Common genetic variants in the 9p21 region and their associations with multiple tumours.

0.36 MB·English

by  Gu, F

FULL PAPER
British Journal of Cancer (2013) 108, 1378–1386 | doi: 10.1038/bjc.2013.7
Keywords: common genetic variants; CDKN2A; 9p21.3
Common genetic variants in the 9p21 region
and their associations with multiple tumours
F Gu1, R M Pfeiffer1, S Bhattacharjee1, S S Han1, P R Taylor1, S Berndt1, H Yang1, A J Sigurdson1, J Toro1,
L Mirabello1, M H Greene1, N D Freedman1, C C Abnet1, S M Dawsey1, N Hu1, Y-L Qiao2, T Ding3,
AVBrenner1,MGarcia-Closas1,4,RHayes1,LABrinton1,JLissowska5,NWentzensen1,CKratz1,LEMoore1,
R G Ziegler1, W-H Chow1, S A Savage1, L Burdette1, M Yeager1, S J Chanock1, N Chatterjee1, M A Tucker1,
A M Goldstein1 and X R Yang*,1
1Division of Cancer Epidemiology and Genetics, National Cancer Institute, Bethesda, MD 20852, USA; 2Cancer Institute and
Hospital,ChineseAcademyofMedicalSciences,Beijing100021,China;3ShanxiCancerHospital,Shanxi30013,China;4Divisionof
Genetics andEpidemiology, Institute of Cancer Research, SuttonSM2 5NG, UKand 5Department of CancerEpidemiology and
Prevention,the M.Sklodowska-Curie Cancer Center andInstitute of Oncology, 02-781Warsaw, Poland
Background: The chromosome9p21.3 region hasbeenimplicated inthepathogenesis ofmultiple cancers.
Methods:Wesystematicallyexaminedupto203taggingSNPsof22geneson9p21.3(19.9–32.8Mb)ineightcase–controlstudies:
thyroid cancer, endometrial cancer (EC), renal cell carcinoma, colorectal cancer (CRC), colorectal adenoma (CA), oesophageal
squamouscellcarcinoma(ESCC),gastriccardiaadenocarcinomaandosteosarcoma(OS).Weusedlogisticregressiontoperform
singleSNPanalysesforeachstudyseparately,adjustingforstudy-specificcovariates.WecombinedSNPresultsacrossstudiesby
fixed-effectmeta-analysesandanewlydevelopedsubset-basedstatisticalapproach(ASSET).Gene-basedP-valueswereobtained
bytheminPmethodusingtheAdaptiveRankTruncatedProductprogram.WeadjustedformultiplecomparisonsbyBonferroni
correction.
Results:Rs3731239incyclin-dependentkinaseinhibitors2A(CDKN2A)wassignificantlyassociatedwithESCC(P¼7(cid:2)10(cid:3)6).The
CDKN2A-ESCCassociationwasfurthersupportedbygene-basedanalyses(P ¼0.0001).Inthemeta-analysesbyASSET,four
gene
SNPs(rs3731239inCDKN2A,rs615552andrs573687inCDKN2Bandrs564398inCDKN2BAS)showedsignificantassociationswith
ESCC and EC (Po2.46(cid:2)10(cid:3)4). One SNP in MTAP (methylthioadenosine phosphorylase) (rs7023329) that was previously
associatedwithmelanomaandneviinmultiplegenome-wideassociationstudieswasassociatedwithCRC,CAandOSbyASSET
(P¼0.007).
Conclusion: Our data indicatethat genetic variants in CDKN2A, andpossibly nearby genes, maybe associated withESCC and
severalothertumours, further highlighting theimportance of9p21.3 genetic variants incarcinogenesis.
The chromosome 9p21.3 region has been identified as a genetic loci are well recognised as tumour-suppressor genes that are
susceptibility locus for multiple disease phenotypes including involved in the regulation of cell cycle, aging, senescence and
coronaryarterydisease,diabetesandcancer(Pasmantetal,2011). apoptosis (Yang et al, 2010b). CDKN2A confers susceptibility to
This region encompasses several tumour suppressor genes familial melanoma and germline mutations in CDKN2A occur in
including cyclin-dependent kinase inhibitors 2A (CDKN2A), about20%ofmelanomafamilies(Goldstein,2004).TheCDKN2A
CDKN2B, a non-coding RNA (CDKN2BAS, or ANRIL) and encodes both p16 (INK4A), a negative regulator of cyclin-
methylthioadenosine phosphorylase (MTAP). The CDKN2A/2B dependant kinases, and p14 (ARF), an activator ofp53. Theexact
*Correspondence:DrXRYang;E-mail:[email protected]
Received23August2012;revised13December2012;accepted18December2012;publishedonline29January2013
&2013CancerResearchUK.Allrightsreserved0007–0920/13
1378 www.bjcancer.com|DOI:10.1038/bjc.2013.7
Genetic variantsin9p21.3 andmultipletumours BRITISH JOURNALOF CANCER
function of CDKN2BAS is unknown, but it has been shown to one benign condition (colorectal adenoma (CA)). Study partici-
regulate gene expression of CDKN2A/2B and SNPs in this locus pants were Chinese (for ESCC and GCA studies) or whites
havebeenassociatedwithcardiovasculardisease,cancerandother (all other studies). The design of these studies included nested
diseases in genome-wide association studies (GWAS) (Yap et al, case–control (RCC (1994); Prorok et al, 2000; Han et al, 2012),
2010; Pasmant et al, 2011). The MTAP, identified by GWAS as a CRC, CA (Gao et al, 2011)), population-based case–control (EC
naevus-andmelanoma-associatedgene(Bishopetal,2009;Falchi (Yangetal,2010a)),hospital-basedcase–control(OS(Troisietal,
et al, 2009), encodes an enzyme that has a role in polyamine 2006; Mirabello et al, 2011)), and case–control studies of mixed
metabolism. Loss of MTAP expression can exert a tumour- design (ThC (Neta et al, 2012), ESCC and GCA (Blot et al, 1993;
promoting effect, and has been observed in a variety of other Gao et al, 2009)). After excluding subjects with a low genotyping
tumours(Stevensetal,2009),suggestingthatMTAPmayfunction completion rate (o80%), the final analysis for each tumour
as a tumour suppressor gene. The 9p21.3 region also includes a outcomeincluded437RCCcasesand1603controls;417ECcases
cluster of type I interferon (IFN) genes, which encode pleiotropic and 407 controls; 344 ThC cases and 452 controls; 393 CRC
cytokines that exhibit strong antiviral, antiproliferative and cases and 434 controls; 1234 CA cases and 1368 controls; 1027
immunomodulatory effects (Stark et al, 1998). ESCC casesand1452controls; 753GCAcasesand1452controls;
In addition to its well established role in melanoma, deletions 96OScasesand1428controls.WepooledcontrolsforESCCand
of the 9p21.3 region have been observed in a variety of cancers GCA (1452 controls total), as these cases were drawn from the
(van der Riet et al, 1994; Okami et al, 1997; Waber et al, 1997; same underlying studies. The RCC and OS shared a subset (1170
Nakanishi et al, 1999; Perinchery et al, 1999; Schmid et al, 2000; and 1363, respectively) of PLCO controls with CA. Detailed
Sanchez-Cespedes et al, 2001; Hu et al, 2004; Hustinx et al, 2005; information for each study is summarised in Table 1 and
Bartolettietal,2007;Guetal,2008),andSNPsin9p21.3havebeen Supplementary Table S1.
associated with breast cancer, melanoma and glioma by GWAS After correction for multiple testing, ESCC, GCA and CA had
(Bishopetal,2009;Sheteetal,2009;Wrenschetal,2009;Turnbull adequate power (94%, 87% and 98%, respectively) to detect an
et al, 2010).These findings are consistent with a broad role for association for a SNP with minor allele frequency (MAF)¼0.35
9p21.3 genes in carcinogenesis. However, whether genetic poly- and an odds ratio (OR) of 1.4, while all other studies were
morphismsin9p21.3confersusceptibilitytoothercancersremains underpowered (powero80%). However, our aim was to identify
unclear.Thegoalofthecurrentstudywastosystematicallyevaluate geneticvariantsinthe9p21regionassociatedwithmultiplecancer/
variants in 9p21.3 with the risk ofmultiple cancers/tumours. tumour outcomes using combined data across studies.
SNP selection, genotyping and quality control. SNP selection,
MATERIALS AND METHODS genotyping and quality control have been described previously
(Yangetal,2010b).Inbrief,252tagSNPsfor22geneslocatedatthe
Study population. This study sample included data from eight chromosome 9p21.3 region (19.9–32.8Mb) were genotyped at the
studies that participated in iSelect, a jointly conducted project in NCI Core Genotyping Facility (Advanced Technology Center,
theDivisionofCancerEpidemiologyandGeneticsoftheNational Gaithersburg, MD; http://snp500cancer.nci.nih.gov) using a cus-
Cancer Institute (NCI), with a goal to evaluate common genetic tom-designed iSelectInfinium assay (Illumina, www.illumina.com).
variants in selected genes and pathways in multiple tumours, Fromtelomeretocentromere,thesegenesincluded:IFNB1,IFNW1,
especiallyrarecancers(Gaoetal,2009;Yangetal,2010a;Gaoetal, IFNA21,10,16,17,14,5, KLHL9, IFNA6, 2, 8, 1, IFNE1, MTAP,
2011;Mirabelloetal,2011;Hanetal,2012;Netaetal,2012).The CDKN2A, CDKN2B, CDKN2BAS, TUSC1, PLAA, IFNK, ACO1.
study samples comprised seven cancers (renal cell carcinoma For each gene, SNPs spanned 20kb 50 of the transcription start
(RCC), endometrial cancer (EC), thyroid cancer (ThC), colorectal point(exon1)to10kb30 ofthelastexon.TagSNPswereselected
cancer (CRC), oesophageal squamous cell carcinoma (ESCC), using a MAF criterion of MAF 45% based upon HapMap data
gastriccardiaadenocarcinoma(GCA)andosteosarcoma(OS))and for whites (CEU) and Yoruba (YRI) samples using a Tagging
Table1.Descriptionofthestudysamples
Study Cases Controls Ethnicity Country Studydesign Covariates
RCC 437 1603 Caucasian US, Nestedcase–controlwithinthePLCOCancerScreeningTrial Age,gender,
Finland (Proroketal,2000)andATBC(1994) studycentre(11)
EC 417 407 Caucasian Poland Population-basedcase–control(Yangetal,2010a) Age,site(2)
ThC 344 452 Caucasian US CasesfromtheUSRTcohort,orUniversityofTexasMDAndersonCancer Age,gender,
Center.ControlsfromUSRT(Netaetal,2012) birthyear
category
CRC 393 434 Caucasian US Nestedcase–controlwithinscreeningarmofPLCO Age
CA 1234 1368 Caucasian US Nestedcase–controlwithinscreeningarmofPLCO(Gaoetal,2011) Age
ESCC 1027 1452 Asian China Neighborhood-basedcase–controlfromtheUGICancerGeneticsProject Age,gender,
(Gaoetal,2009)andnestedcase–controlfromNITs(Blotetal,1993) studyregion(2)
GCA 753 1452 Asian China SameasESCC Age,gender,
studyregion(2)
OS 96 1428 Caucasian US Hospital-basedcase–control(63controls);1365additional Gender
controlsfromPLCO(Mirabelloetal,2011;Troisietal,2006)
Abbreviations:ATBC¼alpha-tocopherol,Beta-CaroteneCancerPrevention;CA¼colorectaladenoma;CRC¼colorectalcancer;EC¼endometrialcancer;ESCC¼oesophagealsquamouscell
carcinoma;GCA¼gastriccardiaadenocarcinoma;NITs¼nutritioninterventiontrials;PLCO¼prostate,lung,colorectalandovarian;RCC¼renalcellcarcinoma;OS¼osteosarcoma;ThC¼
thyroidcancer;UGI¼uppergastrointestinal;USRT¼USradiologictechnologists.
www.bjcancer.com|DOI:10.1038/bjc.2013.7 1379
BRITISHJOURNAL OF CANCER Genetic variants in9p21.3andmultiple tumours
algorithm (Carlson et al, 2004). Selected SNPs are listed in fornumbersofSNPs.Allstatisticalanalyseswereperformedusing
Supplementary Table 2. the R software.
The iSelect panel was validated using all three HapMap
populations (CEU, YRI, Japanese and Chinese). The SNPs with
low (o90%) genotyping completion rate, low (o95%) concor-
dance rate or deviation (Po0.001) from Hardy–Weinberg RESULTS
equilibriumamongcontrolswereexcludedfromeachparticipating
study.ThenumberofSNPsincludedinthefinalanalyseswere:170 When analysing each study separately, we found one SNP in
SNPsforRCC,202SNPsforEC,195SNPsforThC,193SNPsfor CDKN2A(rs3731239)thatwassignificantlyassociatedwithESCC
CRC, 203 SNPs for CA, 139 SNPs for ESCC and GCA and 200 after Bonferroni correction (P¼7(cid:2)10(cid:3)6) (Table 2a). The minor
SNPs for OS. In the ESCC and GCA study, a larger number of allele (G, MAF¼0.12) of this SNP was associated with increased
SNPs were excluded due to low MAF (o5%), likely reflecting ESCC risk (OR¼1.51, 95% CI¼1.25, 1.84, for AG vs AA;
differences between white and Asian populations. OR¼1.88,95%CI¼1.04,3.41,forGGvsAA;Table2a).Figure1
shows that the LD pattern and genotype frequencies among
Statisticalanalyses. Wefirstassessedtheassociationbetweeneach controls were different in the Chinese and Caucasian samples.
SNP and each cancer outcome separately. Unconditional logistic We also found 18 additional SNPs that were associated
regressionwasusedtoestimateORsand95%confidenceintervals with at least one tumour outcome at Po0.01 (Table 2b),
(CIs) and P-values for trend, using additive coding for genotypes although the associations were not significant after Bonferroni
(0,1,2minoralleles).Thehomozygoteofthecommonalleleserved correction.Amongthem,oneSNPinMTAP(rs7023329)thatwas
as the reference group. Heterozygous and homozygous rare previously associated with melanoma and nevi in several GWAS
genotypes were combined when the number of subjects with (Bishop et al, 2009; Barrett et al, 2011), was associated with CA
homozygousminoralleleswaso5,andadominantgeneticmodel (P¼0.0005). Another previously identified SNP (rs4977756) in
was used. Appropriate covariates adjustment was performed for CDKN2BAS from a GWAS for glioma (Shete et al, 2009), was
eachtumouroutcomeperdiscussionwithprincipalinvestigatorsof associated with EC (P¼0.009) and ESCC (P¼0.002).
each study (Table 1). In fixed-effect meta-analyses, only rs7023329 in MTAP showed
To examine whether 9p21 variants were associated with marginal association (fixed effect Po0.05) before correction for
multiple cancer/tumour outcomes, we conducted meta-analyses multiple testing (Table 2c). When using the subset-based approach
combining data from the eight studies. To combine SNP results (ASSET), rs7023329 showed suggestive association with multiple
across studies, we first used a standard fixed-effect meta-analysis tumours (positive effect P¼0.007), with the strongest signal
and then a newly developed subset-based statistical approach obtained from the subset combining data from CRC, CA and OS
(ASSET)(Bhattacharjeeetal,2012).ASSETisamodifiedfix-effect studies (Table 2c). In addition, the subset approach identified
meta-analysisapproachthatallowsforheterogeneityofSNPeffects significant associations between rs3731239 in CDKN2A, rs615552
on different outcomes by exhaustively exploring subsets of the andrs573687inCDKN2B,andrs564398inCDKN2BAS,andECand
studies for the presence of association signals. The ASSET test ESCC after Bonferroni correction (positive effect Po2.46(cid:2)10(cid:3)4,
statisticZ(S)foragivensubsetSofkstudiesisaweightedsumof Table2c),althoughtheseassociationsseemedtobemainlydrivenby
thekstudy-specificteststatistics,Z(S)¼a1Z1þyþakZk,where ESCCbasedonsensitivityanalysesthatexcludedESCC.Theeffects
theajistheproportionofthesamplesizeforthejthstudyrelative of all SNPs with Pp0.01 in the subset analyses showed the same
to the total sample size for the studies in the given subset S. The direction (positive effect) across contributing study outcomes
overall evidence of the association of the SNP is then based on (Table2b andc).
evaluationofZmax¼ maxS|Z(S)|,i.e.themaximumofthesubset- Gene-based analyses showed that the CDKN2A gene was
specificteststatisticsoverallpossiblesubsetsofthestudies.Under significantly associated with ESCC (P ¼0.0001) and the
gene
the null hypothesis the vector of values Z(S) has a multivariate associationremainedsignificantafteradjustingformultipletesting
normal distribution with mean zero and variances equal to one. (Table3).Othergenesinthenearbyregion,MTAP(P ¼0.015),
gene
The correlation between Z(A) and Z(B) for two different subsets CDKN2B (P ¼0.01) and CDKN2BAS (P ¼0.009), also
gene gene
A and B is given in (Bhattacharjee et al, 2012). We computed a showed suggestive associations with ESCC (Table 3). In addition,
two-sidedversionofthetestthatalsoallowsthedetectionofeffects MTAP showed a suggestive association with CA (P ¼0.006).
gene
in opposite directions. Both fixed-effect and subset-based meta-
analyses were performed using the ‘ASSET’ R package, which can
take into account shared controls across studies. Because theSNP
minorallelesmaydifferacrossstudies,westandardisedtheeffects
before combining the data by multiplying the beta-coefficients of
Table2a.Associationbetweenrs3731239andESCCinaChinese
SNPs by 1 or -1. populationa
Gene-based analyses were performed on the 22 genes to assess
thesignificanceofthejointeffectofmultipleSNPsineachgeneon
Case, Control,
each outcome separately. Gene-based P-values (P ) were
gene n¼1027 n¼1452
computed using the minP method by Adaptive Rank Truncated Genotype (%)b (%)b OR 95%CI P-trend
Product(ARTP)program(Yuetal,2009).TheminimumP-value
AA 712(69.3) 1093(75.3) 1.00 —
of each gene was used as the test statistic and its significance was
assessedusingapermutationtestwith10000permutations,taking AG 272(26.5) 294(20.1) 1.51 1.25–1.84
intoaccountthenumberofSNPsgenotypedineachgeneandtheir GG 25(2.4) 22(1.5) 1.88 1.04–3.41
linkage disequilibrium (LD) structure. PerGallele 1.47 1.24–1.75 7(cid:2)10(cid:3)6
We used Bonferroni correction to account for the number of
SNPsorgenesandstudiestested,thereforePforSNPo3.1(cid:2)10(cid:3)5 Abbreviations: CI¼confidence interval; ESCC¼oesophageal squamous cell carcinoma;
(0.05/(203(cid:2)8)) and P o2.8(cid:2)10(cid:3)4 (0.05/(22(cid:2)8)) were used OR¼oddsratio.
to define SNP-based gaennde gene-based statistical significance. In aResultswereobtainedfromunconditionallogisticregression,adjustingforage,gender
meta-analysis, P for combined analysiso2.46(cid:2)10(cid:3)4 (0.05/203) abnPderscteundtyagreegdioone.snotsumupto1owingtomissingvalues.
was considered statistically significant after Bonferroni correction
1380 www.bjcancer.com|DOI:10.1038/bjc.2013.7
Genetic variantsin9p21.3 andmultipletumours BRITISH JOURNALOF CANCER
a
com P 0.53 0.35 0.48 0.15 0.12 0.14 0.75 0.1 0.37 0.27 0.96 0.41 0.34 0.45 0.75 0.34 0.55 0.86
r
a
s
o
e
Ost OR 1.11 1.15 0.84 0.8 1.26 1.25 1.08 0.78 0.87 0.84 1.01 0.88 0.86 0.89 0.95 0.76 0.91 1.03
diaoma P 0.05 0.36 0.12 0.48 0.13 0.16 0.21 0.15 0.8 0.13 0.99 0.98 0.92 0.31 0.32 0.49
rn
ccaarci
trioc
Gasaden OR 0.82 1.06 1.11 0.95 0.91 0.9 0.9 1.1 0.98 1.16 1 1 1.01 0.92 1.08 1.05 200 oldface.
b
n
i
n
w
squamousnoma P 0.87 0.47 0.36 0.002 0.35 0.08 0.002 0.34 0.007 6c(cid:3)710(cid:2) 0.002 0.002 0.001 0.06 0.002 0.31 0.01wasshoo
aguscarci P-value
Oesophcell OR 1.01 1.05 0.95 1.25 1.06 0.89 1.24 0.94 0.84 1.47 1.32 1.32 1.34 0.86 1.25 0.94 139 nTable1.
i
d
e
ctalma P 0.93 0.88 0.009 0.005 0.29 0.0005 0.006 0.84 0.008 0.63 0.56 0.09 0.36 0.09 0.11 0.13 0.65 0.06 0.6 ariateslist
Coloreadeno OR 1.01 0.99 1.22 0.85 1.15 1.22 1.17 1.02 0.86 0.97 0.97 0.91 0.95 0.91 0.91 0.92 1.04 0.9 1.03 139 dy-specificcov
u
ectalcer P 0.006 0.007 0.7 0.08 0.31 0.1 0.19 0.94 0.29 0.67 0.64 0.46 0.27 0.58 0.18 0.92 0.5 0.89 mentofst
Colorcan OR 0.72 0.74 0.95 0.83 1.28 1.19 1.15 1.01 0.9 0.96 1.05 0.93 0.89 0.94 0.87 1.02 0.93 0.99 203 eadjust
h
t
Table2b.SelectedSNP-basedresults,withP-value0.01foratleastonetumouroutcomeo RenalcellEndometrialThyroidcarcinomacancercancer abbORPORPORPSNPGeneA1 rs17692502IFNW1G0.90.230.840.111.180.18 rs10964862IFNW1A0.980.80.990.891.040.7 rs10119678IFNA6A1.150.191.120.390.840.25 rs10757257MTAPA0.950.541.150.150.970.76 rs2039971MTAPT0.920.65 rs7023329MTAPA1.020.790.970.731.040.72 rs7027989MTAPA1.150.070.950.610.98 rs7874112MTAPG0.940.610.780.171.190.34 rs10811629MTAPG0.980.751.150.150.920.44 rs10757261CDKN2AA0.980.750.970.761.190.12 rs3731239CDKN2AG1.030.731.220.050.850.15 rs1063192CDKN2BG1.010.851.310.0090.890.27 rs573687CDKN2BA0.960.611.280.0150.870.19 rs518394CDKN2BC0.940.481.350.003 rs615552CDKN2BC0.940.421.320.0050.830.07 rs564398CDKN2BASC0.970.741.350.0030.870.21 rs11790231CDKN2BASA1.060.640.890.551.630.006 rs4977756CDKN2BASG1.020.791.30.0090.880.26 rs10757274CDKN2BASG0.980.810.760.0061.120.28 TotalSNPno.170202195193 Abbreviations:CIconfidenceinterval;ORoddsratio;SNPsingle-nucleotidepolymorphism.¼¼¼aA1istheeffectallele(minoralleleofColorectalAdenomastudypopulation).bORsandtrendP-valuesforeachSNP-tumourassociationwereobtainedbyunconditionallogisticregressionwithc5(cid:3)SignificantafterBonferronicorrectionfornumberofSNPsandstudies(P0.05/(2038)3.110).(cid:2)¼(cid:2)o
www.bjcancer.com|DOI:10.1038/bjc.2013.7 1381
BRITISHJOURNAL OF CANCER Genetic variants in9p21.3andmultiple tumours
Table2c.Meta-analysesofselectedSNPswithtwo-sidedsubsetsearchPp0.01a
Two-sidesubsetsearchc
Subsetsofstudieswith
Pforpositive strongestsignalwith Pfornegative
SNP Gene Fixed-effectPb CombinedP effect positiveeffect effect
rs7023329 MTAP 0.04 0.01 0.007 CRC,CA,OS 0.23
rs3731239 CDKN2A 0.06 3.13(cid:2)10(cid:3)4 8.7(cid:2)10(cid:3)5d EC,ESCC 0.32
rs615552 CDKN2B 0.91 7.3(cid:2)10(cid:3)4 1.3(cid:2)10(cid:3)4d EC,ESCC 0.53
rs573687 CDKN2B 0.82 2.0(cid:2)10(cid:3)3 2.4(cid:2)10(cid:3)4d EC,ESCC 1
rs4977756 CDKN2BAS 0.38 3.2(cid:2)10(cid:3)3 0.001 EC,ESCC 0.25
rs564398 CDKN2BAS 0.83 8.3(cid:2)10(cid:3)4 2.0(cid:2)10(cid:3)4d EC,ESCC 0.39
Abbreviations:CA¼colorectaladenoma;CRC¼colorectalcancer;EC¼endometrialcancer;ESCC¼oesophagealsquamouscellcarcinoma;OS¼osteosarcoma;SNP¼single-nucleotide
polymorphism.
aResultswerefromASSET,asubset-basedassociationanalysisforcombiningSNP-basedresultsacrosseighttumouroutcomes.
bFixedeffectPwascalculatedbystandardfixed-effectmeta-analyses.
cTwo-sidedsubsetsearchallowedforoppositedirectionsofalleleeffectsacrossdifferentoutcomes.
dSignificantafterBonferronicorrectionfornumberofSNPs(0.05/203¼2.46(cid:2)10(cid:3)4).
rs3731239
A Frequency
1100 77.6%
1 2 61 9 6 1000
7 1 2 3 7 2 8 5
99 41 57 12 68 55 39 77 900
3 7 7 3 3 5 4 7
0 8 0 7 7 1 6 9
2 7 1 3 5 6 5 4 800
s s s s s s s s
r r r r r r r r
700
74 0 5 76 94 95 40 600
0 0 4 75 91 41 500
0 0 4 72 40 400
0 0 4 34 300 20.9%
0 0 11 200
0 0 100
0 1.6%
0
AA AG GG
B Frequency
800 47.8%
1
1 2 6 9 6
7 1 2 3 7 2 8 5
99 41 57 12 68 55 39 77 700
3 7 7 3 3 5 4 7
0 8 0 7 7 1 6 9 37.8%
2 7 1 3 5 6 5 4
rs rs rs rs rs rs rs rs 600
77 0 8 3 6 15 20 500
0 0 3 0 1 0 400
0 0 19 18 3
300
0 1 7 91 14.7%
1 0 7 200
0 0
100
0
0
AA AG GG
Figure1.Linkagedisequilibriumstructuresandgenotypefrequenciesofrs3731239amongcontrolsofChinese(A)andCaucasian(B)samples.
TheLD(indicatedbyr2)mapsweredrawnusingtheHaploviewsoftware,basedonthegenotypingdataofcontrolsamplesforESCC(A)andCA
(B).LDpatternsinotherCaucasianstudiesweresimilartothatinCA.
1382 www.bjcancer.com|DOI:10.1038/bjc.2013.7
Genetic variantsin9p21.3 andmultipletumours BRITISH JOURNALOF CANCER
Table3.Gene-basedP-valuesfor9p21.3genesinassociationwitheighttumouroutcomesa
Oesophageal
Renalcell Endometrial Thyroid Colorectal Colorectal squamous Gastriccardia
Geneb carcinoma cancer cancer cancer adenoma cellcarcinoma adenocarcinoma Osteosarcoma
(473/1603)c (417/407) (344/452) (393/434) (1234/1368) (1027/1452) (753/1452) (96/1426)
IFNB1 0.15 0.94 0.74 0.79 0.39 0.27 0.91 0.99
IFNW1 0.62 0.63 0.56 0.06 0.68 0.82 0.3 0.97
IFNA21 0.65 0.65 0.97 0.78 0.28 0.78 0.9 0.74
IFNA10 0.51 0.76 0.8 0.88 0.07 0.87 0.26 0.9
IFNA16 0.42 0.71 0.35 0.53 0.29 0.95 0.05 0.52
IFNA17 0.58 0.95 0.99 0.28 0.78 0.32 0.86 0.89
IFNA14 0.4 0.47 0.66 0.98 0.51 0.7 0.56 0.65
IFNA5 0.61 0.88 0.82 0.48 0.08 0.77 0.22 0.17
KLHL9 0.31 0.7 0.45 0.46 0.038 0.68 0.16 0.75
IFNA6 0.4 0.57 0.51 0.74 0.026 0.56 0.32 0.81
IFNA2 0.75 0.11 0.73 0.87 0.1 0.66 0.38 0.89
IFNA8 0.76 0.65 0.88 0.94 0.19 0.26 0.31 0.6
IFNA1 0.23 0.66 1 0.99 0.14 0.18 0.56 0.65
IFNE1 0.07 0.14 0.53 0.29 0.37 0.89 0.25 0.57
MTAP 0.37 0.12 0.76 0.31 0.006 0.015 0.52 0.63
CDKN2A 0.88 0.32 0.65 0.47 0.77 0.0001d 0.43 0.19
CDKN2B 0.35 0.02 0.26 0.54 0.41 0.01 0.67 0.21
CDKN2BAS 0.99 0.04 0.07 0.44 0.45 0.009 0.84 0.43
TUSC1 0.75 0.77 0.18 0.14 0.1 0.19 0.79 0.32
PLAA 0.98 0.73 0.21 0.46 0.58 0.36 0.88 0.64
IFNK 0.07 0.82 0.6 0.92 0.3 0.84 0.49 0.86
ACO1 0.46 0.73 0.27 0.14 0.74 0.53 0.3 0.44
aGene-basedPvalueswerecomputedusingtheminPmethod,basedon10000permutations;P-valuep0.01wasshowninboldface.
bGenesareorderedbylocationfromtelomeretocentromere.
cNumberofcasesandcontrols.
dSignificantafterBonferronicorrectionfornumberofgenesandstudies(0.05/(22(cid:2)8)¼2.8(cid:2)10(cid:3)4).
The minor allele of this SNP is more common in Caucasians
DISCUSSION
(0.39 among controls in our study) than in Chinese (0.12 among
controlsinESCC).Inaddition,thetwoethnicpopulationsshowed
In this study, we evaluated associations of up to 203 SNPs in 22 distinctLDpatternsintheregionflankingthisSNP,whichmayalso
geneslocatedonchromosome9p21.3withtheriskofeighttumour contribute tothe differencesinthe associationobserved.
outcomesindatafromeightcase–controlstudies.Whenanalysing Recent studies have suggested that the 9p21.3 region was
each tumour outcome separately, we identified a single SNP in enriched in regulatory sequences such as enhancers that
CDKN2A(rs3731239)thatwassignificantlyassociatedwiththerisk regulate the expression of genes in this region (MTAP-CDKN2A/
of ESCC, after correction for multiple comparisons. Gene-based 2B/CDKN2BAS) and downstream (such as IFNA21), thereby
analyses also suggested that the CDKN2A gene was significantly establishing a functional link between 9p21 genetic variation
associatedwithESCC.Inthesubset-basedmeta-analyses,fourSNPs and immune signalling pathways (Harismendy et al, 2011).
(rs3731239 in CDKN2A, rs615552 and rs573687 in CDKN2B, and Interestingly, rs564398 in CDKN2BAS, which showed suggestive
rs564398 in CDKN2BAS) showed significant associations with associations with both EC and ESCC in our study (see
ESCC and EC. Two previously identified GWAS SNPs, rs7023329 Tables 2b and c), was located within a predicted enhancer
in MTAP for melanoma and nevi and rs4977756 in CDKN2A for sequence. The most significant SNP in our study, rs3731239 in
glioma, showed suggestive associations with CA (for rs7023329) CDKN2A, is located adjacent to the promoter region of CDKN2A
and ECand ESCC(forrs4977756),respectively,inourstudy. Our (about500bpawayfromaCpGislandandpredictedtranscription
findings further highlight the importance of 9p21.3, in particular binding and DNase I sites based on ENCODE data, http://
theMTAP-CDKN2A/2B/CDKN2BASregion,inthepathogenesisof www.genome.ucsc.edu/ENCODE/). A previous study correlating
multiple tumours. 9p21 SNPs with gene expression found that rs3731239 was
Rs3731239 previously demonstrated weak associations with significantly associated with allele-specific expression of
breastcancer(Driveretal,2008;Mavaddatetal,2009)andovarian CDKN2BAS (P¼10(cid:3)25) (Cunnington et al, 2010). Three other
cancer(Goodeetal,2009)inpredominantlyCaucasianpopulations. SNPs (rs1063192, rs564398 and rs11790231) in the CDKN2B/
In our study, this SNP was significantly associated with ESCC CDKN2BAS locus that showed suggestive associations with EC
in Chinese and only weakly associated with EC in Caucasians. (and/or ESCC, ThC) were also significantly associated with
www.bjcancer.com|DOI:10.1038/bjc.2013.7 1383
BRITISHJOURNAL OF CANCER Genetic variants in9p21.3andmultiple tumours
allele-specific expression of CDKN2BAS. CDKN2BAS is a non- chromosomal region in cancer pathogenesis. Further, data on
coding RNA within the CDKN2A/2B locus, which has been somatic alterations of this region (in tumour tissue), such as gene
identified by GWAS of multiple diseases; its expression showed expression, willbeparticularlyhelpful toidentifythemechanisms
the strongest association with the multiple phenotypes (coronary underlying the observed associations.
disease, stroke, diabetes, melanoma and glioma) that were
associated with the 9p21.3 region, as compared with the three
other genes of the cluster (MTAP, CDKN2A, CDKN2B) (Pasmant ACKNOWLEDGEMENTS
etal,2011).TheCDKN2BASisinvolvedinregulatingCDKN2A/2B
expressionthroughacis-actingmechanismaswellasbyregulating
WethankXiaoqinXiongforprovidinghelpwiththedataanalyses.
cell proliferation and senescence through pathways independent
WethankiSelectprincipalinvestigatorsofThC,EC,TGCT,RCC,
from CDKN2A/2B (Visel et al, 2010; Congrains et al, 2012). In
CRC,CA,ESCC,GCAandOSstudiesforkindlyprovidingdatafor
addition to CDKN2BAS, two SNPs in MTAP (rs10757257 and
thisproject.WethankDr.PaulaHylandforherhelpinreviewing
rs7027989), which were suggestively associated with CA in our
ENCODE data. This research was supported by the Intramural
study, were also found to be expression quantitative trait loci
Research Program of the NIH, NCI, DCEG.
for MTAP (Zeller et al, 2010). These data, combined with
previous publications, indicate that common genetic variants in
this region may influence disease risk by regulating gene
expression through a cis-effect. With rapid progress in mapping CONFLICT OFINTEREST
regulatory elements and the growing availability of cell and
tissue-specific gene expression data, future studies should be The authors declare no conflict of interest.
able to evaluate the functional relevance of genetic variants
at 9p21.3.
Somatic 9p21 deletions frequently occur in human cancers REFERENCES
such as bladder cancer, pancreatic cancer, oesophageal cancer,
glioma and melanoma (Schmid et al, 2000; Hu et al, 2004;
ATBCCancerPreventionStudyGroup(1994)Thealpha-tocopherol,beta-
Hustinx et al, 2005; Bartoletti et al, 2007; Gu et al, 2008; Rakosy
carotenelungcancerpreventionstudy:design,methods,participant
et al, 2008). In a previous study conducted in the same Chinese characteristics,andcompliance.AnnEpidemiol4(1):1–10.
population from which the ESCC cases in the current study BarrettJH,IlesMM,HarlandM,TaylorJC,AitkenJF,AndresenPA,
were obtained, the majority (73%) of ESCC tumour specimens AkslenLA,ArmstrongBK,AvrilMF,AziziE,BakkerB,BergmanW,
analysed were found to have LOH at 9p21–22, and 25% Bianchi-ScarraG,Bressac-dePailleretsB,CalistaD,Cannon-AlbrightLA,
(14 of 56) of tumours had CDKN2A mutations (point mutations, CordaE,CustAE,DebniakT,DuffyD,DunningAM,EastonDF,
deletions, insertions) (Hu et al, 2004). In addition, promoters FriedmanE,GalanP,GhiorzoP,GilesGG,HanssonJ,HocevarM,
HoiomV,HopperJL,IngvarC,JanssenB,JenkinsMA,JonssonG,
inCDKN2AaretypicallymethylatedinESCCtumours(Rothetal,
KeffordRF,LandiG,LandiMT,LangJ,LubinskiJ,MackieR,MalvehyJ,
2006).
MartinNG,MolvenA,MontgomeryGW,vanNieuwpoortFA,
Oursisthefirstsystematicevaluationofgeneticvariationinthe
NovakovicS,OlssonH,PastorinoL,PuigS,Puig-ButilleJA,
9p21.3 region in relation to multiple tumour outcomes. The
Randerson-MoorJ,SnowdenH,TuominenR,VanBelleP,
strengths of our study include the careful and comprehensive vanderStoepN,WhitemanDC,ZelenikaD,HanJ,FangS,LeeJE,WeiQ,
selection of genes in the entire 9p21.3 region, the application of a LathropGM,GillandersEM,BrownKM,GoldsteinAM,KanetskyPA,
newly developed subset analysis method to combine SNP data MannGJ,MacgregorS,ElderDE,AmosCI,HaywardNK,GruisNA,
across multiple studies, and use of a gene-based permutation DemenaisF,BishopJA,BishopDT(2011)Genome-wideassociationstudy
analysis method to comprehensively evaluate variation in genes identifiesthreenewmelanomasusceptibilityloci.NatGenet43(11):
with cancer risk. In addition, SNPs were genotyped for all studies 1108–1113.
BartolettiR,CaiT,NesiG,RobertaGirardiL,BaroniG,DalCantoM(2007)
using the same platform and quality control procedures. Our
LossofP16expressionandchromosome9p21LOHinpredictingoutcome
findings suggest that combining data from multiple cancer
ofpatientsaffectedbysuperficialbladdercancer.JSurgRes143(2):
outcomes may provide additional information in understanding
422–427.
disease associations with GWAS variants. BhattacharjeeS,RajaramanP,JacobsKB,WheelerWA,MelinBS,HartgeP,
There are several limitations in our study. First, studies YeagerM,ChungCC,ChanockSJ,ChatterjeeN(2012)Asubset-based
included in this analysis varied by study design, population approachimprovespowerandinterpretationforthecombinedanalysisof
ethnicity and sharing controls in some studies, which posed geneticassociationstudiesofheterogeneoustraits.AmJHumGenet90(5):
challengesforcombining dataaswellasgeneralising thefindings. 821–835.
We therefore applied a new statistical approach, which was BishopDT,DemenaisF,IlesMM,HarlandM,TaylorJC,CordaE,Randerson-
MoorJ,AitkenJF,AvrilMF,AziziE,BakkerB,
specifically designed to handle heterogeneity across studies.
Bianchi-ScarraG,Bressac-dePailleretsB,CalistaD,Cannon-AlbrightLA,
Second, our sample size was in general small, which may limit
ChinAWT,DebniakT,Galore-HaskelG,GhiorzoP,GutI,HanssonJ,
statistical power for identifying significant associations in the
HocevarM,HoiomV,HopperJL,IngvarC,KanetskyPA,KeffordRF,Landi
smaller individual studies. In fact, most associations were not
MT,LangJ,LubinskiJ,MackieR,MalvehyJ,MannGJ,MartinNG,
significant after correcting for multiple testing, with the noted MontgomeryGW,vanNieuwpoortFA,NovakovicS,OlssonH,PuigS,Weiss
exceptionofrs3731239inCDKN2A,withESCC,whichwasamong M,vanWorkumW,ZelenikaD,BrownKM,GoldsteinAM,GillandersEM,
the largest studies. However, the Bonferroni test is conservative, BolandA,GalanP,ElderDE,GruisNA,HaywardNK,LathropGM,Barrett
especiallyforpreviouslyidentifiedGWASSNPs,andthereforethe JH,BishopJA(2009)Genome-wideassociationstudyidentifiesthreeloci
observed associations in our study warrant future investigation in associatedwithmelanomarisk.NatGenet41(8):920–925.
larger samples. BlotWJ,LiJY,TaylorPR,GuoW,DawseyS,WangGQ,YangCS,ZhengSF,
GailM,LiGY,YuY,LiuBQ,TangreaJ,SunYH,LiuFS,FraumeniJF,
In conclusion, our data indicated that genetic variants in the
ZhangYH,LiB(1993)NutritioninterventiontrialsinLinxian,China:
9p21.3 region, particularly near the MTAP-CDKN2A/2B/
supplementationwithspecificvitamin/mineralcombinations,cancer
CDKN2BAS, may be associated with ESCC and possibly several
incidence,anddisease-specificmortalityinthegeneralpopulation.JNatl
other tumours. Our findings further highlight the importance of CancerInst85(18):1483–1492.
the 9p21.3 region in disease susceptibility and cancer aetiology. CarlsonCS,EberleMA,RiederMJ,YiQ,KruglyakL,NickersonDA(2004)
Future studies are needed to further investigate the role of this Selectingamaximallyinformativesetofsingle-nucleotidepolymorphisms
1384 www.bjcancer.com|DOI:10.1038/bjc.2013.7
Genetic variantsin9p21.3 andmultipletumours BRITISH JOURNALOF CANCER
forassociationanalysesusinglinkagedisequilibrium.AmJHumGenet geneticvariantsinthe8q24regionandriskofpapillarythyroidcancer.
74(1):106–120. Laryngoscope122(5):1040–1042.
CongrainsA,KamideK,KatsuyaT,YasudaO,OguroR,YamamotoK, OkamiK,CairnsP,WestraWH,LinnJF,AhrendtSA,WuL,SidranskyD,
OhishiM,RakugiH(2012)CVD-associatednon-codingRNA,ANRIL, JenJ(1997)Detaileddeletionmappingatchromosome9p21innon-small
modulatesexpressionofatherogenicpathwaysinVSMC.BiochemBiophys celllungcancerbymicrosatelliteanalysisandfluorescenceinsitu
ResCommun419(4):612–616. hybridization.IntJCancer74(6):588–592.
CunningtonMS,SantibanezKorefM,MayosiBM,BurnJ,KeavneyB(2010) PasmantE,SabbaghA,VidaudM,BiecheI(2011)ANRIL,along,noncoding
Chromosome9p21SNPsassociatedwithmultiplediseasephenotypes RNA,isanunexpectedmajorhotspotinGWAS.FASEBJ25(2):444–448.
correlatewithANRILexpression.PLoSGenet6(4):e1000899. PerincheryG,BukurovN,NakajimaK,ChangJ,LiLC,DahiyaR(1999)High
DriverKE,SongH,LesueurF,AhmedS,Barbosa-MoraisNL,TyrerJP, frequencyofdeletiononchromosome9p21mayharborseveraltumor-
PonderBA,EastonDF,PharoahPD,DunningAM(2008)Associationof suppressorgenesinhumanprostatecancer.IntJCancer83(5):610–614.
single-nucleotidepolymorphismsinthecellcyclegeneswithbreastcancer ProrokPC,AndrioleGL,BresalierRS,BuysSS,ChiaD,CrawfordED,Fogel
intheBritishpopulation.Carcinogenesis29(2):333–341. R,GelmannEP,GilbertF,HassonMA,HayesRB,JohnsonCC,Mandel
FalchiM,BatailleV,HaywardNK,DuffyDL,BishopJA,PastinenT,Cervino JS,ObermanA,O’BrienB,OkenMM,RaflaS,RedingD,RuttW,
A,ZhaoZZ,DeloukasP,SoranzoN,ElderDE,BarrettJH,MartinNG, WeissfeldJL,YokochiL,GohaganJK(2000)DesignoftheProstate,Lung,
BishopDT,MontgomeryGW,SpectorTD(2009)Genome-wide ColorectalandOvarian(PLCO)CancerScreeningTrial.ControlClin
associationstudyidentifiesvariantsat9p21and22q13associatedwith Trials21(6Suppl):273S–309S.
developmentofcutaneousnevi.NatGenet41(8):915–919. RakosyZ,VizkeletiL,EcsediS,BeganyA,EmriG,AdanyR,BalazsM(2008)
GaoY,HayesRB,HuangWY,CaporasoNE,BurdetteL,YeagerM, Characterizationof9p21copynumberalterationsinhumanmelanoma
ChanockSJ,BerndtSI(2011)DNArepairgenepolymorphismsand byfluorescenceinsituhybridization.CancerGenetCytogenet182(2):
tobaccosmokingintheriskforcolorectaladenomas.Carcinogenesis32(6): 116–121.
882–887. RothMJ,AbnetCC,HuN,WangQH,WeiWQ,GreenL,D’AlelioM,Qiao
GaoY,HuN,HanX,GiffenC,DingT,GoldsteinA,TaylorP(2009)Family YL,DawseySM,TaylorPR,WoodsonK(2006)p16,MGMT,RARbeta2,
historyofcancerandriskforesophagealandgastriccancerinShanxi, CLDN3,CRBPandMT1Ggenemethylationinesophagealsquamouscell
China.BMCcancer9:269. carcinomaanditsprecursorlesions.OncolRep15(6):1591–1597.
GoldsteinAM(2004)Familialmelanoma,pancreaticcancerandgermline Sanchez-CespedesM,DeckerPA,DoffekKM,EstellerM,WestraWH,Alawi
CDKN2Amutations.HumMutat23(6):630. EA,HermanJG,DemeureMJ,SidranskyD,AhrendtSA(2001)Increased
GoodeEL,FridleyBL,VierkantRA,CunninghamJM,PhelanCM, lossofchromosome9p21butnotp16inactivationinprimarynon-small
AndersonS,RiderDN,WhiteKL,PankratzVS,SongH,HogdallE, celllungcancerfromsmokers.CancerRes61(5):2092–2096.
KjaerSK,WhittemoreAS,DiCioccioR,RamusSJ,GaytherSA, SchmidM,SenM,RosenbachMD,CarreraCJ,FriedmanH,CarsonDA
SchildkrautJM,PharaohPP,SellersTA(2009)Candidategeneanalysis (2000)Amethylthioadenosinephosphorylase(MTAP)fusiontranscript
usingimputedgenotypes:cellcyclesingle-nucleotidepolymorphisms identifiesanewgeneonchromosome9p21thatisfrequentlydeletedin
andovariancancerrisk.CancerEpidemiolBiomarkersPrev18(3): cancer.Oncogene19(50):5747–5754.
935–944. SheteS,HoskingFJ,RobertsonLB,DobbinsSE,SansonM,MalmerB,Simon
GuJ,HorikawaY,ChenM,DinneyCP,WuX(2008)Benzo(a)pyrenediol M,MarieY,BoisselierB,DelattreJY,Hoang-XuanK,ElHallaniS,Idbaih
epoxide-inducedchromosome9p21aberrationsareassociatedwith A,ZelenikaD,AnderssonU,HenrikssonR,BergenheimAT,FeychtingM,
increasedriskofbladdercancer.CancerEpidemiolBiomarkersPrev17(9): LonnS,AhlbomA,SchrammJ,LinnebankM,HemminkiK,KumarR,
2445–2450. HepworthSJ,PriceA,ArmstrongG,LiuY,GuX,YuR,LauC,
HanSS,YeagerM,MooreLE,WeiMH,PfeifferR,ToureO,PurdueMP, SchoemakerM,MuirK,SwerdlowA,LathropM,BondyM,HoulstonRS
JohanssonM,SceloG,ChungCC,GaborieauV,ZaridzeD,SchwartzK, (2009)Genome-wideassociationstudyidentifiesfivesusceptibilitylocifor
Szeszenia-DabrowskaN,DavisF,BenckoV,ColtJS,JanoutV,MatveevV, glioma.NatGenet41(8):899–904.
ForetovaL,MatesD,NavratilovaM,BoffettaP,BergCD,Grubb3rdRL, StarkGR,KerrIM,WilliamsBR,SilvermanRH,SchreiberRD(1998)How
StevensVL,ThunMJ,DiverWR,GapsturSM,AlbanesD,WeinsteinSJ, cellsrespondtointerferons.AnnuRevBiochem67:227–264.
VirtamoJ,BurdettL,BrisudaA,McKayJD,FraumeniJr.JF,ChatterjeeN, StevensAP,SpanglerB,WallnerS,KreutzM,DettmerK,OefnerPJ,
RosenbergPS,RothmanN,BrennanP,ChowWH,TuckerMA, BosserhoffAK(2009)Directandtumormicroenvironmentmediated
ChanockSJ,ToroJR(2012)Thechromosome2p21regionharborsa influencesof5’-deoxy-5’-(methylthio)adenosineontumorprogressionof
complexgeneticarchitectureforassociationwithriskforrenalcell malignantmelanoma.JCellBiochem106(2):210–219.
carcinoma.HumMolGenet21(5):1190–1200. TroisiR,MastersMN,JoshipuraK,DouglassC,ColeBF,HooverRN(2006)
HarismendyO,NotaniD,SongX,RahimNG,TanasaB,HeintzmanN, Perinatalfactors,growthanddevelopment,andosteosarcomarisk.BrJ
RenB,FuXD,TopolEJ,RosenfeldMG,FrazerKA(2011)9p21DNA Cancer95(11):1603–1607.
variantsassociatedwithcoronaryarterydiseaseimpairinterferon-gamma TurnbullC,AhmedS,MorrisonJ,PernetD,RenwickA,MaranianM,SealS,
signallingresponse.Nature470(7333):264–268. GhoussainiM,HinesS,HealeyCS,HughesD,Warren-PerryM,Tapper
HuN,WangC,SuH,LiWJ,Emmert-BuckMR,LiG,RothMJ,TangZZ, W,EcclesD,EvansDG,HooningM,SchutteM,vandenOuwelandA,
LuN,GiffenC,AlbertPS,TaylorPR,GoldsteinAM(2004)High HoulstonR,RossG,LangfordC,PharoahPD,StrattonMR,DunningAM,
frequencyofCDKN2Aalterationsinesophagealsquamouscellcarcinoma RahmanN,EastonDF(2010)Genome-wideassociationstudyidentifies
fromahigh-riskChinesepopulation.GenesChromosomesCancer39(3): fivenewbreastcancersusceptibilityloci.NatGenet42(6):504–507.
205–216. vanderRietP,NawrozH,HrubanRH,CorioR,TokinoK,KochW,
HustinxSR,LeoniLM,YeoCJ,BrownPN,GogginsM,KernSE,HrubanRH, SidranskyD(1994)Frequentlossofchromosome9p21-22earlyinhead
MaitraA(2005)ConcordantlossofMTAPandp16/CDKN2Aexpression andneckcancerprogression.CancerRes54(5):1156–1158.
inpancreaticintraepithelialneoplasia:evidenceofhomozygousdeletionin ViselA,ZhuY,MayD,AfzalV,GongE,AttanasioC,BlowMJ,CohenJC,
anoninvasiveprecursorlesion.ModPathol18(7):959–963. RubinEM,PennacchioLA(2010)Targeteddeletionofthe9p21non-
MavaddatN,DunningAM,PonderBA,EastonDF,PharoahPD(2009) codingcoronaryarterydiseaseriskintervalinmice.Nature464(7287):
Commongeneticvariationincandidategenesandsusceptibilitytosubtypes 409–412.
ofbreastcancer.CancerEpidemiolBiomarkersPrev18(1):255–259. WaberP,DlugoszS,ChengQC,TruelsonJ,NisenPD(1997)Genetic
MirabelloL,RichardsEG,DuongLM,YuK,WangZ,CawthonR,BerndtSI, alterationsofchromosomeband9p21-22inheadandneckcancerarenot
BurdettL,ChowdhuryS,TeshomeK,DouglassC,SavageSA(2011) restrictedtop16INK4a.Oncogene15(14):1699–1704.
Telomerelengthandvariationintelomerebiologygenesinindividuals WrenschM,JenkinsRB,ChangJS,YehRF,XiaoY,DeckerPA,BallmanKV,
withosteosarcoma.IntJMolEpidemiolGenet21:19–29. BergerM,BucknerJC,ChangS,GianniniC,HalderC,KollmeyerTM,
NakanishiH,WangXL,ImaiFL,KatoJ,ShiibaM,MiyaT,ImaiY, KoselML,LaChanceDH,McCoyL,O’NeillBP,PatokaJ,PicoAR,Prados
TanzawaH(1999)Localizationofanoveltumorsuppressorgenelocion M,QuesenberryC,RiceT,RynearsonAL,SmirnovI,TihanT,WiemelsJ,
chromosome9p21-22inoralcancer.AnticancerRes19(1A):29–34. YangP,WienckeJK(2009)VariantsintheCDKN2BandRTEL1regions
NetaG,YuCL,BrennerA,GuF,HutchinsonA,PfeifferR,SturgisEM,XuL, areassociatedwithhigh-gradegliomasusceptibility.NatGenet41(8):
LinetMS,AlexanderBH,ChanockS,SigurdsonAJ(2012)Common 905–908.
www.bjcancer.com|DOI:10.1038/bjc.2013.7 1385
BRITISHJOURNAL OF CANCER Genetic variants in9p21.3andmultiple tumours
YangHP,GonzalezBosquetJ,LiQ,PlatzEA,BrintonLA,ShermanME, YuK,LiQ,BergenAW,PfeifferRM,RosenbergPS,CaporasoN,KraftP,
LaceyJrJV,GaudetMM,BurdetteLA,FigueroaJD,CiampaJG,Lissowska ChatterjeeN(2009)Pathwayanalysisbyadaptivecombinationof
J,PeplonskaB,ChanockSJ,Garcia-ClosasM(2010a)Commongenetic P-values.GenetEpidemiol33(8):700–709.
variationinthesexhormonemetabolicpathwayandendometrialcancer ZellerT,WildP,SzymczakS,RotivalM,SchillertA,CastagneR,MaoucheS,
risk:pathway-basedevaluationofcandidategenes.Carcinogenesis31(5): GermainM,LacknerK,RossmannH,EleftheriadisM,SinningCR,
827–833. SchnabelRB,LubosE,MennerichD,RustW,PerretC,ProustC,Nicaud
YangXR,LiangX,PfeifferRM,WheelerW,MaederD,BurdetteL,YeagerM, V,LoscalzoJ,HubnerN,TregouetD,MunzelT,ZieglerA,TiretL,
ChanockS,TuckerMA,GoldsteinAM(2010b)Associationsof9p21 BlankenbergS,CambienF(2010)Geneticsandbeyond–thetranscriptome
variantswithcutaneousmalignantmelanoma,nevi,andpigmentation ofhumanmonocytesanddiseasesusceptibility.PloSone5(5):e10693.
phenotypesinmelanoma-pronefamilieswithandwithoutCDKN2A
mutations.FamCancer9(4):625–633.
Thisworkispublishedunderthestandardlicensetopublishagree-
YapKL,LiS,Munoz-CabelloAM,RaguzS,ZengL,MujtabaS,GilJ,Walsh
MJ,ZhouMM(2010)MolecularinterplayofthenoncodingRNAANRIL ment. After 12 months the work will become freely available and
andmethylatedhistoneH3lysine27bypolycombCBX7intranscriptional the license terms will switch to a Creative Commons Attribution-
silencingofINK4a.Molcell38(5):662–674. NonCommercial-Share Alike 3.0 Unported License.
Supplementary Information accompanies this paper on British Journal of Cancer website (http://www.nature.com/bjc)
1386 www.bjcancer.com|DOI:10.1038/bjc.2013.7