Dysregulated expression of circular RNAs serve as diagnostic and prognostic markers in ovarian and cervical cancer

Abstract Introduction: Recent studies have reported a connection between non-coding RNAs such as circular RNAs (circRNAs) and the prognosis of various cancers. However, the mechanism of circRNA in ovarian cancer and cervical cancer has not been consistent. We evaluated the diagnostic and prognostic roles of circRNAs in ovarian and cervical cancer by meta-analysis. Methods: Pooled hazard ratios with 95% confidence intervals were to estimate overall survival. Diagnostic efficacy was estimated by sensitivity, specificity and area under curve. Results: By searching PubMed, Embase, the Web of Science databases, and other sources, we obtained a total of 22 studies with 2059 patients from Asia population. High expression levels of oncogenic circRNAs were significantly associated with poor prognoses both in ovarian and cervical cancer. However, elevated expression levels of tumor-suppressor circRNAs were linked with favorable survival time in ovarian cancer. As for diagnostic role, the area under the curve value in ovarian cancer and cervical cancer is 0.89 and 0.93, respectively. Conclusions: CircRNAs have the prospect of becoming a promising biomarker for diagnosis and prognosis of ovarian and cervical cancer. Accordingly, circRNAs might be novel indicators and targets of therapy for ovarian and cervical cancer.


Introduction
Ovarian and cervical cancer are the 2 most common tumors in females. [1] Among the tumor-related causes of death in young women, cervical cancer ranks second. [1] Approximately 527,600 new cases and 265,700 mortalities annually and showing an increasing trend with cervical cancer. [2] Besides, ovarian cancer was the 7th most common cancer in 2012 and it was predicted that the number of deaths due to ovarian cancer will increase to 254,000 in 2035 by the Globocan study. [3,4] CircRNAs are a special kind of endogenous noncoding RNAs that have connected 3 0 and 5 0 ends that form a closed covalent ring structure through the cyclization of exons or introns. They are also competitive RNAs that, along with long-chain noncoding RNAs, coregulate microRNAs. [5][6][7][8] CircRNA participates in the growth and development of cancer, diabetes, nervous system disorders, cardiovascular diseases, and other diseases through various biological roles, such as sponge action, protein translation, and binding protein action. [9][10][11][12] In recent years, more and more researchers have found that circRNA plays an important role in the development of ovarian and cervical cancer. However, no consistent results have been obtained regarding the mechanism of circRNA in the 2 cancers. [13][14][15][16][17][18][19][20][21][22][23][24][25][26] Through the study of Zhao et al, [27] oncogenic human papillomaviruses (HPVs) produce circRNA, which inhibits cancer cell growth both in vitro and in tumor xenografts. And, Guan et al [28] reported circPUM1 promoted tumorigenesis and progression of ovarian cancer through sponge miR-615-5p and miR-6753-5p. Enhanced understanding of the role of Editor: Jorddy Neves Cruz.
Declarations: Availability of data and materials: The datasets supporting the conclusions of this article are included in the article (and its additional files).
Funding: This work was supported by grants from the Nantong University Student Innovation Program (2020146) and the National Natural Science Foundation of China (Grant No: 81873978). The funding was used for searching for data. circRNAs in ovarian and cervical cancer survival will provide more accurate prognostic information and could improve clinical decision-making in trial design and cancer treatment. Accordingly, we conducted this meta-analysis based on plenty of original documents to identify the role of circRNAs in ovarian and cervical cancer.

Ethics statement
All analyses were based on previously published studies, this article does not contain any studies with human participants or animals performed by any of the authors; thus, ethical approval and patient consent are not applicable.

Search strategy
Based on the guidelines of the Meta-analysis of Observational Studies in Epidemiology (MOOSE) group and Preferred Reporting Items for Systematic Reviews and Meta-analysis (PRISMA) statement, [29] we searched the Web of Science, EMBASE, PubMed, Cochrane library and CNKI databases up to August 1, 2020. The searching items were: "circRNA," "circular RNA," "hsa circ," "gynecological cancer," "ovarian cancer," "oophoroma," "carcinoma of the ovary," "metrocarcinoma," and "cervical cancer." To avoid missing documents, we manually screened the reference lists of the retrieved articles.

Eligibility criteria
Eligible articles conformed to the following criteria: the subjects were ovarian cancer or cervical patients confirmed by histopathological diagnosis and the clinical data were complete; the article evaluated the relationship between circRNA expression, clinicopathological features, diagnosis and prognosis; and (3) it was a case-control study. The exclusion criteria were: the subjects of the study were not human; the publication was not a primary research publication (eg, a review, correspondence, repeated publication, conference summary), there were no data available in the article.

Quality assessment
The quality of primary diagnostic accuracy studies was assessed by the QUADAS-2 tool. The QUADAS tool consists of 4 key domains, including patient selection, index test, reference standard, and flow of patients. The answer to risk for bias could be rated as "no" (0 points), "yes" (1 point), or "unclear" (0 points). [30] The Newcastle-Ottawa Scale was used to evaluate the quality of casecontrol studies from three aspects: selection, comparability, and results. [31] Publications that were rated <6 points were considered of low quality; high quality was >6 points.

Data extraction
Two researchers (LFY, WXR) separately evaluated the suitability of all retrieved studies and extracted the relevant data. The 2 researchers contacted a third researcher (WF) when there was a disagreement. The following data were extracted: title, first author, ethnicity, year, sample type, patient size, circRNA signature, follow-up (months), TNM stage, expression status, detection methods, sensitivity (SEN), specificity (SPE), cutoff value setting, pooled hazard ratios (HRs), overall survival (OS), disease-free survival, and their corresponding 95% confidence intervals (CIs). When HRs and 95% CIs could not be extracted directly, we applied the methods described by Parmar et al [32] and Tierney et al [33] to estimate the values from the Kaplan-Meier curves in the articles.

Statistical analysis
HRs and 95% CIs were used to estimate OS. Sensitivity, SPE, and area under the curve (AUC) were involved in the diagnostic metaanalysis. Heterogeneity was assessed by the x 2 test and expressed by the I 2 index and was judged to be significant if the I 2 value was >50%. We used SEN analyses to investigate potential sources of heterogeneity. Publication bias was evaluated quantitatively using Deeks funnel plot, Begg tests, and Egger tests. Statistical analyses were performed using Revman 5.3 and Stata 15.1 software (Stata Corporation, College Station, TX).

Cervical cancer studies.
Five studies were included in the analysis of the effect of circRNA on ovarian cancer overall survival. The results showed that high expression of Cervical cancer tumor-promoter circRNAs was associated with poor survival time (HR = 2.53, 95% CI: 1.72-3.73, P < .001) (Fig. 2  C). No significant heterogeneity was found across the studies (I 2 = 0%, P = .931). The relation between tumor suppressor circRNA and overall survival failed to obtain due to the lack of suitable original studies. Table 3 The expression and mechanisms of circRNAs in cervical cancer and ovarian cancer.
3.5.2. Cervical cancer studies. The outcomes of pooled SEN and SPE of cervical cancer were shown in Figure 4. A fixed-effect model was applied due to the no significant heterogeneity between the groups. The summary estimates are as follows: SPE, 0.89 (95% CI 0.82-0.93); SEN, 0.83 (95% CI 0.75-0.88); besides, a summary receiver operator characteristic curve was carried out in Figure 5 and AUC was 0.93 (95% CI 0.90-0.95).

Publication bias
Judged by a Deeks funnel plot, there was no evidence of publication bias (P = .13) in the diagnostic analysis (see Fig. 2, Supplemental Content, http://links.lww.com/MD/G414, which illustrates the publication bias judged by Deeks funnel plot for the diagnostic meta-analysis). Publication bias can be measured using Begg funnel plot and Egger test. A Begg funnel plot (Fig. 6A, P = .976) and an Egger test (Fig. 6B, P = .956), indicated that there was no clear publication bias in the analysis of the tumorsuppressor circRNAs in terms of OS. There was also no publication bias in the analysis of oncogenic circRNAs in the case of OS, as indicated by a Begg funnel plot (Fig. 6C, P = .945) and an Egger test (Fig. 6D, P = .900). These outcomes indicated that circRNAs are likely to be a favorable diagnostic and prognostic biomarker.

Discussion
This is the first meta-analysis explored circRNA for the diagnosis and prognosis of ovarian and cervical cancer. In terms of its mechanism and expression level, we regarded circRNAs as tumor promoters or tumor suppressors. Our outcomes demonstrated that dysregulated expression of circRNAs had a different impact on ovarian and cervical cancer survival. High expression levels of oncogenic circRNAs were significantly associated with poor prognoses both in ovarian and cervical cancer. However, elevated expression levels of tumor-suppressor circRNAs were linked with favorable survival time in ovarian cancer. In addition, the diagnostic significance of circRNAs was investigated as well.
The AUC value in ovarian cancer and cervical cancer is 0.89 and 0.93, respectively. This suggesting that circRNA will be a novel biomarker in diagnosis. Besides, there was no obvious heterogeneity and publication bias performed by a SEN analysis. Previous studies discovered circRNA can affect tumorigenesis, metastasis, and rebuilding of the tumor microenvironment. Through the study of Li et al, [16] overexpression of circ100395 can inhibit the proliferation, migration, and invasion of ovarian cancer cells by regulating the miR-1228/p53/EMT axis; at the same time, Cai et al [40] found that circ0000263 was significantly upregulated in cervical cancer cells and ultimately affected the expression of p53 gene. We found p53 gene was a common gene in the development of ovarian cancer and cervical cancer. According, we inferred the potential use of circRNAs to regulated p53 gene might as therapeutic targets for the treatment of ovarian and cervical cancer. Several previous meta-analyses [41][42] reported circRNA has important diagnostic and prognostic value in tumor. Huang et al [43] have summarized that circRNAs may act as important biomarkers for diagnosis and prognosis in diverse cancers by meta-analysis. There are few studies on ovarian and cervical cancer included in previous study. In our research, 22 studies involving 2059 patients with ovarian or cervical cancer were included, which markedly increased the statistical power and made the pooled results more credible. Additionally, the functions and mechanisms of circRNA in both cancers have been clarified.
Despite the promising data, there are some limitations to our study. First, the samples were all taken from cancerous tissues, and the diagnostic value of plasma samples was higher. More plasma samples are needed for further study. Second, all the patients in our study were selected from an Asian population, so the results may be biased. Third, the sample size in this study was small, so a larger clinical study is needed. Finally, some articles did not directly provide the survival data, so we had to estimate the HRs from Kaplan-Meier curves by the method of Parmar et al. [32]

Conclusions
CircRNAs have the prospect of becoming a promising biomarker for diagnosis and prognosis of ovarian and cervical cancer. Accordingly, circRNAs might be novel indicators and targets of therapy for ovarian and cervical cancer.