What’s a Sponsor to Do?: The Curious Case of “Disputes” Over Phase 3 Study Design

End-of-Phase 2 (EOP2) is, in our view, one of the most critical moments in drug development.  It’s the moment at which a drug sponsor selects, and seeks FDA agreement on, its critical Phase 3 study parameters:  dose, eligibility criteria and endpoints, to name a few.  Errors or miscalculations in choosing these design components can send a safe and effective drug to the very expensive graveyard of failed Phase 3 studies.

So, what’s a sponsor to do when, at EOP2 and in subsequent interactions, it cannot reach agreement with the FDA review division on a critical parameter; let’s say the primary endpoint?  Submit a formal dispute resolution to bring the decision to the Office level, you say?  While that seems like a reasonable approach before spending millions of dollars on Phase 3, that option is not available to sponsors.  You see, in 2015, FDA revised its 15-year old FDRR guidance document to exclude EOP2 disagreements from the FDRR process (i.e., “Advice communicated in meeting minutes and other correspondences is not a regulatory action taken by CDER or CBER; therefore, it would not be an appropriate subject for a formal dispute resolution request (FDRR) by a sponsor.”).

In limiting the scope of FDRRs in this way, the Agency comforted itself that meeting minutes and advice letters (regardless of how strongly worded) convey mere “recommendations and/or advice made to a sponsor” and, as such, “[s]ponsors are not bound by such recommendations and/or advice.”  Of course, that’s true – at least as a technical matter.  The practical truth, however, is that failure to follow the Division’s recommendations for Phase 3 design puts the program at significant risk of failure even if the Phase 3 study is successful.  We’ll spare you the countless references to FDA advisory committee briefing documents in which a review division quotes its own earlier minutes or advice, pointing an accusatory finger toward the recalcitrant applicant who didn’t follow that advice.  Suffice it to say that a Phase 3 program that rebuffs FDA advice is a Phase 3 program fraught with peril – which, not surprisingly, is a Phase 3 program that investors will not fund.

We were so alarmed by the predicament created by the 2015 change to the FDRR guidance that we raised the issue in a comment to the docket.  There, we provided a hypothetical test case in which FDA meeting minutes accurately convey the division’s disagreement with the proposed Phase 3 study endpoint: “No, we do not agree. In order for your studies to provide substantial evidence of efficacy, we strongly recommend that you demonstrate an improvement over placebo in [insert FDA-preferred endpoint].” In our hypo (which, for many clients, reflects reality) further discussions with the division suggest that the disagreement cannot be resolved.  Unable to appeal, the sponsor conducts Phase 3 using its preferred endpoint rather than FDA’s and those studies are clinically and statistically successful.  We observed

the sponsor is already on notice that the review division will not consider these studies sufficient to demonstrate efficacy. It will request a pre-NDA meeting at which the division is likely to restate its view and refer back to the EOP2 minutes noting that the sponsor failed to follow previous advice. The sponsor, still unable to avail itself of the FDR process, must spend the resources necessary to prepare and submit a marketing application, and pay a user fee in excess of $2,000,000. FDA may issue a refusal to file (RTF) letter, which will restate that the endpoint is not appropriate. Because the FDRR process is not available for RTF actions, the sponsor must file over protest. Whether the application is voluntarily filed by FDA or filed over protest, the sponsor must wait 10 additional months, and face the prospect of an advisory committee which will consume significant additional resources, prior to receiving a complete response letter (CRL) stating for at least the fourth time that the endpoint was inappropriate. An appeal is still not available. The sponsor must instead request and attend another meeting with the division at which it will, for the fifth time, be told (likely with some warranted frustration by the review division) that the endpoint is not appropriate. Only at this point, three to four years after the original disagreement with the division, can the sponsor appeal under the 2015 Draft Guidance. If the appeal is successful in determining that the alternative endpoint is appropriate, the sponsor must assemble an NDA resubmission and undergo another six month review clock.

Well, despite our comment, FDA chose to finalize the guidance unchanged and since that moment has refused requests for formal appeals of disputes that arise at EOP2.

But why are we rehashing this 5+ year old history today?  In part, it may be that our egos are bruised or that we fester over having been right every time a potential client approaches us with a failed study that would have succeeded had it been able to convince FDA about a particular design element at EOP2, but mostly, it’s because we have identified a very important loophole.  A loophole so large that we’re not sure why we didn’t see it all along.  You see, what is appealable is a Special Protocol Assessment (SPA) No Agreement Letter.  And while conventional wisdom says that one is most likely to submit a SPA if it is possible to achieve  concurrence with FDA on the adequacy of study design, who’s to say that one can’t submit a SPA request in order to confirm a lack of such concurrence, thereby opening up the route to FDRR.

The truth is, we continue to believe that efficient development of novel therapies would be best served by changing the FDRR guidance to permit formal disputes of Phase 3 study design issues.  Since that doesn’t seem likely at this point, sponsors facing division-level disagreement on Phase 3 study design at EOP2 should consider whether to invest the time and resources necessary for the SPA and FDRR processes prior to initiating Phase 3.  While this could delay Phase 3 initiation by 4-5 months, it could accelerate (or make possible) time to ultimate drug approval.