CAR-T Antigen Escape: CD19-Negative Relapse

CAR-T cell therapy engineers a patient's own T-cells to recognise a single tumor surface marker — usually CD19 in B-cell cancers — and kill any cell that carries it. But that selective pressure is a Darwinian filter: cells with less of the target antigen survive slightly more often, divide, and pass a little less antigen to their offspring. This simulator runs a real stochastic birth–death model of a 3D tumor cell population under CAR-T attack, tracking each cell's individual antigen level as CAR-T cells (green) patrol and preferentially kill the antigen-high cells (red) while antigen-low escape variants (grey) quietly expand underneath the response — the mechanism behind the antigen-loss relapses seen in real CAR-T patients.