Type I vs. Type II Errors: A case study from ALS

Today, the Food and Drug Administration (FDA) approved the drug Relyvrio (sodium phenylbutyrate/taurursodiol; also known as AMX0035), which is indicated for the treatment of patients with amyotrophic lateral sclerosis (ALS), also known as Lou Gehrig’s disease.  Most people with ALS die within 2-5 years of diagnosis.

In justifying the approval, FDA’s press release stated:

“This approval provides another important treatment option for ALS, a life-threatening disease that currently has no cure,” said Billy Dunn, M.D., director of the Office of Neuroscience in the FDA’s Center for Drug Evaluation and Research. “The FDA remains committed to facilitating the development of additional ALS treatments.”

The data for approval came from the Phase 2 CENTAUR trial (see publications here, here) rather than a larger Phase 3 trial. However, some physicians argued convincingly to the FDA that the potential impact of a Type I error (approving a drug with little benefit) was much lower than would be the impact of a Type II error (failing to approve an ALS drug that worked) given the severity of the disease and lack of alternative treatments. The NY Times reports how doctors treating ALS patients urged an approval:

“In your [the FDA’s] difficult job, there’s always going to be a chance of making a mistake; it comes down to which mistake you would rather make,” Dr. Richard Bedlack, director of the A.L.S. clinic at Duke University, testified this month. “To approve AMX0035 and find out in two years that it doesn’t work — I doubt many are going to be very angry because people with A.L.S. got to try something that was safe and appeared promising in 2022.”
But, he added, “Can you imagine the mistake of saying no and then getting confirmatory evidence in two years that this really did work? And realizing all those patients were much more disabled or even dead when they didn’t need to be? I don’t know how you’ll be able to live with yourself if you make that mistake.”

Time will tell how the Relyvrio’s efficacy as demonstrated in clinical trials will translate to real-world effectiveness, but FDA’s deliberation clearly shows that Type I and Type II drug approval errors have different costs depending on the disease as well as available treatments.