University of Groningen

Heredity nonpolyposis colorectal cancerRijcken, Fleur Elise Marie

IMPORTANT NOTE: You are advised to consult the publisher's version (publisher's PDF) if you wish to cite fromit. Please check the document version below.

Document VersionPublisher's PDF, also known as Version of record

Publication date:2006

Link to publication in University of Groningen/UMCG research database

Citation for published version (APA):Rijcken, F. E. M. (2006). Heredity nonpolyposis colorectal cancer: studies on tumor development andpreventive strategies. s.n.

CopyrightOther than for strictly personal use, it is not permitted to download or to forward/distribute the text or part of it without the consent of theauthor(s) and/or copyright holder(s), unless the work is under an open content license (like Creative Commons).

The publication may also be distributed here under the terms of Article 25fa of the Dutch Copyright Act, indicated by the “Taverne” license.More information can be found on the University of Groningen website: https://www.rug.nl/library/open-access/self-archiving-pure/taverne-amendment.

Take-down policyIf you believe that this document breaches copyright please contact us providing details, and we will remove access to the work immediatelyand investigate your claim.

Downloaded from the University of Groningen/UMCG research database (Pure): http://www.rug.nl/research/portal. For technical reasons thenumber of authors shown on this cover page is limited to 10 maximum.

Download date: 02-08-2022

Hereditary

nonpolyposis

colorectal

cancer:

Studies on tumor development and preventive strategies

F.E.M. Rijcken

Rijcken, Fleur Elise Marie

Hereditary nonpolyposis colorectal cancer: Studies on tumor development and preventive

strategies

ISBN 90-367-2731-6

ISBN 90-367-2732-4

© Copyright 2006 FEM Rijcken

All rights are reserved. This publication is protected by copyright. No part of it may be reproduced,

stored in a retrieval system, or transmitted, in any form or by any means without the permission of

the author.

Cover Immunohistochemical staining with Cytodeath (cCK 18, Mab M30)

Cover design and printed by Pasmans Offsetdrukkerij BV, Den Haag

RIJKSUNIVERSITEIT GRONINGEN

Hereditary nonpolyposis colorectal cancer:

Studies on tumor development and preventive strategies

Proefschrift

ter verkrijging van het doctoraat in de Medische Wetenschappen

aan de Rijksuniversiteit Groningen op gezag van de

Rector Magnificus, dr. F. Zwarts, in het openbaar te verdedigen op

maandag 18 september 2006 om 16:15 uur

door

Fleur Elise Marie Rijcken

geboren op 17 mei 1975

te Paramaribo

Promotores Prof dr JH Kleibeuker

Prof dr H Hollema

Prof dr AGJ van der Zee

Beoordelingscommisie

Prof dr RMW Hofstra

Prof dr JHJM van Krieken

Prof dr EGE de Vries

Paranimfen Dr FYFL de Vos

Dr M Oldenrode-Berends

This thesis was financially supported by a grant RUG 98-1660 of the Dutch Cancer Society en

GUIDE.

Aan Marjolein

Contents

Chapter 1 Introduction

Chapter 2 Proximal adenomas in hereditary nonpolyposis colorectal cancer are prone to

rapid malignant transformation

Chapter 3 CASE REPORT

Rapidly progressive adenomatous polyposis in a patient with germline

mutations in both the APC and MLH1 genes: the worst of two worlds

Chapter 4 Hyperplastic polyps in hereditary nonpolyposis colorectal cancer

Chapter 5 Early carcinogenic events in HNPCC adenomas: differences with sporadic

adenomas

Chapter 6 Cell cycle regulators and apoptosis associated proteins in relation to

proliferative activity and degree of apoptosis in HNPCC versus sporadic

endometrial carcinoma.

Chapter 7 Gynecologic screening in hereditary nonpolyposis colorectal cancer

Chapter 8 Sulindac treatment in HNPCC patients

Chapter 9 Summary and future perspective

Samenvatting

Dankwoord

Curriculum vitae

Publicaties

1

Introduction

Chapter 1 |

In 1895, Aldred Warthin, a pathologist, described a family, family G, in which many relatives

had a diagnosis of colorectal cancer. Family G is the first documented family with hereditary

nonpolyposis colorectal cancer (HNPCC), an autosomal dominantly inheriting disorder,

predisposing affected individuals to the development of cancer. HNPCC is characterized by

the development of colorectal and extra-colonic cancer, often at young age, accounting for 2-

5% of all subjects with colorectal cancer and 1-3% of subjects suffering of endometrial

cancer. 1

To achieve uniformity for purpose of investigational protocols, a set of diagnostic criteria was

formulated at a meeting of the International Collaborative Group on Hereditary Nonpolyposis

Colorectal Cancer. The so-called Amsterdam criteria are as follows: 1) at least three relatives

in at least two successive generations should have histologically verified colorectal cancer; 2)

one should be a first degree relative of the other two; 3) one of the colorectal cancers should

be diagnosed under the age of 50 years; 4) familial adenomatous polyposis should be

excluded.2 Some years ago, the criteria were modified to include extra-colonic tumors, i.e.

endometrial, and small-bowel cancer and transitional cell cancer of the renal pelvis or ureter. 3

HNPCC patients have a high rate of synchronous and metachronous tumor development. As

can be derived from the criteria, HNPCC cancers are diagnosed at a relatively early age.

Colorectal cancers occur predominantly in the proximal colon.

Almost a century after the first HNPCC family was recognized clinically, the genetic

background of this disorder was largely unraveled. The first known susceptibility loci,

originally identified in bacteria and yeast, were mapped to chromosome 2p16 (hMSH2) and

chromosome 3p21 (hMLH1)4,5, accounting for the majority of reported HNPCC cases. To

date, two other genes, MSH6 and PMS2, have been associated with (atypical) HNPCC. Three

genes that were previously implicated as the cause of HNPCC in some families, MLH3,

PMS1, and EXO1, have recently been shown to be unlikely causes of HNPCC.6-8 The HNPCC

susceptibility genes, so called mismatch repair (MMR) genes, are part of a complex DNA

repair system which is responsible for correction of mismatches, small insertions or deletions

that arise by mis-incorporations or slippage of the DNA-polymerase during DNA replication.

A hMSH2 heterodimer recognizes and binds directly to mismatched DNA sequences.9 The

heterodimeric complex hMSH2-hMSH6 preferentially recognizes single mismatched base-pair

while hMSH2-hMSH3 recognizes large insertions or deletions.10 A second heterodimeric

10

Introduction |

complex, of hMLH1 and PMS2, is then recruited and binds to the hMSH2-hMSH6 complex

after ATP is bound, and together this group of four proteins target hEXO1 to actually do the

repair. There is some redundancy to the system which helps explain why defects in MLH1

and MSH2 account for the majority of HNPCC. MLH1 can also form heterodimers with either

MLH3 or PMS1 and MSH2 can bind to MSH3. MLH1 is always required for mismatch

repair, but the redundancy of the proteins that can partner with MLH1 lead to modest

phenotype, if any at all.

Short tandem-repeat sequences, microsatellites, are distributed throughout the human genome

and typically consist of DNA repeats of up to six nucleotides, and the total length of the

stretch is fewer than 100 base pairs. Microsatellites are prone to errors, “slippage”, during

DNA replication, resulting in instability of the short repetitive base sequences or so called

microsatellite instability (MSI) if not recognized and corrected by the mismatch repair

system.11 Being due to a deficient MMR system, MSI is the molecular hallmark of hereditary

nonpolyposis colorectal cancer.11,12 To facilitate uniformity, microsatellite instability was

defined at a consensus conference at the National Cancer Institute using a reference panel of

five microsatellite markers and was divided into the two phenotypes: MSI-high (instability at

at least two markers) and MSI-low (instability at one marker).13 No instability at any of the

analyzed markers is called a microsatellite-stable (MSS) phenotype. However, two large

studies have reported that the distinction between MSI-L and MSS is unwarranted as if

enough markers are studied across a large set of tumors, eventually all tumors show instability

at at least one marker.14

Genetic inactivation of a mismatch repair gene is generally associated with loss of

immunohistochemical expression of the corresponding protein.15 Immunohistochemistry for

MLH1 and MSH2 has a high sensitivity and specificity for screening for DNA mismatch

repair defects caused by mutation in MLH1 and MSH2, respectively, and thus indirectly for

identifying colorectal tumors of the MSI-high phenotype.16,17

Cells deficient of MMR, show a 100-1000-fold increase in the rate of spontaneous mutations

as compared to normal cells.18,19 Accumulation of replication errors, microsatellite instability,

may occur within repeat sequences of coding genes relevant for growth control and

differentiation, thereby enhancing malignant transformation. This form of genetic instability

has been regarded as ‘microsatellite mutator phenotype’ and the tumorigenesis follows the so-

11

Chapter 1 |

called microsatellite instability (MIN) or ‘mutator’ pathway. In the majority of sporadic

colorectal cancers the carcinogenesis pathway is characterized by chromosomal instability

(CIN) as evidenced by loss of heterozygosity and referred to as the ‘suppressor’ pathway. The

two distinct genetic instability pathways follow a similar morphological pathway of colorectal

carcinogenesis as described by Kinzler and Vogelstein, the adenoma-carcinoma sequence.20

The premalignant property of adenomas in HNPCC is clinically demonstrated by the

beneficiary results of colonoscopy with polypectomy in the study of Jarvinen et al.21 The

adenomas in the colon of HNPCC patients may occur on a sporadic basis but evidence also

supports the theory that the adenomas are the consequence of dysfunctional DNA mismatch

repair. Irrespective of the initiation process, adenomas in HNPCC seem to have an accelerated

transformation rate to carcinomas. The change in morphology and histology is a reflection of

progressive acquisition of a variety of genomic alternations and a disbalance in proliferation

and apoptosis in neoplastic cells leading to malignancy. The repertoire of genes which, when

mutated, contribute directly to the oncogenic properties and progression of hereditary

nonpolyposis colorectal tumors are partially known. As mentioned above, instability may

occur in coding microsatellites, resulting in frameshift mutations of corresponding genes,

leading to truncated and thus useless proteins. Genes, such as TGFβRII and Bax, have been

demonstrated to be specifically altered in MMR deficient colorectal cancer cells.22,23 The

target gene mutation profile differs between colorectal and endometrial cancers with

mismatch repair deficiency, illustrating the complexity of the disease. Beside the target genes,

the differences between sporadic and HNPCC-specific mutated regulating genes leading to

altered proliferation and apoptosis and subsequently to tumor progression remain to be

clarified.

Preventing cancer in HNPCC is primarily focused on applying genetic screening and

identifying subjects at risk for a variety of life-threatening or otherwise debilitating disease

prior to onset of symptoms or other clinicopathologic manifestations of disease. To modify

the cancer risk, these subjects can subsequently be included in screening programs. Due to the

range of target organs at risk in HNPCC preventive interventions are trivial. Jarvinen et al

have shown that colonoscopies with polypectomy at 3-year intervals and subsequent

polypectomies result in a reduction in the incidence of colorectal tumors and in a significant

survival advantage.21 With regard to screening for endometrial and ovarian cancer comparable

12

Introduction |

data are not available and the effectiveness has not been shown. The Dutch Working Party for

Hereditary Cancer (STOET) has recommended annual gynecologic screening for all female

HNPCC patients with vaginal ultrasound and serum CA125, starting at 25 years of age.

Persons with HNPCC are ideal candidates for cancer prevention strategies that attempt to

modulate their risk; chemoprevention offers one such strategy. Pharmacologic prevention

offers potential benefits: 1) modulation of the need for surgical prophylaxis toward a less

extensive or delayed procedure, 2) less frequent demands for target organ surveillance, 3)

greater personal control over cancer fates and interventive options for those at genetic cancer

risk, 4) the potential for systemic or multi-site modulation of cancer risk. The nonsteriodal

anti-inflammatory drug (NSAID) sulindac, inhibitor of cyclooxygenase (COX), has been

shown to reduce the risk of developing adenocarcinoma of the colon, induce regression and

prevent the development of adenomatous polyps in the colorectum in subjects with familial

adenomatous polyposis (FAP). How sulindac or other NSAIDs exert these effects is not fully

understood. Sulindac has been reported to display profound antiproliferative effects, to alter

the cell cycle distribution, and to induce apoptosis in cell lines and in vivo.24-26 Potential

biomarkers to predict treatment outcome are besides mucosal prostaglandin levels (result of

inhibition of COX) epithelial apoptosis and proliferation.

Outline of this thesis

The first part of this thesis concentrates on further elucidating the carcinogenesis of HNPCC-

related colorectal and endometrial cancer. The second part addresses the care of HNPCC

patients by evaluating an existing endometrial screening program and studying the possible

role for chemopreventive treatment with NSAID sulindac.

As mentioned above colorectal cancers in HNPCC have a right predominance, the majority

occurring proximal of the splenic flexure. In Chapter 2 we investigated whether this proximal

preponderance is due to a proximal preponderance of adenomas or (also) due to differences in

transformation rates from adenoma to cancer in the distal and proximal colon. In Chapter 3

the clinical consequences of simultaneous inheritance of two gene mutations are reported. The

altered carcinogenesis is explored and the complexity of the preventive and curative care in

13

Chapter 1 |

this unfortunate situation is addressed. Adenomas are considered the premalignant lesions of

colorectal cancer in HNPCC, complying with the Vogelstein adenoma-carcinoma sequence of

tumorigenesis. Recently, it was shown that a small percentage of hyperplastic polyps exhibit

microsatellite instability and it is suggested that these are the precursor lesion for

microsatellite instable sporadic colorectal cancers. Whether hyperplastic polyps are possible

premalignant lesions in HNPCC is examined and documented in Chapter 4.

The tumorigenesis in HNPCC differs from sporadic colorectal cancer at an early stage. In

Chapter 5, we studied whether this difference could be explained from disparities in

expression of several cell cycle and apoptosis-related proteins in relation to proliferation and

apoptosis in HNPCC and sporadic adenomas. Furthermore, we studied endometrial cancers, in

Chapter 6, in a similar immunohistochemical manner to identify differences in the

carcinogenetic pathways of HNPCC and sporadic endometrial cancers. In HNPCC, women

have an increased cumulative life time risk for endometrial and ovarian cancer. Therefore

female members of HNPCC-families are offered a gynecologic examination, a transvaginal

ultrasound and serum level CA 125 analysis to be performed annually. However, unlike

colonoscopy, the effectiveness of gynecologic surveillance procedures has not been shown in

either prospective or retrospective studies. Chapter 7 describes a study which evaluates our

10 years experience in endometrial and ovarian cancer screening in women belonging to

HNPCC-families to determine whether our present screening method achieves the aspired

prevention or early detection of gynecologic cancers.

With the increased risk of cancer and the early age at which tumors are diagnosed a search for

more than early detection, namely (primary) prevention is ongoing. The ultimate goal of the

present thesis was- with the knowledge of the carcinogenesis of HNPCC colorectal cancers- to

explore the potential role of sulindac in HNPCC chemoprevention. So the effects of sulindac

were evaluated in HNPCC patients using surrogate end-points for cancer risk including

epithelial cell proliferative activity, degree of apoptosis and expression in normal colonic

epithelium of proliferation-, apoptosis- and cell cycle-involved genes. The results of a

randomised, double-blind, placebo-controlled cross-over study in ascertained MMR gene

mutation carriers and subjects with more than 50% risk to be MMR gene mutation carriers is

described in chapter 8. Finally the results of the thesis are summarized and discussed in

chapter 9.

14

Introduction |

References

1. Lynch HT, Smyrk T, Lynch JF. Molecular genetics and clinical-pathology features of hereditary nonpolyposis colorectal carcinoma (Lynch syndrome): historical journey from pedigree anecdote to molecular genetic confirmation. Oncology 1998;55:103-8.

2. Vasen HF, Mecklin JP, Khan PM et al. The International Collaborative Group on Hereditary Non-Polyposis Colorectal Cancer (ICG-HNPCC). Dis Colon Rectum 1991;34:424-5.

3. Vasen HF, Watson P, Mecklin JP et al. New clinical criteria for hereditary nonpolyposis colorectal cancer (HNPCC, Lynch syndrome) proposed by the International Collaborative group on HNPCC. Gastroenterology 1999;116:1453-6.

4. Peltomaki P, Aaltonen LA, Sistonen P et al. Genetic mapping of alocus predisposing to human colorectal cancer. Science 1993;260:810-2.

5. Lindblom A, Tannergard P, Werelius B et al. Genetic mapping of a second locus predisposing to hereditary non-polyposis colon cancer. Nat Genet 1993;5:279-82.

6. Cannavo E, Marra G, Sabates-Bellver J et al. Expression of the MutL homologue hMLH3 in human cells and its role in DNA mismatch repair. Cancer Res. 2005:65:10759-66

7. Liu T, Yan H, Kuismanen S et al. The role of hPMS1 and hPMS2 in predsposing to colorectal cancer. Cancer Res. 2001:61:7798-7802.

8. Jagmohan-Changur S, Poikonen T, Vilkki S et al. EXO1 variants occur commonly in normal population: Evidence against a role in hereditary nonpolyposis colorectal cancer. Cancer Res 2003;63:154-8.

9. Prolla TA, Pang Q, Alani E et al. MLH1, PMS1, and MSH2 interactions during the initiation of DNA mismatch repair in yeast. Science 1994;265:1091-3.

10. Acharya S, Wilson T, Gardia S et al. hMSH2 forms specific mispair-binding complexes wit hMSH3 and hMSH6. Proc Natl Acad Sci USA. 1996;93:13629-34.

11. Aaltonen LA, Peltomaki P, Leach FS et al. Clues to the pathogenesis of familial colorectal cancer. Science 1993;260:810-2.

12. Thibodeau SN, Bren G, Schaid D. Microsatellite instability in cancer of the proximal colon. Science 1993;260:816-9.

13. Boland CR, Thibodeau SN, Hamilton SR et al. A National Cancer Institute Workshop on Microsatellite Instability for cancer detection and familial predisposition: development of international criteria for the determination of microsatellite instability in colorectal cancer. Cancer Res 1998;58:5248-57.

14. Tomlinson I, Halford S, Aaltonen L et al. Does MSI-low exist? J Pathol 2002;197:6-13.

15

Chapter 1 |

15. Kuismanen SA, Holmberg MT, Salovaara R et al. Genetic and epigenetic modification of MLH1 accounts for a major share of microsatellite-unstable colorectal cancers. Am J Pathol 2000;156:1773-9.

16. Dietmaier W, Wallinger S, Bocker T et al. Diagnostic microsatellite instability: definition and correlation with mismatch repair protein expression. Cancer Res 1997;57:4749-56.

17. Lindor NM, Burgart LJ, Leontovich O et al. Immunohistochemistry versus microsatellite instability testing in phenotyping colorectal tumors. J Clin Oncol 2002;20:1043-8.

18. Bhattacharyya NP, Skandalis A, Ganesh A et al. Mutator phenotypes in human colorectal carcinoma cell lines. Proc Natl Acad Sci U S A 1994;91:6319-23.

19. Parsons R, Myeroff LL, Liu B et al. Microsatellite instability and mutations of the transforming growth factor beta type II receptor gene in colorectal cancer. Cancer Res 1995;55:5548-50.

20. Kinzler KW, Vogelstein B. Lessons from hereditary colorectal cancer. Cell 1996;87:159-70

21. Jarvinen HJ, Aarnio M, Mustonen H et al. Controlled 15-year trial on screening for colorectal cancer in families with hereditary nonpolyposis colorectal cancer. Gastroenterology 2000;118:829-34.

22. Markowitz SD, Wang JY, Myeroff LL et al. Inactivation of the type II TGF-beta receptor in colon cancer cells with microsatellite instability. Science 1995;268:1336-8.

23. Ouyang H, Furukawa T, Abe T et al. The BAX gene, the promoter of apoptosis, is mutated in genetically unstable cancers of the colorectum, stomach, and endometrium. Clin Cancer Res 1998;4:1071-4.

24. Giardiello FM, Hamilton SR, Krush AJ et al. Treatment of colonic and rectal adenomas with sulindac in familial adenomatous polyposis. N Engl J Med 1993;328:1313-6.

25. Labayle D, Fischer D, Vielh P et al. Sulindac causes regression of rectal polyps in familial adenomatous polyposis. Gastroenterology 1991;101:635-9.

26. Ladenheim J, Garcia G, Titzer D et al. Effect of sulindac on sporadic colonic polyps. Gastroenterology 1995;108:1083-7.

16

2

Proximal adenomas in hereditary

nonpolyposis colorectal cancer are

prone to rapid malignant

transformation

F E M Rijcken1, H Hollema2, J H Kleibeuker1

Department of Gastroenterology1 and Pathology2, University

Medical Center Groningen, The Netherlands.

Gut 2003;52:898-9.

Chapter 2 |

Abstract

Background - Hereditary nonpolyposis colorectal cancer (HNPCC) is thought to arise

from adenomas. HNPCC mostly occurs in the proximal colon. We investigated whether

this proximal preponderance is due to a proximal preponderance of adenomas or (also)

due to differences in transformation rates from adenomas to cancer between distal and

proximal colon.

Methods - 100 HNPCC adenomas were evaluated and compared to 152 sporadic

adenomas for location, size and dysplasia. 25 adenomas from patients with a known

mismatch repair gene (MMR) mutation were stained for expression of MLH1 and

MSH2.

Results - HNPCC adenomas were more often located proximally (50% vs. 26%, p=0.018)

and were smaller in comparison to sporadic adenomas. They were similarly dysplastic.

However, all proximal HNPCC adenomas ≥5 mm were highly dysplastic compared to

17% of the larger proximal sporadic polyps (p<0.001). They were also more often highly

dysplastic than larger distal HNPCC adenomas (p<0.001). Small HNPCC adenomas

were, except for their location, not different from sporadic ones. 15 of the 25 'known-

mutation' adenomas showed loss of expression of either MLH1 or MSH2. The 10

adenomas with expression were all small and low-grade dysplastic.

Conclusion - HNPCC adenomas are located mainly in the proximal colon. The

development to high-grade dysplasia is more common in proximal HNPCC adenomas

than in distal ones, indicating a faster transformation rate from early adenoma to cancer

in the proximal colon. MMR-gene malfunction probably does not initiate adenoma

development but is present at a very early stage of tumorigenesis and heralds the

development of high-grade dysplasia.

18

Proximal adenomas |

For the first time, researchers have obtained indisputable data supporting the usage of

contemporary colorectal cancer prevention procedures in hereditary nonpolyposis colorectal

cancer (HNPCC) syndrome.1 These data reveal a significant patient survival advantage and an

incidence reduction of colorectal tumors due to colonoscopic screening and polypectomies.

This complies with the accepted adenoma-carcinoma sequence theory that states that

adenomas are a precursor in the tumorigenesis of malignant HNPCC lesions.2-4

The malignant lesions, often occurring at relatively young age, are well described in HNPCC.

They are located predominantly in the proximal part of the colon, and there is a high incidence

of synchronous and metachronous cases. Microscopically the tumors are characterized by a

Crohn’s like lymphoid reaction, a mucinous component and by poor differentiation.5-10

Publications on the adenomas in HNPCC are less consistent. When calculating the average

distribution of HNPCC adenomas mentioned in the literature, forty-five percent (range: 27-

70%) is located in the proximal colon.1,9,11-14 Some reported an obvious propensity for right-

sided neoplastic lesions while others observed a distribution of adenomas in HNPCC patients

similar to that in the general population. 13,15-17

The adenoma-carcinoma sequence in HNPCC seems to be accelerated. This is especially

illustrated by the relatively frequent occurrence of cancers within the first few years after a

"clean" colon had been confirmed by colonoscopy.14-15,18-19 In addition, several authors

reported that HNPCC adenomas frequently have a villous component and high-grade

dysplasia, two assumed markers of increased risk to develop cancer.9,11,14,20 However, whether

this is a uniform feature for all HNPCC adenomas at every location in the colon has yet to be

determined. The ratio of proximal to distal cancers in HNPCC (7:3) is higher than the reported

ratio for adenomas (4:5). It thus seems that not all HNPCC adenomas have an increased risk

for malignant transformation and that there are regional differences in this respect.

In order to examine the above-mentioned issues, we compared the adenomas resected from

HNPCC patients to sporadic adenomas. More importantly, we investigated whether

differences exist between proximal, distal and rectal HNPCC adenomas.

19

Chapter 2 |

Methods

Patient characteristics

According to the prevention guidelines of the International Collaborative Group on HNPCC

(ICG-HNPCC) people fulfilling the Amsterdam criteria and/or having a germline mismatch

repair gene mutation should undergo a colonoscopy every two years starting at the age of

25.21-23 At the University Hospital of Groningen 136 persons, belonging to 47 families,

participate in a surveillance program. We included all subjects with a positive colonoscopy in

our study group: 46 (24 male and 22 female) patients with a median age of 50 (range 25-78)

years at polypectomy. Sixty-nine colonoscopies, a mean of 1.6 per person (range 1-5), yielded

100 adenomatous polyps (55 from men and 45 from women) in a 12-year period, from 1988

to 2000. Four polyps found in a subtotal colectomy specimen were also included. Seven

persons had previously been diagnosed with cancer and had part of their colon resected.

The data were compared to the findings from a control group consisting of the sporadic

adenomas consecutively removed during sigmoido- and colonoscopy at the Endoscopy Center

of the University Hospital of Groningen in 1997. Lesions from patients with a strong positive

family history of colorectal cancer, or patients having ulcerative colitis, Crohn’s disease or

familial adenomatous polyposis (FAP) were not included. According to protocol patients with

an adenoma detected at sigmoidoscopy should subsequently have a colonoscopy: except for

two patients, the entire large bowel was inspected in each person.24 The group of sporadic

adenomas consisted of 152 adenomas and had a similar male-female ratio as in the HNPCC

group, 84 lesions from men and 68 from women. The average age at polypectomy was 64

(range 24-90) years.

Location

The location of the adenomas was retrieved from endoscopy or pathology reports. The

caecum and the ascending and transverse colon are regarded as the proximal or right-sided

colon while the descending and the sigmoid colon are referred to as the distal or the left-sided

colon. The third location of the adenomas was the rectum.

20

Proximal adenomas |

Histological examination

Thin slides, 3µm, were made of each formalin-fixed paraffin-embedded polyp and stained

with hematoxylin-eosin (HE). Two authors, Rijcken and Hollema, reviewed and scored three

characteristics of the adenomas: 1. Size of the polyp - the microscopic measurement of the

polyp’s circumference, < 5mm (small) and ≥ 5mm (large); 2. Histological subtype- tubular or

having more than a 25% villous component; 3. Grade of dysplasia- low or high (WHO

guidelines).25

The adenomas, 25 in total, which were removed from persons with a known mutation or

belonging to a family with a known mutation were further analyzed for mismatch repair

(MMR) protein status by immunohistochemistry. Monoclonal mouse antibodies against

MLH1 (clone G168-728, Pharmingen, San Diego, USA) and MSH2 (Ab-2, Calbiochem, San

Diego, USA) protein products were used. The paraffin sections (3µm) were fixed onto 3-

aminopropyltriethoxysilane (APES, Sigma-Aldrich, Diesenhofen, Germany) coated slides,

stretched for 30 minutes at 60°C and dried overnight at 37°C. The sections were

deparaffinized in xylene (2x 10 minutes) and rinsed in 100% alcohol. The optimal antibody-

antigen reaction was obtined by immersing the section in 200µl blocking reagent (2% block

and 0.2% SDS in maleic acid, pH 6.0 (Boehringer Mannheim, Germany)) and using a high

pressure cooker for 3 sessions of 5 minutes at 115°C alternating with 5 minutes of incubation

in a humid environment. After cooling for the third time, endogenous peroxidase activity was

quenched by incubation with 30% H2O2 in phosphate buffered sulfate (PBS) for 30 minutes.

Following thorough washing in PBS, the sections were immersed with the specific antibody in

PBS with 1% bovine serum albumin (BSA), at a dilution of 1:500 for MLH1 and 1:100 for

MSH2 antibody, for one hour. Subsequently, the sections were washed three times with PBS

and consecutively incubated for 30 minutes with rabbit antimouse peroxidase and goat

antirabbit peroxidase diluted (1:50) in PBS-1% BSA. The sections were submerged for 10

minutes in a solution of 50 mg 3’-3’diaminobenzidine in PBS and 50 mg of imidazol with

30% H2O2, used as substrate of peroxidase. After rinsing with demi water, the sections were

counterstained with haematoxylin, washed with running water and dehydrated with graded

alcohol, dried and covered with a slide.

21

Chapter 2 |

Statistical analyses

Statistical comparisons between HNPCC and sporadic adenomas and within the two groups

were made using the Mann-Whitney test when comparing a characteristic (size, type and

dysplasia) with two variables and the corrected Chi-squared test was used in the comparisons

of a characteristic (location) with three variables. A p-value of less than 0.05 was considered

to be statistically significant.

Results

Taking the two entire groups of adenomas in consideration (table 1), a significant difference

can be observed between the location, size and histology of HNPCC and that of sporadic

adenomas (p= 0.018, p< 0.001 and p= 0.010, resp.). HNPCC adenomas had a proximal

propensity (50% vs 26% of the sporadic adenomas). The adenomas showed a wide range of

sizes in both groups. The median size of HNPCC adenomas was 2.0 mm (range 0.5-34mm),

while the sporadic adenomas had a medium size of 5.5mm (range 0.5- 45mm). The HNPCC

adenomas were more often tubular in comparison to the sporadic adenomas. Even though the

HNPCC adenomas were more often small and tubular they were similarly dysplastic as the

larger and more often villous sporadic adenomas.

Table 1. The four characteristics of HNPCC and sporadic adenomas by number and percentage

CC S i

n (%) n (%)

Location Proximal 47 (50) 39 (26) Distal 29 (30) 72 (47) P= 0.018 Rectum 19 (20) 41 (27) Size < 5 mm 70 (70) 67 (44) ≥ 5 mm 30 (30) 85 (56)

p< 0.001

Type Tubular 79 (79) 97 (64) Tubulovillous 21 (21) 55 (36)

p= 0.010

Dysplasia Low-grade 68 (68) 114 (76) High-grade 32 (32) 38 (25)

p= 0.567

22

Proximal adenomas |

The proximal propensity was more evident in highly dysplastic HNPCC adenomas: high-

grade dysplastic HNPCC adenomas were more often (55%) located proximal to the splenic

flexure while only 6 (15%) of the high-grade sporadic polyps were right-sided (p< 0.001).

In the proximal colon the HNPCC adenomas were smaller in comparison to sporadic polyps

(p=0.025) but they were more often highly dysplastic (36% vs 13%, p<0.020) (figure 1).

When comparing size as well as dysplasia at one location (table 2A&B), the difference

between the HNPCC and sporadic adenomas was clearly apparent. Especially larger HNPCC

polyps, ≥ 5mm, in the proximal colon were more often highly dysplastic (p<0.001): in fact, all

HNPCC adenomas equal to or larger than 5 mm were highly dysplastic. In the sporadic

lesions this was not observed: only 25% of the large sporadic polyps at this site were high-

grade dysplastic.

Table 2A. Dysplasia of HNPCC adenomas by location and size

Proximal Distal Rectum 47 29 19

Dysplasia low high low high Low high 30 17 25 4 9 10 Size < 5 mm 30 6 17 2 8 3 ≥ 5 mm 0 11 8 2 2 6

* Difference between ≥ 5mm adenomas in the proximal and distal colon and between proximal colon and rectum: p value=0.009 and p=0.07, respectively.

Table 2B. Dysplasia of sporadic adenomas by location and size

Proximal Distal Rectum 39 72 41

Dysplasia low High low high low high 34 5 45 27 33 8 Size < 5 mm 19 2 22 6 17 1 ≥ 5 mm 15 3 23 21 16 7

* Difference between ≥ 5mm adenomas in the proximal and distal colon and between proximal colon and rectum: Both p values are not significant.

23

Chapter 2 |

PROXIMAL ADENOMAS

0

10

20

30

40

50

HNPCC

Sporadic

Low HighDysplasia

Size

(mm

)

Figure 1 Distribution of proximal HNPCC and sporadic adenomas by size and dysplasia

(The line illustrates the median size.)

In the rectum, large HNPCC adenomas were also more often highly dysplastic in comparison

to sporadic adenomas ≥ 5 mm. However, this difference was only borderline significant (p=

0.084). The HNPCC adenomas in the distal colon were smaller than distal sporadic lesions

(66% vs 39% <5mm, p= 0.047), mostly tubular (72% vs 59%, not significant) and nearly

always low-grade dysplastic (86% vs 63%, p=0.020).

A significant difference was not only observed between the HNPCC and sporadic groups but

also between the large HNPCC adenomas at the three locations. The large proximal HNPCC

adenomas were more often highly dysplastic (100% were highly dysplastic) than the HNPCC

adenomas in the distal colon (22%, p=0.001), whereas a borderline significant difference was

detected between proximal and rectal adenomas (75% highly dysplastic, p=0.07). The large

HNPCC adenomas in the rectum were also significantly more often highly dysplastic in

comparison to the HNPCC adenomas in the distal colon (p< 0.03).

24

Proximal adenomas |

Eleven adenomas belonged to patients with a proven MLH1 mutation while 11 adenomas

were removed from persons with a proven MSH2 mutation. Three adenomas were from two

persons who had not undergone genetic testing but in their families an MLH1 mutation had

been detected. In total, 15 adenomas, 8 low-grade and 7 high-grade dysplastic, showed loss of

expression of either MLH1 or MSH2. The three adenomas from the two persons who had not

undergone genetic testing all showed loss of MLH1 staining, corresponding to the known

mutation in their kindreds. Eight adenomas with loss of MLH1 expression were located in the

proximal colon and 2 were situated in the rectum. Loss of MSH2 expression was observed in

3 distal, 1 proximal and 1 rectal polyp. The 10 adenomas, all from proven carriers, that had

normal expression of MMR proteins were small and low-grade dysplastic. Three of these

adenomas were located in the proximal colon, four in the distal colon and the remaining three

were located in the rectum.

Discussion

Our results differ, as will be explained below, from previous studies concerning the

morphological descriptions of adenomas in persons at risk for hereditary nonpolyposis

colorectal cancer. The results support the accelerated adenoma-carcinoma sequence theory of

carcinogenesis in HNPCC but at the same time strongly suggest that lesions at different

locations, proximal or distal in the colon or in the rectum, behave differently in this respect.

More specifically, similarly to the high frequency of hereditary nonpolyposis colorectal

carcinomas proximal to the splenic flexure, HNPCC adenomas have a right preponderance,

though less so than the carcinomas, and these right-sided adenomas are more prone to

malignant conversion in comparison to the left-sided adenomas.

The large group of HNPCC adenomas included in this study is representative for the

Groningen HNPCC population, as they are all the available benign neoplastic lesions

consecutively removed during a twelve year period. The majority of the HNPCC adenomas

were obtained during surveillance examinations, while the sporadic adenomas were acquired

during a colonoscopy to diagnose clinical symptoms such as rectal blood loss, diarrhea and

abdominal pain. The groups are thus not readily comparable and this probably explains why

the HNPCC adenomas were smaller, more often tubular and were not more dysplastic than the

sporadic ones, findings that are in contrast with what has been reported in the

25

Chapter 2 |

literature.9,11,20,26-27 It is noticeable, however, that the small and tubular HNPCC adenomas

were not less dysplastic than the larger and more often villous sporadic cases, as would be

expected.27 To reduce the discrepancy arisen by comparing adenomas obtained through

surveillance and diagnostic colonoscopies we also evaluated only the index HNPCC

adenomas, those adenomas found during the first colonoscopy (data not shown). The

characteristics of these index neoplastic lesions did not differ from the characteristics of all

the HNPCC adenomas reported in the present paper. In accordance to Green's observation,

they were mostly small and tubular.28 The same team of gastroenterologists did all

colonoscopies. The number of small, <5mm, HNPCC and sporadic adenomas in all sections

of the colon illustrates the thorough endoscopic search in both groups and through the entire

colon. Nevertheless, it can not be excluded that even greater care was taken in the

examination of the colons from HNPCC subjects than of those from other patients.

The prevalence and grade of dysplasia of adenomas in the general population differ per

gender, age and geographic region.27,29-35 Any bias was minimized by including the same male

to female ratio of adenomas in both groups. A sufficient amount of age matched subjects,

however, was not available. The histological characteristics of our control group are similar to

the results of Griffioen et al, also a Dutch clinical investigation.36 The results from the control

group also correspond to findings of several other publications.37-38 However, yet other

reports, mainly autopsy and surveillance studies, favor a substantially higher frequency of

low-grade dysplastic sporadic polyps and a lower frequency of villous lesions.17,27,30,33 The

proportion of proximal adenomas varies in the literature between 11% and 40% (Griffioen et

all, 17%).

The essential role of the proximal colon in the pathogenesis of HNPCC is evident from

numerous reports of the high incidence of right-sided (interval) carcinomas in HNPCC

patients.1,6-8,10,12,14,39-42 In accordance with this we found that HNPCC adenomas are located

more often in the proximal colon than sporadic ones. The proximal propensity suggests an

alteration in initiation of neoplastic growth in HNPCC in comparison to the general

population. A recent report on the presence of microsatellite instability in both hyperplastic

and dysplastic aberrant crypt foci in colons from HNPCC-patients also suggests a possible

role of MMR dysfunction in the initiation of neoplastic lesions in HNPCC.43 However, Leach

26

Proximal adenomas |

among others, proposed that the adenomas in HNPCC develop on a sporadic basis only to

provide a substrate for defective DNA mismatch repair genes.44

We therefore tried to substantiate either the Leach theory or the 'initiation' proposal by

performing immunohistochemistry (IHC) for gene products of MLH1 and MSH2 on the

adenomas from persons with a known mutation. In accordance with most reports, the absence

of immunohistochemical staining was seen in two-thirds of the HNPCC adenomas, i.e. in 44%

of the low-grade and in all high grade dysplastic adenomas.45-50 Thus, our data strongly

suggest that DNA repair deficiency is not responsible for the initiation of an adenoma but

determines the subsequent progression of the lesion. Although MMR dysfunction is not the

first event, it is surely a very early one in the tumorigenesis of HNPCC lesions and it heralds

development to high-grade dysplasia.

Our results illustrate that the proposed accelerated adenoma-carcinoma sequence in HNPCC is

probably site-specific i.e. limited to the proximal colon. While the group of HNPCC

adenomas as a whole did not exhibit features of increased susceptibility to malignant

conversion in comparison to sporadic adenomas, a difference was observed within the group

of adenomas proximal to the splenic flexure. The large (≥5mm) HNPCC adenomas in the

proximal bowel were all highly dysplastic suggesting more advancement in the malignant

transformation than sporadic polyps of similar size at this location. This and the fact that 9 of

the 15 HNPCC adenomas that showed loss of protein expression were located in the proximal

colon emphasize the predilection of the proximal colon for tumorigenesis in HNPCC.

Characteristics of HNPCC adenomas at specific sites in the large bowel have had limited

attention so far as authors have reported results on HNPCC adenomas only in general. On the

other hand sporadic adenomas have been described per location. The National Polyp Study

observed an increased frequency of high-grade dysplasia in adenomas located distal to the

splenic flexure, but attributed it mainly to increased size and villous component rather than to

location per se.27 Nusko and colleagues reported that sporadic right-sided adenomas have a

lower risk to become malignant in comparison to sporadic left-sided adenomas. 51 The reasons

for the reverse situation in HNPCC are still unclear.25,52-54

In conclusion, half of the HNPCC adenomas are located in the proximal colon, but this does

not fully explain the proximal propensity of cancers. Our data show that development to high-

27

Chapter 2 |

grade dysplasia is more common in proximal HNPCC adenomas than in distal ones,

indicating a faster transformation rate from early adenoma to cancer in the proximal colon.

MMR-gene malfunction probably does not initiate adenoma development but is present at a

very early stage of tumorigenesis and seems to herald the development of high-grade

dysplasia. At this time there is no ready explanation for this site-specific behavior of HNPCC-

lesions or for the susceptibility of the epithelial cells in the proximal colon for somatic MMR-

gene mutations.

28

Proximal adenomas |

References

1. Jarvinen HJ, Aarnio M, Mustonen H et al. Controlled 15-year trial on screening for colorectal cancer in families with hereditary nonpolyposis colorectal cancer. Gastroenterology 2000;118:829-34.

2. Fearon ER, Vogelstein B. A genetic model for colorectal tumorigenesis. Cell 1990;61:759-67.

3. Watne AL. Colon polyps. J Surg Oncol 1997;66:207-14.

4. Roncucci L, Stamp D, Medline A et al. Identification and quantification of aberrant crypt foci and microadenomas in the human colon. Hum Pathol 1991;22:287-94.

5. Vasen HF, Mecklin JP, Khan PM et al. The International Collaborative Group on Hereditary Non-Polyposis Colorectal Cancer (ICG-HNPCC). Dis Colon Rectum 1991;34:424-5.

6. Messerini L, Mori S, Zampi G. Pathologic features of hereditary non-polyposis colorectal cancer. Tumori 1996;82:114-6.

7. Lynch HT, Smyrk TC, Watson P et al. Genetics, natural history, tumor spectrum, and pathology of hereditary nonpolyposis colorectal cancer: an updated review. Gastroenterology 1993;104:1535-49.

8. Lynch HT, Smyrk T. Hereditary nonpolyposis colorectal cancer (Lynch syndrome). An updated review. Cancer 1996;78:1149-67.

9. Jass JR, Smyrk TC, Stewart SM et al. Pathology of hereditary non-polyposis colorectal cancer. Anticancer Res 1994;14:1631-4.

10. Jass JR. Colorectal adenomas in surgical specimens from subjects with hereditary non-polyposis colorectal cancer. Histopathology 1995;27:263-7.

11. Jass JR, Stewart SM. Evolution of hereditary non-polyposis colorectal cancer. Gut 1992;33:783-6.

12. Lanspa SJ, Lynch HT, Smyrk TC et al. Colorectal adenomas in the Lynch syndromes. Results of a colonoscopy screening program. Gastroenterology 1990;98:1117-22.

13. Jass JR, Pokos V, Arnold JL et al. Colorectal neoplasms detected colonoscopically in at-risk members of colorectal cancer families stratified by the demonstration of DNA microsatellite instability. J Mol Med 1996;74:547-51.

14. Vasen HF, Taal BG, Nagengast FM et al. Hereditary nonpolyposis colorectal cancer: results of long-term surveillance in 50 families. Eur J Cancer 1995;31A:1145-8.

15. Lanspa SJ, Jenkins JX, Cavalieri RJ et al. Surveillance in Lynch syndrome: how aggressive? Am J Gastroenterol 1994;89:1978-80.

29

Chapter 2 |

16. Lynch HT, Smyrk T, Jass JR. Hereditary nonpolyposis colorectal cancer and colonic adenomas: aggressive adenomas? Semin Surg Oncol 1995;11:406-10.

17. Jass JR, Young PJ, Robinson EM. Predictors of presence, multiplicity, size and dysplasia of colorectal adenomas. A necropsy study in New Zealand. Gut 1992;33:1508-14.

18. Jarvinen HJ, Mecklin JP, Sistonen P. Screening reduces colorectal cancer rate in families with hereditary nonpolyposis colorectal cancer. Gastroenterology 1995;108:1405-11.

19. Muto T, Bussey HJ, Morson BC. The evolution of cancer of the colon and rectum. Cancer 1975;36:2251-70.

20. Ponz-de LM, Della CG, Benatti P et al. Frequency and type of colorectal tumors in asymptomatic high-risk individuals in families with hereditary nonpolyposis colorectal cancer. Cancer Epidemiol Biomarkers Prev 1998;7:639-41.

21. Hodgson SV, Bishop DT, Dunlop MG et al. Suggested screening guidelines for familial colorectal cancer. J Med Screen 1995;2:45-51.

22. Vasen HF, Mecklin JP, Watson P et al. Surveillance in hereditary nonpolyposis colorectal cancer: an international cooperative study of 165 families. The International Collaborative Group on HNPCC. Dis Colon Rectum 1993;36:1-4.

23. Vasen HF, van-Ballegooijen M, Buskens E et al. A cost-effectiveness analysis of colorectal screening of hereditary nonpolyposis colorectal carcinoma gene carriers. Cancer 1998;82:1632-7.

24. Schoen RE, Corle D, Cranston L et al. Is colonoscopy needed for the nonadvanced adenoma found on sigmoidoscopy? The Polyp Prevention Trial. Gastroenterology 1998;115:533-41.

25. Hamilton SR. The adenoma-adenocarcinoma sequence in the large bowel: variations on a theme. J Cell Biochem Suppl 1992;16G:41-6.

26. Mecklin JP, Sipponen P, Jarvinen HJ. Histopathology of colorectal carcinomas and adenomas in cancer family syndrome. Dis Colon Rectum 1986;29:849-53.

27. O'Brien MJ, Winawer SJ, Zauber AG et al. The National Polyp Study. Patient and polyp characteristics associated with high-grade dysplasia in colorectal adenomas. Gastroenterology 1990;98:371-9.

28. Green SE, Chapman PD, Burn J et al. Clinical impact of colonoscopic screening in first-degree relatives of patients with hereditary non-polyposis colorectal cancer. Br J Surg 1995;82:1338-40.

29. Eide TJ. The age-, sex-, and site-specific occurrence of adenomas and carcinomas of the large intestine within a defined population. Scand J Gastroenterol 1986;21:1083-8.

30

Proximal adenomas |

30. Rickert RR, Auerbach O, Garfinkel L et al. Adenomatous lesions of the large bowel: an autopsy survey. Cancer 1979;43:1847-57.

31. Johannsen LG, Momsen O, Jacobsen NO. Polyps of the large intestine in Aarhus, Denmark. An autopsy study. Scand J Gastroenterol 1989;24:799-806.

32. Bombi JA. Polyps of the colon in Barcelona, Spain. An autopsy study. Cancer 1988;61:1472-6.

33. Vatn MH, Stalsberg H. The prevalence of polyps of the large intestine in Oslo: an autopsy study. Cancer 1982;49:819-25.

34. Rubio CA, Saito Y, Watanabe M et al. Non-polypoid colorectal neoplasias: a multicentric study. Anticancer Res 1999;19:2361-4.

35. Gaglia P, Atkin WS, Whitelaw S et al. Variables associated with the risk of colorectal adenomas in asymptomatic patients with a family history of colorectal cancer. Gut 1995;36:385-90.

36. Griffioen G, Bosman FT, Verspaget HW et al. Colorectal adenomas: clinical and morphological aspects. A review of 166 polyps from 124 Dutch patients. Anticancer Res 1989;9:1685-9.

37. Johnson DA, Gurney MS, Volpe RJ et al. A prospective study of the prevalence of colonic neoplasms in asymptomatic patients with an age-related risk. Am J Gastroenterol 1990;85:969-74.

38. Cannon AL, Bishop DT, Samowitz W et al. Colonic polyps in an unselected population: prevalence, characteristics, and associations. Am J Gastroenterol 1994;89:827-31.

39. Jass JR, Stewart SM, Stewart J et al. Hereditary non-polyposis colorectal cancer--morphologies, genes and mutations. Mutat Res 1994;310:125-33.

40. Lynch HT, Watson P, Lanspa SJ et al. Natural history of colorectal cancer in hereditary nonpolyposis colorectal cancer (Lynch syndromes I and II). Dis Colon Rectum 1988;31:439-44.

41. Mecklin JP, Jarvinen HJ. Clinical features of colorectal carcinoma in cancer family syndrome. Dis Colon Rectum 1986;29:160-4.

42. Vasen HF, Hartog-Jager FC, Menko FH et al. Screening for hereditary non-polyposis colorectal cancer: a study of 22 kindreds in The Netherlands. Am J Med 1989;86:278-81.

43. Pedroni M, Sala E, Scarselli A et al. Microsatellite instability and mismatch-repair protein expression in hereditary and sporadic colorectal carcinogenesis. Cancer Res 2001;61:896-9.

44. Leach FS, Nicolaides NC, Papadopoulos N et al. Mutations of a mutS homolog in hereditary nonpolyposis colorectal cancer. Cell 1993;75:1215-25.

31

Chapter 2 |

45. Thibodeau SN, French AJ, Roche PC et al. Altered expression of hMSH2 and hMLH1 in tumors with microsatellite instability and genetic alterations in mismatch repair genes. Cancer Res 1996;56:4836-40.

46. Dietmaier W, Wallinger S, Bocker T et al. Diagnostic microsatellite instability: definition and correlation with mismatch repair protein expression. Cancer Res 1997;57:4749-56.

47. Leach FS, Polyak K, Burrell M et al. Expression of the human mismatch repair gene hMSH2 in normal and neoplastic tissues. Cancer Res 1996;56:235-40.

48. Cawkwell L, Li D, Lewis FA, Martin I et al. Microsatellite instability in colorectal cancer: improved assessment using fluorescent polymerase chain reaction. Gastroenterology 1995;109:465-71.

49. Chaves P, Cruz C, Lage P et al. Immunohistochemical detection of mismatch repair gene proteins as a useful tool for the identification of colorectal carcinoma with the mutator phenotype. J Pathol 2000;191:355-60.

50. Marcus VA, Madlensky L, Gryfe R et al. Immunohistochemistry for hMLH1 and hMSH2: a practical test for DNA mismatch repair-deficient tumors. Am J Surg Pathol 1999;23:1248-55.

51. Nusko G, Mansmann U, Altendorf HA et al. Risk of invasive carcinoma in colorectal adenomas assessed by size and site. Int J Colorectal Dis 1997;12:267-71.

52. Bufill JA. Colorectal cancer: evidence for distinct genetic categories based on proximal or distal tumor location. Ann Intern Med 1990;113:779-88.

53. Huang J, Papadopoulos N, McKinley AJ et al. APC mutations in colorectal tumors with mismatch repair deficiency. Proc Natl Acad Sci 1996;93:9049-54.

54. Philips SF, Pemberton J, Shorter RG. The Large Intestine: Physiology, Pathophysiology, and Disease. Mayo Foundation 1991;Published by Raven Press, Ltd., New York.

32

3

CASE REPORT

Rapidly progressive adenomatous

polyposis in a patient with germline

mutations in both the APC and

MLH1 genes: the worst of two

worlds

R Scheenstra1, F E M Rijcken2, J J Koornstra2, H Hollema3,

R Fodde4, F H Menko5, R H Sijmons6, C M A Bijleveld1,

J H Kleibeuker2

Department of Paediatric Gastroenterology1,

Gastroenterology2, Department of Pathology3, Clinical

Genetics6, University Medical Center Groningen; Centre of

Human and Clinical Genetics4, LUMC; Department of

Clinical Genetics and Human Genetics5, Free University

Medical Centre Amsterdam, the Netherlands.

Gut 2003;52:898–899.

Chapter 3 |

Abstract

The two most common inherited forms of colorectal cancer are familial adenomatous

polyposis and hereditary non-polyposis colorectal cancer. Simultaneous inheritance of

both an APC gene mutation and a mismatch repair gene (for example, MLH1) mutation

has never been described. In the present case report, we report rapidly progressive

adenomatous polyposis in a 10 year old boy with a germline frame shift mutation in the

APC gene and a germline splice site mutation in the MLH1 gene. Immunohistochemical

investigations showed abnormal expression of β-catenin in early adenomas with low

grade dysplasia, attributed to the APC gene mutation. Subsequent loss of function of the

MLH1 gene, as shown by absent immunostaining of its protein in adenomas with high

grade dysplasia, may well have caused the rapid progression to high grade dysplasia in

many of the adenomas.

34

Case report |

The two most common inherited conditions predisposing to colorectal cancer are familial

adenomatous polyposis (FAP), caused by germline mutations in the APC (adenomatous

polyposis coli) tumorsuppressor gene, and hereditary non-polyposis colorectal cancer

(HNPCC), caused by germline mutations in one of the DNA mismatch repair (MMR) genes.

Together, FAP and HNPCC account for approximately 2% of colorectal cancer cases.1,2 In

this report, a case is presented of a patient manifesting a severe phenotype who had inherited

germline mutations in both the APC gene and the MMR gene MLH1.

Case report

A 10 year old boy attended the outpatient clinic for evaluation of slimy and bloody stools. His

mother had previously been diagnosed with FAP. At the age of 27 years, she underwent

prophylactic colectomy. Due to the development of extensive abdominal fibromatosis

(desmoid tumour), she died at the age of 32 years. Genetic analysis had revealed a germline

de novo frame shift mutation in exon 15 of the APC gene (a 5 bp deletion at codon 1309). The

father of the boy is healthy but has a classical family history of HNPCC and carries a germline

splice site mutation in the MLH1 gene (codon 226, G to A at position 677). Physical

examination was unremarkable. Colonoscopy revealed hundreds of polyps throughout the

colon. Some were removed and histopathological examination showed small tubular

adenomas with low grade dysplasia. Six months later endoscopic and histopathological

examinations were repeated, again showing multiple tubular adenomas, this time with high

grade dysplasia (see fig 1). Subsequently, the boy underwent proctocolectomy with ileal

pouch-anal anastomosis. The surgical specimen showed a large number of adenomas of which

many contained areas with high grade dysplasia. Genetic analysis revealed that the patient had

inherited both his mother’s APC gene mutation as well as his father’s MLH1 gene defect.

Discussion

FAP is characterised by the development of multiple colorectal adenomas at a young age,

usually before 20 years. Left untreated, one or more of the adenomas will have progressed

to cancer before the age of 45 years. Therefore, most patients have a prophylactic colectomy

in the second or third decade of life. The APC protein, dysfunctional in FAP, regulates the

35

Chapter 3 |

Figure 1. Normal colonic mucosa and areas of adenomatous epithelium with low grade dysplasia (A-C) and adenomatous epithelium with high grade dysplasia (D-F). Serial haematoxylin-eosin staining (A, D) and immunohistochemical (brown) staining of β-catenin (B, E) and MLH1 protein (C, F). In the adenomatous areas with low grade dysplasia, β- catenin staining is normal (membranous) in normal mucosa, in contrast with abnormal (cytoplasmic and nuclear) staining in adenomatous areas; MLH1 protein is expressed in both normal mucosa and adenomatous epithelium. In the adenomatous epithelium with high grade dysplasia, β-catenin staining is similar to the adenomatous areas with low grade dysplasia, whereas MLH1 protein expression is seen in normal mucosa but lost in areas with high grade dysplasia (magnification 100x).

intracellular level of β-catenin, and controls diverse physiological processes in the colon

epithelium including cell growth, adhesion, and apoptosis3. HNPCC is caused by germline

mutations in one of the DNA MMR genes, including MLH1, MSH2, and MSH6.4 MMR

deficient cells rapidly accumulate somatic mutations thus explaining the accelerated

tumorprogression in HNPCC patients.4 Clinical features of HNPCC include colorectal cancer

at an early age (mean 45 years) with a predilection for the proximal colon, an increased

frequency of poorly differentiated mucinous tumours, and a high incidence of extracolonic

tumours (for example, endometrium, ovary, urinary tract).

36

Case report |

To the best of our knowledge, simultaneous inheritance of both types of gene defects in one

patient has not been reported previously. The 5 bp deletion at codon 1309 of the APC gene is

the most commonly reported change in FAP patients and is typically associated with a severe

phenotype with an early onset of disease and the presence of extensive polyposis.5 The MLH1

mutation we found has previously been described and is not associated with a particular

phenotype.6 In our patient, the germline APC mutation, followed by somatic loss of the wild-

type allele, is likely to have triggered early and multiple adenoma formation. This is

illustrated by abnormal expression of β-catenin that was seen in the early adenomas (fig 1).

Loss of MMR function, due to the germline MLH1 mutation, and loss of the wild-type allele,

may well underlie the rapid progression of dysplasia in our patient. The relevance of MMR

dysfunction is supported by the finding that progression from low to high grade dysplasia

observed in the adenomas was associated with loss of MLH1 expression (fig 1).

In addition, evidence has emerged from several animal studies of the important role of MMR

genes in accelerated intestinal tumorigenesis, especially in an APC deficient background.7,8 In

mice carrying both APC and MLH1 mutations, the incidence of intestinal adenomas was

markedly increased compared with mice carrying only one of these mutations.7,8

Alternatively, the severe phenotype in our patient may in fact reflect the marked heterogeneity

which is known to exist in both FAP and HNPCC.9,10 Our patient probably has a higher risk of

developing cancer at other sites than average FAP or HNPCC mutation carriers. Intensive

follow up and surveillance are therefore warranted. In conclusion, our case represents the first

report of a patient with adenomatous polyposis who carried both a pathogenic APC and MLH1

germline mutation. The early adenoma onset and the rapid tumorprogression in this patient

illustrate that the multistep process of colorectal carcinogenesis in this setting was markedly

accelerated.

37

Chapter 3 |

References

1. Potter JD. Colorectal cancer: molecules and populations. J Natl Cancer Inst 1999;91:916–32.

2. Jass JR. Familial colorectal cancer: pathology and molecular characteristics. Lancet Oncol 2000;1:220–6.

3. Fodde R, Smits R, Clevers H. APC, signal transduction and genetic instability in colorectal cancer. Nat Rev Cancer 2001;1:55–67.

4. Lynch HT, Lynch J. Lynch syndrome: genetics, natural history, genetic counseling, and prevention. J Clin Oncol 2000;18:19–31S.

5. Gayther SA, Wells D, SenGupta SB et al. Regionally clustered APC mutations are associated with a severe phenotype and occur at a high frequency in new mutation cases of adenomatous polyposis coli. Hum Mol Genet 1994;3:53–6.

6. Wijnen J, Meera Khan P, Vasen H et al. Majority of hMLH1 mutations responsible for hereditary nonpolyposis colorectal cancer cluster at the exonic region 15–16. Am J Hum Genet 1996;58:300–7.

7. Edelmann W, Yang K, Kuraguchi M et al. Tumorigenesis in Mlh1 and Mlh1/Apc1638N mutant mice. Cancer Res 1999;59:1301–7.

8. Shoemaker AR, Haigis KM, Baker SM et al. Mlh1 deficiency enhances several phenotypes of Apc(Min)/+ mice. Oncogene 2000;19:2774–9.

9. Benatti P, Roncucci L, Ganazzi D et al. Clinical and biologic heterogeneity of hereditary nonpolyposis colorectal cancer. Int J Cancer 2001;95:323–8.

10. Scott RJ, Meldrum C, Crooks R et al. Familial adenomatous polyposis: more evidence for disease diversity and genetic heterogeneity. Gut 2001;48:508–14.

38

4

Hyperplastic polyps in hereditary

nonpolyposis colorectal cancer

F E M Rijcken1, T van der Sluis2, H Hollema2, J H Kleibeuker1

Department of Gastroenterology1 and Pathology2, University

Medical Center Groningen, The Netherlands.

American Journal of Gastroenterology 2003;98:2306-11.

Chapter 4 |

Abstract

Objective - Hereditary nonpolyposis colorectal cancer (HNPCC) is a genetic syndrome

caused by germline mutations in DNA mismatch repair (MMR) genes, in particular

hMLH1, hMSH2 and hMSH6. Dysfunction of MMR genes leads to loss of MMR protein

expression and to microsatellite instability (MSI). MSI is also detected in 10-20% of

sporadic colorectal cancers. Hyperplastic polyps (HP) may serve as precursor for these

MSI+ sporadic colorectal cancers. The aim of this study was to examine whether

hyperplastic polyps are also possible premalignant lesions in HNPCC.

Methods - All hyperplastic polyps resected from (suspected) mismatch repair gene

mutation carriers were retrieved from a screening program database. Clinical

information on the age at colonoscopy and the location of the HP was collected. MLH1,

MSH2, and MLH6 protein expression was evaluated using immunohistochemistry.

Results - 90 HPs were resected from 21 male and 19 female subjects. The mean age at

resection was 45.7 years (44.7 years in male and 46.6 years in female patients). Nineteen

(21%) HPs were resected from the proximal colon, 23 (26%) from the distal colon, and

48 (53%) from the rectum. None of the hyperplastic polyps demonstrated loss of MMR

protein expression.

Conclusion - Mismatch repair dysfunction in hyperplastic polyps of HNPCC patients is

apparently very rare. It seems unlikely that hyperplastic polyps in HNPCC patients are

precursors for (MSI+) cancers in these patients.

40

Hyperplastic polyps |

There is abundant evidence supporting the classical adenoma-carcinoma sequence in the

evolution of colorectal cancer.1,2 Hyperplastic polyps (HPs) have not been included in the

classical model. Traditionally, hyperplastic polyps have been regarded as benign lesions,

lacking the potential for neoplastic progression. Recently, this view is has become

increasingly difficult to sustain given the finding of K-ras mutations3, TGF-βII receptor

mutations4, chromosome 1p deletions5 and DNA microsatellite instability (MSI)6 in

hyperplastic polyps.

With the discovery of the mismatch repair (MMR) genes it has become clear that two major

molecular genetic pathways could be discerned in colorectal carcinogenesis. The first

pathway, the chromosomal instability pathway, is characterized by allelic losses and gains and

aneuploidy. The second pathway, the microsatellite instability pathway, is characterized by an

abundance of subtle DNA mutations and diploidy. Microsatellite instability is caused by

inactivation of a DNA mismatch repair gene. It is the molecular hallmark of hereditary

nonpolyposis colorectal cancer (HNPCC), a genetic syndrome caused by germline mutations

in DNA mismatch repair genes, in particular hMLH1, hMSH2 and hMSH6. Microsatellite

instability seems to be an early event in the carcinogenesis of HNPCC colorectal cancers as up

to 57% of colorectal adenomas from patients with HNPCC show MSI.7,8

Microsatellite instability is also detected in 10-20% of sporadic colorectal cancers.9,10

However, on the basis of morphological observations and epigenetic changes these lesions are

thought to progress to cancer in a way that differs from the microsatellite instability pathway

described for hereditary MSI+ tumors. Jass et al. proposed the serrated pathway in which

hyperplastic polyps serve as precursor for sporadic MSI+ colorectal cancers.6 Under the

influence of promoter methylation of the DNA mismatch repair gene hMLH1 hyperplastic

polyps may develop through serrated dysplasia into sporadic MSI+ colorectal cancer.

In HNPCC all polyps, even minute ones, are resected during screening colonoscopies. Part of

these is hyperplastic and so far, these were considered to be innocent bystanders without any

consequences for further policy. However, in view of the recently described microsatellite

instability and loss of expression of mismatch repair proteins in hyperplastic aberrant crypt

foci (ACF) from MMR gene mutation carriers the innocence of hyperplastic lesions in

HNPCC should be questioned.11 On the other hand, Iino et al. found only two of 17

41

Chapter 4 |

hyperplastic polyps in subjects with HNPCC to be microsatellite unstable.12 Jass et al.

described one HNPCC patient with a mixed hyperplastic polyp/adenoma in contiguity with a

colorectal cancer.13

The aim of this study was to examine whether hyperplastic polyps are possible premalignant

lesions in HNPCC. The loss of expression of mismatch repair proteins is strongly associated

with MMR gene dysfunction and, thus with MSI.14,15 Ninety hyperplastic polyps resected

from (suspected-) MMR gene mutation carriers were evaluated for the presence of the

mismatch repair proteins MLH1, MSH2 and MSH6 using immunohistochemistry.

Materials and Methods

In our hospital, all patients belonging to an HNPCC family are advised to be screened every

two years by means of a colonoscopy. The clinical and the histopathological results are

registered in a database. This database was used to retrieve all polyps, information concerning

the age at the first and every consecutive colonoscopy, age at removal of a polyp, anatomical

location of a polyp and final pathology report. The size of the hyperplastic polyps was

determined by measuring the circumference of the polyp at 10x magnification. An

experienced pathologist (HH) verified the histology.

In all hyperplastic polyps MLH1, MSH2, and MLH6 protein expression was evaluated using

immunohistochemistry as previously described.16,17 Briefly, serial 3µm-thick-sections were

cut from paraffin blocks and fixed onto coated slides. Antigen retrieval was performed by

high pressure cooker treatment of the slides in 200 µl blocking reagent (Boehringer

Mannheim, Germany). After blocking endogenous peroxidase activity, the primary antibodies

diluted in phosphate buffered saline (PBS) with 1% bovine serum albumin (BSA) were

applied for one hour. Slides were stained with MLH1 antibody (dilution: 1:500, clone G168-

728, Pharmingen, San Diego, CA, USA), MSH2 antibody (1:100, clone Ab-2, Calbiochem,

San Diego, CA, USA), and MSH6 antibody (1:200, Transduction, Lexington, KY, USA).

After washing with PBS, the sections were incubated with rabbit antimouse peroxidase

followed by goat antirabbit peroxidase, both diluted (1:50) in PBS-1% BSA. The peroxidase

activity was visualized with diaminobenzidine. The sections were counterstained with

haematoxylin.

42

Hyperplastic polyps |

To verify the correlation between proliferation and location of the mismatch repair protein

immunostaining ten large hyperplastic polyps were assessed by immunohistochemical

staining with a mouse monoclonal antibody to Ki-67 (MIB-1; 1:350, DAKO, Glostrup,

Denmark).

Statistical analyses were performed using SPSS (Statistical Package for the Social Sciences,

Munich, Germany) for Windows. The statistical difference between the location of

hyperplastic polyps and adenomas was calculated using the nonparametric Mann-Whitney U

test. P<0.05 was considered significant. The correlations between the presence of hyperplastic

polyps and adenomas dependent and independent of the location were analyzed using

Pearson's correlation coefficient.

Results

A total of 158 persons, 73 male and 85 female, were included in our screening program. Fifty

(32%) were known mismatch repair gene mutation carriers. Twenty-two had a mutation in

hMLH1, 21 in hMSH2 and 7 in hMSH6. The remaining 108 persons all belonged to families

fulfilling the Amsterdam criteria.18 Five cancer patients (average age at diagnosis: 39 years)

had uninformative genetic analysis. Eight genetic analyses were still in progress. Three of

these eight persons had an HNPCC-related cancer (average age at diagnosis: 36 years), two

had adenomas (average age at first resection: 31 years), while in the remaining three no

neoplastic lesions had been identified (average age now: 41 years). In the remaining 95

persons genetic analysis had not been performed. Sixty-nine persons were from well-

documented HNPCC families in which no germline mutation had been found up to this

moment. Twenty-four of these persons had adenomas removed (average age at first resection:

44 years), in the remaining 45 no neoplastic lesions had been detected (average age now: 41

years). Genetic analysis had not been performed in another nine persons because their affected

family members had yet to be tested. Another 17 persons were from HNPCC-families in

which the mutation was known but the subjects chose not to be genetically tested. In this

group two patients had one or more HNPCC-related tumors removed (average age at

diagnosis of first cancer: 33 years), two women were diagnosed with atypical complex

hyperplasia of the endometrium (average age at diagnosis: 43 years), five patients had

43

Chapter 4 |

adenomas resected (average age: 44 years), and the remaining eight had no neoplastic lesions

detected (average age now: 39 years).

The mean age at the first colonoscopy was 40 years. A total of 540 colonoscopies was done in

994 patient years. An average of 3 colonoscopies was done per patient. 257 colonoscopies in

male patients rendered 41 hyperplastic polyps and 82 adenomas, while in female patients, 283

colonoscopies rendered 49 hyperplastic polyps and 75 adenomas.

Ninety hyperplastic polyps were removed from 40 patients, 21 men and 19 women. The mean

age at which these hyperplastic polyps were removed was 46 years (45 years in male patients

and 47 years in female patients). Nineteen (21%) hyperplastic polyps were resected from the

proximal colon, 23 (26%) from the distal colon, and 48 (53%) from the rectum. The

hyperplastic polyps size ranged from 1.0-14.0 mm with an average of 6.0 mm. The size of the

hyperplastic polyps was not associated with the location in the colon (proximal: 6.2 mm,

distal: 5.8 mm, rectal: 6.0 mm).

One hundred and fifty-seven adenomas were removed from 64 patients, 32 men and 32

women. The mean age at resection was 49.1 years. 82 (52%) adenomas were located in the

proximal colon, 49 (31%) in the distal colon, and 26 (17%) in the rectum. Four adenomas with

adenomatous as well as hyperplastic features from two MLH-1 mutation carriers were

classified as serrated adenomas (figure 1). Two were located in the proximal colon, one in the

distal colon and one in the rectum.

Figure 1. Serrated adenoma from a MLH1 mutation carrier.

44

Hyperplastic polyps |

In 18 colonoscopies synchronously occurring hyperplastic polyps (n=25) and adenomas

(n=28) were resected. Three (12%) of these HPs were located in the ascending colon, 7

(28%) in the descending colon and sigmoid and 15 (60%) in the rectum while 19 (68%)

adenomas were located in de ascending colon, 4 (14%) in the descending colon and sigmoid

and 5 (18%) in the rectum (figure 2). In 387 colonoscopies neither hyperplastic polyps nor

adenomas were observed. In 45 colonoscopies only hyperplastic polyps were resected and in

88 colonoscopies only adenomas. In 2 colonoscopies a cancer was observed. There was no

correlation between the presence of hyperplastic polyps and adenomas or between the location

of the two types of polyps.

Figure 2. Distribution of hyperplastic polyps and adenomas in the colon.

Hyperplastic polyp (●), synchronously occuring hyperplastic polyp (HP) with another HP (●), adenomas (X), synchronously occuring adenomas with another adenoma (x), number of colonoscopies (*). A, B, and C represent colonoscopies in which synchronously occuring hyperplastic polyps and adenomas were resected. If more than one hyperplastic polyp/adenoma were removed than the most distal hyperplastic polyp/adenoma was used as index polyp and marked with a large symbol while the synchronously occuring hyperplastic polyp(s)/adenoma(s) were marked with a smaller symbol. A, colonoscopies with proximal HPs with synchronously occuring adenomas. B, distal hyperplastic polyps with synchronous adenomas. C, rectal HPs with synchronous adenomas. D, colonoscopies in which only hyperplastic polyps were resected. E, colonoscopies in which only adenomas were resected. F, colonoscopies without polyps.

45

Chapter 4 |

A

C Figure 3. Hyperplastic polyp from a MSH2 mutation car(A) MSH2, (B) detail of MSH2, and (C) MIB, (D) detail of

46

B

D

rier demonstrating similar immunostaining pattern for MIB.

Hyperplastic polyps |

Twenty-five (28%) of the 90 hyperplastic polyps were resected from 14 patients with a known

mutation in one of the mismatch repair genes. Nine HPs were from 4 hMLH1 mutation

carriers, 12 HPs were from 6 hMSH2 mutation carriers and 4 HPs were from 4 hMSH6

mutation carriers. The average age at resection was 51.2 years. Three (12%) hyperplastic

polyps of known mutations carriers were found in the proximal colon, seven (28%) in the

distal colon and 15 (60%) in the rectum.

None of the hyperplastic polyps or the serrated adenomas showed loss of expression of one of

the mismatch repair proteins. The immunohistochemical staining in the crypts of the

hyperplastic polyps was similar to that in crypts of normal epithelium of the colorectum. Most

nuclei in the base of the crypts expressed all three MMR proteins intensely while few to none

of the luminal nuclei expressed any of the MMR proteins (figure 3 A and B). Expression of

mismatch repair proteins overlapped with proliferation areas as demonstrated by Ki-67

immunoreactivity (figure 3 C and D).

Discussion

Hyperplastic polyps are possibly precursors of sporadically occurring microsatellite unstable

tumors.6 Even though hyperplastic polyps are common and hyperplasia may occur in

contiguity with adenomatous lesions in HNPCC, the present study illustrates that hyperplastic

polyps most likely do not play a significant role in the carcinogenesis of microsatellite

unstable tumors in subjects with a germline MMR gene mutation.

During an average follow-up of 6 years, 25% of HNPCC patients had hyperplastic polyps

removed. Eight percent of the patients had hyperplastic polyps removed at their first

colonoscopy. In a study by Imperiale et al, 10% of asymptomatic average risk persons 40 to

49 years of age had hyperplastic polyps.19 Hyperplastic polyp prevalence increases with age

and has been reported to be up to 35%.20 In the present study, 185 of the 540 colonoscopies

were performed in patients older than 50 years. The majority (79%) of hyperplastic polyps of

HNPCC patients were located in the distal colon and rectum as is seen in sporadic cases.19, 21

In contrast to sporadic colorectal cancer, hereditary nonpolyposis colorectal cancer has a

predisposition for the proximal colon. We have shown in the past that adenomas in HNPCC

patients also have a predisposition for the proximal colon and that these proximal adenomas

47

Chapter 4 |

are highly dysplastic at a smaller size than sporadic adenomas.16 In the present study, the

presence or location of hyperplastic polyps did not correlate with presence or location of

adenomas. These characteristics are not in line with the theory that hyperplastic polyps play a

role in HNPCC carcinogenesis. On the other hand the detection of four serrated adenomas

could support the premalignant theory of hyperplasia. However, neither the hyperplastic nor

the adenomatous part of the serrated adenomas demonstrated loss of one of the mismatch

repair proteins.

None of the hyperplastic polyps included in this study showed loss of expression of MLH1,

MSH2 or MSH6. Only two other studies have been published concerning hyperplasia in

colorectal lesions of HNPCC patients. In one study, microsatellite instability was reported to

be present in hyperplastic aberrant crypt foci occurring synchronously with MSI+ colorectal

cancer.11 In the other study, 2 (11%) of 17 hyperplastic polyps were found to have a low

degree of microsatellite instability (MSI-low).12 One case report describes a mixed

hyperplastic polyp/adenoma in contiguity with a colorectal cancer in an HNPCC patient.13

Compared to the above studies, our results seem to underestimate the incidence of mismatch

repair dysfunction in hyperplastic polyps in HNPCC patients. However, the studies are not

readily comparable due to the different research populations (HNPCC patients with MSI-high

colorectal tumors vs. an HNPCC screening population) and the different techniques used

(microsatellite analysis vs. immunohistochemistry). Pedroni et al and Jass et al both detected

a low degree of microsatellite instability and not a high degree (MSI-high) in the hyperplastic

polyps. Immunohistochemistry most often can not discern MSI-low hyperplastic polyps as

MSI-low lesions generally do express mismatch repair proteins.22 Whether this should be

considered a limitation of our study is debatable. The significance of MSI-low is

controversial. MSI-low lesions have not been consistently shown to be different from

microsatellite stable lesions.23 HNPCC lesions are most often associated with the MSI-high

phenotype which is undoubtedly different from the microsatellite stable phenotype.4

Genetic inactivation of a mismatch repair gene is generally associated with loss of

immunohistochemical expression of the corresponding protein. 24,25 Lindor et al demonstrated

that immunohistochemistry had a sensitivity of 92.3% and a specificity of 100% for screening

for DNA mismatch repair defects.14 Other recent studies have concluded that

immunohistochemical analysis of MLH1 and MSH2 expression is a rapid and accurate

48

Hyperplastic polyps |

method for identifying colorectal tumors of the MSI-high phenotype.25-28 We also confirmed

the high quality of mismatch repair protein immunohistochemistry in (pre)malignant lesions

in comparison with microsatellite instability analysis. 29

MMR deficiency results in microsatellite instability and it is thought to cause cancer by

failing to repair replication errors within repeat sequences contained in genes relevant for

growth control and differentiation like TGF-βRII, BAX, and IGFRII. Due to the MMR

deficiency, mutation rates in tumor cells are 100-1000-fold as compared to normal cells.30,31

The high mutation rate leads to an accelerated adenoma-carcinoma sequence in HNPCC.2

Pedroni et al demonstrated progressive accumulation of bandshifts in microsatellite analysis in

the sequence normal mucosa-dysplastic aberrant crypt foci-adenoma-carcinoma indicative of

the role of MMR gene dysfunction in the evolution of colonic lesions.11,32,33 Iino proposed that

MSI in hyperplastic sporadic polyps heralds the transformation to serrated adenoma.6

Similarly, we demonstrated that loss of expression of MMR protein heralds development to

high-grade dysplasia in HNPCC adenomas.16

In conclusion, this is the largest series of hyperplastic polyps from subjects with HNPCC or at

50% risk for it, which were evaluated for mismatch repair dysfunction. None of these

hyperplastic polyps nor any of four serrated adenomas displayed loss of expression of the

three MMR proteins. If hyperplastic lesions would play a significant role in the microsatellite

instability carcinogenesis pathway of HNPCC lesions loss of expression of MMR proteins

should have been detected at least in some of the polyps. Our findings do not dictate any

changes in clinical practice. Mismatch repair dysfunction in hyperplastic polyps of HNPCC

patients is apparently very rare and if found it should be considered a coincidence and not a

common step in the carcinogenic pathway of HNPCC.

49

Chapter 4 |

References

1. Vogelstein B, Fearon ER, Hamilton SR et al. Genetic alterations during colorectal-tumor development. N Engl J Med. 1988;319:525-32.

2. Kinzler KW, Vogelstein B. Lessons from hereditary colorectal cancer. Cell 1996;87:159-70.

3. Otori K, Oda Y, Sugiyama K et al. High frequency of K-ras mutations in human colorectal hyperplastic polyps. Gut 1997;40:660-3.

4. Konishi M, Kikuchi YR, Tanaka K et al. Molecular nature of colon tumors in hereditary nonpolyposis colon cancer, familial polyposis, and sporadic colon cancer. Gastroenterology 1996;111:307-17.

5. Bomme L, Bardi G, Pandis N et al. Clonal karyotypic abnormalities in colorectal adenomas: clues to the early genetic events in the adenoma-carcinoma sequence. Genes Chromosomes Cancer 1994;10:190-6.

6. Iino H, Jass JR, Simms LA et al. DNA microsatellite instability in hyperplastic polyps, serrated adenomas, and mixed polyps: a mild mutator pathway for colorectal cancer? J Clin Pathol 1999;52:5-9.

7. Aaltonen LA, Peltomaki P, Mecklin JP et al. Replication errors in benign and malignant tumors from hereditary nonpolyposis colorectal cancer patients. Cancer Res 1994;54:1645-8.

8. Samowitz WS, Slattery ML. Microsatellite instability in colorectal adenomas. Gastroenterology 1997;112:1515-9.

9. Aaltonen LA, Peltomaki P, Leach FS et al. Clues to the pathogenesis of familial colorectal cancer. Science 1993;260:812-6.

10. Lothe RA, Peltomaki P, Meling GI et al. Genomic instability in colorectal cancer: relationship to clinicopathological variables and family history. Cancer Res 1993;53:5849-52.

11. Pedroni M, Sala E, Scarselli A et al. Microsatellite instability and mismatch-repair protein expression in hereditary and sporadic colorectal carcinogenesis. Cancer Res 2001;61:896-9.

12. Iino H, Simms L, Young J et al. DNA microsatellite instability and mismatch repair protein loss in adenomas presenting in hereditary non-polyposis colorectal cancer. Gut 2000;47:37-42.

13. Jass JR, Cottier DS, Pokos V et al. Mixed epithelial polyps in association with hereditary non-polyposis colorectal cancer providing an alternative pathway of cancer histogenesis. Pathology 1997;29:28-33.

50

Hyperplastic polyps |

14. Lindor NM, Burgart LJ, Leontovich O et al. Immunohistochemistry versus microsatellite instability testing in phenotyping colorectal tumors. J Clin Oncol 2002;20:1043-8.

15. Chiaravalli AM, Furlan D, Facco C et al. Immunohistochemical pattern of hMSH2/hMLH1 in familial and sporadic colorectal, gastric, endometrial and ovarian carcinomas with instability in microsatellite sequences. Virchows Arch 2001;438:39-48.

16. Rijcken FE, Hollema H, Kleibeuker JH. Proximal adenomas in hereditary non-polyposis colorectal cancer are prone to rapid malignant transformation. Gut 2002;50:382-6.

17. Berends MJ, Wu Y, Sijmons RH et al. Molecular and clinical characteristics of MSH6 variants: an analysis of 25 index carriers of a germline variant. Am J Hum Genet 2002;70:26-37.

18. Vasen HF, Mecklin JP, Khan PM et al. The International Collaborative Group on Hereditary Non-Polyposis Colorectal Cancer (ICG-HNPCC). Dis Colon Rectum 1991;34:424-5.

19. Imperiale TF, Wagner DR, Lin CY et al. Results of screening colonoscopy among persons 40 to 49 years of age. N Engl J Med 2002;346:1781-5.

20. DiSario JA, Foutch PG, Mai HD et al. Prevalence and malignant potential of colorectal polyps in asymptomatic, average-risk men. Am J Gastroenterol 1991;86:941-5.

21. Williams AR, Balasooriya BA, Day DW. Polyps and cancer of the large bowel: a necropsy study in Liverpool. Gut 1982;23:835-42.

22. Lanza G, Gafa R, Maestri I et al. Immunohistochemical pattern of MLH1/MSH2 expression is related to clinical and pathological features in colorectal adenocarcinomas with microsatellite instability. Mod Pathol 2002;15:741-9.

23. Tomlinson I, Halford S, Aaltonen L et al. Does MSI-low exist? J Pathol 2002;197:6-13.

24. Kuismanen SA, Holmberg MT, Salovaara R et al. Genetic and epigenetic modification of MLH1 accounts for a major share of microsatellite-unstable colorectal cancers. Am J Pathol 2000;156:1773-9.

25. Thibodeau SN, French AJ, Roche PC et al. Altered expression of hMSH2 and hMLH1 in tumors with microsatellite instability and genetic alterations in mismatch repair genes. Cancer Res 1996;56:4836-40.

26. Dietmaier W, Wallinger S, Bocker T et al. Diagnostic microsatellite instability: definition and correlation with mismatch repair protein expression. Cancer Res 1997;57:4749-56.

27. Marcus VA, Madlensky L, Gryfe R et al. Immunohistochemistry for hMLH1 and hMSH2: a practical test for DNA mismatch repair-deficient tumors. Am J Surg Pathol 1999;23:1248-55.

51

Chapter 4 |

28. Chaves P, Cruz C, Lage P et al. Immunohistochemical detection of mismatch repair gene proteins as a useful tool for the identification of colorectal carcinoma with the mutator phenotype. J Pathol 2000;191:355-60.

29. Berends MJ, Hollema H, Wu Y et al. MLH1 and MSH2 protein expression as a pre-screening marker in hereditary and non-hereditary endometrial hyperplasia and cancer. Int J Cancer 2001;92:398-403.

30. Bhattacharyya NP, Skandalis A, Ganesh A et al. Mutator phenotypes in human colorectal carcinoma cell lines. Proc Natl Acad Sci 1994;91:6319-23.

31. Parsons R, Li GM, Longley MJ et al. Hypermutability and mismatch repair deficiency in RER+ tumor cells. Cell 1993;75:1227-36.

32. Heinen CD, Shivapurkar N, Tang Z et al. Microsatellite instability in aberrant crypt foci from human colons. Cancer Res 1996;56:5339-41.

33. Percesepe A, Pedroni M, Sala E et al. Genomic instability and target gene mutations in colon cancers with different degrees of allelic shifts. Genes Chromosomes Cancer 2000;27:424-9.

52

5

Early carcinogenic events in

HNPCC adenomas: differences

with sporadic adenomas

F E M Rijcken1, J J Koornstra1 , T van der Sluis2, W Boersma-

van Ek1, J H Kleibeuker1, H Hollema2

Department of Gastroenterology1 and Pathology2, University

Medical Center Groningen, The Netherlands.

Submitted

Chapter 5 |

Abstract

Background - Tumorigenesis in hereditary nonpolyposis colorectal cancer (HNPCC)

differs from that in sporadic colorectal cancer at an early stage. We examined the

expression of proliferation- and apoptosis-regulating proteins in relation to proliferation

and apoptosis in HNPCC and sporadic adenomas.

Methods - Proliferation and apoptosis were quantified and expression of cyclin B1, D3

and E, p21, p27, bcl-2, bax, p53 and cox-2 were determined by immunohistochemistry in

42 HNPCC and 48 sporadic adenomas.

Results - No differences in proliferation and apoptosis were detected. Low-grade

dysplastic HNPCC adenomas differed from sporadic ones by expressing bcl-2 more

often (69% vs. 42 %) and bax less often (50% vs. 73%). Fewer high-grade dysplastic

HNPCC than sporadic adenomas expressed cyclin B1 and E (50%, 38% vs. 87%, 87%),

p21 (6% vs. 53%) and bax (31% vs. 80%). HNPCC adenomas had less overexpression of

p53 (5% vs. 19%).

Conclusion - The expression of cell cycle and apoptosis related proteins differs between

early through advanced HNPCC and sporadic adenomas though proliferation and

apoptosis are not different. These differences may contribute to the different clinical

behavior of HNPCC and sporadic adenomas.

54

Carcinogenic events in adenomas |

Carcinogenesis is a multi-step process. The steps represent mutations or epigenetic changes

that activate oncogenes or inactivate tumor suppressor genes. As mutations accumulate, cells

loose their balanced homeostasis of cell division (proliferation) and cell death (apoptosis) and

transform into uncontrollably growing cell clones. In the case of colorectal cancer, these

genetic alterations are reflected by the adenoma-carcinoma sequence.1

Similar to the situation in sporadic colorectal cancer and familial adenomatous polyposis,

hereditary nonpolyposis colorectal cancer (HNPCC) arises from adenomas.2 HNPCC is

caused by germline mutations in the mismatch repair (MMR) genes and the lesions are

characterized by microsatellite instability (MSI).3 Dysfunction of one of the MMR genes

results in a rapid accumulation of genetic alterations in susceptible genes, resulting in an

accelerated adenoma-carcinoma sequence in HNPCC compared to sporadic cancer.2

Considerable differences have been shown between MSI-positive and MSI-negative colorectal

tumors, supporting the idea that MSI-positive and MSI-negative colorectal cancers develop

through different pathways.4-6

Proliferation and apoptosis reportedly increase during tumor progression in sporadic and in

HNPCC neoplasms.7,8 However, cell proliferation indices were lower and apoptosis occurred

more frequently in MSI-positive than in MSI-negative colorectal cancers.5 These differences

may be caused by disturbances in gene products stimulating proliferation, for example, cyclin

B1, D3 and E, which push a cell through the G1 and G2 phase of the cell cycle.9 Alternatively,

there may be disturbances in gene products inhibiting proliferation, functioning as

checkpoints in the cell cycle, e.g. cyclin dependent kinase (CDK) inhibitors p21 and p27.9

The p53 tumor suppressor gene is the most commonly mutated gene in human cancer.10 It

normally inhibits the cell cycle through transcriptional activation of p21 and triggers

apoptosis. Bcl-2, a proto-oncogene, can block p53-mediated apoptosis thereby allowing

accumulation of genetic alterations and potentially contributing to neoplastic development.10

Another member of the bcl-2 family, bax, can antagonize the apoptotic blocking ability of bcl-

2 .10 Cox-2 overexpression may enhance expression of bcl-2 and may lead to prostaglandin-

mediated resistance to apoptosis.11

The aim of this study was to explore differences in the carcinogenic pathways between

sporadic and HNPCC neoplasms by examining expression of proliferation- and apoptosis-

55

Chapter 5 |

regulating proteins in adenomas. Furthermore, the expression of these proteins was related to

proliferation and apoptosis indices by means of immunohistochemistry.

Materials and methods

Adenomas were retrieved from the archives of the Department of Pathology at the University

Hospital of Groningen. Those large enough to render 20 3 µm-slides were selected. HNPCC-

associated adenomas had been obtained during colonoscopies of patients who had a known

MMR gene mutation or belonged to a family that fulfilled the Amsterdam criteria and had an

adenoma or cancer diagnosed before 50 years of age.12 Sporadic adenomas were consecutively

removed in 1997 from individuals without a strong family history for colorectal cancer and

without inflammatory bowel disease. Haematoxylin-eosin stained slides were reviewed and

the adenomas were assessed for grade of dysplasia and histopathological subtype according to

the WHO criteria.13

Immunohistochemistry

Proliferation and apoptosis

Proliferation and apoptosis were assessed by immunohistochemical staining of Ki-67 (MIB-1,

Immunotech, Marseilles, France) and of cleavage products of cytokeratin 18 (cCK18, Mab

M30, Boehringer Mannheim, Mannheim, Germany), respectively. Cytokeratin 18 (CK18)

cleavage by activated caspase-3 is an early marker of apoptosis. CK18 is epithelial cell

specific. The optimal antibody-antigen reaction, summarized in table 1, was determined for

proliferation-regulating proteins cyclin B1, D3 and E, CDK inhibitors p21 and p27, and

apoptosis regulators bcl-2, bax, cox-2 and p53. Twenty consecutively numbered sections of 3

µm were cut from formalin-fixed paraffin-embedded adenomas. All sections were fixed onto

3-aminopropyl-triethoxysilane (APES, Sigma-Aldrich, Diesenhofen, Germany) coated slides,

stretched for 30 minutes at 60°C and dried overnight at 37°C. Before staining, the sections

were deparaffinised in xylene (2x 10 minutes) and rinsed in graded alcohol. The entire batch

of adenomas was stained in one session per antibody. If staining was not adequate, sections

were stained again; a well-stained section from the first batch was included as reference point.

Before immunohistochemical staining with MIB-1 antigen retrieval was performed using the

high-pressure cooker. Immersed in 200 µl blocking reagent (2% block and 0.2% SDS in

maleic acid, pH 6.0 (Boehringer Mannheim, Germany)) the section underwent 3 sessions of 5

56

Carcinogenic events in adenomas |

minutes at 115°C in a high pressure cooker alternating with incubation in a humid

environment. The endogenous peroxidase activity was quenched by incubation with 0.3%

H2O2 in phosphate buffered saline (PBS) for 30 minutes. The sections were immersed for one

hour with the MIB-1 antibody in PBS with 1% bovine serum albumin (BSA), at a dilution of

1:400. Subsequently, the sections were consecutively incubated for 30-minute periods with

rabbit antimouse peroxidase (RAMPO; DAKO, Glostrop, Denmark) and goat antirabbit

peroxidase (GARPO; DAKO, Glostrop, Denmark) both diluted (1:50) in PBS-1% BSA. The

sections were submerged for 10 minutes in a solution of 25 mg 3,3’-diaminobenzidine (DAB)

in PBS and 50 mg of imidazol with 50 µl 30% H2O2. After rinsing with demi water, the

sections were counterstained with haematoxylin, washed with running water, dehydrated with

graded alcohol, dried and covered with a slide.

The microwave antigen retrieval method was used for the staining of cleaved cytokeratin 18.

The deparaffinised, rehydrated sections were submersed in preheated 10 mM citrate buffer

(pH 6.0) and heated for 8 minutes at 700 watts in a microwave. After cooling at room

temperature for 15 minutes the sections were thoroughly rinsed with PBS for 5 minutes and a

1:50 solution of M30 was applied for one hour at room temperature. Subsequently, the same

steps as for Ki-67 staining were performed, starting with the incubation of the section with

RAMPO.

Proliferating-regulating proteins and apoptosis-regulating proteins

The methods and dilutions used for immunostaining of all proteins are summarized in table 1.

The high-pressure cooking antigen retrieval method was as described for MIB-1. The

microwave antigen retrieval method was the same as described for M30. However, in four

cases, cyclin B1, D3 and E, and cox-2, 1 mM EDTA (pH 8.0) was used instead of 10 mM

citrate buffer (pH 6.0).

Mismatch repair proteins

Immunohistochemical analysis of MLH1, MSH2, and MSH6 was performed in all HNPCC

adenomas as described for staining with MIB-1 and with the dilutions noted in table 1. The

sporadic adenomas were stained with MLH1 antibody only.

57

Chapter 5 |

Table 1. Antibodies and antigen retrieval methods used for immunohistochemistry.

Protein Antigen retrieval

method

Clone Company Dilution

Ki-67 High-pressure cooker MIB-1 Immunotech, Marseille, France 1:400

Cyclin B1 Microwave EDTA 7A9 Novocastra, Newcastle, UK 1:50

Cyclin D3 Microwave EDTA DCS-22 Novocastra, Newcastle, UK 1:10

Cyclin E Microwave EDTA 13A3 Novocastra, Newcastle, UK 1:10

P21 High-pressure cooker WAF1(Ab-1) Oncogene, Darmstadt, Germany 1:50

P27 High-pressure cooker 1B4 Novocastra, Newcastle, UK 1:50

cCK 18* Microwave citrate MAb M30 Boehringer, Mannheim, Germany 1:50

Bax Microwave citrate B-9 Santa Cruz Biotechnology, Santa Cruz, Ca 1:200

Bcl-2 High-pressure cooker MAb 124 DAKO, Glostrup, Denmark 1:50

Cox-2 Microwave EDTA 33 Transduction Laboratories, Lexington, KY 1:50

P53 High-pressure cooker B-p53-12-1 Biogenex, San Ramon, CA, USA 1:400

* cleaved cytokeratin 18

Evaluation of staining

Two authors (FR and TvdS) scored all stained slides without the knowledge of the clinical

data. In addition, a third investigator (JJK) independently scored all MIB-1 and M30 stained

slides to confirm reproducibility of quantified indices (paired differences: mean 3.3 ± 8.5 %,

95% confidence interval 1.5-5.2 %). To confirm reproducibility of the remaining data, 25 %

of the slides were randomly chosen and scored a second time yielding Kappa-values between

0.844 and 0.954 for the different stainings. The immunoreactivity for every antibody, except

for M30, was quantified in at least three entire crypts of the most dysplastic area of the

adenoma. Proliferation was quantified as labeling index that is the percentage of the positively

stained nuclei. M30 positivity was seen as brown cytoplasmic staining. At least 1000 cells

were counted at x 400 magnification, and the apoptotic index was quantified as percentage of

stained cells of the epithelial cells counted. Only those stained cells that had not been shed and

were morphologically intact within the epithelium were counted. Cyclin B1, bax, bcl-2 and

cox-2 positivity was seen as cytoplasmic staining while cyclin D3 and E, p21, p27 and p53

antibody displayed a nuclear immunoreactivity. For cyclin B1, D3, and E, p21, p27 and cox-2

staining, the percentage of positive cells was assessed semi-quantitatively. Each staining

58

Carcinogenic events in adenomas |

showed a wide range of expression. The median value of each staining in all the adenomas

was calculated and functioned as cut-off value.14 Immunoreactivity was considered low if

percentage of positive cells was below the median value and it was considered high if the

percentage of positive cells was equal to or greater than the median value. In bax and bcl-2

staining, the intensity was judged relative to the expression in normal crypts and graded as

follow: 1, no staining; 2, weaker staining; 3 similar or moderately higher staining; and 4, more

intense staining than in normal crypts. Subsequently, staining category 3 and 4 were

considered positive. An adenoma was scored as overexpressing p53 if unequivocal nuclear

staining was observed in at least 5% of cells.15

Statistics

Statistical analyses were performed using SPSS (Statistical Package for the Social Sciences,

Munich, Germany) for Windows. The statistical difference between HNPCC and sporadic

adenomas per staining was calculated using the nonparametric Mann-Whitney U test. P<0.05

was considered significant. The associations between proliferation- and apoptosis indices on

the one hand and proliferation- and apoptosis-regulating proteins on the other and between the

regulating proteins were analyzed using Chi-square tests.

Results

The adenomas

In total 42 HNPCC adenomas were compared to 48 sporadic adenomas. There were 26 low-

grade dysplastic (LGD) and 16 high-grade dysplastic (HGD) HNPCC adenomas and 33 LGD

and 15 HGD sporadic adenomas. Six (23%) LGD HNPCC and 8 (30%) LGD sporadic

adenomas were tubulovillous while 9 (56 %) HGD HNPCC and 9 (60 %) HGD sporadic

adenomas had a tubulovillous histology. In both groups, presence of tubulovillous histology

correlated with high grade of dysplasia (HNPCC: p=0.029; sporadic: p=0.016). The remaining

adenomas were all tubular. Eighteen HNPCC adenomas had been removed from the proximal

colon, 16 from the distal colon and 8 from the rectum. In the sporadic group, 15 adenomas

had been resected from the proximal colon, 26 from the distal colon and 7 from the rectum.

Mismatch repair protein expression in HNPCC adenomas is depicted in table 2. Sixteen

adenomas were from 13 proven mutation carriers. The genetic status of the subjects (n=18)

59

Chapter 5 |

from whom the remaining 26 adenomas had been resected was not known either because

genetic testing had not been done (n=8) or because a mutation had not been found in their

families (n=10). In HNPCC adenomas from known mutation carriers, the loss of MMR

protein corresponded to the germline mutation. The 48 sporadic adenomas all stained positive

for MLH1.

Table 2. Mismatch repair protein expression of the HNPCC adenomas.

Dysplasia

Low grade High grade

Known MMR mutation carriers

Loss of MLH1* 3/6 4/4 Loss of MSH2* 1/2 2/2 Loss of MSH6* 0/1 1/1

Unknown genetic status (n=17) (n=9)

Loss of MLH1 1 2 Loss of MSH2 3 2 Loss of MSH6 1 2

Expressing all 3 MMR proteins 12 3

Adenomas showing loss of expression of one MMR protein

8/26 13/16

* Number of adenomas demonstrating loss of MLH1/MSH2/MSH6 protein in a MLH1/MSH2/MSH6,

respectively, gene mutation carrier.

Immunohistochemistry

The results of the immunohistochemical staining are summarized in table 3. Figure 1 A-H

depicts examples of an adenoma, stained for proliferation (MIB-1), cyclin D3, cyclin E, p21,

apoptosis (M30), bax, bcl-2 and cox-2.

Proliferation

In normal mucosa adjacent to the adenoma, MIB-1 positive cells were located in the base of

the crypt. Both in HNPCC and sporadic adenomas proliferating cells were located near the

luminal side and in the top one third of the crypt. There was a wide range of proliferative

indices, from 10 to 80 %, in both HNPCC and sporadic adenomas.

60

Carcinogenic events in adenomas |

Table 3. Immunohistochemistry results of the HNPCC and sporadic adenomas.

LOW-grade dysplasia

Statistical difference*

HIGH-grade dysplasia

Statistical difference**

Statistical difference***

Staining categories †

1 2 1 2

Proliferation

HNPCC 46 (4) 48 (5) n.s. Mean prolif. Index %(±SEM) Sporadic 47 (3)

n.s. 49 (5)

n.s. n.s.

HNPCC 22 4 8 8 p=0.017 Cyclin B1 (n) Sporadic 28 5 n.s. 2 13 p=0.032 P<0.001

HNPCC 15 11 12 4 n.s. Cycin D3 (n) Sporadic 27 6 n.s. 8 7 n.s. p=0.042

HNPCC 17 9 10 6 n.s. Cyclin E (n) Sporadic 20 13 n.s. 2 13 p=0.018 p=0.010

HNPCC 20 6 15 1 n.s. P21 (n) Sporadic 21 12 n.s. 7 8 p=0.005 n.s.

HNPCC 17 9 6 10 n.s. P27 (n) Sporadic 18 15 n.s. 10 5 n.s. n.s.

Apoptosis

HNPCC 0.87 (0.28) 1.03 (0,42) n.s. Mean apoptotic index %(±SEM) Sporadic 0.45 (0.07)

n.s. 0.35 (0.10)

n.s. n.s.

HNPCC 13 13 11 5 n.s. Bax (n) Sporadic 9 24 p=0.076 3 12 p=0.007 n.s.

HNPCC 8 18 2 14 n.s. Bcl-2 (n) Sporadic 19 14 p=0.042 12 3 p<0.001 n.s.

HNPCC 15 11 11 5 n.s. Cox-2 (n) Sporadic 23 10 n.s. 9 6 n.s. n.s.

HNPCC 25 1 15 1 n.s. P53 (n) Sporadic 30 3 n.s. 9 6 p=0.021 p=0.012

* Low-grade dysplastic HNPCC adenomas v. low-grade dysplastic sporadic adenomas. ** High-grade dysplastic HNPCC adenomas v. high-grade dysplastic sporadic adenomas. *** Low-grade v. high-grade dysplastic HNPCC adenomas or low-grade v. high-grade dysplastic sporadic

adenomas. † Staining categories:

• cyclin B1, D3 and E, p21 and cox-2: 1) <10%; 2) ≥10% • p27: 1) <20%; 2) ≥20% • bax and bcl-2: 1) no staining and weak; 2) moderate and intense staining • p53: 1) negative, <5% nuclear staining; 2) positive, ≥5% nuclear staining

61

Chapter 5 |

62

D

C

B

A EE

F

H

G

Figure 1. Examples of an adenoma, stained for proliferation (MIB-1) (A), cyclin D3 (B), cyclin E (C), p21 (D), apoptosis (M30) (E), bax (F), bcl-2 (G) and cox-2 (H).

Carcinogenic events in adenomas |

Proliferation-regulating proteins: Cyclin B1, D3 and E, and p21 and p27

Expression of cyclin B1, D3 and E was seen throughout the adenoma crypts. However, a

slight predisposition of the expression of all three cyclins in the luminal side of the crypts was

seen in both HNPCC and sporadic adenomas. Less HGD HNPCC adenomas expressed cyclin

B1 or E compared to HGD sporadic adenomas. In sporadic adenomas a positive correlation

was found between the expression of cyclin B1 and cyclin E and p53 overexpression (p=0.045

and p=0.036, respectively). Proliferation indices were similar in adenomas with low and those

with high immunoreactivity for any of the three proliferation-regulating proteins. This was the

case in HNPCC as well as in sporadic adenomas.

Nuclear immunoreactivity of p21 was seen in the luminal one third of HNPCC and sporadic

adenomatous crypts. The nuclear staining was more intense than in normal adjacent mucosa

but generally less nuclei were stained in adenomatous epithelium than in normal epithelium.

Lower p21 immunoreactivity distinguished HGD HNPCC adenomas from HGD sporadic

adenomas. P27 staining was observed in nuclei in the base of the adenomatous crypts. In

normal epithelium p27 staining was seen in the upper half of the crypt. There was no apparent

difference in p27 expression between HNPCC and sporadic adenomas, irrespective of their

grade of dysplasia.

Apoptosis

M30 positive cells were heterogeneously distributed throughout the adenomas with occasional

small foci. The apoptotic indices in HNPCC adenomas varied from 0 to 6.0 % apoptotic cells

(mean 0.93; median 0.30) and from 0 to 1.37 % cells (mean 0.41; median 0.32) in sporadic

adenomas (not significantly different).

Apoptosis-regulating proteins: bax, bcl-2, cox-2 and p53

Bax staining was more often weak in HGD HNPCC adenomas compared to HGD sporadic

adenomas. Both LGD and HGD HNPCC adenomas expressed more often bcl-2 in comparison

to LGD and HGD sporadic adenomas, respectively. Bax was positively correlated to bcl-2 in

sporadic adenomas (p=0.038) but not in HNPCC adenomas. Apoptotic indices were

comparable between groups with low and high bax expression. Similar results were obtained

for bcl-2. Cox-2 expression was low in the majority of adenomas, however, it was increased

compared to the adjacent normal mucosa. In areas that exhibited features of necrosis or

inflammation cox-2 expression was slightly higher than in the rest of the adenoma. Cox-2

63

Chapter 5 |

expression was similar in HNPCC and sporadic adenomas, irrespective of the grade of

dysplasia. Overexpression of p53 was observed almost exclusively in HGD sporadic

adenomas.

Discussion

It is well known that colorectal neoplasms due to HNPCC differ in several respects, i.e.

clinically, histologically and molecularly, from most sporadic neoplasms. Our present data

expand on these differences and further illustrate that HNPCC adenomas develop along

alternative molecular routes which is exemplified by the variance in proliferation- and

apoptosis-regulating proteins between HNPCC and sporadic adenomas.

Some caution is required regarding our results as not all HNPCC adenomas demonstrated loss

of expression of a mismatch repair protein. In cancers, immunohistochemistry is virtually as

sensitive and specific as microsatellite instability analysis for predicting the presence of MMR

mutations. Whether this is also true for adenomas has not been tested in a large series. Jass

among others, proposed that adenomas in HNPCC develop on a sporadic basis, only then

giving rise to a defective MMR system.2 Even though we classified all our LGD adenomas as

HNPCC adenomas some might well have been sporadic ones. Whether the three HGD

HNPCC adenomas, which expressed all three MMR proteins, are sporadic adenomas is also

unclear.

Proliferation reportedly increases during the carcinogenesis of sporadic colorectal tumors but

no reports exist on this in the carcinogenesis of HNPCC.7 Minor increases in proliferation

between LGD and HGD sporadic adenomas and between LGD and HGD HNPCC adenomas

were observed, but neither was significant. The proliferative index found in the sporadic

adenomas was in accordance with previously reported indices which range from 25% to

55%.7,16 Thus an increase in proliferation does not differentiate HNPCC adenomas from

sporadic ones. Apparently, an increased proliferation rate is not the driving factor that leads to

the accelerated carcinogenesis in HNPCC.

Although the proliferation indices of all HNPCC and sporadic adenomas were similar,

expression of cyclin B1, D3 and E differed, especially between HGD adenomas, being higher

in sporadic ones, and between LGD and HGD sporadic adenomas. An increased cyclin B1

64

Carcinogenic events in adenomas |

expression has been described in sporadic colorectal carcinomas as well as in other

cancers.17,18 Overexpression of cyclin D3 is often found in leukemia but was described only

once in sporadic colorectal cancers and not in colorectal adenomas.19 Yasui et al reported

progressive increase in cyclin E expression from normal mucosa to high-grade dysplastic

adenoma to cancer.16 An increase in expression of cyclins can be interpreted as evidence that

either synthesis of the proteins is deregulated and not limited to a particular phase of the cell

cycle or their proteolytic degradation is impaired. Cyclin B1, D3 and E overexpression could

be a reflection of impaired p53 protein function in sporadic adenomas as illustrated by a

positive correlation between the expression of cyclin B1/cyclin E and p53 overexpression in

our sporadic adenomas.20

In contrast to the strong increase in expression of the three cyclins in sporadic adenomas, only

expression of cyclin B1 increased in HNPCC adenomas. No other data on the

immunoreactivity of cyclin B1, D3, and E in HNPCC lesions have been reported. A

theoretical explanation of the increased cyclin B1, that is, being due to altered p21 expression,

follows below.

In accordance with the largest reported study concerning p21 expression, we observed

approximately half of sporadic adenomas expressing p21.21 The frequently reported

association between downregulation of p21 (being a critical downstream effector in the p53

specific-pathway) and mutant p53 overexpression was not seen in our sporadic adenomas.21 A

striking finding is the decreased expression of p21 in HGD HNPCC adenomas in comparison

to HGD sporadic adenomas. Whether p21 is mutated or downregulated cannot be inferred

from our data. However, mutations in p21 are not frequently found in sporadic human tumors

and as the gene does not carry nucleotide repeat sequences, p21 is probably not more prone to

somatic mutation in mismatch repair deficient tumors than in mismatch repair proficient

tumors. As HNPCC lesions express wild-type p53, which hypothetically should lead to a

normal expression of p21, it can be postulated that p21 in HNPCC is downregulated through a

p53-independent pathway.10,21 A possible pathway through which the downregulation occurs

is the pRB signaling circuit, governed by transforming growth factor beta (TGFβ) and other

extrinsic factors. HNPCC lesions have a high frequency of mutations in the TGFβII receptor

gene that may lead to loss of TGFβ responsiveness. In that case the cytoplasmic Smad protein,

which transduces signals from ligand-activated TGFβ receptors to downstream targets, will

65

Chapter 5 |

receive no signal to induce synthesis of p21, causing the cells to express less p21.22 P21

expression results in cell cycle arrest in G1-phase. With the lack of stimulation of p21 the cell

cycle enters the G2 phase in which cyclin B1 expression increases to promote cell mitosis.

To our knowledge, no data exist on the apoptotic index of HNPCC adenomas. We found that

it did not differ from that of sporadic adenomas. The mean apoptotic index observed in

sporadic adenomas is in agreement with the literature.8 However, the range was large, limiting

the possibility of finding any differences between groups of adenomas or correlations between

the apoptotic index, the expression of apoptosis-regulating proteins and the grade of dysplasia.

Findings on this by others are diverse. Some found that the apoptotic index increases linearly

with the grade of dysplasia, others found no relationship, and even an inverse relationship has

been described .8,23,24

Yagi et al demonstrated that in HNPCC bax mutations occur when adenomas progress to

carcinomas, as in 15% of all adenomas in comparison to 50% of the carcinomas a mutation in

the bax gene could be found.25 In the present study, 14% of the HNPCC adenomas had total

loss of bax expression, compared to only 3 % of the sporadic adenomas (data not specified).

Loss of bax expression may herald the transition from high-grade dysplasia to carcinoma in

HNPCC lesions. Bax and bcl-2 expression were not inversely correlated, nor were the

expression of both correlated to the apoptotic index. Possibly wild type p53 or other

apoptosis-regulating proteins, e.g. PTEN, oppose the resistance to apoptosis due to the altered

bax:bcl-2 ratio.

Cox-2 expression was rather low in both groups of adenomas but in both groups higher than

in the adjacent normal mucosa. Several studies have described increased cox-2 levels at this

early stage of sporadic cancer development.26 Cox-2 expression is reportedly lower in

HNPCC carcinomas than in sporadic ones.27 Data on cox-2 expression in HNPCC adenomas

are restricted to only three adenomas of which two expressed cox-2.27 We found no

differences in cox-2 expression between HNPCC and sporadic adenomas. As cox-2 is induced

by TGFβ and TGFβII receptor mutations are found in HNPCC adenomas it could be deduced

that HNPCC adenomas, similarly to carcinomas, should express less cox-2. However, in

HNPCC carcinomas the reduction of cox-2 was not a direct consequence of TGFβII receptor

mutations.27 The pathway(s) through which cox-2 plays a role should be further elucidated.

66

Carcinogenic events in adenomas |

Unlike sporadic cancers, HNPCC cancers usually carry the wild type p53 tumor suppressor

gene.28 Similarly to HNPCC cancers, we found significantly less p53 overexpression in HGD

HNPCC adenomas in comparison to HGD sporadic adenomas, supporting the hypothesis that

loss of function of bax instead of that of p53 may contribute to the adenoma-carcinoma

transition in HNPCC tumorigenesis.25

The past decade has seen the emergence of new pathways in the development of colorectal

cancer. With the detection of germline mutations in DNA mismatch repair genes the

microsatellite instability pathway in addition to the classical pathway (chromosomal

instability) was identified. MMR dysfunction most probably is not the first event in the

carcinogenesis of HNPCC lesions but heralds development to high-grade dysplasia.29 The

high-grade dysplasia in HNPCC lesions is characterized by a difference in expression of

several proliferation- and apoptosis-regulating proteins in comparison to HGD sporadic

adenomas, supporting the concept of alternative carcinogenic pathways. Whether these

alterations in protein expression give rise to the different clinical behavior of HNPCC and

sporadic adenomas remains to be shown. The fact that changes in expression of proliferation-

and apoptosis-regulating proteins were not associated with changes in proliferation and

apoptosis suggests that also other regulating proteins or pathways play a (possibly more

influential) role in the carcinogenesis of HNPCC-related colorectal cancer.

67

Chapter 5 |

References

1. Fearon ER, Vogelstein B. A genetic model for colorectal tumorigenesis. Cell 1990;61:759-767.

2. Jass JR, Stewart SM, Stewart J, Lane MR. Hereditary non-polyposis colorectal cancer-morphologies, genes and mutations. Mutat Res 1994;310:125-133.

3. Thibodeau SN, French AJ, Roche PC et al. Altered expression of hMSH2 and hMLH1 in tumors with microsatellite instability and genetic alterations in mismatch repair genes. Cancer Res 1996;56:4836-4840.

4. Salahshor S, Kressner U, Pahlman L et al. Colorectal cancer with and without microsatellite instability involves different genes. Genes Chromosomes Cancer 1999;26:247-252.

5. Michael-Robinson JM, Reid LE, Purdie DM et al. Proliferation, apoptosis, and survival in high-level microsatellite instability sporadic colorectal cancer. Clin Cancer Res 2001;7:2347-2356.

6. Edmonston TB, Cuesta KH, Burkholder S et al. Colorectal carcinomas with high microsatellite instability: defining a distinct immunologic and molecular entity with respect to prognostic markers. Hum Pathol 2000;31:1506-1514.

7. Hao X, Du M, Bishop AE, Talbot IC. Imbalance between proliferation and apoptosis in the development of colorectal carcinoma. Virchows Arch 1998;433:523-527.

8. Koornstra JJ, de Jong S, Hollema H, de Vries EGE, Kleibeuker JH. Changes in apoptosis during the development of colorectal cancer: a systematic review of the literature. Crit Rev Oncol Hematol 2003;45:37-53.

9. Hunter T, Pines J. Cyclins and cancer. Cell 1991;66:1071-1074.

10. Levine AJ. p53, the cellular gatekeeper for growth and division. Cell 1997;88:323-331.

11. Tsujii M, Kawano S, DuBois RN. Cyclooxygenase-2 expression in human colon cancer cells increases metastatic potential. Proc Natl Acad Sci U S A 1997;94:3336-3340.

12. Vasen HF, Watson P, Mecklin JP, Lynch HT. New clinical criteria for hereditary nonpolyposis colorectal cancer proposed by the International Collaborative group on HNPCC. Gastroenterology 1999;116:1453-1456.

13. Jass JR, Sobin LH, Watanabe H. The World Health Organization's histologic classification of gastrointestinal tumors. A commentary on the second edition. Cancer 1990;66:2162-2167.

14. Ciaparrone M, Yamamoto H, Yao Y et al. Localization and expression of p27KIP1 colorectal carcinogenesis. Cancer Res 1998;58:114-122.

68

Carcinogenic events in adenomas |

15. Hao XP, Ilyas M, Talbot IC. Expression of Bcl-2 and p53 in the colorectal adenoma-carcinoma sequence. Pathobiology 1997;65:140-145

16. Yasui W, Kuniyasu H, Yokozaki H et al. Expression of cyclin E in colorectal adenomas and adenocarcinomas: correlation with expression of Ki-67 antigen and p53 protein. Virchows Arch 1996;429:13-19.

17. Wang A, Yoshimi N, Ino N, Tanaka T, Mori H. Overexpression of cyclin B1 in human colorectal cancers. J Cancer Res Clin Oncol 1997;123:124-127.

18. Kawamoto H, Koizumi H, Uchikoshi T. Expression of the G2-M checkpoint regulators cyclin B1 and cdc2 in nonmalignant and malignant human breast lesions: immunocytochemical and quantitative image analyses. Am J Pathol 1997;150:15-23.

19. Bartkova J, Thullberg M, Slezak P et al. Aberrant expression of G1-phase cell cycle regulators in flat and exophytic adenomas of the human colon. Gastroenterology 2001;120:1680-1688.

20. Innocente SA, Abrahamson JL, Cogswell JP, Lee JM. p53 regulates a G2 checkpoint through cyclin B1. Proc Natl Acad Sci U S A 1999;96:2147-2152.

21. Doglioni C, Pelosio P, Laurino L et al. p21/WAF1/CIP1 expression in normal mucosa and in adenomas and adenocarcinomas of the colon: its relationship with differentiation. 1996;179:248-253.

22. Datto MB, Li Y, Panus JF et al. Transforming growth factor beta induces the cyclin-dependent kinase inhibitor p21 through a p53-independent mechanism. Proc Natl Acad Sci U S A 1995;92:5545-5549.

23. Nomura M, Watari J, Yokota K et al. Morphogenesis of nonpolypoid colorectal adenomas and early carcinomas assessed by cell proliferation and apoptosis. Virchows Arch 2000;437:17-24.

24. Aotake T, Lu CD, Chiba Y, Muraoka R, Tanigawa N. Changes of angiogenesis and tumor cell apoptosis during colorectal carcinogenesis. Clin Cancer Res 1999;5:135-142.

25. Yagi OK, Akiyama Y, Nomizu T et al. Proapoptotic gene BAX is frequently mutated in hereditary nonpolyposis colorectal cancers but not in adenomas. Gastroenterology 1998;114:268-274.

26. Eberhart CE, Coffey RJ, Radhika A et al. Up-regulation of cyclooxygenase 2 gene expression in human colorectal adenomas and adenocarcinomas. Gastroenterology 1994;107:1183-1188.

27. Sinicrope FA, Lemoine M, Xi L et al. Reduced expression of cyclooxygenase 2 proteins in hereditary nonpolyposis colorectal cancers relative to sporadic cancers. Gastroenterology 1999;117:350-358.

28. Losi L, Ponz de Leon M, Jiricny J et al. K-ras and p53 mutations in hereditary non-polyposis colorectal cancers. Int J Cancer 1997;74:94-96.

69

Chapter 5 |

29. Rijcken FEM, Hollema H, Kleibeuker JH: Proximal adenomas in hereditary non-polyposis colorectal cancer are prone to rapid malignant transformation. Gut 2002;50:382-6.

70

|

6

Cell cycle regulators and apoptosis

associated proteins in relation to

proliferative activity and degree of

apoptosis in HNPCC versus

sporadic endometrial carcinoma

F E M Rijcken1, A G J van der Zee2, T van der Sluis3, W

Boersma-van Ek1, J H Kleibeuker1,

H Hollema3

Department of Gastroenterology1, Gynecology and Obstetrics2,

and Pathology3, University Medical Center Groningen, The

Netherlands.

Histopathology 2006;48:275-85

Chapter 6 |

Abstract

Background - Mismatch repair gene malfunction occurs early in the carcinogenesis of

hereditary nonpolyposis colorectal cancers (HNPCCs) leading to an accelerated

accumulation of mutations and possibly to change in expression of cell cycle proteins.

There is strong evidence that tumorigenesis in HNPCCs differs from sporadic ones.

HNPCC-related endometrial cancers are less well studied. Our aim was to compare

expression of cell cycle and apoptosis-related proteins in relation to proliferation and

apoptosis in HNPCC-related and sporadic endometrial cancers to identify differences in

their carcinogenetic pathways.

Methods and results - 18 HNPCC-related endometrial cancers, each matched by tumor

type, stage and grade with two sporadic endometrial cancers, were examined for

proliferation, apoptosis and the expression of oestrogen and progesterone receptors,

cyclin B1, D3 and E, p21, p27, bcl2, bax, p53 and cox-2. No differences in proliferation

and apoptosis indices were detected between HNPCC and sporadic endometrial cancers.

Cyclin B1 expression in HNPCC was significantly higher than in sporadic cancers. More

HNPCC-related endometrial cancers had total loss of bax expression.

Conclusions - Apart from differences in cyclin B1 and bax expression, HNPCC-related

and sporadic endometrial cancer are comparable. The subtle differences detected are

consistent with the minor clinical diversity between HNPCC-related and sporadic

endometrial cancers.

72

HNPCC vs sporadic endometrial carcinoma |

In hereditary nonpolyposis colorectal cancer (HNPCC) germline mutations in DNA mismatch

repair (MMR) genes, i.e. MLH1, MSH2 and MSH6, lead to an increased risk of colorectal as

well as extracolonic cancers. Endometrial cancer is the most frequently occurring extracolonic

tumorand in genetically predisposed women the cumulative lifetime risk for endometrial

cancer may exceed that of colorectal cancer. HNPCC endometrial cancers are on average

diagnosed at an earlier age than nonhereditary endometrial cancers (further referred to as

sporadic endometrial cancers) but have a prognosis similar to sporadic endometrial cancers.1

In HNPCC patients, loss of mismatch repair function will, among others, lead to mutations in

genes with microsatellites (short tandem repeat sequences) in their coding regions. Mutations

in these genes may lead to alternative pathways of carcinogenesis compared sporadic cases.2

The consequences of MMR dysfunction on the pathogenesis of endometrial cancers, including

proliferative activity and the rate of apoptosis, and the activity of proliferation- and apoptosis-

regulating proteins have received little attention.

The current study was undertaken to explore differences in carcinogenic pathways between

HNPCC-related and sporadic endometrial cancer by evaluating proliferation and apoptotic

indices and the immunohistochemical expression of proliferation- and apoptosis-regulating

proteins. We investigated proliferation-stimulating proteins, cyclin B1, D3 and E, which

regulate the cell cycle at the G2/M, G0/S and G1/S phase. Moreover, cyclin B1 and E have

been shown to be major proliferation regulating genes in sporadic endometrial cancers.3 We

determined the expression of p53, the product of a tumor suppressor gene, as it is an important

gene in the apoptotic route. P53 blocks cell cycle progression from G1 into S phase by G1

arrest through transcriptional regulation of the cyclin-dependent kinase inhibitor p21. P53

overexpression in sporadic endometrial cancer is related to high-grade morphology or a non-

endometrioid morphology.4,5 Whether p53 mutations leading to p53 overexpression play a

role in HNPCC endometrial cancer is not known. P27 is another cyclin-dependent kinase

inhibitor downstream of p21 that also ultimately leads to cell cycle arrest. A second protein

regulated by p53 is bcl-2, which appears to function as an associated heterodimer of bcl-2-bax

that inhibits apoptosis. Bax, alternatively, promotes apoptosis. It is vulnerable to mutations in

case of MMR dysfunction as the bax gene contains a nucleotide repeat sequence.6

Cyclooxygenase-2 (cox-2) has gained increasing attention as it is a potential target for

chemoprevention. Cox-2 expression is elevated in several malignant tumors and probably has

73

Chapter 6 |

a role in programmed cell death.7 Finally, given that the presence of oestrogen receptors (ER)

and progesterone receptors (PR) has been linked to better prognosis and to the carcinogenesis

of endometrial cancers, the expression of ER and PR were also determined.8

Materials and methods

Paraffin-embedded tissue blocks of HNPCC-related and sporadic (control group) endometrial

cancers were retrieved from the archives of the Department of Pathology of the University

Medical Center Groningen. Women fulfilling the Amsterdam Criteria 9 and/or having a

known germline mutation in one of the DNA mismatch repair genes were included in the

HNPCC group. The control group consisted of endometrial cancers from women without a

family history of endometrial cancer or colorectal cancer. Tumors were classified according to

histological type, stage and grade following the current recommendations of the International

Federation of Gynaecology and Obstetrics (FIGO).10 Each HNPCC-related endometrial

cancer was matched with two sporadic endometrial cancers of the same tumor type and FIGO

stage and grade. It was not possible to match the HNPCC and sporadic tumors by age at

diagnosis. Endometrial cancers from patients with a history of hormone replacement therapy

were excluded from the study.

Immunohistochemistry

A slide of each endometrial cancer was stained in one run per antibody. If staining was not

adequate in slides, these specific slides with inclusion of a well-stained slide from the first

batch as reference were stained again. As negative control, the primary antibody was

substituted with 1% bovine serum albumin (BSA) in phosphate buffered saline (PBS).

Table 1 summarizes the 11 primary antibodies, the companies from which they were

purchased, the dilution at which they were used and the corresponding technique for antigen

retrieval. Serial 3µm thick sections were cut from paraffin blocks, fixed onto coated slides and

deparaffinized. For the high-pressure cooker antigen retrieval method, slides were immersed

in 200µl blocking reagent (Boehringer Mannheim, Germany) and underwent 3 sessions of 5

minutes at 115°C. The endogenous peroxidase activity was quenched by incubation with 30%

H2O2 in PBS for 30 minutes. Subsequently, the primary antibody diluted in PBS-1% BSA was

74

HNPCC vs sporadic endometrial carcinoma |

applied for one hour at room temperature. In the microwave antigen retrieval method

deparaffinized, rehydrated sections were submersed either in preheated 10mM citrate buffer

(pH 6.0) or in 1mM EDTA (pH 8.0), as noted in table 1, and heated for 8 minutes at 700 watts

in a microwave. After cooling at room temperature for 7 minutes the sections were thoroughly

rinsed with PBS for 5 minutes and the primary antibody was applied, diluted in PBS-1% BSA

as described in table 1, for one hour at room temperature. In both the high-pressure cooker

and the microwave retrieval method, the sections were consecutively incubated with

secondary antibody, i.e. rabbit antimouse peroxidase (diluted 1:50 in PBS-1% BSA) and

tertiary antibody, i.e. goat antirabbit peroxidase (diluted 1:50 in PBS-1% BSA) for 30

minutes. The peroxidase activity was visualized with diaminobenzidine. The sections were

counterstained with haematoxylin.

Apoptosis was quantified on haematoxylin and eosin-stained sections using light microscopy

with an eyepiece grid. Apoptotic cells were morphologically identified using the criteria as

described by others.11,12 This approach has been shown to have the same yield as techniques

such as TUNEL.13 Apoptotic bodies shed into the gland lumen or in necrotic areas were not

counted.

Scoring Method

Without knowledge of the clinical data, two authors (FR and HH) scored all stained slides.

The immunoreactivity of each antibody was scored in the same area of the carcinoma. MMR

protein expression was scored as positive (present) or negative (absent). Proliferation was

quantified as labelling index, i.e. the percentage of the positively stained nuclei. Using an eye-

piece grid, the apoptotic index was quantified by counting the number of apoptotic cells

crossing the consecutive horizontal lines per total number of counted epithelial cells on the

horizontal lines in 10 high power fields (HPF= 400 x magnification). Estrogen and

progesterone receptors, cyclin B1, D3 and E, p21 and cox-2 were scored quantitatively as a

labelling index. The intensity of p27, bax, and bcl-2 staining was heterogeneous within a

tumor. These assessments were therefore quantified with a weighted score by multiplying the

intensity (1= weak staining, 2= moderate, 3= strong) by the percentage of stained cells. A cut-

off level of 50 % of cells strongly staining for p53 was used to define overexpression. In our

laboratory, this cut-off level correlates well to p53 somatic mutations in endometrial cancer

(data not published).

75

Chapter 6 |

Statistics

Statistical analyses were performed using SPSS (Statistical Package for the Social Sciences,

Munich, Germany) for Windows. The statistical difference between HNPCC and sporadic

endometrial cancers for each protein was calculated using the nonparametric Mann-Whitney

U-test. The relations between each of the investigated parameters were analyzed using

Pearson's correlation coefficient. For both tests, p<0.05 was considered significant.

Table 1. Antibodies and antigen retrieval methods used for immunohistochemistry.

Protein Antigen retrieval Clone Company Dilution

ER Microwave EDTA 6F11 Novocastra, Newcastle, UK 1:400 PR Microwave EDTA 1A6 Novocastra, Newcastle, UK 1:400 Ki-67 High-pressure cooker MIB-1 Immunotech, Westbrook, ME 1:400 Cyclin B1 High-pressure cooker 7A9 Novocastra, Newcastle, UK 1:200 Cyclin D3 Microwave EDTA DCS-22 Novocastra, Newcastle, UK 1:50 Cyclin E Microwave EDTA 13A3 Novocastra, Newcastle, UK 1:10 P21 High-pressure cooker WAF1(Ab-1) Oncogene, Darmstadt, Germany 1:50 P27 High-pressure cooker 1B4 Novocastra, Newcastle, UK 1:50 Bax Microwave citrate B-9 St Cruz Biotechnology, St Cruz, CA 1:200 Bcl-2 High-pressure cooker MAb 124 DAKO, Glostrup, Denmark 1:50 P53 High-pressure cooker B-p53-12-1 Biogenex, San Ramon, CA 1:400 Cox-2 Microwave EDTA 33 Transduction Lab., Lexington, KY 1:50 MLH1 High-pressure cooker G168-728 PharMingen, San Diego, USA 1:500 MSH2 High-pressure cooker Ab-2 Calbiochem, San Diego, USA 1:100 MSH6 High-pressure cooker 44 Transduction Lab., Lexington, KY 1:200

Results

Patient characteristics

A total of 18 HNPCC-related endometrial cancers could be retrieved from the archive and

were included in the study. Six women had a germline mutation in hMLH1, two women had a

mutation in hMSH2 and two in hMSH6. The remaining eight women belonged to families

fulfilling the Amsterdam criteria. Either no mutational study had been done in these patients

or a mutation had not yet been found. One woman belonged to an HNPCC-family in which

the mutation (in MLH1) was known but the patient chose not to be genetically tested. Three

women were postmenopausal, one was in the climacteric phase and the remaining 14 women

were premenopausal at time of diagnosis. The average age at diagnosis was 45 years. In the

76

HNPCC vs sporadic endometrial carcinoma |

control group, the average age at diagnosis was 65 years. Twenty-six of these women were

postmenopausal, one was in the climacteric phase and nine were premenopausal at diagnosis.

Histology

Fifteen HNPCC-related endometrial cancers were endometrioid adenocarcinomas, one was a

pure clear cell endometrial cancer and two were mixed clear cell and endometrioid

adenocarcinomas (clear cell component <50%). Data concerning stage and grade of the

HNPCC carcinomas are summarized in table 2. Two sporadic endometrial cancers were

matched to each HNPCC endometrial cancer for tumor type, stage and grade.

Table 2. Tumortype, stage and grade of the HNPCC endometrial cancers. Tumour

stage & grade

n

Tumour

stage & grade

n

IA GI 3 IIIA GIII 1

IB GI 4 IIIC GIII 1

IB GII 2 III A (clear cell) 3

IIA GII 1

IIB GrII 2

IIB GIII 1

Imunnohistochemistry

Mismatch repair

In proven DNA mismatch repair gene mutation carriers, the endometrial tumors showed loss

of the MMR protein expression corresponding to the mutated gene. Three of the eight

HNPCC-related endometrial cancers from women fulfilling the Amsterdam criteria but from

whom the genetic status was not known stained positive for all three MMR proteins, four

showed loss of MLH1 expression and one of MSH2 expression. All sporadic endometrial

cancers expressed the three MMR proteins.

77

Chapter 6 |

The immunoreactivity of the proliferation- and apoptosis-regulating proteins was

heterogeneous in most HNPCC and sporadic tumors. Figure 1 illustrates several

immunohistochemical results in one HNPCC-related endometrial tumor. The mean

immunoreactivity for each protein is noted in table 3. Estrogen and progesterone receptor

staining in HNPCC-related endometrial cancers was not significantly different from that in

sporadic endometrial cancers. A wide range in Mib-1 labelling indices was observed in

HNPCC-related and sporadic endometrial cancer: 5-80% and 5-70%, respectively. A trend

towards higher proliferation in HNPCC-related endometrial cancers compared to sporadic

endometrial cancers was observed (p=0.084). The immunoreactivity of proliferation-

regulating proteins, cyclin D3 and E, and cyclin-dependent kinase inhibitors, p21 and p27,

was similar in HNPCC-related and sporadic endometrial cancers. However, cyclin B1

expression was significantly higher in HNPCC-related than in sporadic endometrial cancers.

The apoptotic index ranged from 1-22% in HNPCC-related and 1-32% in sporadic

endometrial cancers. The apoptotic index in HNPCC-related endometrial cancers was not

significantly different from that in sporadic cancers. The apoptosis-regulating proteins, bax

and bcl-2, were expressed in a similar number of cells in HNPCC-related as in sporadic

endometrial cancers. However, total loss of bax expression was observed in three HNPCC-

related cancers while all sporadic cancers expressed bax. Bax expression in HNPCC-related

and sporadic endometrial cancer is significantly different when using this classification

method, presence or absence of bax expression (p=0.013). The ratio bax:bcl-2 in HNPCC-

related was not significantly different from that in sporadic endometrial cancers (mean

bax:bcl-2, 46:1 vs. 29:1, p=0.868, respectively). No significant difference was observed in

cox-2 and p53 expression between HNPCC-related and sporadic endometrial cancers. P53

overexpression was seen in two (11%) of the 18 HNPCC-related endometrial cancers. One of

these cancers had pure clear cell histology, the other was a mixed clear cell and endometrioid

tumour. In the control group, 11 (31%) tumours, five clear cell and six endometrioid (five

grade III and one grade I), overexpressed p53.

78

HNPCC vs sporadic endometrial carcinoma |

Table 3. Immunohistochemistry results of the HNPCC and sporadic endometrial

cancers.

Staining categories † p-value

HNPCC Sporadic

Estrogen receptor (%) 45 ± 10 38 ± 15 n.s.

Progesterone receptor (%) 28 ± 9 36 ± 6 n.s.

Mib-1 labelling index (%) 35 ± 6 25 ± 3 n.s.

Cyclin B1 (%) 27 ± 6 12 ± 2 p=0.038

Cycin D3 (%) 20 ± 7 9 ± 3 n.s.

Cyclin E (%) 39 ± 7 27 ± 5 n.s.

P21 (%) 11 ± 4 13 ± 3 n.s.

P27 (%) 48 ± 9 38 ± 6 n.s.

Apoptotic index (%) 4 ± 1 3 ± 1 n.s.

Bax (%) 144 ± 26 190 ± 15 n.s.

Bcl-2 (%) 67 ± 21 81 ± 15 n.s.

Cox-2 (%) 33 ± 8 36 ± 6 n.s.

P53 (n) 2/18 11/36 n.s.

Correlation

In table 4 the correlation between various immunohistochemical results are noted.

In HNPCC-related neoplasms, poor tumor differentiation (high grade) was correlated with

higher stage (p=0.025). PR immunoreactivity was inversely correlated with the grade of the

HNPCC-related endometrial cancers (p=0.002). Absent ER staining (p=0.039) and p53

overexpression (p<0.001) were correlated with clear cell histology. ER immunoreactivity was

related to bcl-2 expression (p=0.018) and inversely related to p53 expression (p=0.022).

Neither the Mib-1 labelling index nor the apoptotic index was related to the grade or stage of

the HNPCC-related endometrial cancers. Cyclin B1 expression was related to stage (p=0.026),

grade (p=0.021) and a higher Mib-1 labelling index (p=0.003). The degrees of expression of

the apoptosis-regulating proteins were not correlated with the apoptotic index.

79

Chapter 6 |

Table 4A.

P53

1.00

0.10

6

<.00

1

-0.2

79

-0.0

22

-0.6

80

-0.3

11

-0.8

23

0.09

6

0.99

2

0.06

7

0.95

1

0.93

9

0.27

2

0.06

6

-

Cox

0.75

7

0.06

5

0.20

3

0.73

9

0.96

3

0.88

5

-0.8

93

-0.8

64

0.05

9

-0.5

19

0.06

3

-0.2

58

-0.4

82

0.70

8

-

Bcl

2

-0.1

25

-0.1

48

0.16

0

0.65

6

0.01

8

-0.2

08

-0.2

36

-0.9

79

-0.6

82

0.38

8

0.57

9

-0.8

85

0.62

2

-

Bax

-0.7

00

-0.0

77

-0.8

56

0.34

5

-0.9

38

-0.2

70

-0.3

52

-0.3

59

-0.9

68

0.85

8

0.82

2

0.18

7

-

Apo

p

0.65

7

0.43

2

0.21

4

-0.1

52

-0.1

42

0.54

7

0.23

1

0.68

9

0.25

4

0.23

6

0.23

5

-

P27

-0.4

74

0.31

3

0.34

4

-0.4

32

0.83

7

-0.0

74

-0.2

58

0.90

1

0.76

6

-0.3

71

-

P21

-0.3

02

-0.4

58

0.88

2

0.42

4

-0.6

77

0.12

1

-0.5

58

0.05

9

0.78

3

-

E

0.53

8

0.10

9

0.19

5

-0.7

04

-0.2

15

0.47

6

0.66

4

-0.4

81

-

D3

-0.4

59

-0.7

11

-0.5

45

0.20

6

-0.8

24

0.18

4

0.22

4

-

B1

0.02

6

0.02

1

-0.6

68

-0.2

92

-0.7

50

0.00

3

-

MIB

0.28

1

0.06

3

0.51

2

-0.9

52

-0.5

67

-

ER

-0.8

05

-0.1

29

-0.0

39

0.20

4

-

PR

-0.1

80

-0.0

02

-0.1

66

-

His

to

-0.6

68

0.03

5

-

Gra

de

0.02

5

-

Stag

e

-

Stag

e

Gra

de

His

to

PR

ER

MIB

B1

D3 E P21

P27

Apo

p

Bax

Bcl

2

Cox

P53

80

HNPCC vs sporadic endometrial carcinoma |

Table 4B. P5

3

0.00

6

<0.0

01

0.00

1

-0.0

75

-0.0

01

0.80

1

0.40

2

-0.6

64

0.04

7

0.07

7

-0.5

38

0.38

6

0.49

3

0.63

2

0.16

8

-

Cox

0.16

2

0.06

3

0.03

6

0.05

2

-0.9

65

0.06

1

0.05

7

0.52

4

0.00

3

0.04

3

0.18

3

0.09

4

0.75

3

0.06

8

-

Bcl

2

-0.5

80

-0.2

09

-0.0

39

0.03

1

0.02

7

-0.2

35

0.43

7

-0.1

29

0.50

3

0.04

2

0.00

6

-0.0

97

0.00

4

-

Bax

0.73

3

-0.5

64

-0.9

56

0.00

1

0.00

8

-0.8

88

-0.9

72

-0.2

25

0.60

9

0.01

4

0.84

2

0.14

2

-

Apo

p

0.75

8

0.58

9

0.25

4

-0.1

45

-0.2

58

0.36

5

-0.6

89

-0.1

29

0.82

5

0.52

1

0.15

6

-

P27

-0.1

59

-0.0

78

-0.4

31

-0.8

30

-0.4

81

-0.2

51

-0.9

47

0.98

8

0.83

5

0.63

5

-

P21

0.05

9

0.07

1

0.05

6

-0.3

37

0.58

9

0.96

0

0.02

3

0.58

8

0.04

2

-

E 0.67

5

0.02

5

0.78

1

-0.0

04

-0.8

24

0.02

8

0.09

2

0.00

7

-

D3

0.71

9

-0.3

07

-0.6

31

-0.0

82

-0.1

71

0.21

2

-0.5

81

-

B1

0.99

0

0.00

7

0.95

9

-0.5

02

-0.9

83

-0.4

01

-

MIB

0.04

3

0.03

8

-0.1

28

-0.5

53

0.16

3

-

ER

0.16

6

-0.0

26

0.25

4

0.00

1

-

PR

-0.4

71

-0.0

44

-0.6

44

-

His

to

0.00

1

0.00

3

-

Gra

de

0.02

5

-

Stag

e

-

Stag

e

Gra

de

His

to

PR

ER

MIB

B1

D3

E P21

P27

Apo

p

Bax

Bcl

2

Cox

P53

81

Chapter 6 |

In sporadic endometrial cancers high stage was correlated with a high Mib-1 labelling index

(p=0.043) and p53 overexpression (p=0.004). PR immunoreactivity was inversely correlated

with tumor grade (p=0.044), while Mib-1 labelling index (p=0.038), cyclin B1 expression

(p=0.007), cyclin E expression (p=0.025) and overexpression of p53 (p<0.001) were

positively correlated with grade in sporadic endometrial cancers. Only cyclin E expression

(p=0.035) was positively correlated with the Mib-1 labelling index. The apoptotic index was

not correlated with stage or grade in sporadic endometrial cancers. The apoptosis-regulating

proteins did not correlate with the apoptotic index. However, an increased bax:bcl-2 ratio was

related to an increased apoptotic index (p=0.030). Bax and bcl-2 expression were both

positively correlated with ER (p=0.008, p=0.027, respectively) and PR (p=0.001, p=0.031,

respectively) immunoreactivity and with p21 expression (p=0.014, p=0.042, respectively).

Cox-2 expression in sporadic endometrial cancers was positively correlated with cyclin B1

(p=0.042) and cyclin E (p=0.003) expression but not with the Mib-1 labelling index. Clear

cell histology was correlated with high cox-2 expression (p=0.036) and overexpression of p53

(p=0.003).

Discussion

Endometrial cancer is the most common extracolonic tumor in HNPCC but little is known

about its carcinogenesis and if it differs from non-hereditary endometrial cancers. The present

study reports proliferation and apoptosis in relation to a number of proliferation- and

apoptosis-regulating proteins in HNPCC-related and sporadic endometrial cancers. In contrast

to a similar immunohistochemical study of colorectal lesions 14, immunohistochemistry

demonstrated few differences between HNPCC-related and sporadic endometrial cancers.

Apparently, the differences in carcinogenesis of HNPCC-related and sporadic endometrial

cancers are subtle.

The endometrial cancers included in our study are representative for those diagnosed in

HNPCC patients. The stage and grade distribution of the HNPCC-related endometrial cancers

are in accordance with the HNPCC endometrial cancers reported by Boks et al. 1 However,

the series of Boks et al. did not include clear cell endometrial cancers 1, whereas we found one

pure clear cell and two mixed endometrioid/clear cell endometrial cancers. Even though no

HNPCC-related clear cell endometrial cancers have been described in the literature, seven

82

HNPCC vs sporadic endometrial carcinoma |

microsatellite unstable sporadic clear cell and mixed endometrioid and clear cell carcinomas

have been described.15-17 Approximately 10% of clear cell carcinomas have been found to

demonstrate the mutator phenotype, microsatellite instability.17 The percentage of clear cell

endometrial cancers in our HNPCC population is similar to the incidence of clear cell

endometrial cancers found in the general population in the northern part of the Netherlands

(Schaapveld, Comprehensive Cancer Center North Netherlands, unpublished). Thus tumor

type does not discern HNPCC-related endometrial cancers from sporadic ones.

A tendency towards an increased proliferation rate was observed in HNPCC-related

endometrial cancers in comparison with sporadic endometrial cancers. HNPCC-related

endometrial cancers differed from sporadic endometrial cancers by increased cyclin B1

expression. The increased cyclin B1 expression possibly indicates that more cells were in the

G2 phase of the cell cycle. Having passed the G1 phase, the cell may become refractory to

external stimuli, such as growth inhibitors. In accordance with the above-mentioned theory,

we found no inverse correlation between p21 or p27 expression and the Mib-1 labeling index.

Cyclin B1, however, correlated with the Mib-1 labeling index, supporting the concept that

cyclin B1 is a major cell cycle proliferation regulator in HNPCC-related endometrial cancers.

The immunohistochemical apoptotic profile of HNPCC-related endometrial cancers did not

differ from that of sporadic endometrial cancers. No significant differences were found in the

apoptotic index or expression of apoptosis regulators. However, when a dual scoring method

(presence versus absence of staining) was applied, bax expression was significantly different

between HNPCC-related and sporadic endometrial cancers. The finding of total loss of bax

expression in three HNPCC-related endometrial cancers while all sporadic endometrial

cancers expressed bax to a certain degree, points to bax as a mutation-susceptible gene in

HNPCC. Vassileva et al. concluded that bax is an early mutational target in the development

of microsatellite unstable endometrial cancers18, while de Leeuw et al. demonstrated

instability of the bax gene in 22% of HNPCC-related endometrial cancers.19 The bax gene

contains a repeat sequence and is thus vulnerable to mutation caused by MMR dysfunction.

Even though bax seems to be a target gene of the microsatellite phenotype in endometrial

cancers, the carcinogenic consequences are disputable. The altered bax expression had a

limited functional role as the apoptotic index was not influenced. Apparently the loss of

83

Chapter 6 |

expression of bax, which theoretically should result in a decrease in apoptosis, was

compensated by other apoptosis-inducing pathway(s).

In general, the immunohistochemical profile of the sporadic endometrial cancers was in

accordance with previous studies.3,19-31 In the present study, the mean Mib-1 labelling index

was 25%, which is slightly lower than that reported by others, 30-40%.11,31-33 The manner of

selection of the control group could explain this slight discrepancy in Mib-1 labelling index.

The Mib-1 labelling index in the sporadic endometrial cancers seemed to be predominantly

regulated by cyclin E. Previously cyclin B1 and E have been reported to be involved in

proliferation regulation in endometrial cancer.3 Cyclin E pushes the cell through the cell cycle

from late G1 phase to DNA synthesis. On the other hand, p21 expression results in cell cycle

arrest in G1 phase. Accumulation of p21 is relatively common in endometrial cancers 20,22 as

also seen in our study but p21 expression was not found to be inversely correlated to

proliferation. Michieli et al. observed that at least two pathways can lead to increased p21

expression; a p53-independent pathway, triggered by growth factors and associated with cell

growth, and a p53-dependent pathway, elicited by DNA damage and resulting in growth

arrest.34 As in most studies, we found no relation between p53 and p21.20,22 The more

abundant cyclin E may neutralize the possible proliferation inhibiting effects of p21. Cox-2,

which is expressed in most endometrial cancers, especially in clear cell carcinomas, was

correlated to increased cyclin E expression and tends to correlate positively with Mib-1

labelling indices. The observed cox-2 immunoreactivity offers perspectives for

chemoprevention with non-steroidal anti-inflammatory drugs of HNPCC-related as well as

sporadic endometrial cancers.

The apoptotic index observed in the sporadic endometrial cancers was in accordance with

previously documented apoptotic indices.11 In the present study, one third of the sporadic

tumors overexpressed p53. In accordance with other studies, P53 overexpression strongly

correlates with the histological type (non-endometrioid versus endometrioid) and among

endometrioid carcinoma with grade and stage. 4,5 P53 overexpression has previously also been

related to hormone receptor negative tumors and poor survival.31 Our results support the

former relation. We found no direct correlation between the apoptosis-regulating proteins and

the apoptotic index, leading us to speculate on other apoptosis-regulating genes which play a

larger role in apoptosis regulation in endometrial cancers. A possible candidate is PTEN, an

84

HNPCC vs sporadic endometrial carcinoma |

essential lipid phosphate that is a negative regulator of anti-apoptosis pathways and has been

found to be mutated in many endometrial cancers. 35

The subtle differences found between HNPCC-related and sporadic endometrial cancers are in

accordance with the clinical study of Boks et al., who found no difference in survival between

HNPCC-related and sporadic tumors.1 A number of factors limited the possibility of detecting

elucid differences if present. First, the present study is restricted by the relatively small

number of HNPCC-related endometrial cancers of MLH1, MSH2 and MSH6 mutation carriers.

Not all HNPCC-related endometrial cancers were from proven mutation carriers or had a

known microsatellite status. Secondly, HNPCC-related endometrial cancers have a lower

mutational rate than HNPCC-related colorectal tumors, which may explain why clear

differences can be found by immunohistochemistry between HNPCC-related and sporadic

colorectal tumors, but not between the two groups of endometrial cancers.6 Furthermore, in

HNPCC-related endometrial cancers, the microsatellite pattern has been shown to be very

heterogeneous within and between tumours.36 Intra- and intertumoral heterogeneity of most

immunohistochemical reactivity was also observed in the HNPCC-related endometrial

cancers, which may reflect diverse microsatellite patterns. Due to the wide range in

immunoreactivity observed for each protein, a larger group of tumors may be necessary to

detect statistical differences.

In conclusion, despite the underlying differences in pathogenesis, dysfunctional and

functional mismatch repair genes, the carcinogenic pathway of HNPCC-related endometrial

cancers differs only in a subtle manner (i.e. higher cyclin B1 expression and more often total

loss of bax expression) from sporadic endometrial cancers. The subtle differences detected are

consistent with the minor clinical disparity between HNPCC-related and sporadic endometrial

cancers.

85

Chapter 6 |

References

1. Boks DE, Trujillo AP, Voogd AC, Morreau H, Kenter GG, Vasen HF. Survival analysis of endometrial carcinoma associated with hereditary nonpolyposis colorectal cancer. Int J Cancer 2002;102;198-200.

2. Risinger JI, Berchuck A, Kohler MF, Watson P, Lynch HT, Boyd J. Genetic instability of microsatellites in endometrial carcinoma. Cancer Res 1993;53;5100-5103.

3. Milde-Langosch K, Bamberger AM, Goemann C et al. Expression of cell-cycle regulatory proteins in endometrial carcinomas: correlations with hormone receptor status and clinicopathologic parameters. J Cancer Res Clin Oncol 2001;127;537-544.

4. Lax SF, Kendall B, Tashiro H, Slebos RJ, Hedrick L. The frequency of p53, K-ras mutations, and microsatellite instability differs in uterine endometrioid and serous carcinoma: evidence of distinct molecular genetic pathways. Cancer 2000;88;814-24.

5. Sherman ME, Bur ME, Kurman RJ. p53 in endometrial cancer and its putative precursors: evidence for diverse pathways of tumorigenesis. Hum Pathol 1995;26;1268-74.

6. Duval A, Reperant M, Compoint A et al. Target Gene Mutation Profile Differs between Gastrointestinal and Endometrial Tumors with Mismatch Repair Deficiency. Cancer Res 2002;62;1609-1612.

7. Eberhart CE, Coffey RJ, Radhika A, Giardiello FM, Ferrenbach S, DuBois RN. Up-regulation of cyclooxygenase 2 gene expression in human colorectal adenomas and adenocarcinomas. Gastroenterology 1994;107;1183-1188.

8. Fukuda K, Mori M, Uchiyama M, Iwai K, Iwasaka T, Sugimori H. Prognostic significance of progesterone receptor immunohistochemistry in endometrial carcinoma. Gynecol Oncol 1998; 69;220-225.

9. Vasen HF, Watson P, Mecklin JP, Lynch HT. New clinical criteria for hereditary nonpolyposis colorectal cancer (HNPCC, Lynch syndrome) proposed by the International Collaborative group on HNPCC. Gastroenterology 1999;116;1453-1456.

10. Scully RE, Bonfiglio TA, Kurman RJ, Silverberg SG, Wilkinson K. International histological classification and typing of female genital tract tumours. New York: Springer-Verlag 1994.

11. Ioffe OB, Papadimitriou JC, Drachenberg CB. Correlation of proliferation indices, apoptosis, and related oncogene expression (bcl-2 and c-erbB-2) and p53 in proliferative, hyperplastic, and malignant endometrium. Hum Pathol 1998;29;1150-1159.

12. Mourits MJ, Hollema H, De Vries EG, Ten Hoor KA, Willemse PH, van der Zee AG. Apoptosis and apoptosis-associated parameters in relation to tamoxifen exposure in postmenopausal endometrium. Hum Pathol 2002;33;341-346.

86

HNPCC vs sporadic endometrial carcinoma |

13. Drachenberg CB, Ioffe OB, Papadimitriou JC. Progressive increase of apoptosis in prostatic intraepithelial neoplasia and carcinoma: comparison between in situ end-labeling of fragmented DNA and detection by routine hematoxylin-eosin staining. Arch Pathol Lab Med 1997;121;54-58.

14. Edmonston TB, Cuesta KH, Burkholder S et al. Colorectal carcinomas with high microsatellite instability: defining a distinct immunologic and molecular entity with respect to prognostic markers. Hum Pathol 2000;31;1506-1514.

15. Tibiletti MG, Furlan D, Taborelli M et al. Microsatellite instability in endometrial cancer: relation to histological subtypes. Gynecol Oncol 1999;73;247-252.

16. Duggan BD, Felix JC, Muderspach LI, Tourgeman D, Zheng J, Shibata D. Microsatellite instability in sporadic endometrial carcinoma. J Natl Cancer Inst 1994;86;1216-1221.

17. Basil JB, Goodfellow PJ, Rader JS, Mutch DG, Herzog TJ. Clinical significance of microsatellite instability in endometrial carcinoma. Cancer 2000;89;1758-1764.

18. Vassileva V, Millar A, Briollais L, Chapman W, Bapat B. Genes involved in DNA repair are mutational targets in endometrial cancers with microsatellite instability. Cancer Res 2002;62;4095-4099.

19. de Leeuw WJ, Dierssen J, Vasen HF et al. Prediction of a mismatch repair gene defect by microsatellite instability and immunohistochemical analysis in endometrial tumours from HNPCC patients. J Pathol 2000;192;328-335.

20. Ito K, Sasano H, Matsunaga G et al. Correlations between p21 expression and clinicopathological findings, p53 gene and protein alterations, and survival in patients with endometrial carcinoma. J Pathol 1997;183;318-324.

21. Lukas J, Groshen S, Saffari B et al. WAF1/Cip1 gene polymorphism and expression in carcinomas of the breast, ovary, and endometrium. Am J Pathol 1997;150;167-175.

22. Backe J, Gassel AM, Hauber K et al. p53 protein in endometrial cancer is related to proliferative activity and prognosis but not to expression of p21 protein. Int J Gynecol Pathol 1997;16;361-368.

23. Burton JL, Stewart RL, Heatley MK, Royds JA, Wells M. p53 expression, p21 expression and the apoptotic index in endometrioid endometrial adenocarcinoma. Histopathology 1999;35;221-229.

24. Taskin M, Lallas TA, Barber HR, Shevchuk MM. bcl-2 and p53 in endometrial adenocarcinoma. Mod Pathol 1997;10;728-734.

25. Henderson GS, Brown KA, Perkins SL, Abbott TM, Clayton F. bcl-2 is down-regulated in atypical endometrial hyperplasia and adenocarcinoma. Mod Pathol 1996;9;430-438.

26. Stewart RL, Royds JA, Burton JL, Heatley MK, Wells M. Direct sequencing of the p53 gene shows absence of mutations in endometrioid endometrial adenocarcinomas expressing p53 protein. Histopathology 1998;33;440-445.

87

Chapter 6 |

27. Caduff RF, Johnston CM, Svoboda-Newman SM, Poy EL, Merajver SD, Frank TS. Clinical and pathological significance of microsatellite instability in sporadic endometrial carcinoma. Am J Pathol 1996;148;1671-1678.

28. Berchuck A. Biomarkers in the endometrium. J Cell Biochem Suppl 1995;23;174-178.

29. Kounelis S, Kapranos N, Kouri E, Coppola D, Papadaki H, Jones MW. Immunohistochemical profile of endometrial adenocarcinoma: a study of 61 cases and review of the literature. Mod Pathol 2000;13;379-388.

30. Lax SF, Pizer ES, Ronnett BM, Kurman RJ. Clear cell carcinoma of the endometrium is characterized by a distinctive profile of p53, Ki-67, estrogen, and progesterone receptor expression. Hum Pathol 1998;29;551-8.

31. Yamauchi N, Sakamoto A, Uozaki H, Iihara K, Machinami R. Immunohistochemical analysis of endometrial adenocarcinoma for bcl-2 and p53 in relation to expression of sex steroid receptor and proliferative activity. Int J Gynecol Pathol 1996;15;202-208.

32. Kuwashima Y, Uehara T, Kishi K, Shiromizu K, Matsuzawa M, Takayama S. Proliferative and apoptotic status in endometrial adenocarcinoma. Int J Gynecol Pathol 1995;14;45-49.

33. Geisler JP, Geisler HE, Miller GA, Wiemann MC, Zhou Z, Crabtree W. MIB-1 in endometrial carcinoma: prognostic significance with 5-year follow-up. Gynecol Oncol 1999;75;432-436.

34. Michieli P, Chedid M, Lin D, Pierce JH, Mercer WE, Givol D. Induction of WAF1/CIP1 by a p53-independent pathway. Cancer Res 1994;54;3391-3395.

35. Myers MP, Pass I, Batty IH et al. The lipid phosphatase activity of PTEN is critical for its tumor supressor function. Proc Natl Acad Sci U S A 1998;95;13513-13518.

36. Kuismanen SA, Moisio AL, Schweizer P et al. Endometrial and colorectal tumors from patients with hereditary nonpolyposis colon cancer display different patterns of microsatellite instability. Am J Pathol 2002;160;1953-1958.

88

7

Gynecologic screening in hereditary

nonpolyposis colorectal cancer

F E M Rijcken1, M J E Mourits2, J H Kleibeuker1, H Hollema3,

A G J van der Zee2

Department of Gastroenterology1, Gynecology and Obstetrics2,

Pathology3, University Medical Center Groningen, The

Netherlands.

Gynecologic Oncology 2003;91:74-80

Chapter 7 |

Abstract

Objective - In hereditary nonpolyposis colorectal cancer (HNPCC), women with a

mismatch repair (MMR) gene mutation, have a 25-50% cumulative life time risk for

endometrial cancer and 8-12% for ovarian cancer. Therefore female members of

HNPCC-families are offered an annual gynecologic and transvaginal ultrasound (TVU)

examination and serum level CA 125 analysis. The aim of the present study was to

evaluate our 10-year experience with this screening program.

Patients and methods - Women who are MMR-gene mutation carriers or who fulfil the

Amsterdam criteria were identified from our HNPCC-database. Information concerning

the screening program was retrospectively collected from patient files.

Results - Forty-one, 35 premenopausal and 6 postmenopausal, women were enrolled in

the program with a median follow-up of 5 years (range 5 months-11 years). In 197

patient years at risk, 17 of 179 TVUs gave reason for endometrial sampling. Three

premalignant lesions, with complex atypical hyperplasia, were discovered. One interval

endometrial cancer was detected as a result of clinical symptoms. No abnormal CA 125

levels were measured and no ovarian cancers were detected.

Conclusion - These results demonstrate that gynecologic screening allows the detection

of premalignant lesions of the endometrium but also illustrate that recognition and

reporting of clinical symptoms by the women themselves is of utmost importance.

90

Gynecologic screening |

In hereditary nonpolyposis colorectal cancer (HNPCC) mutations in mismatch repair (MMR)

genes, i.e. MLH1, MSH2 and MSH6, lead to an increased risk of colorectal as well as extra-

colonic cancers, including endometrial and ovarian cancer.1-3 Effective screening procedures

are essential in the care of people belonging to HNPCC families. Jarvinen et al. have shown

that colonoscopies and subsequent polypectomies result in a significant survival advantage

and a reduction in the incidence of colorectal tumors.4 The cumulative life time risk for

endometrial cancer in HNPCC has been reported to vary between 25% and 50% compared to

3% in the general population.5 In women with HNPCC the cumulative life time risk for

endometrial cancer has been documented to be higher than the cumulative life time risk for

colorectal cancer.1 In comparison to the general population, female members of HNPCC

families also have an up to nine times higher lifetime risk to develop ovarian cancer (8-12%).6

Thus, implementing periodic gynecologic screening for female members of HNPCC families

to reduce the morbidity and mortality due to endometrial and ovarian cancer does seem a

rational approach. However, the effectiveness of gynecologic surveillance procedures has not

been shown in either prospective or retrospective studies. Among different centers, this has

led to a variability in recommendations on the starting age and the preferred methods of

screening.7-10 The aim of the present study was to evaluate our 10 years experience in

endometrial and ovarian cancer screening in women belonging to HNPCC-families and to

determine whether our present screening method achieves the aspired prevention or early

detection of gynecologic cancers.

Patients

The data from HNPCC and HNPCC-suspected families known at the University Hospital of

Groningen are prospectively registered in a database. HNPCC(-suspected) patients are

identified by gynecologists, gastroenterologists or through general physicians who refer

patients to the department of Medical Genetics because of familial cancer. In addition, several

ongoing studies at our hospital actively search for possible HNPCC affected subjects in young

or multiple primary cancer patients by using the database of the National Cancer Institute

which registrates all cancer patients in The Netherlands. For the present study, all female

members of HNPCC families were identified and their medical data were retrieved from the

database and the hospital archive files.

91

Chapter 7 |

In 1991, a program for the surveillance of female members of HNPCC-families was

introduced in our hospital. Women included in the program had a known MMR-gene

mutation and/or belonged to a family fulfilling the Amsterdam criteria II: (1) at least three

relatives with histologically verified colorectal, endometrial, small bowel or urothelial cell

cancer of the ureter or renal pelvis; one of them should be a first-degree relative of the other

two; (2) at least two successive generations should be affected; (3) in one relative the

HNPCC-related cancer should be diagnosed under 50 years of age; and (4) familial

adenomatous polyposis should be excluded.11 The recommended age of enrolment in the

screening program was between 30 and 35 years. If cancer in a relative had been diagnosed at

an earlier age, screening was started at a younger age.

Methods

Surveillance consisted of an annual gynecologic examination, a transvaginal ultrasound

(TVU) and measurement of serum levels of CA 125. Women were asked to report clinical

symptoms. Clinical symptoms such as irregular bleeding or postmenopausal blood loss were

always reason for endometrial sampling. Endometrial sampling was also indicated when the

double-layer endometrium thickness was greater than 12 mm in premenopausal women during

the second week of the menstruation cycle and greater than 5 mm in postmenopausal women

or if the endometrium was irregular or not well assessable by TVU.12 Endometrial samples

were obtained initially with a microcurettage (Pipelle) but if insufficient material was

obtained hysteroscopy and curettage followed. A hysteroscopy was performed at suspicion of

a polyp. For the present study, one pathologist (H.H.) verified the histology of each

endometrium sample. The normal value of serum CA 125 was ≤35 kU/l.

For each patient, data including the age at first screening, duration of the screening, clinical

symptoms, total number of TVUs and extra visits, number of endometrial samplings, verified

pathology reports on the acquired tissue and serum CA 125 levels were collected.

Results

The database for members of HNPCC families contained 103 women, of which 41 women

fulfilled the entrance criteria and who were enrolled in the screening program. Three of the 41

92

Gynecologic screening |

women had undergone colorectal surgery due to colorectal cancer prior to entering the

program. For various reasons 62 women were not enrolled in the gynecologic screening

program. Eighteen women had their surveillance in another hospital. At reevaluation of the

family history seven women were considered to have less than 25 percent risk for being a

gene carrier. Strictly speaking these seven women were members of an HNPCC family but

did not personally fulfill the Amsterdam criteria as their parents were not diagnosed with an

HNPCC related cancer. Four women were still too young (21, 24, 25 and 28 years old)

according to our criteria to start in the surveillance program. In their families no endometrial

cancer had been diagnosed below the age of 35 and thus there was no indication to start

screening earlier. Thirty-three women had undergone a hysterectomy and salpingo-

oophorectomy. Diagnoses in the hysterectomized women were endometrial cancer (n=15,

average age at diagnosis 44 years), ovarian cancer (n=5, average age at diagnosis 50 years),

synchronous ovarian and endometrial cancer (n=3, average age at diagnosis 49 years),

cervical cancer (n=1), complex atypical hyperplasia of the endometrium (n=1), and benign

conditions such as myoma and metrorraghia (n=8). Details regarding the 24 women with

cancer at the time of hysterectomy are described in table 1.

Eight women enrolled in the screening program had a known MLH1 mutation, two women

had a known MSH2 mutation and one woman was a carrier of a MSH6 mutation. Of the 41

participants, 35 were premenopausal and 6 postmenopausal. The women had a total of 179

screening appointments in 197 patient years. The median age at first screening was 37 years

(range: 27-60 years) and the median follow-up was 5 years (range: 5 months to 11 years).

Endometrium

Gynecologic history

At the annual screening, four women reported clinical symptoms, which led to endometrial

sampling even though three of the four TVUs showed no abnormalities. One woman had a

benign polyp, also seen by TVU, while in the other three samplings histological examination

revealed no abnormalities. Clinical and histological characteristics of patients who underwent

endometrial sampling are summarized in table 2.

93

Chapter 7 |

Gynecologic examination

No (pre)malignant abnormalities were detected through speculum or bimanual gynecologic

examination.

Table 1. Details regarding the 24 women with cancer at the time of hysterectomy.

Patient Tumor type Stage Grade Age Reasons leading to diagnosis***

1 Endometrial EA + CC* IA I 45 postmenopausal bloodloss 2 Endometrial EA IA I 45 irregulair vaginal bloodloss 3 Endometrial EA IA I 42 metrorraghia 4 Endometrial EA + CC IA III 45 hysterectomy for myoma 5 Endometrial EA IB I 46 irregulair vaginal bloodloss 6 Endometrial EA IB I 41 metrorraghia 7 Endometrial EA IB I 46 metrorraghia 8 Endometrial EA IB I 53 postmenopausal bloodloss 9 Endometrial EA IB I 37 hysterectomy for myoma 10 Endometrial EA IB - 28 - 11 Endometrial EA IB II 50 postmenopausal bloodloss 12 Endometrial EA IIA II 38 metro-menorraghia 13 Endometrial EA IIB II 50 metro-menorraghia 14 Endometrial EA IIIA III 50 postmenopausal bloodloss 15 Endometrial EA IIIC III 51 irregulair vaginal bloodloss 16 Ovarian SA** IA I 53 - 17 Ovarian EA IC II 40 obstipation 18 Ovarian EA IV II 43 nausea, stomach pains and weightloss 19 Ovarian SA IIC II 53 enlarged abdomen and fatigue 20 Ovarian SA - - 61 - 21 Endometrial EA IB II 50 postmenopausal bloodloss and fatigue Ovarian SA IA I 22 Endometrial EA IB III 46 menorraghia and fatigue Ovarian EA IC I 23 Endometrial EA + CC IIB II 51 weightloss and fatigue Ovarian EA IC II 24 Cervix IB I 49 postcoital bleeding

* EA, endometrioid adenocarcinoma; CC, clear cell component ** SA, serous adenocarcinoma *** In patients where symptoms are listed surgery was performed due to suspicion for malignancy. Some

tumors were diagnosed a long time ago and no details are available anymore, indicated by (-). No prophylactic hysterectomies were performed.

Transvaginal ultrasound

A total of 179 TVUs was performed. Seventeen TVUs (in 11 women) gave reason for

endometrial sampling by means of microcurettage or hysteroscopy and curettage. Three of the

94

Gynecologic screening |

17 endometrial samplings were performed within a year after the previous endometrial

samplings because the previous TVU and histology report necessitated reevaluation (table 2:

patient 2B) or the previous endometrial sampling did not render sufficient material for

adequate diagnosis (patient 3D-E). Of the 17 endometrial samplings, 14 revealed no severe

pathology.

Table 2. Results of the women who underwent endometrial sampling as a consequence of

screening

Patient

Pre/Post meno- pausal

Reason for endometrial sampling

Endo. Thickness (mm)*

Method of sampling

Pathology

1 Pre Polyp?† 4 Hysteroscopy‡ Macroscopical myoma 2 A~ Pre Irregular†† -** VABRA Complex atypical hyperplasia 2 B Pre Follow up - H&C‡‡ No abnormalities 2 C Pre Irregular 3 H&C No abnormalities 3 A Pre Clinical symptoms - Curettage‡‡‡ No abnormalities 3 B Pre Clinical symptoms - Curettage No abnormalities 3 C Post Irregular 6 Pipelle Complex atypical hyperplasia 3 D Post Irregular (extra)††† 3 H&C No abnormalities 3 E Post Irregular (extra) - H&C + biopsy No material 4 Pre Thickness†††† 27 Pipelle Complex atypical hyperplasia 5 Pre Clinical symptoms - H&C Benign polyp 6 A Pre Irregular - H&C Disturbed proliferative endometrium 6 B Pre Not well interpretable - H&C No abnormalities 7 Pre Clinical symptoms 5 H&C No abnormalities 8 Pre Thickness 16 *** Pipelle No abnormalities 9 Pre Thickness and

symptoms >10 H&C Disturbed proliferative endometrium

10 Pre Polyp? 6 Hysteroscopy No material 11 A Pre Polyp? 4 VABRA No abnormalities 11 B Pre Polyp? - H&C Benign polyp 12 Pre Thickness 11 Pipelle No abnormalities 13 Pre Irregular - Hysteroscopy Macroscopically normal

~ One patient underwent more than one endometrial sampling (marked by, A,B,C etc); † Polyp suspected by TV; †† Irregular endometrium seen by TVU; ††† Extra endometrial sampling because of results of prior sampling; ††††

Thickness of endometrium as measured by TVU; * Measurement at the end of the menstruation; ** Thickness not specified in archive; *** Transvaginal ultrasound (TVU) during second half of cycle; ‡ Only a hysteroscopy was done and no curettage as there were no macroscopical abnormalities; ‡‡ Hysteroscopy and curettage; ‡‡‡

Method of endometrial sampling was not specified.

95

Chapter 7 |

Three (18%) premalignant lesions were detected. In one premenopausal woman, the

endometrium measured 27mm by TVU. The pathology of the endometrial epithelium showed

complex atypical hyperplasia and she subsequently had a hysterectomy. In the remaining two

patients, one pre- and one postmenopausal, the transvaginal ultrasound showed a thin, but

focally irregular endometrium. In both cases, hysteroscopy and curettage were performed and

showed focal complex atypical hyperplasia. Both patients choose to have follow-up

hysteroscopy and curettage, which were carried out within two months. The obtained

histology at follow-up showed no abnormalities and both women resumed the surveillance

program.

One interval endometrial cancer was diagnosed outside of the regular screening program. A

postmenopausal woman presented 8 months after a normal transvaginal ultrasound

(endometrium thickness was ≤ 5 mm) with complaints of vaginal bleeding. She was

diagnosed at the age of 61 years with a stage IB grade II (T2N0M0) endometrioid

adenocarcinoma according to the FIGO stages. Five years after total hysterectomy and

bilateral salpingo-oophorectomy followed by radiotherapy, she is alive without signs of

recurrent disease.

Ovaries

Neither bimanual gynecologic examinations nor transvaginal ultrasounds showed

abnormalities of the ovaries. One hundred and fifteen serum CA 125 levels were determined.

CA 125 ranged from 1-24 kU/l with a median of 7 kU/l. No ovarian cancers were detected by

screening or outside of the annual screening.

Discussion

Annual gynecologic surveillance is generally recommended for women who are (suspected)

gene carriers of HNPCC. The theoretical benefit of such programs is early detection of

(pre)malignant lesions, thereby reducing morbidity and mortality from endometrial or ovarian

cancer. However, the value of annual gynecologic cancer surveillance remains unproven in

HNPCC women. The present study analyzed retrospectively, the results of ten-year screening

96

Gynecologic screening |

experience, using pelvic examination, transvaginal ultrasound and CA 125 detection, at our

hospital. Despite annual screening and good patient compliance no asymptomatic malignant

lesions were detected, but asymptomatic premalignant lesions were detected and could be

treated appropriately.

Though small, our research population appears to be representative for the group of women

who are generally enrolled in surveillance programs. The majority of the women was

premenopausal, their genetic status unknown but they all belonged to Amsterdam positive

families. Over the ten-year period and 197 patient years, the protocol was well implemented,

as the average women had a transvaginal ultrasound every 13 months and the great majority

of women reported no objections to screening. During the 10 years of our screening program

four women delivered one child and one woman had two children. These women kept to their

annual screening schedule. We were not confronted with a situation in which future fertility

could affect the choice of treatment.

In the present study, we have limited ourselves to women who fulfilled the Amsterdam

criteria or had a DNA mismatch repair gene mutation. Even though the Amsterdam criteria

have been revised they are possibly still too stringent to recognize all MMR gene mutation

carriers. Especially, unidentified MSH6 mutation carriers might be missed as their phenotype

is different from MLH1 and MSH2 mutation carriers, namely, later age of onset and no

proximal predominance for colorectal cancer. Most importantly in regard to gynecologic

screening, MSH6 carriers are more prone to endometrial cancers than MLH1 and MSH2

mutation carriers and should not be wrongly excluded from the screening program. Several

studies are on their way to realize new criteria for detecting high-risk subjects. At the present

time, all high-risk women not fulfilling the Amsterdam criteria II should be discussed in a

multidisciplinary group. Whether and how they should be screened should depend on the size

of their family, the number of affected family members and the age at which cancers were

diagnosed.

Although, during screening appointments, no malignant lesions were detected through history

and pelvic examination, both are an essential part in the early diagnosis of especially

endometrial cancers. The clinically manifested endometrial cancer outside of regular

97

Chapter 7 |

screening illustrates the importance of good patient instruction for early recognition of alarm

symptoms and the prompt notice (history) to a health professional.

Transvaginal ultrasound as screening method is problematic in premenopausal women due to

false positive findings. The endometrium thickness is dependent on the menstrual cycle,

therefore one well-defined cut-off value can not be applied. Data on the sensitivity, specificity

and predictive values of transvaginal ultrasound in asymptomatic premenopausal women are

scarce. In premenopausal women with abnormal uterine bleeding TVU has been concluded to

be an excellent initial diagnostic method by some but others have concluded that it is of

limited value in comparison to the usage in postmenopausal women.13-15 Using receiver-

operator-characteristic curves, Langer et al concluded, that ultrasound was better for detecting

hyperplasia than malignant abnormalities in postmenopausal women using unopposed-

estrogen but also in women using a placebo.16 In the present study, three benign neoplastic

lesions were detected but no malignant lesions. The one endometrial cancer, which was

detected between screening appointments, was probably missed by transvaginal ultrasound

but could be considered as an interval tumor.

Apart from transvaginal ultrasound, other methods have been suggested as screening devices

for endometrial cancer. Papanicolau smear has been shown to be inadequate as alternative.17

Karlsson and colleagues found endometrial cytology to be a less specific diagnostic tool than

transvaginal ultrasound.18 Jarvinen et al proposed endometrial biopsy every two to three years

as screening protocol for HNPCC women.19 However, endometrial biopsy is an invasive

procedure. Our study shows that only 27% of the women will undergo an invasive procedure

if transvaginal ultrasound is used as triage for biopsy. In their meta-analysis, Dijkhuizen et al.

concluded that endometrial biopsy with the Pipelle is superior to other endometrial sampling

techniques in the detection of endometrial carcinoma and atypical hyperplasia.15 Thus

concluding, the women with an abnormal TVU should, subsequently, be advised an

endometrial sampling by means of Pipelle or, in case of cervical stenosis, dilation and

curettage. If TVU is suspect for a polyp then hysteroscopy and curettage is preferred.

Similarly to our results, a retrospective endometrial cancer surveillance research of the St.

Mark's Hospital London and the Netherlands Foundation for the Detection of Hereditary

Tumors reported two endometrial cancers, which were detected as a result of symptoms and

98

Gynecologic screening |

not by surveillance. However, our results differ from the study of St. Mark's Hospital London,

as in their experience no premalignant lesions were detected. 20 Our histology was evaluated

using the current World Health Organization (WHO) classification of female genital tract

tumors.21 Complex hyperplasia has a low malignant potential while complex atypical

hyperplasia is the direct precancerous lesion of invasive well-differentiated endometrioid

adenocarcinoma.22 Several studies have shown that in curettage samples the diagnosis of

complex atypical hyperplasia does not appear to be highly reproducible among different

observers.23-24 Due to the inconsistency of the diagnosis and uncertainty in predicting the

natural history of individual lesions standardized clinical management is complicated (see our

three cases of complex atypical hyperplasia). Thus far, new classifications have been

proposed but not implemented which may render problems in the surveillance. Regular

follow-up is therefore of utmost important for anyone with a questionable histology.

Screening for any type of carcinoma is aimed primarily at the detection of early-stage disease

before symptoms occur and should result in a significantly improved overall survival. Osmers

et al. concluded that asymptomatic (postmenopausal) endometrial cancer patients detected by

transvaginal ultrasound are likely to have a better prognosis than those patients identified

symptomatically.25 However, others concluded that there was no prognostic advantage for

screened patients compared with symptomatic (postmenopausal) patients, who had vaginal

bleeding shorter than 8 weeks.26 The above mentioned studies all concern postmenopausal

women as similar large studies concerning premenopausal women have not been performed.

Vasen et al. described an 88% 5-year survival rate for endometrial cancer in pre- and

postmenopausal HNPCC-women.27 Whether this relatively high survival rate can be improved

as a result of screening remains unclear.

Similarly to endometrial cancer screening, it remains unclear whether screening for ovarian

cancer improves outcome for women in any risk group.28 The value of CA 125 measurement

as diagnostic instrument for ovarian cancer is controversial and the question should be raised

whether to continue the use of this tumor marker in the screening for ovarian cancer in

HNPCC patients. Vuento et al. concluded that a single CA 125 measurement provides no

advantage in the early detection of ovarian cancer in asymptomatic (postmenopausal) women

compared to transvaginal ultrasound.29 CA 125 with a cut-off value of 30 U/ml has good

sensitivity, but inadequate specificity for detecting preclinical disease. The vast majority of

99

Chapter 7 |

women with elevated CA 125 value have some reason other than an ovarian or endometrial

malignancy for this finding. Transvaginal ultrasound has been used as second-line test in

women with elevated CA 125 to increase specificity of CA125 to acceptable levels. As all

HNPCC women have an annual transvaginal ultrasound the CA 125 measurement appears to

be redundant in our screening program.

Given the limitations described above for endometrial as well as ovarian cancer screening,

HNPCC women who have completed their families or have reached menopause may choose

to undergo prophylactic total abdominal hysterectomy and bilateral salpingo-oophorectomy

(TAH-BSO). Prophylactic TAH-BSO is emerging as the most effective option for cancer risk

reduction. However, the life-time risk for both endometrial and ovarian cancer should be

optimally assessed before considering prophylactic TAH-BSO as this procedure may have

important consequences. Especially for premenopausal women, early onset of menopause,

loss of fertility and the need for hormone replacement therapy may result in physical and

psychological symptoms.30 The ovarian cancer risk in premenopausal HNPCC women

appears to be too low to justify an elective procedure with these possible side effects. TAH-

BSO should be discussed with HNPCC women at time of colon resection, thereby reducing

the physical morbidity of an extra operation. Further research should clarify whether TAH-

BSO is more beneficial and not a significantly larger burden than screening.

It remains controversial whether women belonging to HNPCC-families should have annual

gynecologic surveillance or whether only early recognition and rapid notification of alarm

symptoms (e.g. postmenopausal or irregular vaginal bleeding) will be sufficient to improve

overall survival. Our present study suggests that the potential gain of endometrial surveillance

in HNPCC by means of transvaginal ultrasound lies in the possibility of detecting

premalignant lesions and thereby possibly preventing malignancies to develop and avoiding

extensive treatment i.e. radiotherapy. Ideally, a controlled prospective trial should be

performed to clarify the effectiveness of screening and the method of screening. However, the

number of HNPCC-women, required to detect an association with sufficient power between

an increased survival advantage and surveillance, will be impossible to assemble. Therefore,

our conclusion for the present clinical practice is that annual gynecologic screening with

transvaginal ultrasound as triage for endometrial sampling remains justified for women

100

Gynecologic screening |

motivated for it. Regardless of that, patients should be well instructed for early recognition of

alarm symptoms and rapid notification should be strongly encouraged.

101

Chapter 7 |

References

1. Dunlop MG, Farrington SM, Carothers AD et al. Cancer risk associated with germline DNA mismatch repair gene mutations. Hum Mol Genet 1997, 6:105-10.

2. Aarnio M, Sankila R, Pukkala E et al. Cancer risk in mutation carriers of DNA-mismatch-repair genes. Int J Cancer 1999, 81:214-8.

3. Berends MJW, Wu Y, Sijmons RH et al. Molecular and clinical characteristics of MSH6 variants: an analysis of 25 index carriers of a germline variant. Am J Hum Genet 2002, 70:26-37.

4. Jarvinen HJ, Aarnio M, Mustonen H et al. Controlled 15-year trial on screening for colorectal cancer in families with hereditary nonpolyposis colorectal cancer. Gastroenterology 2000, 118:829-34.

5. Vasen HF, Stormorken A, Menko FH et al. MSH2 mutation carriers are at higher risk of cancer than MLH1 mutation carriers: a study of hereditary nonpolyposis colorectal cancer families. J Clin Oncol 2001, 19:4074-80.

6. Brown GJ, St John DJ, Macrae FA et al. Cancer risk in young women at risk of hereditary nonpolyposis colorectal cancer: implications for gynecologic surveillance. Gynecol Oncol 2001, 80:346-9.

7. Vasen HF, Taal BG, Griffioen G et al. Clinical heterogeneity of familial colorectal cancer and its influence on screening protocols. Gut 1994, 35:1262-6.

8. Jarvinen H, Aarnio M. Surveillance on mutation carriers of DNA mismatch repair genes. Annales Chirurgiae et Gynaecologiae 2000, 89:207-10.

9. Lynch HT, Lynch J. Lynch syndrome: genetics, natural history, genetic counseling, and prevention. J Clin Oncol 2000, 18:19S-31S.

10. Hodgson SV, Bishop DT, Dunlop MG, Evans DG, Northover JM. Suggested screening guidelines for familial colorectal cancer. J Med Screen 1995, 2:45-51.

11. Vasen HF, Mecklin JP, Khan PM, Lynch HT. The International Collaborative Group on Hereditary Non-Polyposis Colorectal Cancer (ICG-HNPCC). Dis Colon Rectum 1991, 34:424-5.

12. Blumenfeld ML, Turner LP. Role of transvaginal sonography in the evaluation of endometrial hyperplasia and cancer. Clin Obstet Gynecol 1996, 39:641-55.

13. Mortakis AE, Mavrelos K. Transvaginal ultrasonography and hysteroscopy in the diagnosis of endometrial abnormalities. J Am Assoc Gynecol Laparosc 1997, 4:449-52.

14. Bakos O, Heimer G. Transvaginal ultrasonographic evaluation of the endometrium related to the histological findings in pre- and perimenopausal women. Gynecol Obstet Invest 1998, 45:199-204.

102

Gynecologic screening |

15. Dijkhuizen FP, Mol BW, Brolmann HA, Heintz AP. The accuracy of endometrial sampling in the diagnosis of patients with endometrial carcinoma and hyperplasia: a meta-analysis. Cancer 2000, 89:1765-72.

16. Langer RD, Pierce JJ, O'Hanlan KA et al. Transvaginal ultrasonography compared with endometrial biopsy for the detection of endometrial disease. Postmenopausal Estrogen/Progestin Interventions Trial. N Engl J Med 1997, 337:1792-8.

17. Gomez-Fernandez CR, Ganjei-Azar P, Behshid K, Averette HE, Nadji M. Normal endometrial cells in Papanicolaou smears: prevalence in women with and without endometrial disease. Obstet Gynecol 2000, 96:874-8.

18. Karlsson B, Granberg S, Wikland M, Ryd W, Norstrom A. Endovaginal scanning of the endometrium compared to cytology and histology in women with postmenopausal bleeding. Gynecol Oncol 1993, 50:173-8.

19. Jarvinen HJ, Aarnio M. Surveillance on mutation carriers of DNA mismatch repair genes. Ann Chir Gynaecol 2000, 89:207-10.

20. Dove-Edwin I, Boks D, Goff S et al. The outcome of endometrial carcinoma surveillance by ultrasound scan in women at risk of hereditary nonpolyposis colorectal carcinoma and familial colorectal carcinoma. Cancer 2002, 94:1708-12.

21. Scully RE, Bonfiglio TA, Kurman RJ, Silverberg SG, Wilkinson K. International histological classification and typing of female genital tract tumors. New York: Springer-Verlag 1994.

22. Widra Ea, Dunton CJ, McHugh M, Palazzo JP. Endometrial hyperplasia and the risk of carcinoma. Int J gynecol Cancer 1995, 5:233-5.

23. Bergeron C, Nogales FF, Masseroli M et al. A multicentric European study testing the reproducibility of the WHO classification of endometrial hyperplasia with a proposal of a simplified working classification for biopsy and curettage specimens. Am J Surg Pathol 1999, 23:1102-8.

24. Kendall BS, Ronnett BM, Isacson C et al. Reproducibility of the diagnosis of endometrial hyperplasia, atypical hyperplasia, and well-differentiated carcinoma. Am J Surg Pathol 1998, 22:1012-9.

25. Osmers RG, Osmers M, Kuhn W. Prognostic value of transvaginal sonography in asymptomatic endometrial cancers. Ultrasound Obstet Gynecol 1995, 6:103-7.

26. Gerber B, Krause A, Muller H et al. Ultrasonographic detection of asymptomatic endometrial cancer in postmenopausal patients offers no prognostic advantage over symptomatic disease discovered by uterine bleeding. Eur J Cancer 2001, 37:64-71.

27. Vasen HF, Watson P, Mecklin JP et al. The epidemiology of endometrial cancer in hereditary nonpolyposis colorectal cancer. Anticancer Res 1994, 14:1675-8.

28. Jacobs IJ, Skates SJ, MacDonald N et al. Screening for ovarian cancer: a pilot randomised controlled trial. Lancet 1999, 353:1207-10.

103

Chapter 7 |

29. Vuento MH, Stenman UH, Pirhonen JP, Makinen JI, Laippala PJ, Salmi TA. Significance of a single CA 125 assay combined with ultrasound in the early detection of ovarian and endometrial cancer. Gynecol Oncol 1997, 64:141-6.

30. Fry A, Busby-Earle C, Rush R, Cull A. Prophylactic oophorectomy versus screening: psychosocial outcomes in women at increased risk of ovarian cancer. Psychooncology 2001, 10:231-241.

104

8

Sulindac treatment in hereditary

nonpolyposis colorectal cancer

F E M Rijcken1, H Hollema 2, AGJ van der Zee3, T van der

Sluis2, W Boersma-van Ek1, JH Kleibeuker1

Departments of 1Gastroenterology, 2Pathology and 3Gynaecology

and Obstetrics. University Medical Center Groningen, The

Netherlands

Submitted

Chapter 8 |

Abstract

Background - Nonsteroidal anti-inflammatory drugs (NSAIDs), e.g. sulindac have

chemopreventive effects on colorectal cancer and adenomas in familial adenomatous

polyposis and the general population. The effects of sulindac on colonic mucosa in

hereditary nonpolyposis colorectal cancer (HNPCC) predisposed subjects have not been

studied before.

Methods and results - We evaluated these effects in HNPCC using surrogate end-points

for cancer risk including epithelial cell proliferative activity, apoptosis and expression of

proliferation-, apoptosis- and cell cycle-involved genes. In a randomized double-blind

cross-over study 22 subjects (9 female; age 30-66 years, mean 44) with a germline

mutation in MLH1 (n=5) or MSH2 (n=8) and/or an Amsterdam criteria positive family

history and an adenoma before age 40, an advanced adenoma before age 50 or an

HNPCC-related cancer in the past were included. Sulindac 150 mg b.i.d. and placebo

were given for four weeks each, with four weeks in between. Colonoscopy was

performed at the end of both periods and biopsies were taken from ascending,

transverse, and sigmoid colon and rectum. Proliferative activity was determined by Ki-

67 staining, apoptosis by staining of cytokeratin 18 cleavage products. Expression of

cyclins B1, D3 and E and p21, p27, bax, bcl2 and cox-2 was studied

immunohistochemically. Proliferative activity was higher during sulindac than placebo

in ascending and transverse colon, but not in sigmoid and rectum. Apoptosis was not

affected. Besides an increase in cyclin D3 no differences were found in expression of

regulating proteins in the proximal colon.

Conclusion - sulindac induces an increase of epithelial cell proliferative activity in the

proximal colon of subjects with HNPCC. Since colorectal cancer predominantly arises in

the proximal colon in HNPCC, these results cast doubts on the potential

chemopreventive effects of sulindac in HNPCC.

106

Sulindac treatment in HNPCC |

Hereditary nonpolyposis colorectal cancer (HNPCC) is an autosomal dominantly inheriting

disorder, predisposing affected individuals to the development of cancer, in particular

colorectal cancer. It is germline mutations in DNA mismatch repair (MMR) genes,

predominantly MLH1, MSH2 and MSH6 that predispose an individual to HNPCC. Mutation

carriers have an up tot 80% lifetime risk to develop colorectal cancer, most prominently in the

proximal colon.1 The cancers arise at a relatively early age and often develop faster than

sporadic colorectal cancers. The carcinogenic process in HNPCC is characterized by the

adenoma-to-carcinoma sequence similar to that described in familial adenomatous polyposis

(FAP) or sporadic cancer, but several of the clinical manifestations, as well as the molecular

pathobiology underlying them, are distinctive.2 Preventive measures, e.g. chemoprevention,

for risk reduction would be welcome aspects in the treatment of patients at high inherited risk

for cancer. Broadly studied in patients with FAP, the efficacy of chemoprevention remains

largely unexplored in HNPCC.

Increasing evidence from cell line, animal and human studies reveals that the administration

of nonsteroidal anti-inflammatory drugs (NSAIDs) represents a viable option for the

chemoprevention of FAP-associated and sporadic colorectal cancer.3-6 The use of NSAIDs is

associated with a lower risk of colorectal adenoma and cancer development, a lower risk of

recurrent colorectal adenomas and carcinomas and these effects are independent of gender,

age and site in the colorectum.6-10

The molecular basis for the chemoprotective action of NSAIDs has not yet been fully

elucidated. The NSAID sulindac has been reported to display profound antiproliferative

effects, to alter the cell cycle distribution, and to induce apoptosis in cell lines and in vivo.11-13

In particular, cyclooxygenase-2 (cox-2), which is inhibited by NSAIDs, is an important target

of these drugs 14, consistent with the fact that colorectal neoplasms often exhibit up-regulation

of this enzyme. Also, other mechanisms may be involved in the chemopreventive effects of

NSAIDs. Sulindac sulfide, the active metabolite of sulindac, has been shown to decrease the

expression of cyclin B1 and E and increase the expression of cyclin D3.15 Several studies

demonstrated markedly induced expression of p21, a cell cycle inhibitor, and bax, a pro-

apoptosis protein, after sulindac therapy but others failed to find these changes.16-18 Also, p27

and bcl-2 have been associated with the possible mechanism of colorectal cancer prevention

by NSAIDs.

107

Chapter 8 |

In order to explore the potential role of sulindac in HNPCC chemoprevention we evaluated

the effects of sulindac in HNPCC patients using surrogate end-points for cancer risk including

epithelial cell proliferation rate, apoptosis and expression in normal colonic epithelium of

proliferation-, apoptosis- and cell cycle-involved genes.

Material and methods

The present study was a randomized, double blind, placebo-controlled cross-over study in

ascertained MMR gene mutation carriers and subjects with more than 50% risk to be MMR

gene mutation carriers. Participants were recruited from a cohort of families with an

established pathogenic mutation in MLH1, MSH2 or MSH6 and/or fulfilling the Amsterdam

criteria 19-20, and who are under regular colonoscopic surveillance at the Department of

Gastroenterology of the University Medical Center Groningen. Subjects from families that

meet the Amsterdam criteria were considered to have more than 50% risk to be a mutation

carrier if they had previously been diagnosed with an HNPCC-related cancer, or with a high

risk colorectal adenoma (larger than 1 cm and/ or villous component and/or high grade

dysplasia) before the age of 50, or with any adenoma before the age of 40 years. Written

informed consent was obtained from each participant. All participants were older than 18

years of age and able to consume a normal diet. Prior colorectal surgery was allowed, but the

subject had to have > 50% of the colorectum left. Subjects with any disease of the colon apart

from adenomatous polyps and diverticulosis were excluded. Further exclusion criteria

consisted of a history of hypersensitivity to NSAIDs, a history of peptic ulcer disease, the

inability to abstain from use of other NSAIDs, and the use of oral adrenocorticosteroids and

cholestyramine during the study period. The study protocol was approved by the Medical

Ethical Committee of the University Medical Center Groningen, and all subjects gave written

informed consent.

After randomization to one of both sequences of treatment subjects were given sulindac 150

mg b.i.d. for four weeks with a cross-over to twice daily placebo for four weeks, or vice versa.

The first treatment period was followed by a washout period of four weeks. During treatment

periods the physician regularly contacted the subject to monitor possible side-effects and to

increase compliance. At the end of the first four weeks treatment period a regularly planned

108

Sulindac treatment in HNPCC |

surveillance colonoscopy was performed after lavage with PEG-containing solution.

Colonoscopy was repeated at the end of the second four weeks of treatment.

During each colonoscopy biopsies were taken from four predefined sites in the colorectum,

namely from (1) the ascending colon, (2) halfway the transverse colon, (3) the sigmoid and (4)

the rectum. Three biopsies of each location were fixed in buffered formalin (pH 7.47) and

embedded in paraffin. Subsequently, consecutively numbered sections of 3 µm were cut from

the biopsies and fixed onto 3-aminopropyl-triethoxysilane (APES, Sigma-Aldrich,

Diesenhofen, Germany) coated slides, stretched for 30 minutes at 60°C and dried overnight at

37°C. Paired samples of each participant were stained in the same batch.

Proliferation

Proliferation was assessed by immunohistochemical staining of Ki67 using MIB1 antibody.

Antigen retrieval was performed using the high-pressure cooker. Immersed in 200µl blocking

reagent (2% block and 0.2% SDS in maleic acid, pH 6.0 (Boehringer Mannheim, Germany))

the section underwent 3 sessions of 5 minutes at 115°C in a high pressure cooker alternating

with incubation in a humid environment. The endogenous peroxidase activity was quenched

by incubation with 0,3% H2O2 in PBS for 30 minutes. The sections were immersed for one

hour with the MIB1 antibody in PBS with 1% bovine serum albumin (BSA), at a dilution of

1:400. Subsequently, the sections were consecutively incubated for 30 minute periods with

rabbit antimouse peroxidase (RAMPO; DAKO, Glostrop, Denmark) and goat antirabbit

peroxidase (GARPO; DAKO) both diluted (1:50) in PBS-1% BSA. The sections were

submerged for 10 minutes in a solution of 25 mg 3,3’-diaminobenzidine (DAB) in PBS and

50 mg of imidazol with 50 µl 30% H2O2. After rinsing with demi water, the sections were

counterstained with haematoxylin, washed with running water, dehydrated with graded

alcohol, dried and covered with a slide.

Apoptosis

Apoptosis was assessed by immunohistochemical staining of cleavage products of cytokeratin

18 (CK18) using Mab M30. CK18 cleavage by activated caspase-3 is an early marker of

apoptosis.21 Antigen retrieval was performed by submersion of the deparaffinised, rehydrated

sections in preheated 10mM citrate buffer (pH 6.0) and heated for 8 minutes at 700 watts in a

109

Chapter 8 |

microwave. After cooling at room temperature for 15 minutes the sections were thoroughly

rinsed with PBS for 5 minutes and a 1:50 solution of Mab M30 was applied for one hour at

room temperature. Subsequently, the same steps as for Ki-67 staining were performed,

starting with the incubation of the sections with RAMPO.

Proliferation- and apoptosis associated proteins

The optimal antibody-antigen reaction was determined for proliferation-regulating proteins

cyclin B1, D3 and E, CDK inhibitors p21 and p27, and apoptosis regulators bcl-2, bax, and

cox-2. Table 1 summarises the primary antibodies that were used in immunohistochemical

studies, the companies from which they were purchased, the dilution at which they were used

and the corresponding technique for antigen retrieval. The high-pressure cooker antigen

retrieval method is as described above for Ki-67 immunostaining and the microwave method

as described for Mab M30 using either 10 mM citrate buffer (pH 6.0) or EDTA.

Table 1. Primary antibodies used in immunohistochemical studies

Protein Antigen retrieval method

Clone Company Dilution

Ki-67 High-pressure cooker MIB1 Immunotech, Marseille, France 1:400 Cyclin B1 Microwave EDTA 7A9 Novocastra, Newcastle, UK 1:50 Cyclin D3 Microwave EDTA DCS-22 Novocastra, Newcastle, UK 1:10 Cyclin E Microwave EDTA 13A3 Novocastra, Newcastle, UK 1:10 P21 High-pressure cooker WAF1(Ab-1) Oncogene, Darmstadt, Germany 1:50 P27 High-pressure cooker 1B4 Novocastra, Newcastle, UK 1:50 cCK 18* Microwave citrate MAb M30 Boehringer, Mannheim, Germany 1:50 Bax Microwave citrate B-9 Santa Cruz Biotechnology, Santa Cruz, Ca 1:200 Bcl-2 High-pressure cooker MAb 124 DAKO, Glostrup, Denmark 1:50 Cox-2 Microwave EDTA 33 Transduction Laboratories, Lexington, KY 1:50

* cCK 18: cleaved cytokeratin 18

Evaluation of staining

Without knowledge of treatment, three authors (FR, TvdS and WBvE) scored all stained

slides. Each set of stains was entirely scored by one author. A second author randomly chose

20 slides to confirm reproducibility of the scoring of each immunostaining. The

110

Sulindac treatment in HNPCC |

immunoreactivity for every antibody in each biopsy was analysed and quantified in at least

five entire crypts. If immunohistochemical staining was inadequate or failed to produce

sufficient quantifiable crypts consecutive slides were stained with the inclusion of a

representative slide from the prior staining session.

Proliferative activity was scored as MIB1 labelling index, where positively stained nuclei

were divided by the total number of counted nuclei x 100 (%). To verify the effect of sulindac

on compartmentalization of proliferation the MIB1 labelling index was quantified of the upper

one third of the crypt in a similar manner as the entire crypt. For apoptosis the number of

apoptotic cells was too low to investigate compartmentalization.

Cyclins B1, D3 and E, p21, p27 and cox-2 were also scored as labelling index. The intensity

of bax and bcl-2 expression was scored as (1) absent, (2) weak, (3) moderate, of (4) strong.

Two slides from the first staining sessions were used in every consecutive staining session as

reference for the intensity scoring. The intensity of bax demonstrated a gradient from the

bottom to the surface of the crypt. The crypt was therefore divided into three compartments. A

weighed score, intensity multiplied by the percentage of cells staining positively, was

calculated for each compartment.

Statistics

Statistical analyses were performed using SPSS 12.0.1 for Windows software (Statistical

Package for the Social Sciences Inc, Chigago, IL). For statistical assessment of changes in

proliferative, apoptotic, cyclins B1, D3 and E, p21, p27 and cox-2 labelling indices following

sulindac treatment, the paired-samples T-test with 95% confidence interval was used. To

evaluate the difference in bax and bcl-2 expression Wilcoxon’s nonparametric two-related-

sample test was used. To determine differences in various colonic regions in HNPCC patients,

the nonparametric Mann-Whitney U test was conducted. Reported p-values are two-tailed and

significance was assumed if p < 0.05.

Results

Research population

In total 22 subjects, 9 women and 13 men were initially included in the study. Thirteen were

proven mutation carriers (5 MLH1 and 8 MSH2). The remaining nine were at high risk of

being a carrier, as defined above. Mutation analysis in this group of patients had not been

111

Chapter 8 |

performed, was inconclusive, or was still in process. Six patients had previously been

diagnosed with an HNPCC-related cancer (ovarian, rectal, endometrial and jejunal cancer).

The average age of the subjects at inclusion was 44 years (range 30-66 years). Four subjects

did not undergo a second colonoscopy. Two subjects, one male (MSH2 mutation carrier) and

one female (mutation status unknown) left the study due to side effects. These patients

complained of gastrointestinal discomfort, stomach cramps and diarrhea. The other two

subjects (two females, an MLH1 and an MSH2 mutation carrier) were not motivated for a

second colonoscopy. So, the analysis of data is based on the remaining 18 subjects.

Table 2. Immunoreactivity of the proteins in the entire colon: after treatment with

placebo and with sulindac.

Protein Placebo Sulindac p-value

Ki67 (%) 42.6 ± 1.3 46.3 ± 1.5 0.004

Ki67 top (%) 1.5 ± 0.2 2.5 ± 0.4 0.011

Cyclin B1 (%) 6.6 ± 0.2 7.0 ± 0.3 0.165

Cyclin D3 (%) 35.1 ± 1.1 36.7 ± 1.0 0.279

Cyclin E (%) 3.6 ± 0.2 3.3 ± 0.2 0.176

P21 (%) 38.1 ± 1.0 39.5 ± 0.9 0.323

P27 (%) 74.3 ± 1.4 75.5 ± 1.4 0.519

cCK18 * (%) 0.68 ± 0.10 0.70 ± 0.08 0.906

cCK18:Ki67 1.7 ± 0.2 1.5 ± 0.2 0.660

Bax (I) 2.0 2.0 0.183

Bcl2 (%) 70.4 ± 2.7 74.5 ± 2.2 0.285

Cox2 (%) 2.7 ± 0.3 2.8 ± 0.2 0.666

* cCK18: cleaved cytokeratin 18

Proliferative and apoptotic indices

On average 7 full crypts (mean of 1070 cells) were evaluated per biopsy. After placebo, the

mean MIB1 labelling index in the biopsies of the proximal colon was significantly higher than

in the biopsies of the distal colon of the same subjects (48.4±1.3% (±SEM) vs 35.4±1.3%,

respectively, p<0.001). The MIB1 labelling index in the entire colon was higher after sulindac

treatment than after placebo (46.3±1.5% vs. 42.6±1.3%, p=0.004, table 2). This increase in

112

Sulindac treatment in HNPCC |

MIB1 labelling index was due to a significant increase in the proximal i.e., ascending and

transverse, colon, which was not evident in the sigmoid and rectum (figure 1). The MIB1

labelling index in the upper crypt compartment in the proximal colon was significantly higher

after sulindac than after placebo (4.2±0.7% vs 2.5±0.2%, p=0.021), whereas this was not the

case in the sigmoid and rectum. Crypt length did not alter after sulindac treatment, neither in

the proximal colon nor in the distal colon. Apoptotic activity was similar in each of the four

locations in the colon both during sulindac and placebo use, and was not measurably affected

by sulindac.

Proliferation after placebo and sulindac treatment

0

10

20

30

40

50

60

70

80

90

100

asc trans sigm rect

MIB

1 la

belli

ng in

dex

(%)

* †

‡

§

Figure 1. Epithelial cell proliferative activity in the colorectum after placebo and

sulindac treatment. Asc = ascending colon; trans = transverse colon; sigm = sigmoid colon; rect: rectum after placebo treatment , after sulindac treatment. * labelling index after placebo versus sulindac in the ascending colon, p=0.048 † labelling index after placebo versus sulindac in the transverse colon, p=0.021 ‡ labelling index after placebo in the proximal versus distal colon, p<0.001 § labelling index after sulindac in the proximal versus distal colon, p=0.004

113

Chapter 8 |

Proliferation- and apoptosis associated proteins

In the ‘untreated’ proximal colon, significantly more epithelial cells expressed cyclin B1 and

less expressed cyclin D3 in comparison to those in the distal colon. Cox-2 expression was

significantly higher in the proximal colon in comparison to the distal colon. No significant

changes in expression of the regulating proteins were observed after sulindac treatment. A

tendency towards more cyclin D3 expression was observed in the proximal colon after

sulindac treatment (35.6±1.6% vs 31.4±1.4%, p=0.055). In the distal colon less cells

expressed bax after sulindac (2.1±0.1% vs. 2.4±0.1%, p=0.032).

Discussion

Experimental and human studies have clearly demonstrated a chemopreventive effect of

NSAIDs, including sulindac, on colorectal carcinogenesis. So far, however, mechanisms by

which this is mediated have not been fully elucidated. Hereditary nonpolyposis colorectal

cancer with its high incidence of colorectal cancer at a young age represents an ideal target for

chemoprevention, however, the effects of sulindac on colorectal epithelium of subjects with

HNPCC have not been studied before. In the present study, the effects of sulindac on normal-

appearing mucosa from HNPCC patients in four regions of the colon were evaluated using

biomarkers for proliferation and apoptosis as endpoints. To identify possible mechanisms of

action of chemoprevention with sulindac proliferation and apoptosis regulating proteins were

identified from previous studies and were evaluated using immunohistochemistry. Sulindac

proved to induce an increase of epithelial cell proliferative activity in the proximal colon,

without affecting this activity in the sigmoid and rectum, and without affecting apoptosis in

any of the regions of the colorectum. Although the clinical value of the biomarkers used is

disputable, the results cast doubts on the chemopreventive effects of sulindac, and NSAIDs in

general, in HNPCC.

As carcinogenesis is associated with a loss of tissue homeostasis by disturbance of both cell

proliferation and apoptotic cell death, both are considered potential biomarkers for activity of

candidate chemopreventive agents. An increase of colorectal epithelial cell proliferative

activity has been considered for a long time now as a marker for an increased susceptibility of

colorectal cancer. Biologically it is thought to increase the risk of spontaneous and

exogenously induced mutations, thus rendering some cells neoplastic properties. Reduction of

114

Sulindac treatment in HNPCC |

proliferative activity by candidate chemopreventive agents has been considered for many

years as an important indication for the potentially beneficial effect of such agents and as a

probable mechanistic explanation for that effect. It is therefore that also the effect of NSAIDs

on cell proliferative activity has been studied extensively. In cell culture studies, sulindac and

sulindac sulphide indeed inhibited proliferation by inducing cell quiescence, with reduction of

the expression of classical biomarkers of proliferation, such as Ki-67, independent of their

ability to inhibit prostaglandin synthesis.3,13,15 However, these results were not consistently

reproducible in in vivo studies and conflicting data concerning effects of NSAIDs on the

colorectal epithelial proliferation have been reported.11,21-25 A majority of the studies

performed on biopsies from the rectosigmoid of patients with FAP failed to demonstrate an

anti-proliferative effect of sulindac in normal colorectal mucosa .11,23,24

A limitation of all previous studies is that only the effects on the epithelium of rectum and

sigmoid were studied. In the present study no effect of sulindac on proliferative activity in

rectum and sigmoid was observed either. However, proliferation in both transverse and

ascending colon increased. The facts that proliferation increased in both regions and that this

was not only apparent in the whole crypts but also in the upper third of the crypts strongly

indicate that these results are not due to chance but represent a true biological effect of

sulindac. Beside a tendency to an increased cyclin D3 expression in the proximal colon upon

sulindac treatment, the present study can not explain the side specific effect of sulindac on

proliferation in the proximal colon of HNPCC patients. Previously it has been suggested that

increased proliferation in sulindac-treated mice could be a compensatory phenomenon

occurring secondary to loss of crypt epithelial cells by apoptosis induced by sulindac.26

Possibly this compensatory effect is more pronounced in the proximal than distal colon in

HNPCC due to the significantly higher proliferative activity in the untreated proximal colon.

It is unknown whether similar effects occur in the proximal colon of subjects without a

predisposition for HNPCC, but the finding of it in an HNPCC population has implications by

its own. The preference for the development of cancers in the proximal colon in HNPCC

makes one reluctant to use agents that affect proliferation in that part of the colon in a way

that is generally considered to mark an increased cancer risk.

Currently, the chemopreventive effects of NSAIDs in the colorectum are thought to be largely

due to induction of apoptosis. This view is supported by the observation of apoptosis

induction in cell lines, including mismatch repair deficient cells, in intestinal epithelium of

115

Chapter 8 |

ApcMin mice and in human colorectal epithelium.3,13,24,27 The apoptotic effect is partially

mediated by cox-2 inhibition but also occurs in cells lacking cox-2.3 A variety of targets of

chemoprevention have been suggested as well as preconditions for NSAID-induced apoptosis.

Bax mutations may be involved in sulindac resistance while Apc1638N mice with inactivation

of p21 did not respond to sulindac .17,18 In the present study, despite expression of bax, p21

and cox-2 in all biopsies, apoptosis did not alter upon sulindac treatment, neither in the

proximal nor in the distal colon of HNPCC patients. Expression of bax, cox-2 as well as

expression of p21 did not alter after treatment with sulindac. Previously, sulindac treatment

demonstrated no effect on rectal epithelial apoptosis in young FAP patients without adenomas 28, whereas in symptomatic FAP patients, regression of adenomas was accompanied by

alteration of the rectal epithelial apoptotic ratio with relative increase in apoptosis in surface

cells compared with the deeper crypt. Our results, in normal appearing mucosa in HNPCC

patients, are in line with the data of phenotypically unaffected FAP patients.27

In summary, in a group of HNPCC-subjects sulindac induced an increase of epithelial cell

proliferative activity, both in the whole crypt and in the upper crypt compartment, in the

proximal colon without affecting proliferation in the distal colon and without affecting

apoptosis anywhere in the colorectum. These results cast doubt on the chemopreventive

potency of sulindac, and NSAIDs in general, in HNPCC.

116

Sulindac treatment in HNPCC |

References

1. Lynch HT, de la Chapelle CA. Hereditary colorectal cancer. N Engl J Med. 2003;348:919-32.

2. Kinzler KW, Vogelstein B. Lessons from hereditary colorectal cancer. Cell. 1996;87:159-70.

3. Hanif R, Pittas A, Feng Y et al. Effects of nonsteroidal anti-inflammatory drugs on proliferation and on induction of apoptosis in colon cancer cells by a prostaglandin-independent pathway. Biochem Pharmacol. 1996;52:237-45.

4. Brown WA, Skinner SA, Malcotenti-Wilson C, Vogiagis D, O'Brien PE. Non-steroidal anti-inflammatory drugs with activity against either cyclooxygenase 1 or cyclooxygenase 2 inhibit colorectal cancer in a DMH rodent model by inducing apoptosis and inhibiting cell proliferation. Gut. 2001;48: 660-6.

5. Giardiello FM, Hamilton SR, Krush AJ et al. Treatment of colonic and rectal adenomas with sulindac in familial adenomatous polyposis. N Engl J Med. 1993;328:1313-6.

6. Baron JA, Cole BF, Sandler RS et al. A randomized trial of aspirin to prevent colorectal adenomas. N Engl J Med. 2003;348: 891-9.

7. Janne PA, Mayer RJ. Chemoprevention of colorectal cancer. N Engl J Med. 2000;342:1960-8.

8. Thun MJ, Henley SJ, Patrono C. Nonsteroidal anti-inflammatory drugs as anticancer agents: mechanistic, pharmacologic, and clinical issues. J Natl Cancer Inst. 2002;94:252-266.

9. Cruz-Correa M, Hylind LM, Romans KE, Booker SV, Giardiello FM. Long-term treatment with sulindac in familial adenomatous polyposis: a prospective cohort study. Gastroenterology. 2002;122: 641-5.

10. Giovannucci E, Egan KM, Hunter DJ et al. Aspirin and the risk of colorectal cancer in women. N Engl J Med. 1995;333:609-14.

11. Labayle D, Fischer D, Vielh, P et al. Sulindac causes regression of rectal polyps in familial adenomatous polyposis. Gastroenterology. 1991;101:635-9.

12. Shiff SJ, Qiao L, Tsai LL, Rigas B. Sulindac sulfide, an aspirin-like compound, inhibits proliferation, causes cell cycle quiescence, and induces apoptosis in HT-29 colon adenocarcinoma cells. J Clin Invest. 1995;96:491-503.

13. Shiff SJ, Koutsos MI, Qiao L, Rigas B. Nonsteroidal antiinflammatory drugs inhibit the proliferation of colon adenocarcinoma cells: effects on cell cycle and apoptosis. Exp Cell Res. 1996;222:179-88.

117

Chapter 8 |

14. Dannenberg AJ, Altorki NK, Boyle JO et al. Cyclo-oxygenase 2: a pharmacological target for the prevention of cancer. Lancet Oncol. 2001;2:544-551.

15. Qiao L, Shiff SJ, Rigas B. Sulindac sulfide alters the expression of cyclin proteins in HT-29 colon adenocarcinoma cells. Int J Cancer. 1998;76:99-104.

16. Poole JC, Thain A, Perkins ND, Roninson IB. Induction of transcription by p21Waf1/Cip1/Sdi1: role of NFkappaB and effect of non-steroidal anti-inflammatory drugs. Cell Cycle. 2004;3:931-40.

17. Zhang L, Yu J, Park BH, Kinzler KW, Vogelstein B. Role of BAX in the apoptotic response to anticancer agents. Science. 2000;290:989-92.

18. Yang W, Velcich A, Mariadason J et al. p21(WAF1/cip1) is an important determinant of intestinal cell response to sulindac in vitro and in vivo. Cancer Res. 2001;61:6297-302.

19. Vasen Vasen HF, Mecklin JP, Khan PM, Lynch HT. The International Collaborative Group on Hereditary Non-Polyposis Colorectal Cancer (ICG-HNPCC). Dis Colon Rectum. 1991;34:424-5.

20. Vasen HF, Watson P, Mecklin JP, Lynch HT. New clinical criteria for hereditary nonpolyposis colorectal cancer (HNPCC, Lynch syndrome) proposed by the International Collaborative group on HNPCC. Gastroenterology. 1999;116:1453-6.

21. Goldberg Y, Nassif II, Pittas A et al. The anti-proliferative effect of sulindac and sulindac sulfide on HT-29 colon cancer cells: alterations in tumor suppressor and cell cycle- regulatory proteins. Oncogene. 1996;12:893-901.

22. Nugent KP, Farmer KC, Spigelman AD, Williams CB, Phillips RK. Randomized controlled trial of the effect of sulindac on duodenal and rectal polyposis and cell proliferation in patients with familial adenomatous polyposis. Br J Surg. 1993;80:1618-9.

23. Spagnesi MT, Tonelli F, Dolara P et al. Rectal proliferation and polyp occurrence in patients with familial adenomatous polyposis after sulindac treatment. Gastroenterology. 1994;106:362-366.

24. Pasricha PJ, Bedi A, O'Connor K et al. The effects of sulindac on colorectal proliferation and apoptosis in familial adenomatous polyposis. Gastroenterology. 1995;109:994-998.

25. Winde G, Schmid KW, Brandt B, Muller O, Osswald H. Clinical and genomic influence of sulindac on rectal mucosa in familial adenomatous polyposis. Dis Colon Rectum. 1997;40:1156-68.

26. Moorghen M, Orde M, Finney KJ, Appleton DR, Watson AJ. Sulindac enhances cell proliferation in DMH-treated mouse colonic mucosa. Cell Prolif. 1998;31:59-70.

27. Keller JJ, Offerhaus GJ, Polak M et al. Rectal epithelial apoptosis in familial adenomatous polyposis patients treated with sulindac. Gut. 1999;45:822-8.

118

Sulindac treatment in HNPCC |

28. Keller JJ, Offerhaus GJ, Hylind LM, Giardiello FM. Rectal epithelial apoptosis does not predict response to sulindac treatment or polyp development in presymptomatic familial adenomatous polyposis patients. Cancer Epidemiol Biomarkers Prev. 2002;11:670-1.

119

9

Summary

Chapter 9 |

Hereditary nonpolyposis colorectal cancer (HNPCC) or Lynch syndrome is a relatively

common, dominantly inheriting cancer syndrome characterized by the development of

neoplastic lesions in a variety of organs (gastrointestinal, endometrial/ovarian, urinary), most

prominently the colorectum. The syndrome is caused by germline mutations in one of the

mismatch repair genes (i.e MLH1, MSH2, MSH6 and PMS2), which are part of a complex

DNA repair system responsible for recognition and excision of mismatched nucleotides. The

ability to identify subjects at high risk for developing cancer has called for better

understanding of the carcinogenesis of the tumors in order to subsequently develop optimal

preventive strategies. The present thesis aimed to explore the carcinogenesis of the most

common HNPCC-associated tumors, namely colorectal and endometrial cancer, and to

evaluate the possible chemopreventive role of the nonsteroidal anti-inflammatory drug

sulindac in the tumorigenesis of colorectal cancer in HNPCC.

Clinically, the colorectal neoplastic process in HNPCC appears to follow an adenoma-

carcinoma progression similar to that described in familial adenomatous polyposis and

sporadic cases, though several of the clinical manifestations, as well as the molecular

pathogenesis underlying them, are distinctive. In most studies, HNPCC tumors are described

to have a right, i.e. proximal of the splenic flexure, predominance and to develop through an

accelerated adenoma-carcinoma sequence. The initiation process of an HNPCC tumor,

however, is a subject of controversy; when does mismatch repair dysfunction occur in the

process of tumorigenesis? And what is the role of the proximal colon in the carcinogenesis of

HNPCC tumors? In chapter 2 one hundred adenomas obtained from patients with HNPCC

were studied to elucidate their role in the carcinogenesis of colorectal tumors in HNPCC.

HNPCC adenomas in comparison to sporadic ones were more often located in the proximal

colon (50% vs. 26%, p=0.018). The proximal propensity was especially evident in highly

dysplastic HNPCC adenomas. All large HNPCC adenomas in the proximal colon were highly

dysplastic in comparison to only 44% in the distal colon and rectum. Loss of mismatch repair

protein expression was observed in all highly dysplastic HNPCC adenomas and in 55% of the

low grade dysplastic HNPCC adenomas. HNPCC adenomas have a proximal propensity

similar to hereditary nonpolyposis colorectal cancers and more importantly the accelerated

adenoma-carcinoma sequence seemed to be site specific in the colon of HNPCC subjects. The

results do not indicate that mismatch repair gene malfunction initiates adenoma development

but it is present at a very early stage of tumorigenesis and heralds tumor progression.

122

Summary |

That loss of function in a mismatch repair gene results in rapid progression to high grade

dysplasia in adenomas is also illustrated in chapter 3. Chapter 3 describes an extremely rare

clinical scenario in which initiation of adenoma growth and tumor progression is effected

through dual inherited germline mutations. Dysfunctional adenomatous polyposis coli

tumorsuppressor gene, demonstrated by abnormal expression of β-catenin, triggered early and

multiple adenoma formation, while loss of mismatch repair function, illustrated by loss of

MLH1 expression, was associated with high grade dysplastic adenomas. The combination of

defect genes in two separate phases of the adenoma-carcinoma sequence led clinically to rapid

clinical progression and drastic preventive and therapeutic measures, namely proctocolectomy

with ileal pouch-anal anastomosis.

Adenomas are indisputably precursor lesions of colorectal cancers in HNPCC. Traditionally,

hyperplastic polyps have been regarded as benign lesions, lacking the potential for neoplastic

progression. However, recently properties, such as K-ras mutations and chromosome 1p

deletions, suggesting potential for neoplastic progression have been demonstrated in

hyperplastic polyps. DNA microsatellite instability has also been described in hyperplastic

polyps and has been associated with microsatellite instable sporadic colorectal cancers. In

chapter 4, the possible role of hyperplastic polyps as precursor lesion in HNPCC was

analyzed. Clinical information on the age at colonoscopy and the location of the hyperplastic

polyps was collected. MLH1, MSH2, and MLH6 protein expression was evaluated using

immunohistochemistry. In our cohort study of 90 hyperplastic polyps none demonstrated loss

of mismatch repair function as demonstrated by loss of immunohistochemical expression of

mismatch repair proteins. Four polyps demonstrated adenomatous as well as hyperplastic

features. A proximal propensity as in HNPCC colorectal cancers and adenomas was not

observed for hyperplastic polyps. Actually the majority of hyperplastic polyps were resected

from the rectum. Apparently hyperplastic polyps, though frequently found at colonoscopy, do

not play a significant role in the carcinogenesis of microsatellite instable tumors in subjects

with a germline mismatch repair gene mutation. Our findings do not dictate any changes in

clinical practice.

Change in morphology and histology is a result of progressive acquisition of genomic

alterations and of a disbalance between proliferation and apoptosis in neoplastic cells. The

differences in morphological and histological aspects of adenomas resected from HNPCC

123

Chapter 9 |

patients (as described in chapter 2) and knowledge concerning the clinical behavior of

HNPCC tumors suggest that tumorigenesis in HNPCC differs from that in sporadic colorectal

cancer at an early stage. In chapter 5 we studied whether this difference could be explained

from disparities in expression of several cell cycle and apoptosis-related proteins in relation to

proliferation and apoptosis in HNPCC (n= 42) and sporadic adenomas (n= 48). Even though

no differences in proliferation and apoptosis indices were detected between HNPCC and

sporadic adenomas, subtle differences in expression of regulating proteins were observed.

Low-grade dysplastic HNPCC adenomas differed from sporadic ones by expressing more

often bcl-2, an anti-apoptotic proto-oncogene, and less often bax, an apoptotic promoter.

Containing a repetitive sequence, the bax gene is a target gene for mutations when mismatch

repair function is impaired. Apparently, change in bax expression occurs early in the

carcinogenesis of HNPCC lesions, before malignant transformation. High grade dysplastic

HNPCC adenomas are different from sporadic ones by expressing less often proliferation

stimulating proteins, cyclin B1, D3 and E. A striking finding in this study is the decreased

expression of p21, not correlating with the proliferation index, in high grade dysplastic

HNPCC adenomas in comparison to sporadic ones (6% vs. 53%, p=0.003). Transforming

growth factor-β type II receptor gene (TGFβRII) has been previously described as target gene

for mutation in cells with dysfunctional mismatch repair genes such as in HNPCC. Mutated

TGFβRII gene may lead to downregulation of p21 through the pRB signaling pathway. The

subtle differences found in adenomas support the concept of alternative carcinogenic

pathways at an early stage. However, the fact that changes in expression of proliferation- and

apoptosis-regulating proteins were not associated with changes in proliferation and apoptosis

suggests that also other regulating proteins or pathways play a (possibly more influential) role

in the carcinogenesis of HNPCC-related colorectal cancer.

The most common extra-colonic tumor in HNPCC-affected persons is endometrial cancer.

Even though in genetically predisposed women the cumulative lifetime risk for endometrial

cancer may exceed that of colorectal cancer, the consequences of mismatch repair dysfunction

on the carcinogenic process of endometrial cancers have received little attention. Chapter 6

describes a study undertaken to explore differences in carcinogenic pathways between

HNPCC and sporadic endometrial cancer by evaluating proliferation and apoptotic indices

and the immunohistochemical expression of proliferation and apoptosis regulating proteins.

Only subtle differences in immunohistochemical expression of proliferation- and apoptosis-

124

Summary |

regulating proteins in relation to proliferation and apoptosis were found between HNPCC and

sporadic endometrial cancers. A tendency towards an increased proliferation rate was

observed in HNPCC in comparison to sporadic endometrial cancers. Cyclin B1 is probably a

major cell cycle proliferation regulator in HNPCC endometrial cancers. Similarly to the

colorectal adenomas, loss of bax expression was observed more often in HNPCC endometrial

cancers than in sporadic ones. However, the altered bax expression apparently had a limited

functional role as the apoptotic index was not influenced. Even though bax seems to be a

target gene of the microsatellite instable phenotype in endometrial cancers the carcinogenic

consequences are disputable and loss of bax function is probably compensated by other,

unknown apoptosis inducing pathway(s). Despite the underlying differences in pathogenesis,

dysfunctional and functional mismatch repair genes, the carcinogenic pathway of HNPCC

endometrial cancers differs only in a subtle manner from that of sporadic cancers. This is in

accordance with the minor clinical diversity between HNPCC and sporadic endometrial

cancers.

Two aspects - high incidence and early age at diagnosis - of HNPCC endometrial cancer has

led to the implementation of a diversity of gynecologic screening programs. The theoretical

benefit of such programs is early detection of (pre)malignant lesions, thereby reducing

morbidity and mortality due to endometrial (and ovarian) cancer. However, the effectiveness

of gynecologic surveillance procedures has not been shown in either prospective or

retrospective studies. In chapter 7, the gynecologic screening program at the University

Medical Center Groningen was retrospectively analyzed. Despite annual screening and good

patient compliance no asymptomatic malignant lesions were detected during 179

appointments, but three asymptomatic premalignant lesions were detected and could be

treated appropriately. One interval endometrial cancer was detected as a result of clinical

symptoms. No abnormal CA 125 levels were measured and no ovarian cancers were detected.

Screening for any type of carcinoma is aimed primarily at the detection of early-stage disease

before symptoms occur and should result in a significantly improved overall survival. The 5-

year survival rate for endometrial cancer in pre- and postmenopausal HNPCC-women is high

(88%) and it remains unclear whether this high survival rate can be improved. Our present

study suggests that the potential gain of endometrial surveillance in HNPCC by means of

transvaginal ultrasound lies in the possibility of detecting premalignant lesions and thereby

possibly preventing malignancies to develop and avoiding extensive treatment i.e.

125

Chapter 9 |

radiotherapy. Our conclusion for the present clinical practice is that annual gynecologic

screening with transvaginal ultrasound as triage for endometrial sampling remains justified for

women motivated for it. Regardless of that, patients should be well instructed for early

recognition of alarm symptoms and rapid notification should be strongly encouraged.

Early detection and treatment of (pre)malignant lesions in the colon as well as the

endometrium is at present the primary mode of care of members of HNPCC families. Primary

prevention of these potentially harmful and mortal lesions would be much better. Increasing

evidence from cell line, animal and human studies reveals that the administration of

nonsteroidal anti-inflammatory drugs (NSAIDs) represents a viable option for

chemoprevention of familial adenomatous polyposis-associated and sporadic colorectal

cancer. The molecular basis for the chemoprotective action of NSAIDs has not yet been fully

elucidated. However, the use of NSAIDs is associated with a lower risk of colorectal adenoma

and cancer development, and a lower risk of recurrent colorectal adenomas and carcinomas.

These effects are independent of gender, age and site in the colorectum. The efficacy of

chemoprevention remains largely unexplored in HNPCC. In search of possibilities for

chemoprevention in HNPCC, the effect of the NSAID sulindac on the epithelium of normal

appearing colon of HNPCC patients is explored in chapter 8. Twenty-two subjects were

included in the randomized double-blind cross-over study and biopsies were taken from four

locations in the colon after the use of sulindac (150 mg twice daily) or placebo during two

separate four weeks periods. To identify possible mechanisms of action of chemoprevention

with sulindac proliferation and apoptosis regulating proteins were identified from previous

studies and were evaluated using immunohistochemistry. Proliferation labeling index (using

MIB-1 antibodies) was higher during sulindac treatment than during placebo use in both

ascending and transverse colon, but not in the sigmoid and the rectum. The apoptotic index

did not alter after sulindac treatment. Except for an increase in cyclin D3 upon sulindac

treatment, no further differences were found in expression of regulating proteins. Sulindac

proved to induce an increase of epithelial cell proliferative activity, both in the whole crypt

and in the upper crypt compartment, in the proximal colon, without affecting this activity in

the sigmoid and rectum, and without affecting apoptosis in any of the regions of the

colorectum. A limitation of all previous studies is that only the effects on the epithelium of

rectum and sigmoid were studied. It is thus unknown whether similar effects occur in the

proximal colon of subjects without a predisposition for HNPCC, but the finding of it in an

126

Summary |

HNPCC population has implications by its own. The preference for the development of

cancers in the proximal colon in HNPCC makes one reluctant to use agents that affect

proliferation in that part of the colon in a way that is generally considered to mark an

increased cancer risk. Although the clinical value of the biomarkers used is disputable, the

results cast doubts on the chemopreventive effects of sulindac, and NSAIDs in general, in

HNPCC.

Future perspectives

At the present, care and treatment of patients with HNPCC are mainly concentrated on

identifying those at risk, identifying germline mutations and, subsequently, entering these

subjects in standard screening programs for early detection of (pre)malignant lesions.

Although screening programs for HNPCC subjects are well-established they do not offer

complete protection. Further understanding of the carcinogenesis of HNPCC tumors and the

development of novel chemoprevention strategies could lead to a reduction of neoplastic

lesions and better care of HNPCC subjects.

In the present thesis, adenomas as well as hyperplastic polyps were studied as (possible)

precursor lesion of HNPCC colorectal cancer. Adenomas are undisputable precursor lesions

which can not be said from hyperplastic polyps. Sessile serrated polyps, containing

hyperplastic and adenomatous features, have been proposed as precursor lesions for sporadic

microsatellite unstable colorectal cancers. The microsatellite instability in these lesions seems

to be a result of inactivation of the MLH1 gene through promoter hypermethylation and, not

of a mutation in a gene as in HNPCC.1 In the study, described in this thesis, neither

hyperplastic nor serrated adenomas exhibited loss of mismatch repair function, thus

suggesting that these lesions are not precancerous in HNPCC-subjects. However, a role for

serrated adenomas in carcinogenesis of HNPCC can not be fully excluded as the number of

adenomas studied was small. The possibility of multiple precursor lesions for HNPCC

colorectal cancer necessitates resection of all macroscopic tumors seen during screening

colonoscopy.

In the near future most of the primary care of HNPCC patients will remain the early detection

and resection of premalignant lesions. Jarvinen et al have demonstrated that regular

127

Chapter 9 |

colonoscopy with polypectomy results in a significant survival advantage and a reduction in

the incidence of colorectal tumors.2 However, there is a need for better endoscopic

visualization as the small polyps prone to malignant transformation may be easily missed by

standard colonoscopy. Recently, Lecomte et al and Hurlstone et al demonstrated that relative

to conventional colonoscopy, high-resolution colonoscopy with chromoendoscopy markedly

improves the detection of adenomas in patients with HNPCC syndrome.3,4 Many more, novel

endoscopic techniques, e.g. narrow band imaging, fluorescence imaging, and elastic (light)

scattering spectroscopy, endocytoscopy and immunoscopy, are currently under investigation

but are not yet available for routine use. The most successful clinical method probably will be

a combination of techniques, providing wide-area surveillance and point detection methods.5

The appropriate regimen for reducing the risk of gynecologic cancer in HNPCC remains

subject of discussion. Gynecologic screening programs as described in chapter 7 may detect

premalignant as well as endometrial cancers. Schmeler et al published overwhelming

evidence to support prophylactic gynecologic surgery, hysterectomy and bilateral salpingo-

oophorectomy, with a prevented fraction of 100% for both ovarian and endometrial cancer.6

However, it remains to be seen whether the costs of the prophylactic surgery (including

surgical complications, premature menopause and its sequelae) will outweigh the benefits. In

addition, the reduced number of diagnosed gynecologic cancers may not translate into reduced

morbidity or mortality. Decision on the most appropriate method for gynecologic risk

reduction in HNPCC awaits the results of a prospective trial.

The underlying differences in pathogenesis, dysfunctional and functional mismatch repair

genes, and the different clinical behavior between HNPCC and sporadic cancers suggest two

distinct carcinogenic pathways. In the present thesis, in which HNPCC cancers were defined

as cancers diagnosed in patients fulfilling the Amsterdam criteria and/or having a germline

MMR gene mutation, only subtle differences could be identified between HNPCC and

sporadic premalignant and malignant lesions. Abdel-Rahman et al demonstrated that

colorectal cancers from patients belonging to families fulfilling the clinical Amsterdam

criteria for HNPCC, but without a germline MMR gene mutation or even without MMR gene

dysfunction in the tumors are distinct from colorectal cancers in patients belonging to families

linked to MMR gene defects and from sporadic cases.7 In the above mentioned study, tumors

from MMR gene mutation positive families have significantly more often active

128

Summary |

Wingless/Wnt signaling as indicated by aberrant β-catenin localization with or without

CTNNB1 mutations compared to tumors from MMR gene mutation negative families. At the

same time a possible role of the Wnt pathway and activating β-catenin mutations in the

(advanced) tumorigenesis of microsatellite instable colorectal tumors in HNPCC was

described by Johnson et al. 8 These studies illustrate the complexity and diversity of HNPCC

tumors. ‘HNPCC colorectal cancer’ is a term based on clinical criteria (e.g. Amsterdam

criteria) and most probably includes a diversity of colorectal tumors which may be subdivided

into separate groups depending on the precise pathogenesis, such as absence or presence and

type of microsatellite instability in combination with absence or presence of MMR gene

mutations. In the future these subgroups of HNPCC colorectal tumors should be studied

individually (when comparing with sporadic colorectal cancer) which most probably will

demonstrate involvement of novel predisposition genes and pathways in their carcinogenesis.

Preventive measures, e.g. chemoprevention, for risk reduction would be welcome aspects in

the treatment of patients at high inherited risk for cancer as in HNPCC. Although the efficacy

of non-steroidal anti-inflammatory drugs has been demonstrated in patients with familial

adenomatous polyposis and also in the general population, one should be reluctant to use

sulindac in HNPCC. In the present thesis we demonstrated that sulindac affects proliferation

in the proximal part of the colon in a way that is generally considered to mark an increased

cancer risk. Despite these results chemoprevention with NSAIDs should not be discarded in

HNPCC. Firstly, additional markers as surrogate end-points which correlate closely to disease

progression should be established to evaluate potential chemopreventive regimens before

drawing a definite conclusion concerning NSAIDs in HNPCC. Secondly, the development of

combinations of drugs could alter/increase the chemopreventive effect of NSAIDs by

targeting specific signaling pathways in (pre-)malignant cells. NSAIDS in combination with

other drugs, such as peroxisome proliferators-activated receptor-γ ligands (PPAR- γ) or 3-

hydroxy-3-methylglutaryl-coenzyme A reductase inhibitors (HRIs), may be efficient in the

chemoprevention of colorectal cancer in HNPCC as both drugs for example modulate the

expression of β-catenin.9 In the future, the mechanism of action of the chemopreventive agent

sulindac in combination with other drugs should be further clarified and the effectiveness

should be examined.

129

Chapter 9 |

The way forward in the care of HNPCC subjects is to attempt to prevent the evolution of

normal epithelium to adenomatous polyps to colorectal cancer through novel combinations of

chemopreventive agents and to maximize the detection rate of premalignant lesion, thereby

most likely further reducing the incidence of colorectal cancer.

130

Summary |

References 1. Jass JR, Whitehall VL, Young J et al. Emerging concepts in colorectal neoplasia.

Gastroenterology 2002;123:862-76.

2. Jarvinen HJ, Aarnio M, Mustonen H et al. Controlled 15-year trial on screening for colorectal cancer in families with hereditary nonpolyposis colorectal cancer. Gastroenterology 2000;118:829-34.

3. Lecomte T, Cellier C, Meatchi T et al. Chromoendoscopic colonoscopy for detecting preneoplastic lesions in hereditary nonpolyposis colorectal cancer syndrome. Clin Gastroenterol Hepatol 2005;3:897-902.

4. Hurlstone DP, Karajeh M, Cross SS et al. The role of high-magnification-chromoscopic colonoscopy in hereditary nonpolyposis colorectal cancer screening: a prospective "back-to-back" endoscopic study. Am J Gastroenterol 2005;100:2167-73.

5. Dekker E, Fockens P. New imaging techniques at colonoscopy: tissue spectroscopy and narrow band imaging. Gastrointest Endosc Clin N Am 2005;15:703-14.

6. Schmeler KM, Lynch HT, Chen LM et al. Prophylactic surgery to reduce the risk of gynecologic cancers in the Lynch syndrome. N Engl J Med 2006;354:261-9.

7. Abdel-Rahman WM, Ollikainen M, Kariola R et al. Comprehensive characterization of HNPCC-related colorectal cancers reveals striking molecular features in families with no germline mismatch repair gene mutations. Oncogene 2005;24:1542-51.

8. Johnson V, Volikos E, Halford SE et al. Exon 3 beta-catenin mutations are specifically associated with colorectal carcinomas in hereditary non-polyposis colorectal cancer syndrome. Gut 2005;54:264-7.

9. Jalving M, Koornstra JJ, de Jong S et al. Review article: the potential of combinational regimen with non-steroidal anti-inflammatory drugs in the chemoprevention of colorectal cancer. Aliment Pharmacol Ther 2005;21:321-39.

131

Samenvatting

Hereditair nonpolyposis colorectaal carcinoom (HNPCC) is een relatief vaak voorkomend

kanker-syndroom. Het wordt gekenmerkt door de ontwikkeling van kwaadaardige laesies in

verschillende organen, namelijk het maag-darmstelsel, baarmoeder en eierstokken en

urinewegen, maar vooral in de dikke darm en de endeldarm. Het syndroom kenmerkt zich

klinisch door het optreden van kanker op jonge leeftijd, het synchroon en metachroon

optreden van kankers, en een snelle groei van tumoren. Er is een autosomaal dominant

patroon van overerving, waarbij 50% van de nakomelingen aangedaan is. Het syndroom

wordt veroorzaakt door kiembaanmutaties in één van de mismatch repair (MMR) genen,i.h.b.

MLH1, MSH2, en MSH6. Door deze genen gecodeerde eiwitten maken deel uit van een

complex DNA-reparatie systeem, dat kleine fouten, zogenaamde nucleotide mismatches, die

ontstaan tijdens replicatie van DNA, herkent en verwijdert. De mogelijkheid om individuen

met dergelijk kiembaanmutaties en dus een sterk verhoogd kankerrisico te identificeren,

vraagt om een beter begrip van de ontstaanswijze van dit type kanker, goede

screeningprogramma’s, en het ontdekken en optimaliseren van medicamenteuze

behandelingsmogelijkheden die het ontstaan van kanker kunnen uitstellen en, idealiter,

tegengaan. De doelstelling van dit proefschrift was om de carcinogenese van de twee meest

voorkomende HNPCC tumoren, dikkedarm- en baarmoederkanker, te onderzoeken, en de

mogelijke chemopreventieve werking van het nonsteroidale anti-inflammatoroire

geneesmiddel sulindac in HNPCC te bestuderen.

Het ontstaan van HNPCC-geassocieerde dikkedarmkanker volgt het klassieke model voor het

ontstaan van darmkanker in familiaire adenomateuze polyposis (FAP) en in de algemene

bevolking. Dit model, de zogenaamde adenoma-carcinoma sequentie, beschrijft het ontstaan

van darmkanker vanuit normaal darmslijmvlies via verschillende stadia van poliepen met

steeds ernstiger afwijkende cytologie en histologie (dysplasie). Echter, het klinische gedrag

van HNPCC tumoren en de onderliggende moleculaire afwijkingen zijn verschillend van

FAP-tumoren en van kanker in de algemene populatie (zonder erfelijke belasting). De meeste

studies beschrijven dat HNPCC-darmtumoren voornamelijk in de rechter helft van de darm,

proximaal van de flexura lienalis, voor komen. In tegenstelling tot FAP kenmerkt HNPCC

zich niet door een overvloed aan darmpoliepen maar de enkele poliepen die voorkomen lijken

zich in een versneld tempo te kunnen ontwikkelen tot kanker via een versnelde adenoma-

carcinoma sequentie. De eerste aanzet tot het ontstaan van darmkanker in HNPCC staat echter

ter discussie; wanneer treedt de versnelling op? Op welk moment in de tumor genese verliest

134

Samenvatting |

het mismatch repair systeem zijn functie? Hoe onderscheidt: het rechter deel van de dikke

darm zich in het ontstaan van (pre-) maligne darmtumoren in HNPCC? In hoofdstuk 2 van dit

proefschrift hebben we 100 HNPCC-gerelateerde adenomen (dysplastische poliepen) van

verschillend graad van dysplasie bestudeerd om de rol van adenomen in de carcinogenese van

HNPCC te doorgronden. De locatie, grootte, histologische kenmerken (tubulair of villeus) en

graad van dysplasie van HNPCC-adenomen werden vergeleken met die van sporadische

adenomen. Sporadische adenomen zijn adenomen, verwijderd uit de darm van patiënten

zonder darmziekte, zoals colitis ulcerosa of ziekte van Crohn, en zonder een sterke erfelijke

belasting voor darmkanker. HNPCC-adenomen zijn vaker dan sporadische adenomen te

vinden in het proximale deel van de dikkedarm (50% vs 26%, p=0.018). De proximale

voorkeur was vooral uitgesproken voor hooggradig dysplastische poliepen. Alle HNPCC-

adenomen in het proximale colon met een diameter groter dan 5 mm waren hooggradig

dysplastisch terwijl maar 44% van de grotere poliepen in het distale deel van de darm

hooggradig dysplastisch waren. Verlies van één van de MMR-eiwitten expressie bij

immuunhistochemisch onderzoek werd gevonden in alle hooggradig dysplastische en in 55%

van de laaggradig dysplastische HNPCC-adenomen. Uit deze bevindingen kan worden

geconcludeerd dat HNPCC-adenomen, net als HNPCC-kankers, vooral proximaal in de dikke

darm voorkomen en, nog belangrijker, de versnelling van de adenoma-carcinoma sequentie

uniek lijkt te zijn voor dit proximale deel van de darm. Mismatch repair dysfunctie initieert

niet de ontwikkeling van een adenoom maar is in een zeer vroeg stadium aanwezig in de

carcinogenese en is een voorbode van tumorprogressie.

Dat verlies van functie van een mismatch repair gen leidt tot een versnelde transformatie van

laaggradige naar hooggradige dysplasie is ook geïllustreerd middels een casuïstische

mededeling in hoofdstuk 3. Dit hoofdstuk beschrijft een uitzonderlijke situatie waarin de

initiatie van de adenoma-carcinoma sequentie en de progressie daarvan in hoge mate werd

bepaald door een tweetal geërfde kiembaanmutatie, te weten in het adenomateuze polyposis

(APC) gen en in het MLH1-gen. Dysfunctie van het APC-gen wordt verantwoordelijk geacht

voor de initiatie van het ontstaan van adenomen en tumoren. Kiembaanmutaties in dit gen zijn

verantwoordelijk voor FAP met de ontwikkeling van multipele adenomen op jonge leeftijd.

Toegenomen nucleaire ophoping van β-catenine is geassocieerd met dysfunctie van het APC-

gen, en werd in alle laaggradig dysplastische adenomen gezien. Verlies van MLH1 expressie

werd gezien in alle hooggradig dysplastische adenomen en wordt daardoor geassocieerd met

135

progressie van tumorformatie. De combinatie van twee kiembaanmutaties leidde tot een zeer

ernstige klinische presentatie op jonge leeftijd waardoor een drastische preventieve en

therapeutische ingreep moest worden verricht, namelijk proctocolectomie met een ileoanale

anastomose met pouch.

Adenomen zijn zonder twijfel voorlopers van (HNPCC-gerelateerde) colorectale tumoren.

Hyperplastische poliepen worden van oudsher beschouwd als onschuldige laesies zonder

neiging om maligne te ontaarden. Recent zijn er aanwijzingen gevonden dat hyperplastische

poliepen mogelijk maligne potentie hebben. K-ras mutaties, chromosoom-1p deleties en zelfs

DNA-microsatellietinstabiliteit zijn beschreven in hyperplastische poliepen. In het bijzonder

de laatste bevinding roept vragen op over een mogelijk neoplastische rol van hyperplastische

poliepen in HNPCC. In hoofdstuk 4 onderzoeken we de mogelijke rol van hyperplastische

poliepen als voorloper van HNPCC-kanker. Retrospectieve informatie werd verzameld

betreffende leeftijd bij poliepresectie en locatie van hyperplastische poliepen in de darm.

MLH1, MSH2, en MSH6 eiwitexpressie werd geanalyseerd middels immuunhistochemie. In

totaal konden 90 HNPCC-gerelateerde hyperplastische poliepen worden geincludeerd. Bij

histologisch herbeoordeling bleek er in vier poliepen spraken te zijn van een adenomateuze

component naast het hoofdzakelijk hyperplastische weefsel. In tegenstelling tot adenomen en

maligne tumoren bij HNPCC waren de meeste hyperplastische poliepen gelokaliseerd in het

distale deel van de darm, het sigmoid en het rectum. Alle 90 hyperplastische poliepen,

inclusief de gemengde poliepen, brachten de mismatch repair genen tot expressie. Al worden

hyperplastische poliepen regelmatig gevonden gedurende de screeningcoloscopieën, zij

blijken geen specifieke rol te spelen bij het ontstaan van darmkanker in HNPCC. Onze

bevindingen geven geen aanleiding om de huidige screeningpraktijk te veranderen.

Progressie naar opeenvolgende stadia in de klassieke adenoma-carcinoma sequentie wordt

gekenmerkt door het optreden van specifieke genetische mutaties. Er treedt een dysbalans op

in proliferatie en apoptose van neoplastische cellen. Klinische eigenschappen van HNPCC-

tumoren en de verschillen gevonden in HNPCC en sporadische adenomen zoals beschreven in

de voorgaande hoofdstukken, pleiten voor verschil in de vroege carcinogenese van HNPCC-

tumoren in vergelijking met sporadische tumoren. In hoofdstuk 5 hebben we bekeken of deze

verschillen zich vertalen in veranderde expressie van specifieke proliferatie- en apoptose-

regulerende eiwitten in relatie tot proliferatie en apoptose-indices in HNPCC (n=42) en

136

Samenvatting |

sporadische (n=48) adenomateuze poliepen. Terwijl geen verschil in proliferatie en apoptose-

indices werd geconstateerd waren er wel kleine verschillen in de expressie van een aantal

proliferatie en apoptose regulerende eiwitten. Laaggradig dysplastische HNPCC-adenomen

brengen bcl-2, een proto-oncogen en apoptoseremmer, vaker tot expressie en bax, een

apoptose-promotor, minder vaak. Aangezien het bax gen een repetitieve sequentie bevat is het

gen kwetsbaar voor fouten als het MMR-systeem niet goed werkt. Mutaties in bax spelen

kennelijk al een belangrijke rol in de vroege carcinogenese van HNPCC-laesies, nog voor het

stadium van ernstige dysplasie. Hooggradig dysplastische HNPCC-adenomen verschillen van

sporadische doordat ze minder vaak de proliferatie stimulerende cyclines B1, D3, en E tot

expressie brengen. Een opmerkelijke bevinding in dit immuunhistochemische onderzoek is de

verminderde expressie van p21, zonder correlatie met de proliferatie index, in hooggradig

dysplastische HNPCC-adenomen ten opzichte van hooggradig dysplastische sporadische

adenomen (6% vs. 53%, p=0.003). P21 is een cycline afhankelijke kinase remmer dat een

negatieve werking heeft op de proliferatie. De downregulatie van p21 kan een uiting zijn van

dysfunctie van het transforming growth factor-β type II receptor gen (TGFβRII). Het

TGFβRII gen is beschreven als een target gen voor mutaties in cellen met een niet

functionerend MMR-systeem zoals in HNPCC. Het TGFβRII gen bevat net als het bax gen

een repetitieve sequentie van nucleotiden. Mutatie van het TGFβRII gen kan leiden tot een

downregulatie van p21 door de pRB signaalroute. De kleine verschillen gevonden tussen

adenomen van beide groepen ondersteunen de theorie dat de carcinogenese van HNPCC-

tumoren al in een zeer vroeg stadium verschillen van die van sporadische colorectale tumoren.

Aangezien de verschillen gevonden in proliferatie en apoptose regulerende eiwitten niet

correleerden met een verandering in proliferatie- of apoptose-indices spelen er waarschijnlijk

nog andere regulerende eiwitten of routes een (belangrijkere) rol in de carcinogenese van

HNPCC-tumoren.

De meest voorkomende extra-intestinale vorm van kanker in HNPCC is baarmoeder- of

endometriumkanker. Vrouwelijke HNPCC-patiënten hebben een verhoogd cumulatief lifetime

risico voor endometriumkanker, dat mogelijk zelfs hoger is dan dat voor darmkanker. Toch is

er veel meer bekend over de carcinogenese van darmkanker dan van endometrium-kanker. In

hoofdstuk 6 beschrijven we een immuunhistochemische studie naar de carcinogenese van

HNPCC-gerelateerde endometriumkanker. De studie heeft een vergelijkbare opzet als die

beschreven in hoofdstuk 5, namelijk het immuunhistochemisch evalueren van proliferatie en

137

apoptose regulerende eiwitten in vergelijking met proliferatie- en apoptose-indices. Subtiele

verschillen in expressie van proliferatie- en apoptose regulerende eiwitten werden gevonden

tussen HNPCC en sporadische endometriumtumoren. De proliferatie index van HNPCC-

endometriumtumoren was marginaal hoger dan die van sporadische tumoren. Cycline B1

expressie was significant hoger in HNPCC-endometriumtumoren in vergelijking met

sporadische tumoren. Mogelijk is cycline B1 de sturende kracht achter proliferatie in HNPCC.

Net als in HNPCC-adenomen werd vaker een verlies in bax expressie geconstateerd in

HNPCC-endometriumtumoren dan in sporadische tumoren. Echter, bax expressie was niet

gecorreleerd met de apoptose-index. Er spelen mogelijk andere apoptose regulerende eiwitten

een grotere rol in de carcinogenese van HNPCC-endometriumtumoren. Ondanks een evident

andere onderliggende pathogenese, namelijk een dysfunctionerend mismatch repair systeem,

is de carcinogenese van HNPCC-endometriumtumoren maar subtiel verschillend van die van

sporadische endometriumtumoren. Dit is in overeenstemming met de marginale klinische

verschillen.

Twee aspecten– hoge incidentie en jonge leeftijd bij ontstaan- van HNPCC-endometrium-

tumoren hebben geleid tot implementatie van een verscheidenheid aan gynaecologische

screeningprogramma’s. De theoretische winst van screeningprogramma’s is vroege detectie

van kwaadaardige tumoren met als vervolg een vermindering in morbiditeit en een

verbetering in overleving. Echter, het gewenste effect van gynaecologisch

screeningprogramma’s in HNPCC is nooit aangetoond in prospectieve of retrospectieve

onderzoeken. In hoofdstuk 7 evalueren we onze 10-jarige ervaring met het gynaecologische

screeningprogramma voor patiënten met HNPCC in het Universitair Medisch Centrum

Groningen. Het screeningprogramma in het UMCG bestaat uit een jaarlijks polikliniekbezoek

waarbij anamnese, gynaecologisch onderzoek en een transvaginale echografie worden

verricht. Tevens wordt CA-125 in het serum bepaald als screeningsmethode voor

ovariumkanker. Ondanks jaarlijkse bezoeken en een goede compliance van de patiënten werd

gedurende 179 reguliere polikliniekafspraken geen endometriumkanker gediagnosticeerd.

Drie asymptomatische premaligne endometriumafwijkingen werden gediagnosticeerd en deze

konden probleemloos worden behandeld. Een interval endometriumkanker werd vastgesteld

aan de hand van klachten zes maanden na een reguliere afspraak. Alle CA-125 waardes waren

normaal. Screening voor alle vormen van kanker heeft als doel vroegdetectie voordat er

klachten zijn en moet resulteren in een betere overleving. De 5-jaars overleving van

138

Samenvatting |

endometriumkanker is hoog (88%) en het is onduidelijk of dit kan worden verbeterd. Ons

onderzoek kan hier geen uitspraak over doen. Echter, het feit dat we middels echografie op het

spoor kwamen van premaligne afwijkingen biedt perspectief en een mogelijkheid om de

verdere ontwikkeling tot kwaadaardige tumoren en de daarbij eventuele noodzakelijk zware

behandeling, zoals radiotherapie, te voorkomen. De huidige gynaecologische

screeningpraktijk voor vrouwelijke HNPCC-patiënten waarbij echografie wordt gebruikt als

indicatiemiddel voor een endometriumbiopsie heeft bestaansrecht voor wie het onderzoek

wensen. Deze patiënten dienen gewezen te worden op de gebreken van het onderzoek. Alle

vrouwelijke HNPCC-patiënten dienen goed geïnstrueerd te worden op het (h)erkennen van

vroege alarmsymptomen en gestimuleerd te worden deze symptomen vroegtijdig en

laagdrempelig mede te delen aan de behandelende specialist.

Het belangrijkste deel van de zorg voor mensen met HNPCC ligt in vroegdetectie van

(pre)maligne afwijkingen van dikke darm en endometrium. Ideaal in de zorg voor HNPCC

zou een medicamenteuze behandelingsmogelijkheid zijn die het ontstaan van kanker zou

tegengaan (chemopreventie). Nonsteroidale anti-inflammatoire middelen, een groep

geneesmiddelen waar ook sulindac en aspirine onder vallen, kunnen de adenoma-carcinoma

sequentie in de darm remmen. Chemopreventieve studies bij personen met FAP toonden

regressie van adenomateuze poliepen onder behandeling met sulindac. Omdat deze poliepen

voorlopers van kanker zijn, impliceert dit dat het ontstaan van kanker wordt geremd. Grote

epidemiologische studies onder de algemene bevolking hebben bevestigd dat regelmatig

gebruik van aspirine het risico op darmkanker vermindert. Het mechanisme dat aan

chemopreventie van darmkanker door NSAIDs ten grondslag ligt is slechts deels bekend. Of

NSAIDs ook een gunstige werking hebben op het darmepitheel van mensen met aanleg voor

HNPCC is ook niet bekend. In hoofdstuk 8 onderzoeken we de mogelijkheid van

chemopreventie met sulindac in HNPCC en bekijken we het effect van kortdurend sulindac

gebruik op het normaal ogende darmepitheel van mensen met aanleg voor HNPCC.

Tweeëntwintig personen werden geincludeerd in een gerandomiseerd dubbel-blind cross-over

onderzoek waarin na een behandeling van vier weken met tweemaal daags 150 mg sulindac of

placebo op vier locaties in de darm biopten werden genomen. Onderzoek naar het

mechanisme van het effect van sulindac op het darmepitheel werd verricht middels

immuunhistochemisch onderzoek naar de mate van proliferatie en apoptose van het colon- en

rectumepitheel en naar proliferatie en apoptose regulerende eiwitten die reeds eerder in de

139

proefschrift zijn beschreven. Na gebruik van sulindac werd een significant hogere proliferatie-

index in het rechter deel van de darm geconstateerd dan na gebruik van placebo. Deze

toename in proliferatie werd niet waargenomen in het linker deel van de darm en rectum. De

apoptose-indices veranderden onder het gebruik van sulindac in geen van de delen van de

darm. Behalve een toename in cycline D3 expressie werden geen verschillen gezien in

expressie van proliferatie of apoptose regulerende eiwitten na het gebruik van sulindac.

Samenvattend veroorzaakte een kortdurende behandeling met sulindac een toename van de

proliferatie-activiteit in het proximale deel van de darm, zowel in de gehele crypte als in het

bovenste deel van de crypte, zonder effect te hebben op de proliferatie in het distale deel van

de darm of op apoptose in de gehele darm van HNPCC patiënten. Niet eerder is dit locatie

afhankelijke effect van sulindac bestudeerd. Het is dus onbekend of sulindac dit effect ook

vertoont in mensen zonder erfelijk aanleg voor HNPCC. Echter deze bevinding heeft mogelijk

wel serieuze implicaties voor HNPCC. De voorkeur van het ontstaan van darmkanker in de

proximale helft van de darm in HNPCC maakt dat men terughoudend zal zijn om een middel

te gebruiken dat de proliferatie beïnvloedt op een wijze die algemeen geassocieerd wordt met

een verhoogd risico op kanker.

Toekomst

De zorg voor en de behandeling van patiënten met HNPCC concentreert zich met name op het

identificeren van mensen met een verhoogd risico, het aantonen van kiembaanmutaties en

vervolgens het includeren van deze mensen in screeningprogramma’s voor vroege detectie

van (pre)maligne afwijkingen. Al zijn de screeningprogramma’s voor HNPCC patiënten

adequaat en goed toegankelijk, ze geven geen volledige bescherming. Uitbreiding van de

kennis over de carcinogenese van HNPCC-tumoren en de ontwikkeling van (nieuwe)

chemopreventieve mogelijkheden zou kunnen leiden tot een vermindering van neoplastische

laesies en een verbeterde zorg voor HNPCC patiënten.

In dit proefschrift zijn zowel adenomen als hyperplastische poliepen bestudeerd als

(mogelijke) precursor laesies van HNPCC-darmkanker. Adenomen zijn zonder twijfel

voorlopers van HNPCC-darmkanker maar dit kan niet worden geconcludeerd over

hyperplastische poliepen. Sessiele serrated poliepen, poliepen met zowel dysplastische als

hyperplastische kenmerken, zijn voorgesteld als mogelijke voorlopers van sporadische

140

Samenvatting |

microsatelliet instabiele darmtumoren. De microsatellietinstabiliteit in deze laesies word

waarschijnlijk veroorzaakt door inactivatie van het MLH1 gen door promoter hypermethylatie

en niet, zoals in HNPCC, door mutatie(s) in het gen.1 In het onderzoek beschreven in

hoofdstuk 4 brengen zowel alle hyperplastische poliepen als alle serrated adenomen MLH1 en

de andere MMR-eiwitten tot expressie, hetgeen de theorie dat deze twee type laesies

voorlopers zijn van HNPCC-darmkanker niet ondersteunt. Echter, een rol voor serrated

adenomen in de carcinogenese van HNPCC kan niet in zijn geheel worden uitgesloten

aangezien het aantal bestudeerde serrated adenomen in onze studie zeer gering was. De

mogelijkheid van verschillende type voorlopers in HNPCC maakt het reseceren van alle

poliepen en poliepjes, gezien bij colonoscopie, noodzakelijk.

In de nabije toekomst blijft vroege detectie van premaligne laesies het grootste deel van de

primaire zorg van HNPCC-patiënten. Jarvinen et al hebben beschreven dat regelmatige

colonoscopie met poliepectomie resulteert in een significante overlevingsverbetering en een

vermindering in de incidentie van darmkanker.2 Echter, is het noodzakelijk om het

endoscopisch zichtbaar maken van kleine poliepen met een verhoogde neiging tot maligne

transformatie, te verbeteren. Lecomte et al en Hurlstone et al hebben aangetoond dat in

vergelijking met conventionele colonoscopie, high resolution colonoscopie met chromoscopie

de detectie van adenomen in patiënten met HNPCC beduidend verbetert.3,4 Vele nieuwe

endoscopische technieken, b.v. narrow band imaging, fluorescence imaging, en elastic (light)

scattering spectroscopie, endocytoscopie en immunoscopie, worden momenteel ontwikkeld en

onderzocht maar zijn nog niet geschikt voor dagelijks gebruik. De meest succesvolle klinische

methode zal waarschijnlijk een combinatie zijn van verschillende technieken.5

De meest geschikte wijze voor het verminderen van het risico op gynaecologische kanker in

HNPCC blijft onderwerp van discussie. Gynaecologische screeningprogramma’s als

beschreven in hoofdstuk 7 kunnen zowel premaligne als maligne laesies van het endometrium

ontdekken maar overlevingsverbetering is nooit aangetoond. Schmeler et al publiceerden een

ruime hoeveelheid bewijs ter ondersteuning van profylactische chirurgie, hysterectomie en

bilaterale salpingo-oophorectomie, leidend tot 100% preventie van baarmoeder- en eierstok-

kanker.6 Echter, het is niet aangetoond of de kosten van profylactische chirurgie (waaronder

chirurgische complicaties, vroegtijdig menopauze en de gevolgen daarvan) opwegen tegen de

winst. Tevens is niet aangetoond of de vermindering van het aantal gediagnosticeerde

141

gynaecologische tumoren leidt tot een vermindering in morbiditeit en/of mortaliteit. Een

definitieve uitspraak over de meest correcte wijze van screening en behandeling ter

vermindering van het risico op gynaecologische kanker in HNPCC kan pas gemaakt worden

na het verrichten van een prospectieve trial.

De onderliggende verschillen in de pathogenese, en klinische kenmerken tussen HNPCC en

sporadische kankers duiden op twee verschillende routes van carcinogenese. In dit

proefschrift werden HNPCC-tumoren geclassificeerd als tumoren gediagnosticeerd in

patiënten die voldoen aan de Amsterdam II criteria en/of met een aangetoonde

kiembaanmutatie in een MMR-gen. Er werden subtiele verschillen gevonden tussen deze

groep premaligne en maligne tumoren en sporadische tumoren. Abdel-Rahman et al toonden

aan dat darmkankers van patiënten uit families die voldoen aan de klinische definitie van

HNPCC volgens de Amsterdam criteria maar zonder bewezen MMR-genmutatie zich

onderscheiden van darmkankers van patiënten uit families met een bewezen mutatie en van

sporadische darmkanker.7 De bovengenoemde studie demonstreerde dat tumoren uit MMR-

genmutatie families significant vaker een geactiveerde Wnt-signaalroute tonen, gekenmerkt

door nucleaire accumulatie van β-catenine met of zonder CTNNB1 mutaties, in vergelijking

met tumoren uit MMR-genmutatie negatieve families. Johnson et al beschreven tevens een

mogelijke rol voor de Wnt-route en activerende β-catenine mutaties in HNPCC.8 Deze studies

illustreren de complexiteit en diversiteit van HNPCC tumoren. De benoeming ‘HNPCC

darmkanker’ is veelal gebaseerd op klinische criteria (bv Amsterdam criteria) en bevat meest

waarschijnlijk een diversiteit aan darmkankers die onderverdeeld kunnen worden in

verschillende groepen afhankelijk van precieze pathogenese, zoals aan- of afwezigheid en

type van microsatelliet instabiliteit in combinatie met aan- of afwijzigheid van MMR-

genmutaties. In de toekomst zullen deze subgroepen van HNPCC-darmkanker afzonderlijk

moeten worden onderzocht (als ze worden vergeleken met sporadische darmkanker) om

evidente verschillen te vinden en mogelijk nieuwe genen en routes in hun carcinogenese te

ontdekken en te belichten.

Preventieve maatregelen, bv. chemopreventie, ter reductie van het kankerrisico zijn een

gewenst onderdeel in de zorg voor patiënten met een hoog risico op kanker zoals in het geval

van HNPCC. Al is het nut van NSAIDs aangetoond in zowel patiënten met FAP als in de

algemene populatie, men moet terughoudend zijn met het gebruik van sulindac in HNPCC. In

142

Samenvatting |

dit proefschrift laten we zien dat sulindac de proliferatie in het rechter deel van het colon van

HNPCC-patiënten beïnvloedt op een wijze, die algemeen geassocieerd wordt met een

verhoogd kankerrisico. Ondanks deze resultaten moet chemopreventie middels NSAIDs in

HNPCC niet worden afgeschreven. Ten eerste, additionele merkers als surrogaateindpunten,

die nauw correleren met ziekteprogressie moeten worden vastgesteld om de

chemopreventieve werking te beoordelen voordat conclusies kunnen worden getrokken. Ten

tweede, het ontwikkelen van combinaties van (chemopreventieve) middelen om de

chemopreventieve werking van sulindac veranderen/verbe-teren door meerdere signaalroutes

te beïnvloeden in premaligne laesies. NSAIDs in combinatie met andere middelen zoals

peroxisome proliferators-activated receptor-γ ligands (PPAR- γ) of 3-hydroxy-3-

methylglutaryl-coenzyme A reductase remmers (HRIs of statines), zijn mogelijke effectief als

chemopreventie in HNPCC aangezien beide middelen bijvoorbeeld invloed hebben op de

expressie van β-catenine.9 In de toekomst, zullen de het mechanismen van de

chemopreventieve werking van sulindac en combinaties met andere middelen verder moeten

worden opgehelderd en de effectiviteit (in HNPCC) moeten worden onderzocht.

De weg vooruit in de zorg voor mensen met HNPCC is het voorkomen van de ontwikkeling

van normaal epitheel naar voorloper laesies naar kanker door nieuwe combinaties van

chemopreventive middelen, en het maximaliseren van de detectie van premaligne laesies.

Hiermee zal waarschijnlijk de incidentie van (darm)kanker in HNPCC nog verder

verminderen.

143

References

1. Jass JR, Whitehall VL, Young J et al. Emerging concepts in colorectal neoplasia. Gastroenterology 2002;123:862-76.

2. Jarvinen HJ, Aarnio M, Mustonen H et al. Controlled 15-year trial on screening for colorectal cancer in families with hereditary nonpolyposis colorectal cancer. Gastroenterology 2000;118:829-34.

3. Lecomte T, Cellier C, Meatchi T et al. Chromoendoscopic colonoscopy for detecting preneoplastic lesions in hereditary nonpolyposis colorectal cancer syndrome. Clin Gastroenterol Hepatol 2005;3:897-902.

4. Hurlstone DP, Karajeh M, Cross SS et al. The role of high-magnification-chromoscopic colonoscopy in hereditary nonpolyposis colorectal cancer screening: a prospective "back-to-back" endoscopic study. Am J Gastroenterol 2005;100:2167-73.

5. Dekker E, Fockens P. New imaging techniques at colonoscopy: tissue spectroscopy and narrow band imaging. Gastrointest Endosc Clin N Am 2005;15:703-14.

6. Schmeler KM, Lynch HT, Chen LM et al. Prophylactic surgery to reduce the risk of gynecologic cancers in the Lynch syndrome. N Engl J Med 2006;354:261-9.

7. Abdel-Rahman WM, Ollikainen M, Kariola R et al. Comprehensive characterization of HNPCC-related colorectal cancers reveals striking molecular features in families with no germline mismatch repair gene mutations. Oncogene 2005;24:1542-51.

8. Johnson V, Volikos E, Halford SE et al. Exon 3 beta-catenin mutations are specifically associated with colorectal carcinomas in hereditary non-polyposis colorectal cancer syndrome. Gut 2005;54:264-7.

9. Jalving M, Koornstra JJ, de Jong S et al. Review article: the potential of combinational regimen with non-steroidal anti-inflammatory drugs in the chemoprevention of colorectal cancer. Aliment Pharmacol Ther 2005;21:321-39.

144

Dankwoord

Dankwoord |

“Er gaat niets boven Groningen!”

Met grappen als ‘moeten we eerst 00 draaien als we je bellen’ en ‘moeten we een paspoort

meenemen als we je komen bezoeken’ reisde ik vol goede moed van de vertrouwde Randstad

naar het verre hoge noorden om onderzoek te gaan doen in toenmalige AZG. Het werd een

boeiende en leerzame reis waarvan het lang verwachte eindproduct nu is afgerond. Veel

mensen kwam ik tegen op mijn reis, velen van hen begeleidden mij en droegen bij aan de

totstandkoming van dit proefschrift. Graag wil ik daarom iedereen bedanken die op enige

wijze heeft bijgedragen aan mijn promotie, waarbij ik een aantal in het bijzonder wil noemen.

Allereerst ben ik alle patiënten die deelnamen in de chemopreventie studie erg dankbaar.

Grote bewondering heb ik voor hun inzet om een extra, meestal niet pijnloos,

screeningonderzoek te ondergaan zonder er direct mogelijk profijt van te hebben.

Mijn promotoren Prof. dr. J.H. Kleibeuker, Prof. dr. H. Hollema en Prof. dr. A.G.J. van der

Zee ben ik bijzonder dankbaar voor hun vertrouwen om mij, ondanks mijn uitgesproken doel

om gynaecoloog te worden, aan te nemen voor dit, m.n. gastro-enterologisch, onderzoek.

Jan, bedankt voor je eeuwige geduld, je positieve houding en de rust waarmee je een

probleem, presentatie of manuscript beoordeelt en bespreekt.

Harry, je was een hele bijzondere (co-)promotor. Je kennis van zaken en de wijze waarop jij

problemen van meerdere kanten wist te belichten was de prikkel die ik regelmatig nodig had

om me verder te verdiepen in het onderzoek. Tussen alle cola’s en koffie’s, grappen, sterke

verhalen (zeker geen roddels) en goede gesprekken was coupes beoordelen met jou nooit saai.

Ate, jij wist met je commentaar van een (in mijn ogen klaar-om-te-versturen) manuscript een

bondig maar toch uitgediept artikel te maken. Als gynaecoloog had je zelfs een kritische blik

op het gastro-enterologische deel van dit proefschrift.

De leescommisie, bestaande uit Prof. dr. R.M.W. Hofstra, Prof. dr. J.H.J.M. van Krieken, en

Prof. dr. E.G.E. de Vries, ben ik zeer erkentelijk voor het beoordelen, hun zinvolle

commentaar, en vanzelfsprekend hun goedkeuring van mijn proefschrift.

147

Marian Mourits, mijn dank is groot voor jouw medewerking aan de gynaecologische

manuscripten. Je aanstekelijke enthousiasme voor zowel de kliniek als het onderzoek, de

wijze waarop jij je leven als arts, moeder, en sportvrouw indeelt is een inspiratie voor mij.

Tineke en Wytske, jullie hebben bergen werk voor mij verzet waarvoor ik jullie erg dankbaar

ben. Ondanks mijn soms wat te chaotische houding kwam al het werk in een georganiseerde

wijze af. Hierbij wil ik ook de laboratoria pathologie en oncologie en het secretariaat van de

pathologie bedanken voor hun medewerking aan mijn proefschrift.

Het endoscopie centrum en alle endoscopisten wil ik bedanken voor de hulp in het uitvoeren

van de sulindac-studie. De afdeling genetica wil ik bedanken voor het (herhaaldelijk)

aanleveren van een up-to-date database van (mogelijke) HNPCC patiënten.

Petra en Gonny verdienen een speciaal dankwoord voor alle rustpauzes, koffie en gebak, maar

vooral voor administratieve ondersteuning, het jagen achter handtekeningen en het versturen

van manuscripten.

De gezelligste kamer op de toenmalige Kidney Alley (thans Maagdarm Passage) was de

MDL-onderzoekerskamer. Maran, bedankt voor alle genetische input in mijn studie. Als

kamergenoot was je de rust zelve tussen je georganiseerde chaos van mappen op de grond.

Met veel plezier hebben we posters en presentaties gezamenlijk voorbereid en deelgenomen

aan (internationale) congressen. De avonden Kolonisten onder her genot van een goed glas

wijn en heerlijk eten mis ik nog steeds. Het was voor mij vanzelfsprekend om je te vragen als

paranimf. Mirjam, het enthousiasme waarmee jij mij de eerste keer rondleidde in het AZG

staat me nog steeds bij. Dat enthousiasme bracht ons van vele gezellige dagen op het werk

naar heel veel rondjes Paterswoldse meer, tot aan de marathon van New York. Ervaringen om

nooit meer te vergeten. Hilde je bent een belofte, je georganiseerdheid en doelmatigheid is een

voorbeeld voor iedereen.

Jan Jacob, jij introduceerde mij in de ins en outs van de onderzoekswereld maar liet met name

zien dat samen onderzoek doen leuker, leerzamer en zeker productiever is.

148

Dankwoord |

Filip, tja ik durf met mijn beperkt literaire en schriftelijke vermogens (m.n. in vergelijking met

jouw talenten) hier bijna geen uitspraken over je te doen. Filip, je bent het synoniem van

Groningen, AZG en promotie-onderzoek. Je was en bent een vriend, collega en getuige voor

het leven en nu ook nog mijn paranimf.

Voor al mijn vrienden uit verre en minder verre oorden; dank jullie voor jullie steun en voor

de heerlijke afwisseling naast het werk. Thanx friends!

Gedurende mijn aanstelling als arts-onderzoeker kreeg ik de gelegenheid om klinische

ervaring op te doen als arts-assistent gynaecologie en verloskunde in het Martini Ziekenhuis

bij opleider dr. W.F.A. Mensink. Ik wil hierbij iedereen met wie ik toen heb gewerkt

bedanken voor de zeer leerzame tijd. Mede vanwege die ervaring ben ik nu in opleiding tot

gynaecoloog. Hierbij wil ik mijn collega’s van de gynaecologie en verloskunde in het Groene

Hart Ziekenhuis en in het Leids Universitair Medisch Centrum bedanken voor hun steun,

geduld en betrokkenheid bij mijn promotie.

“Er gaat niets boven Groningen!” heeft na 3 jaar een heel nieuwe betekenis voor mij

gekregen. De reis naar Groningen heeft mij gesterkt als mens en als arts. Het gaf mij een

rugzak vol mooie herinneringen, ervaringen, wijsheden en heel veel liefde.

149

Curriculum vitae

Curriculum vitae |

Fleur Elise Marie Rijcken werd geboren in Paramaribo, Surimane, op 17 mei 1975, als

dochter van Marjolein Rijcken-Ooms en Guy Rijcken. Zij groeide op in Afrika en Zuid-

Amerika. Van 1988 tot 1992 ging zij naar de International School of Kenya waaraan zij in mei

1992 haar Internationaal Baccalaureaat behaalde. Aansluitend begon zij haar studie

Geneeskunde aan de Katholieke Universiteit van Leuven in België en zette deze voort aan de

(Rijks) Universiteit van Leiden. Gedurende haar studie deed zij een epidemiologische studie

naar Oesophagostomum bifurcum en hookworm infecties in Togo, West Afrika.

Na het behalen van het artsexamen in december 1999 verhuisde zij naar het noorden van

Nederland om te werken als arts-onderzoeker bij de afdeling Maag-, Darm- en Leverziekten

in het Universitair Medisch Centrum Groningen. Het wetenschappelijk onderzoek resulteerde

in dit proefschrift, onder leiding van Prof. dr. J.H. Kleibeuker, Prof. dr. H Hollema en Prof. dr.

A.G.J. van der Zee. Gedurende haar promotie onderzoek was zij enkele maanden arts-

assistent Gynaecologie and Verloskunde in het Martini Ziekenhuis. Op 1 juli 2003 werd

gestart met de opleiding Gynaecologie en Verloskunde in het Groene Hart Ziekenhuis te

Gouda (opleider J.C.M van Huisseling). Sinds 1 juli 2005 volgt ze het academisch deel van de

opleiding in het Leids Universitair Medisch Centrum te Leiden (opleider Prof. dr. H.H.H.

Kanhai en Prof. dr. G.G. Kenter).

153

|

Publicaties

Publicaties |

1. Koornstra JJ, Rijcken FEM, Oldenhuis CN, Zwart N, van der Sluis T, Hollema H, deVries EG, Keller JJ, Offerhaus JA, Giardiello FM, Kleibeuker JH. Sulindac inhibits beta-catenin expression in normal-appearing colon of hereditary nonpolyposis colorectal cancer and familial adenomatous polyposis patients. Cancer Epidemiol Biomarkers Prev. 2005;14:1608-12.

2. Koornstra JJ, Jalving M, Rijcken FEM, Westra J, Zwart N, Hollema H, de Vries EG, Hofstra RW, Plukker JT, de Jong S, Kleibeuker JH. Expression of tumornecrosis factor-related apoptosis-inducing ligand death receptors in sporadic and hereditary colorectal tumours: potential targets for apoptosis induction. Eur J Cancer. 2005;41:1195-202.

3. Koornstra JJ, Rijcken FEM, De Jong S, Hollema H, de Vries EG, Kleibeuker JH. Assessment of apoptosis by M30 immunoreactivity and the correlation with morphological criteria in normal colorectal mucosa, adenomas and carcinomas. Histopathology. 2004;44:9-17.

4. Rijcken FEM, van der Sluis T, Hollema H, Kleibeuker JH. Hyperplastic polyps in hereditary nonpolyposis colorectal cancer. Am J Gastroenterol. 2003;98:2306-11.

5. Rijcken FEM, Mourits MJ, Kleibeuker JH, Hollema H, van der Zee AG. Gynecologic screening in hereditary nonpolyposis colorectal cancer. Gynecol Oncol. 2003;91:74-80.

6. Koornstra JJ, Kleibeuker JH, van Geelen CM, Rijcken FEM, Hollema H, de Vries EG, de Jong S. Expression of TRAIL (TNF-related apoptosis-inducing ligand) and its receptors in normal colonic mucosa, adenomas, and carcinomas. J Pathol. 2003;200:327-35.

7. Scheenstra R, Rijcken FEM, Koornstra JJ, Hollema H, Fodde R, Menko FH, Sijmons RH, Bijleveld CM, Kleibeuker JH. Rapidly progressive adenomatous polyposis in a patient with germline mutations in both the APC and MLH1 genes: the worst of two worlds. Gut. 2003;52:898-9.

8. Rijcken FEM, Hollema H, Kleibeuker JH. Proximal adenomas in hereditary non-polyposis colorectal cancer are prone to rapid malignant transformation. Gut. 2002;50:382-6.

9. Pit DS, Rijcken FEM, Raspoort EC, Baeta SM, Polderman AM. Geographic distribution and epidemiology of Oesophagostomum bifurcum and hookworm infections in humans in Togo. Am J Trop Med Hyg. 1999;61:951-5.

157