Identification and verification of a ten-gene signature predicting overall survival for ovarian cancer
Introduction
Ovarian carcinoma (OC) is the third most common gynecologic cancer in women and the fifth leading cause of cancer death in women worldwide [1,2]. According to the statistics of the American Cancer Society, there will be 21,750 newly diagnosed OC cases and about 13,940 women died of OC in America in 2020. Because of the nonspecific symptoms in the early stage, most women with OC are diagnosed with an advanced stage, having a dismal 25% five-year survival rate [3]. Therefore, it's necessary to identify reliable prognostic models for predicting the clinical outcomes [[4], [5], [6]]. Identification of the subset of patients with worse survival and higher mortality is needed for more rigorous follow-up and post-curative adjuvant therapy, which helps decision-making and personalized treatment.
A variety of biomarkers have been used to predict overall survival (OS), containing clinical parameters, endogenous substances and pathohistological characteristics [[7], [8], [9]]. Besides, more and more single genes have been reported to be associated with the OS of OC patients, such as HE4 [10], LncRNA HOTTIP [11] and miR-23a [12]. Unfortunately, due to wide variability of outcomes and genetic heterogeneity, it is difficult to predict the OS accurately with single parameter in OC patients. Nowadays, the availability of large-scale public database with gene expression data and clinical data makes it possible to establish a more accurate prognostic signature than conventional clinical parameters [13].
In this study, publicly available high-throughput sequence data and clinical information were downloaded from the Cancer Genome Atlas (TCGA) dataset [14,15] and International Cancer Genome Consortium (ICGC) dataset [16]. Multiple analysis methods including univariate and multivariate COX regression analysis and Least Absolute Shrinkage and Selection Operator (LASSO) regression analysis were utilized to build the signature [17,18]. Eventually, we established a novel 10-gene signature that can effectively predict survival of OC patients.
Section snippets
Data collection
Raw RNA-sequencing data and corresponding clinical information of OC patients were acquired from the TCGA cohort (https://portal.gdc.cancer.gov/) and the ICGC cohort (http://dcc.icgc.org), with a mixture of different histologic types of OC included in this study. Normalization and log 2 transformation were implemented with the manufacture-provided R packages [19].
Construction of the prognostic signature
Flowchart of the study was shown in Fig. 1. Kaplan-Meier (KM) survival analysis was performed to evaluate the prognostic value of
Identification and estimation of the gene signature
A total of 464 OC patients were enrolled in our study, of which 371 samples from TCGA cohort and 93 samples from ICGC cohort. As shown in Figs. 1, 60,484 genes from TCGA cohort and 48,633 genes from ICGC cohort was applied for KM survival analysis to filter genes related to the overall survival rate. Taking the cut-off value of P < 0.05, we narrowed down 3127 and 3646 genes in the TCGA and ICGC databases, respectively, among which 260 genes overlapped between two datasets. Univariate Cox
Discussion
A heterogeneous clinical outcome has been observed in OC patients, which spans <5 months to beyond 10 years [23]. Nevertheless, due to lack of accurate predictive models to evaluate the prognosis and select patients with worse survival, all the patients are treated with similar treatment plans. Thus, it is necessary and meaningful to discover a prognostic tool which could accurately identify those patients with refractory disease and worse survival. In this study, using two public ovarian
Conclusion
In this study, we identified a novel 10-gene signature using bioinformatics analysis, which could act as an independent predictive factor for OS in OC patients. The nomogram based on 10 genes in the signature also displayed a favorable predictive efficacy for prognosis in OC. These findings could be utilized to improve clinical decision-making regarding treatment and follow-up regime. Future studies will focus on the validation of this signature in clinical trials and the functional roles of
Funding
This article was funded by the General Project Funds from the Health Department of Zhejiang Province (2017KY199 and 2018KY240).
Availability of data and material
All the data used to support the findings of this study are downloaded the TCGA and ICGC databases. Please contact author for data requests.
Ethics approval and consent to participate
Not applicable.
Consent for publication
Not applicable.
CRediT authorship contribution statement
Jinwei Liu: Data curation, Writing - original draft, Writing - review & editing. Fei Xu: Conceptualization, Methodology, Software. Weiye Cheng: Visualization, Investigation. Leilei Gao: Software, Validation, Supervision.
Declaration of competing interest
None declared.
Acknowledgements
We would like to thank the researchers and study participants for their contributions.
References (30)
- et al.
Detection and monitoring of ovarian cancer
Clin. Chim. Acta
(2013) - et al.
Identification three LncRNA prognostic signature of ovarian cancer based on genome-wide copy number variation
Biomed. Pharmacother. Biomed. Pharmacother.
(2020) - et al.
Identification of a prognostic signature associated with DNA repair genes in ovarian cancer
Front. Genet.
(2019) - et al.
Ovarian cancer
Lancet
(2014) - et al.
A gene signature predictive for outcome in advanced ovarian cancer identifies a survival factor: microfibril-associated glycoprotein 2
Canc. Cell
(2009) - et al.
Effective stimulation of growth in MCF-7 human breast cancer cells by inhibition of syntaxin 18 by external guide sequence and ribonuclease P
Canc. Lett.
(2008) - et al.
Attenuation of tumor suppressive function of FBXO16 ubiquitin ligase activates wnt signaling in glioblastoma
Neoplasia
(2019) - et al.
Cancer statistics, 2020
CA Cancer J Clin
(2020) - et al.
Cancer statistics
CA Cancer J Clin
(2017) - et al.
A nine-gene signature related to tumor microenvironment predicts overall survival with ovarian cancer
Aging
(2020)
Prognostic and predictive biomarkers in cancer
Curr. Cancer Drug Targets
Biomarker analysis for oncology
Biomarkers Med.
Biomarkers in cancer staging, prognosis and treatment selection
Nat. Rev. Canc.
The role of novel biomarker HE4 in the diagnosis, prognosis and follow-up of ovarian cancer: a systematic review
Expert Rev. Anticancer Ther.
Long noncoding RNA HOTTIP is a significant indicator of ovarian cancer prognosis and enhances cell proliferation and invasion
Canc. Biomarkers
Cited by (10)
Clinical and Biological Significances of LINC01940 in Gastric Cancer
2023, Research Square
- 1
Jinwei Liu and Fei Xu contributed equally to this work.