CircSMARCA5 functions as a potential biomarker for clinicopathology and therapy in solid tumors: A systematic review and meta‐analysis

Circular RNAs (circRNAs) have been implicated in cancer proliferation, migration, and invasion. Among various circRNAs, circSMARAC5 attracted our great attention. The research aimed at broadening the knowledge on circSMARCA5 and exploring its function in clinicopathology and therapy in solid tumors. The incorporated research was explored via Web of Science, PubMed, Embase, and Cochrane Library, consisting of 1048 patients. This study was calculated using STATA 12.0 and Review Manager 5.3 software. Clinicopathological characteristics and therapeutic targets were analyzed using pooled odd ratios (ORs) with 95% confidence intervals (CIs). And the ceRNA network of circSMARCA5 was constructed via Cytoscape 3.9.0. Eventually, there were eight studies included in conducting clinicopathological characteristics and four literature used in exploring therapeutic targets. Patients with low expression of circSMARCA5 were closely associated with gender (OR = 1.367, 95% CI = 1.003–1.862, p = .048) and pathological differentiation (OR = 1.627, 95% CI = 1.130–2.343, p = .009). Among these modulating axes, the most commonly microRNA was miR‐432. In the near future, circSMARCA5 may be a promising diagnostic biomarker and a new therapeutic target.


| INTRODUCTION
After heart disease owned the highest death rate, cancer has ranked as the second major cause of death among people all over the country, where the mortality rates have been rising during most of the 20th century. According to cancer statistics in 2021, 1  Therefore, it is critical to implement further studies on molecular biomarkers and develop more effective therapeutic targets. Circular RNAs (circRNAs) possess a circular structure without a conventional 5 0 cap to 3 0 polyadenylated tail, which are a diverse Abbreviations: 95% CI, 95% confidence interval; CS, chemotherapy sensitivity; LNM, lymph node metastasis; NA, not applicable; NOS, Newcastle-Ottawa Scale; OR, odd ratio; PRISMA, preferred reporting items for systemic reviews and meta-analyses; qRT-PCR, quantitative real-time reverse transcription-polymerase chain reaction; RFS, recurrence-free survival. range of RNA transcripts with no protein-coding role. 4 On account of stable circular configuration, RNase R was unable to degrade cir-cRNAs. 5 The preceding studies have revealed that circRNAs are implicated in cancer proliferation, migration, and invasion based on their conservative structure. 6 A study has confirmed that circRNAs could sponge microRNA and RNA-binding protein, regulating the splicing of its parent mRNA and regulating gene expression. 7 Moreover, cir-cRNAs could act as a model, which performed the function of translating or synthesizing peptides or proteins. 8 However, the research about the exact functional mechanism of circRNAs is still unknown.
Different circRNAs could exert inverse modulating effect in solid tumors, such as circularRHOBTB3, 9 circ_0008285, 10 circTADA2A, 11 and circFNDC3B 12 could repress tumor metastasis in colorectal cancer (CRC) cells. Circ5615, 13 circFAT1, 14 circPRTM5, 15 and circCAMSAP1 16 were significantly upregulated in CRC, thus promoting malignant phenotype. Due to different expressions of diverse circRNAs and increasing deaths of solid tumors, it is critical to explore the role of circRNAs in physiological and pathological processes in solid tumors.
Among these diverse circRNAs, circSMARCA5 attracted our great attention. CircSMARCA5 was encoded by SMARCA5 gene, which was located at chr4: 1444662-144465125. 17 CircSMARCA5 has been implicated to be downregulated in hepatocellular carcinoma (HCC) and glioblastoma multiform, but upregulated in prostate cancer (PCa). [17][18][19] Additionally, a study demonstrated that double hydrogen testosterone stimulation increased circSMARCA5 expression depending on the releasing dosage of hormones. 17 Similarly, another study also illustrated that circSMARCA5 upregulation reduced the viability of cells in cisplatin-treated along with gemcitabine-treated cells. 20 Therefore, circSMARCA5 could assume the role of diagnostic biomarker and therapeutic target. The purpose of this study was to broaden the knowledge of circSMARCA5 and explore the link of clinicopathological characteristics and therapeutic targets with circSMARCA5 expression.

| Search strategy and literature collection
The passage was in line with the preferred reporting items for systemic reviews and meta-analyses (PRISMA) declaration criteria. 21 The incorporated research was explored via Web of Science, PubMed, Embase, and Cochrane Library updated to October 31, 2021. The searching terms we used were "circSMARCA5" or "circRNAs SMARCA5" or "circ-SMARCA5" or "cancer" or "tumor" or "carcinoma." We screened references of the retrieved articles in order to prevent articles omission.

| Inclusion and exclusion criteria
Screened studies for eligibility were defined as the following criteria: (1) the included studies explored the association between circSMARCA5 and solid tumors; (2) full text should be available for readers; (3) the expression of circSMARCA5 in solid tumors was measured through a consequence of quantitative real-time reverse transcription-polymerase chain reaction (qRT-PCR); and (4) chemotherapy sensitivity was also described in the studies. Exclusion criteria were exhibited as below: (1) duplicated literature; (2) irrelevant to cir-cSMARCA5 and cancer; (3) meta-analysis, review articles, comments, responses, cases, reports, and clinical guidelines; and (4) studies without usable clinical data.

| Statistical methods
This study was performed using STATA 12.0 and Review Manager 5.3 software. The association between the expression of circSMARCA5 and clinicopathological characteristics was measured via chi-square test.
Pooled odds ratio (OR) > 1 indicated the expression of circSMARCA5 correlated with poor clinicopathological outcomes. A p not more than .05 indicates a significant difference, with a 95% confidence level. 22 The heterogeneity was evaluated through I 2 . A random-effect model should be taken into account, if the I 2 value > 50%. Otherwise, a fixed-effect model needed to be brought into the passage, if the I 2 value < 50%. 23 After that, we omitted the studies one by one and carried out sensitivity analysis to deal with the significant heterogeneity. The publication bias of the included studies was qualitatively tested by a funnel chart and qualitatively tested by Begg's and Egger's tests.

| Quality evaluation of primary studies
The Newcastle-Ottawa Scale (NOS) standard was employed with the attention to evaluating the bias of publication, samples, comparability of case and control, exposure, and outcome. 21 The assessing scores of included studies fluctuated from 0 and 9. Obtaining 7 or more scores were considered high quality.

| Construction of circSMARCA5 mediated the ceRNA network
We screened all studies that described circSMARCA5 and collected the molecular modulating mechanisms of circSMARCA5 in the retrieved studies. As a result, the ceRNA network of circSMARCA5 was constructed via Cytoscape.

| Literature information
The flowchart of included studies is displayed in Figure 1

| Study characteristics
The important characteristics of eight included studies are shown in Table 1. These articles were published ranging from 2017 to 2021. The F I G U R E 1 Flow chart of included studies illustrating the selecting process.
T A B L E 1 Study characteristics in this meta-analysis.

Study
Year

| Association between circSMARCA5 expression and clinicopathological characteristics
As shown in Table 2 and Figure 2, OR and p were applied to estimate the association between circSMARCA5 expression and clinicopathologi-

| Sensitivity analysis and publication bias
The sensitivity analysis results have shown that one study 37 made a great difference on pooled ORs, resulting in huge heterogeneity ( Figure 3). The funnel plot is shown in Figure 4.

| ceRNA network of circSMARCA5 in this meta-analysis
To investigate the mechanisms of circSMARCA5, the ceRNA network was constructed from 20 screened studies that described molecule modulating axis of circSMARCA5. As illustrated in

| CONCLUSION
In conclusion, our study indicates that circSMARCA5 expression is

CONFLICT OF INTEREST STATEMENT
The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT
All analysis data are included in this passage.

ETHICS STATEMENT
Not applicable.