Department of Pathology

Inflammatory Bowel Disease Neoplasia Mary P. Bronner, MD Division Chief of Anatomic Pathology University of Utah

Department of Pathology

Neoplastic Progression in Chronic Inflammatory GI Dz Inflammation Dysplasia Carcinoma

Department of Pathology

Chronic Inflammatory GI Disease & Cancer • Barrett’s Esoph

Esoph CA

• HP Gastritis

Gastric CA

• Hepatitis B & C

HCC

• Ch Pancreatitis

Panc CA

• UC and Crohn’s

Intestinal CA

Department of Pathology

Ulcerative Colitis: A Paradigm

Department of Pathology

Department of Pathology

Managing Cancer Risk in UC •Ignore it •“Prophylactic” colectomy •Colonoscopic surveillance for dysplasia / early carcinoma

Department of Pathology

Optimal Colonic Biomarker • Pancolonic distribution • Predate incurable cancer

• Objective • Sensitive, Specific, PPV, NPV

Gold Standard Biomarker: Dysplasia

Department of Pathology

Dysplasia: Problems • Sampling • Distinction from reactive change

• Observer variation • Natural history incompletely understood

Department of Pathology

Adequate Bx Sampling Histology Dysplasia Cancer

No. Bx’s for 90% confidence

33

34

No. Bx’s for 95% confidence

56

64

From: Rubin CE, et al. Gastroenterology 1992;103:1611

Department of Pathology

UC Surveillance Protocol

10 cm

X

X

X

X

X

X

X

X

X

X

X

X

X X

X

X X

X

X X

X

X

X

X

X

X X X X X X X X X X X X X

X

X

X X

X X

5 cm

Department of Pathology

Rectosigmoid Predominance of Ulcerative Colitis Cancer Location of Colorectal Carcinoma

RS 52%

D T 12% 21%

A/C 15%

Choi PM. Gastroenterology 1993;104:666 Summary of 5 Studies

Department of Pathology

Dysplasia: Problems • Sampling

• Distinction from reactive change • Observer variation • Natural history incompletely understood

Department of Pathology

Department of Pathology

Department of Pathology

Dysplasia: Problems • Sampling

• Distinction from reactive change • Observer variation • Natural history incompletely understood

Department of Pathology

Outcome of 40 UC LGD Patients • 78% no progression, avg f/u 5y (1-13 y)

• 22% HGD, avg f/u 1.5 y (1-3 y) • ≥3 LGD biopsies: 9x progression risk

• 2 non-compliant patients developed Dukes’ A cancer Brentnall, Bronner, et al. Prospective study of progression of LGD in UC. Inflamm Bowel Dis 18:2240-6, 2012.

Dysplastic Field: Limited

Department of Pathology

Better Biomarkers of Cancer Risk Greatly Needed!

Department of Pathology

Department of Pathology

Chromosomal Instability? • FCM Aneuploidy - Detects gross chromosomal instability • CGH - Detects clonal gains and losses of chromosomal regions • FISH - Detects clonal and non-clonal chromosomal abnormalities

Department of Pathology

Biopsy Sampling: Flow Cytometry Dysplasia Cancer

No. Bx for 90% confidence

20

8

No. Bx for 95% confidence

30

14

Rubin CE, et al. Gastroenterology 1992;103:1611

Morphologic + DNA Ploidy Neoplastic Field: Larger

Department of Pathology

Metaphase Comparative Genomic Hybridization in UC 39% (15/38) of diploid bx’s near dysplasia or cancer showed CGH detectable alterations Performed in collaboration with F. Waldman, UCSF

Department of Pathology

Array-based Comparative Genomic Hybridization (CGH) Chromosomes replaced by ordered array of targets Karyotyping of metaphase spreads not necessary Greatly increased resolution

Department of Pathology

Array CGH in UC • 100% (9/9) UC-progressors extensive chromosomal gains and

losses • FISH and PCR targets identified Bronner MP, Mod Pathol 2010;23:1624-33

Department of Pathology

Ulcerative Colitis A-CGH

PROGRESSORS

NON-PROGRESSORS Gain Loss

Bronner MP, Mod Pathol 2010;23:1624-33.

Department of Pathology

BAC CGH Whole Genome Log2-Ratio Plots of All Chromosomes Normal Non-UC Control

UC Progressor

UC Non-progressor

UC Progressor

Bronner MP, Mod Pathol 2010;23:1624-33.

Morphology + DNA Ploidy + CGH Neoplastic Field: Larger Still

Department of Pathology

Non-Clonal Change in UC: Wider Field? • DNA Flow & CGH detect clonally expanded abnormalities only • Larger fields of non-clonal instability? Detectable in negative biopsies, even from rectum? • Assessed by Fluorescence In Situ Hybridization (FISH)?

Department of Pathology

UC FISH Hypothesis: UC progressors differ from UC non-progressors using non-clonal genomic instability biomarkers on single negative rectal biopsies

Department of Pathology

FISH • Interphase nuclear suspensions placed on glass slide • Locus specific probes (Chrom 8, 11, 17, 18) & centromeres (green and red) • Red and green FISH spots counted per 100 nuclei

Normal Cells

Abnormal Cells

Department of Pathology

Control Normal Colon FISH Chrom11 Probe Set 90 % of Nuclei

100 80 60

40 20 0

1

0

2

0

1

2

0

2

2

0r2g 1r1g 1r2g 1r3g 2r1g 2r2g 2r3g 2r4g 3r2g 3r3g Red and Green Signal Counts

Department of Pathology

Diploid Neg Rectal Bx UC Progressor 100 Chrom11 Probe Set 74

% of Nuclei

80 60 40

20 0

4

11

8

2

1

1

0

0

0

0r2g 1r1g 1r2g 1r3g 2r1g 2r2g 2r3g 2r4g 3r2g 3r3g Red and Green Signal Counts

Department of Pathology

FISH in Ulcerative Colitis

% cells with FISH abnormalities

10 8

6

p0.001 p0.001

Non-UC controls N=10 UC non-progressors N=18 UC progressors N=12 p0.001 p=0.001

4 2 0

Arm Centromere Centromere Arm loss gain loss gain Bronner MP, et al. Am J Pathol 2008;173:1853.

0.8 1.0

11q: CyclinD1

0.0

0.2 0.4 0.6

8q: c-myc Sensitivity

0.6 0.8 1.0

ROC Analysis of FISH Biomarkers

0.0 0.2 0.4

Sensitivity

Department of Pathology

0.0

0.2

0.4

0.6

0.8

1.0

0.0

0.2

Sensitivity

18q: DCC

0.0

0.2

0.4

0.6

1-Specificiy

0.6

0.8

1.0

1-Specificity

0.8

1.0

0.0 0.2 0.4 0.6 0.8 1.0

0.0 0.2 0.4 0.6 0.8 1.0

Sensitivty

Specificity 1-Specificity

0.4

17p: p53

0.0

0.2

0.4

0.6

Specificity 1-Specificity

Bronner MP, et al. Am J Pathol 173:1853-1860, 2008

0.8

1.0

Department of Pathology

ROC Analysis of FISH Biomarkers 0.8 0.6 0.4 0.0

0.2

Sensitivity

1.0

All 4 chromosomes combined

0.0

0.2

0.4

0.6

0.8

1.0

1-Specificity

Bronner MP, et al. Am J Pathol 2008;173:1853.

Department of Pathology

Consequences of Shortened Telomeres • Sticky chromosomal ends • Bridge-breakage-fusion cycles

• Chromosomal arm losses/gains and dicentrics Studied by peptide nucleic acid (PNA) probe ISH or RT PCR

Department of Pathology

Department of Pathology

Epithelial: Stromal Telomere Ratio

Telomere Shortening in UC 1.4

1.2 1.0

p=0.001 p=0.08 p=0.02

0.8 0.6 0.4

0.2 0

Non-UC NonProgressors control progressors O’Sullivan J, et al. Nat Genet 2002;32:280-284.

Anaphase Bridges in UC p=0.0002

% Anaphase Bridges

0.03

p=0.011

0.025

0.02 0.015 0.01 0.005 0

NonNon-UC Progressors progressors control Bronner MP, et al. Am J Pathol 173:1853-1860, 2008

Department of Pathology

NGS miRNA bioclassifier of UC patients at increased risk of colon cancer • Why miRNAs? – Small size (~21nt) more stable, less

ribonuclease degradation – Readily detectable in FFPE and stained

slides – Important roles in immune regulation

miRNAs misregulation in UC-P, UC-NP

Department of Pathology

UC_NP UC_P

1000

10000

UC-NP vs. nl (26 miRNAs)

15

11

UC-P vs. nl (29 miRNAs)

18

UC_NP UC_P

1000 10000 10

miRNA Panel

UC-NP 9/10

1001000 1 100

Normalized Read count

10000 100

UC_NP UC_P

• Linear discriminant analysis to predict UC-P vs. UC-NP • Robust candidate panel selected for RT-PCR & additional cohort validation

mir UC_NP+UC_P

UC_NP+UC_P

UC-P 10/10

UC_NP+UC_P

UC_NP+UC_P

UC_NP+

Histology + DNA Ploidy + CGH +FISH +Telomeres +Ana Bridges +miRNA Neoplastic Field: Entire Colon

Department of Pathology

Department of Pathology

UC Polypoid Dysplasia

You’re dalmed if you do, and dalmed if you don’t Teri Brentnall,MD

Department of Pathology

Department of Pathology

Dysplasia in UC vs Adenoma •No clinical features •No endoscopic features •No pathologic features •No molecular tests

Department of Pathology

HOWEVER • If the lesion can be demonstrably completely removed endoscopically • Has only Low-Grade Dysplasia

• There is no other dysplasia on adequate sampling • Then, careful follow-up may be considered

Department of Pathology

UC Dysplasia Management Continue Surveillance with adequate sampling: –Single site LGD while in surveillance –Indefinite of negative for dysplasia

Department of Pathology

UC Dysplasia Management Consider Colectomy: –Multiple LGD sites –LGD on more than one endoscopy –LGD at initial colonoscopy –Excessive inflammatory polyps

Department of Pathology

Inflammatory Polyps

Department of Pathology

UC Dysplasia Management Colectomy Indicated: –HGD –Endoscopically unresectable dysplastic lesion

Department of Pathology

Conclusions • Molecular alterations are widespread in UC, CD, CP, HP, HCV

• Single non-dysplastic bx alterations show promise for reducing sampling error

• Paradigm for cancer in chronic inflammatory disease

Department of Pathology

Further Work: • Reproducibility • Longitudinal analyses • Prospective validation • High throughput • Reduced numbers of markers • Mechanism: why progressors?

Department of Pathology

Thanks To My Colleagues: Bonnie Shadrach Teri Brentnall Peter Rabinovitch Ru Chen David Crispin Rosana Risques Jacintha O’Sullivan

Noah Welker Keith Lai Danielle Elsberry Ryan O’Connell June Round John Valentine

Department of Pathology