When a drug company's New Drug Application runs into some sort of major issue, the FDA sends what's called a CRL, a Complete Response Letter. This details why the application was rejected and what can be done to fix the problem. And all this is, naturally of great interest to the company's investors.
But they can go mow their lawns. Most of the time, that's going to provide as much information as the company will divulge on its own. That's what I took away from this study in the BMJ (here's a summary at FierceBiotech). The authors (who are with the FDA) looked at the full text of dozens of CRLs (from 2008 to 2013) and compared what's in them to the press releases at the time. What results is a sort of regulatory Rashomon.
18% of the time, the companies issued no press release at all, although you would have to characterize a CRL as a "material event", wouldn't you? Another 21% of the time, though, press releases "did not match any statements from the letters", so that's the next best thing to not issuing anything at all. 32 of the CRLs from this period called for new clinical trials to be conducted, but 40% of the press releases never got around to mentioning that. And of the 7 CRLs that brought up a higher mortality in the drug treatment group, only one of the corresponding press releases noted that fact. In none of the 61 cases did the company involved release the complete text of the CRL.
So, to quote that noted philosopher Bullwinkle, if you're waiting for the companies to finish the story, you're going to have a long wait. There have been calls over the years to make CRLs public, and this study shows why that's not a bad idea. Since the authors themselves are from the FDA, and note that this proposal was recommended by the agency's own transparency task force in 2009.
At the very least, a summary of the CRL could be released by the agency in its own words - something that wouldn't disclose any proprietary details, but would at least tell everyone what's going on. Does the FDA need anyone else's permission to do this? The authors say that releasing the full texts "would likely require a change in FDA’s regulations". (Does that make it a matter for Congress to take up?) Someone needs to, because the current system is just not right.