Federal health regulators remain unconvinced about the benefits of a closely watched experimental drug for the debilitating illness known as Lou Gehrig’s disease, even as they prepare to give its drugmaker a rare second opportunity to make a public case for the treatment.
Amylyx Pharmaceuticals’ experimental drug has become a rallying cause for patients with the deadly neurodegenerative disease, their families and members of Congress who are pushing the FDA to approve the drug.
But regulators said Friday that the drugmaker’s new analyses are not “sufficiently independent or persuasive” to establish its effectiveness. The agency posted its review ahead of a Wednesday meeting of its advisers, who will vote on whether to recommend approval.
In March, the same panel of neurological experts voted 6-4 that the company’s data failed to show a convincing benefit for ALS, or amyotrophic lateral sclerosis. It’s extremely rare for the FDA to call a second review meeting after its advisers have already voted.
The FDA will ask the panel to review several new statistical analyses, which the company says strengthen the case that its drug prolongs life and delays hospitalization and other severe complications. The FDA says the experts can take into account “the unmet need in ALS,” the disease’s seriousness and other factors specific to the terminal diseases.
ALS destroys nerve cells needed to walk, talk, swallow and—eventually—breathe. There is no cure and most people die within three to five years.
The FDA’s review reflects some of the biggest questions facing the agency, including: How strict should it be in enforcing approval standards for drugs against rare, fatal diseases? And how much weight, if any, should be given to outside appeals from patients, advocates and their political allies?
Typically, FDA approval requires two large studies or one study with a “very persuasive” effect on survival.
Amylyx’s data comes from one small, mid-stage trial that showed some benefit in slowing the disease, but which was marred by missing data, implementation error