The Expression of Beclin-1 in Hepatocellular Carcinoma and Non-Tumor Liver Tissue: A Meta-Analysis

Background Recently, Beclin-1 expression in hepatocellular carcinoma (HCC) and non-tumor liver tissue have been investigated by several studies. However, the results are controversial. The aim of this study was to clarify the role of Beclin-1 in the occurrence of HCC by comparing the difference of Beclin-1 expression between HCC and non-tumor liver tissue. Methods An electronic retrieve for relevant studies was performed in PubMed, EMBASE, China National Knowledge Infrastructure (CNKI), Wan Fang and Chinese VIP databases updated to December 31, 2017. Newcastle-Ottawa-Scale (NOS) was used to assess the quality of the eligible studies. Sensitivity, subgroup, and publication bias analyses were also carried out in this meta-analysis. Statistical analysis was performed by Review Manager 5.3 and STATA 12.0. Results Six high-quality studies with 357 HCC patients were eligible. There was no significant difference of Beclin-1 expression between HCC and non-tumor liver tissue (OR = 2.48, 95%CI = 0.64-9.58, P = 0.19). However, sensitivity analysis showed that Beclin-1 was lower in HCC than in non-tumor liver tissue after omitting Kang et al.’s study (OR = 4.14, 95% CI = 1.75-9.81, P = 0.001), and heterogeneity was not evident (P = 0.14, I2 = 43%). Subgroup analysis suggested that heterogeneity may stem from ethnicity. The funnel plot and Egger’s test (P = 0.900) demonstrated that no significant publication bias was present in this meta-analysis. Conclusion This meta-analysis indicated that there was no significant difference of Beclin-1 expression between HCC and non-tumor liver tissue.


Introduction
Hepatocellular carcinoma (HCC) is the most frequent type of primary liver cancer and has been a significant health problem worldwide due to its high mortality [1].
Currently, it is generally accepted that hepatitis B/C virus infection, usage of alcohol and aflatoxin B1 exposure are the main risk factors [2]. Since the symptoms of early-stage HCC are atypical, most patients present with metastasis at diagnosis. Therefore the prognosis for HCC patients was poor [3].

9
Autophagy is a metabolic process in which cell components are degraded by lysosomes, thereby maintaining intracellular homeostasis. It is a highly-conserved process regulated by a specific series of autophagy genes (Atg genes) including Atg6 [4].
Beclin-1 was initially thought to be a tumor suppressor.
This viewpoint derived from previous reports that Belin-1 was monoallelically deleted in human prostate, breast and ovarian cancers and was expressed at decreased level in these tumors [6]. The expression of Beclin-1 in HCC has been studied for more than a decade, but controversial results have been obtained.
Qiu et al. [7] found that a significantly low Beclin-1 expression was presented in HCC tissues compared with adjacent non-tumor tissues and Beclin-1 may inhibit tumorigenesis. However, Kang et al. [8] found that there was less Beclin-1 expression in normal liver tissues than that in HCC tissues.
Hence, a meta-analysis investigating the difference of Beclin-1 expression in HCC and non-tumor liver tissue was performed.

Search strategy
This meta-analysis was conducted in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines [9]. PubMed, EMBASE, China National Knowledge Infrastructure (CNKI), Wan Fang and Chinese VIP were searched systematically to identify studies from initial to December 31, 2017. Searches included the terms "Beclin-1" OR "beclin 1" OR "BECN1" OR "ATG6" AND "hepatocellular" OR "liver" OR "hepatic" AND "carcinoma" OR "tumor" OR "neoplasm" OR "cancer".
In addition, the reference lists of relevant reviews and the included literatures were retrieved manually for eligibility.

Criteria for inclusion and exclusion
The criteria for inclusion of eligible studies in this meta-analysis were: (1) English or Chinese studies and full texts were available; (2) case-control studies; (3) HCC tissue and non-tumor liver tissue were pathologically confirmed; (4) Beclin-1 expression in HCC and non-tumor liver tissue were reported; (5) data was available to obtain the odds ratio (OR) and 95% confidence interval (CI).

Data extraction and quality assessment
Two reviewers (Zhiqiang Qin and Xinjuan Yu) separately extracted data from the included studies.
Discrepancies in data extraction were resolved by consensus with a third investigator (Mei Lin). For each study, the following data was extracted: first author, publication year, country, HCC patients' characteristics (number, gender, age range, percentage of Beclin-1 low expression) and detection method of Beclin-1.
The Newcastle-Ottawa-Scale (NOS) was recommended to evaluate the quality of the included studies [10].

Statistical analysis
This meta-analysis was performed by using Review  10 analysis were used to explore sources of heterogeneity.

Publication bias was evaluated by funnel plot and
Egger's test. P < 0.05 was considered statistically significant.

Selection of included studies
The flow diagram of selection steps is displayed in

Studies characteristics and quality assessment
Of the 6 included studies, 4 studies [7,8,12,13] were published in English and the other 2 studies [14,15] were in Chinese. The publication years were from 2013 to 2016 and HCC sample sizes ranged from 35 to 103.
All HCC patients have proven pathological diagnosis.
Among the 6 eligible studies, 4 studies [7,8,14,15] originated from China and the other 2 studies [12,13] were from Egypt. IHC staining was used for investigation of Beclin-1 expression in all the studies.

Beclin-1 expression in HCC and non-tumor liver tissue
As shown in Figure 2, there was no significant

Sensitivity analysis
Sensitivity analysis was performed by omitting 1 study per time in turn. As shown in Tables 3, the difference of Beclin-1 expression in HCC and non-tumor liver tissue was affected by Kang et al.'s study [8]. After omitting this study, Beclin-1 was low in HCC compared with that of non-tumor liver tissue (OR = 4.14, 95% CI = 1.75-9.81, P = 0.001) and heterogeneity was not evident (P = 0.14, I² = 43%).

Subgroup analysis
As shown in Table 4

Publication bias
As shown in Figure 3, funnel plot did not reveal evidence of significant asymmetry in this study. The result was further affirmed by Egger's test (P = 0.900). 13 Figure 3. Funnel plot for publication bias.

DISCUSSION
Currently, it is generally recognized that autophagy is closely linked to a variety of diverse pathologies [16].
Research showed that autophagy may suppress tumorigenesis by protecting the stability of chromosomes and limiting metabolic stress [17].
Beclin-1, an autophagy-related protein, has been found crucial for tumorigenesis [18]. However, the effect of Beclin-1 in the occurrence of HCC is controversial and no meta-analysis is available now. Therefore, we conducted a meta-analysis to clarify the role of Beclin-1 in the carcinogenesis of HCC.
In this meta-analysis, 6  It should be noted that there are several limitations in our meta-analysis. First, articles with positive results are more tended to be published, which may potentially bring publication bias. Second, heterogeneity is a concern. Despite IHC is the only way to detect Beclin-1 expression in all included studies, the judgment standard and cut-off value of Beclin-1 expression was not uniform.
In this meta-analysis, we found that there was no significant difference in Beclin-1 expression between HCC and non-tumor liver tissue.