prognostic signiﬁcance of promoter dna methylation in...

Imaging, Diagnosis, Prognosis

Prognostic Significance of Promoter DNA Methylation inPatients with Childhood Neuroblastoma

DianaT. Lau1, LukeB.Hesson2,MurrayD.Norris1,GlennM.Marshall1,3,MichelleHaber1, andLesley J. Ashton1

AbstractPurpose: To characterize the clinical significance of promoter methylation in a cohort of primary

neuroblastoma tumors and investigate the association between DNA methylation and clinical outcome.

Experimental Design: A customized Illumina GoldenGate methylation assay was used to assess

methylation status of 96 CpG sites within 48 candidate genes in primary neuroblastoma tumors obtained

from 131 children diagnosed in Australia. Genes were selected on the basis of previous reports of altered

DNAmethylation in embryonal cancers. Levels of DNAmethylationwere validated in a subset of 48 patient

samples using combinedbisulfite restriction analysis (CoBRA) andbisulfite sequencing. ACoxproportional

hazards model was used to investigate the association between promoter hypermethylation and the risk of

relapse/death within 5 years of diagnosis, while adjusting for known prognostic factors including MYCN

amplification, age, and stage at diagnosis.

Results: Levels of promoter methylation of DNAJC15, neurotrophic tyrosine kinase receptor 1 or TrkA

(NTRK1), and tumor necrosis factor receptor superfamily,member 10D(TNFRSF10D), were higher in older

patients at diagnosis (P < 0.01), whereas higher levels of methylation of DNAJC15, NTRK1, and PYCARD

were observed in patients withMYCN amplification (P < 0.001). Inmultivariate analysis, hypermethylation

of folate hydrolase (FOLH1), myogenic differentiation-1 (MYOD1), and thrombospondin-1 (THBS1)

remained significant independent predictors of poorer clinical outcome after adjusting for known prog-

nostic factors (P� 0.017). Moreover, more than 30% of patients displayed hypermethylation in 2 genes or

more and were at least 2 times more likely to relapse or die (HR ¼ 2.72, 95% confidence interval ¼ 1.55–

4.78, P ¼ 0.001), independent of MYCN status, age, and stage at diagnosis.

Conclusions: Our findings highlight the potential use of methylation profiling to identify additional

prognostic markers and detect new therapeutic targets for selected patient subsets. Clin Cancer Res; 18(20);

5690–700. �2012 AACR.

IntroductionNeuroblastoma is an embryonal malignancy that

accounts for 8% to 10% of all childhood cancers and ischaracterized by a diversity of clinical behaviors ranging

from spontaneous regression to rapid and fatal tumorprogression (1, 2). In recent years, several genetic changeshave been identified in neuroblastoma tumors that arerelevant to clinical progression, allowing individual tumorsto be classified into distinct subsets. Prognostic markers,such as age at diagnosis, clinical stage, amplification of theMYCN oncogene, DNA ploidy, andmolecular defects, suchas allelic loss of chromosome 1p and 11q are used for riskstratification and treatment assignment. The most promi-nent of these prognostic markers is MYCN, an oncogenethat is amplified in approximately 20% to 25% of allneuroblastoma cases and is strongly associated withadvanced-stage disease (3). However, a significant numberof patients with no MYCN amplification also show poorprognosis (1). Therefore, additional prognostic markers areneeded to further define patient risk groups, particularly inpatients without MYCN amplification.

More recently, it has become clear that the biology ofneuroblastoma is determinednot only by the genetic profilebut also by the epigenetic profile of the tumor. DNAmethylation is a well-characterized epigenetic mechanismand is an essential biochemical process that regulates gene

Authors' Affiliations: 1Children's Cancer Institute Australia for MedicalResearch, Lowy Cancer Research Centre, University of New South Wales,Randwick and 2Adult Cancer Program, Lowy Cancer Research Centre andPrince ofWalesClinical School, University of NewSouthWales, Randwick;and 3Centre for Children's Cancer and Blood Disorders, Sydney Children'sHospital, Randwick, NSW, Australia

Note: Supplementary data for this article are available at Clinical CancerResearch Online (http://clincancerres.aacrjournals.org/).

The Children's Cancer Institute Australia for Medical Research is affiliatedwith the University of New South Wales and Sydney Children's Hospital,Randwick, Australia.

Corresponding Author: Lesley J. Ashton, Children's Cancer InstituteAustralia for Medical Research, Lowy Cancer Research Centre, Universityof New South Wales, P.O. Box 81, Randwick NSW 2031, Australia. Phone:61-2-9385-2162; Fax: 61-2-9662-6583; E-mail:[email protected]

doi: 10.1158/1078-0432.CCR-12-0294

�2012 American Association for Cancer Research.

ClinicalCancer

Research

Clin Cancer Res; 18(20) October 15, 20125690

on April 19, 2018. © 2012 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst August 28, 2012; DOI: 10.1158/1078-0432.CCR-12-0294

http://clincancerres.aacrjournals.org/

transcription and normal cell development. DNA methyl-ation silences gene expression through the addition ofmethyl groups to cytosine residues within CpG-richsequences, known as CpG islands, present in the promoterregion of genes. The availability of methyl groups for DNAmethylation is dependent on folate status, as folate is a keysource of S-adenosyl methionine, a universal methyl donor(4). Folate is also essential for DNA synthesis in rapidlygrowing cells such as that observed in fetal development aswell as cancer. Hence, the bioavailability of folate may alsoenhance the growth of preexisting tumor cells (5). In fact,studies have implicated folate deficiency in several patho-logic diseases, including cancer (6). One of themechanismsby which folate deficiency can promote carcinogenesis is byreduced availability of one-carbon groups required formethylation reactions, which may lead to a decrease inlevels of genomic methylation or DNA hypomethylationand concomitant promoter hypermethylation of specificgenes (7–11). For example, hypomethylation of DNA atspecific sites within the proto-oncogenes c-MYC, FOS, andHRAShas beenobserved in the livers of rats fedwithmethyl-deficient diet (12),whereas other investigationshave shownthat rats with folate ormethyl deficiency induce site-specificmethylationwithin the p53 tumor-suppressor gene, and themethylation of this gene was associated with reducedexpression of p53 (13). Thus, the interplay between folateand DNAmethylation has an important role in normal celldevelopment as well as tumorigenesis.Aberrant DNA methylation at promoter CpG islands is

widely accepted as a common event in a variety of humancancers including neuroblastoma (14). Indeed, a growinglist of aberrantly methylated genes has been described in

neuroblastoma in thepast decade, suggesting a role forDNAmethylation in the tumorigenesis of neuroblastoma.

In this study, themethylation status of 48 candidate genespreviously shown to be the targets of aberrant methylationin embryonal tumors, such as those involved in cell-cycleregulation, apoptosis, and cell differentiation, were deter-mined using a quantitative DNA methylation detectionmethod. Genes involved in the folate-metabolizing path-way were also included because of their role in regulatingthe intracellular pools of folate. We then examined theassociation between levels of DNA methylation in thesegenes and the risk of relapse or death in patients withneuroblastoma to identify additional prognostic markersfor clinical progression.

Materials and MethodsStudy design

Archival DNA was available for 131 children diagnosedwith neuroblastoma in Australia and New Zealand. Treat-ment and clinical data including age at diagnosis, sex,neuroblastoma stage, relapse/death, andMYCN status wereobtained from medical records. Patients were diagnosedbetween1985 and2000and themedian follow-up timewas3 years and 2 months. All children were treated usingstandard protocols according to their tumor stage as previ-ously described (15). Event-free survival (EFS) was definedas the time fromdiagnosis to relapse or deathwithin 5 yearsfrom diagnosis. The study was approved by institutionalethics committees, and informed consent was obtained forpatients enrolled in the study. DNA extraction was con-ducted using QIAamp DNAMini Kit (Qiagen, Inc.) accord-ing tomanufacturer’s instructions. DNAwas eluted in 50 mLof elution buffer.

Selection of candidate genesThe panel of 48 candidate genes examined in this study

was selected on the basis of previous reports of aberrantmethylation in cancer or significant associations with therisk and outcome of cancer, particularly in neuroblastoma.Essential genes involved in the folate-metabolizing pathwaywere also included. Candidate genes were selected usingPubMeth (http://www.pubmeth.org), a publicly accessiblecancer-methylation database that contains a comprehen-sive overview of published information relating to genespreviously reported to be methylated in various cancertypes (16).

Candidate genes examined in the current study are listedin Table 1. Probes selected for the assay were located withinCpG islands, which were identified through the Universityof California Santa Cruz Genome Browser Website (http://genome.ucsc.edu/) and were defined by: (i) GC content of50% or greater, (ii) CpG island length greater than 200 bp,and (iii) the ratio of observed to expected CpG greaterthan 0.6. Where no CpG island was identified for a specificgene, CpG siteswithin 500bpof the transcriptional start siteor promoter region were considered. CpG sites with previ-ously identified polymorphisms listed in public accessible

Translational RelevanceNeuroblastoma is the most common extracranial sol-

id cancer in childhood. Amplification of the MYCNoncogene, tumor stage, and older age at diagnosis areestablished prognostic markers for children with neuro-blastoma. However, only 20% of tumors displayMYCNamplification. Hence, there is a need for additionalmolecular markers for patients lacking MYCN amplifi-cation to enable better stratification of patient riskgroups. Our results showed that hypermethylation offolate hydrolase (FOLH1), myogenic differentiation-1(MYOD1), and thrombospondin-1 (THBS1) are strongpredictors of poorer clinical outcome independent ofMYCN amplification, age, and stage at diagnosis and thata greater number ofmethylated genes increase the risk ofrelapse/death. Our findings highlight the potential useof methylation profiling to identify additional prognos-tic markers in children with neuroblastoma, particularlyin patients without MYCN amplification, and show thepotential prognostic benefit of using a high-throughputcandidate gene approach to rapidly target and quantifylevels of promoter methylation in the clinical setting.

Promoter DNA Methylation in Neuroblastoma

www.aacrjournals.org Clin Cancer Res; 18(20) October 15, 2012 5691




Table 1. A panel of 48 candidate genes examined in this study

GenesMethylationfrequency (%)a

Gene accessionno.

Chromosomelocation

Position of CpGsite relativeto TSSb Gene function

CASP8c 96 NM_001228.3 2q33-q34 120 ApoptosisCCND2 0 NM_001759.2 12p13 �286 Cell-cycle controlCDH1c 90 NM_004360.2 16q22.1 �284 Calcium-dependent cell adhesionCDKN2Ac 68 NM_058195.2 9p21 498 Cell-cycle control; kinase inhibitorCDKN2B 16 NM_004936.3 9p21 302 Cell-cycle control; kinase inhibitorCOL1A2c 100 NM_000089.3 17q21.33 �437 Cell developmentCOMT 0 NM_007310.1 22q11.21 99 Substrate metabolism; catecholamine

metabolism neurotransmitterDAPK1 0 NM_004938.1 9q34.1 �190 ATP-binding, apoptosisDHFR 2 NM_000791.3 5q11.2-q13.2 754 One-carbon metabolismDNAJC15c 70 NM_013238.2 13q14.1 �179 Protein bindingFOLH1c 56 NM_004476.1 11p11.2 122 Folate metabolismGSTP1 30 NM_000852.2 11q13 �322 Metabolic pathwayHIC1 5 NM_006497.2 17p13.3 93 Cell-cycle controlHOXA9 3 NM_152739.2 7p15.2 47 Development regulatorHS3ST2 42 NM_006043.1 16p12 �408 Heparin sulfate glucosaminyl

3-O-sulfotransferase metabolismIGFBP3 1 NM_000598.4 7p13-p12 �231 Growth factorLATS1 8 NM_004690.2 6q25.1 416 Cell-cycle control; kinase activityLATS2 1 NM_014572.1 13q11-q12 �199 Cell-cycle control; kinase activityLHX9 10 NM_020204.2 1q31.1 136 Cell differentiation, brain developmentMGMTc 92 NM_002412.2 10q26 61 DNA repairMTHFR 1 NM_005957.2 1p36.3 2 One-carbon metabolismMTR 5 NM_000254.1 1q43 403 One-carbon metabolismMTRR 24 NM_024010.1 5p15.31 56 One-carbon metabolismMYC 2 NM_002467.3 8q24.21 161 Oncogene, transcription factor activity,

cell differentiation, proliferationMYCN 0 NM_005378.4 2p24.3 �177 Neuroblastoma oncogene, transcription

factor activity, cell differentiationand proliferation

MYOD1 37 NM_002478.3 11p15.4 �124 Transcription regulator; regulatesmuscle cell differentiation

NTRK1c 100 NM_001007792.1 1q21-q22 �16 Tyrosine kinase receptorNTRK2 21 NM_001007097.1 9q22.1 �149 Tyrosine kinase receptorNTRK3c 90 NM_001007156.1 15q25 �37 Tyrosine kinase receptorPYCARDc 86 NM_013258.3 16p11.2 399 ApoptosisRARBc 87 NM_000965.2 3p24 �82 Retinoic acid receptorRASSF1Ac 98 NM_007182.4 3p21.3 18 Tumor suppressor; anaphase inhibitorRB1 0 NM_000321.1 13q14.2 270 Retinoblastoma susceptibility

protein, cell-cycle regulatorS100A6c 99 NM_014624.3 1q21 �34 Cell-cycle regulatorS100A10 1 NM_002966 1q21 22 Cell-cycle regulatorSCGB3A1 14 NM_052863.2 5q35-qter �103 Cell-proliferationSFNc 100 NM_006142.3 1p36.11 �455 Inhibits cell-cycle progressionSLC19A1 22 NM_194255.1 21q22.3 �268 One-carbon metabolismSOCS1 22 NM_003745.1 16p13.13 �64 Kinase bindingSST 14 NM_001048.3 3q28 216 Hormone inhibitor; cell-proliferationTERTc 99 NM_198253.1 5p15.33 �695 Telomeric activityTHBS1 6 NM_003246.2 15q15 �189 Angiogeneis inhibitor

(Continued on the following page)

Lau et al.

Clin Cancer Res; 18(20) October 15, 2012 Clinical Cancer Research5692




database, dbSNP, National Center for BiotechnologyInformation (http://www.ncbi.nlm.nih.gov/projects/SNP/),were excluded from our investigations.For each sample, 1 mg of genomic DNA was modified by

sodium bisulfite using the EZ DNA Methylation Kit (ZymoResearchCorporation) according tomanufacturer’s instruc-tions.Methylation status of all samples was analyzed simul-taneously using a customized GoldenGate Veracode DNAmethylation assay (Illumina), according to manufacturer’sinstructions. Bisulfite-treated DNA was probed at 96 indi-vidual CpG loci within the 48 candidate genes (2 CpG sitesper gene). Fluorescence levels of hybridized samples weredetected using an Illumina BeadXpress platform (Illumina).

Combined bisulfite restriction analysisThe results of the GoldenGate Veracode DNA methyla-

tion assay were validated using combined bisulfite restric-tion analysis (CoBRA) of a selected subgroup of genes in asubset of 48 primary neuroblastoma samples randomlyselected from the original cohort of 131 patients. Briefly,50 ng of bisulfite-modified DNA was amplified by PCRusing 1�AmplitaqGold buffer (Applied Biosystems), 0.5UAmpliTaq Gold (Applied Biosystems), 1.5 mmol/L MgCl2,0.25 mmol/L dNTP, and 1 mmol/L of forward and reverseprimers in a total reaction volumeof 20mL. Seminested PCRwas conducted subsequently using 1 mL of the initial PCRreaction with the same conditions but with 0.4 mmol/L offorward and reverse primers and 1 U of AmpliTaq Gold.CoBRA primer sequences and annealing temperatures arelisted in Supplementary Table S1. Amplified products weresubjected to TaqI or BstUI digestion for the recognition ofTCGA or CGCG sites for 2 hours at 65�C or 60�C, respec-tively, and resolved by gel electrophoresis.

Cloning and direct bisulfite sequencingAs only 2 CpG sites were investigated for each candidate

gene using the GoldenGate Assay, bisulfite sequencing ofclones was used to confirm that the surrounding CpG siteswere also methylated and to examine the level of methyl-

ation heterogeneity, which has been previously reportedacross a range of tumors (17). Methylation status of folatehydrolase (FOLH1), myogenic differentiation-1 (MYOD1),and thrombospondin-1 (THBS1) were confirmed in neu-roblastoma cell lines, such as IMR-32 andNBL-S (AmericanType Culture Collection), as well as a representative subsetof patient samples. Primers were designed to amplify theregion encompassing the CpG site(s) interrogated by theGoldenGate Assay using bisulfite PCR (see SupplementaryTable S2). The PCR products were ligated into the pCR2.1-TOPO vector (Invitrogen), according to manufacturer’sinstructions. Up to 12 individual colonies were chosen forcolony PCRusing theprimers listed in Supplementary TableS2. PCR products were then sequenced to ascertain themethylation status of individual alleles.

Quantitative analysis of methylation levels in CpG-richregions of the genome

Methylation intensity data were evaluated using Geno-meStudio software (Illumina). Background intensityderived from built-in negative controls was subtracted fromeach methylation data point to minimize intra-assay vari-ation. Methylation levels were quantified by the beta value(b), defined as the ratio of fluorescent signal from themethylated allele to the sum of the fluorescent signals ofboth methylated and unmethylated allele. The b-valuerepresented a continuous measure of DNA-methylationlevels in each sample, ranging from 0 in the case ofcompletely unmethylated sites to 1 in completely methyl-ated sites. The average b-value was derived from30 replicatemethylation measurements for each sample.

Statistical analysisStatistical analyses were conducted using STATA version

10 (StataCorp). To see whether methylation levels differedbetween clinical groups, patients were grouped into distinctclinical groups, such as those with MYCN-amplified versusnonamplified tumor, those older than 18months versus 18months or younger, or those with stage IV versus stages I, II,

Table 1. A panel of 48 candidate genes examined in this study (Cont'd )

GenesMethylationfrequency (%)a

Gene accessionno.

Chromosomelocation

Position of CpGsite relativeto TSSb Gene function

TIMP3c 94 NM_000362.4 22q12.3 �579 Tissue inhibitor of metalloproteases,matrix remodeling, tissue invasion

TNFRSF10Ac 54 NM_003844.2 8p21 �91 Death receptor, induce apoptosisTNFRSF10C 38 NM_003841.2 8p22-p21 7 Antiapoptotic decoy receptorsTNFRSF10Dc 76 NM_003840.3 8p21 224 Antiapoptotic decoy receptorsWIF1 0 NM_007191.2 12q14.3 38 Cell signalingZMYND10 19 NM_015896.2 3p21.3 �38 Cell-cycle regulator

aPercentage of samples methylated (b-value > 0.25) in this study. A total of 96 CpG sites or 2 CpG sites per gene were examined. Foreach gene, the CpG site with higher methylation frequency is shown.bAll examined CpG sites were located within CpG islands, except for CASP8 and NTRK1.cGenes that were methylated in more than 50% of all samples.






III, and IVS of tumor. Because the b-value is a continuousmeasure of DNA methylation, the median b-value of eachgroup was compared using Mann–Whitney U tests. Com-parisons between groups with a median difference, |~b|,more than 0.17 and P-values of less than 0.05 were con-sidered significant (18).

For survival analyses, sampleswithb-values of 0.25or lesswere designated as unmethylated, whereas samples withb-values ofmore than0.25were consideredmethylated (19,20). Cumulative EFS was computed by the Kaplan–Meiermethod and compared between subgroups using log-ranktests to determine the association between methylation ofspecific genes and EFS. A Cox proportional hazards modelwas used to examine the influence of hypermethylation ofspecific genes as well as established prognostic factors(MYCN amplification, neuroblastoma stage, and age atdiagnosis) on EFS.

ResultsClinical characteristics of study population

Clinical characteristics of the primary neuroblastomasamples are shown in Supplementary Table S3. Approxi-mately 37% of patients were of ages 18 months or youngerat diagnosis, with amedian age of 18.2months (range: birthto 13 years and 6months). More than 40% of patients werediagnosedwith stage IV neuroblastoma, and 17%of tumorsexhibited amplification ofMYCN. As with previous studies,an increased risk of relapse or death was associated withMYCN amplification [HR ¼ 4.93, 95% confidence interval(CI) ¼ 2.78–8.75, P < 0.001], stage IV disease (HR ¼ 3.96,95% CI ¼ 2.53–6.17, P < 0.001), and being older than 18months at diagnosis (HR¼ 1.85, 95% CI¼ 1.00–3.39, P¼0.048), whereas sex was not predictive of outcome (HR ¼0.75, 95% CI ¼ 0.44–1.29, P ¼ 0.298; refs. 15, 21).

DNA methylation analysesDNAmethylation profiles for replicate samples analyzed

on separate plates displayed highly correlated b-values(Spearman correlation coefficient; r � 0.99). Results fromGoldenGate assays were validated using CoBRA in a subsetof genes (IGFBP3, MTHFR, PYCARD, RASSF1, SFN,SLC19A1, and ZMYND10). The frequencies of DNA meth-ylation were concordant (90%) between the GoldenGateand CoBRA assays (Supplementary Table S4). As shownin Table 1, CASP8, CDH1, CDKN2A, COL1A2, DNAJC15,FOLH1,MGMT, neurotrophic tyrosine kinase receptor 1 orTrkA (NTRK1),NTRK3, PYCARD,RARB,RASSF1A, S100A6,SFN, TERT, TIMP3, TNFRSF10A, and tumor necrosis factorreceptor superfamily, member 10D (TNFRSF10D) werefound to be methylated in more than 50% of primaryneuroblastoma samples. Ninety-eight percent of primarytumors showed hypermethylation (b-value >0.75) of SFNand NTRK1. Other genes including CCND2, CDKN2B,COMT, DAPK1, DHFR, GSTP1, HIC1, HOXA9, HS3ST2,IGFBP3, LATS1, LATS2, human Lim-homeobox 9 (LHX9),MTHFR,MTR,MTRR,MYC,MYCN,MYOD1,NTRK2, RB1,S100A10, SCGB3A1, SLC19A1, SOCS1, SST, THBS1,

TNFRSF10C, WIF1, and ZMYND10 were hypermethylatedin less than 50% of tumor samples. Overall, a total of 15 ofthe 48 genes examined were unmethylated in more than90% of the samples examined (Table 1).

Association between median DNA methylation levelsand clinical characteristics

The levels of methylation observed for the 48 genepromoters were analyzed in patient samples based ontumor stage, age at diagnosis, and MYCN-amplificationstatus. A median level of methylation was determined foreach patient characteristic, with for example, the medianlevel of methylation within MYCN-amplified patients ascompared with the median level of methylation in non-amplified patients. Associations that were statistically sig-nificant at a probability level of more than 0.05 are sum-marized in Table 2. Patients diagnosed at age more than 18months had significantly higher levels of methylation ofDNAJC15, NTRK1, and TNFRSF10D genes, as comparedwith children diagnosed at age 18 months or less (Fig. 1A;P < 0.01). A similar result was also observed for the meth-ylation levels of the DNAJC15,NTRK1, and PYCARD genesin MYCN-amplified samples in comparison with nonam-plified samples (Fig. 1B; P < 0.001). Median levels ofpromoter methylation observed in the remaining genepromoters did not seem to differ based on the individualpatient characteristics examined (P > 0.05).

DNA methylation and patient survivalOverall, patients with promoter hypermethylation of

FOLH1, LHX9,MYOD1, and THBS1 displayed significant-ly lower EFS as compared with those without methylation(log-rank test; P < 0.004; Fig. 2). In patients lackingMYCN amplification, hypermethylation of FOLH1 andMYOD1 was significantly associated with poor outcomeas compared with those without methylation (log-ranktest, P � 0.01; Fig. 2). As shown in Table 3, univariateanalysis showed that patients with overall high levels ofmethylation or hypermethylation of FOLH1, LHX9,MYOD1, or THBS1 had a significantly increased risk ofrelapse or death. In multivariate analyses, associationsremained significant for all genes except LHX9 afteradjusting for MYCN amplification status, age, and stageof disease at diagnosis (Table 3).

Bisulfite sequencing of 2 neuroblastoma cell lines and asubset of primary neuroblastoma samples previously exam-ined in the GoldenGate assay confirmed methylation ofFOLH1,MYOD1, and THBS1within the promoter region orCpG island (Supplementary Fig. S2). We also examinedwhether hypermethylation of 1 or more genes was a stron-ger predictor of outcome in our patient cohort. We focusedon FOLH1,MYOD1, and THBS1 as these genes were strong-ly associated with poorer clinical outcome in the multivar-iate analysis. As shown inTable 4, the risk of relapse or deathwas more than 2 times higher in patients displaying hyper-methylation of at least 2 of these genes after adjusting forMYCN status, stage, and age at diagnosis (HR: 2.72, 95%CI ¼ 1.55–4.78, P ¼ 0.001).

Lau et al.





DiscussionIn this study, we used the GoldenGate Veracode meth-

ylation assay to assess levels of promoter DNAmethylationof 48 genes in 131 patients with neuroblastoma and eval-uated the potential clinical significance of associations

between promoter gene methylation, established prognos-tic risk factors, and risk of relapse or death. We observedhigher levels of promoter methylation of DNAJC15,NTRK1, and TNFRSF10D in older patients, and higherlevels of promoter methylation of DNAJC15, NTRK1, and

1.0

0.8

0.6

0.4

0.2

0.0

DNAJC15cg00948736

A

B

≤18 mo

NMA MA NMA MA NMA MA

>18 mo ≤18 mo >18 mo ≤18 mo >18 mo

NTRK1cg02122575

TNFRSF10Dcg01031400

DNAJC15cg12012021

NTRK1cg02122575

PYCARDcg05898613

-val

ue-v

alue

1.0

0.8

0.6

0.4

0.2

0.0

Figure 1. Comparisons of median b-values or methylation levels by age-group (A) andMYCN amplification status (B; ��,P < 0.001; ��,P < 0.01). Guide for boxplot: top and bottom hinges of the box represent 75th percentile and 25th percentile, respectively; whiskers indicate the highest and lowest values; closedcircles represent outliers; thick horizontal line within the box indicates the median b-value. NMA, non-MYCN amplified; MA, MYCN amplified.

Table 2. Difference in median levels of promoter-DNA methylation based on clinical characteristics

Gene CpG siteStageIV

Stage I, II, IIIand IVS Dba

>18months

�18months Dba

MYCNamplified

Non-MYCNamplified Dba

DNAJC15 cg00948736 0.52 0.47 0.05 0.55 0.28 0.27c 0.86 0.45 0.41d

cg12012021 0.28 0.31 �0.03 0.39 0.23 0.16 0.73 0.26 0.47d

NTRK1 cg02122575 0.61 0.53 0.08 0.63 0.42 0.21d 0.73 0.50 0.23d

cg25827666 0.96 0.96 0.00 0.96 0.96 0.01b 0.97 0.96 0.01PYCARD cg05898613 0.59 0.55 0.04 0.16 0.12 0.04b 0.63 0.10 0.53d

cg03345696 0.16 0.13 0.03 0.52 0.62 �0.11 0.66 0.54 0.12c

TNFRSF10D cg05763426 0.40 0.36 0.04 0.22 0.39 0.05 0.58 0.42 0.16cg01031400 0.42 0.43 �0.01 0.44 0.24 0.48c 0.62 0.28 0.33

NOTE: P-values calculated using the Mann–Whitney U test to compare median b-values from each clinical group.aDb-values shown in bold are comparisons that were considered statistically significant where the differences in b-values >|0.17| andwith P-values of <0.05.bP � 0.05.cP < 0.01.dP < 0.001.






PYCARD in patients with MYCN-amplified tumors. Ourinvestigations also showed that promoter hypermethyla-tion of FOLH1, MYOD1, and THBS1 were independentpredictors of outcome after adjusting for MYCN amplifica-

tion, age at diagnosis, and tumor stage. Moreover, morethan 30%of patients displayedpromoter hypermethylationin at least 2 of these genes andweremore than 2 timesmorelikely to progress than those who did not display promoter

No. at risk Unmethylated Methylated






A B C

D E F

FOLH1 FOLH1

non-MYCN amplifiedFOLH1

MYCN-amplified

LHX9

P = 0.002 P = 0.012 P = 0.03

P = 0.004 P = 0.09

P = 0.93

LHX9 MYCN-amplified

LHX9 non-MYCN amplified


G H I


MYOD1

P < 0.001 P = 0.002


P = 0.02

MYOD1 MYCN-amplified

MYOD1 non-MYCN amplified




J K L THBS1

P = 0.004 P = 0.407 P < 0.001

THBS1 MYCN-amplified

THBS1 non-MYCN amplified

Figure 2. Kaplan–Meier survival curves for patients with neuroblastoma according to methylation status of FOLH1, LHX9,MYOD1, and THBS1 in all patientswith neuroblastoma and in patients with non-MYCN amplified and MYCN-amplified neuroblastoma. Patients with b-values � 0.25 were designated asunmethylated (solid line), whereas b-values >0.25 were considered methylated (dash line).

Lau et al.





hypermethylation after adjusting for known prognosticfactors.As with previous studies, CASP8, CDKN2A, CDH1,

PYCARD, RASSF1A, SFN, and TIMP3 were found to behypermethylated in 68% to 100% of primary neuroblasto-ma samples (22), whereas gene promoters that were notpreviously investigated for methylation levels in neuroblas-toma but shown to bemethylated in other pediatric tumors(23–26) such as COL1A2, DNAJC15, NTRK1, NTRK3,RARB, S100A6, and TERT were also found to be hyper-methylated in 70% to 100% of neuroblastoma samples.Genes previously reported to bemethylated in adult tumorssuch as CCND2, COMT, DAPK1, RB1, and WIF1 were notfound to be hypermethylated in any of the 131 neuroblas-toma tumors examined, suggesting that levels of promotermethylation in pediatric and adults tumors differ.Levels of promoter methylation of DNAJC15, NTRK1,

and TNFRSF10D were significantly higher in older patients

at diagnosis (P < 0.01), whereas higher levels of promotermethylation of DNAJC15, NTRK1, and PYCARD wereobserved in patients with MYCN amplification (P <0.001). Previous studies have shown that the transcriptionalsilencing of DNAJC15, also known as methylation-con-trolled J (MCJ), is epigenetically regulated by methylation(25). Hypermethylation of this gene has also been observedin pediatric brain tumors andWilms’ tumors and in ovariancancers that displayed chemotherapeutic resistance (25,27, 28). Although these cancers are biologically differentfrom neuroblastoma, amplification of the proto-oncogenes,such as c-myc,MYCN, or L-mychave beenobserved in a smallproportion of these tumors (29–31), suggesting a possibleinteraction between proto-oncogenes and methylation thatmay contribute to the tumorigenesis of these cancers.

Our finding that hypermethylation of the NTRK1 pro-moter was positively associated with MYCN amplificationis also consistent with previous reports showing that

Table 3. Associations between DNA methylation and EFS in children with neuroblastoma

Univariate Multivariatec

Variables Total (%) Events (%) HR (95% CI) P HR (95% CI) P

MYCN amplificationAbsent 109 (83.2) 36 (65.5) 1.00 1.00Present 22 (16.8) 19 (34.6) 4.93 (2.78–8.75) <0.001 3.59 (2.27–5.67) <0.001

Neuroblastoma stagea

Stage I, II, III, IVS 75 (59.1) 18 (33.3) 1.00 1.00Stage IV 52 (40.9) 36 (66.6) 3.96 (2.53–6.17) <0.001 3.59 (2.26–5.73) <0.001

Age at diagnosis� 18 months 48 (36.6) 14 (25.5) 1.00 1.00> 18 months 83 (63.4) 41 (74.6) 1.85 (1.00–3.39) 0.048 1.06 (0.66–1.72) 0.804

SexMale 72 (55.0) 33 (60.0) 1.00 1.00Female 59 (45.0) 22 (40.0) 0.75 (0.44–1.29) 0.298 1.32 (0.86–2.03) 0.199

Overall methylationb

Low 66 (50.4) 21 (38.2) 1.00 1.00High 65 (49.6) 34 (61.8) 1.94 (1.13–3.35) 0.017 1.48 (0.83–2.64) 0.181

FOLH1UM 57 (43.5) 15 (27.3) 1.00 1.00M 74 (56.5) 40 (72.7) 2.50 (1.38–4.54) 0.003 2.27 (1.23–4.19) 0.009

LHX9UM 118 (90.1) 45 (81.8) 1.00 1.00M 13 (9.9) 10 (18.2) 2.67 (1.34–5.32) 0.005 1.77 (0.82–3.81) 0.146

MYOD1UM 82 (62.6) 24 (43.6) 1.00 1.00M 49 (37.4) 31 (56.4) 2.91 (1.70–4.98) <0.001 2.28 (1.30–4.00) 0.004

THBS1UM 123 (93.9) 49 (89.1) 1.00 1.00M 8 (6.1) 6 (10.9) 3.31 (1.41–7.76) 0.006 3.05 (1.22–7.63) 0.017

Abbreviations: M, methylated (b-value >0.25); UM, unmethylated (b-value �0.25).aThe exclusion of stage IVS patients did not change the statistical significance of the analysis.bFor each sample, the average b-value was derived from all 96 CpG sites and was grouped into "low" or "high" methylation grouparound the median b-value.cVariables adjusted for MYCN status, neuroblastoma stage, and age at diagnosis.






expression of NTRK1 is negatively correlated with MYCNamplification and thatNTRK1 gene expression is associatedwith favorable neuroblastoma tumors that regress or dif-ferentiate (32, 33). Hypermethylation of the proapoptoticgene PYCARD (PYD andCARDdomain-containing protein,also known as TMS1) has also been previously reported tobe associatedwithMYCN amplification and advanced-stageneuroblastoma (34). However, no associations wereobserved between the levels of methylation of NTRK1 orPYCARD and the clinical outcome in our study. While thereasons for discordant results are not clear, they are likelydue to differences in patient cohorts, variation in methodsto detect methylation, and disparities in the regions ofNTRK1 and PYCARD analyzed.

Although the biologic significance of TNFRSF10D incarcinogenesis is unclear, previous studies have shownmethylation of TNFRSF10D to be associated with reducedEFS and overall survival in patients with neuroblastomaindependent of MYCN amplification (35, 36). Despite theabsence of this association in our study, higher levels ofTNFRSF10D methylation seen in older patients providedsome evidence that TNFRSF10D methylation may have arole in influencing the clinical outcome of older neuroblas-toma patients.

We identified 3 genes that displayed promoter hyper-methylation and independently predicted an increased riskof relapse or death. FOLH1 encodes a protein that hydro-lyses natural food folates from a polyglutamated state to amonoglutamated form before absorption can occur (37).Hypermethylation of FOLH1 has been shown to be corre-lated with chromosome 17q gain, a genetic abnormalityoften observed in neuroblastoma, as well as weakly asso-ciated with an increased risk of death (38). AlthoughFOLH1 is not directly involved in one-carbon folate metab-olism, studies have reported that polymorphisms in theFOLH1 gene can result in impaired intestinal absorption ofdietary folates, leading to lowblood-folate levels and hyper-homocysteinemia (37, 39). Hence, methylation-mediatedinactivation of FOLH1 may provide an alternative mecha-nism for impaired folate absorption, and further studiesexamining the impact of FOLH1-promoter methylation inpatients with neuroblastoma are warranted.

Higher levels of promoter methylation in MYOD1 andTHBS1 also independently predicted an increased risk orrelapse or death in our cohort. MYOD1 encodes for atranscription factor that shares homology to theMYC familyof genes, such as c-myc, which is exclusively expressed infetal- or adult-skeletal muscle (40), whereas THBS1 is aninhibitor of angiogenesis and has previously been shown tobe hypermethylated and silenced in primary neuroblasto-ma tumor samples and cell lines (41, 42). De novo meth-ylation of the MYOD1 CpG islands has been observedduring the establishment of immortal cell lines, suggestingthat silencing of MYOD1 via promoter hypermethylationmay lead to immortalization andoncogenic transformation(43). Although MYOD1 has been reported to be transcrip-tionally inactive in neuroblastoma (44), to our knowledge,promoter methylation of MYOD1 has not yet been exam-ined in this malignancy. Despite in vitro studies showingrestoration of THBS1 gene expression in neuroblastomacells following treatment with a demethylating agent (45),clinical studies have not been able to detect any associationbetween methylation levels of THBS1 and survival inpatients with neuroblastoma (42, 46).

In univariate analysis, higher levels of methylation ofLHX9 were found to be associated with an increased risk ofrelapse or death. However, this apparent association dis-appeared after adjusting for MYCN amplification. Never-theless, the LHX9-gene promoter may be a potential targetfor demethylating agents in patients withMYCN-amplifiedtumors, particularly as it encodes for a transcription factorinvolved in the control of neuronal differentiation aswell asbrain development (47). Moreover, methylation-mediatedsilencing of LHX9 is frequently observed in pediatric malig-nant astrocytomas, themost common form of glioma (48).To our knowledge, analyses of promoter methylation inLHX9 have not been reported in patients with neuroblas-toma. Hence, further studies are required to fully elucidatethe function of LHX9-promoter methylation in MYCN-amplified neuroblastoma tumors.

Of the 3 genes identified, hypermethylation of at least 2genes was associated with an increased risk of relapse ordeath in patients with neuroblastoma. These results suggestthe coordinated methylation of several gene loci or a CpG

Table 4. Combined analysis of FOLH1, MYOD1, and THBS1 methylation and EFS in children withneuroblastoma

Univariate Multivariatec

N of genes methylated Totala (%) Eventsb (%) HR (95% CI) P HR (95% CI) P

�1 84 (64.1) 43 (51.2) 2.42 (1.27–4.59) 0.007 2.16 (1.12–4.17) 0.022�2 43 (32.8) 30 (69.8) 3.62 (2.11–6.12) <0.001 2.72 (1.55–4.78) 0.001¼3 4 (3.1) 4 (100.0) 7.28 (2.57–20.68) <0.001 4.51 (1.56–13.09) 0.006

aNumber of patients with �1, �2, and ¼ 3 genes methylated in a total cohort of 131 patients. There were no patients with all 5 genesmethylated.bThe percentage of events is calculated by the number of events within patient groups of having �1, �2, or ¼ 3 genes methylated.cVariables adjusted for MYCN status, neuroblastoma stage, and age at diagnosis.

Lau et al.





island methylator phenotype (CIMP). Previous investiga-tionshave reported thatmethylationof theprotocadherin-b(PCDHB) gene family, either alone or in combination withmethylation of the hepatocyte growth factor-like protein(HLP) and cytochrome p450 (CYP26C1) genes, is a poten-tial CIMP associated with poorer survival in patients withneuroblastoma (49, 50). Our investigations implicate addi-tional genes that may provide an improved CIMP for pre-dicting the outcome of neuroblastoma.While previous reports have shown the presence ofmeth-

ylation-mediated silencing in neuroblastoma, the frequen-cy ofmethylation has been shown to vary between differentstudies, possibly due to the different techniques usedbetween studies as reviewed in ref. 51. The methylationdetection method used in our study was both sensitive andquantitative, whereas several other techniques, such asmethylation-specific PCR or CoBRA that are commonlyused in research studies are nonquantitative or semiquan-titative, and other quantitative methods, such as pyrose-quencing or bisulfite sequencing can be costly and laborintensive. Hence, uniformmethods or scoring systems needbe established to improve comparison of results betweenlaboratories. The Illumina GoldenGate assay provides astandardized method where specific primers and probeshave been predesigned to interrogate CpG sites that areindividually assigned with a unique identifying code andallow direct comparisons between laboratories. Therefore,this method has potential to be used in a clinical setting forprognostic evaluation of patients.Gene associations found in the present study may con-

tribute to improved prediction of clinical outcomes, espe-cially in patients without MYCN amplification. Our studyprovides strong evidence to support the hypothesis thatepigenetic changes in multiple genes have the capacity to

alter the clinical phenotype of neuroblastoma and that theincreasing number of methylated genes increases the risk ofrelapse or death. While further studies are required todelineate the full phenotypic consequences of DNA meth-ylation in these and other gene promoters, our findingshighlight the potential use of methylation profiling toprovide additional prognostic information and detect newtherapeutic targets for selected patient subsets. The estab-lishment of a rapid standardized molecular approach toassess gene-promoter–methylation status of neuroblasto-ma tumors will be essential for the translation of these andother prognostic findings into the clinical setting.

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

Authors' ContributionsConception and design: D.T. Lau, L.B. Hesson, L.J. AshtonDevelopment of methodology: L.B. Hesson, M. Haber, L.J. AshtonAcquisitionofdata (provided animals, acquired andmanagedpatients,provided facilities, etc.): D.T. Lau, L.J. AshtonAnalysis and interpretation of data (e.g., statistical analysis, biosta-tistics, computational analysis):D.T. Lau, L.B. Hesson, G.M. Marshall, M.Haber, L.J. AshtonWriting, review, and/or revision of the manuscript: D.T. Lau, L.B.Hesson, M.D. Norris, G.M. Marshall, M. Haber, L.J. AshtonAdministrative, technical, or material support (i.e., reporting or orga-nizing data, constructing databases): D.T. LauStudy supervision: L.B. Hesson, L.J. Ashton

Grant SupportD.T. Lau was supported by the National Health and Medical Research

Council Public Health Postgraduate Research Scholarship.The costs of publication of this article were defrayed in part by the

payment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received January 26, 2012; revised July 27, 2012; accepted August 18,2012; published OnlineFirst August 28, 2012.

References1. Brodeur GM, Maris JM.Neuroblastoma. In:Pizzo PA, Poplack DG,

editors. Principles and practice of pediatric oncology. 5th ed. Phila-delphia: Lippincott Williams & Wilkins; 2006. p. 933–70.

2. Schwab M, Westermann F, Hero B, Berthold F. Neuroblastoma:biology and molecular and chromosomal pathology. Lancet Oncol2003;4:472–80.

3. Brodeur GM, Seeger RC. Gene amplification in human neuroblasto-mas: basic mechanisms and clinical implications. Cancer Genet Cyto-genet 1986;19:101–11.

4. Hamid A,WaniNA, Kaur J.Newperspectives on folate transport in relat-ion to alcoholism-induced folate malabsorption–association with epi-genome stability and cancer development. FEBS J 2009;276:2175–91.

5. Ulrich CM, Potter JD. Folate supplementation: too much of a goodthing? Cancer Epidemiol Biomarkers Prev 2006;15:189–93.

6. Giovannucci E. Epidemiologic studies of folate and colorectal neo-plasia: a review. J Nutr 2002;132:2350S–5S.

7. JacobRA,Gretz DM, Taylor PC, JamesSJ, Pogribny IP,Miller BJ, et al.Moderate folate depletion increases plasma homocysteine anddecreases lymphocyte DNA methylation in postmenopausal women.J Nutr 1998;128:1204–12.

8. Crott JW, Liu Z, Keyes MK, Choi SW, Jang H, Moyer MP, et al.Moderate folate depletionmodulates the expression of selected genesinvolved in cell cycle, intracellular signaling and folate uptake in humancolonic epithelial cell lines. J Nutr Biochem 2008;19:328–35.

9. Ciappio ED, Mason JB, Crott JW. Maternal one-carbon nutrient intakeand cancer risk in offspring. Nutr Rev 2011;69:561–71.

10. Dizik M, Christman JK, Wainfan E. Alterations in expression andmethylation of specific genes in livers of rats fed a cancer promotingmethyl-deficient diet. Carcinogenesis 1991;12:1307–12.

11. Choi SW, Mason JB. Folate and carcinogenesis: an integratedscheme. J Nutr 2000;130:129–32.

12. Christman JK, Sheikhnejad G, Dizik M, Abileah S, Wainfan E. Revers-ibility of changes in nucleic acid methylation and gene expressioninduced in rat liver by severe dietarymethyl deficiency.Carcinogenesis1993;14:551–7.

13. Pogribny IP, Miller BJ, James SJ. Alterations in hepatic p53 genemethylation patterns during tumor progression with folate/methyldeficiency in the rat. Cancer Lett 1997;115:31–8.

14. Jones PA, Baylin SB. The fundamental role of epigenetic events incancer. Nat Rev Genet 2002;3:415–28.

15. Ashton LJ, Murray JE, Haber M, Marshall GM, Ashley DM, Norris MD.Polymorphisms in genes encoding drug metabolizing enzymes andtheir influence on the outcome of children with neuroblastoma. Phar-macogenet Genomics 2007;17:709–17.

16. Ongenaert M, Van Neste L, DeMeyer T, Menschaert G, Bekaert S, VanCriekinge W. PubMeth: a cancer methylation database combiningtext-mining and expert annotation. Nucleic Acids Res 2008;36:D842–6.






17. Mikeska T, Candiloro IL, Dobrovic A. The implications of heteroge-neous DNAmethylation for the accurate quantification of methylation.Epigenomics 2010;2:561–73.

18. Bibikova M, Lin Z, Zhou L, Chudin E, Garcia E, Wu B, et al. High-throughput DNA methylation profiling using universal bead arrays.Genome Res 2006;16:383–93.

19. Wilop S, Fernandez AF, Jost E, Herman JG, Brummendorf TH, EstellerM, et al. Array-based DNA methylation profiling in acute myeloidleukaemia. Br J Haematol 2011;155:65–72.

20. Martin-Subero JI, Kreuz M, Bibikova M, Bentink S, Ammerpohl O,Wickham-Garcia E, et al. New insights into the biology and origin ofmature aggressiveB-cell lymphomas by combined epigenomic, geno-mic, and transcriptional profiling. Blood 2009;113:2488–97.

21. AmbrosPF,Ambros IM,BrodeurGM,HaberM,KhanJ,NakagawaraA,et al. International consensus for neuroblastoma molecular diagnos-tics: report from the International Neuroblastoma Risk Group (INRG)Biology Committee. Br J Cancer 2009;100:1471–82.

22. Decock A, Ongenaert M, Vandesompele J, Speleman F. Neuroblas-toma epigenetics: from candidate gene approaches to genome-widescreenings. Epigenetics 2011;6:962–70.

23. Yie SM, Hu Z. Human leukocyte antigen-G (HLA-G) as a marker fordiagnosis, prognosis and tumor immune escape in human malignan-cies. Histol Histopathol 2011;26:409–20.

24. Anderton J, Lindsey J, LusherM, Gilbertson R, Bailey S, Ellison D, et al.Global analysis of themedulloblastoma epigenome identifies disease-subgroup-specific inactivation of COL1A2. Neuro Oncol 2008;10:981–94.

25. Lindsey JC, LusherME, StrathdeeG,BrownR,GilbertsonRJ, Bailey S,et al. Epigenetic inactivation of MCJ (DNAJD1) in malignant paediatricbrain tumours. Int J Cancer 2006;118:346–52.

26. Olk-Batz C, Poetsch AR, Nollke P, Claus R, Zucknick M, Sandrock I,et al. Aberrant DNA methylation characterizes juvenile myelomonocy-tic leukemia with poor outcome. Blood 2011;117:4871–80.

27. Ehrlich M, Jiang G, Fiala E, Dome JS, Yu MC, Long TI, et al. Hypo-methylation and hypermethylation of DNA inWilms tumors. Oncogene2002;21:6694–702.

28. Shridhar V, Bible KC, Staub J, Avula R, Lee YK, Kalli K, et al. Loss ofexpression of a new member of the DNAJ protein family confersresistance to chemotherapeutic agents used in the treatment ofovarian cancer. Cancer Res 2001;61:4258–65.

29. Debiec-Rychter M, Alwasiak J, Papierz W, Liberski PP. Expression ofN-myc, c-myc and c-erbB-1 proto-oncogenes in cerebral primitiveneuroectodermal tumors (PNET). Folia Neuropathol 1994;32:229–31.

30. Williams RD, Al-Saadi R, Chagtai T, Popov S, Messahel B, Sebire N,et al. Subtype-specific FBXW7mutation andMYCN copy number gainin Wilms' tumor. Clin Cancer Res 2010;16:2036–45.

31. Wu R, Lin L, Beer DG, Ellenson LH, Lamb BJ, Rouillard JM, et al.Amplification and overexpression of the L-MYC proto-oncogene inovarian carcinomas. Am J Pathol 2003;162:1603–10.

32. Nakagawara A, Arima M, Azar CG, Scavarda NJ, Brodeur GM. Inverserelationship between trk expression andN-myc amplification in humanneuroblastomas. Cancer Res 1992;52:1364–8.

33. Nakagawara A, Arima-Nakagawara M, Scavarda NJ, Azar CG, CantorAB, Brodeur GM. Association between high levels of expression of theTRK gene and favorable outcome in human neuroblastoma. N Engl JMed 1993;328:847–54.

34. Grau E, Martinez F, Orellana C, Canete A, Yanez Y, Oltra S, et al.Hypermethylation of apoptotic genes as independent prognosticfactor in neuroblastoma disease. Mol Carcinog 2010;50:153–62.

35. Banelli B, Gelvi I, Di Vinci A, Scaruffi P, Casciano I, Allemanni G, et al.Distinct CpGmethylation profiles characterize different clinical groupsof neuroblastic tumors. Oncogene 2005;24:5619–28.

36. Yagyu S, Gotoh T, Iehara T, Miyachi M, Katsumi Y, Tsubai-Shimizu S,et al. Circulating methylated-DCR2 gene in serum as an indicator ofprognosis and therapeutic efficacy in patients with MYCN nonampli-fied neuroblastoma. Clin Cancer Res 2008;14:7011–9.

37. DeVos L, Chanson A, Liu Z, Ciappio ED, Parnell LD, Mason JB, et al.Associations between single nucleotide polymorphisms in folateuptake and metabolizing genes with blood folate, homocysteine, andDNA uracil concentrations. Am J Clin Nutr 2008;88:1149–58.

38. AlaminosM, Davalos V, Cheung NK, GeraldWL, Esteller M. Clusteringof gene hypermethylation associated with clinical risk groups in neu-roblastoma. J Natl Cancer Inst 2004;96:1208–19.

39. Eklof V, VanGuelpen B, Hultdin J, Johansson I, HallmansG, PalmqvistR. The reduced folate carrier (RFC1) 80G > A and folate hydrolase 1(FOLH1) 1561C > T polymorphisms and the risk of colorectal cancer: anested case-referent study. Scand J Clin Lab Invest 2008;68:393–401.

40. EdmondsonDG,Olson EN. A genewith homology to themyc similarityregion of MyoD1 is expressed during myogenesis and is sufficient toactivate the muscle differentiation program. Genes Dev 1989;3:628–40.

41. Yang QW, Liu S, Tian Y, Salwen HR, Chlenski A, Weinstein J, et al.Methylation-associated silencing of the thrombospondin-1 gene inhuman neuroblastoma. Cancer Res 2003;63:6299–310.

42. Yang Q, Zage P, Kagan D, Tian Y, Seshadri R, Salwen HR, et al.Association of epigenetic inactivation of RASSF1A with poor outcomein human neuroblastoma. Clin Cancer Res 2004;10:8493–500.

43. Jones PA,Wolkowicz MJ, Rideout WM III, Gonzales FA, Marziasz CM,Coetzee GA, et al. De novo methylation of the MyoD1 CpG islandduring the establishment of immortal cell lines. Proc Natl Acad SciU S A 1990;87:6117–21.

44. Dias P, ParhamDM, Shapiro DN,Webber BL, Houghton PJ. Myogenicregulatory protein (MyoD1) expression in childhood solid tumors:diagnostic utility in rhabdomyosarcoma. Am J Pathol 1990;137:1283–91.

45. Yang Q, Tian Y, Ostler K, Chlenski A, Guerrero L, Salwen H, et al.Epigenetic alterations differ in phenotypically distinct human neuro-blastoma cell lines. BMC Cancer 2010;10:286.

46. Gonzalez-GomezP,BelloM,Lomas J,ArjonaD,AlonsoM,AminosoC,et al. Aberrant methylation of multiple genes in neuroblastic tumours:relationship with MYCN amplification and allelic status at 1p. Eur JCancer 2003;39:1478–85.

47. Retaux S, Rogard M, Bach I, Failli V, Besson MJ. Lhx9: a novel LIM-homeodomain gene expressed in the developing forebrain. J Neurosci1999;19:783–93.

48. Vladimirova V, Mikeska T, Waha A, Soerensen N, Xu J, Reynolds PC,et al. Aberrant methylation and reduced expression of LHX9 in malig-nant gliomas of childhood. Neoplasia 2009;11:700–11.

49. Abe M, Ohira M, Kaneda A, Yagi Y, Yamamoto S, Kitano Y, et al. CpGislandmethylator phenotype is a strong determinant of poor prognosisin neuroblastomas. Cancer Res 2005;65:828–34.

50. Abe M, Westermann F, Nakagawara A, Takato T, SchwabM, UshijimaT. Marked and independent prognostic significance of the CpG islandmethylator phenotype in neuroblastomas. Cancer Lett 2007;247:253–8.

51. Laird PW.Thepower and thepromise ofDNAmethylationmarkers.NatRev Cancer 2003;3:253–66.

Lau et al.





2012;18:5690-5700. Published OnlineFirst August 28, 2012.Clin Cancer Res Diana T. Lau, Luke B. Hesson, Murray D. Norris, et al. with Childhood NeuroblastomaPrognostic Significance of Promoter DNA Methylation in Patients

Updated version

10.1158/1078-0432.CCR-12-0294doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2012/08/28/1078-0432.CCR-12-0294.DC1

Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/18/20/5690.full#ref-list-1

This article cites 50 articles, 18 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/18/20/5690.full#related-urls

This article has been cited by 3 HighWire-hosted articles. Access the articles at:

E-mail alerts related to this article or journal.Sign up to receive free email-alerts

Subscriptions

Reprints and

[email protected]

To order reprints of this article or to subscribe to the journal, contact the AACR Publications Department at

Permissions

Rightslink site. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC)

.http://clincancerres.aacrjournals.org/content/18/20/5690To request permission to re-use all or part of this article, use this link



http://clincancerres.aacrjournals.org/lookup/doi/10.1158/1078-0432.CCR-12-0294

http://clincancerres.aacrjournals.org/content/suppl/2012/08/28/1078-0432.CCR-12-0294.DC1

http://clincancerres.aacrjournals.org/content/18/20/5690.full#ref-list-1

http://clincancerres.aacrjournals.org/content/18/20/5690.full#related-urls

http://clincancerres.aacrjournals.org/cgi/alerts

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/18/20/5690


prognostic signiﬁcance of promoter dna methylation in...

Documents