The Parkinson's Therapy That Came Back From the Dead
Novo Nordisk killed its entire cell therapy unit last year, leaving the most advanced Parkinson's cell therapy in history without a home. Now a Zuckerberg-backed AI startup has picked it up and is racing toward Phase 2 by year-end.
A Second Life Nobody Expected
In October 2025, Novo Nordisk killed its cell therapy unit. Around 250 employees lost their jobs. Programs in heart failure, Type 1 diabetes, and Parkinson's disease went dark. One of those programs, a brain cell replacement therapy called STEM-PD, looked like it was headed for the pharmaceutical graveyard.
Seven months later, it's alive again. And the unlikely savior? An AI-biotech startup backed by Mark Zuckerberg.
On May 11, 2026, Boston-based Cellular Intelligence announced it had acquired global rights to STEM-PD from Novo Nordisk. The deal includes an equity stake for Novo, potential milestone payments, and royalties. No upfront cash changed hands (at least not publicly). But the real story isn't the deal structure. It's what STEM-PD represents: among the first pluripotent stem cell-derived therapies for Parkinson's disease to reach clinical-stage development.
What STEM-PD Actually Does
Parkinson's disease destroys a very specific type of brain cell. Dopaminergic neurons, the ones that produce dopamine, slowly die off. That's what causes the tremors, the stiffness, the shuffling gait. Current treatments mostly just boost dopamine levels artificially. They manage symptoms. They don't fix the underlying problem.
STEM-PD takes a radically different approach. Think of it like replacing dead batteries instead of just turning up the voltage. Scientists take donor stem cells and coax them into becoming dopamine-producing neuron precursors. Then they transplant those cells directly into the patient's brain, where they (ideally) mature into functional neurons and start producing dopamine on their own.
It's elegant in theory. It's extraordinarily hard in practice. No company has ever gotten an approved cell therapy across the finish line for Parkinson's. STEM-PD is currently the furthest along, sitting in a first-in-human Phase I/II trial with FDA fast-track designation.
Why Novo Walked Away
So why did Novo Nordisk abandon something this promising?
Context matters. In late 2025, new CEO Maziar Mike Doustdar launched a $1.3 billion restructuring. The mandate was clear: double down on obesity, diabetes, and liver disease. Those are the revenue engines. Ozempic and Wegovy print money. Cell therapy for Parkinson's? That's a decade-long bet with massive manufacturing complexity and uncertain returns.
Get tomorrow's biotech intelligence before your competitors.
Join thousands of biotech professionals who start their day with our free, daily briefing.
Novo didn't just kill STEM-PD. They scrapped a heart failure collaboration with Heartseed. They offloaded islet cell technology to Aspect Biosystems in January 2026. The entire cell therapy division was dismantled, piece by piece.
From a shareholder perspective, it made sense. Novo's market cap exceeds $100 billion, driven almost entirely by metabolic disease. Why dilute focus on a $1.3 billion restructuring's savings by chasing high-risk neurology programs?
But from a science perspective, it felt wasteful. STEM-PD had FDA clearance. It had a clinical trial running. It had momentum.
Enter Cellular Intelligence
This is where the story gets interesting.
Cellular Intelligence isn't your typical biotech startup. It's what the founders call "AI-native," meaning artificial intelligence isn't bolted onto their operations; it's woven into everything from manufacturing optimization to development planning. They're backed by Zuckerberg money and headquartered in Boston.
Their pitch: cell therapies are hard to manufacture, hard to scale, and absurdly expensive. AI can solve all three problems.
"This cell therapy Parkinson's program is truly innovative," said Cellular Intelligence CEO Micha Breakstone, PhD. "Optimising and scaling complex cell therapies... is exactly the challenge our AI-native platform was built to solve."
Novo's SVP of Global Research, Jacob Petersen, framed the deal as finding the right custodian: "Finding the right steward for the program was critical, and we are convinced that Cellular Intelligence has the capabilities needed to advance it further."
The plan? Launch a Phase 2 study by the end of 2026. That's aggressive, considering the program just changed hands.
The Bigger Picture: 12 Million Patients, Zero Approved Cell Therapies
Parkinson's disease affects roughly 12 million people worldwide. Treatments haven't fundamentally changed in decades. Levodopa, the gold-standard drug, was introduced in the 1960s. It works, until it doesn't. Over time, patients develop fluctuations, dyskinesia, and diminishing responses.
The field has been chasing disease-modifying therapies for years. GLP-1 receptor agonists (the same class as Ozempic) showed preclinical promise for neuroprotection, but clinical results have been disappointing. A comprehensive 2026 meta-analysis found no significant benefits on standard outcome measures. Novo's own semaglutide trials in Alzheimer's (EVOKE and EVOKE+) failed, further souring the company on neurology bets.
Cell replacement therapy represents a completely different philosophy. Instead of slowing neurodegeneration or managing symptoms, you're restoring the cells that were lost. If it works, it could be transformative.
But "if" is doing a lot of heavy lifting in that sentence.
The Competition Is Real
STEM-PD isn't the only player in this space. Aspen Neuroscience is advancing a patient-derived cell therapy (using the patient's own cells, rather than donor cells) and reportedly moving toward Phase 3. The autologous approach avoids immune rejection issues but creates manufacturing nightmares: you're essentially building a custom product for every single patient.
STEM-PD's allogeneic approach (using donor cells) is more scalable in theory. One manufacturing run could treat many patients. That's exactly the kind of problem AI-driven optimization could help solve, which is presumably why Cellular Intelligence wanted this program.
The Karolinska Institutet also recently secured a 15 million SEK grant from the Novo Nordisk Foundation for preclinical dopamine cell replacement research. The science is gaining momentum across multiple fronts.
What Makes This Deal Unusual
Big pharma companies abandon programs all the time. That's not news. What's unusual here is the ending.
Normally, when a major company shuts down a division, those programs die quietly. Sometimes they get licensed to mid-size biotechs with existing infrastructure. Rarely do they land at a venture-backed startup that's explicitly building new technology to solve the program's biggest challenges.
Novo also structured this deal to maintain skin in the game. The equity stake and milestone payments mean they benefit if STEM-PD succeeds, without bearing the R&D cost or execution risk. It's a free option on a potentially first-in-class therapy.
Novo's communications team confirmed that all clinical and late-preclinical cell therapy assets have now been resolved. The cell therapy chapter is officially closed for the Danish pharma giant.
The Skeptic's Case
Let's be honest about the risks. Phase I/II trials are early. We don't have published efficacy data from STEM-PD yet. The fast-track designation is encouraging (it means the FDA sees potential), but it's not a guarantee of anything.
Cell therapies for the brain face unique challenges. How do transplanted cells survive long-term? Do they integrate properly into neural circuits? Can you manufacture them consistently enough for a commercial product? These questions remain unanswered.
And Cellular Intelligence, for all its AI ambitions, is a startup. They're taking on a profoundly complex program that a $100 billion company decided wasn't worth pursuing. Confidence is great. Execution is harder.
Why This Matters Beyond Parkinson's
The STEM-PD deal is a test case for a new model in biopharma. Big companies increasingly focus on proven, revenue-generating therapeutic areas. The risky, science-heavy programs get spun out, licensed, or killed. If AI-native biotechs can pick up these orphaned programs and actually advance them faster and cheaper, that changes the calculus for the entire industry.
It also raises a philosophical question: should the world's most promising therapies depend on whether they fit a corporate strategy? Novo didn't abandon STEM-PD because the science failed. They abandoned it because their spreadsheet said obesity drugs were a better investment.
That's probably true from a financial returns perspective. But for the millions of Parkinson's patients watching their motor function deteriorate year after year, a spreadsheet isn't much comfort.
The happy ending, if it comes, will be written in Copenhagen and Boston. Not in Novo's quarterly earnings reports.
Clinical & Regulatory
4 min read
The Tiny Biotech That Just Cracked a Problem Big Pharma Couldn't
A small UK biotech just did what no one else could: built an inhaled ENaC blocker that actually works in cystic fibrosis patients. Enterprise Therapeutics' ETD001 is the first drug of its kind to show a real lung function benefit, targeting the 10% of CF patients that Trikafta can't help.
