Ovarian cancer (OC) has the highest mortality rate of all gynaecological malignancies. This is partly due to the limited understanding of the initiating neoplastic transformations culminating in lack of reliable diagnostics for early detection. In vivo animal models have been used to enhance fundamental biological processes of OC development; however, those that can recapitulate initiating pathogenic events are limited. Therefore, there is an urgent need for novel models of early OC to improve the development of early-stage OC detection diagnostics.
We have recently genetically characterised a precursor lesion of OC in the Fanconi anaemia complementation group D2 knock-out (Fancd2-/-) animal model [1-3]. However, its relevance as a model to study early human OC was previously unknown. Therefore, this study firstly compared differential gene expression (c.f. control tissue) of the precursor and late-stage OC phenotype (tubulostromal adenoma) from the Fancd2-/-model to human high-grade serous and serous borderline ovarian tumours by total RNA sequencing of laser capture micro-dissected (LCM) tissue. Subsequently, the Fancd2-/- model was employed to provide proof-of-concept evidence that secreted extracellular vesicle (EV) encapsulated nucleic acid biomarkers of early-staged OC can be detected.
RNA-sequencing analyses of LCM tissue resulted in similar upregulation of key epithelial OC markers, such as Cdh1, Muc16, Keratins, Epcam, Pax8 and Wfdc2, between the mouse precursor lesion and tumour and human OC specimens studied. Then, a comparison of the mouse and human secreted EV sequencing results also revealed shared upregulated EV-derived microRNAs between the mouse OC precursor and adenoma and low- and high-grade human disease specimens. Importantly, these candidate microRNA biomarkers of early OC displayed improved diagnostic value (by receiver-operating characteristic analysis) over the clinical gold-standard CA-125, effectively discriminating between OC and controls. Thus, this research characterised a clinically relevant OC model and identified novel candidate secreted EV biomarkers for earlier detection of human OC.