Should FDA Grant Accelerated Approval to Duchenne Muscular Dystrophy Gene Therapy?

On May 12, independent experts narrowly recommendedopens in a new tab or window that the FDA grant accelerated approval to Sarepta's new gene therapy, SRP-9001 (delandistrogene moxeparvovec), for the treatment of Duchenne muscular dystrophy (DMD). These experts were called to vote after FDA's scientific reviewers raised concernsopens in a new tab or window about the dataopens in a new tab or window Sarepta submitted to support the drug's safety and effectiveness.

While those concerns failed to sway the majority of the advisory committee, we believe the FDA's reviewers were right. The standard for accelerated approval does not appear to have been met here -- to the contrary, the drug's risks may outweigh its benefits. DMD patients need better treatment options, but they also need FDA to uphold strong evidentiary standards so they can be confident in approved drugs.

Background
FDA's accelerated approval pathwayopens in a new tab or window offers a compromise for drugs that meet certain criteria. Drugs can be granted early market access before confirming patient benefit if they demonstrate an effect on surrogate endpoints deemed "reasonably likely" to predict more meaningful clinical outcomes in how patients feel, function, or survive. In return, FDA typically requiresopens in a new tab or window manufacturers to complete additional studies after approval to confirm the predicted clinical benefit.

As a devastating, rare disease affecting approximately 1 out of 3,500 boysopens in a new tab or window globally, DMD is exactly the sort of disease envisaged by early approval pathways. DMD is caused by a mutation in the dystrophin gene, leading to progressive weakening of skeletal muscles and subsequently, the inability to walk, breathing difficulties, and heart complications that lead to an early death. Life expectancy has improved such that patients may live up to their 40sopens in a new tab or window, but DMD has no cure.

Prior Accelerated Approvals for DMD
In 2016, Sarepta's first DMD drug, eteplirsen (Exondys 51), was granted accelerated approvalopens in a new tab or window against the recommendationopens in a new tab or window of another advisory committee and FDA's own scientific staff – probably the most controversial accelerated approval until aducanumab (Aduhelm) for Alzheimer's diseaseopens in a new tab or window. A single, small clinical trial showed that eteplirsen led to a small increase in the protein dystrophin, which became the surrogate endpoint supporting the drug's accelerated approval. However, this same study showed no improvement in clinical progression measured by a functional 6-minute walk test. Increased dystrophin still has not been validated as a measure of clinical benefit. Nevertheless, FDA has awarded three more approvalsopens in a new tab or window for similar DMD drugs using this same surrogate endpoint. Confirmation of clinical benefit for all four treatments remains pending, with none of the required confirmatory trials yet completed. Meanwhile, DMD patients continue to be prescribed these drugs in hopes that more conclusive evidence will emerge showing meaningful, functional improvement.

Shakier Ground
In contrast to these earlier drugs, SRP-9001 is intended to work through the production of a new, genetically-engineered protein called micro-dystrophin. While the drug successfully increases levels of this protein, it is another unvalidated surrogate endpoint: we don't know that micro-dystrophin will actually improve function in boys with DMD or what levels of micro-dystrophin would be needed to do so. Importantly, as notedopens in a new tab or window by FDA reviewers and several advisory committee membersopens in a new tab or window, we don't know that more micro-dystrophin is even "reasonably likely to predict" improved function, the standard for accelerated approval. The only randomized, double-blinded, placebo-controlled studyopens in a new tab or window of SRP-9001 failed to showopens in a new tab or window any correlation between the expression of micro-dystrophin in patients who received the drug and functional outcome. Although analyses of data from other studies show a possible association between micro-dystrophin and function, most participants were enrolled in open-label studies, raising questions around the validity of the results due to expectation bias (i.e., improved performance in patients aware that they received the treatment, and vice versa). While accelerated approvals are expected to entail some uncertainty, there seems to be more here than should be acceptable.