Integrated extracellular microRNA profiling for ovarian cancer screening

• DOI: 10.1038/s41467-018-06434-4
• Publication Date: 2018-10-11T13:04:53Z
• Authors: Akira Yokoi; Juntaro Matsuzaki; Yusuke Yamamoto; Yutaka Yoneoka; Kenta Takahashi; Hanako Shimizu; Takashi Uehara; Mitsuya Ishikawa; Shun-ichi Ikeda; Takumi Sonoda; Junpei Kawauchi; Satoko Takizawa; Yoshiaki Aoki; Shumpei Niida; Hiromi Sakamoto; Ken Kato; Tomoyasu Kato; Takahiro Ochiya
• Keywords: Ovarian Neoplasms; 0303 health sciences; Models, Statistical; Science; Gene Expression Profiling; Q; Carcinoma; Discriminant Analysis; Middle Aged; Article; 3. Good health; MicroRNAs; 03 medical and health sciences; Humans; Mass Screening; Female; Biomarkers

AbstractA major obstacle to improving prognoses in ovarian cancer is the lack of effective screening methods for early detection. Circulating microRNAs (miRNAs) have been recognized as promising biomarkers that could lead to clinical applications. Here, to develop an optimal detection method, we use microarrays to obtain comprehensive miRNA profiles from 4046 serum samples, including 428 patients with ovarian tumors. A diagnostic model based on expression levels of ten miRNAs is constructed in the discovery set. Validation in an independent cohort reveals that the model is very accurate (sensitivity, 0.99; specificity, 1.00), and the diagnostic accuracy is maintained even in early-stage ovarian cancers. Furthermore, we construct two additional models, each using 9–10 serum miRNAs, aimed at discriminating ovarian cancers from the other types of solid tumors or benign ovarian tumors. Our findings provide robust evidence that the serum miRNA profile represents a promising diagnostic biomarker for ovarian cancer.