Tissue microRNA-21 expression predicted recurrence and poor survival in patients with colorectal cancer – a meta-analysis

Introduction

Colorectal cancer (CRC) is one of the most common types of cancers and the leading cause of cancer-related death worldwide. In the US, CRC is the third most common cancer, with >143,000 new cases and >52,000 deaths each year.¹ Moreover, the incidence and mortality of CRC in Asia have also increased rapidly in the past decades.^2,3 There are three frequently used screening modalities, namely fecal occult blood testing, flexible sigmoidoscopy, and total colonoscopy, which have aided in reducing the mortality associated with this disease.³ However, none of these tests has been established as a well-accepted screening tool due to invasiveness, high cost, or low sensitivity. Since treatment of CRC in its nonmetastatic phase increases the survival rate, an insight into molecular pathways of CRC metastasis may be important for developing new prognostic molecular markers facilitating the reduction of CRC metastasis, particularly noninvasive biomarkers in serum or tissue.⁴

MicroRNAs (miRNAs) are small (21- to 25-nucleotide long) RNAs that participate in the regulation of cell differentiation, cell cycle progression, apoptosis, and tumorigenesis.⁵ MicroRNA-21 (miR-21) is an oncogenic miRNA that modulates the expression of multiple cancer-related target genes such as PTEN, TPM1, and PDCD and has been shown to be overexpressed in CRC.^6–10 Several studies have investigated the prognostic effect of miR-21 expression on CRC.^11–20 Most of the research revealed an elevated expression of miR-21 in CRC tissues compared with normal tissues, which is related to a poor survival outcome.^{4,12,16,20,21} However, there were reports showing the insignificant or opposite results.^14,18,19 Therefore, it is essentially necessary to carry out a systematic review and meta-analysis to summarize the published global findings, and get a better understanding on the significance of miR-21 expression in the prognosis of CRC patients.

In the current study, global-related literature were collected to conduct a systematic review and meta-analysis, and the risk of increased miR-21 expression to the survival of CRC patients was successfully assessed.

Materials and methods

Search strategy

The literature published from January 1, 2001 to March 20, 2014 were searched in PubMed, Embase, and Cochrane Library, using key words “microRNA-21 OR miR-21” AND “colon OR colorectal OR rectum” AND “cancer OR carcinoma OR tumor OR tumour OR neoplasm”. A manual review of the references of relevant publications was also performed to obtain additional studies.

Study selection

First, two reviewers primarily examined the titles and abstracts of all literature. Then, full text of the articles was screened separately by two reviewers to determine whether they met the inclusion criteria. We contacted the corresponding authors when the crucial data were not reported in the original papers. Articles were independently read and selected according to the inclusion criteria. Disagreements were resolved through the consensus with a third reviewer.

Inclusion criteria

The primary literature contained expression profiles of miR-21 and the following: (i) survival analysis of CRC patients with overall survival (OS) or disease-free survival (DFS) or recurrence-free survival (RFS); (ii) hazard analysis with hazard ratio (HR), 95% confidence intervals (CIs), and P-value, or relevant data that could be used to calculate the HR and 95% CI.

Exclusion criteria

The following studies were excluded: (i) those that were not published in English and (ii) those not involving the association between miR-21 expression level and CRC prognosis.

Quality assessment

Methodological quality of included articles was assessed using the Newcastle–Ottawa Scale.²² All studies were assessed for quality of selection (representativeness, selection of controls, ascertainment of exposure, absence of asthma at the start of study), comparability (confounding), and outcome (assessment of outcome, length and adequacy of follow-up). Studies could be awarded a maximum score of nine points. Studies with scores of five points or more were considered to be of moderate-to-good quality. Quality assessment was done by two authors using the Newcastle–Ottawa Scale independently. In case of disagreement, the third author was consulted. Quality assessment was completed before data extraction. The detailed assessments of each included article are shown in Table S1.

Data extraction

The extracted data elements included the following: (i) first author’s name and publication year; (ii) characteristics of the studied population including sample size, population, ethnicity, the numbers of death and survival during follow-up, follow-up time, and stage and histological type; (iii) measurement and cut-off values of miRNA expression; and (iv) HRs of miR-21 for OS, or DFS or RFS, along with the 95% CIs. We selected adjusted HR if crude HR and adjusted HR were both provided. The HR and 95% CIs of the study by Shibuya et al¹³ were reciprocals of the reported results of multivariate Cox regression due to inverse comparison.

Statistical methods

The effect size was summarized as HR. Analysis was done for OS and DFS (or RFS). The values were reported by a Forest plot, and uncertainty about the pooled estimates was quantified by 95% CI. Statistical heterogeneity was assessed by means of Cochran’s Q test (significant at P<0.10) and I² test (ranging from 0% to 100%).²³ The degree of heterogeneity with values of 25%, 50%, and 75% was considered low, moderate, and high, respectively.²⁴ Meta-regression was performed to explore the source of heterogeneity. The variables included in the meta-regression were defined as follows:

1. Type of carcinoma: colon cancer, rectal cancer, or CRC
2. Methods of measurement: fluorescence in situ hybridization (FISH) and quantitative real-time polymerase chain reaction (qRT-PCR)
3. Location of study: Asia, Europe, or America.

We used a random-effect model to summarize the overall results and results within subgroups and based statistically significant heterogeneity on a P-value of <0.05.²⁵ We also used Begg’s funnel plots²⁶ and Egger’s test²⁷ to detect possible publication bias. The statistical analyses were performed using STATA 12.0 (StataCorp LP, College Station, TX, USA). A two-tailed P<0.05 was considered statistically significant.

Results

Eligible studies

As shown in Figure 1, totally 107 studies of miR-21 and CRC prognosis were identified. Seventy-six studies were excluded based on manual screening of the title and abstract. Full text of the remaining 31 studies was identified. Of the 31 studies, 12 studies were not directly related to specific outcomes, seven studies lacked sufficient survival data or HR, and one study was found with an overlapping data set and not related to OS or DFS. Finally, eleven studies were considered eligible for inclusion in this review (Figure 1).

Figure 1 Methodological flow chart of the review. Abbreviations: miR-21, microRNA-21; CRC, colorectal cancer; HR, hazard ratio; OS, overall survival; DFS, disease-free survival.

Characteristics of included studies

The characteristics, including the first author, publication year, origin of population, the numbers of patients, source of samples, survival results, types of outcomes, time of follow-up, assay type, and miR-21 expression, are listed in Table 1. As shown in Table 1, our analysis included eleven studies, containing 3,669 subjects. Nine studies^{10,12,13,16–21} provided results of tissue samples, and three studies^14,15,18 provided results of serum samples. Moreover, four studies^13,16,17,19 involved the data of the association between miR-21 and DFS, and ten studies^{10,12–16,18–21} focused on the prognostic value of miR-21 for OS. However, Zhang et al¹⁷ reported three independent results of miR-21 with DFS in one study, while Nielsen et al¹⁶ and Schetter et al¹² reported two independent results of miR-21 with OS in one study each. Each independent result was included in the present meta-analysis.

Table 1 Main characteristics of included studies Abbreviations: HR, hazard ratio; miRNAs, microRNAs; DFS, disease-free survival; CC, colon cancer; qRT-PCR, quantitative real-time polymerase chain reaction; FISH, fluorescence in situ hybridization; RC, rectal cancer; CRC, colorectal cancer; OS, overall survival.

Tissue miR-21 expression and DFS or RFS in patients with CRC

The association of tissue miR-21 expression with DFS in patients with CRC is shown in Figure 2. In the meta-analysis of four studies (seven independent results, 1,567 subjects), high level of tissue miR-21 expression was significantly associated with poor DFS in patients with CRC (HR =1.59, 95% CI =1.20–2.10), with significant heterogeneity (I²=74.2%, P=0.001). Then, meta-regression was performed to explore the source of heterogeneity, and the results showed that different types of carcinomas among different studies may be the major reason of heterogeneity (Table S2). Therefore, stratified analysis was performed. In the sub-analysis of tumor types (Table 2), high level of tissue miR-21 expression significantly predicted poor DFS in patients with colon carcinoma (HR =1.54, 95% CI =1.33–1.79), without significant heterogeneity (I²=1.0%, P=0.40).

Figure 2 Forest plots of studies evaluating HRs of miR-21 with DFS in all studies (tissue). Note: Weights are from random effects analysis. Abbreviations: HRs, hazard ratios; miR-21, microRNA-21; DFS, disease-free survival; ES, effect size; CI, confidence interval.

Table 2 Stratified analysis of pooled HRs of CRC patients with elevated miR-21 expression from tissue Abbreviations: HRs, hazard ratios; CRC, colorectal cancer; miR-21, microRNA-21; CI, confidence interval; DFS, disease-free survival; RC, rectal cancer; OS, overall survival; qRT-PCR, quantitative real-time polymerase chain reaction; FISH, fluorescence in situ hybridization.

Tissue miR-21 expression and OS in patients with CRC

As shown in Figure 3, meta-analysis of eleven studies (14 independent results, 2,102 subjects) showed that high level of tissue miR-21 expression was significantly associated with poor OS in CRC patients (HR =1.53, 95% CI =1.23–1.90, I²=76.8%); however, significant heterogeneity existed. The meta-regression showed that measurement method of miR-21 expression across studies might have contributed to the heterogeneity (Table S2). Hence, the stratified analysis showed that (Table 2) high level of tissue miR-21 expression was significantly related to poor OS among studies that used qRT-PCR for measurement (HR =2.29, 95% CI =1.72–3.04, I²=17.6%), while the association was not significant among studies that used FISH for measurement (HR =1.08, 95% CI =0.98–1.19, I²=37.2%).

Figure 3 Forest plots of studies evaluating HRs of miR-21 with OS in all studies (tissue). Note: Weights are from random effects analysis. Abbreviations: HRs, hazard ratios; miR-21, microRNA-21; OS, overall survival; ES, effect size; CI, confidence interval.

Serum miR-21 expression and OS in patients with CRC

Only three studies reported the association of serum miR-21 expression with prognosis in CRC patients (Figure 4). All three studies focused on serum miR-21 and OS. High level of serum miR-21 expression was not significantly associated with poor OS in patients with CRC (HR =1.34, 95% CI =0.45–4.01, I²=79.3%).

Figure 4 Forest plots of studies evaluating HRs of miR-21 with OS in all studies (serum). Note: Weights are from random effects analysis. Abbreviations: HRs, hazard ratios; miR-21, microRNA-21; OS, overall survival; ES, effect size; CI, confidence interval.

Publication bias

Begg’s funnel plot and Egger’s test were performed to evaluate the publication bias of the studies. P-value of Egger’s regression intercepts for DFS was 0.573 in all studies of tissue samples, and the P-value for OS was 0.776 and 0.544 among studies that used FISH and qRT-PCR, respectively. Egger’s test showed that there was no significant publication bias in studies of serum samples (P=0.500).

Discussion

Recently, the prognostic effect of miR-21 has been observed in the meta-analysis of other cancers, such as lung cancer, especially non-small-cell lung cancer.²⁸ Emerging studies have demonstrated that aberrant expression of miRNAs is associated with prognosis of CRC. Here, we searched more databases and late literature and pooled the prognostic value. In this meta-analysis, elevated miR-21 expression in tumor tissue was found to be prognostic of recurrence, especially among colon cancer patients. qRT-PCR measurement showed that high expression of miR-21 in tumor tissue was significantly associated with poor OS among CRC patients.

DFS is a clinical end point that is a surrogate for the true end point and requires less time to evaluate and may be less costly to assess. Sargent et al²⁹ evaluated DFS with 3 years of follow-up (DFS_{3 years}) as a surrogate for OS with 5 years of follow-up (OS_{5 years}) and concluded that it was an appropriate primary end point to replace OS. Our results indicated that high expression of miR-21 in tumor tissue was a prognostic marker of CRC, especially among colon cancer patients. The reported surgically resected cases of CRC were known to have a 40%–60% recurrence rate in the first 3 years after surgery with the majority in the second year.³⁰ Recurrence of colon cancer still remains a major issue which affects nearly 50% of patients treated by conventional therapeutics.³¹ Detecting the level of miR-21 in tumor tissues of cancer patients, especially among colon cancer patients, might contribute to an apparent decrease in the chance of recurrence, which might have an important potential clinical significance for patient risk stratification, guidance of further follow-up, and additional therapy. Human colon cancer tissues were reported to be more sensitive than rectal cancer tissues to antitumor drugs in vitro.³² This is a potential reason that we found the obvious association between DFS and miR-21 in colon cancer patients in a sub-analysis. Similarly, Hansen et al³³ appealed to introduce a risk index for stratifying patients with stage II colon cancer which may lead to the identification of a considerably smaller group that the traditional approach would have categorized as high risk, sparing a large fraction of the patients of adjuvant chemotherapy with doubtful efficacy. However, to overcome the limitation of the significant heterogeneity across studies of CRC, more multicenter clinical investigations with larger sample sizes are needed.

Similar to Xia et al,³⁴ we also found a significant association between miR-21 expression and OS in CRC patients from studies that used qRT-PCR measurement. However, we specially identified the studies of miR-21 expression in tissues, excluding the studies of miR-21 expression in serum, due to the potential bias between different biological specimens. Moreover, we calculated pooled adjusted HRs,^12,18 which are more reliable than crude HRs. Our results indicated that, regardless of the method of measurement used, high expression of miR-21 in tumor tissue was associated with poor OS among CRC patients but was not significant among studies that used FISH measurement. This may be due to the small numbers of studies that used FISH. Currently, numerous techniques are available for studying miRNA expression levels, in addition to qRT-PCR and FISH. With the recent introduction of locked nucleic acid oligonucleotides as hybridization probes, miRNA-FISH has become a powerful technique for imaging the spatial localization of miRNA at the tissue, cellular, and even subcellular level.^35,36 Further studies are needed to confirm the exact association of the miR-21 expression with CRC in the future.

miRNAs are key players in a wide array of pathological processes, which may partly explain the prognostic associations with CRC. miR-21 has been found to play important roles in suppressing proapoptotic genes and modulating the pivotal components of the Ras/MEK/ERK pathway.³⁷ Additionally, previous studies reported that many tumor suppressors have been experimentally verified as the targets of miR-21 in CRC and other cancers, such as PTEN,^38,39 ITGβ4, and PDCD4.^40,41 miR-21 targets PTEN at the post-transcriptional level and regulates cell proliferation and invasion in CRC. In vitro, miR-21 blockade with ASO inhibits growth and induces apoptosis in CRC cell lines, which is mediated by retrieving PTEN expression. In vivo, injection of miR-21 ASO into CRC xenografts implanted subcutaneously in nude mice suppressed tumor growth.³⁸ Also, miR-21 gene causes downregulation of tumor suppressor genes (eg, PDCD4) and epithelial keygenes, such as ITGβ4, enhancing malignant transformation and epithelial–mesenchymal transition in CRC.⁴¹ Moreover, miR-21 in the tumor microenvironment plays an important role in CRC progression. Deregulation of miR-21 is a stromal phenomenon and supports CRC progression through myofibroblast transdifferentiation, resistance to oxaliplatin cytotoxicity, and promotion of tumor cell invasion. Stromal miR-21-induced invasion is mediated by RECK downregulation and a reciprocal rise in MMP2 activity.^42,43 Hence, miR-21 related to prognosis in CRC may be used as a potential target to identify patients for closer monitoring and therapy.

Limitation

There are several limitations in this paper: (1) It excluded non-English articles and the studies that lacked important survival data (eg, HR). (2) The proportion of cancers and TNM stage were not the same in all studies, and it might result in heterogeneity, especially OS analyses where several studies addressing stage IV patients were pooled with stage I–III. This may not be a safe way as OS in these two settings, the biology of the disease, and the function of miRNA-21 are not comparable. (3) The different cut-off values and other technical issues might also introduce heterogeneity, and it is impossible to identify the association of levels of miR-21 expression with OS or DFS in detail.

Conclusion

Our meta-analysis showed that elevated expression of miR-21 was associated with poor prognosis of CRC, especially among colon cancer patients. More large-scale and standard investigations should be conducted to confirm these findings.

Acknowledgments

This study was conducted by National Basic Research Program of China (973 Program) (grant no 2011CB503706), Zhejiang Natural Science Foundation (grant no LY14H260003), Zhejiang Sport Science Institute (J20131484), and China Scholarship Council.

Authors’ contributions

Zexin Chen, Hui Liu, and Wen Jin participated in the design of the study. Zheyuan Ding and Shuangshuang Zheng searched the articles and extracted data. Zexin Chen performed the statistical analysis and drafted the manuscript. Yunxian Yu conceived of the study, and participated in its design and coordination and helped to modify the manuscript. All authors contributed toward data analysis, drafting and critically revising the paper and agree to be accountable for all aspects of the work.

Disclosure

The authors declare that they have no conflicts of interest in this work.

---

References

1.		Siegel R, Naishadham D, Jemal A. Cancer statistics, 2012. CA Cancer J Clin. 2012;62(1):10–29.
2.		Sung JJ, Lau JY, Goh KL, Leung WK; Asia Pacific Working Group on Colorectal Cancer. Increasing incidence of colorectal cancer in Asia: implications for screening. Lancet Oncol. 2005;6(11):871–876.
3.		Pourhoseingholi MA. Increased burden of colorectal cancer in Asia. World J Gastrointest Oncol. 2012;4(4):68–70.
4.		Tokarz P, Blasiak J. The role of microRNA in metastatic colorectal cancer and its significance in cancer prognosis and treatment. Acta Biochim Pol. 2012;59(4):467–474.
5.		van Kouwenhove M, Kedde M, Agami R. MicroRNA regulation by RNA-binding proteins and its implications for cancer. Nat Rev Cancer. 2011;11(9):644–656.
6.		Meng F, Henson R, Wehbe-Janek H, Ghoshal K, Jacob ST, Patel T. MicroRNA-21 regulates expression of the PTEN tumor suppressor gene in human hepatocellular cancer. Gastroenterology. 2007;133(2):647–658.
7.		Zhu S, Si ML, Wu H, Mo YY. MicroRNA-21 targets the tumor suppressor gene tropomyosin 1 (TPM1). J Biol Chem. 2007;282(19):14328–14336.
8.		Asangani IA, Rasheed SA, Nikolova DA, et al. MicroRNA-21 (miR-21) post-transcriptionally downregulates tumor suppressor Pdcd4 and stimulates invasion, intravasation and metastasis in colorectal cancer. Oncogene. 2008;27(15):2128–2136.
9.		Wu CW, Ng SS, Dong YJ, et al. Detection of miR-92a and miR-21 in stool samples as potential screening biomarkers for colorectal cancer and polyps. Gut. 2012;61(5):739–745.
10.		Bovell LC, Shanmugam C, Putcha BD, et al. The prognostic value of microRNAs varies with patient race/ethnicity and stage of colorectal cancer. Clin Cancer Res. 2013;19(14):3955–3965.
11.		Schetter AJ, Nguyen GH, Bowman ED, et al. Association of inflammation-related and microRNA gene expression with cancer-specific mortality of colon adenocarcinoma. Clin Cancer Res. 2009;15(18):5878–5887.
12.		Schetter AJ, Leung SY, Sohn JJ, et al. MicroRNA expression profiles associated with prognosis and therapeutic outcome in colon adenocarcinoma. JAMA. 2008;299(4):425–436.
13.		Shibuya H, Iinuma H, Shimada R, Horiuchi A, Watanabe T. Clinicopathological and prognostic value of microRNA-21 and microRNA-155 in colorectal cancer. Oncology. 2010;79(3–4):313–320.
14.		Liu GH, Zhou ZG, Chen R, et al. Serum miR-21 and miR-92a as biomarkers in the diagnosis and prognosis of colorectal cancer. Tumour Biol. 2013;34(4):2175–2181.
15.		Menendez P, Padilla D, Villarejo P, et al. Prognostic implications of serum microRNA-21 in colorectal cancer. J Surg Oncol. 2013;108(6):369–373.
16.		Nielsen BS, Jorgensen S, Fog JU, et al. High levels of microRNA-21 in the stroma of colorectal cancers predict short disease-free survival in stage II colon cancer patients. Clin Exp Metastasis. 2011;28(1):27–38.
17.		Zhang JX, Song W, Chen ZH, et al. Prognostic and predictive value of a microRNA signature in stage II colon cancer: a microRNA expression analysis. Lancet Oncol. 2013;14(13):1295–1306.
18.		Toiyama Y, Takahashi M, Hur K, et al. Serum miR-21 as a diagnostic and prognostic biomarker in colorectal cancer. J Natl Cancer Inst. 2013;105(12):849–859.
19.		Kjaer-Frifeldt S, Hansen TF, Nielsen BS, et al. The prognostic importance of miR-21 in stage II colon cancer: a population-based study. Br J Cancer. 2012;107(7):1169–1174.
20.		Oue N, Anami K, Schetter AJ, et al. High miR-21 expression from FFPE tissues is associated with poor survival and response to adjuvant chemotherapy in colon cancer. Int J Cancer. 2014;134(8):1926–1934.
21.		Chen TH, Chang SW, Huang CC, et al. The prognostic significance of APC gene mutation and miR-21 expression in advanced-stage colorectal cancer. Colorectal Dis. 2013;15(11):1367–1374.
22.		Wells GA, Shea B, O’Connell D, et al. The Newcastle-Ottawa Scale (NOS) for assessing the quality of nonrandomised studies in meta-analyses. Available from: http://www.ohri.ca/programs/clinical_epidemiology/oxford.asp. Accessed March 30, 2014.
23.		Higgins JP, Thompson SG. Quantifying heterogeneity in a meta-analysis. Stat Med. 2002;21(11):1539–1558.
24.		Higgins JP, Thompson SG, Deeks JJ, Altman DG. Measuring inconsistency in meta-analyses. BMJ. 2003;327(7414):557–560.
25.		Lau J, Ioannidis JP, Schmid CH. Quantitative synthesis in systematic reviews. Ann Intern Med. 1997;127(9):820–826.
26.		Begg CB, Mazumdar M. Operating characteristics of a rank correlation test for publication bias. Biometrics. 1994;50(4):1088–1101.
27.		Egger M, Davey Smith G, Schneider M, Minder C. Bias in meta-analysis detected by a simple, graphical test. BMJ. 1997;315(7109):629–634.
28.		Yang M, Shen H, Qiu C, et al. High expression of miR-21 and miR-155 predicts recurrence and unfavourable survival in non-small cell lung cancer. Eur J Cancer. 2013;49(3):604–615.
29.		Sargent DJ, Wieand HS, Haller DG, et al. Disease-free survival versus overall survival as a primary end point for adjuvant colon cancer studies: individual patient data from 20,898 patients on 18 randomized trials. J Clin Oncol. 2005;23(34):8664–8670.
30.		Aghili M, Izadi S, Madani H, Mortazavi H. Clinical and pathological evaluation of patients with early and late recurrence of colorectal cancer. Asia Pac J Clin Oncol. 2010;6(1):35–41.
31.		Kanwar SS, Poolla A, Majumdar AP. Regulation of colon cancer recurrence and development of therapeutic strategies. World J Gastrointest Pathophysiol. 2012;3(1):1–9.
32.		Ueo H, Maehara Y, Saito A, Sakaguchi Y, Kohnoe S, Sugimachi K. Human colon cancer tissues are more sensitive than rectal cancer tissues to antitumor drugs in vitro. Oncology. 1991;48(2):158–161.
33.		Hansen TF, Kjaer-Frifeldt S, Christensen RD, et al. Redefining high-risk patients with stage II colon cancer by risk index and microRNA-21: results from a population-based cohort. Br J Cancer. 2014;111(7):1285–1292.
34.		Xia X, Yang B, Zhai X, et al. Prognostic role of microRNA-21 in colorectal cancer: a meta-analysis. PLoS One. 2013;8(11):e80426.
35.		Kloosterman WP, Wienholds E, de Bruijn E, Kauppinen S, Plasterk RH. In situ detection of miRNAs in animal embryos using LNA-modified oligonucleotide probes. Nat Methods. 2006;3(1):27–29.
36.		Silahtaroglu AN, Nolting D, Dyrskjøt L, et al. Detection of microRNAs in frozen tissue sections by fluorescence in situ hybridization using locked nucleic acid probes and tyramide signal amplification. Nat Protoc. 2007;2(10):2520–2528.
37.		Hatley ME, Patrick DM, Garcia MR, et al. Modulation of K-Ras-dependent lung tumorigenesis by MicroRNA-21. Cancer Cell. 2010;18(3):282–293.
38.		Xiong B, Cheng Y, Ma L, Zhang C. MiR-21 regulates biological behavior through the PTEN/PI-3 K/Akt signaling pathway in human colorectal cancer cells. Int J Oncol. 2013;42(1):219–228.
39.		Zhang JG, Wang JJ, Zhao F, Liu Q, Jiang K, Yang GH. MicroRNA-21 (miR-21) represses tumor suppressor PTEN and promotes growth and invasion in non-small cell lung cancer (NSCLC). Clin Chim Acta. 2010;411(11–12):846–852.
40.		Chen Y, Knosel T, Kristiansen G, et al. Loss of PDCD4 expression in human lung cancer correlates with tumour progression and prognosis. J Pathol. 2003;200(5):640–646.
41.		Ferraro A, Kontos C, Boni T, et al. Epigenetic regulation of miR-21 in colorectal cancer: ITGB4 as a novel miR-21 target and a three-gene network (miR-21-ITGBeta4-PCDC4) as predictor of metastatic tumor potential. Epigenetics. 2013;9(1):129–141.
42.		Bullock M, Pickard K, Nielsen BS, et al. Deregulated stromal microRNA-21 and promotion of metastatic progression in colorectal cancer. Lancet. 2014;383:S30.
43.		Bullock MD, Pickard KM, Nielsen BS, et al. Pleiotropic actions of miR-21 highlight the critical role of deregulated stromal microRNAs during colorectal cancer progression. Cell Death Dis. 2013;4:e684.

Supplementary materials

Table S1 The results of meta-regression Notes: aThe distribution of two variables is same. The bold values means that this factor is significant (P<0.05). Abbreviations: DFS, disease-free survival; OS, overall survival.

Table S2 The quality assessment of articles Notes: ‘x’ represents no score in corresponding items. “(a)” and “(b)” in this table represents the choice in first item according to the Newcastle–Ottawa Quality Assessment Scale-Cohort studies.

---

References

1.		Zhang JX, Song W, Chen ZH, et al. Prognostic and predictive value of a microRNA signature in stage II colon cancer: a microRNA expression analysis. Lancet Oncol. 2013;14(13):1295–1306.
2.		Kjaer-Frifeldt S, Hansen TF, Nielsen BS, et al. The prognostic importance of miR-21 in stage II colon cancer: a population-based study. Br J Cancer. 2012;107(7):1169–1174.
3.		Nielsen BS, Jorgensen S, Fog JU, et al. High levels of microRNA-21 in the stroma of colorectal cancers predict short disease-free survival in stage II colon cancer patients. Clin Exp Metastasis. 2011;28(1):27–38.
4.		Shibuya H, Iinuma H, Shimada R, Horiuchi A, Watanabe T. Clinicopathological and prognostic value of microRNA-21 and microRNA-155 in colorectal cancer. Oncology. 2010;79(3–4):313–320.
5.		Oue N, Anami K, Schetter AJ, et al. High miR-21 expression from FFPE tissues is associated with poor survival and response to adjuvant chemotherapy in colon cancer. Int J Cancer. 2014;134(8):1926–1934.
6.		Toiyama Y, Takahashi M, Hur K, et al. Serum miR-21 as a diagnostic and prognostic biomarker in colorectal cancer. J Natl Cancer Inst. 2013;105(12):849–859.
7.		Chen TH, Chang SW, Huang CC, et al. The prognostic significance of APC gene mutation and miR-21 expression in advanced-stage colorectal cancer. Colorectal Dis. 2013;15(11):1367–1374.
8.		Schetter AJ, Leung SY, Sohn JJ, et al. MicroRNA expression profiles associated with prognosis and therapeutic outcome in colon adenocarcinoma. JAMA. 2008;299(4):425–436.
9.		Menendez P, Padilla D, Villarejo P, et al. Prognostic implications of serum microRNA-21 in colorectal cancer. J Surg Oncol. 2013;108(6):369–373.
10.		Liu GH, Zhou ZG, Chen R, et al. Serum miR-21 and miR-92a as biomarkers in the diagnosis and prognosis of colorectal cancer. Tumour Biol. 2013;34(4):2175–2181.
11.		Bovell LC, Shanmugam C, Putcha BD, et al. The prognostic value of microRNAs varies with patient race/ethnicity and stage of colorectal cancer. Clin Cancer Res. 2013;19(14):3955–3965.