A Meta-Analysis of MiRNA-497 and Prognosis of Hepatocellular Carcinoma

Background Recently, microRNA-497 (miR-497) has been reported as a prognostic marker for hepatocellular carcinoma (HCC). However, there is no systematic study summarizing these data. Herein, we elucidated the prognostic role of miR497 in HCC by using meta-analysis. Methods We systematically searched Embase, PubMed, Web of Science, and, China National Knowledge Infrastructure for relevant studies. The two researchers conducted data extraction and quality evaluation independently. We used hazard ratios (HRs), odds ratios (ORs), and their 95% confidence interval (95% CI) to evaluate the relationship between miR-497 expression level and HCC prognosis. Results A total of 6 studies involving 457 participants were included in this meta-analysis. There was a significant association between the lower level of miR-497 expression and the shorter overall survival (HR = 2.17, 95% CI: 1.67–2.84, P < 0.001). Meanwhile, patients with low miR-497 expression were more prone to vascular infiltration (OR = 2.73, 95%: 1.79–4.17, P < 0.001). However, the lower expression level of miR-497 had no significant correlation with TNM (tumor-node-metastasis) stage (OR = 1.47, 95% CI: 0.17–12.49, P=0.47). Conclusions MiR-497 might serve as a prognostic biomarker in HCC, but more clinical studies are needed to confirm this view.


Introduction
Cancer has long been an important factor afecting morbidity and mortality worldwide [1].Hepatocellular carcinoma (HCC) is one of the most common cancers, with mortality ranking fourth among all cancers and second only to pancreatic cancer [2].Although the treatment system for HCC is rapidly developing and several drugs have shown clinical efcacy in phase 3 trials in recent years, the 5-year survival rate of HCC is still very low [3,4].Te difculty of early diagnosis of HCC and the lack of efective therapeutic drugs for advanced HCC are two major problems afecting the survival of patients with HCC [5,6].Terefore, it is necessary to further reveal the pathogenesis of HCC and identify biomarkers with higher specifcity and sensitivity for early diagnosis and prognosis monitoring of HCC.
MicroRNAs (miRNAs) are highly conserved, endogenous nonprotein-encoded small molecules with lengths of 21 to 24 nucleotides, which can bind to the target sequence of the 3′-untranslated region of the target mRNAs, causing degradation or translation inhibition of the target mRNAs at the posttranscriptional level [7].Tey may participate in a variety of cellular processes, including proliferation, differentiation, and apoptosis [8].In recent years, miRNAs have been considered potential biomarkers for cancer prognosis because of their robust expression patterns, stability within cancerous samples, and easy assessment by qRT-PCR [9,10].MicroRNA497 is a highly conserved miRNA located on human chromosome 17p13.1 [11].It can inhibit angiogenesis and metastasis by directly inhibiting vascular endothelial growth factor, thus inhibiting tumor growth, invasion, and metastasis [12].Te main targets of miR-497 include YAP1, WEE1, IGF-1R, IF4E, CCND1, and CCNE1 [13,14].Previous studies have shown that miR-497 acts as a tumor suppressor in a variety of tumors, such as pancreatic cancer, breast cancer, nerve cell cancer, and kidney cancer [14,15].
Te downregulation of miR-497 has been reported in HCC [16].Tus, the role of miR-497 in the prognosis of patients with HCC has attracted much attention.Some studies have shown that the low expression level of miR-497 is signifcantly correlated with the poor prognosis of HCC [13,[17][18][19][20][21].However, these studies have their limitations, such as small sample sizes or controversial results.Terefore, to further clarify the relationship between miR-497 and the prognosis of HCC patients, we conducted this meta-analysis.

Search Strategy.
We systematically searched the databases of PubMed, Web of Science, Embase, and China National Knowledge Infrastructure (CNKI) for studies published in English, Chinese, or other languages up to October 25, 2021, using the following keywords: ("Hepatocarcinoma" or "hepatocellular carcinoma" or "hepatomas" or "liver carcinoma" or "liver cancer" or "HCC") and ("mirna-497" or "mir-497").We also examined the references of retrieved studies to avoid omission.

Inclusion and Exclusion Criteria. Te inclusion criteria included (1) studies focusing on patients pathologically confrmed as HCC patients;
(2) studies focusing on miR-497 expression in HCC tissues; (3) studies with low and high miR-497 expression subgroups; (4) the relationship between the expression level of mi-497 and the prognosis of HCC patients was evaluated; (5) hazard ratio (HR) with 95% CI or survival curves were provided; and (6) the language of the studies was in English or Chinese.Te exclusion criteria were as follows: (1) insufcient data or lack of key information on survival results; (2) review, letter, meeting report, or meta-analysis; and (3) nonhuman experiment or ethical disapproval.

Data Extraction and Quality
Assessment.Te two authors independently extracted useful information from the selected articles.Te following information was extracted: published year, name of the frst author, detection method of miR-497, sample size, patient age, patient sex, cut-of value, follow-up time, the number of patients with diferent TNM stages in low and high miR497 expression level groups, vascular infltration, alpha-fetoprotein (AFP) level, HBV infection, HR, and its 95% CI for overall survival (OS).If the data is not given directly, the data will be extracted through the survival curve.We contacted the authors of the original studies to obtain missing data or clarify ambiguous information.Te two authors independently evaluated the included studies according to the Newcastle Ottawa Scale (NOS).A NOS score ≥6 indicates high quality [22].Any diferences in data extraction and quality evaluation were resolved through consultation.

Statistical Analysis.
Statistical analyses were performed using Review Manager 5.3 (Cochrane Collaboration, Oxford, UK) and Stata 16.0 (Stata Corp., College Station, TX, USA).All tests were 2-tailed.HR and its 95% CI, obtained directly or indirectly from the survival curve using Engauge Digitizer 4.1, were combined and considered statistically signifcant at P < 0.05.Te heterogeneity among studies was tested by a Q-test and I 2 value.If P > 0.05 and I 2 < 50%, the fxed-efect model was used to calculate pooled HR; otherwise, the random efect model was used.Sensitivity analysis was performed by omitting each study at a time to assess the consistency and stability of the pooled results.Te potential publication bias was evaluated using funnel plots, Begg's test, and Egger's test.

Features of the Included Studies.
As shown in Figure 1, 159 studies were preliminarily searched through the above 4 databases.After removing 31 duplicate articles, the remaining 128 articles were screened according to inclusion and exclusion criteria, and 21 full-text articles were further reviewed.Finally, 6 studies with NOS scores greater than 6 were included after detailed screening [13,[17][18][19][20][21].HR and 95% CI of two studies were extracted from the survival curve according to the method provided by Jayne F.T et al.Te six included studies were all published between 2016 and 2021, including 457 patients.All patients were diagnosed with HCC by histopathology (HE staining) and were of Child-Pugh grade A or B. Homogeneous cancer tissues were resected, and RNA in cancer tissues was extracted and amplifed by PCR.With the median level of miR-497 in cancer tissues as the cut-of value, the included patients were divided into the miR-497 high expression group and the miR-497 low expression group.Table 1 shows the basic characteristics and NOS scores of these studies.

Relationship between the Expression Level of miR-497 and OS.
As shown in Figure 2, six studies have reported the relationship between miR-497 levels and OS in HCC patients.Since heterogeneity was not obvious (P � 0.75, I 2 � 0%), we used a fxed-efect model to analyze the low miR-497 expression level and OS.Te combined HRs showed that the low expression level of miR-497 was signifcantly associated with worse OS in patients with HCC (HR � 2.17, 95% CI: 1.67-2.84,P < 0.001).Canadian Journal of Gastroenterology and Hepatology these reports (P < 0.001, I 2 � 89%), we used a random efect model to analyze the combined OR and 95% CI.Te analysis showed that the expression of miR-497 had no signifcant correlation with the tumor-node-metastasis stage (OR � 1.47, 95% CI: 0.17-12.49,P � 0.72) (Figure 3(a)).

Association between miR-497 Expression Level and
Vascular Infltration.As shown in Figure 3(b), fve studies explicitly reported the sample size of vascular infltration.Because the heterogeneity was not obvious (P � 0.52, I 2 � 0%), the fxed-efect model was used to analyze the combined OR and its 95% CI.Te results showed that patients with low expression of miR-497 may be more prone to vascular infltration (OR � 2.73, 95% CI: 1.79-4.17,P < 0.001).

Relationship between miR-497 Expression Level and Other
Clinicopathological Parameters.We also used the combined OR and its 95% CI to analyze the relationship between other pathological indexes of patients with HCC and the expression level of miR-497.In addition to the fxed-efects model for analysis of the age and sex, the random-efects model was selected to analyze all other parameter analyses including AFP level, tumor size, and HBV infection.As shown in Table 2, the combination of OR and 95% CI suggested that the low expression of miR-497 had no signifcant correlation with pathological parameters such as age, sex, AFP level, HBV infection, and tumor size.

Sensitivity Analysis and Publication Bias.
To evaluate the credibility and stability of this meta-analysis, we analyzed the sensitivity of the OS group by omitting one study at a time.Te results showed that no individual study afected the combined HR for OS (Figure 4).Te shape of the funnel plot did not indicate visual evidence of asymmetry (Figure 5).Similarly, Begg's test and Egger's test detected no signifcant publication bias (P > 0.05) (Table 3).

Meta-Regression Analysis.
To evaluate the source of heterogeneity in OS, we performed meta-regression.Te possible covariates including cut-of, sample size (median 76 as the boundary), NOS (mean 6.83 as the boundary), and ethnicity, were analyzed.Te results revealed that these covariates were not the sources of heterogeneity (P > 0.05) (Table 4).Tis indicates that the pooled results may not be afected by these covariates.

Discussion
Increasing evidence has confrmed that the disorder of miRNA expression level is involved in the occurrence and progression of various cancers [23], and miRNA expression level is diferent in normal tissues and tumor tissues [24].Tese abnormally expressed miRNAs can play a variety of biological functions in cancer tissues and directly afect the occurrence, development, and metastasis of tumors [25].Some genes can promote tumorigenesis, such as miR-21 [26], miR155 [27], and miR-222 [28], while others can inhibit tumorigenesis, such as miR-15 [29], miR-139 [30], and mir143 [31].Terefore, it is a perfect choice to detect the changes of miRNA in tumor tissues by various methods and study its impact on the prognosis of cancer patients.
Te occurrence of HCC is a complex process, with unknown etiology, insidious onset, slow early progression, and atypical clinical manifestations.Currently, screening of people at high risk of HCC mainly relies on ultrasonography and AFP detection.However, some small lesions in the early stage are difcult to fnd by ultrasound, and the AFP level does not increase or has a slight increase in 20%-40% of patients with HCC [32].Terefore, HCC is often found to be in an advanced stage, which signifcantly reduces the survival rate of patients with HCC.Liver transplantation, hepatic resection, radiofrequency ablation, transcatheter arterial chemoembolization, targeted therapies based on tyrosine protein kinase inhibitors, and immune checkpoint inhibitors are the most common treatments [2].However, new treatment approaches for HCC are still needed.In recent years, the relationship between HCC and miRNA expression levels is being studied, expecting to clarify the role of miRNA expression in the diagnosis, treatment, and prognosis of HCC patients.
Mir-497 has always been a research hotspot.Previous studies have shown that miR-497 plays a tumor suppressor role in many tumors.Overexpression of miR-497 can inhibit cell proliferation and induce apoptosis in HCC [17].However, miR-497 downregulation can promote tumor angiogenesis and contribute to the occurrence of HCC [32].Although many studies have verifed the close relationship between miR-497 and HCC, it is not convincing because of the small sample size.To further clarify the association between miR-497 and HCC, we performed this metaanalysis.
Te studies on miR-497 and HCC published after 2016 were searched, and 6 articles, involving 457 patients in total, were included in this meta-analysis.We noted that miR497 expression level signifcantly afected the OS of HCC patients, and the OS of patients with low expression of miR497 was shorter.When we studied the relationship between miR-497 and a variety of pathological parameters in patients with HCC, we found that the low expression of miR-497 was not signifcantly correlated with TNM stage, sex, age, AFP level, HBV infection, and tumor size.However, the lower expression level of miR-497 signifcantly means unfavorable vascular infltration for HCC patients.Obvious heterogeneity was discovered when we conducted the relationship between miR-497 and a variety of pathological parameters.To evaluate the heterogeneity, we performed a sensitivity analysis.For example, in the sensitivity analysis of the TNM stage, when we omitted the study of Zhang et al. [17], the heterogeneity became negligible.Tus, this article was the main source of heterogeneity.Te results of the TNM stage by Xie et

Canadian Journal of Gastroenterology and Hepatology
Although this meta-analysis showed that miR-497 level was signifcantly related to the prognosis of HCC and could be used as a marker to judge the prognosis of HCC patients to a certain extent, it still had some defciencies and problems.First, only 6 studies were included in this analysis, including only 457 patients, which is not reliable to a certain extent, so more samples are needed to solve it.Second, the dividing value and follow-up time of the original study we included to distinguish the high and low levels of miR-497 are inconsistent, which will lead to large errors in the combination of HR and OR.Terefore, a clear defnition and a gold standard are needed to solve this problem.Tird, survival data of some eligible studies could not be obtained directly but was extracted from the survival curve, these calculated HRs and their corresponding 95% CIs might also bring several tiny errors.Fourth, this study only proposed that miR-497 was related to the prognosis of liver cancer, but the mechanism and related treatment are still unclear.Fifth, we did not include studies that evaluated the relationship between miR-497 expression and HCC prognosis but did not subdivide miR-497 expression.

. Conclusion
Tis meta-analysis confrmed that the low expression level of miR-497 was signifcantly correlated with low OS and more prone to vascular infltration in patients with HCC.Meanwhile, this study also showed that the expression of miR-497 was not signifcantly related to the tumor metastasis stage, which was contrary to the views of existing studies and required further research.In conclusion, our fndings indicate that the expression level of miR-497 can be used as a marker to predict the OS of HCC patients.More multicenter prospective clinical studies are needed to verify our fndings.

Figure 2 :
Figure 2: Forest plot of pooled HRs of the high expression level of miR-497 for OS.

Figure 3 :
Figure 3: Forest plots evaluating the ORs of miR-497 expression to (a) TNM stage and (b) vascular infltration.

Table 1 :
Te main characteristics of the studies included in this meta-analysis.
Tumor-Node-Metastasis (TNM) Stage.Tree studies reported the TNM classifcation.Due to the obvious heterogeneity ofFigure 1: Flow diagram of the search strategy.Note.Te dashes mean no data.qRT-PCR: quantitative real-time PCR; TNM: Tumor-node-metastasis; OS: overall survival; NOS: Newcastle-Ottawa scale.
al. are not consistent with the combined results of this metaanalysis, which indicates that the relationship between low miR-497 expression level and TNM stage needs further research.

Table 2 :
Te correlation between miR-497 expression and clinicopathologic parameters in HCC.
Figure 4: Te results of sensitivity analysis.