Greenstone Biosciences is using induced pluripotent stem cells derived from DMD patients and generative AI to identify cardiac fibrosis drugs, bypassing animal models where the 90% failure rate in translation reflects fundamental physiological differences between rodent and human hearts. This approach addresses a critical gap: DMD cardiomyopathy lacks approved therapies targeting the underlying fibrosis that drives heart failure, the leading cause of death in this population.
Key Points
- Nine of ten drugs succeeding in mice fail in human trials; cardiac electrophysiology differences acc
- Greenstone screens compounds on patient-derived human cardiomyocytes with DMD genetic background, no
- AI trained on 2,500+ human iPSC lines, including rare disease variants, improves clinical relevance
Longevity Analysis
The translational failure rate in cardiology mirrors a broader challenge in geroscience: animal models cannot faithfully recapitulate the cardiac fibrosis and cellular exhaustion that characterize both DMD and intrinsic aging. By testing compounds directly on human cells carrying the actual genetic substrate of disease, this strategy addresses a fundamental gap in how drug efficacy is predicted. For aging-related cardiac dysfunction—where fibrosis is a hallmark pathology—models trained on patient-derived cells with documented genetic variation offer a more physiologically grounded path to interventions that might otherwise fail in human populations. This represents a shift from extrapolating animal data to humans, toward identifying compounds whose behavior on authentic human tissue predicts clinical outcome.
Original published by Longevity.Technology, by Kyle Umipig.