Integrative Network Analysis Reveals a MicroRNA-Based Signature for Prognosis Prediction of Epithelial Ovarian Cancer

Abstract

Background. Epithelial ovarian cancer (EOC) is a heterogeneous disease, which has been recently classified into four molecular subtypes, of which the mesenchymal subtype exhibited the worst prognosis. We aimed to identify a microRNA- (miRNA-) based signature by incorporating the molecular modalities involved in the mesenchymal subtype for risk stratification, which would allow the identification of patients who might benefit from more rigorous treatments. Method. We characterized the regulatory mechanisms underlying the mesenchymal subtype using network analyses integrating gene and miRNA expression profiles from The Cancer Genome Atlas (TCGA) cohort to identify a miRNA signature for prognosis prediction. Results. We identified four miRNAs as the master regulators of the mesenchymal subtype and developed a risk score model. The 4-miRNA signature significantly predicted overall survival (OS) and progression-free survival (PFS) in discovery (p=0.004 and p=0.04) and two independent public datasets (GSE73582: OS, HR: 2.26 (1.26-4.05), p=0.005, PFS, HR: 2.03 (1.34-3.09), p<0.001; GSE25204: OS, HR: 3.07 (1.73-5.46), p<0.001, PFS, HR: 2.59 (1.72-3.88), p<0.001). Moreover, in multivariate analyses, the miRNA signature maintained as an independent prognostic predictor and achieved superior efficiency compared to the currently used clinical factors. Conclusions. In conclusion, our network analysis identified a 4-miRNA signature which has prognostic value superior to currently reported clinical covariates. This signature warrants further testing and validation for use in clinical practice.

Comprehensive pathway-related genes signature for prognosis and recurrence of ovarian cancer

“…Because OC is usually diagnosed at a late stage and currently lacks effective treatment options, patients with OC have extremely poor prognosis and only a 30–40% 5-year survival rate ( Enroth et al, 2019 ). Despite the moderate improvements that have been made in the diagnosis and treatment of OC over the past 30 years, OC is still a highly malignant disease that is life-threatening ( Li et al, 2019 ). In addition, OC patients have an average relapse-free survival (RFS) of two to three years after first-line therapy and die due to chemotherapy resistance ( Dinh et al, 2008 ).…”

“…A number of studies have shown that combined biomarkers could improve prognostic accuracy compared to single biomarkers ( Barata & Rini, 2017 ). Thus, extensive studies have attempted to establish molecular signatures based on gene expression to predict survival of patients, including mRNA- ( Zhou et al, 2019 ), microRNA- ( Li et al, 2019 ), and long non-coding RNA-based signatures ( Xu et al, 2019 ). However, only a small number of prognostic signatures have been developed, and none have been directly applied in clinical practice.…”

Systematic Profiling of Alternative Splicing Events in Ovarian Cancer

“…Due to the late diagnosis of ovarian cancer and the lack of effective treatment at present, the prognosis of OC patients is very poor, and the 5-year survival rate is only 30–40% ( 2 ). Although the diagnosis and treatment of OC have been improved to some extent in the past 30 years, OC is still a serious malignant tumor, threatening the lives of women ( 3 ). Therefore, the risk assessment of prognosis is of great value for clinical decision-making and patient consultation.…”

“…Many reports have shown that, compared with a single bio-marker, integrated bio-markers can improve the accuracy of prognosis ( 8 ). Therefore, extensive studies have attempted to establish molecular characteristics based on gene expression data to predict the survival and prognosis of patients, including mRNA ( 9 ), microRNAs ( 3 ) and long non-coding RNA (LncRNA) based signatures ( 10 ). However, although these promising traits played an important role in predicting the survival of patients with OV, they mainly focused on changes in gene expression levels, ignoring the diversity of RNA sub-types regulated by alternative splicing (AS).…”

Development of a novel transcription factors-related prognostic signature for serous ovarian cancer

“…So far, autophagy, DNA repair pathway, and the Notch pathway-related prognostic signatures have been developed for OC [14][15][16]. Gene methylation panel, miRNAs-based signature, miRNA-lncRNA signature were established because of the vital role of epigenetics in OC prognosis [17][18][19]. Recently, a novel and non-invasive radiomic prognostic signature was derived based on computed tomography (CT) images, which signi cantly improved prognostic methods and could be exploited to guide personalized therapy of OC [20].…”