9–3 Vote Finds Efficacy Evidence Insufficient, Dealing DMD Cell Therapy Deramiocel a Setback in FDA Review

For patients with Duchenne muscular dystrophy, cardiomyopathy is one of the most serious threats in the later stages of the disease, yet there are no proven treatments that can slow the deterioration of cardiac function. Capricor Therapeutics is seeking to fill this gap with an allogeneic cell therapy, but on July 29, a U.S. Food and Drug Administration (FDA) expert committee voted 9–3 that the company had not yet provided sufficient evidence that deramiocel is effective.

The vote concerned Biologics License Application BLA 125842. Deramiocel is made from cardiosphere-derived cells cultured from donor heart tissue, and the developer hypothesizes that it may treat cardiomyopathy associated with Duchenne muscular dystrophy through anti-inflammatory, antifibrotic, and immunomodulatory effects. The committee’s recommendation is not legally binding, and the final decision remains with the FDA; the current target review date is August 22.

At the heart of the dispute is not a single number, but which analysis adequately represents the Phase 3 HOPE-3 trial. The study enrolled 106 male patients aged at least 10 years, most of whom were no longer able to walk. Participants received deramiocel or placebo once every three months over 12 months. Using an earlier prespecified method, the FDA concluded that neither of the two primary endpoints—the Performance of the Upper Limb 2.0 scale, or PUL 2.0, and left ventricular ejection fraction—reached statistical significance, while the other secondary endpoints also failed to produce a consistent efficacy signal.

Capricor, by contrast, argued that the final statistical analysis plan, SAP 3.0, had been completed before the data were unblinded. Under that method, the PUL 2.0 result reached statistical significance, while the cardiac function data also provided support. The company also said that SAP 1.1, which the FDA used, was merely an incomplete internal draft that later became invalid after a new trial population was added. The FDA’s concern was that as the analysis versions changed, the endpoint definitions, handling of missing data, and statistical methods were also adjusted, causing the magnitude and significance of the observed effect to fluctuate substantially and making it difficult to demonstrate that the results were robust.

The safety data added another layer to the statistical dispute. In HOPE-3, 22 participants in the deramiocel group, or about 42%, experienced hypersensitivity reactions, compared with 8 participants, or about 15%, in the placebo group. The FDA said the distinctly different adverse-reaction patterns between the two groups may have allowed patients or investigators to guess the treatment assignments, resulting in “functional unblinding.” This possibility particularly undermines the credibility of the results when some analysis plans were adjusted only after the randomized, double-blind phase had ended.

The earlier HOPE-2 study also failed to provide the confirmatory evidence the FDA required for the application. Reviewers noted that the study did not demonstrate improvements in skeletal muscle or cardiac function, and a previous application had also received a Complete Response Letter because of insufficient efficacy evidence and an unfavorable benefit-risk assessment. The 9–3 vote does not directly determine deramiocel’s fate, but it clearly shows that most committee members do not believe statistically significant results that were not prespecified in the original plan are sufficient to meet the evidentiary threshold required for approval.