recent article in the BMJ has shown that the data on reboxetine has not proven its effectiveness. This evidence has at least partly been hidden because of publication bias.
However, the FDA never gave reboxetine a licence anyway. Intriguingly, in a rapid response to the BMJ article, the medical officer who reviewed the FDA application hints that we still haven't got all the relevant data. We still don't know why the FDA turned reboxetine down in 2001.
It's tempting to speculate, as GoozNews does, that somehow it was connected with David Healy's overinflated promotion of reboxetine at the time in terms of restoring social interaction. Reboxetine is a NARI rather than a SSRI, which means it is more specific for blocking the reuptake of noradrenaline, rather than serotonin. Serotonin specific reuptake inhibition (SSRI) was marketed as the mechanism of action of a whole new generation of antidepressants, such as fluoxetine. The competitor noradrenaline reuptake inhibition (NARI) hypothesis never really gained ground, not least because the FDA did not approve reboxetine.
So, it could be said there's always been a bias against reboxetine. I suppose this could explain why no amplified placebo effect has been found and therefore reboxetine is no better than placebo. If apparent antidepressant efficacy is due to an amplified placebo effect (see previous post) one might have expected the same to be found with reboxetine. As it hasn't, it is already being said (see post by Neuroskeptic) that the amplified placebo hypothesis must be wrong.
But there could be other explanations, such as the lower expectancy for reboxetine. In other words, the placebo effect was not amplified because there was a lack of belief in reboxetine. Also, in another rapid response to the BMJ article, it is pointed out that the BMJ meta-analysis may have been selective, certainly compared to the NICE analysis.
It'll be interesting to see what NICE make of the BMJ article. Hopefully, we can also find out what was behind the FDA decision in 2001.