Research Focus Area
Neoantigen-Driven Targeting: Moving past CARs
How it works
As cancer cells mutate, they display abnormal proteins called neoantigens, which are distinctive enough that the immune system can, in theory, recognize them as foreign. The challenge is finding them. A single tumor can carry hundreds of mutations, but only a fraction produce neoantigens capable of triggering a real immune response.
Some mutations, like KRAS (one of the most common cancer-driving mutations) show up across many patients and many cancer types, making them attractive shared targets. Others are unique to a single patient's tumor, requiring a personalized approach built from scratch.
Once a neoantigen is identified, TCR (T cell receptor) therapy puts it to use: engineering a patient's T cells with a unique TCR that can recognize that target with precision, then returning them to the body to seek out and destroy the cancer.
Why it matters
Most cancer treatments attack what's common across many tumors. That's also their weakness. Cancer adapts, and therapies that hit a single shared target are often the easiest for it to evade. Neoantigens are different: they come from a patient's own tumor mutations, so no two cancers present quite the same set of targets. That specificity is what makes this approach so promising for achieving real, durable cures rather than temporary remission.
It also signals something the field has long been working towards: cancer treatment built around the patient, not the disease category. Instead of a single drug used broadly, this approach uses each patient's own tumor to build a personalized T cell therapy that will fight it.
What we’re doing
- Leading TESLA (The Tumor Neoantigen Selection Alliance), a global collaboration of 36 biotech, pharma, academic, and nonprofit teams that identified the key characteristics determining whether a neoantigen will actually trigger an immune response — work that, when tested, correctly predicted 75% of effective targets and ruled out 98% of ineffective ones.
- Optimizing TCR binding through catch bond technology, pioneered at Stanford, to make engineered T cells more precise and more durable in their attack on tumor cells.
- Applying AI and computational screening to design TCR-like binders capable of recognizing neoantigen targets that conventional methods can't reach.
- Investigating individual recurring cancer tumor antigens — including KRAS, PAP, and MAGE-A3 — to understand what makes a TCR therapeutically effective, not just theoretically promising.
What’s next
Using AI to predict and design functional, patient-specific TCRs and synthetic binders faster and more reliably.