Prognostic Significance of E-Cadherin Expression in Hepatocellular Carcinoma: A Meta-Analysis

Backgrounds Hepatocellular Carcinoma (HCC) is one of the most common malignancy of liver and HCC-related morbidity and mortality remains at high level. Researchers had investigated whether and how reduced E-cadherin expression impacted the prognosis of patients with HCC but the results reported by different teams remain inconclusive. Methods A systematic literature search was performed in all available databases to retrieve eligible studies and identify all relevant data, which could be used to evaluate the correlation between reduced E-cadherin expression and clinicopathological features and prognosis for HCC patients. A fixed or random effects model was used in this meta-analysis to calculate the pooled odds ratios (OR) and weighted mean differences (WMD) with 95% confidence intervals (CI). Results Total 2439 patients in thirty studies matched the selection criteria. Aggregation of the data suggested that reduced E-cadherin expression in HCC patients correlated with poor 1-, 3- and 5-year overall survival. The combined ORs were 0.50 (n = 13 studies, 95% CI: 0.37–0.67, Z = 4.49, P<0.00001), 0.39 (n = 13 studies, 95% CI: 0.28–0.56, Z = 5.12, P<0.00001), 0.40 (n = 11 studies, 95% CI: 0.25–0.64, Z = 3.82, P = 0.0001), respectively. Additionally, the pooled analysis denoted that reduced E-cadherin expression negatively impacts recurrence-free survival (RSF) with no significant heterogeneity. The pooled ORs for 1-, 3- and 5- year RSF affected by down-regulated E-cadherin were 0.73 (n = 6 studies, 95% CI: 0.54–1.00, Z = 1.95, P = 0.05), 0.70 (n = 6 studies, 95% CI: 0.52–0.95, Z = 2.32, P = 0.02), 0.66 (n = 5 studies, 95% CI: 0.48–0.90, Z = 2.64, P = 0.008). And what’s more, reduced E-cadherin expression tended to be significantly associated with metastasis (OR = 0.31, 95% CI: 0.16–0.60, Z = 3.50, P = 0.0005), vascular invasion (OR = 0.76, 95% CI: 0.59–0.98, Z = 2.14, P = 0.03), advanced differentiation grade (OR = 0.31, 95% CI: 0.21–0.45, Z = 6.04, P<0.00001) and advanced TMN stage (T3/T4 versus T1/T2) (OR = 0.61,95% CI:0.38–0.98, Z = 2.05, P = 0.04). Conclusions Reduced E-cadherin expression indicates a poor prognosis for patients with HCC, and it may have predictive potential for prognosis of HCC patients.


Introduction
Hepatocellular carcinoma (HCC) is not only the seventh most frequent human malignant tumors, but also the second highest cause of cancer-related death from poles to poles. It was estimated that HCC had caused about 746,000 deaths in 2012 [1][2][3][4][5]. Despite the considerable advancement in new-developed therapies, the overall mortality and morbidity for HCC are high and the prognosis of patients remains disappointed [6]. On the one hand, it might be due to that the time of diagnosing HCC is always at the advanced stage; on the other hand, clinicopathological features of HCC, such as differentiation, tumor grade/stage, lymph node status, depth of tumor invasion, and metastasis all influence the prognosis of patients with HCC. Consequently, new biomarkers that could be used effectively to anticipate the prognosis of patients with HCC are in urgent need [7][8][9][10].
Nowadays, the role of cell adhesion molecules, such as cadherin, catenin, selectin, integrin, whose expression levels change dynamically in tumor and have much association with tumor invasion and metastasis, has attached more and more attention [11][12][13][14]. These molecules could serve as potential marker predicting the prognostic significance for patients with HCC.
E-cadherin is the major member of cell adhesion molecule family expressed by epithelial cells [15]. It is a transmembrane calcium-dependent cell adhesion protein with a molecular weight of 120-KD. E-cadherin regulates cell differentiation and maintains cell structure. Detected by immunohistochemistry, reduced Ecadherin expression has been observed in a wide variety of tumors, characterized by decreased epithelial cell adhesion and increased motility and invasiveness of tumor cells [16][17][18][19][20][21]. Vast work has been done to examine the correlation of reduced E-cadherin expression with prognostic significance for patients with HCC but  no concensus was achieved to date [22,23]. Consequently, basing on retrospective cohort studies, we carried out this meta-analysis to systematically and comprehensively investigate whether and how the reduced E-cadherin expression impacted prognosis of HCCs.

Criteria for Inclusion and Exclusion
To make this meta-analysis meet the high standards, studies had to fulfill the following criteria: (1) patients with distinctive hepatocellular carcinoma diagnosis by pathology but without restriction on age or ethnicity; (2) reduced E-cadherin expression was measured by immunohistological chemistry (IHC) or other methods in primary HCC tissues; (3) clinical trials or reports on Ecadherin expression study in HCC were published in English; (4) valid data were provided directly or could be calculated indirectly; (5) the study with the highest quality assessment was enrolled when trials on similar objects were reported many times.
Abstracts, editorials, letters and expert opinions, reviews without original data, case reports and studies lack of control groups were excluded. Studies and data were also excluded if: (1) articles about animals or cell lines; (2) the outcomes or parameters of patients were not clearly reported (e.g. omitting standard deviations (SDs) (3) conference records; (4) no related data required for necessary analysis; (5) overlapping articles.

Data Extraction and Literature Quality Assessment
Independently, valid data were retrieved from eligible studies by two reviewers (JC and JZ) and relevant characteristics were listed as follows: (1) first author's name; (2) publication date; (3) study population characteristics; (4) disease stage; (5) the methods used to evaluate E-cadherin levels; (6) corporations of antibody; (7) percentage of reduced E-cadherin expression ( Table 1). All relevant text, tables and figures were reviewed for data extraction. Any divergence was ironed out by discussion with the third reviewer (RM) for final expectation of consensus. The quality of each included study was assessed by utilizing the Centre of Evidence-Based Medicine.

Statistical Analysis
This meta-analysis was performed using the Review Manager (RevMan) software (version 5.2; Cochrane collaboration, http:ims. cochrane.org/revman/download) and STATA (version 12.0, Stata Corp. College Station, Texas) [24]. Odds ratios (OR), together with 95% confidence intervals (CI), was analyzed to estimate whether and how reduced E-cadherin expression impacts the prognosis of HCCs. Pooled values of ORs and 95% CIs, as the recommended summary statistics for meta-analysis, were calculated using either a fixed-effects or a random-effects model. Heterogeneity among the outcomes of enrolled studies in this meta-analysis was evaluated by using Chi-square based Q statistical test [25]. And I 2 statistic was calculated to quantify the total variation consistent with inter-study heterogeneity, ranging from 0% to 100% Heterogeneity was significant and unacceptable while I 2 statistic was greater than 50%. P,0.05 in Q statistical test was considered statistically significant. Choose fixed-effects models to calculate effect size estimates for those studies lack of heterogeneity with a P value for Q-test higher than 0.10. On the contrary, random-effects models were used when P#0.10. The funnel plots was made by utilizing Egger's test and Begg's test to examine the risk of potential publication bias. Then, trim and fill analyses were used to evaluate the stability of our meta-analysis results if the plots were asymmetrical.

Selection of trials
As shown in Fig. 1, 904 potentially eligible studies were screened out in the preliminary search. 844 articles were excluded for their improper titles and abstracts and 60 ones were captured after reviewing their full text for the relevance with the discussed topic. 30 studies was ultimately excluded due to a lack of clearly quantitative data on E-cadherin expression level in HCC. Thus, 30 studies, with more detailed and sufficient evaluation, met our entry criteria and were retrieved for further analysis. The flow diagram of study selection procedure was depicted in Fig. 1.

Study Characteristics
The related clinical data of the enrolled 30 studies with a total of 2439 patients are depicted in Table 1 Table 2, all the 30 studies were evaluated blindly and the cases were grouped randomly according to the provided parameters without considering their age, gender, stage, pathological type and methods.

Meta-Analysis of Clinicopathology
In this meta-analysis, clinicopathologic features, such as differentiation grade, TMN stages, metastasis, vascular invasion, tumor encapsulation and liver cirrhosis, impacted by reduced or preserved E-cadherin expression was compared comprehensively on the basis of these 30 enrolled studies, in order to assess the association between E-cadherin expression and these clinicopathologic parameters. Some studies stated that lower E-cadherin levels unfavorably impacted clinicopathologic parameters [33,43,51] while the other studies found no significant effect [22,29,46,47,53]. Therefore, we carried out this meta-analysis with expectation of achievement of concensus about the correlation of E-cadherin expression and each clinicopathologic parameter.

Publication Bias
Begg's test indicated that there was seemingly significant publication bias in OS and several other clinicopathologic parameters after assessing the funnel plot ( Figure S1a-c, Figure  S2a-c, Figure S3, Figure S4, Figure S5, Figure S6, Figure S7, Figure S8) for the studies included in our meta-analysis.

Discussion
Meta-analytical technique is a useful tool in clinical researches and has been utilized more and more commonly. It can evaluate previous studies qualitatively and quantitatively, especially for those subjects still with controversial results, by integrating and comparing these results to estimate the outcome of interests. What's worth mentioning, so far, tremendous work dedicated to investigating the relationship of E-cadherin levels and the prognosis of patients with HCC has been done with achieving no concensus. Therefore, we took the first effort to conduct a systematical and comprehensive meta-analysis to assess the relationship between them two. And it would also obviously provide useful information for clinical decision-making and effective targets for clinical therapies to treat HCCs.
Despite new therapies of HCC arising continually, the prognosis remains not very optimistic recently. That's why many researchers have been dedicated to finding out predictors of prognosis. As is known to us all, many prognostic markers, such as surviving and MMP9, have been well studied. These markers could influence   tumor metastasis and recurrence, the main two causes leading to poor prognosis. But unfortunately, all these markers alone could not predict the prognosis of patients with HCC reliably and exactly. So more iconic markers are needed as supplementary.
A newly-developed program, namely epithelial-mesenchymal transition (EMT) has been evidenced to participate in promoting progression and metastases of many epithelium-derived carcinoma including HCC [54]. During the process of EMT, epithelial cells actively downregulate cell-cell adhesion systems, lose polarity, and acquire a mesenchymal phenotype. This phenotype enables tumor cells to infiltrate surrounding tissues, and thus license these cells to metastasize in distant sites [21]. What's more, snail and twist1 are the core transcription inhibitory factors during the process of EMT. The two factors can directly lead to reduced E-cadherin expression which is gradually became a hot spot in the field of cancer research. Downregulated E-cadherin expression indicates worse prognosis in some cancer [21]. E-cadherin is the major member of cell adhesion molecule family expressed by epithelial cells. It plays very important role in cell adhesion and differentiation [55]. According to the latest literatures [41,42,53], reduced E-cadherin expression had an adverse effect on the prognosis of patients with HCC and suggested that E-cadherin might be a factor to predict prognosis of patients with HCC.
We carried out this meta-analysis to examine whether and how E-cadherin level impacts the prognosis of patients with HCC. All available data are extracted from multiple databases including The Pubmed, Elsevier, Embase, Cochrane Library, and Web of Science. Low E-cadherin expression was observed in 41.95% of 2415 tissue samples included in our meta-analysis. Moreover, based on those extracted data, the association of reduced Ecadherin expression with OS, RFS, differentiation grade, metastasis, vascular invasion, TMN stage, tumor encapsulation and liver cirrhosis of HCCs was investigated. It was found that HCC with reduced E-cadherin expression became more aggressive and metastatic. Particularly, the results of this meta-analysis suggested that there was significant correlation between reduced E-cadherin expression and poor OS and RFS, indicating that reduced Ecadherin expression exerted a harmful effect on prognosis of patients with HCC. Moreover, lower E-cadherin level had significant correlation with metastasis, vascular invasion, advanced differentiation grade and TMN stages (T3/T4 versus T1/T2). But no significant association was found between lower E-cadherin level and poor tumor encapsulation and liver cirrhosis. All these results, taken together, denoted that reduced E-cadherin expression significantly correlated with poor prognosis of HCC.
However, some limitations need to be interpreted cautiously for further consideration in this meta-analysis. First, heterogeneity was inevitable among the groups due to impossibility to match patient characteristics in all studies. We used a random-effects model in order to eliminate variations across studies. Although it could not necessarily rule out the effect of heterogeneity among studies, the adverse influence will be weakened to some degree. Second, bias was unavoidable for clinical evidence because the relevant data were extracted from non-randomized controlled trials (NRCTs). The potential risks exist to weaken the results of large sample with better quality and strengthen the effect of the small sample with worse quality. Third, studies performed with positive results or significant outcomes will be apt to be published, suggesting a potential publication bias. Fourth, reports in other languages than English were excluded, so potential language bias may be present in our meta-analysis. Fifth, a significant heterogeneity might also be brought about in this meta-analysis by the difference of the antibodies used to test E-cadherin expression. Besides, other clinical characteristics of patients such as age, sex, different chemotherapies and radiotherapies in each study will obviously lead to bias. Further investigation should be given in determining whether these factors influence the results of the meta-analysis.
The prognostic significance of reduced or constant E-cadherin expression for HCCs was identified by comparing the depth of tumor invasion, lymph node metastasis, and clinical stage, cell differentiation, tumor grade and other clinicopathological features. Eventually, the data showed that reduced E-cadherin levels significantly associated with poor prognosis for patients with HCC. Thus, reduced E-cadherin expression may serve as a potential predictor for prognosis of patients with HCC.