ABSTRACT
Extracellular vesicles (EVs) hold great potential as novel biomarkers for minimally invasive detection of early
stage tumors, since tumors abundantly secrete EVs that accumulate in the circulation and these EVs can
transport factors that regulate tumor initiation, progression, and metastasis. However, most current EV analysis
methods require pre-isolation of EVs prior to analysis and are low-throughput and impractical for clinical use. We
recently developed a rapid, robust, isolation-free and inexpensive assay that directly quantifies tumor-derived
EVs in small volumes (1~5 µL) of serum or plasma. In this assay, EVs that bind probes specific for two different
EV target proteins produce a distance-regulated nanoplasmon-enhanced scattering (nPES) effect that allows
sensitive detection of specific EVs. In a pilot study we used a nPES assay for a pancreatic cancer (PC)-
associated EV marker to distinguish PC cases from non-malignant controls (patients with pancreatitis and
healthy individuals) with high reproducibility, specificity, and sensitivity. This assay also differentiated PC tumor
stages and tumor responses to neoadjuvant, outperforming CA19-9, a biomarker widely used for PC therapy
assessment. Our nPES assay platform has multiple features required for research and clinical translation: 1) It
is rapid, high-throughput, and inexpensive; 2) it does not employ EV isolation, avoiding a major source of EV
assay variation; 3) it robustly and reproducibly quantifies EV biomarkers from small volumes of serum, plasma
or urine, allowing its use in longitudinal analysis of mouse models of human disease; and 4) it can be readily
adapted to diagnose and monitor cancers that express other EV biomarkers. Based on the success of our pilot
study, we propose to develop and validate an automated and highly reproducible nPES EV assay to allow rapid
and accurate PC diagnosis in clinical settings. We hypothesize that a nPES-based digital EV reader will equal
or outperform the analytical performance of our current manual assay. We will build a diagnostic EV assay model
for early PC detection by examining the ability of proteins reported to be enriched on the surface of EVs derived
from PC stem cells or PC-initiating cells (e.g., CD44, CD133 and EpCAM) to diagnose patients with early stage
PC and to differentiate them from patients with pre-malignant pancreatic lesions, hereditary syndromes or family
history of PC, and individuals with normal pancreases. Specifically, we propose to: 1) Development and
fabrication of Chip-nPES platform to achieve single EV resolution; 2) automate and refine our nPES-based
digital EV reader to enhance assay reliability and reproducibility. We will also select and validate candidate
EV capture and detection antibodies for PC diagnosis; 3) establish and evaluate a diagnostic model that
integrates EV biomarkers with known cancer-associations; and 4) perform a pre-clinical validation of this assay
in a third-party laboratory. The successful results of this work would have a significant translational impact in
cancer management, through reliable and accessible screens for early detection of pancreatic cancer.