br According to our results high levels
According to our results high levels of circulating miR34a and miR34b were also associated with ovarian cancer among which miR34b showed better diagnostic properties. MiR34 family members are well known to regulate cell cycle, apoptosis and invasiveness in cancer (Li et al., 2014). These miRNAs proved to have direct role in the devel-opment of breast, prostate, Cyclosporin H or brain cancer and miR34a is
Interactions between miRNAs.A: Correlation between the expression levels of miRNAs measured in the malignant and healthy samples. rs - Spearman’s rank cor-
relation coeﬃcients. # p > 0.05. In all the other cases p < 0.05 B: Number of validated target genes shared by miRNAs. The analysis was performed by miRTargetLink.
Table 4 Cohen’s kappa values (κ) in the pairwise comparison of the agreement between the diagnostic tests based on miRNAs and CA125 or HE4 in malignant samples.
considered to be a promising therapeutic target in cancer treatment (Li et al., 2014). Elevated level of circulating miR34a and miR34b was reported in breast, lung and prostate cancer (Lodes et al., 2009; Roth et al., 2010, 2011). Their aberrant expression was also confirmed in ovarian carcinoma samples (Iorio et al., 2007; Zhang et al., 2008; Lee et al., 2009; Corney et al., 2010; Schmid et al., 2016). However, limited information is available about circulating miR34 family members in ovarian cancer. Only miR34c was under investigation previously and it was upregulated in cancer patients (Ayaz et al., 2014) – in contrast to our results.
We also detected significantly elevated expression level of miR203a in the plasma samples of patients with malignant ovarian tumor. MiR203 is a putative tumor suppressor and regulates tumor cell pro-liferation, invasion and angiogenesis. Its aberrant expression was con-firmed in breast, colorectal, prostate cancer or in melanoma (Shao et al., 2017). Higher circulating miR203 level was reported in breast and colorectal cancer (Madhavan et al., 2016; Hur et al. (2015)). Its aberrant expression was also confirmed in ovarian carcinoma samples (Iorio et al., 2007; Lee et al., 2009; Wyman et al. (2009); Wang et al., 2013; Azizmohammadi et al., 2016). However, the application of cir-culating miR203 as a diagnostic marker in ovarian cancer is under-studied. Circulating exosomal miR203 level was found to be elevated in the sera of patients with malignant tumors compared to the sera of patients with benign disease (Taylor and Gercel-Taylor, 2008). How-ever, miR203 level was not compared with healthy controls in this study (Taylor and Gercel-Taylor, 2008).
One major advantage of the utilization of circulating biomarkers is that when a pelvic mass is present in patients the detection of these markers strengthens the suspicion of ovarian carcinoma. In order to identify those miRNAs that might be applicable in the diﬀerential di-agnosis of non-malignant samples we also compared the expression level of miRNAs between malignant and non-malignant samples. In this comparison less miRNAs proved to be significant than when the sam-ples of healthy volunteers were used as controls. These results suggest the followings: i) elevated miR200a, miR141, miR429 and miR203a levels might be reliable biomarkers in the discrimination of non-ma-lignant masses from malignant tumors; ii) non-malignant samples might not be applicable as controls because comparison of malignant samples to non-malignants might produce diﬀerent results than using healthy samples as controls. However, further studies are needed to strengthen this hypothesis with larger samples size.
We also studied the agreement between the diagnostic tests based on miRNAs and the standard biomarkers CA125 and HE4. Although the agreement of some miRNAs was moderate with HE4 they showed weak agreement with CA125. However, higher agreement was observed be-tween the diagnostic tests based on miRNAs. The Cohen’s kappa coef-ficient was the highest between the diagnostic tests based on miR200b and miR200c. We have to note that this analysis was performed only in the case of malignant samples. Previous publications that studied the correlation between miRNAs and CA125 showed controversial results. According to some publications circulating miRNA levels are not