The biotech that didn't license an immune checkpoint. It discovered one - and built a cancer-and-Alzheimer's pipeline on top of it.
In the autumn of 2025, a roughly 70-person company in a Rockville office park did something most thousand-person pharmas spend a decade chasing. Its lead antibody, gotistobart, posted a 12-month overall survival of 63.1% against 30.3% for chemotherapy in patients with previously treated squamous non-small cell lung cancer. Then, a few weeks later, OncoC4 quietly closed a Series B of nearly $50 million. No billboard. No stadium naming rights. Just data and a wire transfer.
OncoC4 is a clinical-stage biopharmaceutical company. That is the polite, regulatory way of saying it makes medicines that do not exist yet. What separates it from the long gray line of hopeful biotechs is the source of its ideas: the company's own founders discovered the biology it is now drugging. Most startups rent their science. OncoC4 grew its own.
"We develop first-in-class and best-in-class medicines for hard-to-treat cancer and Alzheimer's disease."
- OncoC4 mission statementYour immune system is, in theory, a very good cancer drug. It patrols, it recognizes, it destroys. The trouble is that cancer is an excellent student. Tumors learn the immune system's safety switches - the "checkpoints" that stop healthy cells from being attacked - and flip them on, turning the body's own brakes against it.
The first generation of checkpoint drugs - the anti-CTLA-4 and anti-PD-1 antibodies that won a Nobel Prize and minted blockbusters - went after the brakes the immune system's T cells use. They worked, sometimes spectacularly. But they came with a bill: toxicity. Releasing every brake at once means the immune system can turn on healthy tissue too. Oncologists got more responses and more side effects in the same bottle.
Then there was the other half of the immune system entirely - the innate side, the macrophages that physically eat threats. Tumors had a trick there too. A surface protein called CD24, overexpressed in roughly 70% of all human cancers, sends a chemical whisper that means, more or less, "don't eat me." For years, nobody had a way to silence it.
"CD24 is a potent 'don't eat me' signal used by tumors to evade innate immune destruction - and an oncogene overexpressed in about 70% of all human cancers."
- On the CD24-Siglec-10 checkpointTranslation for the rest of us: the tumor is wearing a hi-vis vest that says "staff only" to the immune system.
Yang Liu and Pan Zheng are the kind of scientists who spend careers on a single pathway and are vindicated late. Liu held faculty posts at Yale, NYU, Ohio State, Children's National Medical Center, and the University of Maryland's Institute of Human Virology before co-founding OncoC4, where he is now CEO and Chief Scientific Officer. Zheng, his longtime collaborator, serves as Chief Medical Officer. Between them they did something unusual for a startup: they identified the CD24-Siglec-10 axis as an innate immune checkpoint in the first place.
That changes the math. A typical biotech in-licenses a target from a university and hopes the original biology holds up. OncoC4's founders are the original biology. When the company says ONC-841 is a potential first-in-class SIGLEC10 antibody, "first" is not marketing - it is a statement about who got there.
Their second, subtler bet was on selectivity. Instead of releasing all the immune brakes, they engineered gotistobart (ONC-392) to delete the immunosuppressive regulatory T cells inside the acidic, hostile tumor microenvironment while sparing the same cells in healthy tissue. A pH-sensitive antibody that does its damage where it should and behaves where it shouldn't. The idea: keep the efficacy, lose the bill.
"Most startups rent their science. OncoC4 grew its own - the founders discovered the very checkpoint they're now drugging."
- The OncoC4 thesis, in one lineOncoC4's pipeline is not a scattershot of unrelated bets. It is the same conviction - that you can be precise with the immune system instead of blunt - applied across cancer and neurodegeneration.
Next-gen anti-CTLA-4 antibody that depletes "bad" Tregs in the tumor and spares the good ones. FDA Fast Track + Orphan Drug. Co-developed with BioNTech.
A dual-blockade program hitting PD-1 and VEGF together for hard-to-treat solid tumors.
Potential first-in-class antibody blocking the innate "don't eat me" checkpoint the founders discovered.
A fusion protein pointed the other way - dialing the immune system down for difficult immune-related adverse events and neuroinflammation.
Earlier-stage antibodies against CD24 itself, with reach into both oncology and Alzheimer's.
Five programs from one pathway. Either it's elegant focus or they really like the letters "C" and "D." (It's the former.)
Ten years, two diseases, one pathway. The kind of timeline that looks obvious only in hindsight.
Conviction is cheap in biotech. Survival curves are not. The clearest argument for OncoC4's selective approach is the gotistobart Phase 3 readout in patients whose squamous NSCLC had already progressed after anti-PD-(L)1 therapy and platinum chemotherapy - a group with few good options left.
Two bars. One of them is the whole pitch.
That signal did not appear in a vacuum. BioNTech - the company the world met during the pandemic - paid $200 million upfront in 2023 to co-develop and commercialize gotistobart, with development, regulatory and commercial milestones and double-digit tiered royalties stacked behind it. Large partners do not write nine-figure checks for hunches; they write them for de-risked science.
The October 2025 Series B - backed by GBA Fund, HM Capital, 3E Bioventures Capital and Kaitai Capital - is the other vote of confidence. It funds the parts of the pipeline OncoC4 still owns outright: the PD-1/VEGF program, the two SIGLEC assets, and the CD24 preclinical candidates aimed at both solid tumors and Alzheimer's.
"Large partners don't write nine-figure checks for hunches. They write them for de-risked science."
- On the BioNTech dealHere is the part that makes OncoC4 genuinely odd, in the good way. The CD24-Siglec-10 checkpoint is not only a cancer story. The same innate-immune signaling that lets tumors hide is implicated in how the brain's immune cells - microglia - behave in neurodegeneration. So the company that is trying to wake the immune system up against cancer is, with closely related biology, trying to modulate it in Alzheimer's.
It is a strange portfolio on a spreadsheet and a coherent one in a microscope. Cancer and Alzheimer's look like unrelated enemies. To OncoC4 they are two outcomes of the same immune miscommunication. The mission - first-in-class and best-in-class medicines for hard-to-treat cancer and Alzheimer's - is less a slogan than a description of one idea pointed in two directions.
Note: programs span discovery through Phase 3. Outcomes in clinical trials are uncertain by nature; figures here reflect public disclosures and are approximate.
Return to where we started. A 70-person company in a Maryland office park, a survival curve, a wire transfer. What looked like a quiet autumn is, read correctly, a thesis being proven in slow motion: that the next wave of immunotherapy is not about pressing harder on the immune system, but about aiming it more carefully.
If gotistobart clears its remaining hurdles, it offers a version of CTLA-4 therapy patients can actually tolerate. If ONC-841 holds up, it opens the innate checkpoint - the macrophage half of the immune system - that the first generation of drugs never touched. And if the CD24 platform translates to the brain, the same discovery that started in a cancer lab could matter to a disease that has humbled the entire industry.
None of that is guaranteed. Biology gets a vote, and it is a tough audience. But OncoC4 is no longer asking whether its idea works. The data already said yes, at least once, loudly. The question now is how far one checkpoint can travel.
"OncoC4 didn't set out to beat the immune system's brakes. It set out to find the brakes nobody else had drugged - and it found them at home."
- The closing argumentNo official OncoC4 demo reel is published, so these searches surface the company's leadership talks, partnership briefings and pipeline explainers.
In the press: BioNTech collaboration · Fierce Biotech: Phase 3 · BioSpace: ONC-841 Phase 1 · CD24-Siglec-10 science
Profile compiled from public sources. Clinical and financial figures are approximate and reflect company and partner disclosures as of late 2025.