Detection of early-stage lung adenocarcinoma using a novel extracellular vesicle and particle-based blood test

Tumor-derived extracellular vesicles and particles (EVPs) represent a promising analyte class for early cancer detection. However, discrimination of tumor-derived EVPs from those shed by healthy tissues represents a major analytical challenge. This work describes the design and characterization of a novel proximity ligation-based immunoassay which targets tumor-derived EVPs in plasma displaying four colocalized surface biomarkers. We demonstrate the functionality of this approach in cancer cell line-derived EVPs and apply the method to lung adenocarcinoma (LUAD) detection. Importantly, we find that requiring four colocalized cancer-associated biomarkers reduces interference from healthy EVPs versus a three-biomarker design, resulting in strong discrimination performance for several biomarker combinations. Using this approach, we developed a prototype assay for LUAD detection which was evaluated in a case-control study composed of 92 LUAD cases and 290 non-cancer controls. The assay and trained classifier exhibited 48.5% (33/68; 95% confidence interval (CI) 37.1-60.2%) stage I sensitivity, 83.3% (15/18; 95% CI 59.8-94.8%) stage II sensitivity, and 100% (6/6; 95% CI 55.2-100%) stage III/IV sensitivity at 90% specificity. Assay signal was significantly correlated with tumor size and uncorrelated with smoking history. These preliminary results demonstrate the technical feasibility of this platform for early-stage lung cancer detection.