Plasma MicroRNAs in Ovarian Cancer—Response

We are pleased that our research ( 1 ) has generated interest in the scientific community and we are happy to provide you with our Reply to the comments submitted by Dr. Haifeng, et al . The major concern addressed by Dr. Haifeng, et al . is that the upregulation of plasma miR-21 level in endometriosis-associated ovarian cancer (EAOC) cases is due to loss of estrogen possibly present in this group of patients. Their argument is based on two lines of reasoning: ( 1 ) miR-21 expression is repressed by estrogen and, ( 2 ) our EAOC patients are significantly older than the healthy controls and those in the endometriosis group, and therefore they are likely postmenopausal women with low estrogen levels. In addition to miR-21, the authors also have concerns regarding the increased expression of miR-16 and miR-195, which have also been previously reported as estrogen targets. Although the estrogen/progesterone imbalance is a well-documented pathogenic factor in ovarian cancer and endometriosis, the role of estrogen in regulating miR-16, 21, and 195 expression remains inconclusive. Studies have demonstrated miR-21 being either repressed ( 2 , 3 ) or induced by estrogen ( 4 ). Furthermore, evidence from several studies on the effect of estrogen on miRNA expression ( 5 ) suggests that miR-21 was not a consistent estrogen-repressed miRNA, and miR-16 and 195 only were demonstrated to be repressed by estrogen in one study ( 6 ) among 19 studies reviewed ( 5 ). Second, in our study the averaged fold change (FC) of miR-21 expression between EAOC and healthy controls is dramatic (FC=147.4) compared to the reported repressive effect of estrogen to miR-21 expression (ranging from 50-70% repression), making it difficult to assume that estrogen has a big effect on our experimental results. Third, the average age of our healthy control group and endometriosis group is comparable, (38.75 yrs and 36.24 yrs, respectively). Although estrogen levels were not measured in our study, the endometriosis patients are likely to have higher estrogen levels that were also suggested by Dr. Haifeng, et al . Thus, if miR-21 expression is repressed by estrogen, miR-21 levels in endometriosis patients should be lower than that of in the healthy controls. However, this was not what we found. Instead, endometriosis patients displayed a much higher plasma miR-21 expression compared to healthy controls (FC=10.65). Fourth, differences in the above miRNAs were also present in comparisons from age-matched serous versus EAOC cases (all menopausal) and from age-matched female mice (tumor-bearing versus controls), further suggesting that factors other than estrogen may have impacted these changes. Lastly, as we have shown in our study, plasma and tissue samples have distinct miRNA expression profiles. Currently, it remains unclear the effect of estrogen levels to plasma miRNA expression profile. In summary, based on our study results and the above arguments, we do not believe that estrogen plays a significant role in regulating the plasma miRNA signature we have identified that distinguishes EAOC, endometriosis patients from healthy controls. Sincerely Yours, Xin Huang, Ph.D. Anda M. Vlad, M.D., Ph.D.