Saturday, May 26, 2018

Bad Blood: Did FDA Inappropriately Leak Non-Public Theranos Information to WSJ?

There is a new boom of Theranos stories tied to the May 22 release of John Carreyou's book on the Theranos saga and the role played by the WSJ.   (See here and here; there are dozens of other stories.  Matthew Herper discusses the book from the viewpoint of a fellow journalist who was also covering Theranos.)

One thing I picked up on were Carreyou's discussion of the senior offical at FDA who provided WSJ with "background" on the FDA technology reviews and Theranos investigations, apparently before they were public. 
  • Before quoting from Carryou's book, let's recall how the government views leaks of FDA and CMS information:  
  • In 2014, WSJ wrote several articles about prepublic leaks of CMS policy information (here and here).  
  • In May 2018, four individuals convicted in federal court of participating in CMS prepublic information leaks (here).   
  • Bloomberg ran a 2018 article about how upset CMS administrator Jonathan Blum was when leaks occurred at his agency, and how he would read his staff the riot act about agency information leaks were illegal (article here, transcript here.)
Now, let's look at what Carreyou writes in his new book.

FDA Source Seems to Make Inappropriate Leak
On page 310, "I called a source of mine who was high up int he FDA's medical products division."  Initially, the FDA official agrees with Carreyou's insight that a then-new FDA approval of a single Theranos herpes test was a one-off approval.  That sounds fine: it's a viewpoint on a public fact.

But then the FDA official added: "the clinical data the company had submitted to the agency for [other tests] were poor and wouldn't pass muster."  OK, stop right there.

That sounds like speaking out of school, by releasing confidential FDA/company information.  Surely release of that information is hardly a safe harbor just because it is "on background" and to a reporter.

I've always understood that FDA can't release prepublic, decisional information about a product under its review.  For example, an FDA reviewer can't tell a reporter or an investor that he's reviewing a Phase III Lilly drug based on its product data files, and that next week, the product will be disapproved by the FDA.

Finally, the source then pointed out to Carreyou that it was hard for FDA to take action against Theranos, because Theranos was chummy with the Obama administration.   I'll give that a pass.  It's not secret information about a company's submissions that couldn't be released, for example, under FOIA.  Senators and other officials regularly remark that can't do "X" because of some aspect of public opinion or party positioning. The remark might make someone at Theranos, at FDA, or at Executive Branch squirm, but it's not secret information.

The FDA Source Comes Back with More Secret Leaks
The source comes back on page 321/322.  Carreyou had had trouble reaching him by phone, we're told, but "this time he picked up the phone."  Now speaking "on deep background" the FDA senior official tells the reporter that the FDA "had recently conducted a surprise inspection of Theranos's facilities in Newark and Palo Alto....the agency had declared the nanotainer an uncleared medical device and forbidden Theranos to continue using it."   If that wasn't yet public, it doesn't sound like it should be made public to reporters who dial up.

On this occasion, the FDA official had more information for the reporter.  The FDA informant told Carreyou that CMS was inspecting Theranos as well.   Carreyou writes that he quickly worked up this information for the next day's paper (October 16, 2015).   (Later, CMS CLIA would require Theranos's lab closed and barred Elizabeth Holmes from lab ownership/management.)

I'm not sure on what day the FDA suppression of the Nanotainer was publicly released by FDA, but the FDA official would hardly be speaking of the issue on "deep background" if it had already been released by the FDA.   (Data on FDA inspections was publicly available by October 27, 2015).

While Theranos may have had a lot to feel guilty about, its paranoia was in fact partly real, if senior FDA officials were giving confidential internal agency decisions  to reporters (or for that matter, to hedge funds). 

Rereading the October 16, 2015 WSJ Article Today:  
Differences Suggest Access to Insider FDA Staff, But No Access to Insider CMS Staff

The October 16 WSJ article is attributed overall to "people familiar with the matter."  At the time, that might have been recent, or even current, Theranos employees; hypothetically, it could have been Elizabeth Holmes' admin.   But now, in the book, we learn the source was literally a senior FDA official.

I also noted this: Regarding inside CMS info, Carreyou made the point on October 16 to tell his readers that he couldn't confirm the CMS investigation status with CMS staff - they refused to talk to him about internal matters of CMS.  But regarding inside FDA info, he didn't feel he needed to tell his readers that the FDA information wasn't confirmed by FDA - which implies by comparison he did have FDA confirmation (albeit off record) of the FDA actions. 

Here's the bottom line.  Regardless of whether the products were those of a publicly traded company (like Siemens) or a private company, I don't think FDA was supposed to release confidential internal file information to anyone, neither a reporter nor an investor.  I suspect from the Journal's viewpoint, it's like leaked Pentagon Papers or leaked Democratic National Committee emails; it might be illegal to leak them but once they are leaked, they become news for all.


For Carreyou book screen shots (p. 310, 321, 322) here.

Carreyou discusses his "deep background" FDA informant in a podcast Inside the Hive, released June 8, at minute 9; see here and here.