Search

Detection of cancer early, when it is most treatable, remains a significant challenge due to the lack of diagnostic methods sufficiently sensitive to detect nascent tumors. Early-stage tumors are small relative to their tissue of origin, heterogeneous, and infrequently manifest in clinical symptoms. Detection of their presence is made more difficult by a lack of abundant tumor-specific indicators (i.e., protein biomarkers, circulating tumor DNA, etc.) that would enable detection using a non-invasive diagnostic assay. To overcome these obstacles, we have developed a liquid biopsy assay that interrogates circulating extracellular vesicles (EVs) to detect tumor-specific biomarkers colocalized on the surface of individual EVs. We demonstrate the technical feasibility of this approach in human cancer cell line-derived EVs where we show strong correlations between assay signal and cell line gene/protein expression for the ovarian cancer-associated biomarkers BST2, FOLR1, and MUC1. Furthermore, we demonstrate that detecting distinct colocalized biomarkers on the surface of EVs significantly improves discrimination performance relative to single biomarker measurements. Using this approach, we observe promising discrimination of high-grade serous ovarian cancer versus benign ovarian masses and healthy women in a proof-of-concept clinical study.

The low incidence of ovarian cancer (OC) dictates that any screening strategy needs to be both highly sensitive and highly specific. This study explored the utility of detecting multiple colocalized proteins or glycosylation epitopes on single tumor-associated extracellular vesicles (EVs) from blood. The novel OC Test employs immunoaffinity capture of tumor-associated EVs followed by proximity-ligation qPCR to detect combinations of up to three biomarkers to maximize specificity and measures multiple combinations to maximize sensitivity. A high-grade serous carcinoma (HGSC) case-control training set of EDTA plasma samples from 397 women was used to lock down the test design, the data interpretation algorithm, and the cut-off between cancer and non-cancer. Performance was verified and compared to CA125 in an independent blinded case-control set of serum samples from 390 women (132 controls, 66 HGSC, 83 non-HGSC OC, 109 benign). In the verification study, the OC Test showed a specificity of 97.0% (128/132; 95% CI: 92.4%-99.6%), a HGSC sensitivity of 97.0% (64/66; 95% CI: 87.8%-99.2%), and an AUC of 0.97 (95% CI 0.93-0.99) and also detected 73.5% (61/83; 95% CI: 62.7%-82.6%) of the non-HGSC OC cases. This test exhibited fewer false positives in subjects with benign ovarian tumors, non-ovarian cancers, and inflammatory conditions when compared to CA125. The combined sensitivity and specificity of this new test suggests it may have potential in OC screening.