You can see it everywhere: on TV and in newspapers and ads. The decades of struggle against cancer are paying off. Even the FDA seems to have set aside its historic preoccupation with large placebo-controlled trials.

The FDA recently awarded breakthrough status to a novel treatment of glioblastoma based merely on promising results among 24 patients enrolled in a Phase I trial. Yet for some children and their families, that FDA initiative will leave a bitter aftertaste.

Earlier this year, an FDA advisory committee solidly rejected a request for accelerated approval of eteplirsen, a promising treatment for a small subset (13%) of Duchenne muscular dystrophy patients.

See also: Zika Crisis Stresses Need for Pharma Industry

The application was based on a long-term Phase I study of 12 patients supplemented with data from European eteplirsen trials and biomarker data. Preliminary results were exciting: After four years of treatment, 10 of the 12 were still walking. In a group of matched historical controls, none were ambulatory after four years of observation. Levels of dystrophin, a biomarker of disease severity, were significantly improved in treated patients, but the FDA quibbled with the testing method.

The appeals of 100 patients and families fell on deaf ears. According to the FDA briefing document, safety wasn’t the concern. The problem was that more efficacy data was needed. So the FDA advisory panel voted against approval. In the words of the FDA briefing document:

“Although FDA is prepared to be flexible with respect to a devastating illness with no treatment options, flexibility does not mean approving drugs for which substantial evidence of effectiveness has not been established.”

Duh! Ten of 12 kids were still walking. And DMD is an orphan disease, diagnosed in perhaps one in 3,600 boys every year. Large clinical trials just aren’t feasible.

Another problem was using a six-minute walking test as the primary endpoint. Because DMD often progresses rapidly, by the time patients were diagnosed and screened, many had already lost the ability to walk, so they were deemed ineligible for trial participation.

See also: A Faster FDA Approval Rate Key to Price Debate

Again, it’s welcome news that the FDA is showing more flexibility with approving cancer treatments. But why can’t that enlightened mind-set be extended to life-threatening orphan diseases as well? In the case of eteplirsen, an obvious solution would have been for the FDA to have granted some sort of provisional approval, reserving the right to change or withdraw the label later.

David Stewart, head of medical oncology at the University of Ottawa, recently proposed some great ideas for shortening development times for potentially lifesaving drugs. His suggestions include creating separate approval processes for lethal diseases and nonlethal ones, increased reliance on Phase I and Phase II data, and allowing investigators more leeway in patient inclusion.

We’re entering a new era of medicine. Let’s make sure that all patients have the opportunity to benefit. Write your congressional representatives and senators.

Sander Flaum is principal of Flaum Navigators.