Is one antidepressant
more effective than another?
I confronted this question last September, when I had breakfast with one of psychiatry's leading scholars and administrators. He was complaining about the influence of pharmaceutical houses on physicians. Why, he asked, did doctors prescribe the newer antipsychotics, like Abilify and Zyprexa, when studies showed that older and much cheaper drugs, like Trilafon and Haldol, were just as effective?
I told my colleague that I shared his concern about the influence of drug companies, but I thought that clinicians had a tough choice. The side effects of the medications were so different that the new and old medications were finally not identical.
All right, the colleague said, but what about antidepressants? Almost all those medications are available as generics, but at a disproportionate rate, doctors prescribe the one antidepressant that still has patent protection, Lexapro. Surely that disparity proves that Big Pharma has too much sway.
I went home thinking that I would write a blog posting on this topic. It turned out that the colleague was right. Competing with 35 to 40 other drugs in its class, Lexapro had upwards of a 13 per cent market share.
But then, I looked at the "comparative outcome" literature. To my surprise, I found studies that showed Lexapro to be especially effective. The relevant drug trials were suspect - many had been underwritten by Lexapro's manufacturer, Forest Laboratories. Besides, I have long thought that Lexapro is more potent than the drug company says. The common wisdom is that 10 milligrams of Lexapro is equivalent to 20 milligrams of Celexa or Prozac. But if that ratio is off - if Lexapro is more than twice as powerful - then certain studies would be skewed in favor of Lexapro, since research subjects would effectively be getting more of that drug than of the comparator.
So, I mistrusted the research, but there it was: data that might give doctors reason to prescribe Lexapro despite its higher cost. And I tend to respect the wisdom of practicing clinicians. Even without perfect objective evidence, they may have a good sense of what works for their patients. This set of circumstances seemed too complicated to write about, and in any case, my colleague's main line of argument had been lost. While it is likely true that drug companies warp doctors' behavior, the prescribing of Lexapro did not make that point in any clear way. I did write on this topic, in a posting that began, "Do drug companies hold too much sway over doctors?" but I restricted the discussion to antipsychotic drugs and left the hazy Lexapro evidence to the side.
Now comes news of a large-scale analysis of research on antidepressant efficacy. Published in The Lancet, it finds a hierarchy, with Remeron, Zoloft, Effexor, and, yes, Lexapro, leading the pack, Cymbalta and Prozac in the middle, and Luvox, Paxil, and (especially) reboxetine, which is marketed outside the US, bringing up the rear. Celexa and Wellbutrin gave statistically fuzzy efficacy results; the two drugs appeared to be about average for the group. In terms of tolerability, Zoloft, Lexapro, Celexa, and Wellbutrin led the pack. So the results give a special place to Zoloft and Lexapro.
Although this study received plenty of press, my sense is that it is reasonably unhelpful except as a stimulus to more and better research. After all, the studies it summarizes suffer from the problems I worried over regarding Lexapro, allegiance bias and questions of dose equivalence. Given the will and resources, we could resolve this question, whether one drug works better than another for major depression, but no one's done it yet.
To my reading, the roundup analysis does one thing: it accords the serotonergic antidepressants or SSRIs (Zoloft, Lexapro, and the rest) new status. Doctors have always suspected that these narrow-target drugs were less effective for serious depression than the broad-spectrum medications like Effexor and Cymbalta. (Oddly, Zoloft had an especially bad reputation for efficacy, though always a good one in terms of its side effects.) But in the new study, the SSRIs look about as effective as the "real antidepressants," and with fewer dropouts due to adverse events. Perhaps, in the widespread prescribing of SSRIs, we again are seeing the wisdom of the clinician in advance of the research results.
For over twenty years now, I have criticized horserace studies. In Moments of Engagement, I wrote about them with regard to psychotherapy. You have to know a great deal before you can set up a fair trial, and then you have to be wary of subtle bias in the way the race is run and judged. It's too soon to award trophies. The important truth is, we need remedies that work better altogether. But if the recent results do hold up, if SSRIs prove to be just as effective as other antidepressants, and if Lexapro (along with Zoloft) continues to hold a subtle edge in one way or another, those findings would suggest that clinicians are not so benighted after all.