Progression of EGFR-Mutant Lung Adenocarcinoma is Driven By Alveolar … · 23(3); 778–88. 2016 AACR. Introduction Non–small cell lung cancer (NSCLC) has an overall 5-year survival

Cancer Therapy: Preclinical

Progression of EGFR-Mutant LungAdenocarcinoma is Driven By AlveolarMacrophagesDon-Hong Wang1, Hyun-Sung Lee2, David Yoon2, Gerald Berry1, Thomas M.Wheeler2,David J. Sugarbaker2, Farrah Kheradmand2, Edgar Engleman1, and Bryan M. Burt2

Abstract

Purpose: Lung adenocarcinomas with mutations in the EGFRhave unprecedented initial responses to targeted therapy againstthe EGFR. Over time, however, these tumors invariably developresistance to these drugs. We set out to investigate alternativetreatment approaches for these tumors.

Experimental Design: To investigate the immunologic under-pinnings of EGFR-mutant lung adenocarcinoma, we utilized abitransgenic mouse model in which a mutant human EGFR geneis selectively expressed in the lungs.

Results: EGFR oncogene–dependent progression and remis-sion of lung adenocarcinoma was respectively dependent uponthe expansion and contraction of alveolar macrophages, andthe mechanism underlying macrophage expansion was localproliferation. In tumor-bearing mice, alveolar macrophagesdownregulated surface expression of MHC-II and costimula-tory molecules; increased production of CXCL1, CXCL2, IL1receptor antagonist; and increased phagocytosis. Depletion of

alveolar macrophages in tumor-bearing mice resulted in reduc-tion of tumor burden, indicating a critical role for these cells inthe development of EGFR-mutant adenocarcinoma. Treatmentof mice with EGFR-targeting clinical drugs (erlotinib andcetuximab) resulted in a significant decrease in alveolar macro-phages in these mice. An activated alveolar macrophage mRNAsignature was dominant in human EGFR-mutant lung adeno-carcinomas, and the presence of this alveolar macrophageactivation signature was associated with unfavorable survivalamong patients undergoing resection for EGFR-mutant lungadenocarcinoma.

Conclusions: Because of the inevitability of failure of targetedtherapy in EGFR-mutant non-small cell lung cancer (NSCLC),these data suggest that therapeutic strategies targeting alveolarmacrophages in EGFR-mutant NSCLC have the potential tomitigate progression and survival in this disease. Clin Cancer Res;23(3); 778–88. �2016 AACR.

IntroductionNon–small cell lung cancer (NSCLC) has an overall 5-year

survival rate of only 18% and is the leading cause of cancer-related deaths in the world (1, 2). The EGFR is the cell surfacereceptor for members of the EGF family of extracellular proteinligands. Mutations that lead to EGFR overexpression or con-stitutive activation have been found in a number of cancers andare associated with increased cell proliferation, migration, andadhesion (3). Ten to thirty percent of NSCLCs have mutationsin exons encoding the tyrosine kinase domain of the EGFRgene, and these tumors represent a unique subset of lung cancer(4, 5). EGFR-mutant NSCLCs display unprecedented responserates of 64% to 82% to targeted molecular therapy with EGFRtyrosine kinase inhibitors (TKI), making these reagents thestandard first-line treatment for patients with metastaticEGFR-mutant NSCLC (6). However, even in combination with

cytotoxic chemotherapy, less than 5% cure rate is expectedbecause nearly all cancers treated with TKIs will develop resis-tance resulting in disease progression (7). There is an obviousneed to develop new strategies to improve long-term outcomesin this disease.

It has become increasingly clear that failure of the immunesystem to recognize transformed cells could also drive theinitiation and progression of cancer. Accordingly, recognitionof aberrant immune responses that promote tumor growth hasidentified a variety of novel therapeutic targets (8). Tumor-associated macrophages are mononuclear phagocytic cells thatare a major cellular component of human NSCLC and theirabundance in human NSCLC tumors portends an overall poorsurvival (9). The importance of tumor-associated macrophagesin lung cancer is also supported by mouse data demonstratingthat macrophages facilitate the development of carcinogen(urethane)-induced (10) and KRAS-driven mouse lung adeno-carcinoma (11). In mice and in humans, macrophages fostertumor progression through angiogenic reprogramming, sup-pression of antitumor immune responses, and production ofsoluble mediators that support the proliferation, survival, andinvasion of malignant cells (12). Therefore, tumor-associatedmacrophages represent an attractive target for cancer immunemodulation and strategies targeting tumor-associated macro-phages are currently being introduced into the clinic by way ofclinical trials (13–16).

To understand the immunologic determinants of tumor pro-gression in EGFR-mutant lung adenocarcinoma, we utilized a

1Stanford University School of Medicine, Stanford, California. 2Baylor College ofMedicine, Houston, Texas.

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

Corresponding Author: Bryan M. Burt, Baylor College of Medicine, 6620 MainSt., Suite 1325, Houston, TX 77030. Phone: 713-798-9012; Fax: 713-798-8131;E-mail: [email protected]

doi: 10.1158/1078-0432.CCR-15-2597

�2016 American Association for Cancer Research.

ClinicalCancerResearch

Clin Cancer Res; 23(3) February 1, 2017778

on July 27, 2021. © 2017 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst August 5, 2016; DOI: 10.1158/1078-0432.CCR-15-2597

http://clincancerres.aacrjournals.org/

genetically engineered mouse model that mirrors the progressionof EGFR-mutant lung adenocarcinoma in humans from preinva-sive lepidic histology that progresses to invasive adenocarcinoma(17, 18). Herein, we have identified a novel mechanism by whichoncogene-dependent tumor progression and remission, respec-tively, require the expansion and contraction of alveolar macro-phages (AM) in an EGFR-mutant model of NSCLC. Specifically,we show that oncogenic EGFR signaling results in the expansionof an overwhelming number of AMs with a functionally immu-nosuppressive phenotype, the elimination of which result inmarkedly decreased tumor burden. Moreover, treatment ofEGFR-mutant mice with oncogene-inactivating agents such aserlotinib decreased the frequency of AMs in the lungs, suggestinga previously unappreciated mechanism for erlotinib-mediatedtumor suppression in this model. Because of the inevitability ofTKI resistance in EGFR-mutant NSCLC, these data suggest thatconcurrent therapeutic strategies that could reduce AMs in EGFR-mediated lung cancer might provide the potential to abrogatetumor progression and prolong survival.

Materials and MethodsAnimal model

Animal experiments were performed in accordance with theInstitutional Animal Care and Use Committees at Stanford Uni-versity (Stanford, CA) and Baylor College of Medicine (Houston,TX). We utilized a genetically engineered bitransgenic mousemodel of lung adenocarcinoma driven by an organ-specifichuman EGFR gene mutation. Transgenic mice carrying a reversetetracycline transactivator (rtTA) driven by the Clara cell secretoryprotein (CCSP) promoter were crossed with mice carrying amutant human EGFR gene governed by a tet operator (TetO)promoter (17, 18). The EGFR transgene had been constructedwitha human EGFR mutant that substitutes arginine for leucineat amino acid 858 (L858R) that results in lung adenocarcinomaswith preinvasive lepidic histology that progress to invasive ade-nocarcinoma. Both strains of mice were gifts from Dr. KaterinaPoliti, Yale School of Medicine (New Haven, CT), and were ofFVB background. Activation of the mutant EGFR transgene wasaccomplished by feeding bitransgenic (CCSP/EGFR) progeny withdoxycycline-impregnated food pellets (625 ppm; Harlan Labora-

tories). Female mice were fed doxycycline when they reached 12to 14 weeks of age, at which time they weighed 20 to 25 g. Controlmice were bitransgenic female mice fed normal chow.

TreatmentClodronate (dichloromethylene diphosphonate) encapsulated

liposomes or PBS control liposomes were purchased from www.clodronateliposomes.org. To deplete AMs, liposomes wereadministered to mice three times per week via an intratrachealroute by established methods (19). Erlotinib (provided by theStanford University Hospital pharmacy) was suspended in 0.5%(w/v)methylcellulose and injected intraperitoneally at 50mg/kg/d. Cetuximab (provided by the Stanford University Hospitalpharmacy) was injected intraperitoneally three times a week ata dose of 10 mg/kg. To deplete circulating monocytes, a neutral-izingmAb against themouse colony-stimulating factor 1 receptor(CSF-1R; BioXCell; clone 5A1)was injected intraperitoneally at 30mg/kg every other day.

Histology and IHCMice were euthanized with a lethal dose of carbon monoxide

per institutional guidelines. Lungs were cleared of circulatingblood cells by perfusion of the right ventricle with PBS, andwhole lungs were excised. For hematoxylin and eosin (H&E)staining, lungs were fixed with 10% paraformaldehyde in PBSovernight at room temperature, and stored at 4�C. Paraffinembedding, sectioning, and H&E staining were performed atHisto-Tec laboratories, and the stained sections were reviewedby board-certified pathologists (G. Berry and T.M. Wheeler). Forimmunofluorescence microscopy, whole lungs were suspendedin OCT embedding compound (Tissue-Tek), immediately frozenon dry ice, and then maintained at �80�C. After cryostat section-ing, the sections were post-fixed in 2% paraformaldehyde andblockedwith 10%goat serum. Sectionswere subsequently stainedwith anti-mouse F4/80 antibody (Biolegend, 123122, 1:300),followed by secondary goat anti-mouse (Life Technologies,A-21235, 1:500), and DAPI was used as nuclear stain. Immuno-fluorescence images were viewed and captured with a ZeissLSM700 microscope.

Flow cytometry and cell sortingLungs were collected and minced into small pieces in 50-mL

conical tubes and digested in 1.0 mg/mL collagenase IV(Worthington, LS-004188) and 100 U/mL DNase I (Sigma-Aldrich, D4263) for 20 minutes in 37�C. Supernatants werepassed through a 100-mm cell strainer and quenched withRPMI1640 with 10% FCS. Single lung cells were washed in a PBSbuffer containing 1% FCS and 2 mmol/L EDTA and red bloodcells were lysed using an ammonium chloride–potassium bufferfor 1.5minutes at roomtemperature. Lung cellswere blockedwithanti-mouse CD16/32 (Biolegend, 101320) and then stained withantibodies against Ly-6G (Biolegend, 127628), CD11b (Biole-gend, 101206), F4/80 (Biolegend, 123110), CD45 (Biolegend,103132), CD11c (Biolegend, 117318), I-Aq (Biolegend, 115208),CD40 (Biolegend, 124614), CD11b (Biolegend, 101218), CSF-1R (Biolegend, 135510), CD206 (Biolegend, 141712), CD64(Biolegend, 139306), CD80 (Biolegend, 104718), CD86 (Biole-gend, 105020), Annexin-V (Biolegend 640905), and DAPI (LifeTechnologies, D1306). Anti-human EGFR was purchased fromBiolegend (352906). Flow cytometric data acquisition was per-formed on a LSRII flow cytometer (BD Biosciences), and data

Translational Relevance

Despite recent advances in treatment of EGFR-mutant lungcancer, narrow therapeutic indices and frequent acquiredresistance limit their overall success rate. We reasoned thatinhibition of immunologic pathways that support tumorgrowth and maintenance in this disease could represent analternative treatment approach to provide long-lived tumordestruction and surveillance. In this communication, we dem-onstrate that selective depletion of alveolarmacrophages fromEGFR-mutant lung tumor–bearing mice results in dramaticreduction of tumor burden indicating that these alveolarmacrophages play a critical role in the growth and survivalof tumors in thismodel. In tumors forwhich targeted therapiesare available, there could be an important role for combiningimmunotherapy-based targeting of tumor-associated macro-phages with targeted therapy to improve cancer survival.

Alveolar Macrophages and EGFR-Mutant Lung Adenocarcinoma

www.aacrjournals.org Clin Cancer Res; 23(3) February 1, 2017 779




analysis was performed using FlowJo software. AM (CD45þF4/80þCD11cþLy6G�) and EGFR-mutant epithelial cells (CD45�

human EGFRþ) were sorted from single lung suspensions ofbitransgenic mice fed doxycycline for 7 weeks using a FACSAriaII Cell Sorter (BD Biosciences).

Phagocytosis assaySingle lung cell suspensions were rinsed in 1% BSA/HBSS (Life

Technologies), and 200 mL of lung cells (1 � 106 cells/mL) wereadded to triplicatewells 96-well V-bottomplates (CorningCostar;Sigma-Aldrich). Fluorescein isothiocyanate (FITC)-dextran(molecular weight 40,000; Sigma, FD40S) was then added fora final concentration of 1.0 mg/mL. Samples were incubated ateither 37�C, or 4�C to determine basal uptake of FITC-dextran.Samples were collected at 30, 60, and 120 minutes. The sampleswere thenwashed twice with 1%BSA/HBSS at 4�C and stained forcell surface markers for analysis by flow cytometry.

Cytokine assaysAMs and EGFR-mutant epithelial cells were sorted from single

lung cell suspensions using a BD FACSAria II (BD Biosciences)and rinsed with 10% FBS/RPMI1640 (Life Technologies). Twohundred and fifty thousand cells were cultured in 200 mL of RPMIwith 10% FCS for 24 hours at 37�C. Supernatants were assayed byLuminex (Life Technologies) assays and Proteome Profiler cyto-kine assays (Mouse Cytokine Array Panel A Kit, R&D Systems)according to the manufacturer's instructions.

Gene expression data and patient cohortsDNA microarray data from AMs of human smokers (n ¼ 15)

and nonsmokers (n ¼ 15), as well as frommurine AMs from twomouse models of emphysema including Integrin-b-6–deficient(Itgb6�/�) mice and transgenic mice with CC10 promoter–drivenoverexpression of IL13 [Tg(cc10-IL13)Stat6þ/�)] were obtainedfrom the National Center for Biotechnology Information (NCBI)Gene Expression Omnibus (GEO) database (www.ncbi.nlm.nih.gov/geo, accession number GSE2125; ref. 20). DNA microarraydata from the lungs of CCSP/EGFR (n ¼ 4), transgene negative(n ¼ 4), and monotransgenic littermate mice (n ¼ 4) fed doxy-cycline for 3 to 6 months were obtained from the NCBI GEO(GSE17373; ref. 21). We obtained mRNA sequencing data (n ¼515) of lung adenocarcinoma from The Cancer Genome Atlas(TCGA) portal (https://gdc-portal.nci.nih.gov/) and the resultsshown here are in whole or part based upon data generated by theTCGA Research Network (http://cancergenome.nih.gov/). Geneexpression data from the tumors of patients undergoing resectionfor early-stage lung cancer from the JapanNational Cancer CenterResearch Institute (JNCC cohort, n¼ 226)were obtained from theNCBI GEO (GSE31210; ref. 22).

Statistical analysis of microarray dataBiometric Research Branch (BRB)-Array Tools were used for

statistical analysis of the gene expression data (23), and all otherstatistical analyseswere performed in the R language environment(http://www.r-project.org). All gene expression data were gener-ated using the Affymetrix platform (U133 plus 2.0). Raw datafrom the Affymetrix platform were downloaded from publicdatabases and normalized using a robust multiarray averagingmethod (24). To predict the class of the independent patientcohort, gene expression data in the training set were combined to

form a series of classifiers according to the compound covariatepredictor (CCP) algorithm as described in previous publications(25) and the robustness of the classifier was estimated by themisclassification rate determined during leave-one-out cross-val-idation (LOOCV) of the training set. When applied to the inde-pendent validation sets, prognostic significance was estimated byevaluating the differences between Kaplan–Meier plots and log-rank tests between the two predicted subgroups of patients. AfterLOOCV, the sensitivity and specificity of the prediction modelswere estimated by the fraction of samples correctly predicted. Toenumerate the frequency of AMs in tumor tissues by the AMactivation (AMac) mRNA signature, we utilized CIBERSORT(http://cibersort.stanford.edu/), a novel analytic method for"in silico flow cytometry" that accurately estimates the abundancesof member leukocyte subsets among a mixed cell populationusing gene expression data (26).

ResultsOncogenic EGFR initiates expansion of AMs

Histologic sections of lungs from bitransgenic (CCSP/EGFR)mice fed doxycycline for 7 weeks demonstrated an increasedpercentage of cells expressing the macrophage marker F4/80compared with control bitransgenic mice fed normal chow(Fig. 1A). Analysis of single lung cell suspensions by flow cyto-metry confirmed increased relative abundance of AMs in the lungsofmice feddoxycycline. At 4weeks, 56.1%�4.3%(mean�SE)ofCD45þ hematopoetic cells in the lung coexpressed F4/80 andCD11c and were negative for the granulocyte marker Ly-6G,consistent with an AM phenotype (Fig. 1B and C). By 7 weeks,81.4% � 1.1% of CD45þ lung cells were AMs. In contrast, incontrolmice fed normal chow for 7weeks, AMs comprised 17.6%� 1.5% of all CD45þ lung cells. When quantified in absolutevalues, AMs increased from an average of 0.5 million AMs in thelungs of control mice, to 11.6 million AMs in doxycycline-fedmice at 7 weeks. The expansion of AMs paralleled an increasein both lung weight (Fig. 1D), and histologic progression toadenocarcinoma (Fig. 1E and F). Human EGFR expression wasdetectedonCD45� cells andnot onAM(Supplementary Fig. S1A)anddoxycycline administration itself did not affect AMnumber orfunction (Supplementary Fig. S1B and S1C).

AMs from tumor-bearing mice exhibit an immunosuppressivephenotype and are functionally distinct from steady-state AMs

To understand whether AMs in this model of lung adenocar-cinoma are phenotypically altered, AMs from doxycycline-fedmice and control mice were compared using flow cytometry. Wefound similar expression of several known macrophage markersincluding CD11b, CD206, and CD64 in AMs isolated from thelungs of doxycycline-fed and control mice. In contrast, we foundreduced expression of costimulatory molecules, such as CD40,CD80, and CD86 in AMs from doxycycline-fed bitransgenic(CCSP/EGFR) mice, as well as downregulated MHC-II expression(Fig. 2A), consistent with an alternatively activated, or M2, mac-rophage phenotype (27). Because the functional properties ofmacrophages may better distinguish AM subsets, we evaluatedAMs from tumor-bearing mice and from control mice, withrespect to phagocytosis and cytokine secretion, two functionsthat characterizemacrophage phenotypes. For example, increasedphagocytic properties of macrophages have been correlated withan M2 phenotype (28). Thus, AMs from tumor-bearing and

Wang et al.

Clin Cancer Res; 23(3) February 1, 2017 Clinical Cancer Research780




control mice were cultured in the presence of FITC-dextran toevaluate their phagocytosis. As expected, control AMs demon-strated greater phagocytic capability compared with lung DCs(CD45þCD11cþF4/80�), which demonstrated little antigenuptake. When compared with AMs from control mice, AMs fromtumor-bearing mice demonstrated increased phagocytosis ofFITC-dextran (Fig. 2B). We next evaluated cytokines in the super-natants of 24-hour cultures of sorted AMs from tumor and controlmice. Compared with AMs from control mice, AMs from tumor-bearingmice produced greater amounts of the cytokines IL1a andTNFa, of the chemokines CXCL1 and CXCL2, as well as the IL1receptor antagonist (IL1RA; Fig. 2C). Furthermore, elevated levelsof CXCL1 were also detected in the brochoalveolar lavage (BAL)fluid of tumor-bearing mice (Fig. 2D).

Local proliferation underlies AM expansionWe sought to determine whether AM expansion in tumor-

bearing mice was caused by infiltration of blood-derived mono-cyte precursors, or local proliferation of AMs. To deplete periph-eral (blood) monocytes, doxycycline-fed bitransgenic CCSP/EGFR mice were injected intraperitoneally every other day for 6

weeks with a neutralizing antibody against CSF-1R. In thismodel,peripheral blood monocytes (CD45þCD11bþLy6Cþ) demon-strated relatively high expression of CSF-1R compared with AMs(Fig. 3A). With this intervention, peripheral blood monocyteswere significantly depleted (Fig. 3B). This depletion of peripheralblood monocytes during doxycycline feeding had no effect onlung weight (Fig. 3C), absolute number of AMs (Fig. 3D), phe-notype of AMs (Supplementary Fig. S2A), or tumor burden(Supplementary Fig. S2B). To determine the contribution of localmacrophage proliferation to the increase of AMs in tumor-bearingmice, we double stained tissue section of lungs from doxycycline-fed CCSP/EGFR and control (no doxycycline) mice for F4/80 andthe proliferation marker Ki-67 (Fig. 3E). In doxycycline-fed mice,44.8% � 6.1% of AMs expressed Ki-67, whereas in control mice3.5% � 8.4% of AMs expressed Ki-67 (Fig. 3F). To investigatewhether EGFR-mutant epithelial cells produced soluble factorsthat could potentiate AM expansion, CD45� human EGFRþ cellswere sorted from doxycycline-fed mice. Detected in the super-natants of these cultured cells was GM-CSF, M-CSF, CXCL12,IL1a, and TNFa (Fig. 3G). Taken together, these data indicate thatlocal proliferation of macrophages, not recruitment of blood

Figure 1.

Oncogenic EGFR initiates expansion of alveolar macrophages. A, Immunostaining was performed on lung sections from bitransgenic mice fed eitherdoxycycline (dox) or normal chow for 7 weeks. F4/80 (red) and DAPI (blue) stains are shown. B, Flow cytometry was performed to identify AMs (CD45þCD11cþF4/80þLy6G� cells) in control mice and mice fed doxycycline for 7 weeks. C, Percent of AMs among CD45þ lung cells in control or doxycycline-fed mice for4 or 7 weeks is shown. Expansion of AMs correlated with increase in lung weight (D) and histologic progression of adenocarcinoma (E and F). These experimentswere performed in triplicate. � , P < 0.05.






monocytes, is the dominant mechanism underlying the expan-sion of AMs in tumor-bearing mice.

AMs contribute to the progression of EGFR-mutant lungadenocarcinoma

Given the striking expansion of AMs and the immunosup-pressive phenotype of AMs in tumor-bearing mice, we hypoth-esized that tumor progression in EGFR-mutant NSCLC is drivenby AMs. To test this hypothesis, we depleted AMs in doxycy-cline-fed CCSP/EGFR mice by intratracheal delivery of clodro-nate-encapsulated liposomes. After 6 weeks, AMs in micereceiving clodronate liposomes were significantly depleted(Fig. 4A), and the lungs of mice receiving clodronate liposomeswere significantly reduced in weight compared with the micetreated the same that received control liposomes (Fig. 4B).Furthermore, mice administered clodronate liposomes demon-strated decreased tumor burden. Evaluation of the lungs ofclodronate liposome–treated mice revealed a significant reduc-

tion in the percentage of lung area that contained preinvasivelepidic adenocarcinoma (Fig. 4C and D). Moreover, whereas 0of 15 mice treated with clodronate liposomes showed evidenceof invasive foci of adenocarcinoma, 2 of 11 mice treated withcontrol liposomes exhibited areas of invasive adenocarcinoma.To eliminate the possibility that clodronate liposomes had adirect effect on the tumor cells in this model, CD45� humanEGFRþ cells were sorted from doxycycline-fed mice and cul-tured in the presence of clodronate-encapsulated lipomes,which did not result in an increase in cell death or apoptosis(Supplementary Fig. S3). To study the effect of AMs on tumorcells, AMs were cocultured with CD45� human EGFRþ cells andthese experiments demonstrated that AMs had a protectiveeffect on tumor cell death and apoptosis (Fig. 4E).

Attenuation of oncogenic EGFR results in contraction of AMsIt has been shown in this model that cessation of EGFR

signaling results in decreased progression of malignant

Figure 2.

AMs from tumor-bearing mice exhibit an immunosuppressive phenotype and are functionally distinct from steady-state AMs. Mice were fed doxycycline orregular diet for 7 weeks. A, Flow cytometry was used to compare the surface phenotype of AMs from doxycycline-fed mice (red, solid line) and controlmice (clear, solid line). B, Phagocytosis of FITC-dextran was evaluated in AMs and dendritic cells (DC; CD45þCD11cþF4/80�) from doxycycline-fed and controlmice. C, AMs were sorted from doxycycline-fed mice and cultured for 24 hours at which time supernatants were assayed for select cytokines and chemokines.D, BAL from control and doxycycline-fed mice were assayed for CXCL1. These experiments were performed in triplicate.

Wang et al.





histology (17, 29). In our hands, lungs from mice fed doxycy-cline for 7 weeks tripled in weight compared with control miceand returned to normal weight when doxycycline was with-drawn (Fig. 5A). Within 2 weeks of discontinuing doxycycline,the frequency of AMs contracted to levels similar to a controlgroup of mice, suggesting that AM expansion is associated withthe expression of mutant EGFR (Fig. 5B). To assess the effect ofclinically active inhibitors of human EGFR on the AMs, wetreated doxycycline-fed mice with either the tyrosine kinaseinhibitor (TKI) erlotinib, or an EGFR-neutralizing antibody,cetuximab. Although cetuximab is particular effective in thisL858R-driven mouse model of EGFR-mutant lung adenocarci-noma, its utility in humans is less convincing (30) and thisagent was used in our studies as a tool compound. Similar todoxycycline withdrawal, AMs contracted to baseline levels after2 weeks of drug delivery (Fig. 5C).

Significance in human lung adenocarcinomaWe investigated the clinical relevance of AM in human EGFR-

mutant lung adenocarcinoma by constructing an "activated"AM mRNA signature (AMac signature) composed of 72 genesoverlapped between AMs from human smokers and mouseAMs from at least one of 2 mouse models of emphysema, andwhich are differentially expressed compared with human AMsfrom nonsmokers (GSE2125). Clustering demonstrated twodistinct patterns when comparing AMs between smokers andnonsmokers (Fig. 6A). Using CIBERSORT, a novel analyticmethod for "in silico flow cytometry" that accurately estimates

the abundances of member leukocyte subsets among a mixedcell population using gene expression data (26), AMs wereenumerated from gene expression profiling data of whole lungfrom EGFR-mutant adenocarcinoma mice. As expected, AMscomprised a large fraction of immune cells in the lungs of thesemice (Fig. 6B). Among 515 human lung adenocarcinomatumors from the TCGA, enumeration of AMs by CIBERSORTdemonstrated higher frequency of AMs within 32 EGFR-mutanttumors compared with 408 EGFR wild-type (WT)/KRAS WTtumors (P ¼ 0.036) and 75 KRAS-mutant tumors (P ¼0.073; Fig. 6C). The impact of the AMac signature on overallsurvival was evaluated in 226 patients with early-stage lungadenocarcinoma undergoing complete surgical resection fromthe Japanese National Cancer Center (GSE31210). Among 127patients with an EGFR mutation, patients whose tumorsexpressed the AMac signature had significantly worse overallsurvival than those who did not; and among 99 patients withEGFR WT tumors, there was no difference in survival of patientswith AMacþ or AMac� tumors (Fig. 6D).

DiscussionEGFR-mutant lung adenocarcinoma is the archetypal "onco-

gene-addicted" tumor in which dramatic and sustained tumorregression results from specific inactivation of a single onco-gene. In patients with human EGFR-mutant lung adenocarci-noma, for example, inhibition of oncogenic EGFR signalingby EGFR-specific tyrosine kinase inhibitors results in clinical

Figure 3.

Local proliferation underlies AMs expansion. A, CSF-1R expression on AMs and on peripheral blood monocytes in mice fed doxycycline or regular diet for 7 weeks.B, Peripheral blood monocytes were depleted in doxycycline-fed mice by intraperitoneal administration of anti-CSF-1R antibodies. Lung weight (C) andabsolute numbers of AMs (D) were measured in doxycycline-fed mice administered intraperitoneal anti-CSF-1R antibodies. These data are pooled from twoseparate experiments. E, Lung sections from control and doxycycline-fed mice were stained for F4/80 (red) and Ki-67 (dark brown) and sections from tworepresentative mice are shown. F, Percent of AMs that stained positive for Ki-67 is compared between control and doxycycline-fed mice. Four representativesections from3mice in eachgroupwere evaluated.G,CD45�humanEGFRþ cells (CD45�huEGFRþ) frommice feddoxycycline for 7weekswere cultured for 24hoursand supernatant cytokines/chemokines were assayed.






responses that are dramatic, but short lived (7). In fact, essen-tially all patients with EGFR-mutant lung adenocarcinomatumors treated with EGFR inhibitors will ultimately progressthrough treatment. It is therefore necessary to understand themechanisms underlying oncogene-dependent and -indepen-dent tumor progression.

Although oncogene inactivation is associated with cellularsenescence and cancer cell death, the precise mechanisms under-lying oncogene addiction are obscure and not well understood.Provocatively, several oncogenes have been shown to influencethe host immune response including tumor MHC expression ortumor differentiation antigens (31, 32), and cytokine secretion orgene expression (33, 34). Furthermore, it has been demonstratedthat an intact immune system, specifically the T-cell compart-ment, is required for sustained tumor regression following onco-gene inactivation in mouse models of lymphoma and leukemia(35), and for potent responses to oncogene-targeted moleculartherapy in mouse gastrointestinal stromal tumors (36).

In a mouse model of EGFR-mutant lung adenocarcinoma, wehave discovered that oncogenic EGFR signaling results in local

expansion of AMs with a distinct phenotype that drives tumorprogression. Macrophages are a dynamic and plastic lineage ofcells that can be categorized, by phenotype and function, asclassically activated (M1) immunostimulatory macrophages thatfacilitate host responses against foreign pathogens and tumorcells or alternatively activated (M2) immunoregulatory macro-phages involved in tissue remodeling, angiogenesis, and facil-itation of protumor immune responses (12, 37). Althoughsuch extreme forms of polarization are conceptually appealing,these often oversimplified and overlapping macrophage defi-nitions may be limiting within the tumor microenvironmentas tumor-associated macrophages (TAM) engage in a widerange of biologic functions that are variable in different tumors(38). For example, in most large-scale transcriptome analysis,macrophages demonstrate a mixed phenotype expressing bothM1 and M2 markers (38).

In this genetically engineered mouse model of mutant EGFRlung adenocarcinoma, AMs in tumor-bearing mice were com-pared with AMs in control mice and demonstrated decreasedcell surface expression of MHC-II and costimulatory molecules

Figure 4.

AMs contribute to the progression of EGFR-mutant lung adenocarcinoma. Frequency of AM (A) and lung weights (B) were measured in mice fed doxycyclinefor 6 weeks and treated with either empty or clodronate-encapsulated liposomes by an intratracheal route. C, H&E staining of lungs from control mice(non-doxycycline fed), and doxycycline-fed mice receiving either empty or clodronate-encapsulated liposomes are shown. D, Percentage of lung area thatwas occupied by adenocarcinoma was determined in doxycycline-fed mice receiving either empty or clodronate-encapsulated liposomes. These data arepooled from two separate experiments. E, DAPI and Annexin V expression were evaluated on CD45� human EGFRþ cells (CD45�huEGFRþ) cultured in thepresence (1:1 ratio) or absence of AMs from mice fed doxycycline for 7 weeks.

Wang et al.





CD40, CD80, and CD86; increased production of CXCL1,CXCL2, and IL1RA; and increased phagocytosis. Taken togeth-er, these characteristics can be considered representative of animmunoregulatory macrophage phenotype. Decreased macro-phage MHC-II expression (27, 39), increased macrophageIL1RA expression (40, 41), and increased macrophage phago-cytic ability (28) have each been associated with an "M2"macrophage phenotype. In addition, AM in EGFR-mutanttumor–bearing mice produced high levels of the chemokinesCXCL1 and CXCL2, each of which has been associated withtumor promotion in breast (42), skin (43), and esophagealcancer (44). Specific to NSCLC, an immunoregulatory macro-phage population may be present in human EGFR-mutantNSCLC and may have clinical importance. In a study of 88patients with advanced lung adenocarcinoma who had beentreated with an EGFR TKI, one subset of TAM with a CD68þ

MMRþ phenotype was correlated with unfavorable overallsurvival and with unfavorable responses to TKIs (45).

Our finding that selective depletion of AMs from mutantEGFR tumor–bearing mice results in dramatic reduction oftumor burden indicates that AMs play a critical role in thegrowth and survival of tumors in this model. Macrophages arean important and abundant cellular component of the tumorstroma in both mouse and human tumors, and in recent years,it has become apparent that nonmalignant cells in the micro-environment provide essential support for the malignant phe-notype. In human lung cancer and other tumors, the abun-dancy of macrophages is often correlated with worse cancer-specific and overall survival outcomes (9, 46). In tumors,macrophages can promote many important features of tumorprogression including tumor cell motility and invasion, angio-genesis, and stimulation of persistent tumor cell growth (38).Substantial evidence indicates that intratumoral macrophages,rather than being tumoricidal or immunostimulatory, adopt aprotumoral phenotype in vivo both in the primary and meta-static sites. These data, together with experimental studiesshowing inhibition of tumor progression and metastasis byablation of intratumoral macrophages, suggest that these cellsmay represent an important therapeutic target for cancer treat-ment, and indeed these strategies are moving forward in theclinic. For example, there are currently ongoing phase I clinicaltrials evaluating neutralizing agents against CSF-1R on intra-tumoral macrophages in patients with metastatic solid tumors

(ClinicalTrials.gov NCT01316822 and NCT01346358). Inaddition, the marine antineoplastic alkaloid, trabectedin,which is used to treat patients with ovarian cancer and softtissue sarcoma, may specifically target TAMs. In mice, thiscompound demonstrates potent antitumor immunity via sup-pression of monocyte recruitment to tumor sites, inhibition ofmacrophage differentiation, and initiation of apoptosis intumor macrophages (14, 47). Furthermore, agonistic anti-CD40 antibodies have been used to activate intratumoralmacrophages in vivo and result in regression of pancreaticcarcinoma in mice and in humans (16, 48).

Our studies of AMs demonstrated that local macrophageturnover was the predominant mechanism by which AMsincreased in the lungs of tumor-bearing mice. Whereas intra-tumoral macrophages can be potentially derived from a circu-lating monocyte pool that infiltrates tumors differentiate intomacrophages (49), local proliferation is another mechanism bywhich macrophages can potentially repopulate resident tissuesincluding tumors. Lung macrophages, for example, have beenshown to repopulate locally throughout adult life, both insteady state and after cell turnover, in a way that is predom-inantly independent of circulating monocytes (50). In ourexperiments, AMs comprised approximately 80% of all lungimmune cells after 7 weeks of doxycycline-induced oncogenicEGFR signaling. Depletion of circulating monocytes by anti-CSF-1R–depleting antibodies did not affect the frequency ofAMs in these mice and AMs in doxycycline-fed mice demon-strated significantly increased nuclear staining of the prolifer-ation protein Ki-67. Notably, where anti-CSF-1R therapy hasbeen used to deplete macrophages in other mouse tumormodels (13), this antibody depleted monocytes but not lungmacrophages in our model, and this was likely secondary todifferential expression of CSF-1R on monocytes and macro-phages in these mice.

Although the focus of cancer treatment has evolved over thepast two decades from relatively nonspecific cytotoxic agentsto selective, mechanism-based therapeutics, these targetedtherapeutics are often limited by a narrow therapeutic indexand frequent acquired resistance, as is the case for EGFR-mutant lung cancer. Although targeted strategies aim toinhibit crucial molecular pathways necessary for tumorgrowth and maintenance, immunotherapy stimulates a hostimmune response aimed at generating long-lived tumor

Figure 5.

Attenuation of oncogenic EGFR results in contraction of AMs. Lung weights (A) and absolute number of AMs (B) from control mice were compared withmice fed doxycycline for 7 weeks, and doxycycline-fed mice who had doxycycline withdrawn for either the last 1 or 2 weeks of this time period. C, Controlmice were compared with doxycycline-fed mice treated with erlotinib or cetuximab for the last 1 or 2 weeks of doxycycline administration. These experimentswere performed in duplicate. � , P < 0.01.






destruction and surveillance. In tumors for which targetedtherapies are available, there could be an important role forcombining immunotherapy with targeted therapy. Collective-ly, our data support the hypothesis that AMs play an impor-tant role in the progression of EGFR-mutant NSCLC. In thistumor, AMs may therefore be a suitable target for immuno-therapy, alone and possibly in combination with EGFR-tar-geted therapy.

Disclosure of Potential Conflicts of InterestT.H. Wheeler is a consultant/advisory board member for PathXL and DNA-

SeqAlliance. No potential conflicts of interest were disclosed by the otherauthors.

Authors' ContributionsConception and design: D.-H. Wang, H.-S. Lee, B.M. BurtDevelopment of methodology: D.-H. Wang, H.-S. Lee, G. Berry, B.M. BurtAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): D.-H. Wang, H.-S. Lee, D. Yoon, G. Berry,D.J. Sugarbaker, B.M. BurtAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): D.-H. Wang, H.-S. Lee, D. Yoon, T.M. Wheeler,F. Kheradmand, B.M. BurtWriting, review, and/or revision of the manuscript: D.-H. Wang, H.-S. Lee,D. Yoon, D.J. Sugarbaker, F. Kheradmand, E. Engleman, B.M. BurtAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases):D.-H.Wang,H.-S. Lee,D. Yoon, D.J. Sugarbaker,B.M. Burt

A B

C TCGA LUAD (n = 515)P = 0.036 P = 0.073

D

60%

Non-smoker

AMac Signature

Smoker

P < 0.001 P < 0.00150%

40%

30%

Est

imat

ed f

ract

ion

of

AM

acR

NA

(%

) by

CIB

ER

SO

RT

20%

10%

0%Transgene negative

(n = 5)

EGFR Mutation (n = 127) EGFR WT (n = 99)

AMac (-)

AMac (+)AMac (-)

AMac (+)

P = 0.559P = 0.037

Monotransgeniclittermates

(n = 5)

CCSP/EGFR(n = 4)

1.00

0.75

0.50

Su

rviv

al p

rob

abili

ty

Su

rviv

al p

rob

abili

ty

0.25

0.00

1.00

0.75

0.50

0.25

0.00

0

AMac (+)KRAS Mutation(n = 75)

EGFR Mutation(n = 32)

EGFR/KRAS WT(n = 408)

50%

40%

30%

20%

10%

0%Est

imat

ed f

ract

ion

of

AM

acR

NA

(%

) by

CIB

ER

SO

RT

AMac (-)

67 67

60 59

64 53

55

43 34

35

14

15

9

7

6

3

2

146

Numbers at risk Numbers at risk

60

AMac (+)

AMac (-)

60 57

38 38

50 44

33

36 25

16

15

5

9

3

6

2

3

00

1

2639

12 24 36 48 60

OS (months)72 84 96 108 0 12 24 36 48 60

OS (months)72 84 96 108 120

Figure 6.

Clinical significance of AMs in human EGFR-mutant lung adenocarcinoma. An activated AM mRNA signature (AMac signature) comprised of 72 genesoverlapped between AMs from human smokers and mouse AMs from at least one of two mouse models of emphysema, and which are differentially expressedcompared with human AMs from nonsmokers (GSE2125). A, The signature is demonstrated by two distinct clusters of genes differentially expressed in the AMsof human smokers and nonsmokers, with red representing high expression and green representing low expression. B, CIBERSORT was utilized to enumerateAMs from gene expression profiling data of whole lung from EGFR/CCSP bitransgenic mice, from transgene-negative mice, and from monotransgeniclittermate mice fed doxycycline (GSE31210). C, Among 515 human lung adenocarcinoma tumors from the TCGA, the proportion of activated AMs wereestimated by CIBERSORT in tumors with an EGFR mutation (n ¼ 32), in EGFR WT and KRAS WT tumors (n ¼ 408), and in tumors with a KRAS mutation(n ¼ 75). D, The Japanese National Cancer Center cohort (GSE31210) was used to determine the association of the AMac signature and overall survivalwas evaluated in 127 patients undergoing resection for an EGFR-mutant early-stage lung adenocarcinoma, and in 99 patients with an early-stage EGFR WTlung adenocarcinoma.

Wang et al.





Study supervision: H.-S. Lee, B.M. BurtOther (reviewed microscopic slides and provided grading of specimens):G. Berry

Grant SupportThis work was supported by the Thoracic Surgery Foundation for Research

and Education of Baylor College of Medicine to Bryan M. Burt.

The costs of publication of this article were defrayed in part by thepayment of page charges. This article must therefore be hereby markedadvertisement in accordance with 18 U.S.C. Section 1734 solely to indicatethis fact.

Received November 20, 2015; revised June 24, 2016; accepted July 17, 2016;published OnlineFirst August 5, 2016.

References1. GoldstrawP, Ball D, Jett JR, LeChevalier T, LimE,NicholsonAG, et al. Non-

small-cell lung cancer. Lancet 2011;378:1727–40.2. American Cancer Society. Cancer facts and figures 2015. Available from:

http://www.cancer.org/acs/groups/content/@editorial/documents/docu-ment/acspc-044552.pdf.

3. Jutten B, Rouschop KM. EGFR signaling and autophagy dependencefor growth, survival, and therapy resistance. Cell Cycle 2014;13:42–51.

4. Shigematsu H, Lin L, Takahashi T, Nomura M, Suzuki M, Wistuba II,et al. Clinical and biological features associated with epidermal growthfactor receptor gene mutations in lung cancers. J Natl Cancer Inst2005;97:339–46.

5. Bell DW, Lynch TJ, Haserlat SM, Harris PL, Okimoto RA, Brannigan BW,et al. Epidermal growth factor receptormutations and gene amplification innon-small-cell lung cancer: molecular analysis of the IDEAL/INTACTgefitinib trials. J Clin Oncol 2005;23:8081–92.

6. Ettinger DS, Akerley W, Borghaei H, Chang AC, Cheney RT, Chirieac LR,et al. Non-small cell lung cancer. J Natl Compr Cancer Netw 2012;10:1236–71.

7. Ohashi K, Maruvka YE, Michor F, PaoW. Epidermal growth factor receptortyrosine kinase inhibitor-resistant disease. J Clin Oncol 2013;31:1070–80.

8. Coussens LM, Zitvogel L, Palucka AK. Neutralizing tumor-promotingchronic inflammation: a magic bullet? Science 2013;339:286–91.

9. Welsh TJ, Green RH, Richardson D, Waller DA, O'Byrne KJ, Bradding P.Macrophage and mast-cell invasion of tumor cell islets confers a markedsurvival advantage in non-small-cell lung cancer. J Clin Oncol 2005;23:8959–67.

10. Zaynagetdinov R, Sherrill TP, Polosukhin VV, HanW, Ausborn JA, McLoedAG, et al. A critical role formacrophages in promotion of urethane-inducedlung carcinogenesis. J Immunol 2011;187:5703–11.

11. Ji H, Houghton AM, Mariani TJ, Perera S, Kim CB, Padera R, et al. K-rasactivation generates an inflammatory response in lung tumors. Oncogene2006;25:2105–12.

12. Ruffell B, Affara NI, Coussens LM. Differential macrophage programmingin the tumor microenvironment. Trends Immunol 2012;33:119–26.

13. Hume DA, MacDonald KP. Therapeutic applications of macrophage col-ony-stimulating factor-1 (CSF-1) and antagonists of CSF-1 receptor(CSF-1R) signaling. Blood 2012;119:1810–20.

14. Germano G, Frapolli R, Belgiovine C, Anselmo A, Pesce S, Liguori M, et al.Role of macrophage targeting in the antitumor activity of trabectedin.Cancer Cell 2013;23:249–62.

15. Tang X, Mo C, Wang Y, Wei D, Xiao H. Anti-tumour strategies aiming totarget tumour-associated macrophages. Immunology 2013;138:93–104.

16. Beatty GL, Chiorean EG, Fishman MP, Saboury B, Teitelbaum UR, Sun W,et al. CD40 agonists alter tumor stroma and show efficacy against pan-creatic carcinoma in mice and humans. Science 2011;331:1612–6.

17. Politi K, Zakowski MF, Fan PD, Schonfeld EA, Pao W, Varmus HE. Lungadenocarcinomas induced in mice by mutant EGF receptors found inhuman lung cancers respond to a tyrosine kinase inhibitor or to down-regulation of the receptors. Genes Dev 2006;20:1496–510.

18. Ji H, Li D, Chen L, Shimamura T, Kobayashi S, McNamara K, et al. Theimpact of human EGFR kinase domain mutations on lung tumorigen-esis and in vivo sensitivity to EGFR-targeted therapies. Cancer Cell2006;9:485–95.

19. DuPage M, Dooley AL, Jacks T. Conditional mouse lung cancer modelsusing adenoviral or lentiviral delivery of Cre recombinase. Nat Protoc2009;4:1064–72.

20. Woodruff PG, Koth LL, Yang YH, Rodriguez MW, Favoreto S, Dolga-nov GM, et al. A distinctive alveolar macrophage activation state

induced by cigarette smoking. Am J Respir Crit Care Med 2005;172:1383–92.

21. Regales L, Gong Y, Shen R, de Stanchina E, Vivanco I, Goel A, et al. Dualtargetingof EGFR canovercome amajor drug resistancemutation inmousemodels of EGFR mutant lung cancer. J Clin Invest 2009;119:3000–10.

22. Okayama H, Kohno T, Ishii Y, Shimada Y, Shiraishi K, Iwakawa R, et al.Identification of genes upregulated in ALK-positive and EGFR/KRAS/ALK-negative lung adenocarcinomas. Cancer Res 2012;72:100–11.

23. Simon R, Lam A, Li MC, Ngan M, Menenzes S, Zhao Y. Analysis of geneexpression data using BRB-ArrayTools. Cancer Inform 2007;3:11–7.

24. Irizarry RA, Hobbs B, Collin F, Beazer-Barclay YD, Antonellis KJ, Scherf U,et al. Exploration, normalization, and summaries of high density oligo-nucleotide array probe level data. Biostatistics 2003;4:249–64.

25. Oh SC, Park YY, Park ES, Lim JY, Kim SM, Kim SB, et al. Prognostic geneexpression signature associated with two molecularly distinct subtypes ofcolorectal cancer. Gut 2012;61:1291–8.

26. Newman AM, Liu CL, Green MR, Gentles AJ, Feng W, Xu Y, et al. Robustenumeration of cell subsets from tissue expression profiles. Nat Methods2015;12:453–7.

27. CavnarMJ, Zeng S, KimTS, Sorenson EC,Ocuin LM, BalachandranVP, et al.KIT oncogene inhibition drives intratumoralmacrophageM2polarization.J Exp Med 2013;210:2873–86.

28. Rey-Giraud F, Hafner M, Ries CH. In vitro generation of monocyte-derivedmacrophages under serum-free conditions improves their tumor promot-ing functions. PLoS One 2012;7:e42656.

29. Pao W. Monitoring EGFR-mutant lung cancers by means of the blood.J Thorac Oncol 2006;1:199–200.

30. Cho J, Chen L, Sangji N, Okabe T, Yonesaka K, Francis JM, et al. Cetuximabresponse of lung cancer-derived EGF receptor mutants is associated withasymmetric dimerization. Cancer Res 2013;73:6770–9.

31. Boni A, Cogdill AP, Dang P, Udayakumar D, Njauw CN, Sloss CM, et al.Selective BRAFV600E inhibition enhances T-cell recognition of melanomawithout affecting lymphocyte function. Cancer Res 2010;70:5213–9.

32. Zheng P, Guo Y, Niu Q, Levy DE, Dyck JA, Lu S, et al. Proto-oncogene PMLcontrols genes devoted to MHC class I antigen presentation. Nature1998;396:373–6.

33. Sparmann A, Bar-Sagi D. Ras-induced interleukin-8 expression plays acritical role in tumor growth andangiogenesis. CancerCell 2004;6:447–58.

34. Ancrile B, Lim KH, Counter CM. Oncogenic Ras-induced secretion of IL6 isrequired for tumorigenesis. Genes Dev 2007;21:1714–9.

35. Rakhra K, Bachireddy P, Zabuawala T, Zeiser R, Xu L, Kopelman A, et al.CD4(þ) T cells contribute to the remodeling of the microenvironmentrequired for sustained tumor regression upon oncogene inactivation.Cancer Cell 2010;18:485–98.

36. Balachandran VP, Cavnar MJ, Zeng S, Bamboat ZM, Ocuin LM, Obaid H,et al. Imatinib potentiates antitumor T cell responses in gastrointestinalstromal tumor through the inhibition of Ido. NatMed 2011;17:1094–100.

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

38. Qian BZ, Pollard JW. Macrophage diversity enhances tumor progressionand metastasis. Cell 2010;141:39–51.

39. Mantovani A, Sica A, Sozzani S, Allavena P, Vecchi A, Locati M. Thechemokine system in diverse forms of macrophage activation and polar-ization. Trends Immunol 2004;25:677–86.

40. Mantovani A. From phagocyte diversity and activation to probiotics: backto Metchnikoff. Eur J Immunol 2008;38:3269–73.

41. Stahl M, Schupp J, Jager B, Schmid M, Zissel G, Muller-Quernheim J, et al.Lung collagens perpetuate pulmonary fibrosis via CD204 and M2 macro-phage activation. PLoS One 2013;8:e81382.






42. Acharyya S, Oskarsson T, Vanharanta S,Malladi S, Kim J,Morris PG, et al. ACXCL1 paracrine network links cancer chemoresistance and metastasis.Cell 2012;150:165–78.

43. Cataisson C, Ohman R, Patel G, Pearson A, Tsien M, Jay S, et al. Induciblecutaneous inflammation reveals a protumorigenic role for keratinocyteCXCR2 in skin carcinogenesis. Cancer Res 2009;69:319–28.

44. Wang B, Hendricks DT, Wamunyokoli F, Parker MI. A growth-relatedoncogene/CXC chemokine receptor 2 autocrine loop contributes to cel-lular proliferation in esophageal cancer. Cancer Res 2006;66:3071–7.

45. Zhang B, Zhang Y, Zhao J, Wang Z, Wu T, Ou W, et al. M2-polarizedmacrophages contribute to the decreased sensitivity of EGFR-TKIs treat-ment in patients with advanced lung adenocarcinoma. Med Oncol2014;31:127.

46. Bingle L, Brown NJ, Lewis CE. The role of tumour-associated macrophagesin tumour progression: implications for new anticancer therapies. J Pathol2002;196:254–65.

47. Allavena P, Signorelli M, Chieppa M, Erba E, Bianchi G, Marchesi F, et al.Anti-inflammatory properties of the novel antitumor agent yondelis (tra-bectedin): inhibition of macrophage differentiation and cytokine produc-tion. Cancer Res 2005;65:2964–71.

48. Beatty GL, Torigian DA, Chiorean EG, Saboury B, Brothers A, Alavi A,et al. A phase I study of an agonist CD40 monoclonal antibody(CP-870,893) in combination with gemcitabine in patients withadvanced pancreatic ductal adenocarcinoma. Clin Cancer Res 2013;19:6286–95.

49. Franklin RA, LiaoW, Sarkar A, KimMV, BivonaMR, Liu K, et al. The cellularand molecular origin of tumor-associated macrophages. Science2014;344:921–5.

50. Hashimoto D, Chow A, Noizat C, Teo P, Beasley MB, Leboeuf M, et al.Tissue-resident macrophages self-maintain locally throughout adult lifewith minimal contribution from circulating monocytes. Immunity2013;38:792–804.


Wang et al.




2017;23:778-788. Published OnlineFirst August 5, 2016.Clin Cancer Res Don-Hong Wang, Hyun-Sung Lee, David Yoon, et al. Alveolar MacrophagesProgression of EGFR-Mutant Lung Adenocarcinoma is Driven By

Updated version

10.1158/1078-0432.CCR-15-2597doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2016/08/05/1078-0432.CCR-15-2597.DC1

Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/23/3/778.full#ref-list-1

This article cites 49 articles, 19 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/23/3/778.full#related-urls

This article has been cited by 1 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/23/3/778To 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-15-2597

http://clincancerres.aacrjournals.org/content/suppl/2016/08/05/1078-0432.CCR-15-2597.DC1

http://clincancerres.aacrjournals.org/content/23/3/778.full#ref-list-1

http://clincancerres.aacrjournals.org/content/23/3/778.full#related-urls

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

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/23/3/778


Documents

Progression of EGFR-Mutant Lung Adenocarcinoma is Driven By Alveolar … · 23(3); 778–88. 2016 AACR. Introduction Non–small cell lung cancer (NSCLC) has an overall 5-year survival