Todd Hembrough, PhD
Title: Identification of early-stage NSCLC biomarkers from extracellular vesicle microenvironment identifies tumor and immune response biomarkers associated with cancer.
Abstract:
Detection of early-stage cancer is challenging as small tumor volumes limit biomarker release into circulation for analysis. Plasma extracellular vesicles (EVs) represent a promising target for tumor detection and monitoring. Our novel EV isolation methods stabilize the EV corona, which allows detection of tumor and TME specific proteins, as well as biomarkers representing the host immune, inflammatory and metabolic responses to cancer.
To identify novel biomarkers, we enriched EVs using size exclusion chromatography and a proprietary buffer system. DIA mass spectrometry analysis was performed on EVs from 78 NSCLC patients (1A/B (35), 2A/B (8); III (17); IV (18)) and 50 normal plasma samples. Raw data was subjected to machine learning analysis to identify cancer predictive protein multiplexes with high predictive accuracy; hits were validated using ELISAs.
Comparing tumor vs normal samples using the Biognosys TrueDiscovery platform (avg=1877 proteinIDs) identified 520 differentially expressed proteins (Log2FC>0.58; q<0.01). When individual stages were compared to control plasma, a large number of differentially expressed proteins were seen: Stage 1 – 388, Stage 2 – 149, Stage 3 – 358, and Stage 4 – 337 proteins (Log2FC> +/-0.58; q=0.01). Differentially expressed proteins could also distinguish early from late stage NSCLC tumors (Stage 1 vs Stage 3/4). A large number of these differentially expressed proteins represented host immune, inflammatory and metabolic pathway proteins likely associated with the EV corona. We also identified biomarkers associated with antigen presenting cells and checkpoints. These targets suggest that plasma EVs may recapitulate the immune contexture of the TME. Interestingly, when these biomarkers are assessed in complete plasma, no cancer specific differences are seen.
The use of EVs to detect differentially expressed tumor and host immune response biomarkers from even the smallest stage 1 NSCLC tumors supports the hypothesis that EVs have a unique EV microenvironment proteins that can be a highly sensitive and specific platform for detection of early stage and recurrent tumors.