We scan new podcasts and send you the top 5 insights daily.
Bostwick tells preclinical teams that mice are poor proxies and human genetics is more trustworthy. His advice is to chart the fastest, most agile route to first-in-human testing, treat that trial as the starting line, and prioritize biomarker and regulatory strategy over leaning on animal data.
The high failure rate of drugs in human trials after passing animal tests stems from a fundamental biological reality: a "mouse is not a small human." This "structural mismatch" is especially severe for modern, human-specific therapies like CAR-T and RNA, rendering animal models poor proxies.
Bostwick says AI can help molecule design and preclinical packages, but claiming discovery is the main bottleneck misses the cost center. He notes 90% of costs and failures sit in clinical development, so he wants companies using AI to compress biomedical timelines and trials, not just generate many compounds.
Current drug development heavily relies on animal testing. However, significant biological differences mean we may be filtering out effective human medicines that fail in animal models, creating a hidden opportunity cost for medical breakthroughs.
Startups often rush to publicize their first patient enrollment. A better approach is to treat the entire first-in-human study as a rigorous experiment focused on learning and optimization, delaying major marketing until after the trial is successfully completed.
The push away from animal models is a technical necessity, not just an ethical one. Advanced therapeutics like T-cell engagers and multispecific antibodies depend on human-specific biological pathways. These mechanisms are not accurately reproduced in animal models, rendering them ineffective for testing these new drug classes.
A common failure mode is sprinting forward with a molecule after seeing any signal in an animal model. Isaac Stoner advises overinvesting in robust in-vitro assay development first. This ensures a deep understanding of a molecule's functional activity on human biology, preventing costly translational failures later.
For CNS diseases, where animal models are notoriously unreliable predictors of efficacy, the most pragmatic R&D model is to quickly move promising new chemical entities into human trials. The focus shifts from extensive preclinical validation to early biological experimentation in humans for proof-of-concept.
Beyond establishing safety and efficacy, first-in-human trials provide crucial, practical insights unavailable from preclinical models. They reveal how a product fits into real surgical practice: if it's easy for surgeons to use, integrates into existing procedures, and requires additional training. This real-world validation is invaluable.
The "Valley of Death" where most biotech companies fail is not due to bad science but to the crippling cost of animal trials, which are poor predictors of human outcomes. Parrish argues for shifting focus directly to human data, as mice are not a reliable biological proxy for humans.
Unlike big pharma, capital-constrained biotechs can't afford long, expensive trials. MindImmune’s CEO advocates for designing clever Phase 1b studies that use biomarker endpoints to get an early efficacy signal in months, not years, thereby de-risking the program for investors much faster.