Antidepressants and Efficacy

In some cases, it's worse than we thought.

Posted Apr 08, 2016

We have long known that so-called “antidepressants” are not highly effective because they fail so many of their licensing trials for the Food and Drug Administration. When a company develops a drug, they have to submit it, along with all of their clinical trial data, to the FDA. Routinely, these antidepressant drugs fail about half of their trials.

This has long been known. What was not known, until recently, was just how badly some of those drugs that received FDA approval did in their trials. This came out almost inadvertently at the December 1, 2015 meeting of the Psychopharmacologic Drugs Advisory Committee of the FDA to consider for approval the Fabre-Kramer Pharmaceuticals Company’s New Drug Application for its proposed antidepressant gepirone, the extended-release version of which the company hopes to market as Travivo.

Well, in the trials, gepirone had not done very well. It had passed with flying colors in only two of the 12 trials (all conducted with random control groups; and in five of the 12, gepirone went head to head against a standard antidepressant already on the market.)

Now, two out of 12 is not fabulous. But, hey, the FDA only insists that a candidate drug passes two trials. What is interesting in all this: FDA officials brought out the statistics on the 18 drugs already approved for “major depressive disorder.” These 18 drugs had received 46 trials. In 15 of the 48, neither the control drug nor the candidate drug beat placebo. In eight of the 48, the test drug beat placebo but the active control didn’t.

When you add all the “negative trial” and “failed trial” data together, these 18 fabulous antidepressants failed to beat placebo in 23 of the 46 trials. That is 50 percent. These data have not, to my knowledge, ever been made public and I extracted them by reading the minutes of the December 1 meeting.

For four of the big-name antidepressants, the percent of positive studies was only around 30 percent. In other words, in almost three-quarters of the trials, these subsequently marketed drugs were unable to improve on sugar pills, either because the trial population was subsequently judged incapable of responding properly (lacking in “assay sensitivity”) or because the candidate drug failed to best placebo. The four were vilazodone, duloxetine, desvenlafaxine, and citalopram. All, as I said, were marketed.

The members of the PDAC voted nine to four against approval on December 1. This speaks well to the composition of these committees. Their members are, in general, respected figures. 

Nobody went into too much detail about the history of gepirone, because it’s kind of scary. The Bristol-Myers Company (later Bristol-Myers-Squibb) synthesized it in 1983 and sought a patent in 1987 for the treatment of “panic disorders”. Bristol-Myers did some clinical trials, including depression, noted the poor results, and sold the compound to Fabre-Kramer, which in turn sold it to the NV Organon Company in the Netherlands. Organon did some more trials, then submitted the drug to the FDA, where it was turned down. Fabre-Kramer then re-acquired the rights in 2005. By this time the “panic” drug had turned into a “depression” drug.

On three separate occasions, in 2002, 2004, and 2007 the FDA turned hands down on gepirone. 

So, after that fateful December 1, 2015, meeting of the PDAC, it looked as though gepirone was toast for a fourth time.

But on March 3, 2016, John Jenkins, the director of the agency’s Office of New Drugs, decided to ignore the advice of the Psychopharmacologic Drugs Advisory Committee and approve gepirone.

The funny shift from a panic to a depression drug, the litany of failed and negative trials—all no matter. Fabre-Kramer was free to market gepirone. Why Jenkins took this decision is unknown.

But there is a larger issue here for people out there who display a credulous confidence in the wisdom of our regulatory authorities. Drug development is an exercise in marketing, not in science. Industry decided that the depression market was bigger than the panic market. As well, drug development is an exercise in brutal, plodding persistence. 

FDA is starting to smell like a pile of dead fish after a week in the sun.