Background

As a class of endogenous non-coding RNAs with closed-loop structure, circular RNAs (circRNAs) are receiving more and more attention. CircRNAs have been reported to be widely expressed in various human cancers and are implicated in tumorigenesis and progression. The present study aimed to systematically evaluate the clinicopathological, diagnostic and prognostic values of circRNAs in lung cancer.

Methods

We searched literature from PubMed, Web of science, Cochrane Library, EMBASE and Ovid online databases up to May 29, 2020. Statistical analyses were undertaken based on Stata 11.0, Meta-DiSc 1.4, and RevMan 5.3 software.

Results

Finally, a total of 63 eligible articles were included in our meta-analysis, including 18 studies for diagnosis, 22 studies for prognosis and 57 studies for clinicopathological features. In terms of diagnostic values, circRNAs could discriminate between lung cancer patients and the normal individuals with a relatively high pooled area under the curve (AUC) of 0.83 (95%CI, 0.80–0.86). For the prognostic values, we found that elevated expression of oncogenic circRNAs could predict poor survival outcomes based on multivariate analysis (HR = 2.430, 95%CI = 2.003–2.948, P < 0.001 for OS; HR = 2.228, 95%CI = 1.289–3.853, P = 0.004 for DFS) while tumor-suppressor circRNAs was correlated with better OS in univariate analysis (HR = 0.627, 95%CI = 0.519–0.757, P < 0.001). The pooled results suggested that elevated expression of carcinogenic circRNAs was associated with tumor size (OR = 1.676, 95%CI = 1.209–2.323, P = 0.002), smoking statue (OR = 1.260, 95%CI = 1.062–1.494, P = 0.008), TNM stage (OR = 2.345, 95%CI = 1.617–3.399, P < 0.001), differentiation grade (OR = 1.843, 95%CI = 1.228–2.765, P = 0.003), and lymphatic metastasis (OR = 2.097, 95%CI = 1.482–2.967, P < 0.001). Moreover, the expression of tumor-suppressor circRNAs was related to the improved clinicopathological features (lymphatic metastasis: OR = 0.536, 95%CI = 0.311–0.926, P = 0.025).

Conclusions

Our meta-analysis demonstrated that circRNAs could be used as feasible and important biomarkers for diagnosis, prognosis and clinicopathological features in lung cancer.

中文翻译：

---

CircRNAs 作为肺癌诊断和预后的有前途的生物标志物：一项更新的荟萃分析

背景

作为一类具有闭环结构的内源性非编码RNA，环状RNA（circRNA）越来越受到关注。据报道，CircRNA 在各种人类癌症中广泛表达，并与肿瘤发生和进展有关。本研究旨在系统评价 circRNA 在肺癌中的临床病理学、诊断和预后价值。

方法

我们从 PubMed、Web of science、Cochrane 图书馆、EMBASE 和 Ovid 在线数据库中检索了截至 2020 年 5 月 29 日的文献。基于 Stata 11.0、Meta-DiSc 1.4 和 RevMan 5.3 软件进行了统计分析。

结果

最后，共有 63 篇符合条件的文章被纳入我们的荟萃分析，包括 18 篇诊断研究、22 篇预后研究和 57 篇临床病理学特征研究。在诊断价值方面，circRNAs 可以区分肺癌患者和正常个体，曲线下面积（AUC）相对较高，为 0.83（95%CI，0.80-0.86）。对于预后值，我们发现，基于多变量分析，致癌性 circRNA 的表达升高可以预测较差的生存结果（HR = 2.430, 95%CI = 2.003–2.948, P  < 0.001 for OS; HR = 2.228, 95%CI = 1.289 –3.853， DFS 的P  = 0.004），而在单变量分析中抑癌 circRNA 与更好的 OS 相关（HR = 0.627，95%CI = 0.519–0.757，P < 0.001)。汇总结果表明致癌性 circRNA 表达升高与肿瘤大小（OR = 1.676, 95%CI = 1.209–2.323, P  = 0.002）、吸烟状况（OR = 1.260, 95%CI = 1.062–1.494, P  = 0.008)、TNM 分期 (OR = 2.345, 95%CI = 1.617–3.399, P  < 0.001)、分化程度 (OR = 1.843, 95%CI = 1.228–2.765, P  = 0.003) 和淋巴转移 (OR = 2.097) , 95%CI = 1.482–2.967, P  < 0.001)。此外，抑癌基因 circRNA 的表达与改善的临床病理特征有关（淋巴转移：OR = 0.536, 95%CI = 0.311–0.926, P  = 0.025）。

结论

我们的荟萃分析表明，circRNA 可用作肺癌诊断、预后和临床病理学特征的可行且重要的生物标志物。