An FDA-approved ALS drug doesn’t work. Now what?

In a surprise to biotech industry watchers and people with amyotrophic lateral sclerosis (ALS), Amylyx Pharmaceuticals’ drug Relyvrio failed in a Phase 3 study, the company announced Friday.

The 664-person study was already underway in 2022, when Health Canada and the US Food and Drug Administration approved Relyvrio based on a successful Phase 2 trial.

Although a Phase 3 trial wasn’t technically a condition of the approvals, patient advocacy groups and regulators both put pressure on Amylyx to run a confirmatory study. Billy Dunn, then the director of the FDA’s neuroscience office, went so far as to ask the company’s cofounders during a public meeting in 2022 if they’d be willing to pull their drug from the market should it fail in the Phase 3 study.

The cofounders—Justin Klee and Joshua Cohen, who are also Amylyx’s co-CEOs—agreed. “If the drug’s not helping patients, then why would we want to be giving it to them?” Klee said at the time.

But that’s now the situation Amylyx finds itself in.

Relyvrio, also known as called AMX0035, is a combination drug made up of sodium phenylbutyrate and taurursodiol. Taken together as a powder mixed into a drink, the components are meant to reduce stress in the endoplasmic reticula and mitochondria, which can lead to misfolding and aggregation of proteins in people with ALS (N. Engl. J. Med. 2020, DOI: 10.1056/NEJMoa1916945). Sodium phenylbutyrate blocks enzymes called histone deacetylases and upregulates heat shock proteins, which is thought to soften any toxic effects from distressed endoplasmic reticula, while taurursodiol stops the apoptosis regulator BAX protein from moving into the mitochondrial membrane, which should reduce the membrane’s permeability and restore the organelle’s energy-producing function.

In the 137-person Phase 2 study that paved the way for the US and Canadian approvals, Relyvrio showed a functional benefit for patients with ALS, based on the ALS Functional Rating Scale. A post hoc analysis revealed that it also helped patients live longer.

The Phase 3 trial went differently. The later study enrolled nearly five times as many patients as the Phase 2 study did, and it monitored them over 48 weeks—almost twice as long as the 6 months in the previous trial. The larger sample size is meant to better reflect real-world patients.

In the larger trial, Relyvrio did not meet its primary goal of improving function, and it did not meet its secondary goals either—changes to metrics on physical mobility, independence, eating and drinking, communication, and emotional reactions were not statistically significant. Amylyx will present the full data at a scientific meeting and publish their findings in a peer-reviewed journal later this year.

“This trial clearly was not successful,” Klee says. “Hearing from many in the [ALS patient] community, I think it’s 100 times harder for them, which is always so important to keep in mind. There’s a lot we don’t know about ALS. There’s still a lot more to learn.”

Klee and Cohen have already decided to stop advertising the drug. They’ll make a decision on whether to stop selling it altogether sometime in the next 8 weeks.

If Relyvrio is pulled from the market, patients with ALS will have just a few treatment options. Managing the disease comes down largely to supportive therapies, including speech therapy, dietary support, and physical and occupational therapies. There are two pharmaceutical ALS interventions: riluzole, an oral medication that can increase life expectancy by about 25%, and edaravone, an intravenous drug that slows the rate of functional decline. The average survival time for people with ALS is just 2–5 years after diagnosis.

Thousands of patients have taken Relyvrio since the drug launched in 2022. The FDA had approved the drug despite skepticism from some agency reviewers, arguing that, given ALS’s deadliness, “this level of uncertainty is acceptable in this instance.” Listed at $158,000 a year before insurance, the drug has brought in $403 million in net revenue for Amylyx, helping the company turn a $49 million profit in 2023.

“We supported the early approval of Relyvrio because there were no questions about its safety and the Phase 2 trial data was positive,” the ALS Association said in a statement Friday. “Had the FDA not approved early and then the [Phase 3] PHOENIX trial results had been positive, thousands of people living with ALS would have not had access to a life-extending treatment that was safe for over 2 years. Our community expressed to the FDA that they were willing to take the risk if it turned out to be ineffective. Now that the confirmatory trial has failed, we hope Amylyx will consult with all stakeholders in the ALS community on next steps.”

It’s possible that Relyvrio might be helpful in subgroups of patients, but while Amylyx researchers plan to examine that possibility, the company isn’t in the business of cherry-picking data for commercial reasons, Cohen says.

“As a company, we’ve always tried to be extremely rigorous in the science,” Cohen says. “Our intention is to evaluate the rigorous data, the prespecified data, not to create some sort of post hoc narrative. The point here is to evaluate the real science, as it stands, and make the decision on that.”

Cohen and Klee are not abandoning AMX0035. Cohen says the fundamental belief that targeting the endoplasmic reticula and mitochondria reduces neuronal stress is still a valid hypothesis, and it underpins a few Amylyx programs beyond Relyvrio.

The company is testing AMX0035 as a treatment for Wolfram syndrome, another neurodegenerative disorder, which involves diabetes and vision loss. Data from a midstage study are expected in the second quarter of this year. It’s also testing AMX0035 in midstage studies for Alzheimer’s disease and progressive supranuclear palsy, operating on the hypothesis that the drug could treat both by targeting the tau protein.

“We know that these drugs are biologically active,” Cohen says. “It may be the case that in ALS, or in certain populations of [people with] ALS, we don’t have quite the activity that we wanted, but that does not mean that it will not work in other indications.”