Evaxion Just Killed Its Tumor Vaccine. What It's Betting On Instead Is Wild.

EVX-05 is an off-the-shelf vaccine (meaning it doesn't need to be custom-made for each patient) that targets something called ERV-derived antigens. Think of endogenous retroviruses as ancient viral DNA that's been hiding in the human genome for millions of years. In glioblastoma tumors, some of these viral remnants get reactivated and produce proteins on the tumor surface. EVX-05 is designed to train the immune system to recognize and attack those proteins.

It's like giving your immune system a mugshot of a suspect it didn't know existed.

The science is still early. Evaxion presented preclinical work on its ERV-derived neoantigen strategy for glioblastoma at the AACR (a major cancer research conference), but there's no clinical data yet. The plan is for Duke University School of Medicine, under Professor Mustafa Khasraw, to run the initial phase 1 trial. That academic partnership is smart: it offloads some of the clinical costs while lending institutional credibility to a very small company.

Evaxion's Track Record: Actually Not Bad

Before you write off this pivot as desperate flailing, consider what Evaxion has already done with its lead program, EVX-01, a personalized cancer vaccine for melanoma.

The numbers are genuinely impressive. In a phase 2 trial combining EVX-01 with Merck's blockbuster checkpoint inhibitor Keytruda (pembrolizumab), Evaxion reported a 75% overall response rate across 16 patients. Four of those were complete responses, meaning the tumors disappeared entirely on scans. After two years, 92% of responders were still responding with no relapses.

Those are small-sample results, so take them with appropriate caution. But for a company this size, they demonstrate that the underlying AI platform can actually pick the right targets. Evaxion says the learnings from EVX-03 (the program it just killed) and EVX-01 both fed into the design of EVX-05.

The AI platform, called AI-Immunology, is really the common thread here. It's the engine that predicts which antigens a vaccine should target. In the EVX-01 trial, Evaxion reported a statistically significant correlation between the AI's predictions and actual immune responses (p=0.00013). That's a genuinely striking number for a prediction tool in oncology.

The Elephant in the Room: Competition

Evaxion isn't the only company chasing cancer vaccines. Not even close.

Moderna and Merck are the 800-pound gorillas in this space. Their personalized mRNA cancer vaccine, mRNA-4157/V940, showed a 49% sustained reduction in recurrence or death risk in high-risk melanoma over five years, reported in January 2026. That's the kind of data that moves markets and earns FDA attention.

BioNTech is another heavyweight, with deep neoantigen and mRNA capabilities. Then there's a long tail of smaller players: CureVac, Transgene, IO Biotech, PDS Biotech, and others. The cancer vaccine market is projected to balloon from roughly $10 billion in 2025 to over $23 billion by 2033.

But here's the thing that makes Evaxion's pivot interesting: most of the big players are focused on personalized vaccines, which require custom manufacturing for every single patient. That's expensive, slow, and logistically nightmarish. EVX-05 is designed as an off-the-shelf product, targeting antigens shared across glioblastoma patients. If it works, you skip the whole custom-manufacturing headache.

And glioblastoma? It's a different battlefield entirely. The competition for brain cancer vaccines is far thinner than in melanoma or lung cancer. Evaxion is essentially moving from a crowded bar fight to a quieter corner where it might actually be heard.

The Real Question: Can $18 Million Buy a Miracle?

Let's be blunt about the math. Evaxion has $18 million, burns roughly $3.5 million a quarter, earns nothing, and is now betting on a preclinical program with no human data. The company's total equity is $13.2 million. Its total assets are $23.6 million.

This is not a company with margin for error. One failed experiment, one delayed trial, one bad data readout, and the fundraising conversation gets very uncomfortable very fast.

But small-cap biotechs don't survive by playing it safe. They survive by making sharp, focused bets on programs with high unmet medical need and differentiated science. Glioblastoma checks both boxes. The AI platform has real (if early) proof points. And the Duke partnership gives EVX-05 a path into the clinic without Evaxion having to bankroll the entire trial itself.

The pivot makes strategic sense. Whether the biology cooperates is a completely different question.

For Evaxion, the clock is ticking. Eighteen million dollars, one preclinical brain cancer vaccine, and roughly a year to show the world it picked the right fight. In biotech, that's either the beginning of a great comeback story or the opening line of an obituary. We'll know which one soon enough.

Amylyx's Unlikely Comeback Started With a Sugar Crash

Amylyx was left for dead after its ALS drug imploded. Two years later, a $35 million acquisition just delivered Phase 3 data so strong it could create the first approved treatment for a post-surgical complication nobody's been able to fix.