mg mirtazapine into 30 mg fluoxetine and give people 50 mg fluoxetine. for one thing that would be a very high dose of fluoxetine.
What you could do, however, would be to compare a normal dose of fluoxetine to a low dose of fluoxetine + a low dose of mirtazapine. If the two drugs really "add up", that would be equivalent.
I call BS.
Along the same lines as the second commenter, what they've basically done is a sloppy dose-response study, in which the combination arm got a higher aggregate dose than the monotherapy arms.
There's a right way to do studies like this, but it's difficult and expensive. It's called isobolometric analysis.
Also, strictly speaking, all the first 6 weeks shows is that mirtazapine is better than placebo. Everyone in the study got a non-mirtazapine antidepressant, so any improvement in the non-mirtazapine group (i.e. the fluoxetine alone group) could have been placebo, regression to the mean etc.
The only placebo-controlled aspect was that some people got placebo mirtazapine and some people got real mirtazapine.
The appetite/sleep score concern makes intuitive sense. Why doesn't anyone present analysis based on each question in the scale to address concern like these? Do people seriously think that improvement in appetite and mood should carry the same weight?
Stopping an antidepressants quickly is associated with withdrawal symptoms that mimic depression itself. Could this explain the higher rate of relapse that occurred in those switched from two ADs, to one AD and placebo (I'm assuming they didn't taper the AD)?
I'm also a bit confused as to how mirtazepine works. Have I got this right? It's an antagonist of 5HT2 and 5HT3 receptors, this allowing for more serotonin uptake at 5HT1 receptor?
Art: People have proposed using just the 6 or 7 main mood items of the Hamilton scale instead of the full 17 item scale. For the SSRI paroxetine this has been shown to give bigger effects vs placebo which suggests it really does improve mood. But SSRIs don't cause sleepiness and weight gain.
For mirtazapine, it's been shown to improve the mood item vs placebo but I'd really want to see the same analysis in this study. All of the "other" antidepressants in this study cause weight loss and insomnia (if anything), and I'm worried that the "added benefit" of the mirtazapine is just cancelling that out.
NeuroPsych,
Psychiatrists generally don't go on the lecture circuits unless sponsored by a pharmaceutical company, not that there anything wrong with that. And medications plus psychotherapy studies do not necessarily show advantage over medication alone with exception of CBT. I am not saying that this is not the case, only that studies designs and limitations preclude any conclusions in this matter.
The reasons why that could not be easily discerned are a) heterogeneous etiology of depression not taken into consideration when the studies are designed, b) cyclical nature of mood disorders independent from interventions, c) limitations on choice, dose, and number of medications allowed into the study (regardless of individual differences), d) subjective and faulty interpretation of the results.
Anonymous,
Fluoxetine and mirtazapine are proverbial apples and oranges. We can't compare doses mg per mg for unrelated meds. And the dose is just one of dozens of variables that we can control, not to mentions hundreds that we can't.
Paul,
Mirtazapine is a complex preparation indeed. It has several distinct mechanisms in addition to those mentioned: sedating anti H1 effect, NE reuptake inhibition (more prominent in higher doses), mostly presynaptic α2-adrenergic receptor antagonism, weak α1 adrenergic, and mild muscarinic.
If you feel confused, you are in the good company.
Anyone,
There is a strong trend in old psychotherapeutic community to equate effect of psychotherapy to meds, that is after years of failed attempts to dismiss meds as harmful and useless.
There are myriads of reasons for anti-pharma (Big or small) bias. As I wrote earlier, it's a turf battle. Psychiatry is still a field where emotions and personal preferences overwhelm facts and logic. Not for a long, I hope.
The absence of placebo group is poor design clearly...
The great problem is that in clinical research there will be less and less placebo for the future since this is the ethical idea of the regulators.
Having in ming how exactly the research are done (poorly and by generally uninterested practitioners all over the world [or should I say interested only in the payment]) the clinical research future is grim in any disease area and most so in the psychiatric, where the subjective factor is the strongest.
To Anonymous:
Let me challenge you to find a psychodynamic psychotherapy effectiveness study with good placebo control. While at it, let us know about practicing psychotherapists who are not "interested in the payment" and offer to work for substantially reduced or no fee.
Ho-ho-ho!
Thanks for this post. It can be challenging dealing with troubled teens especially. A lot of the problems you mentioned should not be taken lightly. Depression is a serious illness that affects adolescents just like it does adults. I’ve found that Silver Hill Hospital offers a series of great resources for Adolescent Depression patients and is a useful source of information about treatments for adolescents.
I surmise these trial results to be appalling and obtuse. The presynaptic receptors are nothing to toy with. Mitazipine, fluoxetine, and all these fancy names fail to amuse me. Results are what we seek. Hear me, big pharma?
Post a Comment