Background
INOVIQ is using its EXO-NET® exosome isolation technology to isolate exosomes from biofluids, discover novel exosomal biomarkers and develop more accurate and reliable diagnostics for cancer. Our first exosome diagnostic, the EXO-OC screening test, is in development for the early detection of ovarian cancer. Currently, there is no approved screening test for ovarian cancer, resulting in a critical unmet need for an accurate and reliable test for earlier detection of ovarian cancer to enable timely intervention and help save women's lives.
Objective
The purpose of this biomarker validation study was to validate our previously discovered ovarian cancer biomarkers and diagnostic performance in an independent patient cohort of over 530 plasma samples from ovarian cancer patients across all-stages, benign disease and healthy women.
Methods
In this case-control study, exosomes were isolated from more than 500 blood samples, using INOVIQ’s EXO-NET® on a fully-automated high-throughput robotic platform. Exosome ovarian cancer protein biomarkers, previously identified in the OC97 study, were measured using targeted mass spectrometry (multiple reaction monitoring, MRM) performed by The University of Queensland’s (UQ) Centre for Extracellular Vesicle Nanomedicine. All targeted biomarkers were identified in ovarian cancer samples and their diagnostic performance was confirmed using ROC curve analysis and multivariate modelling. EXO-NET-isolated exosomes also enabled the identification of additional informative cancer biomarkers.
Results
When these high performing EXO-NET-isolated biomarkers were combined in 10-fold cross validated machine learning algorithms overall test accuracy exceeded 94%. When test specificity was set at 96%, sensitivity was 92% for all stages of disease and 91% for Stage I alone.
Figure 1: Receiver Operating Characteristic Curve for the EXO-OC test. The EXO-OC test combine EXO-NET isolated ovarian cancer biomarkers using a cross-validated machine learning algorithm. The area under the curve (AUC) = 0.98, indicative of very high accuracy.
Conclusion
The results obtained in this study are outstanding and evidence the robustness and reproducibility of EXO-NET-isolated ovarian cancer biomarkers that deliver high diagnostic performance in multivariate algorithms. This biomarker panel showed exceptional performance in detecting early-stage ovarian cancer (Stages I), where accurate diagnosis is most critical for improving patient outcomes. Early detection enables timely intervention, which is crucial for increasing survival rates and reducing disease progression.
Further development and optimisation of the EXO-OC blood test will be required including analytical validation on a commercial MRM instrument platform that can be routinely used in pathology laboratories worldwide and robust clinical studies to commercialise the test as an LDT or IVD in an accredited clinical laboratory.
