The FDA AdComm Reboot: Part 1; Lessons from the Meeting on Capricor’s Deramiocel

After more than a year with nary an adcomm (just 2 drug approval-related meetings in the Makary era), CBER’s Cellular, Tissue, and Gene Therapies Advisory Committee meetings came back with a vengeance last week with back-to-back meetings to discuss two seemingly controversial cell and gene therapies that received complete response letters (CRLs) and are now under review following BLA resubmissions: Capricor’s Deramiocel for DMD and Replimune’s RP1 for advanced melanoma.

Covered in Part 1 of this FDA Law Blog series is the July 29, 2026 meeting on Capricor Inc.’s Deramiocel for the treatment of cardiomyopathy in patients in Duchenne muscular dystrophy (DMD).  Capricor previously submitted a BLA on December 31, 2024, based on results from a Phase 2 randomized trial (HOPE-2) and open label extension (HOPE-2-OLE).  FDA granted priority review but ultimately issued a CRL on July 9, 2025, finding a lack of substantial evidence of effectiveness.  At issue in the recent adcomm was Capricor’s BLA resubmission, and whether the data from its Phase 3 randomized, double-blind, placebo-controlled HOPE-3 study provided substantial evidence of effectiveness. A 9 to 3 vote by the advisory committee opined that it did not, at least for the cardiomyopathy indication (versus the broader impact on skeletal muscle which was not asked directly).  The PDUFA date for Capricor’s BLA resubmission is set for August 22, 2026.

The FDA’s review principally focused on data analysis and the statistical analysis plan (SAP).  They highlighted that decisions made on imputation, exclusion of certain patient data, and changing the primary endpoint from an absolute change to percent change, with key changes in primary endpoint and methodology made after the last subject visit.  FDA further noted the potential for functional unblinding from the AE profile and cross-over into the open-label extension.

Because of these concerns, in its briefing book and presentation, FDA relied on version 1.1 of the SAP, prior to when substantial data collection began (i.e., first subject, first visit), to analyze the data and found a lack of statistical significance.  While shifting to version 1.1 of the SAP seems to be a draconian and overly conservative analysis of the data, particularly in a rare and devastating disease like DMD where unmet needs are substantial and regulatory flexibility is warranted, the advisory committee members appeared concerned by these issues.  In their discussion, the committee members noted that whether statistical significance was reached appeared to hinge on the inclusion or exclusion of certain data or imputations from just a few patients – leading to what the committee members dubbed as data “fragility.”

Without opining on the merits of FDA’s assertions about Capricor’s program, the critiques the agency raised provide important insights for sponsors:

• Focus on the SAP from the get-go. Now this may seem obvious, but for many small companies with limited resources, most of the time and attention is spent on the clinical trial protocol – debating the sample size, single-arm vs. randomized, open-label vs. blinded, which controls to use, and what endpoints to measure.  The statistical analysis plan at this stage is mainly focused on ensuring the trial is adequately powered.  It’s often not until later, much (MUCH) later, when many of the important details for the SAP are hashed out.  Questions like how to handle missing data, what imputations (if any) should be made, and when it may be appropriate to exclude data, should all be considered carefully and pre-specified in the SAP, ideally before the start of the trial, but at least as early as is feasible.  Of course, not all issues can be forecasted but as issues arise during the conduct of the study, the SAP should be evaluated and updated in much the same way as the protocol.
• Avoid major changes to the SAP once the trial is underway, if at all possible. If major changes are unavoidable, ensure there is a very strong scientific justification for making the change and have that documented with the agency (in addition to the change itself).  Prior to submitting the BLA, sponsors should consider running the analysis prior to the SAP changes as an internal exercise to prepare for what the agency may do during a BLA review and prepare in advance for questions that may arise.
• Consider getting FDA alignment on the SAP – especially if major changes are made – or at least try to. While FDA does not need to approve the SAP, it may nevertheless be in a company’s best interest to actively seek alignment, particularly if there are aspects of the SAP the company predicts may raise questions.  As noted by Capricor, versions of the SAP were submitted to its IND and the agency did not provide comments objecting to the changes at the time, however this didn’t preclude the agency from later rejecting the changes. Seeking FDA feedback requires time, as FDA has taken 6+ months to return comments on a SAP submitted to an IND.

The reality is that we learn about rare diseases as we study them, and medical science advances at a rapid pace in parallel.  However, incorporating these learnings in analysis plans in real time needs to be weighed against the real risk that became apparent on July 29: that FDA may void those changes outright.