Expression, Mutation, and Amplification Status of EGFR...

Research ArticleExpression Mutation and AmplificationStatus of EGFR and Its Correlation with Five miRNAsin Salivary Gland Tumours

Emanuela BoštjanIiI1 Nina Hauptman1 Aleš Grošelj2

Damjan GlavaI1 andMetka Volavšek3

1Department of Molecular Genetics Institute of Pathology Faculty of Medicine University of LjubljanaZaloska Cesta 4 1000 Ljubljana Slovenia2Department of Otorhinolaryngology and Cervicofacial Surgery University Medical Centre LjubljanaZaloska Cesta 2 1000 Ljubljana Slovenia3Institute of Pathology Faculty of Medicine University of Ljubljana Korytkova 2 1000 Ljubljana Slovenia

Correspondence should be addressed to Metka Volavsek metkavolavsekmfuni-ljsi

Received 14 October 2016 Revised 9 January 2017 Accepted 23 January 2017 Published 9 March 2017

Academic Editor Gang Liu

Copyright copy 2017 Emanuela Bostjancic et al This is an open access article distributed under the Creative Commons AttributionLicense which permits unrestricted use distribution and reproduction in any medium provided the original work is properlycited

Malignant salivary gland tumours are rare histologically and clinically heterogeneous group of tumours missing prognostic factorsand therapeutic targets MicroRNAs (miRNAs) small noncoding RNAs and posttranscriptional regulators of mRNA are poorlydescribed in different subtypes of salivary gland tumours Epidermal growth factor receptor (EGFR) an important therapeutictarget and target of certain miRNAs (ie miR-133b) shows variable degrees of expression in salivary gland tumours Our studyincluded 70 parotid gland tumours of different histological subtypes Expression mutations and copy number variations (CNVs)of EGFR were determined using immunohistochemistry single-stranded conformation polymorphism quantitative polymerasechain reaction (qPCR) and fluorescence in situ hybridization Expression of miR-99b miR-133b miR-140 miR-140-3p and let-7awas analysed using qPCR Expression of EGFR was observed in 37 of tumours with low and 40 of tumours with high malignantpotential There were no mutations with the majority of samples showing polysomy of chromosome 7 Based on histologicalsubtypes we found differential expression of all five miRNAs We confirmed association of reactivity of EGFR miR-133b miR-140 miR-140-3p and let-7a with CNV of EGFR and a positive association between miR-133blet-7a and reactivity of EGFR Ageand need for postoperative radiotherapy were characterized as significant in multivariate survival analysis

1 Introduction

Malignant salivary gland tumours are rare tumours with anannual incidence of 04ndash26100000 and represent sim6 of allhead and neck tumours [1] This is a very heterogeneousgroup of tumours in terms of histological subtypes and div-erse clinical behaviour and prognoses [2] The primary treat-ment option in localized disease is surgery with decision onpostoperative radiotherapy depending upon histological typeand tumour stage Since in some patients relapsing ormetastatic disease develops there is a need for new therapeu-tic strategies [3]

Abnormalities within human genome that might con-tribute to a wide range of cancer types may include small-and large-scale genomic alterations Oncogenes and tumour-suppressors including protein-coding genes [4] as wellas noncoding RNAs [5] can be affected by numerousgenetic alterations These alterations include chromosomaltranslocations and rearrangements copy number variationsmutations within genes expansion of nucleotides andsingle-nucleotide polymorphisms Recently chromosomerearrangements in diagnostic pathology of salivary glandmalignancies have been summarized [6] However thereare still unexplained cases missing prognostic factors and

HindawiBioMed Research InternationalVolume 2017 Article ID 9150402 11 pageshttpsdoiorg10115520179150402

2 BioMed Research International

therapeutic targets that have to be defined in salivary glandtumour pathology

Epidermal growth factor receptor (EGFR) is a trans-membrane tyrosine kinase an oncogene that is often con-tinuously activated and consequently activates a series ofintracellular signalling cascades Activation of cascades oftenaffects gene transcription resulting in angiogenesis celldifferentiation proliferation and tumour progression ofmany cancers Overexpression of EGFR is usually due toamplification of gene or activatingmutations within the geneTherefore in non-small cell lung cancer and colorectal cancerit is important also as a therapeutic target where in caseof mutation either tyrosine kinase inhibitors or monoclonalantibodies against EGFR are used in routine practice [2]Several experiments performed were describing expressionof EGFR in salivary gland tumours however with variabledegrees of expression being reported so far from 17 to 100of cases depending on histological subtypes [1 3]

Other important players in tumorigenesis are miRNAssmall noncoding RNAs In associationwith proteinsmiRNAsform miRNA-ribonucleoprotein complex bind to 3-UTR oftarget mRNA and thus sufficiently block translation miRNAmay be ubiquitously expressed or might possess cell- ortissue-specific expression pattern miRNAs thus function innumerous biological processes from development to adultlife including proliferation differentiation division of stemcells apoptosis and epithelial-to-mesenchymal transition(EMT) [7 8] Studying cancer phenotype it was also shownthat approximately sim50 of the miRNA genes are encodedwithin sites that are prone to cancer-associated rearrange-ments [9] Genomic abnormalities often cause abnormalmiRNA expression profiles Numerous expressions profilingusing microarrays and next gene sequencing approachesof miRNA analysis suggests that miRNAs are associatedwith various types of cancer [10] including salivary glandtumours One of the first studies described expression ofDicer a miRNA maturation enzyme in mucoepidermoidcarcinoma (MEC) a subtype of a malignant salivary glandtumour Abnormal immunoexpression of Dicer suggesteda role for miRNAs in tumour progression of MEC [11]Another study described alterations in miRNA expressionin pleomorphic adenomas (PAs) the most frequent benignsalivary gland tumoursmiR-140 andmiR-99bwere describedamong mostly upregulated miRNAs and let-7a was mostlydownregulated [12] miR-133b among other miRNAs targetEGFR in different cancer types especially non-small cell lungcancer [13 14]

Taken together aims of this study were (i) to investigateexpression of EGFR in different subtypes of salivary glandtumours (ii) to define whether overexpression is correlatedwith the copy number variation of EGFR (amplification ofEGFR or polysomy of chromosome 7) or mutation withinEGFR gene (iii) to analyse whether expression of miR-99bmiR-140 and let-7a is different between different salivarygland carcinoma subtypes and finally (iv) to test whetherexpression of miR-133b is in correlation with the expressionof their target EGFR also in salivary gland carcinoma

2 Materials and Methods

21 Patients and Tissue Samples The study comprised 70patients with carcinomas of the parotid gland diagnosed

at the Institute of Pathology Faculty of Medicine at theUniversity of Ljubljana between 1995 and 2008 At diag-nosis patients were staged according to the 6th edition ofUICCTNM system [15] Tumour surgery was performed atthe Department for Otorhinolaryngology and CervicofacialSurgery at University Clinical Centre Ljubljana

Haematoxylin-eosin (HE) slides from formalin-fixedparaffin-embedded (FFPE) tumours were available for allcases and reviewed by pathologist who chose representativetissue blocks for further analysis

For all samples histological subtypes were determinedAdditionally eleven normal salivary glands were included astissue controls All the tumour samples contained more than50 of tumour tissue with majority of samples between 70and 80

The study was approved by Republic of Slovenia NationalMedical Ethics Committee (number 770413)

22 Immunohistochemistry FFPE tissue samples were cut at4 120583m for immunohistochemistry (IHC) All reagents werefrom Ventana Medical Systems Inc Tucson USA exceptotherwise indicated We used commercially available anti-bodies against EGFR (CONFIRM anti-EGFR clone 3C6)Deparaffinization antigen retrieval (including 8 minutes ofProtease 1) and staining with antibodies were performedin an automatic immunostainer (Benchmark XT) in combi-nation with treating sections with secondary antibody andcolour developmentwith horseradish peroxidase (ultraVIEWDAB Detection Kit) The sections were then counterstainedwith haematoxylin Tissue sections of normal salivary glandsserved as positive and sections treated without primaryantibodies as negative controls

The immunostaining of EGFR was semiquantitativelyevaluated based on intensity of membrane reactivity with athreshold of 10 immune-positive cells [1] 0 negative (noreactivity or reactivity in lt10 of cells) 1+ weak reactivityin gt10 of cells 2+ moderate reactivity in gt10 cells and3+ strong reactivity in gt10 cells Tumours with 3+ stainingwere considered as positive Immunostaining pattern wasalso documented in normal salivary gland tissues as controlsAll tissue samples were stained and analysed in duplicate

23 DNA Isolation Tissue samples were cut at 10 120583m fromFFPE tissue blocks and for the isolation procedure six toeight 10 120583m sections were used Total DNA isolation wasperformed using QIAamp DNA FFPE Tissue Kit (Qiagen)according to the manufacturerrsquos protocol The DNA waseluted in 60 120583l of elution buffer The yield was measuredfluorescently using Quant-It (Invitrogen) according to man-ufacturer instruction and Rotor Gene Q (Qiagen)

24 Copy Number Variation (CNV) The copy numbers forEGFR were determined using commercially available andpredesigned TaqMan Copy Number Assays according tothe manufacturerrsquos instructions (Applied Biosystems) Allreagents were from Applied Biosystems except otherwiseindicated The primer ID used for EGFR was as followsEGFR1 Hs00997424 cn [16] The RNaseP and TERT locuswere used as the Copy Number Reference Assays Human

BioMed Research International 3

Table 1 Sequences of primer pairs used for PCR

Exon Primer sequence Genomics position18 51015840-CATGGTGAGGGCTGAGGTGA-31015840 chr7 55173871ndash55173890

51015840-GTCCCTGGCACAGGCCTCTGG-31015840 chr7 55174049ndash5517406919 51015840-ATGTGGCACCATCTCACAATTGCC-31015840 chr7 55174667ndash55174690

51015840-AAAGGTGGGCCTGAGGTTCAGA-31015840 chr7 55174856ndash5517487720a 51015840-AAGCCACACTGACGTGCCTCT-31015840 chr7 55181260ndash55181280

51015840-ACCGTGCAGCTCATCACGC-31015840 Chr7 55181362ndash5518138020b 51015840-CCTCCACCGTGCACCTCATC-31015840 chr7 55181357ndash55181377

51015840-CCCGTATCTCCCTTCCCTGA-31015840 chr7 55181483ndash5518150221 51015840-GGATGCAGAGCTTCTTCCCATGAT-31015840 chr7 55191659ndash55191682

51015840-AAATGCTGGCTGACCTAAAGCCAC-31015840 chr7 55191883ndash55191906

Genomic Control DNA (10 ng120583l) was used as a normalcontrol (calibrator sample) Prior to CNV PCR reactionefficiency was tested for all three genes as well as for HumanGenomic Control DNA and pooled FFPE samples Real-timegenomic PCRwas performed as duplex PCR in a total volumeof 10120583l in eachwell containing 5120583l of TaqManUniversal PCRmaster mix 10 ng of genomic DNA and 05 120583l of each pair ofprimers (20x TaqManCopyNumber Assay for EGFR and 20xTaqMan Copy Number Assay for either TERT or RNaseP)Each reaction was performed in duplicate using Rotor GeneQ (Qiagen)The PCR conditions were 95∘C for 10min and 40cycles of 95∘C for 15 s and 60∘C for 1min Data were analysedaccording to Applied Biosystems recommendation Sinceefficiencies were comparable each replicate was normalizedto RNAaseP or TERT to obtain aΔCt (CtTERTRNasePminusCtEGFR)and then averaged for each sample All samples were thennormalized to a calibrator sample to obtain ΔΔCt Copynumber was calculated as follows 2xRQ where RQ standsfor relative quantity (2minusΔΔCt)

25 Fluorescent In Situ Hybridization (FISH) For FISH anal-ysis 20 samples were randomly selected to compare to resultsfrom Copy number variation 1-2 120583m sections were mountedon charged slides (Thermo Scientific Super Frost Glass) HEslides were used for reference histology For FISH analysisreagents fromVysis AbbotMolecular was used except other-wise indicated After deparaffinization using xylene (Merck)slides were incubated 20min in 02MHCl (Merck) followingwashing in water and 2xSSC and incubation in NaSCN for30min at 80∘C Prior to protease digestion for 30min at37∘C (Paraffin Pretreatment Kit II) washing in water and2xSSC was repeated After protease digestion and washingin 2xSSC slides were dried for 5min at 57∘C and incubatedfor 10min in 10 buffered formalin Following washing in2xSCC and water slides were incubated in fixative methanolacetic acid (4 1) and washed in 70 and 100 ethanolFluorescence in situ hybridization (FISH) was performedwith the use of directly labelled ZytoLight SPECEGFRCEN7dual colour probes (ZytoVision) according to manufacturerinstruction Codenaturation and hybridization were per-formed in ThermoBrite using following conditions 73∘C5min and 37∘C overnight After probe hybridization slideswere washed in 2xSSC01NP40 for 2min in 2xSSC03NP40

for 2 minutes at 73∘C and for 1min in 2xSSC01NP40Nuclei were counterstained with antifading 410158406-diamidino-2-phenylindole (DAPI) and were analysed using the EclipseE600 microscope (Nikon) Hybridization signals of 30ndash100nonoverlapped nuclei were manually counted on single cellbasis

For EGFR samples were grouped as normal disomy le2centromere signals in ge40 of cells low polysomytrisomyge3 centromere signals in ge40 of cells excluding caseswith high polysomy or gene amplification high polysomyge4 centromere signals in ge40 of cells excluding caseswith gene amplification and gene amplification ratio ofgenechromosome ge2 or clusters of probes (gt10 copies pertumour cell) in ge40 of cells

26 Polymerase Chain Reaction (PCR) Five pairs ofnucleotide primers were selected in order to obtainmolecular analysis of exons 18 19 20 and 21 of EGFR gene(Table 1)The primers were prepared with IDT tools and generunner program and made at Eurofins and Qiagen Operon

27 Single-Stranded Conformation Polymorphism (SSCP) Itwas performed on the thin (04mm) 6 or 8 nondenat-urating polyacrylamide gel with 26 cross-linking Priorto conformation analysis six microliters of PCR productswere mixed with 10 microliters of SSCP loading dye (95formamide 20mM EDTA 005 bromophenol blue 005xylene cyanol and 20mM NaOH) and heated to 90∘C for2min After loading on the polyacrylamide gel the sampleswere electrophoresed in a 1xTBE (TrisBorateEDTA) buffer ina cold room (4∘C) for 4-5 hours at 50W and silver stained

28 Sequence Analysis ABI Prism 310 Genetic Analyzer(Applied Biosystems) was used to analyse sequences Thereaction was performed in a volume of 10 microliters usingBigDye Terminator v11 Cycle Sequencing Kit (AppliedBiosystems) containing dideoxynucleotide triphosphateAfter reaction was completed the products were cleaned withsodium acetate method

29 RNA Isolation from FFPE Tissue Samples Tissue sampleswere cut at 10 120583m from FFPE tissue blocks and for theisolation procedure six to eight 10 120583m sections were used


Total RNA isolation was performed using miRNA easy FFPEKit (Qiagen) according to the manufacturerrsquos protocol TheRNA was eluted in 30 120583l of nuclease-free water The yieldwas measured spectrophotometrically using NanoDrop-1000(Thermo Scientific USA) and the quality was evaluated onBioanalyzer 2100 (Agilent Technologies USA)

210 Quantitative Real-Time PCR (qPCR) All the reagentswere from Qiagen except where otherwise indicated Quan-titative PCR (qPCR) was carried out using the Rotor Gene QReal-Time PCR System

Prior to qPCR analysis two pools of RNA samples werecreated from FFPE tissue samples (salivary gland tumour andnormal salivary gland tissue) After reverse transcription ofpooled RNA and reference RNA the cDNA was diluted infive steps and the probes were tested for qPCR efficiency Allthe qPCR efficiency reactions were performed in triplicate

For reverse transcription the miScript reverse transcrip-tion kit was used Briefly a 10 120583l reaction master mix wasperformed containing 50 ng of total RNA 2 120583l 5x miScriptHiFlex RT buffer 1 120583l miScript reverse transcriptase mix 1120583lmiScript nucleic mix and 10 units (033 120583l) RNase inhibitorAfter incubation for 60min at 37∘C and 5min at 95∘Cthe cDNA was diluted 100-fold and 3 120583l was used for eachqPCR reaction Ten 120583l PCR master mix contained 5120583l 2xQuantiTect SYBR Green PCR Master Mix 1 120583l miScriptuniversal primer and 1120583l 10x miScript Primer Assay Asthe reference genes RNU5A RNU6B and SNORD25 wereused based on pretesting results All the qPCR reactionswere performed in duplicate The signal was collected atthe endpoint of every cycle Following amplification meltingcurves analysis of PCRproductswas performed to verify theirspecificity and identity Melting curves were acquired on theSYBR channel using a ramping rate of 07∘C60 s for 60ndash95∘CTested microRNAs were miR-99b miR-133b miR-140 miR-140-3p and let-7a To present a relative gene expression the2minusΔΔCt method was used in which the fold changes of thetested groups were calculated relative to the calibrator group[17 18]

211 Statistical Analysis For expression analysis of indepen-dent group of samples (ie salivary gland tumour normalsalivary gland) ΔCt was used for MannndashWhitney test Thedata were presented as a fold change in a graph with errorbars representing the calculated fold change error using theSD of the Ct triplicates

The expression level of EGFR was determined by IHCwhich was scored by multiplying the intensity (range 0ndash3) with the percentage of positive cells (range 0ndash100)Scored expression levels of EGFR expression of miRNAand copy number of EGFR were tested for correlation usingpolycor package in R (Drasgow 1986) Pearsonrsquos correlationcoefficient and 119901 values were calculated Polyserial corre-lation between ordinal variables (stage tumour size andpresenceabsence of metastasis) and the expression of EGFRmiRNAs and CNV of EGFR were calculated using thepolyserial function in the polycor library in R

Overall survival analysis was performed in software Rwith the latest data of patients obtained in year 2015 Survival

Table 2 Tumour characteristics at diagnosis

Characteristics (119899 = available data) Number ofpatients ()

Histological subtype (119899 = 70)Acinic cell carcinoma (ACCC) 16 (229)Mucoepidermoid carcinoma (MEC) 16 (229)Carcinoma ex pleomorphic adenoma

(Ca ex PA) 9 (129)

Adenoid cystic carcinoma (ACC) 8 (114)Poorly differentiated carcinoma 7 (100)Epithelial myoepithelial carcinoma

(EMC) 5 (71)

Adenocarcinoma NOS (ACNOS) 4 (57)Salivary duct carcinoma (SDC) 2 (29)Small cell carcinoma (SCC) 1 (14)Carcinosarcoma 1 (14)Adenosquamous carcinoma 1 (14)

Classification (119899 = 70)Favourable clinical prognosis 35 (500)Poor clinical prognosis 35 (500)

Tumour sizeT stage (119899 = 67)Tis 4 (60)T1 15 (224)T2 18 (269)T3 14 (209)T4 16 (239)

Lymph node statusN stage (119899 = 64)N0 45 (703)N1 5 (78)N2 14 (219)

Surgical resection (119899 = 63) 5963 (937)Postoperative radiotherapy (119899 = 63) 3963 (619)

was estimated using Cox proportional hazards model Astepwise Coxrsquos proportional hazards model was used tosimultaneously account for all potential prognostic factorsover time Hazard ratios and two-sided 95 confidenceinterval (95 CI) were estimated

3 Results

31 Patients All tumours were classified according tothe contemporary WHOrsquos classification of salivary glandtumours (Barnes et al 2005) The histological subtype oftumour samples is presented in Table 2 together with pro-gnostic classification and pTN stage

Among patients with salivary gland cancer there were39 men and 31 women aged 616 plusmn 139 and 620 plusmn 125respectively Among control samples there were 6 womenand 4 men aged 450 plusmn 151 and 555 plusmn 93 respectively

The tumours were divided into two groups (i) favourableclinical prognosis and (ii) poor clinical prognosis In agroup with favourable clinical prognosis there were 38


(a) (b)

Figure 1 EGFR expression in adenoid cystic carcinoma (ACC) in two different patients (a) Intense positivity in all cells of parotid glandACC (EGFR times20) (b) EGFR negative ACC cells surround and infiltrate the EGFR positive ductal structure of parotid gland (EGFR times20)

patients including histological subtypes of salivary glandtumours with lowmalignant potential acinic cell carcinomas(ACCCs) mucoepidermoid carcinoma (MEC) grades I andII epithelial myoepithelial carcinomas (EMCs) and carci-noma ex pleomorphic adenoma (Ca ex PA ) ldquoin siturdquo (Ca exPA = 4 cases) In a group with poor clinical prognosis therewere 32 patients including histological subtypes of salivarygland tumours with high malignant potential MEC gradeIII adenoid cystic carcinomas (ACCs) adenocarcinomaNOS (ACNOS) poorly differentiated carcinoma salivaryduct carcinomas (SDCs) carcinosarcoma adenosquamouscarcinoma invasive Ca ex PA (Ca ex PA = 5 cases) and smallcell carcinoma (SCC)

32 Immunohistochemical Expression of EGFR Immunohis-tochemical (positivenegative) expression of EGFR amongdifferent histological subtypes is presented in Table 3 Expres-sion of EGFR was found in 14 (37) samples amongfavourable clinical prognosis and in 13 (40) samples amongpoor clinical prognosis Examples of cases with positive andnegative expression are represented in Figure 1

33 Mutation Status and Copy Number Variation of EGFRIn salivary gland tumours we have not found any nucleotidechanges being responsible for activation of EGFR and itselevated expression pattern

According to results of copy number variation (CNV)analysis using qPCR we found in the majority of samplespolysomy of chromosome 7 There were similar resultsbetween groups with favourable and poor clinical prognosisThere were no sample with disomy 28 samples (80) withpolysomy and 5 samples (14) with amplification of EGFRin a group with poor clinical prognosis There were 1 samplewith disomy (3) 25 samples (71) with polysomy and 8samples (23) with amplification of EGFR in a group withfavourable clinical prognosis There was no statistically sig-nificant change in proportion of samples with disomy poly-somy or amplification between the histological subtypes ofpoor and favourable clinical prognosis

However in three samples CNV analysis using qPCRfailed to give the reliable results (1 ACC 1 SDC and 1 MEC)

FISH analysis of these three samples gave the following resultACC showed 80 of disomic cells 15 of cells with highpolysomy and 5 with amplification SDC showed 56 ofdisomic cells 24 of cells with low polysomy 16 of cellswith high polysomy and 4with amplificationMEC showed80 of disomic cells 8 of cells with low polysomy 4of cells with high polysomy and 8 with amplificationTherefore all three samples were treated as disomic

FISH analyses gave similar results as CNV using qPCRIn 3 samples (15) we observed positive result with amplifi-cation in one case and combined amplification and polysomyin two cases All other samples were negative with majorityshowing low polysomy and a small proportion with disomy

34 microRNA Expression Analysis Analysis of expressionof selected miRNA showed that among all salivary glandtumoursmiR-140was downregulated andmiR-133b andmiR-99b were upregulated when compared to normal salivaryglands (Figure 2(a)) When group with favourable clinicalprognosis was compared to normal salivary glands theexpression pattern of miR-133b was upregulated and expres-sion of miR-140 and let-7a was downregulated (Figure 2(b))Comparison of favourable versus poor clinical prognosisgroup yields no statistically significant difference in expres-sion of miRNAs

Further analysis of most common entities of salivarygland tumours revealed that miR-140 and let-7a are down-regulated in ACCCs Although the number of samples wassmall we also observed miR-140 downregulation in EMCs(Figure 2(c))

When group with poor clinical prognosis was comparedto group of normal salivary gland there were two differ-entially expressed miRNAs miR-133b and miR-99b whichwere both upregulated (Figure 2(b)) Further subdivisionaccording to the histological subtypes of tumours revealeddownregulation of miR-140 in ACCs Although the numberof samples was small we also observedmiR-99b andmiR-133bupregulation in ACNOS as well asmiR-140 and let-7a down-regulation in poorly differentiated carcinoma (Figure 2(d))

Although some of the samples of MEC and Ca ex PAwere included in group with favourable and some in group


Table 3 Number of samples with positive and negative EGFR expression pattern and copy number variation of EGFR among histologicalsubtypes of salivary tumours

Histological subtype EGFR 3+ ()

CNV EGFRAmplification 119899 ()Polysomy 119899 ()Disomy 119899 ()

MECs (119899 = 16) 8 (50)4 (250)11 (688)1 (63)

ACCCs (119899 = 16) 2 (111)3 (188)12 (75)1 (63)

Ca ex PA (119899 = 9) 6 (667)1 (111)8 (889)

0

ACCs (119899 = 8) 4 (50)2 (250)5 (625)1 (125)

Poorly differentiated carcinoma (119899 = 7) 3 (429)2 (286)5 (714)

0

EMCs (119899 = 5) 4 (800)0

5 (100)0

ACNOS (119899 = 4) 01 (250)3 (750)

0

SDCs (119899 = 2) 00

1 (50)1 (50)

SCC (119899 = 1) 00

1 (100)0

Carcinosarcoma (119899 = 1) 00

1 (100)0

Adenosquamous carcinoma (119899 = 1) 00

1 (100)0

Total 2770 (386)1370 (186)5370 (757)470 (57)

with poor clinical prognosis we observed miR-140 and let-7a downregulation and miR-133b upregulation in MEC andupregulation ofmiR-140miR-140-3pmiR-99b andmiR-133bin Ca ex PA (Figure 2(e))

All the values of fold changes are provided in supplemen-tary Table S1 (in Supplementary Material available online athttpsdoiorg10115520179150402)

35 Correlation of EGFR between Immunostaining Amplifica-tion Status Expression Analyses of miRNAs and Clinical andPathohistological Data First we were able to confirm veryweak but statistically significant association between CNV ofEGFR and IHC expression of EGFR We also observed weakpositive association between expression of let-7a miR-133bandmiR-140 and reactivity of EGFR as well as between CNV


minus400minus300minus200minus100

000100200300400

miR-99b miR-133b miR-140 miR-140-3p let-7a

Fold

chan

ge

MicroRNAsSalivary gland tumor

lowastlowast

lowast

lowast

(a)


050150250350450550


Fold

chan

ge

MicroRNAsPoor clinical prognosisFavourable clinical prognosis

lowastlowast

lowastlowast

lowastlowastlowast

(b)

minus650minus550minus450minus350minus250minus150minus050

050150250350


Fold

chan

ge

MicroRNAsACCCsEMCs

lowastlowast

lowast

lowast

(c)

minus1000

minus800

minus600

minus400

minus200

000

200

400

600


Fold

chan

ge

MicroRNA

ACCsPoorly differentiated CaACNOS

lowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

(d)

minus400

minus200

000

200

400

600

800

1000

1200


Fold

chan

ge

MicroRNAsMECsCa ex PA

lowast

lowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

(e)

Figure 2 Expression of selected miRNA in salivary gland tumours Fold changes of miRNAs (a) in salivary gland tumour compared tonormal salivary gland (b) in favourable and poor clinical prognosis compared to normal salivary gland (c) in ACCC and EMC comparedto normal salivary gland (d) in ACCs poorly differentiated Ca and ACNOS compared to normal salivary gland (e) in MEC and Ca exPA compared to normal salivary gland Significant expression is marked with asterisk (lowastlowast119901 lt 001 lowast119901 lt 005) bars represent the standarddeviation


Table 4 Correlationsassociations between CNV of EGFR and IHC expression of EGFR and microRNA expression Significantcorrelationsassociations are marked with asterisk (lowastlowast119901 lt 001 lowast119901 lt 005)

CNV of EGFR IHC of EGFR ΔCt miR-99b ΔCt miR-133b ΔCt miR-140 ΔCt miR-140-3p ΔCt let-7aCNV of EGFRIHC of EGFR 013lowastlowast

ΔCt (miR-99b) minus007lowastlowast

ΔCt (miR-133b) 011lowastlowast 012lowastlowast 042lowast

ΔCt (miR-140) 025lowast 003lowast 024lowastlowast

ΔCt (miR-140-3p) 027lowast 028lowast

ΔCt (let-7a) 021lowastlowast 031lowastlowast minus008lowast 050lowast 048lowast

Table 5 Association between tumour characteristics and overall survival Univariate analysis (Cox proportional hazards regression model)is shown in the left panels and multivariate analysis (Cox regression with stepwise backwards selection) in the right panels Statisticallysignificant119901 values in the univariate analysis and in the last step of themultivariate analysis are shown in boldHR hazard ratio CI confidenceinterval

Characteristics (119899 = available data) Univariate model Multivariate modelHR 95 CI 119901 value HR 95 CI 119901 value

Male (119899 = 70) 057 028ndash116 012Age gt 60 years (119899 = 70) 288 129ndash643 001 222 097ndash508 005Tumour size ge 2 cm (119899 = 63) 156 067ndash363 036Lymph node infiltration (119899 = 63) 184 125ndash273 0001Radiotherapy (119899 = 61) 360 148ndash876 0005 291 117ndash725 002Poor clinical prognosis (119899 = 70) 201 100ndash401 0049ICH EGFR (119899 = 70) 100 099ndash100 010CNV EGFR (119899 = 70) 093 080ndash108 035ΔCt miR-99b 105 080ndash138 071ΔCt miR-133b 093 082ndash106 027ΔCt miR-140 104 087ndash124 070ΔCt miR-140-3p 101 084ndash122 089ΔCt let-7a 101 076ndash134 093

of EGFR and almost all investigated miRNAs The strongestcorrelations were observed between the expression patternsof let-7a and expressions of miR-140 and miR-140-3p Allthese statistically significant associations and correlations aresummarized in Table 4

Unsurprisingly we confirmed a moderate positive asso-ciation between tumour sizestage and the number ofmetastases (119903119904 = 054 119901 lt 001) We have further obser-ved increasing expression of all miRNAs except miR-140with tumour sizeT stage (from T1 to T3) however theonly statistically different expression was observedfor miR-133b between T1 and T2 (119901 = 003) and accordingly miR-133b showed weak positive association with tumour size(119903119904 = 023 119901 = 005) Distribution of EGFR reactivity andpolysomyamplification was similar between different tum-our sizes

We have also observed increasing expression of miR-133b with the presence of metastases (from N0 to N1 andN2) however the difference in expression did not reachthe statistical significance Distribution of EGFR reactivityand polysomyamplification was similar independent of thepresence or absence of metastases

We did not observe any statistically significant changein expression between patients that survived and those that

died due to the presence of neoplasm More interestingly wedid observe the statistically significant higher percentage ofEGFR reactive samples (58) among patients that died dueto the presence of salivary glade tumours compared to thosethat were still alive (30) at the time of the time of analysis(119901 = 005)

36 Tumour and Patient Characteristics Associated with Over-all Survival The follow-up data was taken on September25 2015 Most of the patients have been followed for morethan 10 years and for all patients the follow-up period ismore than 5 years Results on univariate and multivariateanalyses are summarized in Table 5 In the univariate modelradiotherapy shows the strongest association with overallsurvival followed by age (gt60) poor clinical prognosis andlymph node infiltration Results are summarized in Figures3(a)ndash3(d) and Table 5 In the multivariate analysis the overallsurvival is associated with radiotherapy and age (over 60years at the time of diagnosis) Results are summarized inFigures 3(e) and 3(f) and Table 5 The measured molecularcharacteristics do not seem to have an effect on the overallsurvival of the patients


10

08

06

04

02

Cum

surv

ival

Survival function for patterns 1ndash3

RT RTNo RT

Overall survival200150100500

(a)

10

08

06

04

Cum

surv

ival

Survival function for patterns 1-2


0=le601=ge60

0

1

(b)

10

08

06

04

02

Cum

surv

ival

PoorClinical prognosis

Good



(c)

10

08

06

04

02

00

Cum

surv

ival



Metastases

AbsentN1

N2N3

(d)

10

08

06

04

02

Cum

surv

ival


RT RTNo RT


(e)

10

09

08

07

06

05

04

Cum

surv

ival



0=le601=ge60

0

1

(f)

Figure 3 Overall survival of patients based on Cox proportional hazards regressionmodel Univariate model with impact of (a) radiotherapy(blue line is without data on RT) (b) age (gt60 y at the time of diagnosis) (c) poor clinical prognosis and (d) lymph node infiltration on overallsurvival Multivariate model with impact of (e) radiotherapy and (f) age (gt60 y at the time of diagnosis) RT radiotherapy ge more than 60years le less than 60 years N1 N2 and N3 nodal status of TNM tumour classification no data of lymph node metastases

4 Discussion

One of the aims of our study was to investigate expression ofEGFR in different subtypes of salivary gland tumours In 27of 70 salivary gland tumours we found positive reactivity (3+IHC staining) of EGFR Compared to the study performed onACCs and non-ACCs tumours the percentage on ACCs washigher in our study (4 of 8) compared to 5 of 20 [3] howeverthe overall percentage of EGFR positive reaction was lower(27 of 70) compared to 16 of 39 On a larger cohort of salivarygland tumours [2] 134 of 663 salivary gland carcinomas werefound with 3+ EGFR reaction and 80 of 189 salivary glandadenomas with 3+ EGFR reaction [2]

In our studywewere able to associate immunohistochem-ical expression of EGFR with the copy number variation ofEGFR (amplification of EGFR or polysomy of chromosome

7) although we did not find any mutation being responsiblefor activation of EGFR and its elevated expression patternSimilar to previous research [3] we have observed amplifiedEGFR in only 2 of 8 ACCs and small proportion amongother tumour entities although half of samples stained 3+ byimmunohistochemistry However polysomy of chromosome7 was observed in the majority of the samples Similar toour results no samples with amplification and only polysomywere detected on large cohort of salivary gland tumours withonly 2 mutations found in 2 samples out of 107 [2]

Distribution of EGFR reactivity and polysomyamp-lification was similar between different tumour sizes andindependent of metastatic disease These results are in con-trast to the studies performed on larger cohort of patientswhere EGFR positivity was associated with tumour size [1]and lymph node metastases [1 19] However one of these


two studies similar to our results observed that copy numbergain predictedworse survival for the patients [1] Accordinglyour results showed statistically significant higher percentageof EGFR reactive samples (58) among patients that died dueto the presence of salivary gland tumours compared to thosethat were still alive (30) at the time of analysis

In the second part of the study we have analysed expres-sion of five different miRNAs miR-99b miR-133b miR-140miR-140-3p and let-7a since miR-99b miR-140 and let-7ahave been shown to be differently expressed in PA of salivarygland [12] The aim of our study was to analyse what is theirexpression in other salivary gland tumour subtypes We fur-ther tested whether expression of miR-133b is in correlationwith the expression of its target EGFR which was accord-ing to miRTarBase (httpmirtarbasembcnctuedutw adatabase of confirmed and validated miRNA targets) vali-dated using Reporter Assay Western blot and qPCR

We observed differences in expression patterns of miR-133b miR-140 and let-7a when group with poor clinicalprognosis was compared to normal salivary glands Whengroup with favourable clinical prognosis was compared togroup of normal salivary gland there were two differentiallyexpressed miRNAs miR-133b and miR-99b Further analysisof most common entities of salivary gland tumours revealeddownregulation of miR-140 in MECs ACCCs EMCs andACCs and in poorly differentiated carcinoma let-7a inMECsandACCCs and in poorly differentiated carcinoma andmiR-99b in poorly differentiated carcinoma We further observedupregulation of miR-133b in MEC and in ACNOS andmiR-99b in ACNOS Similar to a previous described studyon PAs [12] differentially expressed miRNAs in Ca ex PAappears to be upregulated miR-140 miR-140-3p miR-99band additionally miR-133b However downregulation of let-7a was not observed

In association of miRNAs expression with the immuno-histochemical expression of EGFR we observed positiveassociation between expression of miR-133b and let-7a andreactivity of EGFR Let-7a has been already implicated intumourigenesis and similarly as in MECs ACCCs andpoorly differentiated carcinoma of salivary glands it has beenshown to be reduced in lung cancer Its two targets appearto be RAS and HMGA2 oncogenic proteins in a varietyof tumours [12] It has been recently observed that WIF1(Wnt inhibitory factor 1) is downregulated in Ca ex PA andthat when overexpressed it increases expression of pri-let-7a a primary transcript of let-7a [20] miR-133b has beendownregulated in many tumours and upregulated only incervical carcinoma where it acts as an oncogene to promotetumorigenesis and metastasis Its targets are MST2 CDC42and RHOA which results in activation of AKT and ERKsignalling pathways [21] In research done on non-small-celllung cancer the expression ofmiR-133b also influenced EGFRsignalling pathway [14]

Additionally we have found positive association betweenmiR-133b miR-140 miR-140-3p and let-7a and CNV ofEGFR miR-140 is believed to be cartilage specific miRNAhowever it appears to be important in certain subtypesof salivary gland tumours and for differential diagnosis ofhistological subtypes of salivary gland tumours Pronounced

positivity of miR-140 in Ca ex PA which most probablyreflects the presence of remaining chondroid matrix ausual component of PA could therefore in the differentialdiagnostic process implicate a malignant transformation ofPA in comparison with a tumour arising ldquode novordquo wherewe would anticipate negative values of miR-140 Usingintersection of mRNAs repressed by miR-140 overexpressionand of mRNAs expressed by miR-140 silencing 49 geneswere identified as potential miR-140 targets with 22 of thesepossessing seed sequence [22] Therefore which of these isimplicated in salivary gland tumour is yet to be identified

We have also observed increasing expression of all miR-NAs except miR-140 with tumour size (from T1 to T3) how-ever the only statistically different expression was observedfor miR-133b between T1 and T2 and accordingly miR-133bshowed positive correlation with tumour size There was alsoan increasing expression of miR-133b with the presence ofmetastases (from N0 to N1 and N2) although the differencein expression did not reach the statistical significance

Overall survival analysis showed that statistically signif-icant parameters influencing overall survival are of clinicalnature In univariate analysis those factors are age above 60years at the time of the diagnosis lymph node metastasespoor clinical prognosis and radiotherapy Multivariate anal-ysis showed that radiotherapy and age above 60 years havesignificant impact on overall survival These results are simi-lar to those previously published although they also includedthe proliferative marker Ki67 [23] We did not observe anystatistically significant change in expression patterns betweenpatients that survived and those that died due to the presenceof neoplasm however we did observe upregulation of miR-133b in patients that died due to other reasons compared tothose that were alive at the time of analysis We were able toconfirm the correlation of expression of miR-133b with theimmunohistochemical expression of its target

5 Conclusions

In summary we observed expression of EGFR in 37 oftumours with lowmalignant potential and in 40 of tumourswith high malignant potential There were no mutations inEGFR in themajority of samples (76) polysomy of chromo-some 7 was detected and amplification was present in 19 ofsamples In demonstrated absence of EGFR gene mutationsin malignant salivary gland tumours we can conclude thatvariability in EGFR expression occurs as a consequence ofgene amplification and chromosome 7 polysomy Althoughno significant differences were observed among differenthistologic types further studies are needed to eliminateEGFR as a possible therapeutic target in malignant salivarygland tumours Further ldquonegativerdquo results would push us andother researchers toward identification of novel biomarkersand biological targets in salivary gland tumours

The expression of miR-99b and miR-133b was upregu-lated while miR-140 expression was downregulated in sali-vary gland tumours compared to normal salivary glandsmiR-133b was upregulated in poor and favourable clinicalprognosis compared to normal tissues additionally in thepoor clinical prognosis group upregulation was observed in


expression of miR-99b and in favourable clinical prognosisgroup the downregulation of miR-140 In all histologicalsubtypes miR-140 was significantly down-regulated exceptin Ca ex PA where it was significantly up-regulatedmiR-133bwas significantly upregulated in ACNOS Ca ex Pa andMECsubtypesmiR-99bwas also upregulated in ACNOS andCa exPA whilemiR-140-3p expressionwas only significant in Ca exPA Downregulated expression profile of let-7a was observedin ACCC MEC and poorly differentiated Ca

We observed positive correlation between EGFR reactiv-ity miR-133b miR-140 miR-140-3 and let-7a and CNV ofEGFR and a positive correlation between miR-133b and let-7a and reactivity of EGFR

Univariate overall survival analysis showed that agelymph node infiltration radiotherapy and clinical progno-sis have significant effect on survival while multivariateapproach yields age and radiotherapy as significant factorsThere was also a statistical difference in expression of miR-133b between T1 and T2 tumour size

Differential expression of various microRNAs in malig-nant salivary gland tumours especially upregulation ofmiR-140 appears to be important in the differential diagnosispointing to possible malignant transformation of PA

Competing Interests

The authors declare that they have no competing interests

Authorsrsquo Contributions

Emanuela Bostjancic and Nina Hauptman contributed equa-lly to this work

Acknowledgments

This work was supported by the Slovenian Research Agency(P3-054)

References

[1] T Ettl K Baader C Stiegler et al ldquoLoss of PTEN is associatedwith elevated EGFR and HER2 expression and worse prognosisin salivary gland cancerrdquo British Journal of Cancer vol 106 no4 pp 719ndash726 2012

[2] T S Clauditz A Gontarewicz P Lebok et al ldquoEpidermal gro-wth factor receptor (EGFR) in salivary gland carcinomaspotentials as therapeutic targetrdquo Oral Oncology vol 48 no 10pp 991ndash996 2012

[3] L Vidal M S Tsao G R Pond et al ldquoFluorescence in situ hyb-ridization gene amplification analysis of EGFR and HER2 inpatients with malignant salivary gland tumors treated withlapatinibrdquo Head and Neck vol 31 no 8 pp 1006ndash1012 2009

[4] BVogelstein andKWKinzler ldquoCancer genes and the pathwaysthey controlrdquoNature Medicine vol 10 no 8 pp 789ndash799 2004

[5] R P Alexander G Fang J RozowskyM Snyder andM B Ger-stein ldquoAnnotating non-coding regions of the genomerdquo NatureReviews Genetics vol 11 no 8 pp 559ndash571 2010

[6] R HW Simpson A Skalova S Di Palma and I Leivo ldquoRecentadvances in the diagnostic pathology of salivary carcinomasrdquoVirchows Archiv vol 465 no 4 pp 371ndash384 2014

[7] R W Carthew and E J Sontheimer ldquoOrigins and mechanismsof miRNAs and siRNAsrdquo Cell vol 136 no 4 pp 642ndash655 2009

[8] M Esteller ldquoNon-coding RNAs in human diseaserdquo Nature Rev-iews Genetics vol 12 no 12 pp 861ndash874 2011

[9] A EWilliams ldquoFunctional aspects of animalmicroRNAsrdquoCell-ular and Molecular Life Sciences vol 65 no 4 pp 545ndash5622008

[10] A Gaur D A Jewell Y Liang et al ldquoCharacterization of mic-roRNA expression levels and their biological correlates inhuman cancer cell linesrdquo Cancer Research vol 67 no 6 pp2456ndash2468 2007

[11] S I Chiosea E L Barnes S Y Lai et al ldquoMucoepidermoid car-cinoma of upper aerodigestive tract clinicopathologic study of78 cases with immunohistochemical analysis of Dicer expres-sionrdquo Virchows Archiv vol 452 no 6 pp 629ndash635 2008

[12] X Zhang M Cairns B Rose et al ldquoAlterations in miRNA pro-cessing and expression in pleomorphic adenomas of the salivaryglandrdquo International Journal of Cancer vol 124 no 12 pp 2855ndash2863 2009

[13] J Tao D Wu B Xu et al ldquomicroRNA-133 inhibits cell pro-liferation migration and invasion in prostate cancer cellsby targeting the epidermal growth factor receptorrdquo OncologyReports vol 27 no 6 pp 1967ndash1975 2012

[14] L Liu X ShaoW Gao et al ldquoMicroRNA-133b inhibits the gro-wth of non-small-cell lung cancer by targeting the epidermalgrowth factor receptorrdquo FEBS Journal vol 279 no 20 pp 3800ndash3812 2012

[15] LH Sobin CWittekind and InternationalUnion against Can-cer TNMClassification of Malignant Tumours Wiley-Liss NewYork NY USA 6th edition 2002

[16] H Kato T Arao K Matsumoto et al ldquoGene amplification ofEGFR HER2 FGFR2 and MET in esophageal squamous cellcarcinomardquo International Journal of Oncology vol 42 no 4 pp1151ndash1158 2013

[17] G Chen W Zhu D Shi et al ldquoMicroRNA-181a sensitizeshumanmalignant gliomaU87MG cells to radiation by targetingBcl-2rdquo Oncology Reports vol 23 no 4 pp 997ndash1003 2010

[18] G J Latham ldquoNormalization of microRNA quantitative RT-PCR data in reduced scale experimental designsrdquo Methods inMolecular Biology vol 667 pp 19ndash31 2010

[19] T Ettl M Gosau G Brockhoff et al ldquoPredictors of cervicallymphnodemetastasis in salivary gland cancerrdquoHead andNeckvol 36 no 4 pp 517ndash523 2014

[20] I Ramachandran VGanapathy E Gillies et al ldquoWnt inhibitoryfactor 1 suppresses cancer stemness and induces cellular senes-cencerdquo Cell Death amp Disease vol 5 Article ID e1246 2014

[21] W Qin P Dong C Ma et al ldquoMicroRNA-133b is a key pro-moter of cervical carcinoma development through the activa-tion of the ERK and AKT1 pathwaysrdquo Oncogene vol 31 no 36pp 4067ndash4075 2012

[22] F ENicolas H Pais F Schwach et al ldquoExperimental identifica-tion of microRNA-140 targets by silencing and overexpressingmiR-140rdquo RNA vol 14 no 12 pp 2513ndash2520 2008

[23] J Cros E Sbidian S Hans et al ldquoExpression and mutationalstatus of treatment-relevant targets and key oncogenes in 123malignant salivary gland tumoursrdquo Annals of Oncology vol 24no 10 pp 2624ndash2629 2013

Submit your manuscripts athttpswwwhindawicom

Stem CellsInternational

Hindawi Publishing Corporationhttpwwwhindawicom Volume 2014


MEDIATORSINFLAMMATION

of


Behavioural Neurology

EndocrinologyInternational Journal of



Disease Markers


BioMed Research International

OncologyJournal of



Oxidative Medicine and Cellular Longevity


PPAR Research

The Scientific World JournalHindawi Publishing Corporation httpwwwhindawicom Volume 2014

Immunology ResearchHindawi Publishing Corporationhttpwwwhindawicom Volume 2014

Journal of

ObesityJournal of



Computational and Mathematical Methods in Medicine

OphthalmologyJournal of


Diabetes ResearchJournal of



Research and TreatmentAIDS


Gastroenterology Research and Practice


Parkinsonrsquos Disease

Evidence-Based Complementary and Alternative Medicine

Volume 2014Hindawi Publishing Corporationhttpwwwhindawicom


therapeutic targets that have to be defined in salivary glandtumour pathology

Epidermal growth factor receptor (EGFR) is a trans-membrane tyrosine kinase an oncogene that is often con-tinuously activated and consequently activates a series ofintracellular signalling cascades Activation of cascades oftenaffects gene transcription resulting in angiogenesis celldifferentiation proliferation and tumour progression ofmany cancers Overexpression of EGFR is usually due toamplification of gene or activatingmutations within the geneTherefore in non-small cell lung cancer and colorectal cancerit is important also as a therapeutic target where in caseof mutation either tyrosine kinase inhibitors or monoclonalantibodies against EGFR are used in routine practice [2]Several experiments performed were describing expressionof EGFR in salivary gland tumours however with variabledegrees of expression being reported so far from 17 to 100of cases depending on histological subtypes [1 3]

Other important players in tumorigenesis are miRNAssmall noncoding RNAs In associationwith proteinsmiRNAsform miRNA-ribonucleoprotein complex bind to 3-UTR oftarget mRNA and thus sufficiently block translation miRNAmay be ubiquitously expressed or might possess cell- ortissue-specific expression pattern miRNAs thus function innumerous biological processes from development to adultlife including proliferation differentiation division of stemcells apoptosis and epithelial-to-mesenchymal transition(EMT) [7 8] Studying cancer phenotype it was also shownthat approximately sim50 of the miRNA genes are encodedwithin sites that are prone to cancer-associated rearrange-ments [9] Genomic abnormalities often cause abnormalmiRNA expression profiles Numerous expressions profilingusing microarrays and next gene sequencing approachesof miRNA analysis suggests that miRNAs are associatedwith various types of cancer [10] including salivary glandtumours One of the first studies described expression ofDicer a miRNA maturation enzyme in mucoepidermoidcarcinoma (MEC) a subtype of a malignant salivary glandtumour Abnormal immunoexpression of Dicer suggesteda role for miRNAs in tumour progression of MEC [11]Another study described alterations in miRNA expressionin pleomorphic adenomas (PAs) the most frequent benignsalivary gland tumoursmiR-140 andmiR-99bwere describedamong mostly upregulated miRNAs and let-7a was mostlydownregulated [12] miR-133b among other miRNAs targetEGFR in different cancer types especially non-small cell lungcancer [13 14]

Taken together aims of this study were (i) to investigateexpression of EGFR in different subtypes of salivary glandtumours (ii) to define whether overexpression is correlatedwith the copy number variation of EGFR (amplification ofEGFR or polysomy of chromosome 7) or mutation withinEGFR gene (iii) to analyse whether expression of miR-99bmiR-140 and let-7a is different between different salivarygland carcinoma subtypes and finally (iv) to test whetherexpression of miR-133b is in correlation with the expressionof their target EGFR also in salivary gland carcinoma

2 Materials and Methods

21 Patients and Tissue Samples The study comprised 70patients with carcinomas of the parotid gland diagnosed

at the Institute of Pathology Faculty of Medicine at theUniversity of Ljubljana between 1995 and 2008 At diag-nosis patients were staged according to the 6th edition ofUICCTNM system [15] Tumour surgery was performed atthe Department for Otorhinolaryngology and CervicofacialSurgery at University Clinical Centre Ljubljana

Haematoxylin-eosin (HE) slides from formalin-fixedparaffin-embedded (FFPE) tumours were available for allcases and reviewed by pathologist who chose representativetissue blocks for further analysis

For all samples histological subtypes were determinedAdditionally eleven normal salivary glands were included astissue controls All the tumour samples contained more than50 of tumour tissue with majority of samples between 70and 80

The study was approved by Republic of Slovenia NationalMedical Ethics Committee (number 770413)

22 Immunohistochemistry FFPE tissue samples were cut at4 120583m for immunohistochemistry (IHC) All reagents werefrom Ventana Medical Systems Inc Tucson USA exceptotherwise indicated We used commercially available anti-bodies against EGFR (CONFIRM anti-EGFR clone 3C6)Deparaffinization antigen retrieval (including 8 minutes ofProtease 1) and staining with antibodies were performedin an automatic immunostainer (Benchmark XT) in combi-nation with treating sections with secondary antibody andcolour developmentwith horseradish peroxidase (ultraVIEWDAB Detection Kit) The sections were then counterstainedwith haematoxylin Tissue sections of normal salivary glandsserved as positive and sections treated without primaryantibodies as negative controls

The immunostaining of EGFR was semiquantitativelyevaluated based on intensity of membrane reactivity with athreshold of 10 immune-positive cells [1] 0 negative (noreactivity or reactivity in lt10 of cells) 1+ weak reactivityin gt10 of cells 2+ moderate reactivity in gt10 cells and3+ strong reactivity in gt10 cells Tumours with 3+ stainingwere considered as positive Immunostaining pattern wasalso documented in normal salivary gland tissues as controlsAll tissue samples were stained and analysed in duplicate

23 DNA Isolation Tissue samples were cut at 10 120583m fromFFPE tissue blocks and for the isolation procedure six toeight 10 120583m sections were used Total DNA isolation wasperformed using QIAamp DNA FFPE Tissue Kit (Qiagen)according to the manufacturerrsquos protocol The DNA waseluted in 60 120583l of elution buffer The yield was measuredfluorescently using Quant-It (Invitrogen) according to man-ufacturer instruction and Rotor Gene Q (Qiagen)

24 Copy Number Variation (CNV) The copy numbers forEGFR were determined using commercially available andpredesigned TaqMan Copy Number Assays according tothe manufacturerrsquos instructions (Applied Biosystems) Allreagents were from Applied Biosystems except otherwiseindicated The primer ID used for EGFR was as followsEGFR1 Hs00997424 cn [16] The RNaseP and TERT locuswere used as the Copy Number Reference Assays Human




























(Ca ex PA) 9 (129)


(EMC) 5 (71)







3 Results





(a) (b)

















MECs (119899 = 16) 8 (50)4 (250)11 (688)1 (63)

ACCCs (119899 = 16) 2 (111)3 (188)12 (75)1 (63)

Ca ex PA (119899 = 9) 6 (667)1 (111)8 (889)

0

ACCs (119899 = 8) 4 (50)2 (250)5 (625)1 (125)


0

EMCs (119899 = 5) 4 (800)0

5 (100)0

ACNOS (119899 = 4) 01 (250)3 (750)

0

SDCs (119899 = 2) 00

1 (50)1 (50)

SCC (119899 = 1) 00

1 (100)0


1 (100)0


1 (100)0

Total 2770 (386)1370 (186)5370 (757)470 (57)






000100200300400


Fold

chan

ge


lowastlowast

lowast

lowast

(a)


050150250350450550


Fold

chan

ge


lowastlowast

lowastlowast

lowastlowastlowast

(b)


050150250350


Fold

chan

ge

MicroRNAsACCCsEMCs

lowastlowast

lowast

lowast

(c)

minus1000

minus800

minus600

minus400

minus200

000

200

400

600


Fold

chan

ge

MicroRNA


lowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

(d)

minus400

minus200

000

200

400

600

800

1000

1200


Fold

chan

ge


lowast

lowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

(e)




















10

08

06

04

02

Cum

surv

ival


RT RTNo RT


(a)

10

08

06

04

Cum

surv

ival



0=le601=ge60

0

1

(b)

10

08

06

04

02

Cum

surv

ival


Good



(c)

10

08

06

04

02

00

Cum

surv

ival



Metastases

AbsentN1

N2N3

(d)

10

08

06

04

02

Cum

surv

ival


RT RTNo RT


(e)

10

09

08

07

06

05

04

Cum

surv

ival



0=le601=ge60

0

1

(f)


4 Discussion














5 Conclusions








Competing Interests




Acknowledgments


References





























of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of











































(Ca ex PA) 9 (129)


(EMC) 5 (71)







3 Results





(a) (b)

















MECs (119899 = 16) 8 (50)4 (250)11 (688)1 (63)

ACCCs (119899 = 16) 2 (111)3 (188)12 (75)1 (63)

Ca ex PA (119899 = 9) 6 (667)1 (111)8 (889)

0

ACCs (119899 = 8) 4 (50)2 (250)5 (625)1 (125)


0

EMCs (119899 = 5) 4 (800)0

5 (100)0

ACNOS (119899 = 4) 01 (250)3 (750)

0

SDCs (119899 = 2) 00

1 (50)1 (50)

SCC (119899 = 1) 00

1 (100)0


1 (100)0


1 (100)0

Total 2770 (386)1370 (186)5370 (757)470 (57)






000100200300400


Fold

chan

ge


lowastlowast

lowast

lowast

(a)


050150250350450550


Fold

chan

ge


lowastlowast

lowastlowast

lowastlowastlowast

(b)


050150250350


Fold

chan

ge

MicroRNAsACCCsEMCs

lowastlowast

lowast

lowast

(c)

minus1000

minus800

minus600

minus400

minus200

000

200

400

600


Fold

chan

ge

MicroRNA


lowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

(d)

minus400

minus200

000

200

400

600

800

1000

1200


Fold

chan

ge


lowast

lowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

(e)




















10

08

06

04

02

Cum

surv

ival


RT RTNo RT


(a)

10

08

06

04

Cum

surv

ival



0=le601=ge60

0

1

(b)

10

08

06

04

02

Cum

surv

ival


Good



(c)

10

08

06

04

02

00

Cum

surv

ival



Metastases

AbsentN1

N2N3

(d)

10

08

06

04

02

Cum

surv

ival


RT RTNo RT


(e)

10

09

08

07

06

05

04

Cum

surv

ival



0=le601=ge60

0

1

(f)


4 Discussion














5 Conclusions








Competing Interests




Acknowledgments


References





























of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of

















(a) (b)

















MECs (119899 = 16) 8 (50)4 (250)11 (688)1 (63)

ACCCs (119899 = 16) 2 (111)3 (188)12 (75)1 (63)

Ca ex PA (119899 = 9) 6 (667)1 (111)8 (889)

0

ACCs (119899 = 8) 4 (50)2 (250)5 (625)1 (125)


0

EMCs (119899 = 5) 4 (800)0

5 (100)0

ACNOS (119899 = 4) 01 (250)3 (750)

0

SDCs (119899 = 2) 00

1 (50)1 (50)

SCC (119899 = 1) 00

1 (100)0


1 (100)0


1 (100)0

Total 2770 (386)1370 (186)5370 (757)470 (57)






000100200300400


Fold

chan

ge


lowastlowast

lowast

lowast

(a)


050150250350450550


Fold

chan

ge


lowastlowast

lowastlowast

lowastlowastlowast

(b)


050150250350


Fold

chan

ge

MicroRNAsACCCsEMCs

lowastlowast

lowast

lowast

(c)

minus1000

minus800

minus600

minus400

minus200

000

200

400

600


Fold

chan

ge

MicroRNA


lowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

(d)

minus400

minus200

000

200

400

600

800

1000

1200


Fold

chan

ge


lowast

lowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

(e)




















10

08

06

04

02

Cum

surv

ival


RT RTNo RT


(a)

10

08

06

04

Cum

surv

ival



0=le601=ge60

0

1

(b)

10

08

06

04

02

Cum

surv

ival


Good



(c)

10

08

06

04

02

00

Cum

surv

ival



Metastases

AbsentN1

N2N3

(d)

10

08

06

04

02

Cum

surv

ival


RT RTNo RT


(e)

10

09

08

07

06

05

04

Cum

surv

ival



0=le601=ge60

0

1

(f)


4 Discussion














5 Conclusions








Competing Interests




Acknowledgments


References





























of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of




















MECs (119899 = 16) 8 (50)4 (250)11 (688)1 (63)

ACCCs (119899 = 16) 2 (111)3 (188)12 (75)1 (63)

Ca ex PA (119899 = 9) 6 (667)1 (111)8 (889)

0

ACCs (119899 = 8) 4 (50)2 (250)5 (625)1 (125)


0

EMCs (119899 = 5) 4 (800)0

5 (100)0

ACNOS (119899 = 4) 01 (250)3 (750)

0

SDCs (119899 = 2) 00

1 (50)1 (50)

SCC (119899 = 1) 00

1 (100)0


1 (100)0


1 (100)0

Total 2770 (386)1370 (186)5370 (757)470 (57)






000100200300400


Fold

chan

ge


lowastlowast

lowast

lowast

(a)


050150250350450550


Fold

chan

ge


lowastlowast

lowastlowast

lowastlowastlowast

(b)


050150250350


Fold

chan

ge

MicroRNAsACCCsEMCs

lowastlowast

lowast

lowast

(c)

minus1000

minus800

minus600

minus400

minus200

000

200

400

600


Fold

chan

ge

MicroRNA


lowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

(d)

minus400

minus200

000

200

400

600

800

1000

1200


Fold

chan

ge


lowast

lowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

(e)




















10

08

06

04

02

Cum

surv

ival


RT RTNo RT


(a)

10

08

06

04

Cum

surv

ival



0=le601=ge60

0

1

(b)

10

08

06

04

02

Cum

surv

ival


Good



(c)

10

08

06

04

02

00

Cum

surv

ival



Metastases

AbsentN1

N2N3

(d)

10

08

06

04

02

Cum

surv

ival


RT RTNo RT


(e)

10

09

08

07

06

05

04

Cum

surv

ival



0=le601=ge60

0

1

(f)


4 Discussion














5 Conclusions








Competing Interests




Acknowledgments


References





























of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of


















000100200300400


Fold

chan

ge


lowastlowast

lowast

lowast

(a)


050150250350450550


Fold

chan

ge


lowastlowast

lowastlowast

lowastlowastlowast

(b)


050150250350


Fold

chan

ge

MicroRNAsACCCsEMCs

lowastlowast

lowast

lowast

(c)

minus1000

minus800

minus600

minus400

minus200

000

200

400

600


Fold

chan

ge

MicroRNA


lowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

(d)

minus400

minus200

000

200

400

600

800

1000

1200


Fold

chan

ge


lowast

lowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

lowastlowast

(e)




















10

08

06

04

02

Cum

surv

ival


RT RTNo RT


(a)

10

08

06

04

Cum

surv

ival



0=le601=ge60

0

1

(b)

10

08

06

04

02

Cum

surv

ival


Good



(c)

10

08

06

04

02

00

Cum

surv

ival



Metastases

AbsentN1

N2N3

(d)

10

08

06

04

02

Cum

surv

ival


RT RTNo RT


(e)

10

09

08

07

06

05

04

Cum

surv

ival



0=le601=ge60

0

1

(f)


4 Discussion














5 Conclusions








Competing Interests




Acknowledgments


References





























of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of


































10

08

06

04

02

Cum

surv

ival


RT RTNo RT


(a)

10

08

06

04

Cum

surv

ival



0=le601=ge60

0

1

(b)

10

08

06

04

02

Cum

surv

ival


Good



(c)

10

08

06

04

02

00

Cum

surv

ival



Metastases

AbsentN1

N2N3

(d)

10

08

06

04

02

Cum

surv

ival


RT RTNo RT


(e)

10

09

08

07

06

05

04

Cum

surv

ival



0=le601=ge60

0

1

(f)


4 Discussion














5 Conclusions








Competing Interests




Acknowledgments


References





























of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of

























5 Conclusions








Competing Interests




Acknowledgments


References





























of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of





















Competing Interests




Acknowledgments


References





























of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of





















of






Disease Markers



OncologyJournal of





PPAR Research



Journal of

ObesityJournal of
















Expression, Mutation, and Amplification Status of EGFR...

Documents

Transcript of Expression, Mutation, and Amplification Status of EGFR...