Every so often there comes a truly “headdesk” moment in science. A moment where you sit there, stunned by a new finding, and thinking, blankly…”OK, now what?”

For psychiatry and behavioral pharmacology, one of those moments came a few weeks ago with the findings of a meta-analysis published in the British Medical Journal (Eyding et al., 2010). The meta-analysis showed that an antidepressant, reboxetine (marketed by Pfizer in Europe, but not in the U.S., under the names Edronax, Norebox, Prolift, Solvex, Davedax or Vestra) doesn’t work. Not only does it not work, it really doesn’t work, and it turns out that Pfizer hadn’t published data on the putative antidepressant from 74% of their patients. Some people have reported that the study found that reboxetine was even “possibly harmful,” but that’s not quite true. What the study did find is that reboxetine produced more side effects (noted as “adverse events”) than placebo (as might be expected), but with no positive effects at all. While many antidepressants on the market today are not great, most are effective in around 60% of patients; reboxetine turns out to be even worse than that.

It turns out that publication bias was rampant. Pfizer and Lundbeck, the two companies running the studies, didn’t publish a lot of their data, especially the data showing no effect and unfortunate side effects. A bit nefarious, that. But bad science will out.

While sales of reboxetine never compared to sales of more traditional antidepressants like citalopram and fluoxetine, the study still puts a major kink in the pharmacotherapies currently available for depression. Whereas drugs like citalopram and fluoxetine primarily target the neurotransmitter serotonin, reboxetine targets the neurotransmitter norepinephrine. So it was hoped when drugs like reboxetine came on to the market that the different chemical focus might prove more effective or change the side-effect profiles normally associated with antidepressants. But obviously the side effects got worse, and reboxetine turns out to not be so effective in patients after all.

And this is a rough moment for scientists studying depression. Why? Because reboxetine works beautifully in our animal models. It’s practically a poster-child antidepressant. It produces acute effects in tests such as forced-swim tests and tail-suspension tests (which use changes in struggle as a measure of antidepressant efficacy). It produces neurogenesis in the hippocampus, which is thought to be correlated with antidepressant effects. When behavioral pharmacologists are doing comparisons between older antidepressants and newer ones, reboxetine is often used as a positive control, a drug known to have an effect in the behavioral test of choice.

But it doesn’t work in patients. And patients are what matters. Now, scientists are stuck with a difficult question: What went wrong? This is more than just an issue with an antidepressant that didn’t work, it’s an issue with the tests we are using to study depression. How effective are they, really? Are we in fact modeling the right things? Do the tail-suspension test and forced-swim test detect antidepressant activity after all? And if they aren’t detecting antidepressant activity, what are they actually doing? What does this mean for both the neurochemical theory of depression and the neurogenesis theory? Reboxetine affects both but still has no clinical effect. Does this mean that both of these theories are wrong? Or does it mean that they are incomplete? And where, exactly, do we go from here?

We may need new models to study depression, or we may need to simply redefine and reexamine the ones that we have. But the latest findings on reboxetine raise more questions than those about pharma companies, scientific conduct and efficacy in patients. They raise questions about the way we study depression and what it is we need to measure to come up with the therapies that patients need. And it makes it more important than ever to study the possible mechanisms behind depression and other mental disorders, to understand how they work and what behaviors and changes we need to detect, to gain new insights into how to combat depression with more success and less…reboxetine.