research article the clinicopathological...

Research ArticleThe Clinicopathological Significance of MicroRNA-155 inBreast Cancer A Meta-Analysis

Hui Zeng1 Cheng Fang1 Seungyoon Nam2 Qing Cai3 and Xinghua Long1

1 Zhongnan Hospital of Wuhan University Wuhan 430071 China2 Cancer Genomics Branch National Cancer Center Goyang-si Gyeonggi-do 410-769 Republic of Korea3Hubei University of Chinese Medicine Wuhan 430061 China

Correspondence should be addressed to Xinghua Long xlong888yahoocom

Received 15 May 2014 Accepted 3 July 2014 Published 3 August 2014

Academic Editor Elisa Giovannetti

Copyright copy 2014 Hui Zeng et alThis is an open access article distributed under theCreative CommonsAttribution License whichpermits unrestricted use distribution and reproduction in any medium provided the original work is properly cited

Objective Previous studies demonstrated that the associations between expression level of microRNA-155 (miR-155) andclinicopathological significance of breast cancer remained inconsistent Therefore we performed a meta-analysis based on eligiblestudies to summarize the possible associations MethodsWe identified eligible studies published up toMay 2014 by a comprehensivesearch of PubMed EMBASE CNKI and VIP databases The analysis was performed with RevMan 50 software Results A totalof 15 studies were included The results of meta-analysis showed that miR-155 was positively correlated with breast cancer withstandardized mean difference (SMD) = 122 Elevated miR-155 was found in Her-2 positive or lymph node metastasis positive orp53mutant type breast cancer But the result showed to be insignificant inTNMcomparisonWith respect to estrogen receptor alpha(ER) and progesterone receptor (PR) status both of them showed significant associations with SMD = minus12 and minus185 respectivelyConclusionMiR-155 detection might have a diagnostic value in breast cancer patients It might be used as an auxiliary biomarkerfor different clinicopathological breast cancer

1 Introduction

Development of tumors is considered to be a complexmultistep process including the activation of oncogenesand the silencing of tumor suppressors [1] MicroRNAs(miRNAs) are kinds of highly conserved approximately 20sim22 nucleotide noncodingRNAs EndogenousmiRNAs inhibitthe expression of target genes in the translation level byassembling to RNA-induced silence complex (RISC) throughparts of base pairing with the 31015840 untranslated region (UTR)of target mRNAs [2ndash4]

Recent studies have shown that miRNAs were involvedin various physiological and pathological processes and anincreasing number of evidences prove that the abnormalexpression ofmiRNAs has a direct correlationwith the occur-rence and development of cancers The abnormal expressionand dysfunction ofmiRNAsmay lead to cellular disorder andfinally result in diseases or even cancers [5] Thus miRNAsmay be exploited as new kinds of molecular targets fordiagnosing and treating cancers

At present in studies that focused on the associationbetween miRNAs and tumors abnormal expression of miR-155 in patients with breast cancer has gained substantialattention As an oncogene high expression of miR-155 wasconsidered as a breast cancer risk factor Suppressor cytokinesignaling-1 (socs1) is a constant target gene of miR-155 inbreast cancer cells in the human evolutionary process Theexpression ofmiR-155 and socs1 is negative correlated [6]Thestudy by Kong et al has shown that miR-155 participates inthe process of epithelial to mesenchymal transition (EMT)and infiltration in NMuMG cells [7] suggesting that miR-155 could be used as a diagnostic marker for breast cancermetastasis Our group has reportedmultiplemiRNAnetworkclusters involving in antiestrogen resistance in breast cancermiR-155 was one of the most important dysregulated miR-NAs indicating thatmiR-155might be utilized as a diagnosticmarker for breast cancer endocrine therapy [8]

As the most common malignant tumor in the worldamong women it had been anticipated that there would be15 million women being newly diagnosed by 2010 in 2004

Hindawi Publishing CorporationBioMed Research InternationalVolume 2014 Article ID 724209 7 pageshttpdxdoiorg1011552014724209

2 BioMed Research International

[9] In Chinese population breast cancer has an incidencerate of 1639 per 100000 Chinese women and seriously affectspeoplersquos lives and health Among cancers that happened inChinese women of economically developed provinces andcities breast cancer has the highest incidence and is thefourth most common cause of cancer death [10] Recentlythere were many investigations referring to the relationshipbetween miR-155 and breast cancer but because of the smallsample size discrepancy exists among different studies moreor less Therefore in order to provide a more considerableclearer and more systematic recognition of the expression ofmiR-155 in breast cancer we collected the data related tomiR-155 expression in breast cancer to carry out themeta-analysis

2 Materials and Methods

21 Search Strategy The literature retrieval was performedby two independent reviewers (Hui Zeng and Cheng Fang)All studies included in the meta-analysis were selected bysearching the PubMed EMBASE CNKI and VIP databasesup to May 2014 using the following keywords ldquomiR-155rdquo orldquomicroRNA-155rdquo and ldquobreast cancerrdquo All references in thesestudies were examined to identify additional research thatwas not indexed by the databases We selected publishedarticles written in English or Chinese

22 Including and Excluding Criteria Including criteriainclude (1) being related to breast cancer and miR-155 (2)patients that are confirmed by pathology (3) sufficient dataincluding mean value and standard deviation or other datathat can result in mean value and standard deviation and(4) measurementmethods and experiment group that are thesame or almost same

Excluding criteria include (1) reviews comments orletters (2) low-quality or incomplete data (3) abstract onlyand lack of the full text without authorrsquos reply and (4)reduplicate publication For duplicate articles only the mostrecent or largest data set was selected

23 Data Extraction Data were extracted from eligible arti-cles independently by two of the authors with any disagree-ment resolved by consensus The following information wascollected in a predefined data collection form first authorrsquosname publication year country sample type total numberof cases and controls quantitative methods and publicationlanguage

24 Literature Quality Evaluation Based on the results ofthe system we used grading method [11] recommended bythe GRADE system to evaluate quality of evidence Evidencequality classification is as follows (I) high quality furtherresearch would not change the credibility of evaluationresults about the curative effect (II) medium quality furtherresearch is likely to affect the credibility of evaluation resultsabout the curative effect and may change the assessmentresults (III) low quality further research is likely to affect thecredibility of evaluation results about the curative effect andthe assessment results are very likely to change (IV) very lowquality any curative effect evaluation results are uncertain

25 Statistic Analysis Rev-Man 50 software which wasrecommended by Cochrane collaboration was used in thismeta-analysis Rev-Man which is short for review manageris the software used for preparing and maintaining Cochranereviews Results can be presented graphically with the soft-ware

Firstly heterogeneity between studies was assessed by 1205942-based 119876-tests and 1198682 tests where 1198682 () gt 50 or 119875 lt 010was considered significantly heterogeneous [12]The randomeffect model (DerSimonian-Laird) [13] was used to assesspooled odds ratios (ORs) when significant heterogeneitywas observed Otherwise the fixed effect model (Mantel-Haenszel) [14] was used Sensitivity analysis was used toanalyze the stability of the text results by omitting one studyat a time For continuous data if quantitative method is thesame we adopted the weighted mean difference (MD) as ouranalysis index If they used different measuring instrumentsor units for the same variable or there was large differenceamong the mean value of numerical analysis the standard-ized mean difference (SMD) was adopted for analysis Wecalculated 95 confidence interval (CI) of all analysis At thesame time the funnel chart is used to determine publicationbias Data input and monitor were done by two researchers

3 Results

31 Study Characteristics For the initial inspection 65related English articles and 53 Chinese articles were obtainedby literature search from the PubMed EMBASE CNKIand VIP databases After titles and abstracts were screened86 articles were excluded because of irrelevant or dupli-cate records The full texts of the remaining 32 recordswere carefully reviewed Among these articles seven arti-cles were abandoned for overlapped data and five articleswere excluded because of review papers Another five wereexcluded due to data that were incomplete or inappropriatelycalculated or original data that could not be obtained despiteattempts to contact the authors Therefore 15 [15ndash29] articleswere considered in the present meta-analysis Among them13 discussed the different expression level ofmiR-155 betweenbreast cancer samples and normal samples and another 2only discussed it between different subtype breast cancersOne [27] was included in ER and PR analysis and another[28] was included in TNM and p53 analysis The specificretrieval process was shown in Figure 1 The study character-istics included in the meta-analysis were listed in Table 1

32 Results of Meta-Analysis A total of 13 studies including791 breast cancer samples and 509 normal samples werecollected in this sectionAs shown in Figure 2 we observed anelevated miR-155 expression in breast cancer samples (SMD= 122 95 CI = 065ndash178 119875 lt 000001)

However high heterogeneitywas observed in the analysisand sensitivity analyses indicated that the study from Lu et al[22] was mainly responsible for the observed heterogeneityWhen we excluded this study the high heterogeneity was sig-nificantly decreased and the association was still significantlydifferent (for SMD = 058 95 CI = 046ndash071 119875 lt 000001119875 for heterogeneity = 015 1198682 = 31) Sensitivity analyses

BioMed Research International 3

Table 1 Characteristics of studies included in this meta-analysis

Reference Year Origin Sample size (casecontrol) Quantitative method LanguageChen et al [15] 2012 China Tissue 9292 qRT-PCR EnglishFu and Zhang [16] 2011 China Tissue 3838 qRT-PCR ChineseHafez et al [17] 2012 Egypt Tissue 4040 qRT-PCR EnglishHan et al [18] 2013 China Serum 4522 qRT-PCR ChineseHeneghan et al [19] 2010 Ireland Blood 8363 qRT-PCR EnglishHuang et al [20] 2013 China Plasma 5530 qRT-PCR ChineseIorio et al [21] 2005 Italy Tissue 766 microarray EnglishLu et al [22] 2012 China Tissue 6767 qRT-PCR EnglishOuyang et al [23] 2014 China Tissue 33 microarray EnglishMar-Aguilar et al [24] 2013 Mexico Serum 6110 qRT-PCR EnglishShao et al [25] 2013 China Serum 165120 qRT-PCR ChineseSun et al [26] 2012 China Serum 10355 qRT-PCR EnglishWang et al [27] 2010 China Tissue 5858 qRT-PCR EnglishWang and Zhang [28] 2011 China Serum 2010 qRT-PCR ChineseZheng et al [29] 2012 China Tissue 4545 qRT-PCR Chinese

86 irrelevant articles or duplicate records excluded

15 articles available for meta- analysis (5 Chinese articles and

10 English articles)

7 articles with overlapped data 5 articles were review papers3 with data incomplete2 with data inappropriately calculated

32 of full-text articles assessed for eligibility

118 retrieved articles (65 English articles and 53 Chinese articles)

Figure 1 Flow diagram of study identification

indicated that the pooled SMD was consistently significantby omitting one study at a time for the last studies We thendid a subgroup analyses to investigate the expression level ofmiR-155 in different sample types As the study of Lu et al [22]mainly contributed to the high heterogeneity we excluded itin the subgroup analyses In both tissue sample and bloodsample higher miR-155 expression level was detected inbreast cancer samples than that in normal samples (with SMD= 063 95 CI = 039ndash087 119875 lt 000001 SMD = 056 95CI = 041ndash071 119875 lt 000001 resp Figure 3)

In subtype analysis the expression level of miR-155 wassignificantly correlated with estrogen receptor alpha (ER)progesterone receptor (PR) Her-2 lymph node metastasistumor size and p53 status In subtype analysis of differentER and PR status breast cancer miR-155 was significantly lessexpressed in ER+ or PR+ breast cancer But it was highlyexpressed inHer-2+ or lymph nodemetastasis positive breast

cancer compared with Her-2minus or lymph node metastasisnegative breast cancer Comparing breast cancer with tumorsize gt2 cm with that lt2 cm the one with larger tumor maybe along with higher miR-155 expression level When breastcancer with wild p53 type and breast cancer with mutantp53 type were compared higher miR-155 expression wasdetected inmutant p53 type breast cancer However themiR-155 expression level in different TNM grades breast cancersamples showed no significant difference (Table 2)

33 Evaluation of Publication Bias We assessed the publica-tion bias of the literature by the funnel plot The funnel plotshowed thatmost of the researches lay in the top of the funneland rare in the base The shape of the funnel plot did notreveal any evidence of obvious asymmetry (Figure 4) Eggerrsquostest and Beggrsquos test showed 119875 gt 005

4 Discussion

Although the pathogenesis of breast cancer has not beencompletely understood the occurrence development andmetastasis of cancer are sure to be genes participated Inrecent years miRNA studies started a new field for cancerresearch there were studies showing [21 30ndash32] that miR-155 might be closely related with breast cancer and played acrucial role in the development of it Zhu et al [33] found thatmiR-155 expression in breast cancer tissues was higher thanthat in normal tissue lymph node metastasis and the level ofestrogen receptor alpha (ER) and progesterone receptor (PR)were associated with the expression of miR-155 level in thestudy Wang et al [27] also showed significant associationsbetween the expression of miR-155 level and the status of ERand PR In this meta-analysis we analyzed the correlationbetween expression level of miR-155 and characteristics ofbreast cancer We did our best to do a comprehensive searchto avoid publication bias and closely followed both includingand excluding criteria We evaluated the publication bias by


Chen et al 2012 095710352

01109minus0707

minus0444

31001908

3970006349

0599618218

1202

0193052302981284

06000635

2620000875

019201125

11230560054

1038404583

067570149

00076

29minus1966

01133038

01221201

07000461

02300162

1435576673

6101 003

000018750001401

104066

01640911

16510345

2125148023minus087

10 68 170 [065 276]

045 [001 089]099 [045 153]

049 [004 094]099 [015 184]778 [677 878]

073 [049 097]039 [006 072]065 [023 108]

122 [065 178]

384022

85

858788

85

857570

79

8783

23

1000509

Favours experimental Favours control

Fu and Zhang 2011Hafez et al 2012Han et al 2013Heneghan et al 2010

Iorio et al 2005Lu et al 2012Ouyang et al 2014Mar-Aguilar et al 2013Shao et al 2013Sun et al 2012Zheng et al 2012

Total (95 CI) 791

Study or subgroupMean Mean

BC NBC

SD SDTotal Total Weight

63306

673

101205545

031 [minus002 064]

044 [minus002 089]

288 [minus033 609]050 [minus018 117]

Heterogeneity 1205912 = 093 120594

2

= 21020 df = 12 (P lt 000001) I

2

= 94

Test for overall effect Z = 423 (P lt 00001)

Standard mean difference Standard mean differenceIV random 95 CI IV random 95 CI

minus2minus4 0 2 4

Huang et al 2013

Figure 2 Random effects standardized mean difference (SMD) for the association of miR-155 expression level and breast cancer BC breastcancer and NBC nonbreast cancerThe central of the square means the study-specific SMD and the horizontal lines correspond to the study-specific 95 CI

095710352

01109

minus0707

minus0444

31001908

397059

9618218

1202

019305230298

128406

000635

262019

201125112

3056

0054

103840

103840

45

83

067570149

00076

29minus1966

01133038

0122

120107

000461

023

00162

143

55

763

61

01 003

104066

0164

0911

165103

45

2125148023

minus087

10

170 [065 276]

045 [001 089]

099 [045 153]

049 [004 094]

099 [015 184]

073 [049 097]039 [006 072]056 [041 071]

065 [023 108]063 [039 087]

058 [046 071]

22

1000442


Total (95 CI) 724


BC NBC


6330

63

12055

45

031 [minus002 064]

044 [minus002 089]

288 [minus033 609]

050 [minus018 117]




IV fixed 95 CI IV fixed 95 CI

minus2minus4 0 2 4

147680

87280

2202

142212Subtotal (95 CI)

Subtotal (95 CI) 512 300 720

54

7735

266144

145

Heterogeneity 1205942

= 788 df = 5 (P = 016) I

2

= 37


= 773 df = 5 (P = 017) I

2

= 35


= 1584 df = 11 (P = 015) I

2

= 31

612 Blood

611 Tissue

Test for subgroup differences 1205942

= 023 df = 1 (P = 063) I

2

= 0

Chen et al 2012Fu and Zhang 2011Hafez et al 2012

Han et al 2013Heneghan et al 2010Huang et al 2013

Iorio et al 2005Ouyang et al 2014

Mar-Aguilar et al 2013Shao et al 2013Sun et al 2012

Zheng et al 2012

Standard mean difference Standard mean difference

Figure 3 Fixed effects standardized mean difference (SMD) for the association of miR-155 expression level and subgroup breast cancer BCbreast cancer and NBC nonbreast cancerThe central of the square means the study-specific SMD and the horizontal lines correspond to thestudy-specific 95 CI

using the Eggerrsquos test and Beggerrsquos test resulting in 119875 gt 005It indicated that no significant publication bias existed In theanalysis of the homogeneity of the literatures that had beenincorporated we found that most had heterogeneity In thiscase we preferred random effects model

According to the results we found that the expressionlevel of miR-155 in breast cancer sample was greater thanthat in nonbreast cancer sample Results of most articlesreported are almost consistent Liu et al [34] showed in theirarticle that the expression level in breast cancer group was


Table 2 Analysis of miR-155 expression level and different subtype breast cancer

Comparisons 119873 Heterogeneity Modela Effect sizeStudies 119868

2 () 119875(119876) SMD 119875(119885)ER+ Versus ERminus 7 [16ndash18 20 22 27 29] 96 lt000001 R minus12 003PR+ Versus PRminus 7 [16ndash18 20 22 27 29] 97 lt000001 R minus185 001Her2+ Versus Her2minus 4 [20 22 25 29] 50 011 F 032 0007LNM+ Versus LNMminus 7 [15ndash18 20 22 29] 97 lt000001 R 262 00001TNM III Versus IIIIV 8 [15ndash18 22 26 28 29] 96 lt000001 R minus094 008TS 1 Versus 23 5 [16 18 20 22 29] 92 lt000001 R minus058 lt00001P53 WT Versus MT 3 [16 25 28] 65 006 R minus079 002119873 number of studies 119875(119876) 119875 value of119876 test for heterogeneity 119875(119885) 119875 value of119885 test for significant test LNM lymph node metastasis TS tumor size WT wildtype MT mutant typeaR random-effect model F fixed-effect model

0

05

1

15

2

minus4 minus2 0 2 4

SMD

SE(S

MD

)

Figure 4 Funnel plots of studies included in the overall analysis

32-fold higher than that in nonbreast cancer group Otherstudies also have shown the miR-155 expression in breastcancer sample are at least 2-fold higher than that in nonbreastcancer sample for example Hui et al [35] reported 24-foldoverexpression in breast cancer Sun et al [26] reported 294-fold overexpressionWhen analyzing the different expressionfolds of miR-155 in breast cancer samples relative to normalsamples of available data [6 17 26 34ndash37] we discoveredthat the expression folds of miR-155 in breast cancer sampleswere higher than that in normal samples with the averageexpression about 6 times folds In the subgroup analysis bysample type the associations between miR-155 expressionlevel and breast cancer showed to be significant in both tissueand blood sample Therefore the detection of the expressionlevel of miR-155 in blood can be used in diagnosing breastcancer as an auxiliary molecular marker

According to our analysis the higher expression level ofmiR-155 in breast cancer samples than that in normal sampleswas detected However the level in different clinical pathol-ogy breast cancer samples also showed to be inconsistent Sowe merged into available literatures for the analysis Pooledresults for different subtypes showed that the expression levelof miR-155 was not statistically significant associated withTNM-staging but the expression level of miR-155 in lymphnode metastasis positive group was significantly higher thanthat in lymph node metastasis negative group

ER and PR were proteins which played important rolesin the regulation of the growth and differentiation of breastcancer [38] According to our analysis significantly higherexpression level of miR-155 was detected in both ER- and PR-group

Gene p53 as a tumor suppressor is located on the shortarm of chromosome 17 It can inhibit cell transformation andactivity of cancer gene But mutant gene p53 can cause celltransformation unlimited cell growth and cancer Gene p53mutation may be the most important deterioration factor ofbreast cancer [39] In the study it showed that miR-155 wassignificantly overexpressed in breast cancer with p53 mutanttype compared with breast cancer with p53 wild type

Her-2 is a prooncogene which is also located on chromo-some 17 It is recognized to be one of the most closely relatedgene to breast cancer MiR-155 is overexpressed in Her-2+breast cancer compared with Her-2minus breast cancer

Although our meta-analysis represented a quantifiedsynthesis of all available studies some limitations shouldbe noticed First this meta-analysis was conducted basedon case-control studies which might encounter recall andselection bias Second in subgroup analyses by sample typeand subtype analyses the number of the studies was relativelysmall A further analysis in the subtype analysis could not beperformedThird lack of the original data of available studieslimited our further evaluation of potential associations

In conclusion we can demonstrate that miR-155 is one ofthe most significant altered miRNAs in breast cancer Andthe overexpression of miR-155 is not so related to TNMstage but it is closely related to lymph node metastasis p53status and hormone receptor status of breast cancer patientsMore sizable sample based clinical investigations should beconducted before miR-155 can be applied as an auxiliarydiagnostic biomarker

Conflict of Interests

The authors declare that there is no conflict of interestsregarding the publication of this paper

Authorsrsquo Contribution

Hui Zeng and Cheng Fang contribute equally to this work


Acknowledgments

Theproject is sponsored by SRF for ROCS SEMThis work issupported by National Natural Science Foundation of China(Grants nos 30873044 and 81272372 to XL) and KoreanNational Cancer Center (NCC)Grant (NCC-1210460 to SN)

References

[1] W C Hahn and R A Weinberg ldquoRules for making humantumor cellsrdquoThe New England Journal of Medicine vol 347 no20 pp 1593ndash1603 2002

[2] D P Bartel ldquoMicroRNAs genomics biogenesis mechanismand functionrdquo Cell vol 116 no 2 pp 281ndash297 2004

[3] G Meister and T Tuschl ldquoMechanisms of gene silencing bydouble-stranded RNArdquo Nature vol 431 no 7006 pp 343ndash3492004

[4] P D Zamore and B Haley ldquoRibo-gnome the big world of smallRNAsrdquo Science vol 309 no 5740 pp 1519ndash1524 2005

[5] G D Leva G A Calin and C M Croce ldquoMicroRNAsfundamental facts and involvement in human diseasesrdquo BirthDefects Research Part CmdashEmbryo Today Reviews vol 78 no 2pp 180ndash189 2006

[6] X Zhang L Zhen X Han J Shi X Qiu and W SongldquoExpression of three microRNAs in plasma of breast cancerpatientsrdquo Chinese Journal of Clinical Oncology vol 39 no 3 pp136ndash140 2012

[7] W Kong H Yang L He et al ldquoMicroRNA-155 is regu-lated by the transforming growth factor 120573Smad pathway andcontributes to epithelial cell plasticity by targeting RhoArdquoMolecular amp Cellular Biology vol 28 no 22 pp 6773ndash67842008

[8] S Nam X Long C Kwon S Kim and K P Nephew ldquoAnintegrative analysis of cellular contexts miRNAs and mRNAsreveals network clusters associated with antiestrogen-resistantbreast cancer cellsrdquo BMC Genomics vol 13 no 1 article 7322012

[9] J Ferlay F Bray P Pisani and D M ParkinGLOBOCAN 2002Cancer Incidence Mortality and Prevalence Worldwide IARCCancer Base No 5 Version 20 IARC Press Lyon France 2004

[10] W Cao XWang and J Li ldquoHereditary breast cancer in theHanChinese populationrdquo Journal of Epidemiology vol 23 no 2 pp75ndash84 2013

[11] D Atkins D Best P A Briss et al ldquoGrading quality of evidenceand strength of recommendationsrdquo British Medical Journal vol328 p 1490 2004

[12] J P T Higgins S G Thompson J J Deeks and D G AltmanldquoMeasuring inconsistency in meta-analysesrdquo British MedicalJournal vol 327 no 7414 pp 557ndash560 2003

[13] R DerSimonian and R Kacker ldquoRandom-effects model formeta-analysis of clinical trials an updaterdquo Contemporary Clini-cal Trials vol 28 no 2 pp 105ndash114 2007

[14] J P T Higgins and S GThompson ldquoQuantifying heterogeneityin ameta-analysisrdquo Statistics inMedicine vol 21 no 11 pp 1539ndash1558 2002

[15] J Chen B Wang and J Tang ldquoClinical significance ofMicoRNA-155 expression in human breast cancerrdquo Journal ofSurgical Oncology vol 106 no 3 pp 260ndash266 2012

[16] S M Fu and H W Zhang ldquoThe expression of miR-155 andmiR-21 in breast cancer tissuerdquo Chinese Journal of ExperimentalSurgery vol 28 article 1240 2011

[17] M M Hafez Z K Hassan A R N Zekri et al ldquoMicroRNAsand metastasis-related gene expression in egyptian breast can-cer patientsrdquo Asian Pacific Journal of Cancer Prevention vol 13no 2 pp 591ndash598 2012

[18] X D Han X Y Zhang L L Zhen et al ldquoCirculatingmiR-155 inserumof patients with breast cancerrdquoChinese Journal of SurgicalOncology vol 5 no 1 pp 47ndash49 2013

[19] H M Heneghan N Miller R Kelly J Newell and M JKerin ldquoSystemic miRNA-195 differentiates breast cancer fromother malignancies and is a potential biomarker for detectingnoninvasive and early stage diseaserdquo Oncologist vol 15 no 7pp 673ndash682 2010

[20] G L Huang X Y Xue R C Zeng et al ldquoExpression ofmicroRNAs in plasma of patients with breast cancer and itsclinical significancerdquo Journal of Wenzhou Medical College vol43 no 5 pp 286ndash290 2013

[21] M V Iorio M Ferracin C Liu et al ldquoMicroRNA gene expres-sion deregulation in human breast cancerrdquoCancer Research vol65 no 16 pp 7065ndash7070 2005

[22] Z Lu Y Ye D Jiao J Qiao S Cui and Z Liu ldquoMiR-155 andmiR-31 are differentially expressed in breast cancer patientsand are correlated with the estrogen receptor and progesteronereceptor statusrdquo Oncology Letters vol 4 no 5 pp 1027ndash10322012

[23] M Ouyang Y Li and S Ye ldquoMicroRNA profiling implies newmarkers of chemoresistance of triple-negative breast cancerrdquoPlos ONE vol 9 no 5 Article ID e96228 2014

[24] F Mar-Aguilar J A Mendoza-Ramırez I Malagon-Santiago etal ldquoSerum circulating microRNA profiling for identification ofpotential breast cancer biomarkersrdquo Disease Markers vol 34no 3 pp 163ndash169 2013

[25] Y B Shao S Zhang Y Liu et al ldquoAnalysis of serum microR-NAs differentially expressed in Breast cancer tissuesrdquo ChineseGeneral Practice vol 16 no 5 pp 1736ndash1739 2013

[26] Y Sun M Wang G Lin et al ldquoSerum microRNA-155 as apotential biomarker to track disease in breast cancerrdquo PLoSONE vol 7 no 10 Article ID e47003 2012

[27] F Wang Z Zheng J Guo and X Ding ldquoCorrelation andquantitation of microRNA aberrant expression in tissues andsera from patients with breast tumorrdquo Gynecologic Oncologyvol 119 no 3 pp 586ndash593 2010

[28] X Q Wang and J Zhang ldquoAbnormal expression of miR-155 inserum of breast cancer patientsrdquo Tian Jin Medical College 2011

[29] S Zheng G GuoW Zhang et al ldquoClinical significance of miR-155 expression in breast cancer and effects of miR-155 ASO oncell viability and apoptosisrdquoOncology Reports vol 27 no 4 pp1149ndash1155 2012

[30] E A C Wiemer ldquoThe role of microRNAs in cancer no smallmatterrdquo European Journal of Cancer vol 43 no 10 pp 1529ndash1544 2007

[31] L Silveri G Tilly J Vilotte and F Le Provost ldquoMicroRNAinvolvement in mammary gland development and breast can-cerrdquo Reproduction Nutrition Development vol 46 no 5 pp549ndash556 2006

[32] M L Si S Zhu H Wu Z Lu F Wu and Y-Y Mo ldquomiR-21-mediated tumor growthrdquo Oncogene vol 26 no 19 pp 2799ndash2803 2007

[33] J Zhu X Hu G Guo et al ldquoExpression and its clinicalsignificance of miR-155 in human primary breast cancerrdquoZhonghua Wai Ke Za Zhi vol 48 no 3 pp 205ndash208 2010


[34] J Liu Q Mao and Y Liu ldquoAnalysis of miR-205 and miR-155 expression in the blood of breast cancer patientsrdquo ChineseJournal of Cancer Research vol 25 pp 46ndash54 2013

[35] A BHuiW Shi P C Boutros et al ldquoRobust globalmicro-RNAprofiling with formalin-fixed paraffin-embedded breast cancertissuesrdquo Laboratory Investigation vol 89 no 5 pp 597ndash6062009

[36] G F Bian and A J Chen ldquoCorrelation study of the expressionof miR-21 andmiR-155 in the blood and tumor tissue of patientswith breast cancerrdquo Hainan Medical Journal vol 23 no 18 pp1ndash3 2012

[37] X D Huang X R Zhou H Li L Mao and F Liu ldquoTheanalysis of expression of miR-155 in Breast cancer tissue and therelationship between the expression and clinical Pathologicalfeaturesrdquo Chinese Journal of Cancer Prevention and Treatmentvol 18 no 22 pp 1783ndash1801 2011

[38] D C Allred J M Harvey M Berardo and G M ClarkldquoPrognostic and predictive factors in breast cancer by immuno-histochemical analysisrdquo Modern Pathology vol 11 no 2 pp155ndash168 1998

[39] T Norberg S Klaar G Karf H Nordgren L Holmberg andJ Bergh ldquoIncreased p53 mutation frequency during tumorprogressionmdashresults from a breast cancer cohortrdquo CancerResearch vol 61 no 22 pp 8317ndash8321 2001

Submit your manuscripts athttpwwwhindawicom

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Anatomy Research International

PeptidesInternational Journal of


Hindawi Publishing Corporation httpwwwhindawicom

International Journal of

Volume 2014

Zoology


Molecular Biology International

GenomicsInternational Journal of


The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014


BioinformaticsAdvances in

Marine BiologyJournal of



Signal TransductionJournal of


BioMed Research International

Evolutionary BiologyInternational Journal of



Biochemistry Research International

ArchaeaHindawi Publishing Corporationhttpwwwhindawicom Volume 2014


Genetics Research International


Advances in

Virolog y

Hindawi Publishing Corporationhttpwwwhindawicom

Nucleic AcidsJournal of

Volume 2014

Stem CellsInternational



Enzyme Research



Microbiology


[9] In Chinese population breast cancer has an incidencerate of 1639 per 100000 Chinese women and seriously affectspeoplersquos lives and health Among cancers that happened inChinese women of economically developed provinces andcities breast cancer has the highest incidence and is thefourth most common cause of cancer death [10] Recentlythere were many investigations referring to the relationshipbetween miR-155 and breast cancer but because of the smallsample size discrepancy exists among different studies moreor less Therefore in order to provide a more considerableclearer and more systematic recognition of the expression ofmiR-155 in breast cancer we collected the data related tomiR-155 expression in breast cancer to carry out themeta-analysis

2 Materials and Methods

21 Search Strategy The literature retrieval was performedby two independent reviewers (Hui Zeng and Cheng Fang)All studies included in the meta-analysis were selected bysearching the PubMed EMBASE CNKI and VIP databasesup to May 2014 using the following keywords ldquomiR-155rdquo orldquomicroRNA-155rdquo and ldquobreast cancerrdquo All references in thesestudies were examined to identify additional research thatwas not indexed by the databases We selected publishedarticles written in English or Chinese

22 Including and Excluding Criteria Including criteriainclude (1) being related to breast cancer and miR-155 (2)patients that are confirmed by pathology (3) sufficient dataincluding mean value and standard deviation or other datathat can result in mean value and standard deviation and(4) measurementmethods and experiment group that are thesame or almost same

Excluding criteria include (1) reviews comments orletters (2) low-quality or incomplete data (3) abstract onlyand lack of the full text without authorrsquos reply and (4)reduplicate publication For duplicate articles only the mostrecent or largest data set was selected

23 Data Extraction Data were extracted from eligible arti-cles independently by two of the authors with any disagree-ment resolved by consensus The following information wascollected in a predefined data collection form first authorrsquosname publication year country sample type total numberof cases and controls quantitative methods and publicationlanguage

24 Literature Quality Evaluation Based on the results ofthe system we used grading method [11] recommended bythe GRADE system to evaluate quality of evidence Evidencequality classification is as follows (I) high quality furtherresearch would not change the credibility of evaluationresults about the curative effect (II) medium quality furtherresearch is likely to affect the credibility of evaluation resultsabout the curative effect and may change the assessmentresults (III) low quality further research is likely to affect thecredibility of evaluation results about the curative effect andthe assessment results are very likely to change (IV) very lowquality any curative effect evaluation results are uncertain

25 Statistic Analysis Rev-Man 50 software which wasrecommended by Cochrane collaboration was used in thismeta-analysis Rev-Man which is short for review manageris the software used for preparing and maintaining Cochranereviews Results can be presented graphically with the soft-ware

Firstly heterogeneity between studies was assessed by 1205942-based 119876-tests and 1198682 tests where 1198682 () gt 50 or 119875 lt 010was considered significantly heterogeneous [12]The randomeffect model (DerSimonian-Laird) [13] was used to assesspooled odds ratios (ORs) when significant heterogeneitywas observed Otherwise the fixed effect model (Mantel-Haenszel) [14] was used Sensitivity analysis was used toanalyze the stability of the text results by omitting one studyat a time For continuous data if quantitative method is thesame we adopted the weighted mean difference (MD) as ouranalysis index If they used different measuring instrumentsor units for the same variable or there was large differenceamong the mean value of numerical analysis the standard-ized mean difference (SMD) was adopted for analysis Wecalculated 95 confidence interval (CI) of all analysis At thesame time the funnel chart is used to determine publicationbias Data input and monitor were done by two researchers

3 Results

31 Study Characteristics For the initial inspection 65related English articles and 53 Chinese articles were obtainedby literature search from the PubMed EMBASE CNKIand VIP databases After titles and abstracts were screened86 articles were excluded because of irrelevant or dupli-cate records The full texts of the remaining 32 recordswere carefully reviewed Among these articles seven arti-cles were abandoned for overlapped data and five articleswere excluded because of review papers Another five wereexcluded due to data that were incomplete or inappropriatelycalculated or original data that could not be obtained despiteattempts to contact the authors Therefore 15 [15ndash29] articleswere considered in the present meta-analysis Among them13 discussed the different expression level ofmiR-155 betweenbreast cancer samples and normal samples and another 2only discussed it between different subtype breast cancersOne [27] was included in ER and PR analysis and another[28] was included in TNM and p53 analysis The specificretrieval process was shown in Figure 1 The study character-istics included in the meta-analysis were listed in Table 1

32 Results of Meta-Analysis A total of 13 studies including791 breast cancer samples and 509 normal samples werecollected in this sectionAs shown in Figure 2 we observed anelevated miR-155 expression in breast cancer samples (SMD= 122 95 CI = 065ndash178 119875 lt 000001)

However high heterogeneitywas observed in the analysisand sensitivity analyses indicated that the study from Lu et al[22] was mainly responsible for the observed heterogeneityWhen we excluded this study the high heterogeneity was sig-nificantly decreased and the association was still significantlydifferent (for SMD = 058 95 CI = 046ndash071 119875 lt 000001119875 for heterogeneity = 015 1198682 = 31) Sensitivity analyses















4 Discussion



Chen et al 2012 095710352

01109minus0707

minus0444

31001908

3970006349

0599618218

1202

0193052302981284

06000635

2620000875

019201125

11230560054

1038404583

067570149

00076

29minus1966

01133038

01221201

07000461

02300162

1435576673

6101 003

000018750001401

104066

01640911

16510345

2125148023minus087

10 68 170 [065 276]

045 [001 089]099 [045 153]

049 [004 094]099 [015 184]778 [677 878]

073 [049 097]039 [006 072]065 [023 108]

122 [065 178]

384022

85

858788

85

857570

79

8783

23

1000509




Total (95 CI) 791


BC NBC


63306

673

101205545

031 [minus002 064]

044 [minus002 089]

288 [minus033 609]050 [minus018 117]

Heterogeneity 1205912 = 093 120594

2

= 21020 df = 12 (P lt 000001) I

2

= 94



minus2minus4 0 2 4

Huang et al 2013


095710352

01109

minus0707

minus0444

31001908

397059

9618218

1202

019305230298

128406

000635

262019

201125112

3056

0054

103840

103840

45

83

067570149

00076

29minus1966

01133038

0122

120107

000461

023

00162

143

55

763

61

01 003

104066

0164

0911

165103

45

2125148023

minus087

10

170 [065 276]

045 [001 089]

099 [045 153]

049 [004 094]

099 [015 184]

073 [049 097]039 [006 072]056 [041 071]

065 [023 108]063 [039 087]

058 [046 071]

22

1000442


Total (95 CI) 724


BC NBC


6330

63

12055

45

031 [minus002 064]

044 [minus002 089]

288 [minus033 609]

050 [minus018 117]





minus2minus4 0 2 4

147680

87280

2202


Subtotal (95 CI) 512 300 720

54

7735

266144

145


= 788 df = 5 (P = 016) I

2

= 37


= 773 df = 5 (P = 017) I

2

= 35


= 1584 df = 11 (P = 015) I

2

= 31

612 Blood

611 Tissue


= 023 df = 1 (P = 063) I

2

= 0





Zheng et al 2012









0

05

1

15

2

minus4 minus2 0 2 4

SMD

SE(S

MD

)














Acknowledgments


References
















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology















4 Discussion



Chen et al 2012 095710352

01109minus0707

minus0444

31001908

3970006349

0599618218

1202

0193052302981284

06000635

2620000875

019201125

11230560054

1038404583

067570149

00076

29minus1966

01133038

01221201

07000461

02300162

1435576673

6101 003

000018750001401

104066

01640911

16510345

2125148023minus087

10 68 170 [065 276]

045 [001 089]099 [045 153]

049 [004 094]099 [015 184]778 [677 878]

073 [049 097]039 [006 072]065 [023 108]

122 [065 178]

384022

85

858788

85

857570

79

8783

23

1000509




Total (95 CI) 791


BC NBC


63306

673

101205545

031 [minus002 064]

044 [minus002 089]

288 [minus033 609]050 [minus018 117]

Heterogeneity 1205912 = 093 120594

2

= 21020 df = 12 (P lt 000001) I

2

= 94



minus2minus4 0 2 4

Huang et al 2013


095710352

01109

minus0707

minus0444

31001908

397059

9618218

1202

019305230298

128406

000635

262019

201125112

3056

0054

103840

103840

45

83

067570149

00076

29minus1966

01133038

0122

120107

000461

023

00162

143

55

763

61

01 003

104066

0164

0911

165103

45

2125148023

minus087

10

170 [065 276]

045 [001 089]

099 [045 153]

049 [004 094]

099 [015 184]

073 [049 097]039 [006 072]056 [041 071]

065 [023 108]063 [039 087]

058 [046 071]

22

1000442


Total (95 CI) 724


BC NBC


6330

63

12055

45

031 [minus002 064]

044 [minus002 089]

288 [minus033 609]

050 [minus018 117]





minus2minus4 0 2 4

147680

87280

2202


Subtotal (95 CI) 512 300 720

54

7735

266144

145


= 788 df = 5 (P = 016) I

2

= 37


= 773 df = 5 (P = 017) I

2

= 35


= 1584 df = 11 (P = 015) I

2

= 31

612 Blood

611 Tissue


= 023 df = 1 (P = 063) I

2

= 0





Zheng et al 2012









0

05

1

15

2

minus4 minus2 0 2 4

SMD

SE(S

MD

)














Acknowledgments


References
















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


Chen et al 2012 095710352

01109minus0707

minus0444

31001908

3970006349

0599618218

1202

0193052302981284

06000635

2620000875

019201125

11230560054

1038404583

067570149

00076

29minus1966

01133038

01221201

07000461

02300162

1435576673

6101 003

000018750001401

104066

01640911

16510345

2125148023minus087

10 68 170 [065 276]

045 [001 089]099 [045 153]

049 [004 094]099 [015 184]778 [677 878]

073 [049 097]039 [006 072]065 [023 108]

122 [065 178]

384022

85

858788

85

857570

79

8783

23

1000509




Total (95 CI) 791


BC NBC


63306

673

101205545

031 [minus002 064]

044 [minus002 089]

288 [minus033 609]050 [minus018 117]

Heterogeneity 1205912 = 093 120594

2

= 21020 df = 12 (P lt 000001) I

2

= 94



minus2minus4 0 2 4

Huang et al 2013


095710352

01109

minus0707

minus0444

31001908

397059

9618218

1202

019305230298

128406

000635

262019

201125112

3056

0054

103840

103840

45

83

067570149

00076

29minus1966

01133038

0122

120107

000461

023

00162

143

55

763

61

01 003

104066

0164

0911

165103

45

2125148023

minus087

10

170 [065 276]

045 [001 089]

099 [045 153]

049 [004 094]

099 [015 184]

073 [049 097]039 [006 072]056 [041 071]

065 [023 108]063 [039 087]

058 [046 071]

22

1000442


Total (95 CI) 724


BC NBC


6330

63

12055

45

031 [minus002 064]

044 [minus002 089]

288 [minus033 609]

050 [minus018 117]





minus2minus4 0 2 4

147680

87280

2202


Subtotal (95 CI) 512 300 720

54

7735

266144

145


= 788 df = 5 (P = 016) I

2

= 37


= 773 df = 5 (P = 017) I

2

= 35


= 1584 df = 11 (P = 015) I

2

= 31

612 Blood

611 Tissue


= 023 df = 1 (P = 063) I

2

= 0





Zheng et al 2012









0

05

1

15

2

minus4 minus2 0 2 4

SMD

SE(S

MD

)














Acknowledgments


References
















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology





0

05

1

15

2

minus4 minus2 0 2 4

SMD

SE(S

MD

)














Acknowledgments


References
















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology


Acknowledgments


References
















































Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology















Volume 2014

Zoology






















Advances in

Virolog y



Volume 2014




Enzyme Research



Microbiology

research article the clinicopathological...

Documents