An article recently published in The BMJ (formerly known as the British Medical Journal), titled “Comparison of content of FDA letters not approving applications for new drugs and associated public announcements from sponsors: cross sectional study,” and co-authored by several FDAers, drew a lot press (see, e.g., here, here, and here). The article describes the results of an internal FDA review of 61 Complete Response Letters (“CRLs”) issued between August 11, 2008 and June 27, 2013 (48 for NDAs and 13 for BLAs) attempting to compare the reasons identified in those CRLs for not approving a product with the respective applicants’ public statements regarding the CRL. The article includes a lot of statistics and little context; for example:
- NDAs were less likely than BLAs to have a press release associated with a CRL, with a relative risk of 0.79 (95% confidence interval 0.66 to 0.95); however no information was provided on whether the CRL was a material event to the applicant in more of the BLA cases than the NDA cases.
- Of the 33 CRLs mentioned in both press releases and U.S. Securities and Exchange Commission filings, 7 of the filings (21%) included a statement with more information from the CRL than the press release; however, no information was provided on how many of the CRL applicants were subject to SEC reporting requirements.
- Of 32 CRLs calling for a new clinical trial (for safety or effectiveness), 19 (59%) had “matching” press release statements; “matching” was defined as “[a]ny press release statements that covered the same issues as statements in corresponding [CRLs]”;
- Press releases were not issued for 11 of 61 CRLs (18%) sent between August 2008 and June 2013; however, again, no information was provided on whether the CRL was a material event to the applicant in more of the BLA cases than the NDA cases;
- Press releases did not match CRL statements in 13 instances; and
- Press releases matched 93 of the 687 reasons (14%) FDA stated in a CRL for not approving a product; however, no information was provided on how many of the FDA’s stated reasons were material.
Such a lack of context renders these statistics largely uninformative – but perhaps the article is more focused on advancing FDA’s years-long push to circumvent applicable FOIA exemptions by pressuring industry to make CRLs public. The authors acknowledge as much near the end of the paper in a section titled “Policy considerations.” There, the authors state:
Our analysis of the content of press releases indicates that they are incomplete substitutes for the detailed information contained in [CRLs]. Disclosure of [CRLs] would allow the FDA to increase the overall transparency of its regulatory processes, providing greater awareness of the agency’s role in protecting health and combating misperceptions regarding the basis for non-approval of a drug. It would also allow for broader and more informed public discussion by relevant stakeholders (such as patients, clinicians, researchers, and public health advocates) of the scientific and regulatory reasons for the FDA’s actions. The need for increased transparency, however, must take into consideration the legal requirement to protect sponsors’ trade secrets and confidential business information.
FDA’s campaign to obtain public release of CRLs spans at least 5 years but appears to be growing in intensity. Back on May 19, 2010, FDA’s Transparency Task Force, chaired by then-Principal Deputy Commissioner Joshua Sharfstein, issued a report with 21 proposals that would allegedly improve transparency across a variety of FDA activities. Proposals 13 states:
FDA should disclose the fact that the Agency has issued a refuse-to-file or [CRL] in response to an original NDA, BLA, or an efficacy supplement for an NDA or BLA at the time the refuse-to-file or [CRL] is issued, and should, at the same time, disclose the refuse-to-file or [CRL], which contains the reasons for issuing the letter.
As legal experts noted after the May 2010 release of FDA’s Transparency Task Force report, the Agency’s “application and data disclosure proposals in particular raise serious concerns for many regulated companies which consider product application information to constitute commercially sensitive trade secret information, the disclosure of which could give competitors significant advantages.” They continued, stating that “[t]he Task Force largely dismisses such concerns by concluding that such information is not competitively useful, and stating that even though FDA regulations have historically shielded application-related information from disclosure, ‘a substantial amount of [such] information…does not fall under FDA’s definition of trade secrets as confidential commercial information.’” Not competitively useful? We beg to differ. We seem to recall a similar position during discussions as to whether or not certain bioequivalence studies intended to support ANDA approval should be reported on ClinicalTrials.gov (see our previous post here). Some bioequivalence studies are currently reported and their listing and status are regularly monitored for competitive purposes. CRLs, if they’re made public, would also be used to gain a competitive advantage.
Since May 2010, FDA has made clear its continued interest in the topic of releasing CRLs. In 2011, FDA released, with the sponsor’s permission, parts of a CRL ahead of an advisory committee meeting, causing a feeding frenzy among some in the press (see here). The issue popped up again in February 2013 when then-FDA Commissioner Margaret Hamburg briefly indicated during a congressional briefing sponsored by Faster Cures and the Friends of Cancer Research that FDA is working with industry to find a way for it to make CRLs public (see here and here).
The interest, however, may stem more from Agency frustration at having to remain moot in the face of some applicants’ statements regarding overly conservative or apparently arbitrary Agency action rather than from a public-health-driven concern about transparency, as is apparent both in CDER presentations (see here) and remarkable recent FDA statements skirting the lines established by FOIA (see, e.g., here). While FDA’s frustration is palpable, and maybe even justified in some situations, it’s certainly not sufficient reason to undercut one of the basic tenets of drug development – that interactions with the Agency on an unapproved product (including the very submission of an NDA) are not disclosable by the Agency.
The BMJ authors suggest 3 possible approaches to greater parity between CRLs and press releases: (1) sponsors could release CRLs themselves; (2) sponsors could issue “more complete press releases; or (3) FDA could release CRLs” – but note that option 3, they note, “would likely require a change in FDA’s regulations.” Indeed.
The easier path then would be for FDA to exert pressure on applicants (or their trade groups) to arrive at option 1 – voluntary release of CRLs. And perhaps The BMJ paper, rather than being an unbiased review of all applicable considerations by experts in FDA matters, SEC disclosures, fiduciary duties and competitive harm, is more of a directed advocacy piece by a hardly unbiased agency.
Beware of caving into the pressure readers! CDER has already signaled that it will not stop at CRLs but that it wants authority to release meeting minutes next.