Putting the “Complete” Back into Complete Response Letters

A biotech company facing a complete response letter (CRL) action on its NDA/BLA has no greater goal than to quickly and fully understand the deficiencies that FDA has identified in the application.  Such an understanding is critical to addressing the review division’s findings through additional data or analyses, and is even more essential should the company choose to appeal those findings through Formal Dispute Resolution (FDR).

The purpose of a CRL is to communicate to the applicant that FDA will not approve the application in its present form, and, with limited exception, the CRL describes all of the deficiencies that must be satisfactorily addressed before the application can be approved.  21 C.F.R. § 314.3.  A CRL is, by its nature, a summary document that abbreviates the many months of review and independent analyses performed by a number of FDA disciplines such as medical, statistical, and clinical pharmacology, into a handful of pages.  The actual detailed work performed by the FDA reviewers is embodied in various highly informative review documents that, by contrast, typically span several hundred pages.

The applicant receives the CRL but is not provided the more instructive underlying reviews.
Because of its typical brevity, the CRL is limited to a high-level description of deficiencies and suggested actions for addressing them.  It cannot encompass all of the nuanced information needed to fully appreciate the division’s view or the basis for that view.  FDA regulations offer the opportunity for a subsequent End-of-Review (EOR) meeting, which the Center for Drug Evaluation and Research (CDER) requires an applicant to attend as a prerequisite for appeal under the FDR process.  Despite consuming significant Agency resources that are already stretched thin, these meetings are also often just too short to satisfactorily communicate the details of what may be more than one complex issue.  FDA reviewers simply cannot be expected to articulate hundreds of pages of reviews, including the methods and results of any statistical or pharmacokinetic modeling, in one hour.  In addition, the Q&A format of CDER meetings hampers the exchange, making it dependent on the applicant having sufficient understanding from the CRL to articulate questions that will elicit detailed responses about critical issues.

The lack of clarity can result in deep frustration and misunderstanding as applicants address what they have understood to be the basis of FDA’s concern, only to learn that there are one or more additional bases.  Regulated companies, having spent months attempting to address a deficiency, can feel as though FDA is constantly “moving the goalposts.”  The reviewers for their part can become frustrated with a company that “just doesn’t get it.”  In our experience, these perceptions often don’t reflect reality.  Instead, the FDA reviewers are acting in good faith, but the clarity of the direction they can provide (and therefore the ability of the applicant to understand it) is hampered by the brevity of the CRL and EOR meeting.  While not the FDA reviewers’ intent, the applicant may find itself trapped in a game of regulatory whack-a-mole at a moment when resources are dwindling, and investors are losing faith.

What an applicant really needs, in addition to the CRL, are the FDA reviews themselves which, conveniently, have already been drafted and finalized and which are likely the only documents that can communicate exactly what FDA is seeing in the data.  Failure to gain access to these comprehensive reviews necessarily handicaps an applicant’s appeal.  Access to the reviews could aid some would-be appellants in more fully appreciating the reviewer’s point and choosing not to appeal.  In other cases, such access would aid the appellant in understanding the emphasis being placed on various analyses and pointing out any flaws in those assessments.

Without being overly dramatic about it, failure to provide the reviews to the applicant strikes us as fundamentally unfair in addition to being inefficient.  As a general legal matter, it is a well-accepted principle of administrative law that when an agency relies on scientific and technical data, it must provide adequate information regarding those data to allow critique of them.  Banner Health v. Price, 867 F.3d 1323, 1335 (D.C. Cir. 2017); United States Lines, Inc. v. Federal Maritime Com., 584 F.2d 519, 534 (D.C. Cir. 1978).  For that reason, when the Agency elects to rely on, for instance, a statistical simulation or a correlation it discovered among different adverse events, it must disclose the details of it.  Unlike a citation to a publicly available study, a reference in the CRL or EOR minutes to an FDA-conducted analysis which exists only in FDA’s files provides inadequate notice and is improper and unlawful.  National Classification Comm. v. U.S., 779 F.2d 687, 695 (D.C. Cir. 1985) (“The agency cannot, however, rely on data known only to the agency . . . .”).

The point is perhaps best made by considering CDER written responses to FDR requests (whether granted or denied) which uniformly list those documents that form the basis of the appellate officer’s thinking.  In our experience, those responses contain a near boilerplate sentence that reads something like this: “I have carefully reviewed the materials you submitted in support of your appeal, as well as the reviews, meeting minutes, and decision memoranda prepared by FDA staff along with the CRL” (emphasis ours).  To be clear, this indicates that the deciding official has been presented information about the case from one side in the dispute and that the opposing side has not been granted access to that material.  By its very nature, this suggests that all facts needed to understand whether the review division appropriately denied approval were not housed in the CRL and EOR minutes and were not made available to the applicant.

Our understanding (based to some extent on Agency lore) is that CDER does not share the underlying reviews with the applicant because it believes that disclosing them to the applicant would make the documents disclosable, at least in some respects, to third parties, under the requirements of the Freedom of Information Act (FOIA). We believe such an interpretation is incorrect, and that FOIA case law does not require that outcome. Moreover, we struggle to understand this interpretation by the Agency and to distinguish how it applies to other CDER-generated documents such as, for instance, the summary minutes of the EOR meeting which are uniformly provided to the applicant and not to the broader public.  Both the FDA summary reviews and the minutes seem to fall within 21 C.F.R. § 314.430 and yet their release to the applicant is handled differently.

Without the benefit of access to the complex analyses and thinking that underlie a CRL, an applicant may be denied the ability to efficiently move its program forward, and may spend significant time and money, or make the decision to abandon a program, based on incomplete information – despite the existence of fully developed and internally vetted detailed reviews.  FDA’s public health mission is not promoted by unnecessarily withholding information that could be used to more efficiently move new drugs into an approvable position (or have sponsors make fully informed decisions to halt programs for products that are destined to not be approved).

We believe that a modification in CDER policy to allow the applicant access to the underlying reviews could change the post-CRL process for the better for CDER and for the CRL recipients.  At a minimum, that information would reduce the multiple requests to review divisions to provide further clarification, thereby reducing the drain on resources.

We would welcome a public dialogue regarding such a potential policy change as part of the Agency’s thinking on increased transparency.