Imagine testing a really good drug for HER2+ breast cancer in someone with liver cancer. Would it be any surprise when that drug fails?
Pictured: Multiple heads representing patient diversity/iStock, wildpixel
Disappointing results are nothing new in clinical trials for amyotrophic lateral sclerosis (ALS), and yet, the news that Amylyx’s Relyvrio failed its Phase III trial still landed like a piano on pavement.
Amylyx had hoped the latest trial would confirm the drug’s promising Phase II results, which showed a 33% slower rate of progression and significant survival benefit for the devastating disease. In 2022, the FDA granted approval to Relyvrio, citing these results, its strong safety profile and the company’s pledge to withdraw the medication from the market if confirmatory studies find it doesn’t work. The drug’s fate is now in limbo.
Sadly, this is the third ALS drug to fail a Phase III trial in the past 12 months—and the second FDA-approved ALS drug. In September 2023, Mitsubishi Tanabe stopped a Phase IIIb trial of daily administered oral edaravone (Radicava), which had been approved the previous year, saying it didn’t appear more effective than an already FDA-approved on/off regimen. In January, Ferrer similarly announced that its different formulation of oral edaravone also failed in Phase III. As an ALS patient, this has been excruciating to watch.
Evidence for edaravone’s efficacy was always shaky. The drug failed two out of three of its original Phase III trials, and while additional analyses found that it might slow progression in a subset of patients with early disease, an independent analysis failed to find real-world evidence of this effect. These latest failures, therefore, only add more doubt.
This all means that almost 30 years after the FDA approved riluzole—which extends life for a few months at best—it remains the only drug we believe unequivocally works. As the founder of an advisory firm dedicated to accelerating ALS research and an ALS Association trustee who helps oversee a portfolio of 150 translational research projects, I know we can do better. While those of us in the field like to be optimistic, this moment calls for sober reflection. Our one-size-fits-all approach to ALS drug development isn’t working.
Understanding the Diverse Nature of ALS
ALS, like other neurodegenerative conditions, is highly heterogeneous. Patterns of onset and progression vary widely, and while some cases have hereditary origins, 80%–90% occur without an obvious genetic cause. This has led some researchers to suggest that it may not be a single disease. Rather, like cancer, ALS may actually be a mosaic of overlapping conditions that share common characteristics. This means we’ll probably never find a singular cure.
The cancer lessons extend further. A major reason why five-year survival rates are now above 70%—up from 50% in 1975—is because we started tailoring treatments to cancer subtypes. In ALS, however, most researchers and drugmakers treat heterogeneity as an afterthought. We rely on cell and animal models that aren’t genetically representative of most patients with ALS, and we include people in clinical trials without knowing if they’re the best biological fit for the drug. Imagine testing a really good drug for HER2+ breast cancer in someone with liver cancer. Would it be any surprise when that drug fails?
One reason so many scientists cling to the idea that ALS is a single condition is because 97% of cases involve malfunction of a protein called TDP43. TDP43 normally resides in the cell’s nucleus, where it plays an important quality control function, ensuring that DNA gets accurately transcribed into RNAs. RNAs act as blueprints for protein manufacturing. When TDP43 stops working, proteins that should get made don’t, or they get made incorrectly. And now, thanks to recent findings from the National Institutes of Health, we know an additional outcome of faulty TDP43: rogue proteins that shouldn’t get made do.
This “protein soup” causes a host of downstream problems, eventually culminating in cell death. But the makeup of that soup is highly variable, as it’s influenced by a person’s genetics and epigenetics. This means that while there may be overlap across patients in terms of TDP43’s role and the rogue proteins created as a result, there are likely important differences too.
Indeed, in December, a team from King’s College London found that among those with non-hereditary forms of ALS, gene expression patterns fall into three distinct categories, each corresponding to known disease mechanisms—suggesting there are different “flavors” of soup, if you will. This may also explain why, when a team from the University of Southern California last year tested 50 drugs shown to boost survival in motor neurons grown from patients with a genetic form of ALS in cells grown from patients with non-genetic forms of the disease, they saw highly variable responses—even though all the neurons exhibited TDP43 malfunction. Even riluzole was suggested to work only in a subset of patients.
Toward Better Drug Development
So, did Relyvrio and edaravone fail their most recent trials because they don’t work for anyone, or because they only work for some? We don’t know.
What we do know is that ALS is not a molecularly monolithic disease, so we therefore need to stop treating it as one. This means advancing methods of modeling non-hereditary ALS, using RNA sequencing and high-throughput proteomics to understand disease subtypes and developing better diagnostics to enable trials designed around those subtypes.
In September 2023, Takeda announced a $580 million investment in AcuraStem, which is developing therapies for a target identified with these tools. This move highlights the potential value of prioritizing heterogeneity. Last spring, the FDA approved Biogen and Ionis’ Qalsody for the 1% of patients with the SOD1 genetic form of ALS, further validating precision approaches. But we need more industry leaders to follow suit.
Until we stop treating ALS as a single disease, I fear we’re in for more Relyvrio-type disappointments.
Bernie Zipprich is the founder of Zipprich Ventures LLC, an advisory firm focused on accelerating treatments for neurodegenerative diseases. He is also a trustee of the ALS Association and a research ambassador for the Northeast ALS Consortium. The views expressed here are his own.
