Novel stool-based protein biomarkers for improved colorectal

cancer screening Meike de Wit

DCCG middag 6 juli

The MEDOCC plan WP1 – Biomarker based Screening WP2 – ctDNA liquid biopsy for

residual disease detection

Better assays Better markers

Ye

ar

1 &

2

Ye

ar

3 &

4

Increase analytical sensitivity for DNA methylation assays

- Improved DNA extraction - nanotechnology based methylation detection

Protein markers* - Clinical validation & selection

- Antibody based assay development *separate funding

WP3 - Turning into practice

1) Better CRC stool-based screening tests ready to be implemented in CRC screening programs 2) Validated ctDNA liquid biopsy for residual disease detection & monitoring assay ready for intervention trial 3) Validated Wnt target gene methylation assay as prognostic marker, ready for intervention trial 4) Direct comparison of the prognostics value of liquid biopsy versus tissue based biomarker

Prospective screening trial in the context of the Dutch CRC population based

screening program with a paired design

Deliverables

Compare biomarker test to state of the art test (= FIT)

Better assays Observational study

ctDNA - Rearrangements - Mutations - Promoter methylation

- Within PICNIC registry - Prospective observational design - Tissue&plasma collection - Repeated ctDNA assays - Clinical follow up compare to Wnt target gene methylation in tissue* *separate funding

Patient follow up

Correlate ctDNA & tissue methylation markers to: Disease free survival at 3 years Overall survival at 5 and 7 years

> Started in 2014

> All persons 55-75 year: biennial FIT

> Expected number of lives saved yearly 2400

Fecal immunochemical test (FIT) Colonoscopy

+ FIT

CRC screening in the Netherlands

FIT performance at specificity ~95%* Sensitivity: CRC: ~65% Precursor lesions (advanced adenomas): ~27% * Van Veen W et al, Colorectal cancer screening: advice from the Health Council of the Netherlands, Ned.Tijdschr.Geneesk., 2009

Room for improvement!!

DNA

Proteins

Molecular changes as biomarkers in

stool

10

co

ntr

ol

CR

C

12

LTQ-FTMS 20 30 40 50 60 0

100

129 43

A

AA

co

ntr

ol

CR

C

40 79

QExactive 29

4-protein marker panels By logistic regression and classification and regression tree (CART) analysis

Workflow:

MS-based protein identification in stool Sample series 1

N=22 Sample series 2

N=291

CR

C

co

ntr

ol

Hemoglobin: sensitivity of 43% for detection of CRC at 95% specificity.

HBA1

CR

C

95% Specificity

Marker panel: sensitivity of 71% versus 43% Hemoglobin alone at

95% specificity.

HBA1 Marker panel

co

ntr

ol

Hemoglobin: sensitivity of 8% for detection of advanced adenomas at 95% specificity.

HBA1

co

ntr

ol

AA

AA

95% Specificity

co

ntr

ol

Marker panel: Sensitivity of 40% versus 8% Hemoglobin alone at

95% specificity.

HBA1 Marker panel

biomarker proteins

FIT ~ 10 mg stool

Does it work in a small FIT sample?

Multiplexed antibody-based assay

*Not part of the top list of 29

• Off the shelf assay for 4 out of top 50 (NOT top 4) proteins

1 2 3 4

biomarker proteins

FIT ~ 10 mg stool

Does it work in a small FIT sample?

*Not part of the top list of 29

1 2 3 4

• Off the shelf assay for 4 out of top 50 (NOT top 4) proteins

22 20

16 14

0

5

10

15

20

25

Normal Adenoma Advanced adenoma

CRC

FIT samples (50 µl) n = 72

Multiplexed antibody-based assay

biomarker proteins

CRCs vs controls: P=2.0E-08 CRCs vs controls: P=1.1E-05 n

g/m

l

ng

/ml

pg

/ml

pg

/ml

controls As AAs CRCs controls As AAs CRCs

controls As AAs CRCs controls As AAs CRCs

Validation in FIT samples

CRCs vs controls: P=0.1 CRCs vs controls: P=6.9E-05

1 2 3 4

biomarker proteins

Assay development for 10 selected candidates

GLP / CLIA compliant quality

29

*Outsource assay development

10

FIT and biomarker test

Prospective cross-sectional screening trial within the Dutch CRC screening program

n= 10,000

biomarker test

Prospective cross-sectional screening trial within the Dutch CRC screening program

n= 10,000

marker validation

assay development

prospective validation

cost effectiveness

clinical trial

marker identification

test approval & implementation

Protein markers

Retrospective stool collections

Robust technology &

automation

Prospective stool collections

Dutch screening program

from 2014

Department of Pathology

Gerrit Meijer

Beatriz Carvalho

Remond Fijneman

Linda Bosch

Meike de Wit

Annemieke Hiemstra

Sandra Mongera

Department of Pathology

Manon van Engeland

Veerle Melotte

Kathleen Daenen

Department of Gastroenterology

Chris Mulder

Jochim Terhaar sive Droste

Frank Oort

Sietze van Turenhout

Ilhame Ben Larbi

Graham Lidgard

Barry Berger

Michael Domanico

Department of Gastroenterology

Ad Masclee

Silvia Sanduleanu

Carolina Khalid-de Bakker

Daisy Jonkers

Joost Louwagie

Wim van Criekinge

Department of Gastroenterology

Evelien Dekker

Thomas de Wijkerslooth

Joep Ijspeert

Clasine de Klerk

Department of Gastroenterology

Ernst Kuipers

Manon Spaander

Esther Stoop

Els Wieten

Acknowledgements

Department of Gastroenterology

René van der Hulst

Marcia Horn Huig Schippers

Oncoproteomics Laboratory dept. of Medical Oncology Connie Jimenez Thang Pham Sander Piersma

Department of Epidemiology

and Biostatistics

Veerle Coupé

Marjolein Greuter

biomarker proteins

Validation FIT sample set for Phase 1

833

252 166

47

Control Adenomas Advancedadenomas

CRCs

Total FITs @ AVL/NKI

93

32 27 25

Control Adenomas Advancedadenomas

CRCs

FF samples: Referral population n=177

Control Adenomas Advanced adenomas CRCs

716

202 123

8

Control Adenomas AA/ASP CRCs

Cocos samples: CRC screening study n=1049

Control Adenomas AA/ASP CRCs

24

18 16 14

Control Adenomas Advancedadenomas

CRCs

FKG samples: Referral population n=72

Control Adenomas Advanced adenomas CRCs