The number that killed Elevidys's European approval was 0.65.
That is how many points separated boys who received the gene therapy from those who received a placebo on a standard test of motor function called the North Star Ambulatory Assessment — a 34-point scale used to track how Duchenne muscular dystrophy affects walking and movement. The difference failed to meet the threshold for statistical significance, which is the point at which scientists can be confident the result is not just chance. Europe's top drug regulator called it proof that the therapy did not work.
But placed against what the disease does without treatment, 0.65 points is not nothing. Duchenne muscular dystrophy is a progressive, fatal condition that steals the ability to walk from boys typically by age 12. Published natural history studies show that once boys enter the decline phase of the disease, they lose between 3 and 8 points on the NSAA every year. A 0.65-point advantage is roughly two to three months of function that the therapy appears to have preserved — and that European regulators decided not to count.
Roche announced Thursday that it will run a new Phase 3 trial in approximately 100 boys at the earliest stages of Duchenne, a 72-week study designed to generate the kind of data the European Medicines Agency says it needs to reconsider approval. The trial represents Roche's second attempt to bring Elevidys to the roughly 30 European countries where it is currently unavailable, after the agency's Committee for Medicinal Products for Human Use issued a negative opinion last July.
The NSAA math is the crux of the regulatory dispute. Elevidys, which Roche markets outside the United States under a partnership with Sarepta Therapeutics, works by delivering a truncated version of the dystrophin gene into muscle cells. Boys with Duchenne lack functional dystrophin; the therapy is designed to produce a shortened but functional form of the protein. The theory is that restoring dystrophin slows the disease. The problem is that the connection between dystrophin levels and functional improvement has been difficult to demonstrate in trials.
In the EMBARK study, the primary endpoint — a statistically significant difference in NSAA scores at one year — was not met. But secondary endpoints, including time to rise from the floor and walking speed, did show clinically meaningful differences favoring the treatment group. Longer-term data released in January, showing three-year results from the same trial, indicated that functional benefits appeared to sustain over time. Roche argues this longer follow-up window is what European regulators should be weighing, not the one-year snapshot.
The EMA did not see it that way. The agency's opinion noted that while Elevidys produced measurable levels of micro-dystrophin in muscle tissue, those levels could not be linked to measurable improvement in movement. The 0.65 NSAA point difference, the agency concluded, was not sufficient evidence of benefit, according to its published assessment of the drug.
The new trial will focus exclusively on early-stage patients — boys who retain ambulation and whose disease has not yet inflicted the most severe damage. Roche argues this population is most likely to benefit, and that the original EMBARK enrollment included a broader range of disease stages that may have diluted the signal. Critics of the approach note that this was the same population in which the strongest signal was expected, and where the EMA's critique was most acute.
The trial launch comes as the gene therapy's safety profile remains under cloud. The FDA placed a clinical hold on Elevidys in July of last year after three patient deaths, all attributed to acute liver failure. Two of the deaths were in non-ambulatory patients treated with the therapy; a third was a boy in Brazil who received Elevidys outside of a clinical trial and died in June. Brazilian health authorities determined the death was most likely caused by influenza compounded by immunosuppression, and both the FDA and Roche have said the evidence does not implicate the therapy. The FDA lifted its voluntary pause for ambulatory patients the following month, though non-ambulatory patients remain on hold and the agency's investigation is ongoing.
Elevidys remains approved in the United States and five Middle Eastern countries, but Europe and Japan have not approved it, and the European regulatory outcome is closely watched by investors and developers as a test of how the EMA handles gene therapies granted accelerated approval in the United States based on surrogate endpoints.
Whether a new trial with a narrower enrollment can satisfy regulators who have already rejected the drug once will depend partly on whether the NSAA math becomes a story the field has to reckon with. For now, Roche is betting that longer data and younger patients will be enough. The EMA will be watching to see if the numbers add up.
