Abstract

Background: MicroRNA-135 (miR-135) is a well-known non-coding RNA that has been demonstrated to participate in tumorigenesis and cancer development; however, the clinical prognostic value of miR-135 in digestive system cancers remains controversial. This meta-analysis aims to explore the potential value of miR-135 as a prognostic marker for digestive system cancers.

Methods: The PubMed, Embase, Cochrane Library, and Web of Science databases were searched for eligible articles published before 31 August 2019. Stata 12.0 software was used to analyze the overall survival (OS), disease-free survival (DFS), and recurrence-free survival (RFS) rates to access the prognostic value of miR-135 in digestive system cancers. We then used The Cancer Genome Atlas (TCGA) datasets to validate the meta-analysis results.

Results A total of 1470 patients from 17 studies were included in this meta-analysis. The pooled results showed that enhanced miR-135 expression was significantly associated with poor OR (hazard ratio (HR): 1.790; 95% confidence interval (95% CI): 1.577–2.031; P=0.000), DFS (HR: 1.482; 95% CI: 0.914–2.403; P=0.110), and RFS (HR: 3.994; 95% CI: 1.363–11.697; P=0.012) in digestive system cancers. A sensitivity analysis confirmed the reliability of our findings, and no significant publication bias was observed.

Conclusion: MiR-135 can be used as a novel biomarker for patients with digestive system cancers. We look forward to future large-scale clinical studies that will investigate the prognostic value of miR-135.

Introduction

Digestive system cancers are one of the most common malignancies throughout the world. Such digestive system cancers include esophageal, gastric, colorectal, liver, gallbladder, and pancreatic cancers (PCs). According to the latest reports, there are approximately 4.9 million new diagnosed cases and 3.5 million deaths for digestive system cancers that occur worldwide each year [1]. Although surgical techniques and immunotherapy have been developed, the prognosis of patients with digestive system cancers remains poor. With increased in-depth research into long non-coding RNA, miRNA, and epigenetics, we have gained a better understanding of cancer at the genetic level [2–4]. Therefore, we expect to identify reliable biomarkers that can be used as prognostic factors from these in digestive system cancers. Recently, miRNAs have also attracted increased attention and are considered to be potential biomarkers for cancer prognosis [5,6].

The iRNAs are short non-coding RNAs (ncRNAs) comprising approximately 18−23 nucleotides that mediate gene silencing by guiding Argonaute (AGO) proteins to target sites in the 3′ untranslated region (UTR) of mRNAs [7]. Since their serendipitous discovery in nematodes, miRNAs have emerged as key regulators of various biological processes in animals [8]. To date, over 30000 miRNAs have been identified. Moreover, miRNAs have been applied to both basal experimental studies, as well as Phase I clinical trials for treating cancer [9]. MicroRNA-135 (MiR-135) is a member of the miRNA family that plays an important role in a variety of cancers, and has two isoforms, miR-135a and miR-135b. Although these isoforms are located on different chromosomes, mature miR-135a and miR-135b only differ in one nucleotide, which is not within the miRNA binding region. Thus, it has been reported that both miRNAs target the same genes [10,11].

Studies have reported that miR-135 is overexpressed in several human cancers and cancer cell lines, including PC [12], breast cancer [13], colorectal cancer (CRC) [14], gastric cancer [15], among others. However, the relationship between miR-135 expression and digestive system cancers remains controversial. Despite this discrepancy, an increasing number of recent studies have emphasized the role of miR-135 in predicting poor tumor prognosis. In addition, based on a bioinformatics analysis, miR-135 was considered worthy of further investigation as a potential novel biomarker for the diagnosis and prognosis of esophageal squamous cell carcinoma (ESCC) and PC [16,17]. In contrast, the prognostic value of miR-135 in gastric cancer was associated with the opposite result [18]. Therefore, the aim of the present study was to evaluate the association of miR-135 expression with the overall survival (OS), disease-free survival (DFS), and recurrence-free survival (RFS) of patients with digestive system cancers.

Methods

Search strategy

In this meta-analysis, the PubMed, Embase, Web of Science, and Cochrane Library databases were systematically searched using: ‘microRNA 135 OR microRNA 135a OR microRNA 135b OR miRNA 135 OR miRNA 135a OR miRNA 135b OR miR-135 OR miR-135a OR miR-135b OR miR-135a-3p OR miR-135a-5p OR miR-135b-3p OR miR-135b-5p’ and ‘tumor or cancer or carcinoma or neoplasm’ for studies published up to 31 August 2019. For the Cochrane Library, we searched the keywords ‘microRNA or miRNA’ and ‘tumor or cancer or carcinoma or neoplasm’ to avoid missing suitable studies. Moreover, the reference of eligible studies and review articles was also manually reviewed for suitable studies. While searching and screening the eligible studies, the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) [19] and Assessing the Methodological Quality of Systematic Reviews (AMSTAR) Guidelines were followed.

Inclusion and exclusion criteria

The inclusion criteria were as follows: (1) the studies researched digestive system cancers; (2) the studies investigated the association between miR-135 and clinical prognosis; (3) the study provided available and usable data for estimating the hazard ratios (HRs) and 95% confidence intervals (95% CIs) for survival. Furthermore, studies were excluded if they conformed to one of the following criteria: (1) insufficient data to estimate the prognostic correlation; (2) basic medical experiment, animal studies, case reports, and review articles; and (3) duplicated data.

Quality assessment

Two investigators (Ce Chao andChen Sang) independently scored the included studies according to the Newcastle–Ottawa Quality Assessment Scale [20]. The final scores were presented from 0 to 10, with higher scores indicating better methodological quality.

Data extraction

Two investigators (Min Wang and Yanfei Li) independently extracted all the data from eligible studies. The following data were extracted: the first author, year of publication, type of cancer, size of sample, country of origin, follow-up time, detection methods, treatment measure, patient survival, and statistical methods. If the survival data were not reported, it was extracted from Kaplan–Meier curves for further calculation. Finally, the third investigator was consulted to review all the data .

Extraction and analysis of The Cancer Genome Atlas datasets

The level of miR-135 expression and clinical data for The Cancer Genome Atlas (TCGA) digestive system cancer (including cholangiocarcinoma, colon adenocarcinoma, esophageal carcinoma, liver hepatocellular carcinoma (HCC), pancreatic adenocarcinoma, rectum adenocarcinoma, and stomach adenocarcinoma) were extracted from the UCSC Xena website (https://xenabrowser.net/hub/) [21]. The level of miRNA expression was measured using the IlluminaHiseq platform. There were 1488 patients with digestive system cancers who had both survival and miR-203 expression data in the TCGA Pan-cancer datasets for this subject [22]. The level of miR-135 expression was distinguished based on receiver operating characteristic (ROC) curves with a Youden’s index correction [23]. The survival curves were plotted and tested using the Kaplan–Meier method and a log-rank test using Stata software version 12.0 (Stata Corporation, College Station, TX).

Statistical analysis

In this meta-analysis, all pooled results were analyzed with Stata software version 12.0 (Stata Corporation, College Station, TX). Pooled HRs with the corresponding 95% CIs were used to estimate the strength of the association between miR-135 and the clinical prognosis of cancer patients. The optimal method was to obtain survival data from the articles or calculate the HRs from O-E statistics and variance. Otherwise, Kaplan–Meier curves were analyzed using the Engauge Digitizer version 4.1 according to the method introduced by Parmar et al. [24]. The level of heterogeneity was assessed by the inconsistency index I2 and χ2-based Q test. I2 statistics with values >50% and a χ2 test with P<0.05 indicated a strong heterogeneity across the studies. If the heterogeneity was not significant, the fixed-effect model was appropriate; otherwise, the random-effect model was adopted. Subgroup analysis was conducted for tumor type, digestive function, biomarker subtype, publication year, sample size, the method of HRs extraction, and quality score. The meta-analysis results were displayed as forest plots. Sensitivity analysis was performed to test the impact of an individual study on the pooled data by omitting each single study. Begg’s funnel plots and an Egger’s linear regression test were performed to estimate any potential publication bias [25]. False-positive report probability (FPRP) analysis and trial sequential analysis (TSA) were performed to assess the significant associations [26,27].

Results

Study characteristics

A total of 1952 potentially relevant studies were retrieved from the electronic databases using the search strategy. After removing any duplications and carefully screening the abstracts and full-text of these articles, a total of 17 articles [14,15,18,28–41] were finally included in the current quantitative analysis of the prognostic value of miR-135 for digestive system cancers. The selection process of the literature search is presented in Figure 1. The general characteristics of these studies are summarized in Table 1. The 17 included studies were reports from around the world, including 14 from China [14,15,18,28,30,31,33,34,36–41], 1 from the U.K. [29], and 2 from Germany [32], all of which were published between 2012 and 2019. Among the included studies, five focused on CRC [14,29,30,35,37], four were on HCC [28,32,39,41], three were about gastric cancer [15,18,31], one was on ESCC [34], three were regarding PC [33,38,40], and one investigated gastric cardia adenocarcinoma (GCAC) [36]. There were 14 studies [14,15,18,28–31,33,34,36–38,40,41] involving 1225 patients that reported the OS, 7 [14,15,28,33,35,39,40] that reported DFS, and 2 [29,32] reported the RFS.

Flow chart for selection of studies for inclusion in this meta-analysis

Figure 1
Flow chart for selection of studies for inclusion in this meta-analysis
Figure 1
Flow chart for selection of studies for inclusion in this meta-analysis
Table 1
Characteristics of the studies included in the meta-analysis
AuthorTimeCountrySample sizeCancer typeQualityFollow-upMethodsTreatmentSurvival dataSubtypeCut-offSource of HR
Gaedcke 2012 Germany 116 RC 120 qPCR Multiple DFS, CSS Median Reported 
Liu 2012 China 50 HCC 25 qPCR Surgery DFS, OS Median Reported 
Valeri 2014 U.K. 62 CRC NR qPCR Surgery RFS, OS Median Curve 
Gao 2016 China 58 GCAC NR qPCR Surgery OS Median Curve 
Kan 2016 China 105 CRC 60 qPCR Surgery OS Median Reported 
Li 2016 China 47 CRC NR qPCR Multiple OS Mean Curve 
Yan 2016 China 280 GC 60 qPCR Surgery DFS, OS Median Reported 
Cheng 2017 China 176 GC 60 qPCR Surgery OS Mean Curve 
Han 2017 China 36 PC NR qPCR Surgery OS Median Reported 
Huang 2017 China 103 HCC NR qPCR Surgery DFS Normal Curve 
Felden 2017 Germany 26 HCC NR qPCR Multiple RFS Median Reported 
Zhang 2017 China 85 PC 48 qPCR Surgery DFS, OS Median Curve 
Jiang 2018 China 37 PC NR qPCR Chemotherapy OS, DFS Mean Curve 
Qiao 2018 China 94 CRC 80 qPCR Surgery DFS, OS Median Curve 
Zhang 2018 China 105 ESCC NR qPCR Surgery OS Mean Reported 
Xie 2019 China 40 GC 102 qPCR NR OS Median Curve 
Yang 2019 China 50 HCC NR qPCR Surgery OS Normal Curve 
AuthorTimeCountrySample sizeCancer typeQualityFollow-upMethodsTreatmentSurvival dataSubtypeCut-offSource of HR
Gaedcke 2012 Germany 116 RC 120 qPCR Multiple DFS, CSS Median Reported 
Liu 2012 China 50 HCC 25 qPCR Surgery DFS, OS Median Reported 
Valeri 2014 U.K. 62 CRC NR qPCR Surgery RFS, OS Median Curve 
Gao 2016 China 58 GCAC NR qPCR Surgery OS Median Curve 
Kan 2016 China 105 CRC 60 qPCR Surgery OS Median Reported 
Li 2016 China 47 CRC NR qPCR Multiple OS Mean Curve 
Yan 2016 China 280 GC 60 qPCR Surgery DFS, OS Median Reported 
Cheng 2017 China 176 GC 60 qPCR Surgery OS Mean Curve 
Han 2017 China 36 PC NR qPCR Surgery OS Median Reported 
Huang 2017 China 103 HCC NR qPCR Surgery DFS Normal Curve 
Felden 2017 Germany 26 HCC NR qPCR Multiple RFS Median Reported 
Zhang 2017 China 85 PC 48 qPCR Surgery DFS, OS Median Curve 
Jiang 2018 China 37 PC NR qPCR Chemotherapy OS, DFS Mean Curve 
Qiao 2018 China 94 CRC 80 qPCR Surgery DFS, OS Median Curve 
Zhang 2018 China 105 ESCC NR qPCR Surgery OS Mean Reported 
Xie 2019 China 40 GC 102 qPCR NR OS Median Curve 
Yang 2019 China 50 HCC NR qPCR Surgery OS Normal Curve 

Abbreviations: CCS, cancer-specific survival; GC, gastric cancer; NR, not recorded; qPCR, real-time quantitative PCR detection system.

Prognostic value of miR-135 for the OS of digestive system cancers

In this meta-analysis, a total of 14 studies, including a total of 1225 patients, reported HRs and associated 95% CIs for OS. As shown in Figure 2, the pooled results indicate that elevated miR-135 expression was associated with a poor OS in patients with digestive system cancer (HR: 1.790; 95% CI: 1.577–2.031; P=0.000). Moreover, a heterogeneity analysis revealed that there was no between-study heterogeneity in the eight studies (χ2 = 12.31; d.f. = 13; P=0.503; I2 = 0.0%).

Meta-analysis of the pooled HRs of OS of patients with miR-135 overexpression

Figure 2
Meta-analysis of the pooled HRs of OS of patients with miR-135 overexpression
Figure 2
Meta-analysis of the pooled HRs of OS of patients with miR-135 overexpression

Although no inter-study heterogeneity was observed in this meta-analysis, a subsequent analysis was performed based on the cancer type, digestive function, biomarker subtype, publication year, sample size, the method of HR extraction, and quality score. As shown in Table 2, a significant correlation between miR-135 overexpression and a poor OS was shown in patients with HCC (HR: 3.536; 95% CI: 1.604−7.796; P=0.002), CRC (HR: 1.814; 95% CI: 1.257−2.619; P=0.001), PC (HR: 1.885; 95% CI: 1.521−2.336), and ESCC (HR: 1.820; 95% CI: 1.446–2.291). There was a significant association between miR-135 expression and the patient OS in the digestive gland (HR: 1.931; 95% CI: 1.591−2.341; P=0.000) and digestive tract (HR: 1.689; 95% CI: 1.428−1.997; P=0.000). Both miR-135a (HR: 1.624; 95% CI: 1.290−2.044; P=0.000) and miR-135b (HR: 1.902; 95% CI: 1.557−2.323; P=0.000) were significantly associated with poor OS in digestive system cancer patients. In addition, miR-135 was found to be significantly associated with the OS of patients reported in studies published from 2012 to 2016 (HR: 1.800; 95% CI: 1.403−2.309; P=0.000) and 2017 to 2018 (HR: 1.786; 95% CI: 1.543−2.068; P=0.000). The association between miR-135 and the OS of patients was significant in studies with a sample size less than 100 (HR: 1.879; 95% CI: 1.571−2.247; P=0.000) and more than 100 (HR: 1.706; 95% CI: 1.427−2.038; P=0.000). We also detected a significant relationship between the level of miR-135 expression and the OS of patients in both studies with a quality score less than 6 (HR: 1.717; 95% CI: 1.403−2.100; P=0.000) and those with a quality score greater than 6 (HR: 1.839; 95% CI: 1.563−2.162; P=0.000). Finally, a significant relationship was found between miR-135 and the OS of patients in both studies with HRs reported in the text (HR: 1.846; 95% CI: 1.548−2.202; P=0.000) and studies with HRs extracted from survival curves (HR: 1.732; 95% CI: 1.444–2.076; P=0.000).

Table 2
Results of subgroup analysis of pooled HRs of OS of patients with increased miR-135 expression
Stratified analysisNumber of studiesNumber of patientsHeterogeneityPooled HR (95% CI)Meta-regression P-valueStatistical powerFPRP P-valueFPRP
I2 (%)P-value
All 14 1225 0.503 1.790 (1.577–2.031)  
Tumor type         
HCC 100 0.939 3.536 (1.604–7.796)  0.017 0.002 0.911 
CRC 308 0.817 1.814 (1.257–2.619)  0.155 0.001 0.486 
GC 496 61.5 0.075 1.031 (0.477–1.917)  0.882 0.923 0.990 
PC 158 0.657 1.885 (1.521–2.336)  0.018 0.000 0.000 
ESCC 105 1.820 (1.446–2.291)  0.050 0.000 0.001 
GCAC 58 1.790 (0.718–4.465)  0.352 0.212 0.983 
Digestive function      0.257    
Digestive grand 258 0.539 1.968 (1.600–2.421)  0.005 0.000 0.000 
Digestive tract 967 0.442 1.692 (1.443–1.984)  0.069 0.000 0.000 
Biomarker subtype      0.588    
miR-135a 687 43.1 0.118 1.624 (1.290–2.044)  0.249 0.000 0.014 
miR-135b 433 0.874 1.902 (1.557–2.323)  0.010 0.000 0.000 
Publication year      0.957    
2012–2016 602 0.666 1.800 (1.403–2.309)  0.076 0.000 0.005 
2017–2019 623 22.9 0.247 1.786 (1.543–2.068)  0.010 0.000 0.000 
Sample size      0.453    
<100 10 559 0.600 1.879 (1.571–2.247)  0.007 0.000 0.000 
≥100 666 31.6 0.223 1.706 (1.427–2.038)  0.180 0.000 0.074 
Quality score      0.603    
≤6 343 28.9 0.207 1.717 (1.403–2.100)  0.094 0.000 0.000 
>6 556 0.732 1.839 (1.563–2.162)  0.007 0.000 0.000 
HRs extraction      0.619    
Reported 675 0.478 1.846 (1.548–2.202)  0.011 0.000 0.000 
Curve 649 6.5 0.381 1.732 (1.444–2.076)  0.060 0.000 0.000 
DFS 765 86.9 0.000 1.482 (0.914–2.403)  0.520 0.111 0.995 
RFS 88 0.879 3.994 (1.363–11.697)  0.037 0.012 0.969 
Stratified analysisNumber of studiesNumber of patientsHeterogeneityPooled HR (95% CI)Meta-regression P-valueStatistical powerFPRP P-valueFPRP
I2 (%)P-value
All 14 1225 0.503 1.790 (1.577–2.031)  
Tumor type         
HCC 100 0.939 3.536 (1.604–7.796)  0.017 0.002 0.911 
CRC 308 0.817 1.814 (1.257–2.619)  0.155 0.001 0.486 
GC 496 61.5 0.075 1.031 (0.477–1.917)  0.882 0.923 0.990 
PC 158 0.657 1.885 (1.521–2.336)  0.018 0.000 0.000 
ESCC 105 1.820 (1.446–2.291)  0.050 0.000 0.001 
GCAC 58 1.790 (0.718–4.465)  0.352 0.212 0.983 
Digestive function      0.257    
Digestive grand 258 0.539 1.968 (1.600–2.421)  0.005 0.000 0.000 
Digestive tract 967 0.442 1.692 (1.443–1.984)  0.069 0.000 0.000 
Biomarker subtype      0.588    
miR-135a 687 43.1 0.118 1.624 (1.290–2.044)  0.249 0.000 0.014 
miR-135b 433 0.874 1.902 (1.557–2.323)  0.010 0.000 0.000 
Publication year      0.957    
2012–2016 602 0.666 1.800 (1.403–2.309)  0.076 0.000 0.005 
2017–2019 623 22.9 0.247 1.786 (1.543–2.068)  0.010 0.000 0.000 
Sample size      0.453    
<100 10 559 0.600 1.879 (1.571–2.247)  0.007 0.000 0.000 
≥100 666 31.6 0.223 1.706 (1.427–2.038)  0.180 0.000 0.074 
Quality score      0.603    
≤6 343 28.9 0.207 1.717 (1.403–2.100)  0.094 0.000 0.000 
>6 556 0.732 1.839 (1.563–2.162)  0.007 0.000 0.000 
HRs extraction      0.619    
Reported 675 0.478 1.846 (1.548–2.202)  0.011 0.000 0.000 
Curve 649 6.5 0.381 1.732 (1.444–2.076)  0.060 0.000 0.000 
DFS 765 86.9 0.000 1.482 (0.914–2.403)  0.520 0.111 0.995 
RFS 88 0.879 3.994 (1.363–11.697)  0.037 0.012 0.969 

Abbreviation: GC, gastric cancer.

A sensitivity analysis was performed to determine the reliability of the pooled results. The results proved to be solid by sequentially omitting each study (Figure 3). A Begg’s funnel plot and Egger’s test were conducted to assess the publication bias for the pooled OS. No publication bias was observed for the OS (Begg’s test: P=0.622; Egger’s test: P=0.932) (Figure 4).

Sensitivity analysis of the effect of the individual study on the pooled HRs for the correlation between miR-135 expression and OS in patients with digestive system cancers

Figure 3
Sensitivity analysis of the effect of the individual study on the pooled HRs for the correlation between miR-135 expression and OS in patients with digestive system cancers
Figure 3
Sensitivity analysis of the effect of the individual study on the pooled HRs for the correlation between miR-135 expression and OS in patients with digestive system cancers

Begg’s funnel plot analysis for publication bias

Figure 4
Begg’s funnel plot analysis for publication bias
Figure 4
Begg’s funnel plot analysis for publication bias

Prognostic value of miR-135 for the DFS and RFS of digestive system cancers

Seven studies that included a total of 765 patients reported the level of miR-135 expression and DFS of digestive system cancers. Two studies consisting of 88 patients reported HRs with the 95% CI for RFS. As shown in Figure 5, miR-135 overexpression was associated with a poor RFS (HR: 3.994; 95% CI: 1.363–11.697; P=0.012), but DFS (HR: 1.482; 95% CI: 0.914–2.403; P=0.110) in patients with digestive system cancers.

Meta-analysis of the pooled HRs of DFS and RFS of patients with miR-135 overexpression

Figure 5
Meta-analysis of the pooled HRs of DFS and RFS of patients with miR-135 overexpression
Figure 5
Meta-analysis of the pooled HRs of DFS and RFS of patients with miR-135 overexpression

FPRP analysis

We performed an FPRP assay for OS, DFS, and RFS to uncover potential false-positive results among significant associations, with a prior probability of 0.01 and an FPRP cut-off value of 0.2. Because the P-value of FPRP for OS was equal to 0, the corresponding FPRP value was infinitely close to 0 [26]. Thus, there was a significant association between OS and digestive system cancer. Moreover, there was a significant association between OS and digestive system cancer observed in the subgroups of cancer type (PC and ESCC), cancer classification (digestive gland and tract), source of control (curve and reported), sample size (large and small), biomarker subtype (miR-135a and miR-135b), publication year (2012−2016 and 2017−2019), and quality score (more than 6 and less than 6). The FPRP values for DFS and RFS were all >0.2, revealing that these associations were not truly significant (Table 2).

TSA

The cumulative Z-curve (blue line) was found to cross both the traditional boundary line and the trial sequential monitoring boundary (red line), and the cumulative information reached the required information size (RIS) for OS and RFS (Figure 6A,C). Thus, additional studies for OS were not required. Additional studies for RFS were necessary due to the insufficient number of samples; however, the cumulative Z-curve (blue line) crosses the traditional boundary line rather than crossing the trial sequential monitoring boundary (red line), while the cumulative information reaches the RIS for DFS (Figure 6B).

Trial sequential analysis (TSA) for the primary outcomes

Figure 6
Trial sequential analysis (TSA) for the primary outcomes

TSA for OS (A), DFS (B), and RFS (C).

Figure 6
Trial sequential analysis (TSA) for the primary outcomes

TSA for OS (A), DFS (B), and RFS (C).

Validation by independent TCGA digestive system cancer datasets

To verify the meta-analysis results, we performed an analysis of TCGA database regarding the level of miR-135 expression and corresponding survival data. This analysis included seven types of cancer: cholangiocarcinoma, colon adenocarcinoma, esophageal carcinoma, pancreatic adenocarcinoma, liver HCC, rectum adenocarcinoma, and stomach adenocarcinoma. Moreover, the results showed that high miR-135 expression in the tissues was associated with a poorer OS for digestive system cancers (HR: 1.217; 95% CI: 1.083–1.368; P=0.002). And, the relationship between miR-135 subtype and prognosis of digestive system cancers is shown in Figure 7. In the analysis of single tumor types, we found that high miR-135 expression was associated with a poor OS for colon adenocarcinoma (HR: 1.526; 95% CI: 1.051–2.214; P=0.026), pancreatic adenocarcinoma (HR: 1.403; 95% CI: 1.042–1.888; P=0.026), liver HCC (HR: 1.654; 95% CI: 1.221–2.241; P=0.001), and stomach adenocarcinoma (HR: 1.303; 95% CI: 1.010–1.681; P=0.042) (Table 3). In the other tumor types, miR-135 expression was not associated with the prognosis of patients with cholangiocarcinoma, esophageal carcinoma, and rectum adenocarcinoma (Table 3). Therefore, TCGA results were consistent with the findings of our meta-analysis.

The relationship between miR-135 subtypes and prognosis of digestive system cancers using TCGA data

Figure 7
The relationship between miR-135 subtypes and prognosis of digestive system cancers using TCGA data
Figure 7
The relationship between miR-135 subtypes and prognosis of digestive system cancers using TCGA data
Table 3
HRs and corresponding 95% CIs of miR-135 overexpression in digestive system cancers based on TCGA datasets
Cancer typeSample sizesmiR-135 assayHR (95% CI)P-value
Digestive system cancer 1488 IlluminaHiseq 1.217 (1.083–1.368) 0.002 
Cholangiocarcinoma 36 IlluminaHiseq 1.432 (0.808–2.539) 0.218 
Colon adenocarcinoma 249 IlluminaHiseq 1.526 (1.051–2.214) 0.026 
Esophageal carcinoma 182 IlluminaHiseq 1.344 (0.967–1.869) 0.078 
Pancreatic adenocarcinoma 369 IlluminaHiseq 1.403 (1.042–1.888) 0.026 
Liver HCC 177 IlluminaHiseq 1.654 (1.221–2.241) 0.001 
Rectum adenocarcinoma 88 IlluminaHiseq 2.284 (0.893–5.482) 0.085 
Stomach adenocarcinoma 387 IlluminaHiseq 1.303 (1.010–1.681) 0.042 
Cancer typeSample sizesmiR-135 assayHR (95% CI)P-value
Digestive system cancer 1488 IlluminaHiseq 1.217 (1.083–1.368) 0.002 
Cholangiocarcinoma 36 IlluminaHiseq 1.432 (0.808–2.539) 0.218 
Colon adenocarcinoma 249 IlluminaHiseq 1.526 (1.051–2.214) 0.026 
Esophageal carcinoma 182 IlluminaHiseq 1.344 (0.967–1.869) 0.078 
Pancreatic adenocarcinoma 369 IlluminaHiseq 1.403 (1.042–1.888) 0.026 
Liver HCC 177 IlluminaHiseq 1.654 (1.221–2.241) 0.001 
Rectum adenocarcinoma 88 IlluminaHiseq 2.284 (0.893–5.482) 0.085 
Stomach adenocarcinoma 387 IlluminaHiseq 1.303 (1.010–1.681) 0.042 

Discussion

This meta-analysis represents the first systematic evaluation of the relationship between the level of miR-135 expression and the prognosis of patients with digestive system cancers. A total of 17 studies consisting of 1470 patients were included in this meta-analysis. Our results revealed that a greater level of miR-135 expression was associated with a poorer OS and RFS in patients with digestive system cancers. Stratified analyses were performed, and significant results pertaining to the relationship between miR-135 and a poor OS were some of tumor type, digestive function, biomarker subtype, publication year, sample size, the method of HR extraction, and quality score. There was no significant heterogeneity and publication bias in this pooled result, which suggested that our results were reliable. In addition, the TSA indicated that the sample size was sufficient and the OS results were reliable due to the incalculable probability of false-positive reports in the FPRP analysis. In addition, the results were verified using TCGA database analysis. These findings suggested that miR-135 overexpression is indicative of a poor prognosis and higher recurrence in patients with digestive system cancers. However, the results required further validation for DFS and RFS due to the high probability of false-positive reports in the FPRP analysis.

Stratified analyses showed that article factors, including publication time, sample size, studies quality, and methods of HRs, had no effect on the association between miR-135 expression and the OS of patients with digestive system cancers. Although mature miR-135a and miR-135b only differ regarding one nucleotide, this did not impact their prognostic effect because both of the miRNAs target the same genes. In the subgroup analysis, there was significant heterogeneity associated with the gastric cancer group due to the two contrary results of the incorporated studies. While Cheng et al. [31] found that decreased miR-135 expression was indicative of a poor OS, a study by Yan et al. [15] reported the opposite findings. However, TCGA database analysis revealed that miR-135 expression had prognostic significance for gastric cancer. Thus, future large-scale studies should be performed to verify this speculation and consolidate the pooled results. In addition, we used the data from TCGA datasets to validate the meta-analysis results. Similar results were found for pancreatic, colon, and hepatocellular cancers.

An increasing number of studies have shown that the deregulation of miRNA function is associated with an increasing number of human diseases, particularly cancer. Moreover, miRNA occupies an important position in regulating cancer proliferation, growth, invasion, and migration. Conventional non-invasive serological biomarkers, including carcinoembryonic antigen (CEA) and cancer antigen 19-9 (CA19-9), can be screened using blood tests that are currently being applied in the clinic. However, these biomarkers lack specificity for the early diagnosis of digestive system cancer [42,43]. Although some studies have found that CEA and CA19-9 could predict the prognosis of patients with digestive system cancer, including CRC and gastric cancer [44,45], the positive rate is too low. In contrast, miRNAs in digestive cancer patients may represent promising biomarkers that are stable and reproducible in both the peripheral blood and tissues. In the peripheral blood, circulating miRNAs primarily bind to proteins or are capsulated in exosomes [46]. Although measuring plasma or serum miRNA is non-invasive, the level of circulating miRNAs is susceptible to perturbations due to blood cells or hemolysis compared with tissue miRNA [47]. In addition, it is not difficult to measure tissue miRNA in patients after surgery. Moreover, the detection of tissue miRNA was more accurate, with fewer influencing factors. It was convincing that changes in miRNA expression and patient survival was consistent in both the peripheral blood and tissues. However, the results may differ on occasion; for example, when Zhang et al. [48] elucidated the prognostic value of miR-20a in human cancers, high expression of both circulating and tumorous miR-20a were associated with an unfavorable outcome only in gastrointestinal cancers. In this meta-analysis, all the included studies were related to tissue miRNAs with reliable results.

Kawaguchi et al. [49] proposed the use of miRNA as a next-generation biomarker for digestive system cancers. In addition, the authors thought that several miRNAs should be used in combination to evaluate the prognosis of one or more cancers. Recently, a database analysis demonstrated that a panel of miRNAs, including miR-135, was associated with a poor prognosis for both pancreatic and esophageal cancers [16,17,50]. Therefore, although meta-analyses have shown that several miRNAs, including miR-17-5p [51] and miR-133a [52], could predict the prognosis of digestive system tumors, a meta-analysis was required to assess the relationship between miR-135 and digestive system cancers. Moreover, miR-135 has been found to be involved in mediating the metabolic stress response in cancer. For example, miR-135 can modulate critical signaling pathways (e.g., JAK/STAT [14], P38MAPK/NF-κB [53], PTEN/PI3k/AKT [54], and RERG [34]) that indirectly regulate metabolism. Altered metabolism has long been recognized as a central hallmark of cancer. Yang et al. [12] established a key role for miR-135 in regulating glycolysis in vivo and found that miR-135 suppressed glycolysis and promoted PC cell adaptation to metabolic stress by targeting phosphofructokinase-1. In addition, a study by Wang et al. [14] found that miR-135 increased the levels of phosphorylated key kinases in the JAK/STAT pathway and functioned as a tumor promoter by targeting TRIM16. Furthermore, the deregulation of miR-135 function has been associated with both breast cancer [13] and Hodgkin lymphoma [55]. These findings provide evidence for the function of miR-135 as a potential prognostic factor and treatment for tumors.

There were some limitations associated with the present research. First, due to insufficient data, only 17 studies were recruited, comprising a total of 1470 patients in this meta-analysis. Moreover, some studies that yielded negative results were generally not published. Due to these factors, some subgroups only included one study in the subgroup analysis. Therefore, this result may be an outlier because the data were insufficiently comprehensive. Second, most of the included patients were Asian, which might reduce the universal applicability of the results. Thus, additional large-scale studies should be performed among different ethnicities. Third, the cut-off value of miR-135 was variable in different studies, which did not reach a consensus value in this meta-analysis. Fourth, HRs with associated 95% CIs and P-value extracted from the curve differed from the actual value because some of the studies did not provide HRs with 95% CIs.

In conclusion, the results of this meta-analysis indicate that miR-135 expression might be a novel prognostic factor for the survival of patients with digestive system malignancies. Further studies with a large-scale sample size should be performed to identify the association between miR-135 expression and digestive system cancer prognosis.

Acknowledgments

We thank International Science Editing (http://www.internationalscienceediting.com) for editing this manuscript.

Competing Interests

The authors declare that there are no competing interests associated with the manuscript.

Funding

This work was supported by The National Natural Science Fund of China [grant number 81701584].

Author Contribution

Ce Chao, Chen Sang, Min Wang, Zijin Wang, and Yanfei Li were responsible for collecting, analyzing, and interpreting the data. Ce Chao was responsible for writing the manuscript. Guanghua Luo provided critical revisions. Xiaoying Zhang approved the final version of the manuscript.

Abbreviations

     
  • CA19-9

    cancer antigen 19-9

  •  
  • CEA

    carcinoembryonic antigen

  •  
  • CRC

    colorectal cancer

  •  
  • DFS

    disease-free survival

  •  
  • ESCC

    esophageal squamous cell carcinoma

  •  
  • FPRP

    false-positive report probability

  •  
  • HCC

    hepatocellular carcinoma

  •  
  • HR

    hazard ratio

  •  
  • miR-135

    microRNA-135

  •  
  • OS

    overall survival

  •  
  • PC

    pancreatic cancer

  •  
  • RFS

    recurrence-free survival

  •  
  • RIS

    required information size

  •  
  • TCGA

    The Cancer Genome Atlas

  •  
  • TSA

    trial sequential analysis

  •  
  • 95% CI

    95% confidence interval

References

References
1.
Bray
F.
,
Ferlay
J.
,
Soerjomataram
I.
,
Siegel
R.L.
,
Torre
L.A.
and
Jemal
A.
(
2018
)
Global cancer statistics 2018: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries
.
CA Cancer J. Clin.
68
,
394
424
[PubMed]
2.
Sanchez Calle
A.
,
Kawamura
Y.
,
Yamamoto
Y.
,
Takeshita
F.
and
Ochiya
T.
(
2018
)
Emerging roles of long non-coding RNA in cancer
.
Cancer Sci.
109
,
2093
2100
[PubMed]
3.
Lin
S.
and
Gregory
R.I.
(
2015
)
MicroRNA biogenesis pathways in cancer
.
Nat. Rev. Cancer
15
,
321
333
[PubMed]
4.
Kinnaird
A.
,
Zhao
S.
,
Wellen
K.E.
and
Michelakis
E.D.
(
2016
)
Metabolic control of epigenetics in cancer
.
Nat. Rev. Cancer
16
,
694
707
[PubMed]
5.
Nassar
F.J.
,
Nasr
R.
and
Talhouk
R.
(
2017
)
MicroRNAs as biomarkers for early breast cancer diagnosis, prognosis and therapy prediction
.
Pharmacol. Ther.
172
,
34
49
[PubMed]
6.
Yao
L.
,
Liu
Y.
,
Cao
Z.
,
Li
J.
,
Huang
Y.
,
Hu
X.
et al.
(
2018
)
MicroRNA-493 is a prognostic factor in triple-negative breast cancer
.
Cancer Sci.
109
,
2294
2301
[PubMed]
7.
Li
J.
,
Tan
S.
,
Kooger
R.
,
Zhang
C.
and
Zhang
Y.
(
2014
)
MicroRNAs as novel biological targets for detection and regulation
.
Chem. Soc. Rev.
43
,
506
517
[PubMed]
8.
Lee
R.C.
,
Feinbaum
R.L.
and
Ambros
V.
(
1993
)
The C. elegans heterochronic gene lin-4 encodes small RNAs with antisense complementarity to lin-14
.
Cell
75
,
843
854
[PubMed]
9.
Rupaimoole
R.
and
Slack
F.J.
(
2017
)
MicroRNA therapeutics: towards a new era for the management of cancer and other diseases
.
Nat. Rev. Drug Discov.
16
,
203
222
10.
Issler
O.
,
Haramati
S.
,
Paul
E.D.
,
Maeno
H.
,
Navon
I.
,
Zwang
R.
et al.
(
2014
)
MicroRNA 135 is essential for chronic stress resiliency, antidepressant efficacy, and intact serotonergic activity
.
Neuron
83
,
344
360
[PubMed]
11.
Petracco
R.
,
Grechukhina
O.
,
Popkhadze
S.
,
Massasa
E.
,
Zhou
Y.
and
Taylor
H.S.
(
2011
)
MicroRNA 135 regulates HOXA10 expression in endometriosis
.
J. Clin. Endocrinol. Metab.
96
,
E1925
E1933
[PubMed]
12.
Yang
Y.
,
Ishak Gabra
M.B.
,
Hanse
E.A.
,
Lowman
X.H.
,
Tran
T.Q.
,
Li
H.
et al.
(
2019
)
MiR-135 suppresses glycolysis and promotes pancreatic cancer cell adaptation to metabolic stress by targeting phosphofructokinase-1
.
Nat. Commun.
10
,
809
[PubMed]
13.
Jiang
D.
,
Zhou
B.
,
Xiong
Y.
and
Cai
H.
(
2019
)
miR-135 regulated breast cancer proliferation and epithelial-mesenchymal transition acts by the Wnt/beta-catenin signaling pathway
.
Int. J. Mol. Med.
43
,
1623
1634
14.
Qiao
P.
,
Li
S.
,
Zhang
H.
,
Yao
L.
and
Wang
F.
(
2018
)
Farnesoid X receptor inhibits proliferation of human colorectal cancer cells via the miR135A1/CCNG2 signaling pathway
.
Oncol. Rep.
40
,
2067
2078
[PubMed]
15.
Yan
L.H.
,
Chen
Z.N.
,
Li
L.
,
Chen
J.
,
Wei
W.E.
,
Mo
X.W.
et al.
(
2016
)
miR-135a promotes gastric cancer progression and resistance to oxaliplatin
.
Oncotarget
7
,
70699
70714
[PubMed]
16.
Li
C.Y.
,
Zhang
W.W.
,
Xiang
J.L.
,
Wang
X.H.
,
Li
J.
and
Wang
J.L.
(
2019
)
Identification of microRNAs as novel biomarkers for esophageal squamous cell carcinoma: a study based on The Cancer Genome Atlas (TCGA) and bioinformatics
.
Chin. Med. J.
132
,
2213
2222
[PubMed]
17.
Zhou
J.
,
Hui
X.
,
Mao
Y.
and
Fan
L.
(
2019
)
Identification of novel genes associated with a poor prognosis in pancreatic ductal adenocarcinoma via a bioinformatics analysis
.
Biosci. Rep.
39
,
18.
Xie
Y.
,
Li
F.
,
Li
Z.
and
Shi
Z.
(
2019
)
miR-135a suppresses migration of gastric cancer cells by targeting TRAF5-mediated NF-κB activation
.
Onco Targets Ther.
12
,
975
984
19.
Moher
D.
,
Liberati
A.
,
Tetzlaff
J.
and
Altman
D.G.
(
2009
)
Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement
.
PLoS Med.
6
,
e1000097
[PubMed]
20.
Stang
A.
(
2010
)
Critical evaluation of the Newcastle-Ottawa scale for the assessment of the quality of nonrandomized studies in meta-analyses
.
Eur. J. Epidemiol.
25
,
603
605
[PubMed]
21.
Goldman
M.
,
Craft
B.
,
Swatloski
T.
,
Cline
M.
,
Morozova
O.
,
Diekhans
M.
et al.
(
2015
)
The UCSC Cancer Genomics Browser: update 2015
.
Nucleic Acids Res.
43
,
D812
D817
[PubMed]
22.
Cancer Genome Atlas Research Network
Weinstein
J.N.
,
Collisson
E.A.
,
Mills
G.B.
,
Shaw
K.R.
,
Ozenberger
B.A.
et al.
(
2013
)
The Cancer Genome Atlas Pan-Cancer Analysis Project
.
Nat. Genet.
45
,
1113
1120
[PubMed]
23.
Ruopp
M.D.
,
Perkins
N.J.
,
Whitcomb
B.W.
and
Schisterman
E.F.
(
2008
)
Youden Index and optimal cut-point estimated from observations affected by a lower limit of detection
.
Biom. J.
50
,
419
430
[PubMed]
24.
Parmar
M.K.
,
Torri
V.
and
Stewart
L.
(
1998
)
Extracting summary statistics to perform meta-analyses of the published literature for survival endpoints
.
Stat. Med.
17
,
2815
2834
[PubMed]
25.
Egger
M.
,
Davey Smith
G.
,
Schneider
M.
and
Minder
C.
(
1997
)
Bias in meta-analysis detected by a simple, graphical test
.
BMJ
315
,
629
634
[PubMed]
26.
Wacholder
S.
,
Chanock
S.
,
Garcia-Closas
M.
,
El Ghormli
L.
and
Rothman
N.
(
2004
)
Assessing the probability that a positive report is false: an approach for molecular epidemiology studies
.
J. Natl. Cancer Inst.
96
,
434
442
[PubMed]
27.
Wetterslev
J.
,
Thorlund
K.
,
Brok
J.
and
Gluud
C.
(
2008
)
Trial sequential analysis may establish when firm evidence is reached in cumulative meta-analysis
.
J. Clin. Epidemiol.
61
,
64
75
[PubMed]
28.
Liu
S.
,
Guo
W.
,
Shi
J.
,
Li
N.
,
Yu
X.
,
Xue
J.
et al.
(
2012
)
MicroRNA-135a contributes to the development of portal vein tumor thrombus by promoting metastasis in hepatocellular carcinoma
.
J. Hepatol.
56
,
389
396
[PubMed]
29.
Valeri
N.
,
Braconi
C.
,
Gasparini
P.
,
Murgia
C.
,
Lampis
A.
,
Paulus-Hock
V.
et al.
(
2014
)
MicroRNA-135b promotes cancer progression by acting as a downstream effector of oncogenic pathways in colon cancer
.
Cancer Cell
25
,
469
483
[PubMed]
30.
Li
G.D.
,
Li
X.L.
,
Dang
S.W.
,
Li
Z.S.
,
Lin
L.Q.
and
Bai
M.H.
(
2016
)
Investigation of the clinical significance and prognosis of increased microRNA-135a expression in colorectal cancer tissues
.
Chin. J. Cancer Prev. Treat.
23
,
1169
1173
31.
Cheng
Z.
,
Liu
F.
,
Zhang
H.
,
Li
X.
,
Li
Y.
,
Li
J.
et al.
(
2017
)
miR-135a inhibits tumor metastasis and angiogenesis by targeting FAK pathway
.
Oncotarget
8
,
31153
31168
[PubMed]
32.
von Felden
J.
,
Heim
D.
,
Schulze
K.
,
Krech
T.
,
Ewald
F.
,
Nashan
B.
et al.
(
2017
)
High expression of micro RNA-135A in hepatocellular carcinoma is associated with recurrence within 12 months after resection
.
BMC Cancer
17
,
60
[PubMed]
33.
Zhang
Z.
,
Che
X.
,
Yang
N.
,
Bai
Z.
,
Wu
Y.
,
Zhao
L.
et al.
(
2017
)
miR-135b-5p Promotes migration, invasion and EMT of pancreatic cancer cells by targeting NR3C2
.
Biomed. Pharmacother.
96
,
1341
1348
[PubMed]
34.
Zhang
Y.
,
Ren
S.
,
Yuan
F.
,
Zhang
K.
,
Fan
Y.
,
Zheng
S.
et al.
(
2018
)
miR-135 promotes proliferation and stemness of oesophageal squamous cell carcinoma by targeting RERG
.
46
,
1
10
[PubMed]
35.
Gaedcke
J.
,
Grade
M.
,
Camps
J.
,
Sokilde
R.
,
Kaczkowski
B.
,
Schetter
A.J.
et al.
(
2012
)
The rectal cancer microRNAome–microRNA expression in rectal cancer and matched normal mucosa
.
Clin. Cancer Res.
18
,
4919
4930
[PubMed]
36.
Gao
S.
,
Zhou
F.
,
Zhao
C.
,
Ma
Z.
,
Jia
R.
,
Liang
S.
et al.
(
2016
)
Gastric cardia adenocarcinoma microRNA profiling in Chinese patients
.
Tumour Biol.
37
,
9411
9422
[PubMed]
37.
Kan
S.-F.
,
Yang
J.-S.
,
Sun
G.-X.
and
Sun
J.-J.
(
2016
)
MicroRNA-135b is associated with tumor progression in colorectal cancer
.
Int. J. Clin. Exp. Med.
9
,
6533
6538
38.
Han
X.
,
Saiyin
H.
,
Zhao
J.
,
Fang
Y.
,
Rong
Y.
,
Shi
C.
et al.
(
2017
)
Overexpression of miR-135b-5p promotes unfavorable clinical characteristics and poor prognosis via the repression of SFRP4 in pancreatic cancer
.
Oncotarget
8
,
62195
62207
[PubMed]
39.
Huang
K.-T.
,
Kuo
I.Y.
,
Tsai
M.-C.
,
Wu
C.-H.
,
Hsu
L.-W.
,
Chen
L.-Y.
et al.
(
2017
)
Factor VII-induced microRNA-135a inhibits autophagy and is associated with poor prognosis in hepatocellular carcinoma
.
Mol. Ther. Nucleic Acids
9
,
274
283
40.
Jiang
W.
,
Zhao
S.
,
Shen
J.
,
Guo
L.
,
Sun
Y.
,
Zhu
Y.
et al.
(
2018
)
The MIR-135b-BMAL1-YY1 loop disturbs pancreatic clockwork to promote tumourigenesis and chemoresistance article
.
Cell Death Dis.
9
,
149
41.
Yang
L.
and
Jiang
J.
(
2019
)
GAS5 regulates RECK expression and inhibits invasion potential of HCC cells by sponging miR-135b
.
Biomed Res. Int.
2019
,
2973289
[PubMed]
42.
Scarpa
M.
,
Noaro
G.
,
Saadeh
L.
,
Cavallin
F.
,
Cagol
M.
,
Alfieri
R.
et al.
(
2015
)
Esophageal cancer management: preoperative CA19.9 and CEA serum levels may identify occult advanced adenocarcinoma
.
World J. Surg.
39
,
424
432
[PubMed]
43.
Feng
F.
,
Tian
Y.
,
Xu
G.
,
Liu
Z.
,
Liu
S.
,
Zheng
G.
et al.
(
2017
)
Diagnostic and prognostic value of CEA, CA19-9, AFP and CA125 for early gastric cancer
.
BMC Cancer
17
,
737
[PubMed]
44.
Nozoe
T.
,
Rikimaru
T.
,
Mori
E.
,
Okuyama
T.
and
Takahashi
I.
(
2006
)
Increase in both CEA and CA19-9 in sera is an independent prognostic indicator in colorectal carcinoma
.
J. Surg. Oncol.
94
,
132
137
[PubMed]
45.
Wang
W.
,
Chen
X.L.
,
Zhao
S.Y.
,
Xu
Y.H.
,
Zhang
W.H.
,
Liu
K.
et al.
(
2016
)
Prognostic significance of preoperative serum CA125, CA19-9 and CEA in gastric carcinoma
.
Oncotarget
7
,
35423
35436
[PubMed]
46.
Chakraborty
C.
and
Das
S.
(
2016
)
Profiling cell-free and circulating miRNA: a clinical diagnostic tool for different cancers
.
Tumour Biol.
37
,
5705
5714
[PubMed]
47.
Pritchard
C.C.
,
Kroh
E.
,
Wood
B.
,
Arroyo
J.D.
,
Dougherty
K.J.
,
Miyaji
M.M.
et al.
(
2012
)
Blood cell origin of circulating microRNAs: a cautionary note for cancer biomarker studies
.
Cancer Prev. Res.
5
,
492
497
48.
Zhang
Q.
,
Wang
Q.
,
Sun
W.
,
Gao
F.
,
Liu
L.
,
Cheng
L.
et al.
(
2018
)
Change of circulating and tissue-based miR-20a in human cancers and associated prognostic implication: a systematic review and meta-analysis
.
Biomed Res. Int.
2018
,
6124927
[PubMed]
49.
Kawaguchi
T.
,
Komatsu
S.
,
Ichikawa
D.
,
Tsujiura
M.
,
Takeshita
H.
,
Hirajima
S.
et al.
(
2016
)
Circulating microRNAs: a next-generation clinical biomarker for digestive system cancers
.
Int. J. Mol. Sci.
17
,
50.
Wen
J.
,
Wang
G.
,
Xie
X.
,
Lin
G.
,
Yang
H.
,
Luo
K.
et al.
(
2019
)
Prognostic value of a four-miRNA signature in patients with lymph node positive locoregional esophageal squamous cell carcinoma undergoing complete surgical resection
.
Ann. Surg.
[PubMed]
51.
Wang
Z.
,
Zhang
J.
,
Zhang
Z.
,
Jiang
Y.
,
Li
M.
,
Li
Q.
et al.
(
2018
)
Prognostic value of miR-17-5 p in gastrointestinal cancers: a systematic review and meta-analysis
.
Onco Targets Ther.
11
,
5991
5999
[PubMed]
52.
Zhu
W.
and
Ji
X.
(
2019
)
The impact of microRNA-133a on prognosis and clinicopathological parameters for digestive system cancers: a comprehensive study based on meta-analysis and TCGA Database
.
Pathol. Oncol. Res.
53.
Zheng
G.
,
Pan
M.
,
Jin
W.
,
Jin
G.
and
Huang
Y.
(
2017
)
MicroRNA-135a is up-regulated and aggravates myocardial depression in sepsis via regulating p38 MAPK/NF-kappaB pathway
.
Int. Immunopharmacol.
45
,
6
12
[PubMed]
54.
Zhou
Y.
,
Li
S.
,
Li
J.
,
Wang
D.
and
Li
Q.
(
2017
)
Effect of microRNA-135a on cell proliferation, migration, invasion, apoptosis and tumor angiogenesis through the IGF-1/PI3K/Akt signaling pathway in non-small cell lung cancer
.
Cell. Physiol. Biochem.
42
,
1431
1446
[PubMed]
55.
Hu
N.
(
2018
)
Molecular mechanism of miR-135b-targeting APC gene in promoting invasion and metastasis of Hodgkin’s lymphoma cells
.
Lat. Am. J. Pharm.
37
,
651
656
This is an open access article published by Portland Press Limited on behalf of the Biochemical Society and distributed under the Creative Commons Attribution License 4.0 (CC BY).