Advanced-stage ovarian cancer is one of the most lethal gynecologic malignancies. To improve prognosis of patients with ovarian cancers, a predictive biomarkers leading to personalized treatments are required. In this large-scale cross-platform study of six microarray datasets consisting of 1054 ovarian cancer patients, we developed a novel risk classification system based on a 126-gene expression signature for predicting overall survival by applying elastic net7 and 10-fold cross validation to a Japanese dataset A (n = 260). We further validated its predictive ability with the five other datasets using multivariate analysis. Also, through gene ontology and pathway analyses of 1109 high-risk ovarian cancer specific transcripts, we identified a significant reduction of expression of immune-response related genes, especially on the antigen presentation pathway. Furthermore, an immunohistochemical analysis demonstrated that the number of CD8 T lymphocytes infiltrating into tumor tissue was significantly decreased in high-risk ovarian cancers. These predictive biomarkers based on the 126-gene expression signature will identify high-risk ovarian cancer patients who need novel immune-activating therapeutic approaches, leading to improved outcomes for such patients.
Overall design
40 patients who were diagnosed as advanced-stage high-grade serous ovarian cancer were recruited in this study.