Predictive Analysis in Oral Cancer Immunotherapy: Profiling Dual PD-L1-Positive Extracellular Vesicle Subtypes with Step-Wedge Microfluidic Chips.

PD-L1-positive extracellular vesicles (PD-L1 EVs) play a pivotal role as predictive biomarkers in cancer immunotherapy. These vesicles, originating from immune cells (I-PD-L1 EVs) and tumor cells (T-PD-L1 EVs), hold distinct clinical predictive values, emphasizing the importance of deeply differentiating the PD-L1 EV subtypes for effective liquid biopsy analyses. However, current methods such as ELISA lack the ability to differentiate their cellular sources. In this study, a novel step-wedge microfluidic chip that combines magnetic microsphere separation with single-layer fluorescence counting is developed. This chip integrates magnetic microspheres modified with anti-PD-L1 antibodies and fluorescent nanoparticles targeting EpCAM (tumor cell marker) or CD45 (immunocyte marker), enabling simultaneous quantification and sensitive analysis of PD-L1 EV subpopulations in oral squamous cell carcinoma (OSCC) patients' saliva without background interference. Analysis results indicate reduced levels of I-PD-L1 EVs in OSCC patients compared to those in healthy individuals, with varying levels of heterogeneous PD-L1 EVs observed among different patient groups. During immunotherapy, responders exhibit decreased levels of total PD-L1 EVs and T-PD-L1 EVs, accompanied by reduced levels of I-PD-L1 EVs. Conversely, nonresponders show increased levels of I-PD-L1 EVs. Utilizing the step-wedge microfluidic chip allows for simultaneous detection of PD-L1 EV subtypes, facilitating the precise prediction of oral cancer immunotherapy outcomes.