FDA Panel Votes 9-3 Against Capricor’s DMD Heart Drug

FDA's Cellular, Tissue, and Gene Therapies Advisory Committee voted 9 to 3 on 29 July 2026 that the available evidence does not provide substantial evidence of effectiveness for deramiocel, Capricor Therapeutics' cell therapy for cardiomyopathy in Duchenne muscular dystrophy (DMD). The meeting, held at FDA's White Oak campus, was the second time this application has come before the agency's reviewers — and the second time the underlying data has drawn open dispute, this time over which statistical analysis plan governs the reading of the pivotal trial's cardiac results.

What the committee was actually asked

FDA's voting question was narrower than the indication Capricor originally proposed: it asked only whether the evidence supports effectiveness of deramiocel for cardiomyopathy in DMD, not whether the product's overall benefit-risk profile favors approval. Nine of twelve voting members said no. Three said yes. None abstained. That framing matters because a narrow, unfavorable answer on effectiveness forecloses less than a broad one would — but it still lands squarely on the question FDA's July 2025 Complete Response Letter had already flagged as unresolved: whether Capricor's data meets the statutory substantial-evidence-of-effectiveness standard at all.

The fight over which analysis plan counts

Deramiocel's case rests on HOPE-3, a randomized, placebo-controlled Phase 3 trial that Capricor has said met its primary endpoint, a measure of upper-limb function, and showed benefit on cardiac secondary endpoints including left ventricular ejection fraction. The 29 July meeting reopened how those cardiac results should be read. FDA's review relied on an earlier, prespecified statistical analysis plan; Capricor's position, restated at the meeting, is that a later version, SAP 3.0, was finalized before the trial was unblinded and should be the version that controls interpretation of the cardiac endpoints. Which SAP governs is not a cosmetic dispute — different prespecified plans can produce different answers to the same question from the same dataset, and the committee's discussion of missing data and secondary-endpoint handling ran alongside that disagreement rather than resolving it.

A second rejection, on a different ground

This is deramiocel's second pass through FDA review. The original BLA drew a Complete Response Letter in July 2025 that cited both insufficient substantial evidence of effectiveness and unresolved chemistry, manufacturing, and controls (CMC) items. Capricor resubmitted with the new HOPE-3 data; FDA classified the filing as a Class 2 resubmission, restarted its review clock, and lifted the CMC-specific concerns on the way to setting the current PDUFA target action date of 22 August 2026. The 29 July vote does not revisit CMC — it is squarely about whether the clinical evidence, on FDA's preferred statistical footing, clears the effectiveness bar. A second rejection on efficacy grounds, even a non-binding one, is a harder problem for a sponsor to out-negotiate than a manufacturing deficiency: there is no protocol amendment that retroactively changes which analysis plan was locked before unblinding.

What happens next

FDA is not bound by the committee's vote and can still approve, issue a further Complete Response Letter, or ask for more information by the 22 August PDUFA date. The agency's own briefing document ahead of the meeting had already raised efficacy concerns consistent with Tuesday's outcome, which makes the vote confirmatory of the agency's stated doubts rather than a surprise reversal of them. For sponsors watching this docket, the operative fact is procedural: a narrow voting question, a split that broke against the sponsor on the precise issue the prior CRL raised, and three weeks left on the clock before FDA has to decide what a non-binding 9-3 vote is worth.