One Phase 3 Trial, Two Statistical Conclusions: FDA Questions Efficacy of DMD Cell Therapy

For patients with Duchenne muscular dystrophy, cardiomyopathy is one of the most dangerous aspects of disease progression. However, as a cell therapy intended to protect cardiac and muscle function enters a pivotal stage of U.S. regulatory review, the FDA and developer Capricor Therapeutics have reached starkly different conclusions about whether the efficacy evidence holds up.

In review documents released on July 27, the FDA said the Phase 3 HOPE-3 trial of deramiocel failed to meet its prespecified primary or secondary efficacy endpoints after 12 months of treatment; none of the comparisons with placebo showed a statistically significant difference. Reviewers therefore concluded that neither this trial nor the earlier HOPE-2 study was sufficient to constitute substantial evidence of effectiveness, and raised questions about whether the overall benefit-risk profile was favorable.

deramiocel is made from cells derived from the heart tissue of healthy donors, and its proposed indication is the treatment of cardiomyopathy associated with Duchenne muscular dystrophy. HOPE-3 also used the PUL 2.0 scale to assess upper-limb function. The FDA noted that Capricor changed the endpoint definition, statistical methods, and strategy for imputing missing data after the randomized trial was completed, making the subsequent positive conclusions difficult to regard as the direct result of the originally planned analysis.

Capricor countered that version 1.1 of the statistical analysis plan relied upon by the FDA was merely an unsigned, incomplete internal draft and had become obsolete after Group B participants were added to the trial. The company argues that version 3.0, finalized before unblinding, should be the governing plan; under that methodology, HOPE-3 achieved a statistically significant improvement on the primary PUL 2.0 endpoint, with cardiac function data also providing support. The dispute therefore concerns not only the figures themselves, but also which analysis plan had prospective validity and whether changes to the trial design preserved the credibility of the statistical inference.

Safety may also affect the assessment. FDA documents show that 42% of patients receiving deramiocel experienced hypersensitivity reactions, compared with 15% in the placebo group. Beyond the reactions themselves, the difference between the groups could also have allowed patients or investigators to guess treatment assignments, resulting in “functional unblinding” and potentially affecting functional assessments that contain subjective elements. The available data cannot yet determine how much bias this factor actually caused, but it adds uncertainty to the interpretation of efficacy.

Background

The FDA’s Cellular, Tissue, and Gene Therapies Advisory Committee is scheduled to publicly discuss deramiocel’s biologics license application on July 29. The central question facing the committee is whether the functional and cardiac signals obtained through reanalysis can offset concerns arising from the failure to meet prespecified endpoints and the changes to the analysis. The experts’ recommendations are not legally binding, and the FDA will make the final approval decision after completing its overall review.