Dead brains kept alive for drug testing just helped fast-track a Parkinson's trial
A US startup is testing drugs on human brains kept biologically active after death — and the FDA has already used the resulting data to greenlight a clinical trial. Bexorg, a spinout from Yale, has processed more than 700 donated brains over five years using its BrainEx perfusion system. The approach is now producing results that animal models simply can't match.
How it works
BrainEx acts as an artificial body for a harvested brain: it pumps in oxygen, nutrients, and a blood substitute, keeping cellular metabolism running for up to 24 hours. There is no consciousness involved. The brains show no coordinated electrical activity — the kind required for any form of awareness — and the system doses tissue with propofol, a general anesthetic, as an added precaution. Bexorg employs six bioethicists to oversee the work. After the 24-hour window closes, each brain is sliced into hundreds of small fragments for molecular analysis.
The key advantage is biological realism. A lab mouse has a genetically uniform, sterile life history. A donated human brain belongs to someone who lived seven or eight decades — accumulating infections, environmental exposures, medication histories, and the cellular wear that actually underlies diseases like Alzheimer's and Parkinson's. That context matters enormously when you're trying to predict how a drug will behave in a real patient.
From bench to trial
Pharmaceutical company Biohaven tested roughly 130 Bexorg brain samples on compounds targeting Parkinson's disease and other conditions. One candidate drug worked at a dose 20 times lower than mouse models had predicted — a gap that, left undiscovered until human trials, would cost years of development time and potentially millions of dollars, reports Science (AAAS).
That data supported Biohaven's FDA application for BHV-8100, a Phase 1 trial that began human dosing in June 2026. It marks one of the first times whole-brain perfusion data has contributed directly to a US regulatory approval. The FDA Modernization Act 2.0 explicitly allows these kinds of New Approach Methods — alternatives to animal testing — in drug applications, and Bexorg is now positioned as the first whole-brain proof of concept under that framework, per AdvanceCT.
What comes next
Bexorg's current process is largely manual. The company plans to automate using robotics, with a stated target of processing up to 1,600 brains per year. Reaching that scale would make the platform viable as a routine screening step for central nervous system drugs — potentially cutting one to two years off development timelines before any human volunteer is enrolled.
The main caveat: Bexorg has not yet published peer-reviewed papers specifically on its human brain work. The strongest evidence so far comes from partner announcements and the FDA's own decision to accept BHV-8100's application. Independent validation will be essential before the approach becomes an industry standard.
