Low STMN1 is associated with better prognosis in Asian patients with esophageal cancers: a meta-analysis

Background: The role of STMN1 in the development and progression of esophageal carcinoma is not yet determined. The present study aimed to systematically evaluate the correlation between STMN1 and prognosis of patients with esophageal carcinoma. Methods: Electronic databases including PubMed, Embase, the Cochrane library, and Chinese Biomedical Literature Database (CBM) were searched to identify studies evaluating the impact of STMN1 on the survival of esophageal cancer patients, without the language limitation. Two investigators screened the literature according to the inclusion and exclusion criteria and evaluated the quality of the included studies. The combined analysis was performed using RevMan5.3 software. Results: A total of 8 studies, involving 1240 esophageal carcinoma patients, were included in this retrospective design. Meta-analysis showed that esophageal carcinoma patients with low STMN1 had a superior overall survival (OS) and disease-free survival (DFS) than those with high expression of STMN1. Compared to the high expression of STMN1, the 5-year survival rate was significantly higher in patients with low level of STMN1. Patients with high STMN1 expression had a higher risk of experiencing clinical grade III-IV disease, lymph node metastasis, and tumor invasion than those with low STMN1. Conclusion: STMN1 is an indicator for the prognosis of esophageal carcinoma patients.


BACKGROUND
Esophageal cancer is a common malignant tumor of the digestive system, with high morbidity and mortality that severely threatens the health and the quality of life of patients [1]. Esophageal cancer includes esophageal squamous cell carcinoma and esophageal adenocarcinoma; the esophageal squamous cell carcinoma is the most common type in esophageal cancers [1]. Invasion and metastasis are the primary causes of esophageal cancer-related deaths and the key factors threatening the survival of these patients [2]. The transformation is a complicated process and controlled by multiple genes, involving several steps and stages [1].
The microtubule destabilizing phosphoproteins are kinetics of microtubules.
These are expressed in a variety of cancer tissues and participate in cell proliferation, cell movement, microtubule dynamics regulation, and cycle regulation [3]. STMN1, also known as stathmin 1, is encoded by the STMN1 gene localized on human chromosome 1p36.1 [3,4]. The expression of the STMN1 protein can be altered during mitosis and cell division [3]. Thus, it might play a role in the invasion and metastasis of tumors owing to its role in the cellular movement. In recent years, STMN1 has been reported to promote the metastasis of bladder cancer, gastric cancer, colorectal cancer, breast cancer, nasopharyngeal carcinoma, and sarcoma [5,6]. However, the 6 impact of STMN1 on the prognosis of esophageal cancer is yet under investigation, albeit the prognosis of esophageal cancer is closely related to the degree of tumor differentiation, staging, and complication [1]. Poor prognosis of esophageal cancer is often associated with invasiveness and metastasis. Therefore, biological indicators related to the development and prognosis of esophageal cancer are a prerequisite. Nevertheless, STMN1 has been shown to play a key role in the prognosis of esophageal cancer, deeming it as a cancer-associated gene [7-10].
To date, no meta-analysis has evaluated the specific correlation between STMN1 and prognosis of esophageal cancer patients. Thus, the present study used evidence-based literature to statistically analyze the correlation between STMN1 expression and survival and prognosis in esophageal cancer patients and to explore the correlation between STMN1 and the clinicopathological parameters of esophageal cancer in order to provide the prognostic evidence for the management of the disease.

Literature search
Electronic databases including PubMed, Embase, the Cochrane library, and 7 Chinese Biomedical Literature Database (CBM) were searched to identify the potentially eligible without language limitation. The retrieval period was from the inception of the database to September 2018. Studies evaluating the impact of STMN1 on the survival of esophageal cancer patients were collected. The search terms were as follows and were used in various combinations: STMN1, stathmin 1, esophageal carcinoma, esophageal cancer, esophageal squamous cell carcinoma, esophageal adenocarcinoma, survival, prognosis, and metastasis. Exclusion criteria (1) STMN1 detection method and the positive judgment criteria were not mentioned; (2) Duplicate reports or studies that were unusable, such as similar data and poor-quality data; (3) Summary, review, conference summary, and case report articles; (4) The study model was not human.

Data extraction and literature quality evaluation
The relevant potential studies were retrieved from the above search. Then, the titles and abstracts of the articles were reviewed, and those that failed to meet the inclusion criteria were excluded. The full-text was reviewed if the literature consisting of titles and abstracts was difficult to judge. The process of study selection was completed by two researchers, independently.
In the event of discrepancies in the conclusions, a third researcher was involved.
The quality of the selected studies was evaluated according to the non-randomized controlled Newcastle-Ottawa Scale (NOS) [11] by two researchers, independently. The inconsistencies were resolved with the help from a third researcher. After assessing the quality of the literature, the 9 data of interest were extracted. The baseline characteristics of the included studies such as the first author, publication time, sex, sample size, detection method, positive expression cutoff value, and sample source were recorded.
The primary outcomes were overall survival (OS) and disease-free survival (DFS). The secondary outcomes including the 5-year survival rate and various parameters of esophageal cancer such as Tumor Node Metastasis (TNM) stage, lymph node metastasis, differentiation degree, and invasion were also extracted.

Statistical analysis
Meta-analysis was performed using RevMan 5.3 software [12]. The heterogeneity between each included study was assessed using the χ 2 test.
When the results of each study were not statistically heterogeneous (P>0.1, I 2 <50%), a fixed effects model was used, whereas, in the case of statistical heterogeneity of each study (P<0.1, I 2 >50%), the source of heterogeneity was analyzed by a subgroup analysis of the preceding factors. When the results of the study were not significantly heterogeneous, the random effects model was applied. The sensitivity analysis was performed to test the heterogeneity of the low-quality studies. Descriptive analysis was used if the data were limited. Begg's and Egger's tests were used to construct the funnel plot, and determine the publication bias. 10 Calculation and transformation of lnHR and its variance: If the included study already provided HR values and 95% CI, the natural logarithm of HR and the 95% CI can be calculated directly; if the HR value was not provided directly, the HR for OS of high STMN1 vs. low STMN1 was calculated from the survival curve as described previously [13,14]. For the 5-year survival rate, the number of surviving patients in each group was calculated. If the combined HR and the 95% CI were >1, the high expression of STMN1 was deemed as a risk factor for the prognosis of esophageal cancer.
Furthermore, the patients were classified into different groups based on various parameters of esophageal cancer, such as TNM stage, lymph node metastasis, differentiation degree, and invasion. For clinical stages, the patients were grouped into I+II and III+IV groups. For lymph node metastasis and invasion, the yes and no groups were formed. For differentiation degree, the patients were classified into well and intermediate and poor groups. The risk ratio (RR) of high vs. low STMN1 expression in the III+IV, yes, and poor groups was calculated. A difference of P<0.05 was considered as statistically significant.

Search results and quality evaluation
The initial search retrieved a total of 70 articles. The titles and the abstracts were read, and 58 articles including repetitions, animal experiments, and reviews were excluded. Subsequently, the full-texts were analyzed, and four articles that did not meet the inclusion criteria were excluded. Of these, two studies were discarded as they used the data from the same patient population, and the other two studies did not provide sufficient data. Finally, a total of eight studies [7, [15][16][17][18][19][20][21] retrospective studies were included, which encompassed a total of 1240 patients with esophageal cancer. The study selection process is presented in Fig. 1.  Table 1.
The NOS criteria were employed to assess the quality of the included studies. The quality evaluation demonstrated that the eight studies had distinct diagnostic criteria, sufficient sample size, and data of interest and were ascribed six stars, indicating good quality (Table 2).

OS
In order to determine the effect of STMN1 on OS in patients with esophageal cancer, we extracted the HR of high STMN1 vs. low STMN1 from six studies [7, 15-17, 19, 21]. As shown in Fig. 2A 13

5-year survival rate
Five studies [7, 15,17,19,21] reported the 5-year survival rate in patients with different expression levels of STMN1, and slightly significant heterogeneity was detected (I 2 =56%). Therefore, a random effects model was used for the meta-analysis. A total of 132/443 patients exhibited high STMN1 expression, and 264/428 patients had low STMN1 (Fig. 2C)
Therefore, a random effects model was used. The results from Fig. 3B showed that low STMN1 expression was in favor of well or intermediate differentiation, albeit without significant difference (RR=1.88, 95% CI: 0.86-4.10; P=0.11). Due to identical sample sources, protein detection method, and disease, it is difficult to conduct a subgroup analysis. Therefore, we performed sensitivity analysis, and the results showed that the study by He et al. introduced a significant heterogeneity (Supplemental Fig. 1).

STMN1 and invasion
Three studies [7, 18,20] reported a number of patients with different expression of STMN1 in the positive invasion group, and no statistical heterogeneity was observed between the studies (I 2 =8%). Therefore, a fixed effects model was applied. The results showed that a 1.75-fold risk of experiencing invasion was found in patients with high STMN1 as compared to those with low expression of STMN1 (RR=1.75, 95% CI: 1.43-2.14; P<0.00001).

STMN1 and lymph node metastasis
Six studies [7, 16,[18][19][20][21] reported a number of esophageal cancer patients with high vs. low expression of STMN1 with positive lymph node metastasis, and significant heterogeneity was detected between the included studies (I 2 =87%); thus, the random effects model was applied. As shown in Fig. 3D, the combined result showed that patients with high STMN1 expression had a significantly high risk of developing lymph node metastasis than those with low STMN1 expression (RR=2.33, 95% CI: 1.44-3.77; P=0.0006).
Because of significant heterogeneity, we performed a sensitivity analysis to determine the responsible study. As illustrated in Supplemental Fig. 2, the study of Suzuki et al. might be the main source of differences.

Publication bias analysis
A funnel plot was used to detect the publication bias based on the data of OS and 5-year survival. The included studies were presented on the respective plots and spread out with the combined values as the center. The distribution and symmetry of the plots suggested the absence of significant 16 publication bias (Fig. 4)

DISCUSSION
In this study, we performed a systematic review and comprehensive search to identify clinical studies that fulfilled the inclusion criteria. These retrospective studies had determined the potential correlation between STMN1 expression and the prognosis of patients with esophageal carcinoma, as well as clinical stage, lymphatic metastasis, tissue differentiation, and invasion. Finally, eight studies comprising of a total of 1240 cases were included for quantitative analysis. The results of the meta-analysis showed that the high expression of STMN1 was an indicator of poor OS and DFS in patients with esophageal cancer. High STMN1 expression was associated with an increased risk of poor clinical stage, invasion, lymph node metastasis, and poor tissue differentiation grades, suggesting that STMN1 might be a potential biomarker for esophageal cancer patients after surgery. 17 The poor prognosis of esophageal cancer is often associated with invasiveness and metastasis [1,11]. Therefore, the physiological indicators related to the development and prognosis of esophageal cancer are required.
In recent years, STMN1 has been reported to play a key role in the prognosis of esophageal cancer, suggesting that it is a cancer-associated gene [5,9,22,23]. In the present meta-analysis, six studies compared the OS of esophageal cancer patients with high vs. low expression of STMN1, and the results suggested that high STMN1 was associated with poor OS.
Therefore, it can be speculated that high STMN1 expression might play a vital role in predicting the outcome of esophageal cancer. Subsequently, we found that high expression of STMN1 was involved in advanced stage, lymphatic metastasis, and invasion.
The study included 26 studies involving 5335 patients, and the results showed that high STMN1 yielded a poor OS (HR=2.17), and DFS (HR=2.46).
These findings were in accordance with the results of the current study.
Both studies suggested a negative correlation between the expression of STMN1 and survival of various cancer patients. However, the specific link between STMN1 and prognosis of esophageal cancer is not elucidated in these studies. Herein, we provided additional evidence about the prognostic role of STMN1 in esophageal cancer patients.
Nevertheless, the present study has some limitations. First, although we conducted a relatively comprehensive search, the lack of gray literature might lead to publication bias. All the included studies were conducted in Asia, which might restrict the reliability. Second, the quality of the retrospective studies was moderate. Some of the included studies might have a selection bias. Two studies did not provide survival data and might introduce selective reporting bias. Third, the number of retrieved studies 19 was small with a limited number of patients. Four, heterogeneities were detected in terms of baseline characteristics, tumor differentiation, and clinical staging, as well as the definition of high or low STMN1. In order to decrease the impact of these differences on the pooled outcomes, the random effects model was used for improved credibility. In addition, the sensitivity analysis was performed to explore the source of heterogeneity.

CONCLUSION
In summary, the expression of STMN1 in esophageal cancer is a prognostic factor for survival. The high STMN1 expression is associated with poor clinical stage, positive lymph node metastasis, and poor histological grades, indicating that STMN1 may play a critical role in the development and progression of esophageal carcinoma. However, additional rigorous, meticulous, high-quality studies are essential to substantiate these findings.

Ethics approval and consent to participate
This is a meta-analysis of data from published articles, it is necessary ethical approval

Consent for publication
Not applicable.

Availability of data and material
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Competing interests
All authors declare that they have no any conflict of interests.

Funding
The study Supported by the National Natural Science Fund of China.