Prognostic Role of MicroRNA-21 in Non-small Cell Lung Cancer : a Meta-analysis

Lung cancer is the most common cause of cancer mortality in the world. Approximately 1.6 million new cases of lung cancer will be diagnosed and 1.4 million deaths will occur from lung cancer during 2008. Among all cases suffered lung cancer, non-small-cell lung cancer (NSCLC) approximately accounts for 80% (Ramalingam et al., 1998; Jemal et al., 2011) MicroRNAs (miRNAs) are endogenous, small noncoding and have a length of 18-25 nucleotides RNAs. These miRNAs could identify post transcriptional gene regulators that paired to complementary sequences in the 3’ untranslated region (3’ UTR) of target mRNAs, leading to mRNA degradation or translational repression (Bartel 2004; Bartel and Chen 2004). Many studies have reported that over expression of microRNA-21 (miR-21) play important roles in increasing cell proliferation, migration, invasion and survival (Lu et al., 2008; Yang et al., 2011). Relatively speaking suppression or knock-down of miR-21 could induce apoptosis and repress cell proliferation and invasion (Zhang et al., 2010). The miR-21 was also found to be elevated in many cancers such as lymphoma, prostate cancer, colorectal cancer and breast cancer (Lawrie et al., 2007; Siva et al., 2009; Qian et al., 2009; Nielsen et al., 2011). However, the expression of miR-21 in NSCLC and


Introduction
Lung cancer is the most common cause of cancer mortality in the world.Approximately 1.6 million new cases of lung cancer will be diagnosed and 1.4 million deaths will occur from lung cancer during 2008.Among all cases suffered lung cancer, non-small-cell lung cancer (NSCLC) approximately accounts for 80% (Ramalingam et al., 1998;Jemal et al., 2011) MicroRNAs (miRNAs) are endogenous, small noncoding and have a length of 18-25 nucleotides RNAs.These miRNAs could identify post transcriptional gene regulators that paired to complementary sequences in the 3' untranslated region (3' UTR) of target mRNAs, leading to mRNA degradation or translational repression (Bartel 2004; Bartel and Chen 2004).
Many studies have reported that over expression of microRNA-21 (miR-21) play important roles in increasing cell proliferation, migration, invasion and survival (Lu et al., 2008;Yang et al., 2011).Relatively speaking suppression or knock-down of miR-21 could induce apoptosis and repress cell proliferation and invasion (Zhang et al., 2010).The miR-21 was also found to be elevated in many cancers such as lymphoma, prostate cancer, colorectal cancer and breast cancer (Lawrie et al., 2007;Siva et al., 2009;Qian et al., 2009;Nielsen et al., 2011).However, the expression of miR-21 in NSCLC and The aim of this study is to comprehensively and quantitatively summarize the evidence for the use of miR-21 to predict the clinical results of NSCLC patients.And we also want to evaluate the overall risk of elevated miR-21 for survival in patients with NSCLC.

Study inclusion/exclusion criteria
Studies were considered eligible if they met all of the following inclusion criteria, (i) discussed patients with NSCLC (ii) measured the miR-21 expression in tumor and serum;(iii) investigated the survival outcome or the correlation between miR-21 expression and the clinical variables.Studies were excluded based on any of the following reasons, (i) were review articles, laboratory articles or letters (ii), described the survival outcome of other tumors or other markers, (iii) lacked key information for calculation with methods developed by Parmar, Williamson , and Tierney (Parmar et al., 1998;Williamson et al., 2002;Tierney et al., 2007), (iv) the articles from one author and the studies brought into the repeated samples from the same patients .

Data Extraction
Eligible articles were reviewed independently by two investigators (Ma XL and Liu Lei).Disagreements were resolved by consensus.Multivariate Cox hazard regression analysis reported in the article was included in the our analysis; if these data were not available, we extracted univariate Cox hazard regression analysis or log-rank p value and Kaplan-Meier survival curves of survival outcomes instead.Above primary information had been extracted by two investigators (Ma XL and Liu XX) independently.Additional data were extracted from the studies included first author, publication year, study size, patients age and sexuality, smoker or not, TNM stage, lymph status, histological classification, methods to detect miR-21, positive miR-21 definition, the attitude conclusion and other clinical characteristics.

Statistical Methods
All these HRs and 95% confidence interval(CI) were calculated following Tierney's method and the logHR and SE (logHR) (SE) were used for aggregation of the survival results, but these statistical variables were not given directly in most studies.We calculated the necessary statistics on the basis of a available numerical data with methods developed by Parmar, Williamson, and Tierney.Calculation was accomplished by the software designed by Matthew Sydes and Jayne Tierney with these methods (Medical Research Council Clinical Trials Unit, London, UK) (Tierney et al., 2007).
We also examine the correlation between miR-21 expression and the clinical variables including TNM stage, lymph node status, histological type, sexuality and smoking status.According to clinical characteristics, Stage I and Stage II were combined and Stage III and Stage IV were combined.Odds ratio (OR) was used as the measure index to describe the correlation (CORNFIELD 1951).
Forrest plots were used to estimate the effect of miR-21 expression on survival outcome and the correlation between miR-21 expression and the clinical variables.Heterogeneity was defined as p<0.10 or I2>50%.When homogeneity was fine (p≤0.10,I 2 ≤50%), a fixed effect model was used for secondary analysis.If not, a random effect model was used (Higgins et al., 2003).An observed HR>1 indicated worse outcome for the positive group relative to the negative group and would be considered statistically significant if the 95% CI did not overlap 1.The Begg's rank correlation also was applied to assess the potential publication bias, p>0.05 was considered that there was no potential publication bias (Begg, 1994).All above calculations were performed using RevMan 5.1 (Cochrane collaboration, Oxford, UK) Publication biases were evaluated using the Begg's funnel plot by STATA 11.0 (STATA Corporation, College Station, TX).

Eligible Studies
The initial search returned 42 studies in PubMed and EMBASE.Following review of these abstracts, 13 potentially relevant studies were identified as eligible for full-text review.5 studies were excluded, because these studies were short of the necessary data for calculation.One article from Saito M (Saito et al., 2011) gave the CSS for Maryland/Norway cohort, and the recurrence free survival (RFS) for Japan cohort.Because there was only one study for CSS, we just used the RFS for Japan cohort in our meta-analysis.We combined the results disease free survival (DFS) and RFS together as RFS/ DFS.There were 3 articles from Gao W, one article used RFS to describe the survival outcome (Gao et al., 2012), while another two used OS (Gao et al., 2010(Gao et al., , 2011)).We exclude one article for overall survival (OS) in the meta-analysis, because Gao W referred in the article  that they used the repeated data have been defined and described previously (Gao et al. 2011).Liu XG's study (Liu et al., 2011) gave the HRs and CI both in the tumor and the serum, we used the correspond data in different meta-analysis.Finally, we enrolled 8 eligible articles containing survival outcomes (Markou et al., 2008;Gao et al., 2010Gao et al., 2011Gao et al., , 2012;;Voortman et al., 2010;Liu et al., 2011;Saito et al., 2011;Wang et al., 2011), respectively, 6 articles (Markou et al., 2008;Gao et al., 2010Gao et al., , 2011Gao et al., , 2012;;Voortman et al., 2010;Saito et al., 2011) for tumor, and 2 articles (Liu et al., 2011;Wang et al., 2011) for serum in our meta-analysis (Figure 1).These eligible studies were published from 2008 to 2012.The 8 eligible studies included a total of 1163 patients with a median number of 145 patients per study.The patients' clinical characteristics and other useful information have been extracted in Table 1.

Assessment of publication bias
Begg's test was used to examine publication bias.No significant publication biases were found in results of meta-analyses of miR-21 prediction value for OS both using tumor and serum samples (P=0.497 and P=0.317 respectively).There was also no publish bias in the studies for RFS/DFS (P=0.602)(Figure 4).

Discussion
As we know, it was the first time that a comprehensive and detailed meta-analysis revealed the prognostic role of tumor and serum miR-21 for NSCLC.The prognostic role of miR-21 expression was still a great puzzle according to the evidence-based medicine in our study.
The prognostic values of miR-21 have been proven in the recent meta-analysis in head and neck squamous cell carcinoma (HNSCC) and carcinomas in digestion system.Our result of HR for OS was 2. 19 [0.76, 6.30].The interval of HR overlapped 1, it showed that the miR-21 expression had no prognostic significance on NSCLC.The subgroup analysis showed that the HR for Asian group was 5. 49 [2.46, 12.27], it suggested that the miR-21 expression played significant prognostic role on NSCLC.Meanwhile, the combined HR for non-Asian group was 1. 25 [0.48, 3.25],it suggested that the miR-21 was no find the effect to predict survival outcome in non-Asian group.
The HR for RFS/DFS to predict the survival outcome was statistically significant.As referred in Hayes (Hayes et al. 2001), a prognostic factor with RR>2 is considered as useful practical value.All the meta-analysis except the pooled result for OS had a well result.It may suggest that detected miR-21 expression in NSCLC patients could predict their prognosis practically.However, Voortman J's study (Voortman et al. 2010) indicated that the miR-21 expression patterns examined were neither predictive nor prognostic.And this study have the greatest patients cohort with radically resected NSCLC.This clinical discovery was opposite to the recent laboratory studies, these studies most considered the miR-21 as a risky microRNA for lung cancer (Diederichs and Haber 2006;Wang et al., 2009;Hatley et al., 2010).This puzzle could be solved when much more studies were conducted to confirm clinical value of the miR-21 expression.And the prognostic role of miR-21 for Asian group could be confirmed by adequately multi-center designed prospective studies in Asian country in future.We could also find the serum miR-21 could play prognostic role on NSCLC (HR 2.08[1.55,2.80]).This result indicated that serum miR-21could be a diagnostic tool to detect the NSCLC.Some clinical studies and review have been proven the application of the microRNA (Liang 2008;de Planell-Saguer M and Rodicio, 2011;Wei et al., 2011).
In the correlation study of miR-21 expression with patients' clinical characteristics, TNM stage, sexuality and smoking status showed complete no correlation with miR-21 expression while ORs for lymph metastasis and histological type were significant.The miR-21 have been proven to regulates the metastatic behavior in both mouse and human by activating many signal ways (Gao et al., 2012).These mechanism indicated that miR-21might be correlated with the metastasis.Much more clinical studies were required to research the correlation between miR-21 and metastatic factors.
Significant heterogeneity was found in the metaanalysis for OS of the prognostic role of miR-21 (I 2 = 88%, P<0.0001).To exclude technique biases, subgroup analyses were performed by the country group, detecting

Figure 1 .
Figure 1.Selection of Studies

Figure 2 .
Figure 2.Estimated Hazard Ratios (HRs) Summary for (A) overall survival with miR-21 expression in tumor tissue, (B) disease free survival or recurrence free survival with miR-21 expression in tumor tissue, and (C) overall survival with miR-21 expression in serum

Figure 4 .
Figure 4. Funnel Plots of Publication Bias Summary for Corresponding Meta-analysis in Figure 2. Orderly, they are Funnel plots of publication bias for meta-analysis of hazard ratios (HRs) for (A) overall survival with miR-21 expression in tumor tissue (B)disease free survival or recurrence free survival with miR-21 expression in tumor tissue, and (C) overall survival with miR-21 expression in serum

Figure 3 .
Figure 3.Estimated Odds Ratios (ORs) Summary for Correlation of (A) miR-21 expression and lymph node status, (B) miR-21 expression and histological differentiation