For two decades, the pharmaceutical industry has been running the same experiment on ALS and watching it fail the same way. More than 160 drugs have moved from animal models into human clinical trials. Almost all of them bombed. The reason, increasingly clear to researchers, is that the models themselves are broken: they are built almost entirely on rare familial forms of the disease, while roughly 85 percent of ALS cases arise spontaneously, with no clear inherited mutation. PRISM press release
That gap is what a new initiative called PRISM aims to close. The ALS Therapy Development Institute, LifeArc, and Axol Bioscience launched the program on March 31, 2026, announcing a collaboration to generate and characterize a large panel of induced pluripotent stem cell lines derived from patients with sporadic ALS. The goal is to give drug hunters better human tissue to work with, and ultimately to improve the historically dismal hit rate in clinical trials. PRISM press release
"The stem cell models we produce can unleash a new generation of treatments that could be effective against this disease by slowing its progression and, ultimately, curing it," said Paul Wright, head of motor neuron disease at LifeArc. PRISM press release
ALS kills people in under two and a half years on average. There are two approved drugs that modestly extend survival, but nothing that halts progression. The industry has poured resources into the problem, but the attrition curve tells the story: a drug looks promising in a mouse, goes into humans, and disappears. A study published in Nature Neuroscience earlier this year put a sharp number on the translational failure. Researchers screened 93 previously tested compounds in motor neurons derived from sporadic ALS patients. Ninety-seven percent did not improve motor neuron health. Only three showed any efficacy.
One combination did stand out: riluzole, memantine, and baricitinib together produced a response across all donor lines in the screen. The finding is early, and a dish result does not a drug make, but it illustrates what is possible when models actually reflect the disease being studied. Nature Neuroscience study
Fernando Vieira, CEO and chief scientific officer at ALS TDI, said the field has known about the sporadic-familial mismatch for years without adequately addressing it. "We know that ending ALS will require delivering the right treatments to the right individuals," he said. PRISM press release
ALS TDI has spent more than a decade building one of the largest collections of ALS-specific iPSCs in the world, drawn from patients enrolled in the ARC Study, which has recruited more than 1,800 participants contributing samples and longitudinal data. What PRISM adds is scale and diversity on the sporadic side, and a formalized partnership to make those lines available to the broader research community through Axol Bioscience's commercial platform. PRISM press release
Sapna Vyas, head of scientific programs at Axol, said the initiative is designed to address a specific technical bottleneck. iPSC models of ALS have been in use since the mid-2010s, but most labs work with a handful of lines, usually carrying mutations in genes like SOD1, C9orf72, or TDP-43, which are rare outside the familial subset. PRISM will attempt to generate lines from patients who represent the full geographic and clinical spectrum of sporadic ALS. ALS.net
The three organizations are presenting PRISM as a precompetitive collaboration, which is notable in a space where companies typically guard cell lines tightly. Whether the model produces usable data within a year or five is an open question. Drug discovery based on iPSC screens has delivered some high-profile failures alongside its successes, and moving from a hit in a dish to a clinical candidate requires its own gauntlet of safety, formulation, and dosing work.
But the logic is straightforward: if the models are wrong, the hits will be wrong too. Fix the models, and at least one layer of attrition becomes addressable. For a disease that gives researchers fewer than 30 months to work with, that is not a small thing. MDPI review
† Add footnote: Source-reported; not independently verified.
†† Consider adding footnote: Source-reported; not independently verified. Alternatively, clarify that this finding comes from a specific iPSC screen rather than broader clinical trial history.
