Translational modelling challenges receptor‑occupancy‑based dosing of PD‑1 inhibitors like pembrolizumab.

backgroundOptimal dosing of the programmed death-1 (PD-1) inhibitor pembrolizumab remains poorly defined. Dosing has been guided by achieving complete peripheral receptor occupancy, which may not reflect immune activation. We investigated relationships between dose, systemic and intratumoural pharmacokinetics, receptor occupancy, and immune activation in non-small cell lung cancer.

methodsA physiological pharmacokinetic model describing plasma and intratumoural kinetics was extended with receptor occupancy and interleukin-2 (IL-2) induction parameters derived from ex vivo experiments. Simulations compared three approved regimens-2 mg/kg Q3W, 200 mg Q3W, and 400 mg Q6W.

resultsAll regimens produced near-complete PD-1 receptor occupancy in both compartments. Despite this, IL-2 concentrations were predicted to change throughout the dosing interval, tracking pembrolizumab concentration rather than receptor occupancy. The 400 mg Q6W regimen yielded the highest IL-2 levels. Changes in intratumoural IL-2 concentration were less pronounced than in plasma.

conclusionReceptor occupancy did not predict immune activation, suggesting it's an inadequate pharmacodynamic surrogate for pembrolizumab. These findings challenge current PD-1 inhibitor dose selection strategies and highlight the need for robust pharmacodynamic markers and clearer definition of the immune activation required for optimal efficacy.