Biomarkers of High Grade Dysplasia in Barrett's Esophagus

Biomarkers of High GradeDysplasia in Barrett’sEsophagusMurray Resnick, MD/PhDDirector of Surgical Pathologyand Co-Director of the Molecular PathologyCore Facility, Rhode Island Hospital andAlpert Medical School of Brown University

Barrett’s Esophagus DefinitionA premalignant condition in which the normal squamousepithelium of the distal esophagus undergoes metaplasiato a specialized intestinal epithelium.

Barrett’s Esophagus• 5-15% of all patients with reflux esophagitis(up to 20% of population have GERD)• For every patient with BE another 20 go undetected(Olmstead County autopsy study)• Much more common in white males even in populationswith high incidence of reflux

BE Pathogenesis• Reflux induces inflammation and mucosal injury• Healing occurs by in-growth of stem cells andre-epitheliazation•Cells differentiate into intestinal mucosa mayoffer more resistance to acid damage

Barrett’s Esophagus SequelaeUlcerationBleedingStrictureDysplasia Adenocarcinoma

BE and AdenocarcinomaOnly patients with specialized columnar epithelium are atrisk for developing Barrett’s related EAC.EAC develop in BE at the rate of 1 cancer per 125yrpatient follow-up (annual incidence rate of 800/100,000)Increased frequency in EAC in white malesobserved in last 20yrs (increases 5-10% per year)Why is EAC increasing?Older populationObesityDecrease in helicobacter infection

Histopathologic Dx of DysplasiaDysplasia: Presence of neoplastic epithelium thatremains confined within the basement membrane.Occurs in up to 10-20% of cases if left untreated.Identification of dysplasia by light microscopy isstill considered the “gold standard” by which patientfollow-up and treatment is coordinated.

Significance of Dysplasia in BEDysplasia in BE:Presence of dysplasia (particularly HGD) is a risk factorfor synchronous and metachronous EAC.Dysplasia is frequently seen adjacent to and distant fromBarrett’s related EAC.Dysplasia is both a marker of adenocarcinoma and isclearly the pre-invasive lesion.

Dysplasia in Barrett’s Esophagus•Negative for dysplasia•Positive for dysplasiaLow grade dysplasiaHigh grade dysplasia•Indefinite for dysplasia

Grading Dysplasia in BEFeature Negative IND LGD HGDSurface + + - -maturationArchitecture Normal Normal or mild Mild alteration Marked alterationalterationCytology Normal or Mild alterations or Mild alterations (D) Marked alteration;reactive focal marked with or marked (F); loss of polarityinflammation maintained polarity(Montgomery et al Human Pathology 2001)

LGD

Surveillance GuidelinesInitial screening of individuals with highest risk:white men over 50 with chronic symptoms of heartburnNo dysplasia- Every 2-3yrs after two negative endoscopiesLGD - Every 6mths for 1yr, then yearlyHGD- Second confirmation followed by surgery orevery 3mth surveillanceSurveillance = 4 quarter biopsies at 2cm intervals

Dysplasia in BEPitfalls• Sampling errors• Diagnostic interpretationreactive vs dysplasticlow-grade vs high gradehigh grade vs intramucosal ca

Barrett’s Related DysplasiaInterobserver AgreementCategory%AgreementHGD+IMC vs others 85-87%Negative vs others 71-72%Negative+Ind vs others 75-77%Neg vs ind/LG vs HGD/IMC 58-61%Reid BJ et al Human Pathology 1988

Ancillary Techniques in theScreening and DX of BEFlow cytometry Aneuploidy more common in HGD and maybe a marker for progressionMolecular P53 (late)Cyclin D1 expression (early)Telomerase (early and non-dysplastic)p16 hypermethylation or mutation (early)Brush cytology and balloon cytology“Light” biopsy (fluorescence, optical coherence tomography)

ObjectivesIdentify differentially expressed genes in HGD compared withnon-dysplastic Barrett’s esophagus through gene microarrayanalysis of epithelial cells microdissected from archival tissuespecimens by LCM.Identify markers of dysplasia suitable for immunohistochemistry

Advantages over other microarray studies screening forbiomarkers of BE associated dysplasia:LCM of epithelium as opposed to whole biopsies allows for:Identification of epithelial associated genes.Confirms by direct visualization presence of dysplasia.Only study to use formalin-fixed paraffin embedded archivaltissues.One of only two studies to use samples of dysplasia and nondysplasticBE from the same patient (other n=1).

Methods (1)• Esophageal biopsies (11 cases)– Formalin-fixed, paraffin-embedded. 2004-7– Inclusion criteria• Good tissue orientation• Easily defined areas of HGD and normal BE (SAME PATIENT!!)• HGD confirmed by 2-3 GI pathologists• Must have high quality mRNA–10μm sections deparaffinized, scraped– RNA extracted with Paradise kit (Arcturus)– DNA removed (DNAse)– RNA quality determined by» Agilent 2100 Bioanalyser» 5’ and 3’ β-actin RT-PCR (oligo-dT)–3’:5’β -actin < 10; >15ng

Specimen Evaluation40 cases (1-6 blocks of biopsies per case) received fromcollaborators26 had both NBE and HGD available for LCM11 had sufficient quantities of RNA and passed 3’5’ QCevaluation for intact RNA

Methods (2)• RNA Extraction/Preparation– 2500 surface esophageal epithelial cells (AutoPix LCM)– RNA extracted (Paradise kit, Arcturus)– RNA quality determined by 5’ and 3’ β-actin QT-PCR– Amplified and Reverse transcribed (Arcturus T7 kit)– cDNA → biotin-labeled cRNA labeled (Enzo kit)– Chip hybridization, staining with streptavidin bound tophycoerytrin, expression assayed (Keck Center, Yale)– Affymetrix custom U133-X3P expression array• As for standard U133+2 array– 61,000 probe sets targeting 47,000 transcripts• But, focused on 300 bases at 3’ end– Because FFPE mRNA fragmentation, chemical modification, esp 5’

Biopsy of Normal BE

Biopsy of HGD

Use of Aperio ScanScope for Remote Online Annotationof Slides for LCMAperioScanScopeCS SystemEntire slide scanned in labAccessed by internet link from pathologist’s officeSmall biopsies with minor areas of specific cellsidentified and annotated with pathologist’s expertiseAnnotated image displayed on Autopix screenLab tech uses as reference to guide dissectionLaser Micro Dissection

Laser Capture MicrodissectionBarrett’s EpitheliumHigh Grade DysplasiaABABCDCDEsophageal epithelium stained with Histogene staining solution (A), marked for capture (B),after capture (C) and captured cells after dissection (D)

BioinformaticsExpression signals normalized using the Standardization andNormalization of Microarray Data web-based softwareConcordant absent expression signals were removed from analysisStatistical comparison of HGD to ND was done using the Wilcoxonmatched-pairs signal ranks test (>2.5 fold cutoff, p

Results132 genes with 2.5 > increased expression in HGD vs ND BE13 genes with 2.5 > decreased expression in HGD vs ND BE

Dysregulated Genes in HGD by Functional CategorySignal TransductionCell Communication and MotilityImmune ResponseECM and ProteaseProtein BindingCalcium Binding and Ion ChannelMiscellaneousCell Adhesion and MembraneCell Proliferation and Cell DeathStructural and CytoskeletonMetabolism and MitochondriaNucleic Acid Binding &Transcription Factors-10 -5 0 5 10 15 20 25Down-regulated in HGD Up-regulated in HGD

Table 2. Differentially Expressed Genes Confirmed by Real-Time Quantitative PCR*GeneChipFunctional Gene Group GeneBank ID Fold p valueChangeUpregulated genesComplement component 3 Miscellaneous NM_000064 13.1 0.016Topoisomerase (DNA) IIα 170kDa Cell proliferation and cell death AL561834 12.6 0.01Matrix metallopeptidase 1 (interstitial collagenase) Extracellular matrix and protease NM_002421 10.8 0.018Matrix metallopeptidase 12 (macrophage elastase) Extracellular matrix and protease NM_002426 9.3 0.006TIA1 cytotoxic granule-associated RNA binding protein-like 1 Cell proliferation and cell death NM_003252 8.3 0.005Protein tyrosine phosphatase, non-receptor type 7 Metabolism and mitochondria NM_002832 7.3 0.010Nurim (nuclear envelope membrane protein) Cell adhesion, motility and membrane proteins NM_014641 7 0.016Histone cluster 1, H2ag Nucleic acid binding and transcription factors NM_021064 7 0.004Lipocalin-2 (oncogene 24p3) Ca+ binding and ion channels NM_005564 6.8 0.003Fidgetin-like 1 Cell proliferation and cell death AK023411 6.8 0.01Protein phosphatase 2 (formerly 2A), catalytic subunit, β isoform Cell proliferation and cell death AI379894 6.7 0.003Formyl peptide receptor 1 Cell adhesion, motility and membrane proteins NM_002029 6.40.003Minichromosome maintenance complex component 10 Cell proliferation and cell death AL136840 6.3 0.005Secernin 1 Miscellaneous NM_014766 6 0.004S100 calcium binding protein A9 Ca+ binding and ion channels NM_002965 4.1 0.003Downregulated genesFatty acid binding protein 1 liver Metabolism and mitochondria NM_001443 -2.5 0.003Trefoil factor 1 Cell adhesion, motility and membrane proteins NM_003225 -2.6 0.003Mucin 5AC, oligomeric mucus/gel-forming Extracellular matrix and protease AW192795 -2.6 0.01Aldolase B, fructose-bisphosphate Metabolism and mitochondria NM_000035 -2.8 0.003Calcium/calmodulin-dependent protein kinase II inhibitor 1 Signal transduction N75559 -3 0.006Hydroxysteroid (17-β) dehydrogenase 2 Metabolism and mitochondria NM_002153 -3.1 0.003Hydroxyprostaglandin dehydrogenase 15-(NAD) Metabolism and mitochondria J05594 -3.3 0.004Meprin A, alpha (PABA peptide hydrolase) Extracellular matrix and protease NM_005588 -4.7 0.003Alanyl (membrane) aminopeptidase Extracellular matrix and protease NM_001150 -4.7 0.003

CAMK2N1MUC5ACHSD17BPGDHAMEP1AALDOBANPEPFABP1TFF1HIST2H2AA3FPR1MMP12FIGNLMCM10LCN2HIST1H2AGPTPN7TIALPPP2CBSCRN1NRMS100A9TOP2A151050Fold Change-5-10-15Confirmation of the Expression Levels of SelectedDysregulated Genes by real-time PCR andCorrelation with Microarray DataR S = 0.75, p

Protein Expression Confirmed by ImmunohistochemistryUpregulatedTopoisomerase II alpha•DNA replication in a number of cancersLipocalin-2•Cell survival factor upregulated in GI cancersS100-A9•Chemotaxis and cell proliferationDownregulatedTrefoil Factor –1•Putative tumor suppressor gene in GI cancers

Topoisomerase II alpha

Lipocalin-2

S100-A9

Trefoil factor 1

Immunohistochemical stains for topoisomerase IIa (Topo), lipocalin-2 (Lipo),S100A9 (S-100), and TFF1. ADCA, adenocarcinomaTopoNumber of cases15105012(100%)p < 0.01 for:NDBE vs AllLGD vs HGD075 6(58%) 5 4(45%)(55%)(42%) (33%)0 0 0(8%)1NDBE LGD HGD ADCA7(58%)Number of casesLipo12108642011(92%)p < 0.02 for:NDBE vs All21(18%)(8%)05(46%) 4(36%)5(42%)5(42%)2(16%)7(58%)5(42%)NDBE LGD HGD ADCA0Number of casesS-10012108642012(100%)0 0p < 0.03 for:NDBE vs AllHGD vs ADCA7(64%)4(36%)07(58%)5(42%)01(8%)NDBE LGD HGD ADCA6(50%) 5(42%)Number of casesTFF1IHC Score = +1 IHC Score = +2 IHC Score = +31210864200 012(100%)4(36%)p < 0.03 for:NDBE vs AllLGD vs HGD9(82%)1(9%)6(55%)1 1(9%) (9%)11(92%)01(8%)NDBE LGD HGD ADCA

Next Steps……1. Continue testing different commercially available markers2. Generate Abs towards novel proteins3. Determine utility of the expression of these markers on a large groupof BE associated dysplasia cases.a. LGD vs HGDb. Indefinite vs LGD (or HGD)4. Determine whether these markers can predict progression from LGDor HGD to adenocarcinoma.

Thanks to…….Patricia A. MeitnerEdmond SaboRosemarie TavaresSteven F.MossCOBRE Core lab Rhode Island HospitalPathology, Rambam Medical CenterHaifa, IsraelCOBRE Core lab Rhode Island HospitalGI/Medicine, Brown University, ProvidenceChristopher L. Corless Pathology, OHSU Portland OregonGregory Y. LauwersPathology, Mass General HospitalSupported in part by the Rhode Island Hospital COBRE CCRD.Director: Douglas C. Hixson, PhD

Molecular Pathology CoreRonald DeLellis, MDMurray Resnick, MD/PhDPatricia Meitner, PhDRose TavaresLelia NobleLaura GanttAldrich Building, 6 th FloorRhode Island HospitalStaffCore DirectorCore Co-Director, Tissue BankMolecular Biology & LCMIHC and TMAsTissue Bank and TMAsAdministrative/BillingSupported in part by the Rhode Island Hospital COBRE CCRD. Director: Douglas C. Hixson, PhD