1774.full

Ethnicity and Risk for Colorectal Cancers ShowingSomatic BRAF V600E Mutation or CpG IslandMethylator Phenotype

Dallas R. English,1,4 Joanne P. Young,6,7 Julie A. Simpson,1,4 Mark A. Jenkins,1

Melissa C. Southey,2 Michael D. Walsh,6 Daniel D. Buchanan,6 Melissa A. Barker,6

Andrew M. Haydon,3 Simon G. Royce,5 Aedan Roberts,6 Susan Parry,8

John L. Hopper,1 Jeremy J. Jass,9 and Graham G. Giles1,4

1Centre for Molecular, Environmental, Genetic and Analytic Epidemiology, School of Population Health, and 2Genetic EpidemiologyLaboratory, Department of Pathology, The University of Melbourne; 3Department of Medical Oncology, Alfred Hospital, Melbourne,Victoria, Australia; 4Cancer Epidemiology Centre, The Cancer Council Victoria, Carlton, Victoria, Australia; 5Murdoch Children’sResearch Institute, Parkville, Victoria, Australia; 6Familial Cancer Laboratory, Queensland Institute of Medical Research;7Faculty of Health Sciences, University of Queensland, Brisbane, Queensland, Australia; 8Department of Gastroenterologyand Hepatology, Middlemore Hospital, Auckland, New Zealand; and 9Department of Cellular Pathology, St. Mark’sHospital and Imperial College, Harrow, Middlesex, United Kingdom

Abstract

Colorectal cancers arising from serrated polyps arecharacterized by the CpG island methylator phenotype(CIMP) and somatic mutation (V600E) in the BRAFproto-oncogene. Few epidemiologic studies haveinvestigated risk factors for these tumors. We con-ducted a cohort study of 41,328 residents of Melbourne,Australia that included 9,939 participants of southernEuropean origin and 31,389 of Anglo-Celtic origin.Colorectal adenocarcinomas were identified frompopulation-based cancer registries. BRAF V600E muta-tion in tumors was determined using a PCR-basedallelic discrimination method. Tumors were classifiedas CIMP positive when at least three of five markers(RUNX3, CACNA1G, SOCS1, NEUROG1 , and IGF2)were methylated according to MethyLight analysis.Hazard ratios (HR) and 95% confidence intervals(95% CI) were estimated by Cox regression with

adjustment for risk factors for colorectal cancer. Duringfollow-up, 718 participants were diagnosed withcolorectal cancer. CIMP assays were done for 579 andBRAF V600E mutation testing for 582. After adjustmentfor other risk factors, when compared with people ofAnglo-Celtic origin, those of southern European originhad lower incidence of colorectal cancer that hadCIMP (HR, 0.32; 95% CI, 0.16-0.67) or BRAF mutations(HR, 0.30; 95% CI, 0.16-0.58) but similar incidenceof colorectal cancer without CIMP (HR, 0.86; 95% CI,0.70-1.05) or BRAF (HR, 0.90; 95% CI, 0.74-1.11). Peopleof southern European origin had lower risk ofcolorectal cancers with CIMP and BRAF mutation thanpeople of Anglo-Celtic origin, which may in part bedue to genetic factors that are less common in people ofsouthern European origin. (Cancer Epidemiol Bio-markers Prev 2008;17(7):1774–80)

Introduction

Colorectal cancer is a heterogeneous disorder for whichmolecular pathologic examination has identified distinctsubsets that evolve through different pathways. Onesubset that arises from advanced serrated polyps ratherthan from adenomas has been identified (1-9). Thesetumors are characterized by activating somatic mutationsin the BRAF proto-oncogene (10), predominantly V600E,and widespread hypermethylation of CpG islands,especially in the promoters of genes, referred to as theCpG island methylator phenotype (CIMP; refs. 11, 12).BRAF mutation and CIMP frequently co-occur in

serrated polyps and their malignant counterparts(11-17). In addition, BRAF mutation has been identifiedin aberrant crypt foci, which show evidence of epithelialserration (18), suggesting that these changes are estab-lished early in colorectal carcinogenesis and inducecommitment to the serrated neoplasia pathway (19).Different molecular subsets of colorectal cancer may

have different environmental, genetic, and lifestylerisk factors, and investigation of possible risk factorsseparately for different molecular subtypes may leadto a better understanding of how to prevent disease.For example, in Australia, men have higher incidenceof colorectal cancer than women (20), but CIMP andBRAF-related tumors are more common in women (21).Smoking was associated with risk of CIMP andBRAF-related tumors in a case-control study but notwith other colorectal cancers (22), although there were noclear differences in dietary risk factors between thesubtypes (23).The Melbourne Collaborative Cohort Study consists

mostly of people of British or Irish descent (usuallyreferred to in Australia as of ‘‘Anglo-Celtic’’ descent) and

Cancer Epidemiol Biomarkers Prev 2008;17(7). July 2008

Received 1/29/08; revised 4/3/08; accepted 4/22/08.

Grant support: National Health and Medical Research Council grants 209057, 251533,299955, and 442916; National Health and Medical Research Council AustraliaFellowship 466668 (J.L. Hopper); and Cancer Council Queensland Senior ResearchFellowship (J.P. Young). Cohort recruitment was partly funded by VicHealth.

Requests for reprints: Dallas R. English, Centre for Molecular, Environmental,Genetic and Analytic Epidemiology, School of Population Health, The University ofMelbourne, Level 1, 723 Swanston Street, Parkville, Victoria 3010, Australia.Phone: 61-3-8344-0742; Fax: 61-3-9349-5815. E-mail: [email protected]

Copyright D 2008 American Association for Cancer Research.

doi:10.1158/1055-9965.EPI-08-0091

1774

migrants to Australia from Greece or Italy (southernEuropeans). The migrants were included to increase theheterogeneity of the cohort with respect to lifestyle(especially diet), genes, and rates of disease outcomes.When southern European migrants arrive in Australia,they initially have substantially lower mortality fromcolorectal cancer, but the differences diminish over time(24-26), suggesting that environmental factors aftermigration might be important risk factors.Within the Melbourne Collaborative Cohort Study, we

determined CIMP and BRAF mutation status for colorec-tal cancers diagnosed during follow-up with a view toidentifying those arising from the serrated pathway anddetermining whether their risk factors differed from themore common cancers that arise from adenomas. Here, wereport on the incidence of colorectal cancer with CIMP andBRAF mutation in relation to ethnic origin and gender.

Materials and Methods

Subjects. The Melbourne Collaborative Cohort Studyis a prospective cohort study of 41,528 participants, ages27 to 75 years at recruitment from 1990 to 1994 (almost allwere ages 40-69 years), and includes 5,425 migrants fromItaly and 4,535 from Greece. For this analysis, 200participants who had a colorectal cancer diagnosedbefore baseline were excluded, leaving 41,328.Subjects were recruited via the electoral rolls (regis-

tration to vote is compulsory for adults in Australia),advertisements, and community announcements in localmedia (e.g., television, radio, and newspapers). Compre-hensive lists of Italian and Greek surnames were alsoused to target southern European migrants listed in thephone books and on electoral rolls. The study protocolwas approved by the Cancer Council Victoria’s HumanResearch Ethics Committee. Participants gave writtenconsent for participation and for the investigators toobtain their medical records.

Baseline Data Collection. A structured interviewschedule was used to obtain information on potential riskfactors including education, country of birth, alcoholconsumption, smoking habits, current physical activityduring leisure time, education, and, for women, reproduc-tive history, menopausal status including age atmenopause,and use of hormone replacement therapy. Information oncurrent dietwas obtained fromadietary questionnaire thatcontained a 121-item food frequency questionnaire thatwas developed for this study (27). Height, weight, andwaist and hip circumferences were measured.

Cohort Follow-up and Case Ascertainment. Caseswere participants who had a first diagnosis of invasivecancer of the colon or rectum during follow-up toDecember 31, 2004 and were identified by linkage topopulation-based cancer registries in all Australianstates. Addresses and vital status of the subjects weredetermined by record linkage to electoral rolls, theNational Death Index, Victorian death records, fromelectronic phone books, and from responses to mailedquestionnaires and newsletters.

Molecular Pathology. Archival tumor tissue wassought for all primary, histopathologically confirmedadenocarcinomas diagnosed in Victoria. Diagnosis wasverified and pathology was reviewed by an experienced

histopathologist (J.R.J.). Immunohistochemistry for DNAmismatch repair genes MLH1, MSH2, MSH6 , and PMS2was done as reported by Lindor et al. (28). Microsatelliteinstability (MSI) was examined using 10 microsatellitemarkers (28). BRAF V600Emutation analysis was done bya real-time PCR-based allelic discrimination method (29).CIMP status and MLH1 methylation was determined byMethyLight analysis of five markers (RUNX3, CACNA1G,SOCS1, NEUROG1 , and IGF2), and MLH1 , respectively(21). Exon 1 of KRAS was analyzed by direct sequencing(21). Positive methylation measurements were deter-mined by the presence of amplification product for thefive CIMP markers andMLH1 . Samples were considerednegative for methylation if no marker-specific probeamplification was seen together with the Alu referenceprobe displaying amplification with a Ct value <25,indicating the initial amount of bisulfite convertedtemplate was significant enough to allow the amplifica-tion and detection of the methylation markers. Tumorswere classified as CIMP positive when at least threemarkers were positive for methylation. For participantswho had both colon and rectal tumors, results for colontumors are reported here.

Statistical Analysis. Ethnicity was defined basedon country of birth and was classified into two groups:southern European (born in Italy or Greece) andAnglo-Celtic (born in Australia, New Zealand, UnitedKingdom, or Ireland).Cox regression, with age as the time scale, was used to

estimate hazard ratios (HR), 95% confidence intervals(95% CI), and P values (30). To estimate HR separatelyfor molecular subtypes and to test their difference, Coxmodels based on competing risks were fitted using a dataduplication method (31). Calculation of person-timebegan at baseline and ended at the earliest of the dateof diagnosis of colorectal cancer, date of diagnosis ofcancer of unknown primary site, date of death, date lastknown to be in Australia, or December 31, 2004. Testsbased on Schoenfeld residuals and graphical methodsusing Kaplan-Meier curves (32) showed no evidence thatproportional hazard assumptions were violated for anyanalyses.Level of education, smoking status (never, former, cur-

rent), alcohol intake (g/d), waist circumference, physicalactivity, frequency of red meat consumption, intakes ofcereal fiber, calcium, dietary folate and energy, multivi-tamin use, calcium supplements, and, for women,hormone replacement therapy use were all consideredas potential confounders. Only waist circumferenceand alcohol intake changed the HR for sex or ethnicorigin by >10% and were retained in the final models.Differences between characteristics of tumors for

which molecular measurements were done and thosefor which no measurements were made were testedusing m2 tests for categorical variables and the Mann-Whitney rank sum test for continuous variables. Allstatistical tests are two sided, with P < 0.05 consideredstatistically significant. All statistical analyses were donein Stata version 9.2 (StataCorp).

Results

Baseline characteristics of the cohort by ethnic origin areshown in Table 1. Southern Europeans had larger mean

Cancer Epidemiology, Biomarkers & Prevention


1775

waist circumference than Anglo-Celts and were morelikely to be lifetime abstainers from alcohol, to havelower educational attainment, less likely to engage inrecreational physical activity, were more likely to eat redmeat frequently, had lower intakes of cereal fiber,calcium, and folate, and were less likely to use multi-vitamins, calcium supplements, or hormone replacementtherapy. Most of the migrants had lived in Australia for>30 years (median, 33 years; interquartile range, 28-37years).During an average of 11 years of follow-up per person

to December 31, 2004, 55 participants were known tohave left Australia and 718 were diagnosed withcolorectal cancer, including 449 with colon tumors (236proximal and 173 distal), 269 with rectal tumors (threesubjects had both colon and rectal tumors), and 40 withcolorectal cancers of unspecified site.Archival tissue was obtained for 656 cases (92%); of the

remaining cases, 9 were diagnosed in states other thanVictoria, and for 53, the material could not be found orwas not provided by the pathology laboratory. Molecular

analysis of archival tumor tissue was done for 585 casesfor which CIMP status was determined for 579 and BRAFV600E mutation status for 582. Of those patients forwhom no molecular analysis was attempted, 10 did nothave a histopathologic diagnosis, and on review of thediagnostic slides, 25 were found to have metastaticdisease, 11 were classified as having adenomas only, and25 did not have tissue suitable for molecular analysis.A slightly lower proportion of tumors from men than

from women had no molecular measurements (79%versus 84%; P = 0.05), but there was little differencein the proportions with molecular measurements byethnicity (P = 0.8), age at diagnosis (P = 0.4), waistcircumference (P = 0.1), smoking status (P = 0.3), alcoholconsumption (P = 0.6), education (P = 0.4), physicalactivity (P = 0.3), red meat consumption (P = 0.8), or forthe site (rectum, proximal, or distal colon; P = 0.7).CIMP and BRAF mutation status were strongly

associated; 54 (74%) of 73 CIMP-positive tumorshad BRAF mutations compared with 37 (7%) of506 CIMP-negative tumors, giving an odds ratio of 36

Table 1. Baseline demographic, dietary, and lifestyle characteristics by ethnic origin

Ethnic origin

Anglo-Celtic Southern European

Total 31,389 9,939Age (y)* 55.1 (8.9) 55.9 (7.8)Waist circumference (cm)* 84 (13) 90 (12)Females 18,923 (60) 5,459 (55)Smoking status

c(%)

Never 17,878 (57) 5,855 (59)Former 10,198 (32) 2,709 (27)Current 3,310 (11) 1,368 (14)Alcohol intake

c

Lifetime abstainer 7,694 (25) 4,138 (42)Former drinker 1,280 (4) 369 (4)1-19 g/d 15,612 (50) 3,351 (34)20-39 g/d 4,326 (14) 1,210 (12)40+ g/d 2,452 (8) 857 (9)Education

c

Primary school 1,369 (4) 6,653 (67)Some secondary school 13,936 (44) 1,847 (19)Completed secondary school 7,437 (24) 1,093 (11)Degree/diploma 8,644 (28) 340 (3)Recreational physical activity

c

None 5,638 (18) 3,551 (36)Low 6,079 (19) 2,206 (22)Medium 11,461 (37) 3,238 (33)High 8,211 (26) 940 (9)Multivitamin use

c6,052 (19) 652 (7)

Calcium supplement usec

3,888 (12) 497 (5)Hormone replacement therapy use

c(women only)

Never 13,353 (71) 4,667 (87)Former 1,762 (9) 329 (6)Current 3,717 (20) 341 (6)Red meat consumption

c(times a week)

0-4 7,642 (24) 2,334 (24)5-6 7,363 (24) 2,125 (21)7-9 8,679 (28) 2,357 (24)10+ 7,678 (24) 3,104 (31)Energy intake (kJ/d)

b8,899 [7,090-11,101] 8,092 [6,287-10,414]

Cereal fiber intake (g/d)b

10.5 [7.5-14.5] 8.1 [5.5-11.5]Calcium intake (mg/d)

b826 [633-1,064] 680 [509-908]

Folate intake (mg/d)b

319 [249-405] 255 [187-344]

* Mean (SD).cNumbers do not always add up to total numbers due to missing data.bMedian [interquartile range].

Ethnicity and Risk for Colorectal Cancer


1776

(95% CI, 19-71; P < 0.001). Both features were morecommon in tumors from females than from males(P < 0.005) and in tumors of the right colon (P < 0.005),were positively associated with MSI-high tumor status(P < 0.001) and MLH1 methylation (P < 0.001), and wereinversely associated with KRAS mutations (P < 0.001;Table 2).Table 3 shows HR by sex and ethnic origin. Overall,

women had a slightly lower incidence rate of colorectalcancer than did men (HR, 0.88; 95% CI, 0.74-1.05) andalso had slightly lower incidence of tumors that wereCIMP negative (HR, 0.84; 95% CI, 0.69-1.02) or that didnot have BRAF mutations (HR, 0.83; 95% CI, 0.68-1.01)but had higher incidence rates for tumors that wereCIMP positive or that had BRAF mutations (CIMPpositive: HR, 1.83; 95% CI, 1.11-3.01; BRAF mutation:HR, 1.67; 95% CI, 1.08-2.58). For both molecular features,the differences between the HR for the two subtypeswere significant.After adjustment for sex, alcohol intake, and waist

circumference, people of southern European origin hadf20% lower overall incidence of colorectal cancer thanthose of Anglo-Celtic origin (HR, 0.78; 95% CI, 0.65-0.93),14% lower incidence rate of CIMP-negative tumors (HR,0.86; 95% CI, 0.70-1.05) and 10% lower incidence oftumors without BRAF mutations (HR, 0.90; 95% CI, 0.74-1.11). In contrast, they had about one-third the rate ofcolorectal cancers that were CIMP positive (HR, 0.32;95% CI, 0.16-0.67) or that had BRAF mutations (HR, 0.30;95% CI, 0.16-0.58). For both molecular features, the

differences between the HR for the two subtypes weresignificant (Table 3).The HR for ethnicity in relation to tumors of the right

colon were similar to those for all sites, although the95% CI were wider (data not shown); P values for thedifferences between HR for the two subtypes were 0.07for CIMP status and 0.008 for BRAF mutation status.

Discussion

Overall, people of southern European origin had f20%lower incidence rate of colorectal cancer than did peopleof Anglo-Celtic origin. This was largely due to theirmuch lower incidence rate of tumors with BRAFmutations or the CIMP, because both ethnic groups hadsimilar incidence rates of other colorectal cancer. Likeothers, we also found that BRAF mutation and CIMPwere strongly associated with each other and were morecommon in tumors from females and the proximal colon(10, 21). In addition, also as reported previously, BRAFmutation was positively associated with MSI-high statusand MLH1 methylation and inversely associated with thepresence of KRAS in a tumor (33).Our study has several strengths and limitations. We

had almost complete ascertainment of colorectal cancerand little loss to follow-up but were unable to obtainarchival tissue from the primary lesion for 19% of cases.However, this is unlikely to have biased the observedassociations because the proportions of cases with no

Table 2. Characteristics of participants by colorectal cancer status during follow-up

Colorectal cancer diagnosed during follow-up

No CIMP BRAF V600E mutation status

Negative Positive Negative Positive

Total 40,610 506 73 487 95Ethnic origin (%)Anglo-Celtic 76 74 89 73 89Southern European 24 26 11 27 11Sex (%)Female 59 48 67 47 65Male 41 52 33 53 35Alcohol intake (%)Lifetime abstainer 29 24 30 23 36Former drinker 4 3 7 2 8Low 46 42 36 42 34Medium 13 17 18 18 12High 8 15 10 15 10Waist circumference (cm)*Male 93.4 (10.3) 96.3 (9.5) 95.2 (7.6) 96.2 (9.4) 95.6 (9.5)Female 79.9 (12.0) 82.4 (12.1) 83.5 (12.4) 82.7 (12.3) 82.7 (11.8)Age at diagnosis (y)* 66.8 (7.9) 68.3 (6.8) 66.7 (8.0) 68.5 (6.6)Site (%)Proximal colon 26 84 28 62Distal colon 26 8 25 18Rectum 44 7 43 16Unknown 4 1 4 4MSI statusHigh 5 49 6 38Intermediate 5 0 4 4Low 14 7 13 13Stable 76 44 77 45

KRAS mutation 71 15 31 4MLH1 methylation 4 62 4 47

*Mean (SD).



1777

suitable tissue varied little by ethnicity or sex. Ourmeasurement of CIMP used a panel of recently reportedmarkers with high sensitivity and specificity, and theassociations between CIMP and BRAF mutation status,MSI status, MLH1 methylation, KRAS mutations, andsubsite are supported by earlier studies, indicating thatour measurements of CIMP and BRAF mutation status arein general agreementwith those of other investigators (21).Despite residing in Australia for many years, the

migrants maintained substantial differences in dietarypatterns from those born in Australia (34) and alsodiffered from the Australian-born participants in theirsmoking habits, alcohol intake, waist circumference, andeducation. Nevertheless, control for these risk factors hadlittle effect on the results. It is unlikely that familyhistory, which was not recorded, could account for astrong inverse association between southern Europeandescent and risk of tumors with CIMP or BRAF mutationstatus but little or no association with risk of tumorsnot displaying these features. We cannot rule out chanceas an explanation for the ethnic differences in risk ofBRAF-related colorectal cancer.The HR for all colorectal cancer for the southern

European migrants (0.78) is consistent with descriptivestudies of mortality. Compared with Australian-bornresidents, migrants who had lived in Australia for at least30 years had mortality rate ratios between 0.7 (25) and0.8 (95% CI, 0.7-1.0; ref. 26). This suggests that althoughmigrants’ risk for the more common form of colorectalcancer converges to that of the Australian born withincreasing duration of residence, the same is not true fortumors with BRAF mutation or the CIMP. Risk factorsthat act primarily before migration, or genetic factors(35, 36), could explain the lack of convergence for thesetumor types.

Evidence for a genetic predisposition for tumorsbearing BRAF mutations comes from a case-controlstudy of colorectal cancer, where microsatellite-stabletumors with BRAF mutation were over four times morelikely than tumors without BRAF mutation to beassociated with a family history of colorectal cancer(37). In addition, familial syndromes associated withBRAF mutation-bearing tumors have been describedfrom Australia and Sweden (29, 38). There is also indirectevidence regarding an association between ethnicity andthe risk of colorectal cancer with CIMP and BRAFmutation from a study of hyperplastic polyposis syn-drome, which is a rare condition characterized bymultiple serrated polyps and tumors with CIMP andBRAF mutation. In a case series from a gastroenterologyclinic in New Zealand, all 24 cases were people ofEuropean origin, whereas only 46% of patients attendingthe clinic were of European descent (39). Thus, thepossibility exists that a sequence variant in the Anglo-Celtic population may underlie the development ofCIMP-related colorectal cancer either directly or throughmodulation of an environmental influence; however, atpresent, none has been definitively identified.Two studies have examined associations between

common genetic variants and CIMP status for patientswith colorectal cancer, including genes in the folatepathway. van Rijnsoever et al. found that overallthere was no association between a variant in the5,10-methylenetetrahydrofolate reductase gene (C677T;rs1801133) and type of colorectal cancer, but that femaleshad a higher proportion of CIMP-positive tumors(40). CIMP was defined as methylation of two of threegenes (CDKN2A, MDR1 , andMINT2). Paz et al. observedthat the mean proportion of methylation in six genes(CDKN2A, p14ARF, MGMT, APC, LKB1 , and MLH1)

Table 3. HR for sex and ethnic origin according to type of colorectal cancer

Type of colorectal cancer Sex Ethnic origin

Male Female P* Anglo-Celtic Southern European P*

Person-years 187,654 277,942 346,149 119,447All colorectal cancerCases 371 347 545 173Partially adjusted

c1 0.65 (0.57-0.80) 1 0.84 (0.70-0.99)

Fully adjustedb

1 0.88 (0.74-1.05) 1 0.78 (0.65-0.93)CIMP statusPositiveCases 24 49 65 8Partially adjusted

c1 1.37 (0.84-2.24) 1 0.35 (0.17-0.72)

Fully adjustedb

1 1.83 (1.11-3.01) 1 0.32 (0.16-0.67)NegativeCases 266 240 374 132Partially adjusted

c1 0.63 (0.53-0.75) 1 0.92 (0.76-1.12)

Fully adjustedb

1 0.84 (0.69-1.02) 0.003 1 0.86 (0.70-1.05) 0.011BRAF mutation statusPositiveCases 33 62 85 10Partially adjusted

c1 1.26 (0.83-1.93) 1 0.33 (0.17-0.62)

Fully adjustedb

1 1.67 (1.08-2.58) 1 0.30 (0.16-0.58)NegativeCases 257 230 356 131Partially adjusted

c1 0.63 (0.52-0.75) 1 0.96 (0.79-1.18)

Fully adjustedb

1 0.83 (0.68-1.01) 0.003 1 0.90 (0.74- 1.11) 0.002

* P value for test that fully adjusted HR for two tumor subtypes are equal.cHR (95% CI) adjusted for both variables in the table and with age as the time scale.bHR (95% CI) adjusted for both variables in the table plus waist circumference and alcohol intake and with age as the time scale.



1778

was higher for colorectal cancer patients with a poly-morphism in the gene coding 5-methyltetrahydrofo-late-homocysteine methyltransferase, also known asmethionine synthase reductase (41). Other genes in-volved in the folate pathway, those encoding including5-methyltetrahydrofolate-homocysteine methyltransfer-ase and thymidylate synthetase, have been shown tobe associated with colorectal cancer in general (42).Colorectal cancers with CIMP and BRAF mutation

arise in a subset of serrated polyps called sessile serratedadenomas (3). Given our findings, it is likely that sessileserrated adenomas, lesions with high levels of BRAFmutation (10), may also be more common in peopleof Anglo-Celtic descent than in people of southernEuropean descent. The prevalence of these lesions inpatients undergoing colonoscopy ranges from 4% to 9%(43, 44) and is associated with increased polyp burden(44). CIMP tumors are frequently located in the proximalcolon (40), and the detection of their sessile precursorlesions may prove difficult (45). CIMP-related colorectaltumors have been shown to account for almost allnonfamilial MSI-high colorectal cancers (11) and areassociated with improved survival for this subgroup (46).In contrast, tumors that show somatic BRAF mutation inthe absence of a MSI-high phenotype may have arelatively poor outcome (37).In summary, we have confirmed the findings of

Weisenberger et al. (21) that their five-marker paneleffectively identifies CIMP colorectal cancer and itsassociation with BRAF mutation, female sex, proximallocation in the colon, MSI-high tumor status, and MLH1methylation. Further, we report the novel finding thatindividuals with Anglo-Celtic ancestry were more likelyto develop CIMP-related colorectal cancer than wereindividuals of Italian or Greek origin, and this differencewas unlikely to be due to environmental factorscommonly associated with colorectal cancer.

Disclosure of Potential Conflicts of InterestNo potential conflicts of interest were disclosed.

AcknowledgmentsThe costs of publication of this article were defrayed in part bythe payment of page charges. This article must therefore behereby marked advertisement in accordance with 18 U.S.C.Section 1734 solely to indicate this fact.We thank Peter Laird for extensive advice in the perfor-

mance of MethyLight assays, Veronika Gavdik for laboratoryassistance, Lesley Jaskowski for the preparation of data for thisarticle, and the pathology laboratories that provided archivedtissue for the project.This study was made possible by the contribution of many

people, including the original investigators, the ProgramManager, Georgina Marr, and the diligent team who recruitedthe participants and who continue working on follow-up. Theauthors express their gratitude to the many thousands ofMelbourne residents who continue to participate in the study.

References1. Biemer-Huttmann AE, Walsh MD, McGuckin MA, et al. Immuno-

histochemical staining patterns of MUC1, MUC2, MUC4, andMUC5AC mucins in hyperplastic polyps, serrated adenomas, andtraditional adenomas of the colorectum. J Histochem Cytochem 1999;47:1039–48.

2. Biemer-Huttmann AE, Walsh MD, McGuckin MA, et al. Mucin coreprotein expression in colorectal cancers with high levels of micro-satellite instability indicates a novel pathway of morphogenesis. ClinCancer Res 2000;6:1909–16.

3. Goldstein NS. Clinical significance of (sessile) serrated adenomas:another piece of the puzzle. Am J Clin Pathol 2005;123:329– 30.

4. Goldstein NS, Bhanot P, Odish E, Hunter S. Hyperplastic-like colonpolyps that preceded microsatellite-unstable adenocarcinomas. Am JClin Pathol 2003;119:778– 96.

5. Jass JR. Serrated route to colorectal cancer: back street or superhighway? J Pathol 2001;193:283 –5.

6. Jass JR. Hyperplastic polyps of the colorectum—innocent or guilty?Dis Colon Rectum 2001;44:163–6.

7. Jass JR. Hyperplastic-like polyps as precursors of microsatellite-unstable colorectal cancer. Am J Clin Pathol 2003;119:773– 5.

8. Jass JR, Iino H, Ruszkiewicz A, et al. Neoplastic progression occursthrough mutator pathways in hyperplastic polyposis of the color-ectum. Gut 2000;47:43–9.

9. Jass JR, Young J, Leggett BA. Hyperplastic polyps and DNAmicrosatellite unstable cancers of the colorectum. Histopathology2000;37:295 –301.

10. Kambara T, Simms LA, Whitehall VL, et al. BRAF mutation isassociated with DNA methylation in serrated polyps and cancers ofthe colorectum. Gut 2004;53:1137– 44.

11. Toyota M, Ahuja N, Ohe-Toyota M, Herman JG, Baylin SB, Issa JP.CpG island methylator phenotype in colorectal cancer. Proc NatlAcad Sci U S A 1999;96:8681 –6.

12. Toyota M, Ho C, Ahuja N, et al. Identification of differentiallymethylated sequences in colorectal cancer by methylated CpG islandamplification. Cancer Res 1999;59:2307– 12.

13. Beach R, Chan AO, Wu TT, et al. BRAF mutations in aberrant cryptfoci and hyperplastic polyposis. Am J Pathol 2005;166:1069– 75.

14. Chan AO, Issa JP, Morris JS, Hamilton SR, Rashid A. ConcordantCpG island methylation in hyperplastic polyposis. Am J Pathol 2002;160:529 –36.

15. O’Brien MJ, Yang S, Clebanoff JL, et al. Hyperplastic (serrated)polyps of the colorectum: relationship of CpG island methylatorphenotype and K-ras mutation to location and histologic subtype.Am J Surg Pathol 2004;28:423 –34.

16. O’Brien MJ, Yang S, Mack C, et al. Comparison of microsatelliteinstability, CpG island methylation phenotype, BRAF and KRASstatus in serrated polyps and traditional adenomas indicates separatepathways to distinct colorectal carcinoma end points. Am J SurgPathol 2006;30:1491– 501.

17. Yang S, Farraye FA, Mack C, Posnik O, O’Brien MJ. BRAF and KRASMutations in hyperplastic polyps and serrated adenomas of thecolorectum: relationship to histology and CpG island methylationstatus. Am J Surg Pathol 2004;28:1452 –9.

18. Rosenberg DW, Yang S, Pleau DC, et al. Mutations in BRAF andKRAS differentially distinguish serrated versus non-serrated hyper-plastic aberrant crypt foci in humans. Cancer Res 2007;67:3551–4.

19. Jass JR. Classification of colorectal cancer based on correlation ofclinical, morphological and molecular features. Histopathology 2007;50:113–30.

20. Australian Institute of Health and Welfare and AustralasianAssociation of Cancer Registries. Cancer in Australia: an overview,2006. Cancer series no. 37. Cat. no. CAN 32. Canberra: AustralianInstitute of Health and Welfare; 2007.

21. Weisenberger DJ, Siegmund KD, Campan M, et al. CpG islandmethylator phenotype underlies sporadic microsatellite instabilityand is tightly associated with BRAF mutation in colorectal cancer.Nat Genet 2006;38:787–93.

22. Samowitz WS, Albertsen H, Sweeney C, et al. Association ofsmoking, CpG island methylator phenotype, and V600E BRAFmutations in colon cancer. J Natl Cancer Inst 2006;98:1731 –8.

23. Slattery ML, Curtin K, Sweeney C, et al. Diet and lifestyle factorassociations with CpG island methylator phenotype and BRAFmutations in colon cancer. Int J Cancer 2007;120:656 –63.

24. McMichael AJ, McCall MG, Hartshorne JM, Woodings TL. Patternsof gastro-intestinal cancer in European migrants to Australia: the roleof dietary change. Int J Cancer 1980;25:431–7.

25. Balzi D, Khlat M, Matos E. Cancer in Italian migrant populations.Australia: mortality study. IARC Sci Publ 1993;123:125 –37.

26. McCredie M, Williams S, Coates M. Cancer mortality in migrantsfrom the British Isles and continental Europe to New South Wales,Australia, 1975-1995. Int J Cancer 1999;83:179– 85.

27. Ireland P, Jolley D, Giles GG, et al. Development of the MelbourneFFQ: a food frequency questionnaire for use in an Australianprospective study involving an ethnically diverse cohort. Asia Pac JClin Nutr 1994;3:19 –31.

28. Lindor NM, Burgart LJ, Leontovich O, et al. Immunohistochemistry



1779

versus microsatellite instability testing in phenotyping colorectaltumors. J Clin Oncol 2002;20:1043 –8.

29. Young J, Barker MA, Simms LA, et al. Evidence for BRAFmutation and variable levels of microsatellite instability in asyndrome of familial colorectal cancer. Clin Gastroenterol Hepatol2005;3:254 –63.

30. Korn EL, Graubard BI, Midthune D. Time-to-event analysis oflongitudinal follow-up of a survey: choice of the time-scale. Am JEpidemiol 1997;145:72–80.

31. Lunn M, McNeil D. Applying Cox regression to competing risks.Biometrics 1995;51:524–32.

32. Therneau TM, Grambsch PM. Modeling survival data: extendingthe Cox model. Statistics for biology and health. New York:Springer; 2000.

33. Yuen ST, Davies H, Chan TL, et al. Similarity of the phenotypicpatterns associated with BRAF and KRAS mutations in colorectalneoplasia. Cancer Res 2002;62:6451 –5.

34. Harriss LR, English DR, Powles J, et al. Dietary patterns andcardiovascular mortality in the Melbourne Collaborative CohortStudy. Am J Clin Nutr 2007;86:221–9.

35. Young J, Jass JR. The case for a genetic predisposition to serratedneoplasia in the colorectum: hypothesis and review of the literature.Cancer Epidemiol Biomarkers Prev 2006;15:1778 –84.

36. Young J, Jenkins M, Parry S, et al. Serrated pathway colorectal cancerin the population: genetic consideration. Gut 2007;56:1453–9.

37. Samowitz WS, Sweeney C, Herrick J, et al. Poor survival associatedwith the BRAF V600E mutation in microsatellite-stable colon cancers.Cancer Res 2005;65:6063–9.

38. Vandrovcova J, Lagerstedt-Robinsson K, Pahlman L, Lindblom A.

Somatic BRAF-V600E mutations in familial colorectal cancer. CancerEpidemiol Biomarkers Prev 2006;15:2270– 3.

39. Yeoman A, Young J, Arnold J, Jass J, Parry S. Hyperplastic polyposisin the New Zealand population: a condition associated withincreased colorectal cancer risk and European ancestry. N Z Med J2007;120:31 –39.

40. van Rijnsoever M, Grieu F, Elsaleh H, Joseph D, Iacopetta B.Characterisation of colorectal cancers showing hypermethylation atmultiple CpG islands. Gut 2002;51:797– 802.

41. Paz MF, Avila S, Fraga MF, et al. Germ-line variants in methyl-group metabolism genes and susceptibility to DNA methylation innormal tissues and human primary tumors. Cancer Res 2002;62:4519–24.

42. Ulrich CM, Curtin K, Potter JD, Bigler J, Caan B, Slattery ML.Polymorphisms in the reduced folate carrier, thymidylate synthase,or methionine synthase and risk of colon cancer. Cancer EpidemiolBiomarkers Prev 2005;14:2509 –16.

43. Higuchi T, Sugihara K, Jass JR. Demographic and pathologicalcharacteristics of serrated polyps of colorectum. Histopathology2005;47:32 –40.

44. Spring KJ, Zhao ZZ, Karamatic R, et al. High prevalence of sessileserrated adenomas with BRAF mutations: a prospective study ofpatients undergoing colonoscopy. Gastroenterology 2006;131:1400–7.

45. Farrar WD, Sawhney MS, Nelson DB, Lederle FA, Bond JH.Colorectal cancers found after a complete colonoscopy. Clin Gastro-enterol Hepatol 2006;4:1259–64.

46. Wright CM, Dent OF, Barker M, et al. Prognostic significance ofextensive microsatellite instability in sporadic clinicopathologicalstage C colorectal cancer. Br J Surg 2000;87:1197–202.



1780

Documents

1774.full