EpCAM⁺ Extracellular Vesicle PD-L1 Dynamics as a Predictive Biomarker of Immune Checkpoint Blockade Response

Programmed death-ligand 1 (PD-L1) blockade improves outcomes in patients with various malignancies; however, biomarkers for monitoring treatment responses are lacking. This study presents a surface-enhanced Raman scattering (SERS)-based platform for the ultrasensitive detection of PD-L1 on circulating epithelial cell adhesion molecule-positive (EpCAM⁺) extracellular vesicles (EVs). The platform is validated using interferon-gamma-treated epithelial tumor cell line-derived EVs. The platform is further applied to paired pre- and post-treatment plasma samples from patients with non-small-cell lung cancer (n = 140) and head and neck squamous cell carcinoma (n = 73) receiving anti-PD-(L)1 immunotherapy. Dynamic changes in PD-L1 expression on EV are shown to correlate with clinical outcomes. A post-treatment decrease in PD-L1 expression on EpCAM⁺ EVs (EpCAM⁺ EV PD-L1) is associated with improved 5-year progression-free and overall survival. These results establish EpCAM⁺ EV PD-L1 as a dynamic, non-invasive biomarker for monitoring immunotherapy responses and demonstrate the utility of SERS-based profiling in predicting long-term responses to immune checkpoint blockade.