Prognostic and Clinicopathological Value of Slug Expression in Breast Cancer: A Meta-analysis

Slug as a member of the epithelial-mesenchymal transition(EMT) leads to the decrease of adhesion between cancer cells and the increase of cell migration. Many studies have been reported the relationship between prognosis and slug expression in patients with breast cancer, yet the results are controversial. Therefore, it is necessary to conclude the relationship between slug and breast cancer by a meta-analysis. We looked for studies on breast cancer and slug via PubMed, Scopus and Web of Science .A total of 1458 patients from the nal eight studies were included in this meta-analysis. Overall survival (OS) and Disease-free survival (DFS) were the primary endpoints. Pooled hazard ratio (HR), pooled odds ratio (OR), and 95% condence interval (CI) were calculated to assess the association between slug, prognosis, and clinicopathological parameters (age, tumor size, histological grade, axillary lymph nodes status(LN),TNM stage, ER(estrogen receptor) status, PR(progesterone receptor) status, HER-2(human epidermal growth factor receptor 2)status). Data analysis was carried out by using STATA version 14.0. (Stata Corporation, TX, USA)software. from eight studies. The nal results show that high slug expression leads to poor OS (cid:0)pooled HR = 2.21;95%CI=1.47-3.33;p(cid:0)0.001) and DFS (cid:0)pooled HR = 2.03;95%CI=1.26-3.28 ; P=0.004) in breast cancer. In terms of clinicopathological parameters, the results show that breast cancer patients with high slug expression have higher TNM stage(I-II/III-IV; pooled OR=0.42;95%CI=0.25-0.70;P=0.001), more prone to axillary lymph node metastasis(N+/N0;pooled OR =2.16; 95%CI =1.31-3.56; P=0.003) and more severe ER deciency (positive /negative; pooled OR=0.67;95% CI=0.45-0.99; P=0.042).However, slug is not related to age, histological grade, tumor size, PR status and Her-2 status. This meta-analysis results show that high slug expression in breast cancer is associated with poor OS, DFS and axillary lymph node status, TNM stage and ER status, but not related to age, histological grade, tumor size, PR status and Her-2 status. of the high expression of slug. Second, there were differences in the type and dilution of immunohistochemical antibodies. Third, HR reliability of some prognostic parameters obtained from Kaplan-Meier curve was poor. Fourth , even though we screened through 1,186 articles, we ended up with only eight available studies, and only one of them from a non-Asian population, so more studies were need in non-Asians in the future .


Introduction
Breast cancer is the most common cancer among women worldwide, which is the leading cause of death among women aged 20-50 [1].In the 2018 global cancer statistics, breast cancer accounted for about 11.6% of all cancer cases, tied with lung cancer. There were 0.63 million breast cancer deaths a year [2]. The annual incidence rate is on the rise [3].Which suggests that breast cancer has a signi cant impact on women's health, so it is important to reduce the incidence and improve the prognosis. There are many common clinicopathological parameters include age, axillary lymph nodes status,tumor size, histological grade, TNM stage, PR status, ER status, HER-2 status [4].But none of them could accurately predict the prognosis [5].In the context of precision medicine, we need to identify new prognostic biomarkers in breast cancer, to classify groups of patients with poor clinical outcomes who may bene t from new individualized treatments [6].
Slug (also termed snail2) is a C 2 H 2 zinc-nger transcriptional repressor belonging to the three-member family of snail proteins (Snail, Slug, and Smuc), which mediates sequence-speci c interactions with DNA [7], which has many biological functions, such as cell migration, cell invasion, cell cycle regulation, and stem cell characteristics in tumor cells [8]. What's more, slug is highly expressed in a variety of cancer cells ,including lung cancer, breast cancer, ovarian cancer, esophageal cancer, pancreatic cancer and colorectal cancer etc [9] . High expression of slug in breast cancer is associated with higher TNM stage and increased metastatic potential [10,11].In addition, many studies have found that high expression of slug in breast cancer cells is associated with multiple drug resistance, including resistance to chemotherapy and endocrine therapy [12,13].This makes slug a potential predictor of poor prognosis in breast cancer Many studies have explored the role of slug in the clinicopathological parameters and prognosis of breast cancer, but the results were inconsistent. It has been reported by Wan et al that high expression of slug in breast cancer was correlated with poor survival [14].
However, it is worth noting that Imani et al [15]and Huang et al [16] found that the expression of slug in breast cancer was not related to the prognosis [17].In the face of controversy, it is necessary to do a meta-analysis to determine the relationship between high slug expression in breast cancer and prognosis and clinicopathological parameters (age, tumor size, lymph node metastasis, histological grade, TNM stage, ER, PR, HER-2).

Publication search
Search PubMed, Scopus and Web of Science for articles (for dates up to December 01,2020) with the Keyword: "SLUG" or "SNAIL2" or "SNAI2","breast cancer"or"breast neoplasm" or "breast tumor". In addition, a further manual search of the reference list of eligible studies was conducted to identify additional relevant studies.

Study selection
Studies that meet the following criteria may be included: (1)All studies had prognostic or clinicopathological results.(2) Slug expression in breast cancer cells was detected by immunohistochemistry(IHC). (3) A prognostic indicator of HR is extracted directly or extrapolated indirectly. Summaries of meetings, letters, reviews, meta-analyses, animal model studies, and studies that did not provide su cient data were excluded.

Data extraction
The data of the papers were extracted by two researchers (Zhang and Yang). When disagreements arose, they resolved through discussion. Data extraction includes: Authors, country, year of publication, including number of patients, mean age(year) and median follow-up time(months), survival data, and clinicopathological parameters. In addition, we also extracted the slug antibody dilution, location and cut-off value. For prognosis of HR we extract directly from the paper, if the HR can't be extracted directly, we choose to extract from Kaplan-Meier Curve via Engauge Digitizer Version4.1 (http:// markum mitc hell. github. io/engauge-digi tizer/). We evaluated the quality of articles using the Newcastle Ottawa Scale (Nos)[18] , which has a maximum score of 9, and we will discard articles with a score of 5 or less.

Statistical analysis
We used STATA version 14.0 (Stata Corporation, TX, USA) software to analyze data, the clinicopathological parameters were eventually combined as OR, and the prognostic indicators were nally combined as HR. If we can't extract HR directly, according to Tierney et Al, Kaplan-Meier curve was used to extract the data [19]. P value 0.05 was considered statistically signi cant Cochran Q test and I 2 statistics were used to measure heterogeneity, P value < 0.1 and I 2 value > 50% represented strong heterogeneity [20] at this point the random model was used. Otherwise, the xed-effects model was used [21]. Sensitivity analysis was used to check data stability, The Egger's test was used to detect publication bias, with P value 0.05 indicating publication bias.

Search results
A total of 1,186 articles were retrieved using the search strategy Fig.1 649 records saved after duplicates removed. Next, 596 were excluded after screening of the titles and abstracts. By reading the full text of 53 articles, 43 articles were found to have no available outcome indicators, and 2 articles were bioinformatics analysis. A total of 1458 patients were included in the nal eight studies [10,14,[22][23][24][25][26][27]. The detailed parameters of these articles are listed in Table 1 and Table 2. The quality of all articles was listed (Supplementary

Correlation between slug expression and clinicopathological parameters
A total of 7 articles provide clinicopathological related data, the speci c characteristics were showed in Table 2. The clinicopathological parameters pooled OR values were showed in Table

Sensitivity analysis and publication bias
Our analysis of publication bias using Egger's test correlation test revealed no bias for OS (P =0.751) ( Fig.3a) DFS (P = 0.596) (Fig.3b) and all clinicopathological parameters. In addition, the sensitivity analysis showed that the results were reliable for OS DFS and all clinicopathological parameters.

Discussion
Many studies have explored the role of slug in the prognosis and clinicopathological parameters of breast cancer, but the results were inconsistent. Therefore, we determined the relationship between high slug expression in breast cancer and prognosis and clinicopathological parameters by this meta-analysis. Our meta-analysis shows that high expression of slug in breast cancer is Slug is associated with prognosis in a variety of cancer cells. Liu et al. [30] and Song et al. [31]found that slug is expressed more in lung cancer cells than in normal lung tissue, and high expression of slug in lung tumor cells is associated with poor survival and more aggressive clinicopathological parameters. Chang et al. [32] and Gu et al. [33] found that slug is highly correlated with the invasiveness and drug resistance of ovarian cancer cells. Toiyamo et al. [34] found that slug expression was signi cantly increased in colorectal cancer with high T stage, liver metastasis and lymph node metastasis, it may be a potential prognostic marker for colorectal cancer. The relationship between slug and solid tumor was analyzed in a previous study [16] , interestingly ,it found that slug is associated with a poor prognosis in lung cancer head and neck carcinoma urinary cancer gastrointestinal tract carcinoma, but not in breast cancer.
Metastasis of tumor cells and the drug resistance to anti-tumor therapy are the main causes of poor prognosis in cancer patients [35]. Pan [37]demonstrated that slug is an important factor in promoting the metastasis of breast cancer cells and may be an important marker of metastasis potential. Slug, as a member of EMT, was initially recognized for its involvement in the EMT project,. During the progression of breast cancer, cells and cell adhesion are lost in the EMT process, resulting in migration and invasion [38,39].There are many molecular mechanisms that slug promotes the metastasis of cancer underlying EMT. Liu et al [40] found that slug represses the expression of miR-200b and miR-1,and the inhibition of miR-200b and miR-1 promote EMT and tumor cell invasiveness. Fazilaty et al [41]found that slug can induce TNC and Postn by signal cascade, thus promoting the invasiveness of tumor cells. What's more, Lamouille et al. [42] found that the high expression of slug decrease the expression of epithelial genes and activate the expression of mesenchymal genes, which promotes the metastasis of the tumor cells.
High expression of slug can lead to multiple drug resistance [12,13]. Slug has recently been found to play an important role in tamoxifen resistance of breast cancer [29,43]. The expression of slug prevented the killing effect of tamoxifen on ER(+) breast cancer cells [44]. Slug has been shown to induce endocrine therapy resistance in breast cancer cells by altering the cell survival signaling pathways, leading to worse DFS[45].Musgrove et Al. found that the loss of ER expression due to high expression of slug is the main cause of drug resistance in tamoxifen [46]. Some studies suggested that slug induces tamoxifen resistance by increasing EGFR expression and Erk phosphorylation [47]. what's worse, Li et al. [48] found that slug can induce chemotherapy resistance in cancer cells via PI3K/Akt/GSK3b pathway. These studies con rm that slug leads to poor prognosis in breast cancer, consistent with our Metaanalysis results.
Our meta-analysis shows that slug is a key biomarker for predicting prognosis in breast cancer patients, which is the main nding of this study. No heterogeneity or publication bias was found in this meta-analysis, and sensitivity analysis suggested that our results were reliable. There were limitations in our meta-analysis. First, there was no uniform scoring criterion to de ne the cut-off of the high expression of slug. Second, there were differences in the type and dilution of immunohistochemical antibodies. Third, HR reliability of some prognostic parameters obtained from Kaplan-Meier curve was poor. Fourth , even though we screened through 1,186 articles, we ended up with only eight available studies, and only one of them from a non-Asian population, so more studies were need in non-Asians in the future .

Conclusion
In conclusion, this meta-analysis showed a relationship between slug and poor prognosis in breast cancer. We think slug could be used not only as a biomarker for predicting prognosis, but perhaps as a therapeutic tool in the future. All of which suggests we should do some more research on slug.  Tables   Table 1 haracteristics of studies included. Table 2 Data extracted from studies included.   Figure 1 Flow chart of study selection process.

Figure 2
Forest plot depicting association between slug expression and OS(2a) and DFS(2b) in breast cancer.