Tumor-AssociatedMacrophagesinSHHSubgroup of … · Medulloblastoma is the most common malignant brain tumor in children. Craniospinal irradiation still remains an essential component

Biology of Human Tumors

Tumor-AssociatedMacrophages in SHHSubgroupof MedulloblastomasAshley S. Margol1,2, Nathan J. Robison1,2, Janahan Gnanachandran1, Long T. Hung1,Rebekah J. Kennedy1, Marzieh Vali1, Girish Dhall1,2, Jonathan L. Finlay1,2,Anat Erdreich-Epstein1,2,3, Mark D. Krieger1,2, Rachid Drissi4, Maryam Fouladi4,Floyd H. Gilles1,2, Alexander R. Judkins1,3, Richard Sposto1,2, andShahab Asgharzadeh1,2,3

Abstract

Purpose: Medulloblastoma in children can be categorizedinto at least four molecular subgroups, offering the potentialfor targeted therapeutic approaches to reduce treatment-relatedmorbidities. Little is known about the role of tumor microen-vironment in medulloblastoma or its contribution to thesemolecular subgroups. Tumor microenvironment has beenshown to be an important source for therapeutic targets inboth adult and pediatric neoplasms. In this study, we investi-gated the hypothesis that expression of genes related to tumor-associated macrophages (TAM) correlates with the medullo-blastoma molecular subgroups and contributes to a diagnosticsignature.

Methods: Gene-expression profiling using human exon array(n ¼ 168) was analyzed to identify medulloblastoma molecularsubgroups and expression of inflammation-related genes. Expres-sion of 45 tumor-related and inflammation-related genes wasanalyzed in 83 medulloblastoma samples to build a gene signa-

ture predictive of molecular subgroups. TAMs in medulloblasto-mas (n ¼ 54) comprising the four molecular subgroups wereassessed by immunohistochemistry (IHC).

Results: A 31-gene medulloblastoma subgroup classificationscore inclusive of TAM-related genes (CD163 and CSF1R) wasdeveloped with a misclassification rate of 2%. Tumors in theSonic Hedgehog (SHH) subgroup had increased expressionof inflammation-related genes and significantly higher infiltra-tion of TAMs than tumors in the Group 3 or Group 4 subgroups(P < 0.0001 and P < 0.0001, respectively). IHC data revealed astrong association between location of TAMs and proliferatingtumor cells.

Conclusions: These data show that SHH tumors have a uniquetumor microenvironment among medulloblastoma subgroups.The interactions of TAMs and SHH medulloblastoma cells maycontribute to tumor growth revealing TAMs as a potential ther-apeutic target. Clin Cancer Res; 21(6); 1457–65. �2014 AACR.

IntroductionThe role of inflammation in promoting tumor growth and

regulation of the antitumor immune response is an importantcharacteristic of cancer (1–3). Tumor-associated macrophages(TAM) are major contributors to the tumor microenvironmentand are present in a variety of human cancers (4, 5). TAMs are nowknown to promote cancer via multiple mechanisms, includingeffects on tumor cell growth, survival, invasion,metastasis, angio-genesis, inflammation, and immunoregulation (6, 7). The pres-

ence of TAMs has been described in many adult malignancies,including central nervous system tumors (8–12). The first evi-dence of TAMs in childhood tumors and its correlation withtumor stage was recently shown in neuroblastoma, a peripheralnervous system tumor (13). However, little is known about therole of TAMs and the tumor microenvironment in childhoodbrain tumors.

Medulloblastoma is the most common malignant braintumor in children. Craniospinal irradiation still remains anessential component of multimodality therapy for many youngchildren putting these patients at risk for developmental neu-rotoxicity (14, 15). Medulloblastomas are a molecularly het-erogeneous group of tumors that can be classified into at leastfour distinct subgroups: WNT, SHH (Sonic Hedgehog), Group3, and Group 4 (16–19). Recent data have provided insightinto the biology (20) and prognosis of medulloblastomas(21, 22); however, relatively little is known about the role ofthe tumor microenvironment with respect to these molecularsubgroups. Large-scale genomic and gene-expression analysesusing a number of different platforms have been shownto accurately identify medulloblastoma subgroups, but imple-mentation in real-time for clinical application remains a chal-lenge (16–19, 23–26).

In this study, we developed a novel and clinically applicable31-gene TaqMan Low Density Array (TLDA) signature that

1Children's Hospital Los Angeles and The Saban Research Institute,Los Angeles, California. 2Department of Pediatrics, Keck School ofMedicine of University of Southern California, Los Angeles, California.3Department of Pathology, Keck School of Medicine of University ofSouthern California, Los Angeles, California. 4Cincinnati Children'sHospital Medical Center, Cincinnati, Ohio.

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

Corresponding Author: Shahab Asgharzadeh, Children's Hospital Los Angelesand The Saban Research Institute, Keck School of Medicine of University ofSouthern California, 4650 Sunset Boulevard, MS 57, Los Angeles, CA 90027.Phone: 323-361-8643; Fax: 323-361-4902; E-mail: [email protected].

doi: 10.1158/1078-0432.CCR-14-1144

�2014 American Association for Cancer Research.

ClinicalCancerResearch

www.aacrjournals.org 1457

on February 7, 2021. © 2015 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst October 24, 2014; DOI: 10.1158/1078-0432.CCR-14-1144

http://clincancerres.aacrjournals.org/

allows accurate identification of medulloblastoma subgroupsand examined the expression of inflammation-related geneswith respect to each subgroup. We identified evidence of intra-tumoral inflammation and presence of TAMs in the SHH sub-group of medulloblastomas correlating with regions of prolifer-ation. Our findings provide the first report of TAMs in pediatricbrain tumors and introduce the possibility of using targetedtherapies against TAMs, thereby decreasing the reliance on currentradiation-based therapies, which are associated with potentiallydevastating long-termneurocognitive sequelae in young children.

Materials and MethodsSamples were collected from patients with medulloblastoma

(primary samples n ¼ 85 and relapse samples n ¼ 2) treated atChildren's Hospital Los Angeles (CHLA; Los Angeles, CA) orCincinnati Children's Hospital Medical Center (Cincinnati, OH)between 1989 and 2012 with available adequate fresh-frozentissue for evaluation. All samples underwent pathologic review bytwo neuropathologists to confirm the diagnosis. The patient andtumor characteristics are provided in Supplementary Table S1.Sixty-five samples underwent Affymetrix Human Exon 1.0 STArray (HuEx) analysis. The data from these 65 HuEx data wereanalyzed in combination with data from a previously publishedcohort of 103 samples (Supplementary Fig. S1 and Supplemen-tary Table S3; ref. 18). Additional 36 samples and a subset ofHuEx samples with sufficient RNA (n ¼ 47 of 65) were analyzedusing a custom medulloblastoma-specific TLDA assay (totaln ¼ 83, Supplementary Table S2). The details of analyses per-formed on the HuEx microarray and the custom TLDA assay dataare provided in the Supplementary Materials and Methods. Inbrief, themolecular subgroups were identified in an unsupervisedmanner using the HuEx data and performing 1,000 runs ofnonnegative matrix factorization (NMF) clustering on several

subsets of genes with high coefficient of variation (27). Silhouetteanalysis (28) was used to identify samples with high silhouettewidth for a given subgroup's cluster, indicating higher similarityto their own subgroup than to any other molecular subgroup(Supplementary Table S3). These samples with large silhouettewidth along with two samples designated as the WNT groupbased on mutational analysis of the CTNNB1 (exon 3) gene(29, 30) were used as core samples or true positives in construct-ing and validating the TLDA signature. The molecular subgroupof the remaining samples was predicted using the TLDA signa-ture. Supplementary Fig. S1 provides a schematic outline of theexperimental design and samples used for generating the TLDAsignature.

Macrophages were identified using immunohistochemical(IHC) analysis of 54 of the 83 medulloblastoma samples forwhich molecular subgroups had been determined using an anti-body directed against CD163 as previously described (13). Par-affin tissue section scores ranged from 0 to 3, with higher scores,indicating a greater proportion of positive cells. Two neuropatho-logists independently scored all samples and the mean of thescores was used for further analyses. Twenty-three samples werealso stained using an antibody against Ki-67 to assess associationof macrophages and cell proliferation.

Statistical analysisCommon statistical analyses, including ANOVA with linear

contrast, c2, and Spearman rank correlation coefficient, were usedwhere appropriate and are indicated in the text and Supplemen-tary Data. Statistical computations were performed using theR project (http://www.r-project.org) or Stata 11 (StataCorp.2009. Stata Statistical Software: Release 11: StataCorp LP).

ResultsIdentification of the molecular subgroup using HuEx assay

Molecular subgroups of medulloblastomas were identifiedusing HuEx microarray data from 168 samples (65 studypatients and 103 from a previously published cohort) usingan algorithm based on NMF (Fig. 1A and SupplementaryFig. S2). There was an extremely high concordance (92%)between the molecular subgroup designation of the 103 patientspreviously published and results obtained with our combinedanalysis (Supplementary Table S3), with majority of discordantfindings occurring between Groups 3 and 4. The patient char-acteristics and distribution of tumors among molecular sub-groups for the 65 CHLA patients were similar to previouspublished reports (Supplementary Table S3; refs. 17–19).

Inflammation-related genes in medulloblastoma molecularsubgroups

We sought to identify inflammation and immunology-relat-ed genes that were differentially expressed among the molec-ular subgroups using HuEx gene-expression data (n ¼ 168). Weidentified greater expression of inflammation-related genes(CD14, PTX3, CD4, CD163, CSF1R, and TGFB2) in tumorsof the SHH molecular subgroup compared with those of theGroup 3 and Group 4 subgroups (Fig. 1C). Several of these geneshave been shown to play an important role in the microenviron-ment of the developing cerebellumor other tumors types.CD14, amonocyticmarker present onboth circulating and residentmono-cytes, has been show to correlate with tumor grade in gliomas

Translational Relevance

Medulloblastoma is the most common malignant child-hood brain tumor. Approximately 30% of patients remainincurable and current radiotherapy containing treatmentprotocols cause significant adverse long-term neurocogni-tive effects and endocrine dysfunction. The role of tumormicroenvironment as an enabling characteristic of cancerand development of novel immunotherapeutics invokes thepossibility of tumor-associated inflammatory cells as ther-apeutic targets. Here, we report distinct tumor microenvir-onments of medulloblastoma molecular subgroups and thepresence of tumor-associated macrophages (TAM) in theSonic Hedgehog subgroup of medulloblastomas. We devel-oped a 31–gene-expression signature inclusive of inflam-mation-related genes that is clinically applicable and highlyaccurate in classifying medulloblastoma subgroups. Weconfirm presence of TAMs using immunohistochemistryand demonstrate their proximity to proliferating cells. Ourwork sheds light on the importance of the tumor microen-vironment in childhood brain tumors and inhibition ofTAMs, possibly through the CSF1R inhibitor, as a potentialnew therapeutic target in medulloblastomas.

Margol et al.

Clin Cancer Res; 21(6) March 15, 2015 Clinical Cancer Research1458




(11). Murine experiments demonstrate that increased levels ofTGFB2 are associated with the presence of proliferating, undiffer-entiated cerebellar neurons (31).CD163 is a well-describedmark-er of TAMs, andCSF1R is an important receptor that alongwith itsligandCSF1 controls theproduction, differentiation, and functionof TAMs (4, 6, 32, 33). PTX3 is produced by macrophages thathave beenpolarized to theM2-like phenotype via their interactionwith CD4þ T regulatory cells (32).

Expression levels of TAM markers, CD163 and CSF1R, variedsignificantly among molecular subgroups (CD163 P < 0.0001;CSF1R P < 0.0001) and were significantly greater in tumors ofthe SHH and WNT subgroups compared with those in Groups 3and 4 (CD163, ANOVA with linear contrast P < 0.0001 for SHHor WNT compared with Groups 3 and 4; CSF1R, ANOVA withlinear contrast P < 0.0001 for SHH or WNT compared withGroups 3 and 4). There was no statistically significant differencein CD163 or CSF1R expression between SHH tumors and WNTtumors (CD163 P ¼ 0.97; CSF1R P ¼ 0.50; Fig. 1B andSupplementary Fig. S3). Additional unbiased analysis wasperformed to assess association between expression levels ofmacrophage-related genes to medulloblastoma subgroups.Unsupervised clustering of 40 genes identified through GeneOntology search to be related to macrophage biology demon-strated distinct clustering of SHH and WNT subgroups from

Group 3 and 4 samples (Supplementary Fig. S9). These datasuggest that expression of inflammation-related genes, espe-cially those related to TAMs, can distinguish the tumor micro-environment of the SHH and WNT subgroups of medulloblas-tomas from Groups 3 and 4.

Expression of inflammation- and tumor cell–related genescomprises a molecular subgroup signature

To identify the subgroups in a larger cohort of medulloblas-toma patients and to validate expression of inflammation-relatedgenes, we developed a robust and clinically applicable assay usingthe TLDA technology, a system currently being evaluated inneuroblastoma and used in breast cancer clinical trials (34,35). We built a medulloblastoma-specific TLDA card containing39 tumor-related and six inflammation-related genes (CD163,CSF1R, MMD, CD4, ALCAM, and CXCR4) that were observed assignificantly deregulated among medulloblastoma subgroups inour HuEx microarray analysis (Supplementary Table S2). TheTLDA gene-expression profiles of medulloblastomas were thenused to build and validate a 31-gene signature that could accu-rately predict the four molecular subgroups in 83 samples,including two matched relapse cases (Table 2 and Fig. 2A). Theestimated leave-one-out cross-validated error rate of the 31-genesignature was 2% with classification errors occurring only in

10

0

−10

−10 0 10

Principal component 1

Prin

cipa

l com

pone

nt 2

Molecular group

Molecular group

Group 3

Group 4

SHH

WNT

12

10

8

6

4

2WNT SHH Group 3 Group 4

(n = 9) (n = 53) (n = 23) (n = 51)

CD

163

HuE

x ex

pres

sion

(lo

g 2)

NSP < 0.0001

P < 0.0001

Molecular groupHistology

Histology

10

8

6

4

CD14CD163C1STGFB2CSF1RCD4PTX3

Group 3Group 4SHHWNT

AnaplasticClassicDesmoplastic

A B

C

Figure 1.A, separation of 168 samples into 4 molecular subgroups (WNT, SHH, Group 3, and Group 4) using principal component analysis of 369 highly variable genesof the HuEx array data. B, HuEx expression levels of CD163, a macrophage specific marker, showed highest expression in the SHH subgroup of tumors. C,heatmap of gene-expression levels of top seven differentially expressed inflammation-related genes compared across molecular subgroups. Heatmap colors reflectexpression values (log2). Molecular group and histology of tumors are indicated by the color codes.

Tumor Microenvironment in Medulloblastoma Subgroups

www.aacrjournals.org Clin Cancer Res; 21(6) March 15, 2015 1459




samples identified asGroup4 (Supplementary Tables S4–S6). Thepatient characteristics and distribution of tumors among molec-ular subgroups for the 81 CHLA patients were again similar toprevious published reports with 4% WNT, 31% SHH, 26%Group 3, and 39% Group 4 (Table 1). The molecular subgroupsof the two relapse cases were the same as their diagnostic counter-parts. All patients identified as the WNT subgroup in our studycohort enjoyed long-term progression-free survival (PFS), similarto previously published reports (Supplementary Fig. S4).

Among the 31 genes included in our signature, increasedexpression of WIF1, DKK2, PGYL, and TNC was associated withWNT tumors, HHIP, PDLIM3, SFRP1, and GLI1 associated withSHH tumors, MYC, IMPG2, NPR3 associated with Group 3, andKCNA,MPP3, and EOMES associatedwithGroup 4 tumors. Thesedata are consistent with previous microarray-based publications,indicating differential expression of these genes amongmoleculargroups of medulloblastomas (17–19, 23, 36).

We also identified several novel genes, which were differen-tially expressed among medulloblastoma molecular subgroupsand contributed to their accurate identification (Table 2).Notably, the TAM-associated genes CD163 and CSF1R weredifferentially expressed among molecular subgroups withincreased expression in tumors of the SHH subgroup comparedwith those in Groups 3 and 4 (CD163 P < 0.0001; CSF1R P <0.0001 for all pairwise comparisons) and contributed to the 31-gene signature predictive of molecular subgroups (Fig. 2B andC). A gene–gene correlation was observed between CD163 andCSF1R (Spearman r¼ 0.67, Fig. 2D), suggestive of coexpressionof these two genes most likely by TAMs. There was no differencein CD163 expression in SHH tumors with desmoplastic his-

tology compared with those with classic histology (Supple-mentary Fig. S5). The median gene expression of CD163 amongthe 22 patients with SHH medulloblastoma was used to definelow and high CD163 expressers. There was no difference in the10-year PFS for patients in these two groups (P ¼ 0.57);however, the 10-year overall survival (OS) for patients in thesetwo groups trended toward but did not reach statistical signif-icance (Supplementary Fig. S6; CD163 high vs. low expresser,58% vs. 100%, respectively, P ¼ 0.08).

With regard to WNT tumors, there was a statistically significantdifference in the expression of CD163 when compared withGroup 4 (P ¼ 0.04), but not compared with Group 3 tumors(P ¼ .18) and no difference in expression of CSF1R (P ¼ 0.83 forGroup 3 and P¼ 0.48 for Group 4). Only threeWNT tumors wereavailable for analysis with TLDA, so we cannot conclude whetherthere is increased expression of CD163 or CSF1R in WNT tumorscompared with Group 3 or Group 4 tumors.

Pattern of infiltration of macrophages in SHHmedulloblastoma

We next performed IHC analysis of 54 paraffin-embeddedmedulloblastoma tumors using antibodies directed againstCD163, to assess the extent and pattern of TAM infiltration inmedulloblastomas. There was a significant difference in macro-phage infiltration among molecular subgroups (P < 0.0001)with significantly greater numbers of macrophages observed intumors of the SHH subgroup compared with those in Groups 3and 4 (P < 0.0001 for both comparisons; Fig. 3). There was astatistically significant difference in the number of intratumoralCD163þ macrophages between WNT tumors and Group 3 and

0

−4

−8

−2.5 0.0 2.5 5.0 7.5

−2

−4

−6

−8

−10

−12

CD

163

expr

essi

on (

ΔCt)

WNT SHH Group 3 Group 4

WNT SHH Group 3 Group 4

(n = 3)

(n = 3) (n = 26) (n = 21) (n = 33)

(n = 26) (n = 21) (n = 33)

Principal component 1

Prin

cipa

l com

pone

nt 2

CS

F1R

exp

ress

ion

(ΔC

t)

CS

F1R

exp

ress

ion

(ΔC

t)

NS P < 0.0001P < 0.0001

−6

−8

−10

−12

−14

−16

P = 0.01P < 0.0001 P < 0.0001

−9

−11

−13

−10.0 −7.5 −5.0

Group 3

Group 4

SHH

WNT

Group 3

Group 4

SHH

WNT

CD163 expression (ΔCt)

A

C D

B

Figure 2.A, separation of medulloblastomasamples (n ¼ 83) using the first twoprincipal components of the genesrepresented in the TLDA 31-genesignature. B, the RT-PCR–based TLDAassay validated HuEx resultsdemonstrating significant increase inexpression of CD163 (B) and CSF1R (C)in the SHH subgroup compared withGroups 3 and 4 subgroups. D, CD163expression was correlated with CSF1R(Pearson r ¼ 0.64, P < 0.001).

Margol et al.





Group 4 tumors (P ¼ 0.04 for both groups) and no statisticallysignificant difference was found between the WNT andSHH tumors (P ¼ 0.80); however, there were only two WNT

samples available for evaluation. Among tumors with an IHCscore �2.5 (n ¼ 16), 94% were SHH tumors and the remaining6% were WNT tumors. The SHH tumors with desmoplastic

Table 1. Patient characteristics by the molecular subgroup

WNTa,b SHH Group 3 Group 4Number of Patients (%) Number of Patients (%) Number of Patients (%) Number of Patients (%) P

Total patients 3 (4) 25 (31) 21 (26) 32 (39)Age group, y <0.01c

<3 0 (0) 15 (60) 5 (24) 2 (6)3–6 0 (0) 5 (20) 6 (29) 11 (34)6–10 2 (67) 4 (16) 6 (29) 10 (31)>10 1 (33) 1 (4) 4 (19) 9 (28)

Sex <0.05c

Male 0 (0) 13 (52) 14 (67) 25 (78)Female 3 (100) 12 (48) 7 (33) 7 (22)

M Stage 0.41c

M0 3 (100) 21 (84) 14 (67) 23 (72)M1 0 (0) 1 (4) 2 (10) 0 (0)M2 0 (0) 0 (0) 1 (5) 0 (0)M3 0 (0) 3 (12) 4 (19) 9 (28)

Histology <0.001c

Desmoplastic 0 (0) 13 (52) 1 (5) 2 (6)Classic 3 (100) 11 (44) 12 (57) 22 (69)Anaplastic 0 (0) 1 (4) 8 (38) 8 (25)Number of events 0 (0) 5 (20) 7 (33) 11 (34) d

Number of deaths 0 (0) 3 (12) 6 (29) 9 (28) d

aThe molecular group based on TLDA analysis.bTwo designated WNT based on presence of CTNNB1 mutation.cOn the basis of the c2 test.dOn the basis of the Kaplan–Meier method and log-rank test.

Table 2. TLDA 31-gene signature

Gene symbols Gene name Gene location AUC

Tumor relatedTERC Telomerase RNA component 3q26 0.78FOXG1 Forkhead box G1 14q13 0.95PPP1R17 Protein phosphatase 1, regulatory subunit 17 7p15 0.91SLC6A5 Solute carrier family 6 (neurotransmitter transporter), member 5 11p15.1 0.82BCAT1 Branched chain amino-acid transaminase 1, cytosolic 12p12.1 0.72CBLN3 Cerebellin 3 precursor 14q12 0.81PID1 Phosphotyrosine interaction domain containing 1 2q36.3 0.98ERG1 Early growth response protein 1 5q23-q31 0.63WIF1 WNT inhibitory factor 1 12q14.3 0.51DKK2 Dickkopf WNT signaling pathway inhibitor 2 4q25 0.71PYGL Phosphorylase, glycogen, liver 14q21-q22 0.51TNC Tenascin C 9q33 0.82PDLIM3 PDZ and LIM domain 3 4q35 0.89HHIP Hedgehog interacting protein 4q28-q32 0.93SFRP1 Secreted frizzled-related protein 1 8p11.21 0.90GLI1 GLI family zinc finger 1 12q13.2-q13.3 0.96NPR3 Natriuretic peptide receptor C/guanylate cyclase C (atrionatriuretic peptide receptor C) 5p14-p13 0.75MYC V-myc avian myelocytomatosis viral oncogene homolog 8q24.21 0.52IMPG2 Interphotoreceptor matrix proteoglycan 2 3q12.2-q12.3 0.88GABRA5 Gamma-aminobutyric acid (GABA) A receptor, alpha 5 15q12 0.88EOMES Eomesodermi 3p24.1 0.88MPP3 Membrane protein, palmitoylated 3 (MAGUK p55 subfamily member 3) 17q21.31 0.95FSTL5 Follistatin-like 5 4q32.3 0.84PDGFRA Platelet-derived growth factor receptor, alpha polypeptide 4q12 0.96OTX2 Orthodenticle homeobox 2 14q22.3 0.96

Inflammation relatedCD163 CD163 molecule 12p13.3 0.94CSF1R Colony stimulating factor 1 receptor 5q32 0.78MMD Monocyte to macrophage differentiation associated 17q22 0.97CD4 CD4 molecule 12p13.31 0.74CXCR4 Chemokine (C–X–C motif) receptor 4 2q21 1.00ALCAM Activated leukocyte cell adhesion molecule 3q13.1 0.71

NOTE: AUC values for distinguishing WNT and SHH from the Group 3 and 4. Four-way ANOVA P < 0.001 for all genes in signature.






histology exhibited a distinct pattern of macrophage infiltrationin the internodular, poorly differentiated areas while sparing themore differentiated nodules (Fig. 3 and Supplementary Fig. S7).Interestingly, in the subset of SHHmedulloblastomas with classichistology, CD163þ macrophages sometimes loosely recapitulat-ed this lobular organization. We examined the extent and patternof tumor cell proliferation as an increased proliferation index hasbeen described in cells in the presence of macrophages (6). Theareas ofmacrophage infiltration in the SHH tumors correspondedwith areas of increased proliferation as evidenced by positivestaining for Ki-67, a nuclear marker of cell proliferation, whichwas performed in a subset of 23 tumor samples (Fig. 4 and Sup-plementary Fig. S8).

DiscussionTreatment strategies aimed at improving survival of young

children with medulloblastoma by avoiding radiotherapy andits neurocognitive sequelae require identification of novelsubgroup-specific targets. Our study suggests for the first timethat TAMs contribute to the microenvironment of a childhoodbrain tumor and demonstrates their prevalence in tumors ofchildren with the SHH subgroup of medulloblastoma. Weshow that expression of inflammation-related genes, includingTAM-related genes, CD163 and CSF1R, is higher in SHH ascompared with the other medulloblastoma subgroups. Theincreased expression of CD163 and CSF1R suggests that theTAMs seen on IHC are of the M2 phenotype, and thereforeassociated with tumor progression (4, 6, 7, 10, 37). A 31-genesignature, inclusive of both inflammatory and tumor cell genes,enables proper identification of molecular subgroups with 98%accuracy. The 31–gene-expression scoring model has clinicalapplicability and could be of use for risk stratification, whereas

identification of TAMs in SHH tumors uncovers a previouslyunrecognized potential target for therapy. CD163 expressionwas observed in the limited number WNT samples, suggestingthat macrophages may also play a role in the WNT subgroupof medulloblastomas; however, given the lack of number ofWNT samples (nine in HuEx, three in TLDA, and two IHC),we do not feel that we have enough sufficient evidence to drawa strong conclusion.

Although targeted therapy with SHH pathway inhibitorsshows tremendous promise, it has become clear that noveltreatments that overcome mechanisms of resistance to theseinhibitors need to be identified to improve OS (38). In recentyears, the concept of inflammatory cells in the tumor micro-environment as critical participants in tumor progressionhas gained acceptance. Large numbers of infiltrating TAMs arepredictive of a poor prognosis in many adult cancers (39–41),and a 14-gene signature inclusive of five genes representingTAMs has been shown to predict PFS in patients with metastaticneuroblastoma (13). Our study also suggests a prognostic rolefor expression of CD163 among patients with SHH medullo-blastoma, but validation with larger number of samples isneeded. The tumor microenvironment also plays an importantrole in drug resistance mechanisms of tumors. Coculture ofleukemia cells with stromal cells allows for environment-medi-ated drug resistance (EMDR) to tyrosine kinase inhibitors(42). This EMDR is associated with differential regulation ofinflammation-related genes. TAMs produce cytokines thatactivate STAT3 and Hedgehog signals in colon and lung cancerstem cells rendering them resistant to chemotherapy (43). Thissuggests that combination therapy aimed at targeting the micro-environment in addition to the tumors cells may improve theresponse to chemotherapy and decrease the risk of develop-ment of EMDR.

3.5

3.0

2.5

2.0

1.5

1.0

0.5WNT SHH Group 3 Group 4

(n = 2) (n = 22) (n = 9) (n = 21)

NSP < 0.0001

P < 0.0001

A B

C D

E

Ave

rage

IHC

sco

re

Figure 3.Evidence of TAMs acrossmedulloblastoma subgroups. Representative CD163 IHC images in tumor samples from the SHH subgroupwith desmoplastic histology (A),SHH with classic histology (B), the Group 3 subgroup (C), and the Group 4 subgroup (D). E, average CD163 IHC score is significantly higher in SHH tumorscompared with Group 3 or Group 4 subgroups (P < 0.0001, respectively).

Margol et al.





In this study, we show that the expression of inflammation-related genes, especially macrophage markers CD163 andCSF1R, is highest in the SHH subgroup of medulloblastomas,which was validated by the IHC analyses. The protumor effectsof TAMs on tumor pathogenesis have been shown in de novoepithelial carcinogenesis in mice through production of cyto-kines such as IL6, IL10, and IL4 that stimulate tumor growth andangiogenesis. Coculture of neuroblastoma cells with peripheralblood monocytes or mesenchymal cells also increases tumorcell proliferation through IL6- and STAT3-dependent mechan-isms (44, 45). In a prostate cancer model, macrophages induceCCL4 production that promotes tumorigenesis through STAT3activation (46), whereas glioblastoma conditioned media pro-tect TAM survival demonstrating the cross-talk between tumorcells and TAMs (47). The TAMs in SHH medulloblastoma arelocated near the proliferating tumor cells, as identified by Ki-67marker, and points to their likely role in creating a progrowthtumor microenvironment.

Macrophages of the M2 phenotype have been shown to pro-mote tumor progression via a variety of mechanisms, includingimmunosuppression and angiogenesis (4, 6, 7, 37, 40). CSF1Rresides on the surface of tumor-promoting macrophages, and ithas been demonstrated that inhibition of CSF1R can reduce theprotumor effects of TAMs in the tumor microenvironment(37, 48). In a transgenic murine model of mammary adenocar-cinoma, blockade of CSF1R signaling leads to a decrease inintratumoral TAMs resulting in increased sensitivity to chemo-therapy. Administration of a CSF1R antagonist in combinationwith paclitaxel leads to a decrease in primary tumor progression aswell as decreased rates of pulmonary metastasis and OS whencomparedwithmice treatedwith paclitaxel alone (41). Inhibitionof CSF1R in several preclinical models of proneural glioblastomamultiforme repolarizes TAMs from theM2 phenotype toward the

M1 phenotype resulting in cessation of their tumorigenic func-tions (46). Our finding of TAMs with high expression of CSF1R inSHH medulloblastomas provides a novel therapeutic target. Inthe future, therapies aimed at blocking pathways mediatingmacrophage recruitment, polarization, and cross-talk with tumorcells could be combined with current or reduced intensity che-moradiation strategies.

Our study also defines a clinically applicable 31–gene-expres-sion signature that identifies the four molecular subgroups ofmedulloblastomas with a 2% misclassification rate. Our groupand others have demonstrated the clinical utility of these TLDAassays in childhood and adult clinical trials (34, 35). Althoughcurrent techniques such as IHC, FISH, and cytogenetics could beused to identify only a portion of medulloblastomas subgroups(21, 23), our proposed 31-gene signature provides a rapid andhighly accurate assay for determining these subgroups. Imple-mentation of this assay could also be used in protocols aimed atavoiding or delaying radiotherapy in children with WNT or SHHtumors, or identification of children with Group 3 of Group 4tumors for novel therapies. As the molecular subgroup becomespart of risk stratification, it will be important to have tools adept atmaking that determination.

In summary, our study reports the first evidence of the presenceof TAMs in pediatric medulloblastoma and provides a 31-genesignature, inclusive of macrophage-associated genes, that accu-rately determines medulloblastoma subgroups. The increase inexpression of macrophage markers can be used as a biomarkerto identify subgroups of patients who may benefit from adjunc-tive treatments targeting TAMs and the tumormicroenvironment.The success of therapies directed at reversing the suppressive roleof immune cells in adult cancers (49) and the recent developmentof anti-CSF1Randother antibodies (48, 50) suggest opportunitiesfor their application in pediatric SHH medulloblastoma.

A B

C D

Figure 4.Representative IHC images of tumorsstained with anti–Ki-67 antibody inSHH subgroup with desmoplastichistology (A), SHH with classichistology (B), the Group 3 subgroup(C), and the Group 4 subgroup (D).Increased cell proliferation in theinternodular areas corresponded tothe presence of macrophages in theSHH subgroup with desmoplastichistology.






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

Authors' ContributionsConception and design: A.S. Margol, N.J. Robison, J.L. Finlay, M.D. Krieger,S. AsgharzadehDevelopment of methodology: A.S. Margol, N.J. Robison, S. AsgharzadehAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): A.S. Margol, N.J. Robison, L.T. Hung, G. Dhall,M.D. Krieger, R. Drissi, M. Fouladi, F.H. Gilles, S. AsgharzadehAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): A.S. Margol, J. Gnanachandran, M. Vali, G. Dhall,M.D. Krieger, F.H. Gilles, A.R. Judkins, R. Sposto, S. AsgharzadehWriting, review, and/or revision of themanuscript: A.S. Margol, N.J. Robison,G. Dhall, J.L. Finlay, A. Erdreich-Epstein, M.D. Krieger, R. Drissi, M. Fouladi,F.H. Gilles, A.R. Judkins, R. Sposto, S. AsgharzadehAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): A.S. Margol, N.J. Robison, J. Gnanachandran,L.T. Hung, R.J. Kennedy, S. AsgharzadehStudy supervision:A.S.Margol, J.L. Finlay,M.D.Krieger, A.R. Judkins, S. Asgharzadeh

Other (items relating to PID1): A. Erdreich-EpsteinOther (specimen contribution): R. Drissi

AcknowledgmentsThe authors thank Karen Miller and John Harbert for their technical assis-

tance in IHC, and Esteban Fernandez for his assistance in photomicroscopy.

Grant SupportThis work was supported by a grant (to S. Asgharzadeh) from the American

Cancer Society, Stop Cancer Foundation, St. Baldrick's Foundation, Alex'sLemonade Stand Foundation, Hope Funds for Cancer Research (to S. Asghar-zadeh and N.J. Robison), Brad Kaminsky Foundation (to S. Asgharzadeh),National Institutes of Health T32 CA009659-18 training grant (to A. Margol),and National Institute of Child Health and Human Development (K12-CA60104).

The costs of publication of this articlewere defrayed inpart by the payment ofpage charges. This article must therefore be hereby marked advertisement inaccordance with 18 U.S.C. Section 1734 solely to indicate this fact.

ReceivedMay 6, 2014; revised August 23, 2014; accepted September 19, 2014;published OnlineFirst October 24, 2014.

References1. Hanahan D, Weinberg RA. Hallmarks of cancer: the next generation. Cell

2011;144:646–74.2. Mantovani A, Allavena P, Sica A, Balkwill F. Cancer-related inflammation.

Nature 2008;454:436–44.3. Coussens LM, Werb Z. Inflammation and cancer. Nature 2002;420:860–7.4. Pollard JW. Opinion: tumour-educated macrophages promote tumour

progression and metastasis. Nat Rev Cancer 2004;4:71–8.5. ZhangQ-W, Liu L,GongC-Y, ShiH-S, Zeng Y-H,WangX-Z, et al. Prognostic

significance of tumor-associated macrophages in solid tumor: a meta-analysis of the literature. PLoS ONE 2012;7:e50946.

6. Lewis CEC, Pollard JWJ. Distinct role of macrophages in different tumormicroenvironments. Cancer Res 2006;66:605–12.

7. Candido J, Hagemann T. Cancer-related inflammation. J Clin Immunol2012;33:79–84.

8. Hussain SF. The role of human glioma-infiltrating microglia/macro-phages in mediating antitumor immune responses. Neuro Oncol 2006;8:261–79.

9. Engler JR, Robinson AE, Smirnov I, Hodgson JG, Berger MS, Gupta N, et al.Increased microglia/macrophage gene expression in a subset of adult andpediatric astrocytomas. PLoS ONE 2012;7:e43339.

10. LiW,GraeberMB. Themolecular profile ofmicroglia under the influence ofglioma. Neuro Oncol 2012;14:958–78.

11. Prosniak M, Harshyne LA, Andrews DW, Kenyon LC, Bedelbaeva K,Apanasovich TV, et al. Glioma grade is associated with the accumulationand activity of cells bearing m2 monocyte markers. Clin Cancer Res2013;19:3776–86.

12. Carvalho da Fonseca AC, Badie B.Microglia andmacrophages inmalignantgliomas: recent discoveries and implications for promising therapies. ClinDev Immunol 2013;2013:264124.

13. Asgharzadeh S, Salo JA, Ji L, Oberthuer A, Fischer M, Berthold F, et al.Clinical significance of tumor-associated inflammatory cells in metastaticneuroblastoma. J Clin Oncol 2012;30:3525–32.

14. Packer RJ, Sposto R, Atkins TE, Sutton LN, Bruce DA, Siegel KR, et al.Quality of life in children with primitive neuroectodermal tumors(medulloblastoma) of the posterior fossa. Pediatr Neurosci 1987;13:169–75.

15. Ris MD, Packer R, Goldwein J, Jones-Wallace D, Boyett JM. Intellectualoutcome after reduced-dose radiation therapy plus adjuvant chemother-apy for medulloblastoma: a Children's Cancer Group Study. J Clin Oncol2001;19:3470–6.

16. ThompsonMC, FullerC.Genomics identifiesmedulloblastoma subgroupsthat are enriched for specific genetic alterations. J Clin Oncol 2006;24:1924–31.

17. Kool M, Koster J, Bunt J, Hasselt NE, Lakeman A, van Sluis P, et al.Integrated genomics identifies five medulloblastoma subtypes with dis-

tinct genetic profiles, pathway signatures and clinicopathological features.PLoS ONE 2008;3:e3088.

18. Northcott PA, Korshunov A,Witt H,Hielscher T, Eberhart CG,Mack S, et al.Medulloblastoma comprises four distinct molecular variants. J Clin Oncol2011;29:1408–14.

19. Cho Y-J, Tsherniak A, Tamayo P, Santagata S, Ligon A, Greulich H, et al.Integrative genomic analysis of medulloblastoma identifies a molecularsubgroup that drives poor clinical outcome. J Clin Oncol 2011;29:1424–30.

20. Gibson P, Tong Y, Robinson G, Thompson MC, Currle DS, Eden C, et al.Subtypes ofmedulloblastoma have distinct developmental origins. Nature2010;468:1095–9.

21. Ellison DW, Kocak M, Dalton J, Megahed H, Lusher ME, Ryan SL, et al.Definition of disease-risk stratification groups in childhood medulloblas-toma using combined clinical, pathologic, and molecular variables. J ClinOncol 2011;29:1400–7.

22. Taylor MD, Northcott PA, Korshunov A, Remke M. Molecular subgroupsof medulloblastoma: the current consensus. Acta Neuropathol 2012;23:465–72.

23. SchwalbeEC, Lindsey JC, StraughtonD,Hogg TL,ColeM,MegahedH, et al.Rapid diagnosis of medulloblastoma molecular subgroups. Clin CancerRes 2011;17:1883–94.

24. Parsons DW, Li M, Zhang X, Jones S, Leary RJ. The genetic landscape of thechildhood cancer medulloblastoma. Science 2011;331:435–9.

25. de Haas T, Hasselt N, Troost D, Caron H, Popovic M, Zadravec-Zaletel L,et al. Molecular risk stratification of medulloblastoma patients based onimmunohistochemical analysis of MYC, LDHB, and CCNB1 expression.Clin Cancer Res 2008;14:4154–60.

26. Kunder R, Jalali R, Sridhar E, Moiyadi A, Goel N, et al. Real-timePCR assay based on the differential expression of microRNAsand protein-coding genes for molecular classification of formalin-fixed paraffin embedded medulloblastomas. Neuro Oncol 2013;15:1644–51.

27. Gaujoux R, Seoighe C. A flexible R package for nonnegative matrixfactorization. BMC Bioinformatics 2010;11:367.

28. Rousseeuw P. Silhouettes: A graphical aid to the interpretation and vali-dation of cluster analysis. J Comput Appl Math 1987;20:53–65.

29. Eberhart CG, Tihan T, Burger PC. Nuclear localization and mutation ofbeta-catenin in medulloblastomas. J Neuropathol Exp Neurol 2000;59:333–7.

30. Clifford SC, Lusher ME, Lindsey JC, Langdon JA, Gilbertson RJ,Straughton D, et al. Wnt/Wingless pathway activation and chromo-some 6 loss characterize a distinct molecular sub-group of medullo-blastomas associated with a favorable prognosis. Cell Cycle 2006;5:2666–70.

Margol et al.





31. Constam DB, Schmid P, Aguzzi A, Schachner M, Fontana A. Transientproduction of TGF-b2 by postnatal cerebellar neurons and its effect onneuroplast proliferation. Eur J Neurosci 1994;6:766–78.

32. Condeelis J, Pollard JW. Macrophages: obligate partners for tumor cellmigration, invasion, and metastasis. Cell 2006;124:263–66.

33. Biswas SK, Mantovani A. Macrophage plasticity and interaction withlymphocyte subsets: cancer as a paradigm. Nat Immunol 2010;11:889–96.

34. Espinosa E, S�anchez-Navarro I, G�amez-Pozo A, Marin AP, Hardisson D,Madero R, et al. Comparison of prognostic gene profiles using qRT-PCR inparaffin samples: a retrospective study in patients with early breast cancer.PLoS ONE 2009;4:e5911.

35. Vermeulen J, De Preter K, Naranjo A, Vercruysse L, Van RoyN, Hellemans J,et al. Predicting outcomes for children with neuroblastoma using a multi-gene-expression signature: a retrospective SIOPEN/COG/GPOH study.Lancet Oncol 2009;10:663–71.

36. KoolM, Korshunov A, RemkeM, Jones DTW, SchlansteinM,Northcott PA,et al. Molecular subgroups of medulloblastoma: an international meta-analysis of transcriptome, genetic aberrations, and clinical data of WNT,SHH, Group 3, and Group 4 medulloblastomas. Acta Neuropathol 2012;123:473–84.

37. Ruffell B, Affara N I, Coussens LM. Differential macrophage programmingin the tumor microenvironment. Trends Immunol 2012;33:119–126.

38. Rudin CM, Hann CL, Laterra J, Yauch RL. Treatment of medulloblastomawith hedgehog pathway inhibitor GDC-0449. N Engl J Med 2009;361:1173–78.

39. Steidl C, Lee T, Shah SP, Farinha P, Han G, Nayar T, et al. Tumor-associatedmacrophages and survival in classic Hodgkin's lymphoma. N Engl J Med2010;362:875–85.

40. Zhang BC, Gao J, Wang J, Rao ZG, Wang BC, Gao JF. Tumor-associatedmacrophages infiltration is associated with peritumoral lymphangiogen-esis and poor prognosis in lung adenocarcinoma. Med Oncol 2011;28:1447–52.

41. DeNardo DG, Brennan DJ, Rexhepaj E, Ruffell B, Shiao SL, Madden SF,et al. Leukocyte complexity predicts breast cancer survival and func-tionally regulates response to chemotherapy. Cancer Discov 2011;1:54–67.

42. Feldhahn N, Arutyunyan A, Stoddart S, Zhang B, Schmidhuber S, Yi S-J,et al. Environment-mediated drug resistance in Bcr/Abl-positive acutelymphoblastic leukemia. Oncoimmunology 2012;1:618–29.

43. JinushiM, Chiba S, YoshiyamaH. Tumor-associatedmacrophages regulatetumorigenicity and anticancer drug responses of cancer stem/initiatingcells. Proc Natl Acad Sci U S A 2011;108:12425–30.

44. Song L, Asgharzadeh S, Salo J, Engell K,WuH-W, Sposto R, et al. Valpha24-invariant NKT cells mediate antitumor activity via killing of tumor-asso-ciated macrophages. J Clin Invest 2009;119:1524–36.

45. Ara T, Song L, Shimada H, Keshelava N, Russell HV, Metelitsa LS, et al.Interleukin-6 in the bonemarrowmicroenvironment promotes the growthand survival of neuroblastoma cells. Cancer Res 2009;69:329–37.

46. Fang L-Y, Izumi K, Lai K-P, Liang L, Li L, Miyamoto H, et al.Infiltrating macrophages promote prostate tumorigenesis via modu-lating androgen receptor-mediated CCL4-STAT3 signaling. Cancer Res2013;73:5633–46.

47. Pyonteck SM, Akkari L, Schuhmacher AJ, Bowman RL, Sevenich L, QuailDF, et al. CSF-1R inhibition alters macrophage polarization and blocksglioma progression. Nat Med 2013;19:1264–72.

48. Hume DA, MacDonald KPA. Therapeutic applications of macrophagecolony-stimulating factor-1 (CSF-1) and antagonists of CSF-1 receptor(CSF-1R) signaling. Blood 2012;119:1810–20.

49. Robert C, Thomas L, Bondarenko I. Ipilimumab plus dacarbazine forpreviously untreated metastatic melanoma. N Engl J Med 2011;364:2517–26.

50. Horton HM, Bernett MJ, Peipp M, Pong E, Karki S, Chu SY, et al. Fc-engineered anti-CD40 antibody enhances multiple effector functions andexhibits potent in vitro and in vivo antitumor activity against hematologicmalignancies. Blood 2010;116:3004–12.






2015;21:1457-1465. Published OnlineFirst October 24, 2014.Clin Cancer Res Ashley S. Margol, Nathan J. Robison, Janahan Gnanachandran, et al. MedulloblastomasTumor-Associated Macrophages in SHH Subgroup of

Updated version

10.1158/1078-0432.CCR-14-1144doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2014/10/25/1078-0432.CCR-14-1144.DC1

Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/21/6/1457.full#ref-list-1

This article cites 50 articles, 16 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/21/6/1457.full#related-urls

This article has been cited by 6 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/21/6/1457To 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-14-1144

http://clincancerres.aacrjournals.org/content/suppl/2014/10/25/1078-0432.CCR-14-1144.DC1

http://clincancerres.aacrjournals.org/content/21/6/1457.full#ref-list-1

http://clincancerres.aacrjournals.org/content/21/6/1457.full#related-urls

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

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/21/6/1457


Documents

Tumor-AssociatedMacrophagesinSHHSubgroup of … · Medulloblastoma is the most common malignant brain tumor in children. Craniospinal irradiation still remains an essential component