VEGF Neutralization Plus CTLA-4 BlockadeAlters Soluble and ... · Cancer Immunol Res; 4(10); 858–68. 2016 AACR. Introduction Escape from immune surveillance is a hallmark of metastatic

Research Article

VEGF Neutralization Plus CTLA-4 Blockade AltersSoluble and Cellular Factors Associated withEnhancing Lymphocyte Infiltration and HumoralRecognition in MelanomaXinqi Wu1, Anita Giobbie-Hurder2,3, Xiaoyun Liao2,4, Donald Lawrence5,David McDermott6, Jun Zhou1, Scott Rodig2,4, and F. Stephen Hodi1,7

Abstract

Immune recognition of tumor targets by specific cytotoxiclymphocytes is essential for the effective rejection of tumors. Aphase I clinical trial of ipilimumab (an antibody that blocksCTLA-4 function) in combination with bevacizumab (an anti-body that inhibits angiogenesis) in patients with metastaticmelanoma found favorable clinical outcomes were associatedwith increased tumor endothelial activation and lymphocyteinfiltration. To better understand the underlying mechanisms,we sought features and factors that changed as a function oftreatment in patients. Ipilimumab plus bevacizumab (Ipi-Bev)increased tumor vascular expression of ICAM1 and VCAM1.Treatment also altered concentrations of many circulatingcytokines and chemokines, including increases of CXCL10,IL1a, TNFa, CXCL1, IFNa2, and IL8, with decreases inVEGF-A in most patients. IL1a and TNFa induced expressionof E-selectin, CXCL1, and VCAM1 on melanoma tumor-asso-

ciated endothelial cells (TEC) in vitro and promoted adhesionof activated T cells onto TEC. VEGFA inhibited TNFa-inducedexpression of ICAM1 and VCAM1 and T-cell adhesion, whichwas blocked by bevacizumab. CXCL10 promoted T-cell migra-tion across TEC in vitro, was frequently expressed by mela-noma cells, and was upregulated in a subset of tumors intreated patients. Robust upregulation of CXCL10 in tumorswas accompanied by increased T-cell infiltration. Ipi-Bev alsoaugmented humoral immune responses recognizing targets inmelanoma, tumor endothelial, and tumor mesenchymal stemcells. Our findings suggest that Ipi-Bev therapy augmentsimmune recognition in the tumor microenvironment throughenhancing lymphocyte infiltration and antibody responses.IL1a, TNFa, and CXCL10, together with VEGF neutralization,contribute to Ipi-Bev–induced melanoma immune recognition.Cancer Immunol Res; 4(10); 858–68. �2016 AACR.

IntroductionEscape from immune surveillance is a hallmark of metastatic

cancer (1). Tumor-infiltrating lymphocytes (TIL) are good prog-nostic markers in a number of tumor types, including melanoma(2–7). Lymphocyte trafficking across tumor vascular endotheliainto the tumor is a multi-step process that involves rolling andhoming of lymphocytes onto endothelial cells (EC) with subse-quent transmigration across the endothelia. Tumors have devel-

oped a number of uniquemechanisms to suppress the infiltrationof immune cells (8, 9). Expression of adhesion molecules such asE-selectin, ICAM1, and VCAM1 in the tumor vasculature is inhib-ited, in addition to the reduced expression of chemokines criticalfor lymphocyte recruitment into tumors (8–14).

Clinical studies show that ipilimumab, a fully human mono-clonal antibody that blocks the immune checkpoint CTLA-4,improves overall survival in a subset of patients with metastaticmelanoma (15, 16). The addition of bevacizumab, a neutralizingantibody for VEGFA, to ipilimumab treatment generated favor-able outcomes in patients with metastatic melanoma (17). Com-pared with ipilimumab alone, ipilimumab plus bevacizumab(Ipi-Bev) led to striking tumor endothelial activation and infil-tration of lymphocytes, including CD8þ T cells (17). Here, weexamine the mechanisms by which Ipi-Bev facilitates lymphocyteinfiltration in tumors.

Given the critical roles of circulating factors that influencelymphocyte homing and trafficking across tumor endothelia, weevaluated cytokines/chemokines as well as humoral immuneinfluences on the tumor microenvironment in Ipi-Bev–treatedpatients with metastatic melanoma.

Materials and MethodsPatients

Patients with metastatic melanoma enrolled in the phase I trialof Ipi-Bev (NCT00790010) have been described (17). Patients

1Department of Medical Oncology, Dana-Farber Cancer Institute andHarvardMedical School, Boston, Massachusetts. 2Center for Immuno-oncology, Dana-Farber Cancer Institute and Harvard Medical School,Boston, Massachusetts. 3Department of Biostatistics, Dana-FarberCancer Institute, Boston, Massachusetts. 4Department of Pathology,Brigham and Women's Hospital, Boston, Massachusetts. 5Massachu-setts General Hospital Cancer Center, Boston, Massachusetts. 6BethIsrael-Deaconess Medical Center, Boston, Massachusetts. 7MelanomaDisease Center, Dana-Farber Cancer Institute and Harvard MedicalSchool, Boston, Massachusetts.

Note: Supplementary data for this article are available at Cancer ImmunologyResearch Online (http://cancerimmunolres.aacrjournals.org/).

Corresponding Author: F. Stephen Hodi, Dana-Farber Cancer Institute, 450Brookline Avenue, Boston, MA 02215. Phone: 617-632-5053; Fax: 617-582-7992;E-mail: [email protected]

doi: 10.1158/2326-6066.CIR-16-0084

�2016 American Association for Cancer Research.

CancerImmunologyResearch

Cancer Immunol Res; 4(10) October 2016858

on February 23, 2021. © 2016 American Association for Cancer Research. cancerimmunolres.aacrjournals.org Downloaded from

Published OnlineFirst August 22, 2016; DOI: 10.1158/2326-6066.CIR-16-0084

http://cancerimmunolres.aacrjournals.org/

were enrolled onDana-Farber/HarvardCancer Center Institution-al Review Board–approved protocols.

Serum collectionsBlood samples were centrifuged at 1,000x g for 15 minutes at

room temperature. The supernatants were collected and stored at� –20�C in aliquots.

Measurement of serum VEGFA, cytokines, and chemokinesVEGFA in the pre- andposttreatment (5–10weeks after initiating

therapy) serum samples were determined using MILLIPLEX MAPHumanAngiogenesis/Growth FactorMagnetic Bead Panel–CancerMultiplex Assay (Millipore). The levels of 39 cytokines andchemokines (EGF,CCL11/Eotaxin, FGF2, FLT3LG/Flt-3L,CX3CL1/Fractalkine, GM-CSF, CXCL1/GRO, CSF3/G-CSF, IFNa2, IFNg ,IL1a, IL1b, IL2, IL3, IL4, IL5, IL6, IL7, IL8, IL9, IL10, IL12B/IL12p40,IL12A/IL12p70, IL13, IL15, IL17, IL1R1/IL1ra, CXCL10/IP-10,CCL2/MCP-1, CCL7/MCP-3, CCL22/MDC, CCL3/MIP-1a, CCL4/MIP-1b, sCD40L, sIL2RA, TGFa, TNFa, TNFb, and VEGF) in theserum samples were measured using the MILLIPLEX MAP HumanCytokine/ChemokineMagneticBeadPanel I and II (Millipore).Theinstructions provided by the manufacturer were followed.

Isolation and culture of TEC and TMSCTumor samples were obtained from patients on Dana-Farber/

Harvard Cancer Center Institutional Review Board–approvedprotocols. Tumor-associated endothelial cells (TEC) and mesen-chymal stem cells (TMSC) were isolated from melanoma tissuesusing CD31þ selection in 2010. Briefly, tumor tissues were cutinto small pieces and digested with collengase, DNase, anddispase in DMEM. After filtration through a cell strainer, cellsuspension was incubated with CD31 Dynabeads as guided bythe manufacturer (Life Technologies). CD31þ cells were expand-ed in EGM2 (Lonza) and purified once more using the CD31Dynabeads. The isolated CD31þ ECs were confirmed to expressCD31 and VEGFR2 on cell surface by FACS analysis and used asTEC. The CD31� cells isolated from the expanded cells were usedas TMSC, as they expressed mesenchymal stem cell surface mar-kers CD105, CD73, andCD146, andwere induced to differentiateinto adipocytes.

Melanoma cell lines and cultureMelanoma cell lines utilized were established approximately

25 years ago from patients on Dana-Farber/Harvard CancerCenter Institutional Review Board–approved protocols. Theywere not authenticated. These cells have confirmed expressionof MITF and melanocytic markers. Their culture has beendescribed (18).

Immunoblot analyses of cell lysates with patient plasmaCells were lysed in 1x lysis buffer (Cell Signaling Technology)

supplemented with proteinase inhibitor cocktail (Roche), andcentrifuged for 10 minutes at 14,000 rpm. Supernatants werecollected, run on SDS gels, and transferred onto membranes. Themembranes were blocked with milk and incubated with pretreat-ment and posttreatment patient plasma samples diluted 500- to1,000-fold in PBS containing milk. Antibodies recognizing pro-teins were detected using a horseradish peroxidase (HRP)–con-jugated goat antibody to human IgG.

T-cell adhesion and transendothelial migration assaysPeripheral blood mononuclear cells (PBMC) were isolated

from normal donor blood samples using Ficoll density gradientseparation. T cells were isolated from PBMC using DynabeadsUntouched Human T Cells Kits, activated and expanded usingDynabeads Human T-Activator CD3/CD28 in R-PS10 (RPMIsupplemented with 10% FBS, 50 mg/mL penicillin, 100 mg/mLstreptomycin) and 30 U/mL IL2 as guided by the manufacturer(Life Technologies).

ECs were grown to confluence in fibronectin-coated 12-wellplates and treated with VEGF (100 ng/mL), TNFa (0.05 or1 ng/mL), IL1a (2 ng/mL), bevacizumab (25 mg/mL), VEGF plusTNFa, or VEGF plus TNFa and bevacizumab overnight. T cellswere labeled with 5 mmol/L calcein AM (Life Technologies) byincubation at 37�C for 30 minutes followed by washing with R-PS10 three times. Labeled T cells (106 cells/mL) were added to ECand incubated for 15 to 20 minutes. Unattached T cells wereremoved by washing with PBS. Adhered T cells were recordedunder an invert fluorescencemicroscope. The numbers of adheredT cells were counted and averaged from four to five randomlyselected fields. Alternatively, EC and attached T cells weredetached from the plates with trypsin and analyzed by FACS toestimate the ratio of labeled (T cells)/unlabeled (EC) cells.

T-cell transendothelial migration was assessed using TranswellInserts. TEC (1 � 105 cells in 300 mL EMG-2) were added ontofibronectin-coated inserts and incubated for 2 days. The insertscoveredwith ECmonolayerwerewashedwithR-PS10, and0.1mLT cells (107 cells/mL in R-PS10) were added onto the inserts. Thelower chamber contained 600 mL of R-PS10 with indicatedcytokines or chemokines (R&D Systems). After incubation for 4hours, migrated cells in the lower chamber were counted using ahemocytometer.

Detection of CXCR3þ CD4 and CD8 T cells in patient PBMCsFrozen PBMCs from Ipi-Bev–treated patients were thawed and

stained with APC anti-human CXCR3 (Biolegend), FITC anti-human CD4, and PE anti-human CD8 (BD Bioscience) in PBScontaining 1% PBS, 1.5 mmol/L EDTA, and human FcR blockingreagent (Miltenyl Biotech Inc.) for 30 minutes at 4�C. CXCR3þ

CD4 and CD8 T cells were detected by FACS.

Phenotyping and enumeration of circulating endothelial cellsin whole blood sample

A standard cytometry protocol for phenotypic identificationand enumeration of circulating endothelial cells (CEC) in wholeblood samples using four surfacemarkerswas followed (19). CECwere identified as CD31brightCD34þCD45�CD133� cells.

Expression of E-selectin, ICAM1, and VCAM1 in ECsECs were collected by trypsinization and incubated with PE-

conjugated anti–E-selectin, anti-ICAM1, anti-VCAM1, or an iso-type antibody (Biolegend) in PBS containing 1.0% BSA for 30minutes at 4�C. ECs were then washed with PBS, and surfaceexpression of E-selectin, ICAM1, and VCAM1 was determined byFACS analysis. Alternatively, ICAM1 and VCAM1 in ECs weredetermined by immunoblot analyses using anti-ICAM1 and anti-VCAM1 (Abcam), respectively.

IHC staining of ICAM1, VCAM1, and CXCL10Formalin-fixed paraffin-embedded blocks from pre- and post-

treatment tumor biopsies (approximately 12 weeks after

Ipilimumab Plus Anti-VEGF Augments Tumor Immune Recognition

www.aacrjournals.org Cancer Immunol Res; 4(10) October 2016 859




initiation of therapy) were collected. For antigen retrieval, 5-mm-thick sections were heated in a steamer (VCAM1) or pressurecooker (ICAM1 and CXCL10) for 30 minutes in citrate buffer pH6.0 (Invitrogen). Slides were incubated at room temperature withprimary antibodies to ICAM1 [rabbit monoclonal (mAb),ab109361; Abcam; 1:250] for 1 hour, or against VCAM1 (rabbitmAb; ab134047; Abcam; 1:200) for 25 minutes, or to CXCL10(Goat polyclonal; AF-266-NA; R&D Systems; 1:20) at 4�C over-night. For the secondary antibody, Envision anti-rabbit or rabbitanti-goat (1:200) HRP-labeled polymer (DAKO) was applied for30 minutes. The slides were visualized with diaminobenzidine(DAKO), hematoxylin counterstained, and mounted. Positiveand negative controls were included for all markers. All the slideswere evaluated and scored by a pathologist (X. Liao) blinded toclinical data. The intensity (0, negative; 1, weak; 2, moderate; 3,intense) and the percentage of staining (0%–100%) wereassessed. Scores were based on the results of staining intensitymultiplied by staining percentage. The scores of tumor cells andECswere recorded separately and added up to obtain a final score.All the scoring was performed twice with at least 7-day interval.

Statistical analysesData for 39 cytokines/chemokines collected from 43 of the 46

patients enrolled in the Ipi-Bev trial (17)were analyzed. Several ofthe cytokines had a considerable number of values below thelower limit of detection (LLD) of the assay. If 10 or fewermeasurements (less than 20%) within amarker had values belowthe LLD, then the LLDwas substituted and a quantitative analysiswas conducted. If more than 10 measurements were below theLLD, then a qualitative analysis was done, classifying the markeras either above or below the LLD. A parallel analysis was con-ducted for sCD40L that had a large number of values above theupper limit of detection (ULD).

Quantitative analyses were conducted for 16 cytokines (EGF,CCL11, FGF2, GM-CSF, CCL11, CSF3, IFNg , IL8, IL1a, IFNa2,CXCL10, CCL2, CCL22, MIP-1b, TNFa, and VEGFA). Qualitativeanalyses were conducted for the remaining 23 cytokines (IL1b,IL2, IL3, IL4, IL5, IL6, IL7, IL9, IL10, IL12B, IL2A, IL13, IL15, IL17,TGFa, FLT3LG, CX3CL1, TNFb, MCP-3, sCD40L, IL1R1, sIL2RA,and MIP-1a). For quantitative analysis, three analyses were per-formed. The first explored the prognostic relationship betweenpretreatment expression and response—complete response/par-tial response (CR/PR) versus stable disease/progressive disease(SD/PD)/uneval—or disease control (CR/PR/SD vs. PD/uneval).The second analysis calculated the fold change in expression(post/pre) and related change in expression with either responseor disease control. Both sets of analyses used the Wilcoxon rank-sum test. The third analysis, which was univariate, looked forevidence that treatment with Ipi-Bev was associated with changein expression and compared the fold change with one using theWilcoxon signed-rank test. A Bonferroni correction was used tocontrol the overall type-1 error. Based on 39 cytokines, statisticalsignificance was defined as P < 0.0013.

For qualitative analysis, each cytokine was classified as above/below the LLDat each timepoint (pre or post), except for sCD40L,which was classified into above/below the ULD. Three analyseswere also performed. The first analysis compared pre- and post-treatment for each marker and assessed change using McNemartest. A statistically significant P value for this test would suggestthat the proportion of patients with measurements below LLD/above ULD at pretreatment was different from the proportion of

patients below the LLD/above ULD posttreatment (or thereverse). The "above/below" classes at each time point were thencombined into four groups (Pre-/Posttreatment both below LLD/ULD, Pre-/Posttreatment both above LLD/ULD, Pretreatmentbelow and Posttreatment above LLD/ULD, and Pretreatmentabove and Posttreatment below the LLD/ULD). In the subsequentanalyses, these four groups were compared according to responseand according to disease control using the Fisher exact test. Astatistically significant P value would suggest that there was arelationship between cohort (or response, or disease control) andthe behavior of the biomarker. For sCD40L, qualitative analysiswas conducted with two classes: below the ULD/above the ULD.Statistical significance was defined as P < 0.0013, which adjustedthe overall type-1 error using a Bonferroni correction. Significanceof difference observed in in vitro studies was analyzed using two-tailed t test, and a P value < 0.05 was considered significant.

Immune-related adverse events (irAE) include pruritus, rash,vitiligo, fatigue, diarrhea, colitis, increased alanine aminotrans-ferase (ALT) or aspartate aminotransferase (AST), hepatitis,hypothyroidism, hypopituitarism, hypophysitis, adrenal insuf-ficiency, increased thyrotropin, or decreased corticotropin. Toassess any association of cytokines and chemokines with irAEs,patients were divided into two groups: severe (grade 3 orhigher) irAEs or not. The "not" group includes patients withno AEs (few or none), non-irAEs, or irAEs of grade 2 or less.Pretreatment cytokine/chemokine levels and fold changes werecompared according to subsequent severe irAE developmentusing Wilcoxon rank-sum tests.

ResultsIncreased tumor vascular expression of adhesion molecules

Endothelial E-selectin expression is upregulated, and lympho-cyte infiltration increases in tumors derived from melanomapatients on Ipi-Bev therapy (17).We therefore examined the effectof Ipi-Bev on the expression of adhesion molecules VCAM1 andICAM1 in the tumors derived from7patients on the trial (Fig. 1A–D; Supplementary Table S1). IHC staining revealed that VCAM1was not expressed in endothelia and melanoma cells of pretreat-ment tumors of all 7 patients, but was induced in posttreatmenttumor vasculatures of 3 patients and melanoma tumor cells of 1patient (Fig. 1A, C, and D; Supplementary Table S1). ICAM1 wasnot detected in pretreatment tumor endothelia of 5 patients,although high ICAM1 expression was observed in the pretreat-mentmelanoma cells of all 7 patients and tumor vasculatures of 2patients (Fig. 1B, C, and D; Supplementary Table S1). Among the5 patients displaying no tumor endothelial ICAM1 expression inthe pretreatment samples, Ipi-Bev induced endothelial ICAM1expression in all of them, with four of them exhibiting strongexpression (Fig. 1B and D; Supplementary Table S1). We alsoanalyzed ICAM1 and VCAM1 expression in melanoma tumorsderived from 4 ipilimumab only–treated patients, aiming toclarify the specific effect of bevacizumab on the expression ofadhesion molecules. The expression patterns of ICAM1 andVCAM1 in pretreatment tumors were overall similar to those ofIpi-Bev patients (Fig. 1C and D; Supplementary Table S1). Ipili-mumab therapy induced VCAM1 and ICAM1 expression in thetumor vasculature of 1 patient and reduced tumor endothelialICAM1 expression in 2 patients (Fig. 1C and D; SupplementaryTable S1). In addition, VCAM1 and ICAM1 expressions were notdetected in the posttreatment tumor endothelia of two additional

Wu et al.

Cancer Immunol Res; 4(10) October 2016 Cancer Immunology Research860




ipilimumab-treated patients forwhichpretreatment sampleswerenot available (Supplementary Table S1). These findings suggestthat, similar to E-selectin (17), the addition of VEGFA blockadewithbevacizumabmayhave apotent stimulatory effect on ICAM1and VCAM1 expression in tumor endothelia.

Altered circulating cytokine and chemokine concentrationsCytokines and chemokines can induce adhesion molecule

expression and influence lymphocyte infiltration in tumors.Pretreatment and posttreatment circulating concentrations of apanel of 39 cytokines/chemokines in 43 patients receiving Ipi-Bev were analyzed (Supplementary Fig. S1). The concentrationsof many circulating cytokines/chemokines were eitherincreased or decreased as a result of Ipi-Bev therapy (Supple-mentary Fig. S1). Because a significant proportion of patientshad pretreatment and/or posttreatment cytokines/chemokinesbelow the detection threshold, only 16 cytokines/chemokineshad sufficient data sets for quantitative analysis (Table 1).Among these, concentrations of TNFa, IL8, CXCL10, CXCL1,

IL1a, and IFNa2 were statistically significantly increased as afunction of treatment (Table 1; Fig. 2A; Supplementary Fig. S2).An increase of 50% or more was considered significant. Basedon this cutoff, approximately 37% to 58% of the patientsdisplayed increases in these six cytokines/chemokines (Fig.2B). Although these cytokines/chemokines have different andcomplex roles in tumor biology that are often context-depen-dent, they were loosely classified as factors of promoting(IFNa2, TNFa, IL1a, and CXCL10) or inhibiting (IL8 andCXCL1) antitumor immune activity. Approximately 14%,20%, 26%, and 16% of the patients had increases in 1, 2, 3,and 4 cytokines/chemokines of the first group, respectively, andapproximately 35% and 28% of the patients exhibited increasesin 1 and 2 cytokines/chemokines of the second group (Fig. 2C).In addition, approximately 14% of the patients did not showincreases in any of these six cytokine/chemokines, and close to7% of the patients had increases in all six cytokines/chemo-kines in response to Ipi-Bev. For cytokines/chemokines withinsufficient data for quantitative analysis, qualitative analyses

Figure 1.

Ipi-Bev therapy upregulated ICAM1 and VCAM1 expression in tumor endothelia. Paired pretreatment and posttreatment tumor biopsies from Ipi-Bev– or ipilimumab-treated patients with metastatic melanoma were stained with VCAM1– and ICAM1–specific antibodies. Inserts highlight tumor vascular staining.A, VCAM1 was not detected in the pretreatment sample from an Ipi-Bev–treated patient but highly expressed in ECs of the posttreatment sample withvery weak immunoreactivity in tumor cells. B, ICAM1 was highly expressed in tumor cells but not in the endothelia in pretreatment biopsy from an Ipi-Bev–treatedpatient, and significantly upregulated in ECs of the posttreatment sample. Representative images are shown. C, quantitative analyses of VCAM1 andICAM1 inmelanoma cells of pre- and posttreatment tumor samples.D, quantitative analyses of VCAM1 and ICAM1 in tumor endothelial cells of pre- and posttreatmenttumor biopsies. The expression of ICAM1 and VCAM1 was indicated by H-scores that were calculated by multiplying the percentage of positivecells with their intensity (1, 2, or 3).






were performed, classifying the marker as either above or belowthe LLD. Qualitative analyses found that circulating IL5 andsIL2RA levels increased upon Ipi-Bev treatment (Supplemen-tary Fig. S3).

IL1a and TNFa promote T-cell adhesionTo examine the potential influence of the six soluble factors

that increased with Ipi-Bev treatment on tumor endothelialactivation and lymphocyte trafficking, their effects on expressionof E-selectin, ICAM1, and VCAM1 in cultured TEC were assessed.E-selectin and VCAM1 had limited expression, whereas ICAM1was more highly expressed on the surface of TEC (Fig. 2D) andhuman dermal microvascular endothelial cells (HDMEC;Supplementary Fig. S4A). IL1a and TNFa significantly increasedE-selectin, ICAM1, and VCAM1 expression and promoted T-celladhesion onto TEC (Fig. 2D and E) andHDMEC (SupplementaryFig. S4A and S4B). Approximately 72% of the patients had anincrease in IL1a and/or TNFawith treatment, and approximately33% had an increase in both cytokines in response to Ipi-Bevtherapy (Fig. 2F). In addition to IL1a and TNFa, IL1b alsoincreased E-selectin, ICAM1, and CCL22 expression on HDMECand enhanced T-cell adhesion (Supplementary Fig. S4A and S4B).However, circulating IL1b was increased in only approximately23% of patients, significantly less frequently compared with IL1aand TNFa.

VEGF inhibits TNFa-induced ICAM1 and VCAM1 expressionand adhesion on TEC

Ipi-Bev treatment reduced circulating VEGFA levels of thepatients (Table 1). This prompted us to examine if VEGFA hada direct effect on basal and cytokine-induced adhesion moleculeexpression in TEC and on lymphocyte adhesion. Immunoblotanalysis demonstrated that TNFa induced ICAM1 and VCAM1expression in TEC (Fig. 3A). VEGFA substantially inhibited TNFa-induced ICAM1 and VCAM1upregulationwhile VEGFA itself hadlittle effect on the expression of both molecules (Fig. 3A). Bev-acizumab prevented the inhibition of TNFa-induced ICAM1 andVCAM1 expression by VEGFA, although bevacizumab itself hadlittle effect on ICAM1 and VCAM1 expression (Fig. 3A). Theinhibitory effect of VEGFA on TNFa-induced ICAM1 expression

and its prevention by bevacizumab was also demonstrated byFACS analysis (Fig. 3B; Supplementary Fig. S5A). As a result of itsinhibitory effect on adhesionmolecule expression, VEGFA causeda small but statistically significant reduction in TNFa-induced T-cell adhesion onto TEC, which was reversed by bevacizumab (Fig.3C). Similarly, VEGFA inhibited TNFa-induced ICAM1 expres-sion and T-cell adhesion onto human umbilical vein endothelialcell, and this inhibitory effect of VEGFA was blocked by bevaci-zumab (Supplementary Fig. S5B and S5C). These findings suggestthat VEGFA may have a direct inhibitory role in T-cell traffickingby repressing adhesion molecule expression on endothelia, andits inhibitory effect can be halted by VEGFA blockade.

CXCL10 induces T-cell transendothelial migrationT-cell recruitment into tumors requires chemoattractant(s) that

are expressed in and secreted from tumor and/or tumor stromalcells. To examine which of the six cytokines/chemokines elevatedby Ipi-Bev treatmentmight function as the chemoattractant(s) forlymphocytes, their effect on lymphocyte transendothelial migra-tion was examined. CXCL10 was the only chemokine that signif-icantly facilitated T-cell transmigration across HDMEC (Supple-mentary Fig. S6). CXCL10 also induced T-cell transmigrationacross TEC (Fig. 4A).

CXCL10 expression in melanoma tumors and upregulationTo further address the role of CXCL10 in tumor lymphocyte

recruitment in Ipi-Bev–treated patients, CXCL10 expressionwas examined in the pretreatment and posttreatment tumorbiopsies derived from the same 7 patients whose samples wereanalyzed for adhesion molecule expression. IHC stainingrevealed that CXCL10 was expressed in the pretreatment tumorcells of 5 patients (Fig. 4B; Supplementary Table S1). RobustCXCL10 upregulation was seen in the posttreatment tumors of2 patients where the expressions of VCAM1 and/or ICAM1 intumor vasculature were concomitantly increased (Fig. 4B; Sup-plementary Table S1). Increased tumor infiltration of CD3þ

and CD8þ T cells was observed in the posttreatment tumorbiopsies of 1 of the 2 patients with upregulated tumor CXCL10expression (Fig. 4C). However, the effect of Ipi-Bev on tumorCXCL10 expression was heterogeneous with reduced CXCL10

Table 1. Aggregated fold changes of circulating cytokines/chemokines in the Ipi-Bev–treated melanoma patients

Cytokine/chemokine N Min Median Max Wilcoxon signed-rank P value

TNFa 43 0.36 1.7 14.29 <0.0001IL8 43 0.04 1.67 19.5 0.0007CXCL10 43 0.42 1.52 30.68 <0.0001CXCL1 43 0.56 1.29 194.8 <0.0001IL1a 43 0.43 1.24 16.14 <0.0001IFNa2 43 0.26 1.24 10.45 0.0001CCL4 43 0.21 1.15 14.7 0.002CCL11 43 0.59 1.09 37.85 0.03EGF 43 0.06 1.06 18.83 0.29FGF2 43 0.28 1 50.11 0.17GM-CSF 43 0.32 1 4.15 0.52CSF3 43 0.15 1 8.13 0.03IFNg 43 0.09 1 41.41 0.08CCL2 43 0.15 0.98 142.3 0.73CCL22 43 0.33 0.89 142.2 0.57VEGFA 43 0.03 0.1 3.02 <0.0001NOTE: Cytokines/chemokines were quantitatively analyzed. A Bonferroni correction was used to control the overall type-1 error. Based on 39 cytokines, statisticalsignificance is defined asP <0.0013. Cytokines/chemokineswith significant increases are depicted in red. VEGF exhibited significant decrease and is depicted in blue.All others are presented in black.

Wu et al.





expression being observed in three cases where VCAM1 and/orICAM1 expression was increased (Supplementary Table S1). Ofnote, CXCL10 was not detected in tumor endothelia of all thepatients. Similar heterogeneous CXCL10 expression patternswere observed in tumors analyzed from patients treated withipilimumab alone (Supplementary Table S1).

Effect of Ipi-Bev on T-cell CXCR3 expressionCXCR3 is the receptor for CXCL10 and is expressed on

activated T cells (20). In light of increased lymphocyte infil-tration into tumors after Ipi-Bev treatment, CXCR3 expressionon T cells in Ipi-Bev–treated patients was analyzed. Pretreat-ment and posttreatment PBMCs isolated from 11 patients onIpi-Bev treatment, including 1 CR, 3 PR, 2 SD, and 5 PDpatients, were phenotyped. Among these patients, patientP12, the only patient who achieved CR, displayed lowCXCR3-positive CD4þ and CD8þ cells before the treatmentand robust increases in CXCR3þ populations as a function ofIpi-Bev therapy: from 2.7% to 63.1% in CD4þ T cells and from10.1% to 91.6% in CD8þ T cells (Fig. 4D–F). Nonetheless, the

percentages of CXCR3þ cells did not significantly change incirculating CD4þ and CD8þ T cells in the remaining patients(Supplementary Table S2).

Serum cytokine/chemokine expression and clinical outcomesIpi-Bev achieved a clinical response rate of approximately 20%

and a disease-control rate of 67.4% in patients with metastaticmelanoma (17). We next examined if pretreatment concentra-tions and changes of circulating cytokines/chemokines were asso-ciatedwith clinical outcomes to Ipi-Bev therapy. Clinical responseor disease control did not correlate with baseline concentrationsof the cytokines/chemokines analyzed here (SupplementaryTables S3 and S4), although PD patients appeared to have higherbaseline IFNa2 and CCL2 in comparison with CR/PR/SDpatients [median IFNa2 (pg/mL): 15.5 (4.3–94.9) vs. 6.3 (0.8–26.3), P ¼ 0.01; median CCL2 (pg/mL): 449.0 (168.5–755.9) vs.359.2 (3.2–705.7), P ¼ 0.07; Supplementary Table S4]. Noassociation could be found between fold changes of serumcytokines/chemokines and clinical response or disease controlafter treatment (Supplementary Tables S5–S8), except for CCL7

Figure 2.

Ipi-Bev treatment altered circulating CXCL10, TNFa, IL1a, IFNa2, CXCL1, and IL8 concentrations. A, pretreatment and posttreatment levels of selected cytokines/chemokines in a representative patient.B, proportion of patients displaying an increase by 50%ormore in the indicated cytokines/chemokines in response to Ipi-Bevtreatment. Forty-three Ipi-Bev–treated patients were analyzed. C, proportions of the patients had increases in the indicated numbers of cytokines/chemokines withpromoting (IFNa2, TNFa, IL1a, and CXCL10) or inhibiting (IL8 and CXCL1) antitumor immune activity. D, IL1a and TNFa induced adhesion molecule expressionon tumor-associated endothelial cells (TEC). Representative data from at least two independent experiments are shown. E, IL1a and TNFa promoted T-cell adhesiononto TEC. Calcein AM–labeled CD3/CD28-activated T cells were incubated with TNFa- or IL1a-treated TEC, and adhered cells were counted. Mean � SDof three independent experiments are shown. F, percentage of patients who had an increase in either IL1a or TNFa, or both as a function of Ipi-Bev treatment.






which appeared to be associated with disease control (P ¼ 0.02;Supplementary Table S8). In addition, an increase by 50% ormore in circulating IL1a, IFNa2, IL8, CXCL1, or CXCL10 was notassociated with overall survival of the patients (SupplementaryFig. S7A–S7E). However, an increase by 50% or more in circu-lating TNFa was unexpectedly found to be associated withreduced overall survival (Supplementary Fig. S7F). This associa-tion became insignificant when the fold change cutoff point wasset as 2 (data not shown). Nonetheless, a significantly higherproportion of CR/PR patients had increases in circulating IL1a by100% or more (or fold change � 2) compared with SD and PDpatients (Supplementary Fig. S8A). A similar trend was alsoobserved with increases in circulating IFNa2 by 50% or more,although it did not reach statistical significance (SupplementaryFig. S8A). Patients displaying increases in both circulating IL1aand IFNa2 appeared to have better overall survival and responsecompared with the remainder of the patient population (Sup-plementary Fig. S8B and S8C).

Serum cytokine/chemokine expression and irAEsIn the Ipi-Bev–treated patients, severe (grade 3 or higher) irAEs

include rash, fatigue, colitis, increased ALT, or AST adrenal insuf-ficiency. Twelve of 43 patients (28%) reported at least one of theseirAEs, and the remaining 31 patients had no irAEs reported or anirAE of grade 2 or less. Among the six cytokines/chemokines withsignificant increase in response to Ipi-Bev, only pretreatmentIFNa2 levels were different according to subsequent developmentof severe irAEs, with patients who developed severe irAEs hadsignificantly lower IFNa2 (P ¼ 0.01; Supplementary Table S9).Fold changes in these cytokines/chemokines were not associatedwith severe irAEs (Supplementary Table S10).

Antibody responses elicited to melanoma tumor cell andstromal cell targets

Although our findings indicate that IL1a and TNFa play a keyrole in lymphocyte homing to tumor endothelia, no associationwas seen between increases in circulating IL1a and TNFa andbetter clinical outcomes. This reiterates the importance of break-ing down immunosuppressive barriers within the tumor micro-environment so that infiltrating lymphocytes can execute their fullantitumor activity. We hypothesized that humoral immuneresponses to targets with immunosuppressive and/or angiogen-esis activities in the tumor microenvironment might be involvedin the long-term tumor destruction of responding patients treatedwith Ipi-Bev. To test this hypothesis, whole cell lysates made fromcultured melanoma cells, TECs, and TMSCs isolated from post-treatment biopsies were probed with pretreatment and posttreat-ment plasma samples derived from two of the long-term respond-ing patients (Supplementary Table S11). A number of proteins inmelanoma cells, TEC, and TMSCwere recognized by antibodies inthe pretreatment plasma samples (Fig. 5A and B). Of note, newand enhanced antibody recognitions were detected with theposttreatment samples (Fig. 5A and B), indicating that antibodyresponses to tumors, TEC, and TMSCwere elicited as a function ofIpi-Bev therapy.

DiscussionThe success of antitumor immune responses depends on effec-

tor T-cell infiltration into tumors and the inhibition of immuno-suppressive activity within the tumor microenvironment. Ipi-Bevtherapy in metastatic melanoma leads to robust upregulation ofthe adhesion molecule E-selectin in the tumor vasculature andpronounced T-cell infiltration into tumors (17). In the current

Figure 3.

VEGFA inhibits TNFa-mediated ICAM1and VCAM1 expression and T-celladhesion onto TEC. TEC were treatedwith VEGFA, TNFa, and/orbevacizumab as indicated for16 hours. A, immunoblot analysis ofICAM1 and VCAM1 expression inTEC. B, FACS analysis of ICAM1expression on TEC. C, T-cell adhesiononto TEC. Data are presented asmean � SD of two (C) or three (B)independent experiments.

Wu et al.





study, we investigated the possible mechanism(s) responsible forIpi-Bev–induced tumor lymphocyte infiltration. Ipi-Bev treat-ment increased tumor vascular expression of ICAM1 and VCAM1,adhesion molecules that are associated with T-cell infiltration inmelanoma (8). Furthermore, CXCL10, a chemokine critical for T-cell trafficking in melanoma (9, 21), was frequently expressed inmelanomaandupregulated by Ipi-Bev in aproportion of patients.These findings suggest that Ipi-Bev therapy might promote tumorT-cell infiltration by enhancing the expression of key adhesionmolecules on tumor endothelia and/or chemokines in the tumormicroenvironment. Ipi-Bev also elicited humoral immuneresponses to targets in melanoma tumor and tumor stromal cellsthat may further enhance immune recognition in the tumormicroenvironment.

Cytokines play a crucial role in controlling endothelial expres-sion of various adhesion molecules (22). TNFa and IL1a weresignificantly elevated in a high percentage of patients as a function

of Ipi-Bev therapy. Both of these cytokines induced adhesionmolecule expression on TEC isolated from tumor tissues. Thesefindings suggest that TNFa and IL1amay be critical for treatment-induced tumor endothelial expression of adhesion molecules.Many studies have shown that TNFa and IL1a are among themajor inflammatory cytokines that increase the expression ofadhesionmolecules on EC in tumor vasculature (23–30). Similarpercentages of ipilimumab- and Ipi-Bev–treated patients dis-played increases by 50%ormore in the circulating levels of TNFa,IL1a, and the other four cytokines/chemokines (SupplementaryTable S12), suggesting that ipilimumab by itself may play amajorrole in influencing serum cytokine/chemokine levels. This alsosuggests that the neutralization of VEGFA by bevacizumab maycomplement the increased lymphocyte infiltration into tumors ofIpi-Bev–treated patients. VEGFA inhibition of TNFa-inducedexpression of adhesion molecules in TEC and ability of bevaci-zumab to block the inhibitory effect of VEGFA support this

Figure 4.

CXCL10 contributes to Ipi-Bev–induced tumor lymphocyte infiltration.A,CXCL10 inducesmigration of activated T cells across TEC. Data are presented asmean� SDof three independent experiments. B, CXCL10 expression was upregulated in the posttreatment tumor biopsy from an Ipi-Bev–treated patient. C, infiltrationof CD3þ and CD8þ T cells in the pretreatment and posttreatment tumor biopsies of the same patient shown inB.D–F, Ipi-Bev increased the expression of CXCR3, thereceptor for CXCL10, on CD4þ (D) and CD8þ (E) T cells in the peripheral blood of a patient who achieved a CR. Bar presentations of percent CXCR3þ cellsin CD4þ and CD8þ T-cell populations before and after Ipi-Bev treatment are shown in F. Representative data from two independent experiments are shown.






complementarity. Our results are consistent with previouslyreported VEGFA inhibition of TNFa-induced ICAM1 and VCAM1expression in normal ECs (13, 31–33), and enhanced T-cellinfiltration into tumors by anti-VEGF in animal studies(33, 34). High pretreatment serum VEGFA is associated withworse clinical outcomes in advanced melanoma treated withipilimumab (35), and VEGFA upregulation in advanced mela-noma is associated with innate resistance to anti–PD-1 blockade(36). These findings together suggest that VEGFAmay function asa suppressor of T-cell trafficking by reducing cytokine-inducedadhesion molecule expression and T-cell homing, thus contrib-uting to resistance to immune therapy. VEGF neutralization mayresult in enhanced adhesion molecule expression and tumorlymphocyte recruitment, providing a possible synergistic mech-anism to checkpoint blockade.

CXCL10 functions as a major chemoattractant for activated Tcells through its putative receptor CXCR3 (37–39). CXCL10 is oneof the chemokines preferentially expressed in primary and met-astatic melanoma tumors with lymphocyte infiltration (21, 40).High tumor CXCL10 expression is associated with better prog-nosis in colorectal cancer (41, 42) and upregulated with ipilimu-mab therapy (43). We found that CXCL10 was expressed inmelanoma tumor cells of most patients, and that Ipi-Bev therapyupregulated tumor CXCL10 expression in association withincreased lymphocyte infiltration in a proportion of patients.Given its activity to promote T-cell migration across TEC, ourfindings suggest that CXCL10 may play an important role inmediating treatment-induced tumor T-cell infiltration. Nonethe-less, other chemoattractants such as CXCL9, CXCL11, CCL5, andSDF-1 can also promote T trafficking in tumors (8, 9, 14, 21, 43).Specifically, tumor expression of CXCL11 is associated withhigher T-cell infiltration in melanoma and favorable responsesto ipilimumab in melanoma patients (14, 43). The roles of thesecomplex chemokines in Ipi-Bev–induced T-cell infiltrationrequire further investigation.

CXCR3 is absent on na€�ve T cells but is rapidly induced uponactivation, andpreferentially remains highly expressed onCD4þTcells, effector CD8þ T cells, and innate-type lymphocytes (20). Anassociation between CXCR3 expression by peripheral T cells andfavorable clinical outcome in stage III melanoma patients hasbeen suggested (44). Interestingly, Ipi-Bev significantly increasedCXCR3 expression on circulating CD8þ and CD4þ T cells in theonly patient who achieved a CR. An increase in T-cell CXCR3expression would be expected to facilitate T-cell trafficking into

tumors. Nonetheless, it appears that Ipi-Bev did not alter T-cellCXCR3 expression frequently in patients, with additionalmechanisms needed to be explored.

Although our data suggest important roles for IL1a andTNFa in Ipi-Bev–mediated tumor T-cell infiltration, their pre-treatment concentrations or fold changes as a function oftreatment do not appear to be associated with clinical out-comes. This is not surprising because our sample size isrelatively small, and accumulating evidence has shown com-plex effects of cytokines in the tumor microenvironment,ranging from support to inhibition of tumor growth (23,45). IL1 and TNFa can be highly inflammatory in the tumormicroenvironment and facilitate tumor growth, angiogenesis,and metastasis (27). IL1a can also upregulate the expression ofimmunosuppressive COX-2, PD-L1, and PD-L2 in tumor-asso-ciated fibroblasts (46). The function of tumor-infiltrating Tcells may be suppressed by multiple mechanisms within thetumor microenvironment, such as the presence of FoxP3þ

regulatory T cells, and the expression of PD-L1 and indolea-mine-2,3-dioxygenase (47). Together, these findings under-score the importance of inhibition of immunosuppressiveactivity in the tumor microenvironment in order for infiltrat-ing T cells to exert their antitumor function. In this context, it isintriguing that increases in IL1a and IFNa2 occurred morefrequently in CR/PR patients than SD and PD patients and thatincreases in both IFNa2 and IL1a were associated with betterclinical outcomes. IFNa is known to have potent antitumorimpact, ranging from antiangiogenic to potent immunoregu-latory, antiproliferative, and proapoptotic effects. Type I IFNscan be critical to the innate immune recognition of a growingtumor in animal studies (48). Clinical studies have alsosuggested that IFNa has significant immunomodulatory andantitumor activity in metastatic melanoma and may haveadditive or synergistic effects in promoting tumor eliminationwith anti–CTLA-4 (49).

Many soluble factors such as the galectin family members areexpressed in and secreted from tumors, and possess immuno-suppressive activity. TMSC can also cause immune suppressionand protect tumor cells from immune attack (50). As a result,antibodies targeting immunosuppressive factors in the tumormicroenvironment may enhance antitumor activity of infiltratingtumor-specific T cells. Ipi-Bev could elicit antibody responses totargets in tumor cells as well as stromal cells in patients receivinglong-term clinical benefit. Antibody responses to galectins, for

Figure 5.

Ipi-Bev inducedhumoral immune responses inmelanomapatients with long-term PR. Whole cell lysates ofmelanomacells,melanoma tumor endothelial cells (TEC),and TMSC were probed with pretreatment andposttreatment plasma samples from Ipi-Bev–treatedpatients. Antibodies specific to the posttreatmentsamples and antibodies with increased levels in theposttreatment samples are indicated by arrows. A,plasma from patient P8. B, plasma from patient P6. Bothpatients achieved PRs and long-term survival (> 4 years).Representative images from two independentexperiments are shown.

Wu et al.





example, are associated with Ipi-Bev treatment in melanomapatients (17). Functional relevance for these antibody responsesand potential contribution to the increased infiltration of CD8þ Tcells in theposttreatment tumors of Ipi-Bevpatients (17) is an areaof continued investigation. Potential influences of antibodyresponses to TMSC and TEC in immune checkpoint therapy alsowarrant further exploration, in light of the reported association ofupregulation of genes involved in mesenchymal transition andangiogenesis with innate resistance of metastatic melanomatumors to anti–PD-1 therapy (36).

In summary, Ipi-Bev therapy influenced circulating immunecells (17) and CECs (Supplementary Fig. S9A and S9B), as well asmany soluble factors, including angiogenic factors, cytokines/chemokines, and antibodies. Increases in IL1a and TNFa togetherwith VEGF neutralization and tumor CXCL10 may contribute totumor lymphocyte recruitment. Antibody responses to bothtumor and stromal elements in the tumor microenvironmentmay enhance immune recognition and antitumor activity of Ipi-Bev. Our findings support continued investigation of antiangio-genesis combinations with immune checkpoint therapy toenhance activity and improve understanding of the significancein targeting stromal elements in the tumor microenvironment.

Disclosure of Potential Conflicts of InterestD. McDermott is a consultant/advisory board member for BMS. S. Rodig

reports receiving commercial research grant from Bristol Myers Squibb and hashonoraria from the speakers bureau of Bristol Myers Squibb and Perkin ElmerInc. F.S. Hodi reports receiving commercial research grant from Bristol-MyersSquibb and is a consultant/advisory board member for Genentech and Merck.No potential conflicts of interest were disclosed by the other authors.

Authors' ContributionsConception and design: X. Wu, F.S. HodiDevelopment of methodology: X. Wu, X. Liao, D. Lawrence, F.S. HodiAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): X. Wu, X. Liao, D. Lawrence, D. McDermott, S. Rodig,F.S. HodiAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): X. Wu, A. Giobbie-Hurder, X. Liao, D. Lawrence,S. Rodig, F.S. HodiWriting, review, and/or revision of themanuscript: X.Wu, A. Giobbie-Hurder,X. Liao, D. Lawrence, D. McDermott, S. Rodig, F.S. HodiAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): D. Lawrence, F.S. HodiStudy supervision: D. Lawrence, F.S. HodiOther (provided patient sera samples): J. Zhou

Grant SupportThis study was supported by NIH CA143832 (F.S. Hodi), the Melanoma

Research Alliance (F.S. Hodi), the Sharon Crowley Martin Memorial Fund forMelanoma Research (F.S. Hodi), and theMalcolm and EmilyMacNaught FundforMelanomaResearch (F.S.Hodi) atDana-Farber Cancer Institute,Genentech/Roche, and Bristol-Myers Squibb. This was also supported by a Stand Up ToCancer–Cancer Research Institute Cancer Immunology Dream Team Transla-tional ResearchGrant (SU2C-AACR-DT1012). StandUp ToCancer is a programof the Entertainment Industry Foundation administered by the AmericanAssociation for Cancer Research.

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.

Received April 19, 2016; revised July 29, 2016; accepted July 29, 2016;published OnlineFirst August 22, 2016.

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

2011;144:646–74.2. Cipponi A, Wieers G, van Baren N, Coulie PG. Tumor-infiltrating lympho-

cytes: Apparently good for melanoma patients. But why? Cancer ImmunolImmunother 2011;60:1153–60.

3. Ibrahim EM, Al-Foheidi ME, Al-Mansour MM, Kazkaz GA. The prognosticvalue of tumor-infiltrating lymphocytes in triple-negative breast cancer:A meta-analysis. Breast Cancer Res Treat 2014;148:467–76.

4. Galon J, Costes A, Sanchez-Cabo F, Kirilovsky A, Mlecnik B, Lagorce-PagesC, et al. Type, density, and location of immune cells within humancolorectal tumors predict clinical outcome. Science 2006;313:1960–4.

5. Erdag G, Schaefer JT, Smolkin ME, Deacon DH, Shea SM, Dengel LT, et al.Immunotype and immunohistologic characteristics of tumor-infiltratingimmune cells are associated with clinical outcome in metastatic melano-ma. Cancer Res 2012;72:1070–80.

6. Santoiemma PP, Powell DJ Jr.Tumor infiltrating lymphocytes in ovariancancer. Cancer Biol Ther 2015;16:807–20.

7. GoodenMJ, de BockGH, LeffersN,Daemen T,NijmanHW. The prognosticinfluence of tumour-infiltrating lymphocytes in cancer: A systematic reviewwith meta-analysis. Br J Cancer 2011;105:93–103.

8. Peske JD,Woods AB, Engelhard VH. Control of CD8 T-Cell infiltration intotumors by vasculature and microenvironment. Adv Cancer Res 2015;128:263–307.

9. Slaney CY, Kershaw MH, Darcy PK. Trafficking of T cells into tumors.Cancer Res 2014;74:7168–74.

10. Griffioen AW, Damen CA, Martinotti S, Blijham GH, Groenewegen G.Endothelial intercellular adhesion molecule-1 expression is suppressed inhuman malignancies: The role of angiogenic factors. Cancer Res 1996;56:1111–17.

11. Piali L, Fichtel A, Terpe HJ, Imhof BA, Gisler RH. Endothelial vascular celladhesion molecule 1 expression is suppressed by melanoma and carcino-ma. J Exp Med 1995;181:811–6.

12. MadhavanM, Srinivas P, Abraham E, Ahmed I, VijayalekshmiNR, BalaramP. Down regulation of endothelial adhesion molecules in node positivebreast cancer: Possible failure of host defence mechanism. Pathol OncolRes 2002;8:125–8.

13. Dirkx AE, Oude Egbrink MG, Kuijpers MJ, van der Niet ST, Heijnen VV,Bouma-ter Steege JC, et al. Tumor angiogenesismodulates leukocyte-vesselwall interactions in vivo by reducing endothelial adhesion moleculeexpression. Cancer Res 2003;63:2322–9.

14. KunzM, Toksoy A, Goebeler M, Engelhardt E, Brocker E, Gillitzer R. Strongexpression of the lymphoattractant C-X-C chemokine Mig is associatedwith heavy infiltration of T cells in human malignant melanoma. J Pathol1999;189:552–8.

15. Hodi FS,O'Day SJ,McDermott DF,Weber RW, Sosman JA,Haanen JB, et al.Improved survival with ipilimumab in patientswithmetastaticmelanoma.N Engl J Med 2010;363:711–23.

16. Robert C, Thomas L, Bondarenko I, O'Day S, Weber J, Garbe C, et al.Ipilimumab plus dacarbazine for previously untreated metastatic mela-noma. N Engl J Med 2011;364:2517–26.

17. Hodi FS, Lawrence D, Lezcano C, Wu X, Zhou J, Sasada T, et al. Bevaci-zumab plus ipilimumab in patients with metastatic melanoma. CancerImmunol Res 2014;2:632–42.

18. Wu X, Marmarelis ME, Hodi FS. Activity of the heat shock protein 90inhibitor ganetespib in melanoma. PLoS One 2013;8:e56134.

19. Duda DG, Cohen KS, Scadden DT, Jain RK. A protocol for phenotypicdetection and enumeration of circulating endothelial cells andcirculating progenitor cells in human blood. Nat Protoc 2007;2:805–10.

20. Groom JR, Luster AD. CXCR3 ligands: Redundant, collaborative andantagonistic functions. Immunol Cell Biol 2011;89:207–15.

21. Harlin H, Meng Y, Peterson AC, Zha Y, Tretiakova M, Slingluff C, et al.Chemokine expression in melanoma metastases associated with CD8þ T-cell recruitment. Cancer Res 2009;69:3077–85.






22. Ley K, Laudanna C, Cybulsky MI, Nourshargh S. Getting to the site ofinflammation: The leukocyte adhesion cascade updated.Nat Rev Immunol2007;7:678–89.

23. Balkwill F.Tumour necrosis factor and cancer. Nat Rev Cancer 2009;9:361–71.

24. Locksley RM, Killeen N, Lenardo MJ. The TNF and TNF receptor super-families: Integrating mammalian biology. Cell 2001;104:487–501.

25. Sethi G, Sung B, Aggarwal BB. TNF: A master switch for inflammation tocancer. Front Biosci 2008;13:5094–107.

26. Mantovani A, Sica A. Macrophages, innate immunity and cancer: Balance,tolerance, and diversity. Curr Opin Immunol 2010;22:231–7.

27. Voronov E, Carmi Y, Apte RN. The role IL-1 in tumor-mediated angio-genesis. Front Physiol 2014;5:114.

28. Thornton P, McColl BW, Greenhalgh A, Denes A, Allan SM, Rothwell NJ.Platelet interleukin-1alpha drives cerebrovascular inflammation. Blood2010;115:3632–9.

29. True AL, Rahman A,Malik AB. Activation of NF-kappaB induced byH(2)O(2) and TNF-alpha and its effects on ICAM1 expression in endothelial cells.Am J Physiol Lung Cell Mol Physiol 2000;279:L302–11.

30. Ledebur HC, Parks TP. Transcriptional regulation of the intercellularadhesion molecule-1 gene by inflammatory cytokines in human endothe-lial cells. Essential roles of a variant NF-kappa B site and p65 homodimers.J Biol Chem 1995;270:933–43.

31. HuangH, Langenkamp E,GeorganakiM, Loskog A, Fuchs PF, Dieterich LC,et al. VEGF suppresses T-lymphocyte infiltration in the tumor microenvi-ronment through inhibition ofNF-kappaB-induced endothelial activation.FASEB J 2015;29:227–38.

32. Griffioen AW, Damen CA, Blijham GH, Groenewegen G. Tumor angio-genesis is accompanied by a decreased inflammatory response of tumor-associated endothelium. Blood 1996;88:667–73.

33. Dirkx AE, oude EgbrinkMG,Castermans K, van der Schaft DW, Thijssen VL,Dings RP, et al. Anti-angiogenesis therapy can overcome endothelial cellanergy and promote leukocyte-endothelium interactions and infiltrationin tumors. FASEB J 2006;20:621–30.

34. Shrimali RK, Yu Z, Theoret MR, Chinnasamy D, Restifo NP, Rosenberg SA.Antiangiogenic agents can increase lymphocyte infiltration into tumor andenhance the effectiveness of adoptive immunotherapy of cancer. CancerRes 2010;70:6171–80.

35. Yuan J, Zhou J, Dong Z, Tandon S, Kuk D, Panageas KS, et al. Pretreatmentserum VEGF is associated with clinical response and overall survival inadvanced melanoma patients treated with ipilimumab. Cancer ImmunolRes 2014;2:127–32.

36. Hugo W, Zaretsky JM, Sun L, Song C, Moreno BH, Hu-Lieskovan S, et al.Genomic and transcriptomic features of response to anti-PD-1 therapy inmetastatic melanoma. Cell 2016;165:35–44.

37. Singh UP, Singh S, Iqbal N, Weaver CT, McGhee JR, Lillard JW Jr.IFN-gamma-inducible chemokines enhance adaptive immunity and colitis.J Interferon Cytokine Res 2003;23:591–600.

38. Neville LF, Mathiak G, Bagasra O. The immunobiology of interferon-gamma inducible protein 10 kD (CXCL10): A novel, pleiotropic memberof the C-X-C chemokine superfamily. Cytokine Growth Factor Rev 1997;8:207–19.

39. Lazzeri E, Romagnani P. CXCR3-binding chemokines: Novel multifunc-tional therapeutic targets. Curr Drug Targets Immune Endocr MetabolDisord 2005;5:109–18.

40. Monteagudo C, Martin JM, Jorda E, Llombart-Bosch A. CXCR3 chemokinereceptor immunoreactivity in primary cutaneous malignant melanoma:Correlation with clinicopathological prognostic factors. J Clin Pathol2007;60:596–9.

41. Jiang Z, XuY,Cai S. CXCL10 expression andprognostic significance in stageII and III colorectal cancer. Mol Biol Rep 2010;37:3029–36.

42. Mlecnik B, Tosolini M, Charoentong P, Kirilovsky A, Bindea G, Berger A,et al. Biomolecular network reconstruction identifies T-cell homing factorsassociated with survival in colorectal cancer. Gastroenterology 2010;138:1429–40.

43. Ji RR, Chasalow SD, Wang L, Hamid O, Schmidt H, Cogswell J, et al. Animmune-active tumor microenvironment favors clinical response to ipi-limumab. Cancer Immunol Immunother 2012;61:1019–31.

44. Mullins IM, Slingluff CL, Lee JK, Garbee CF, Shu J, Anderson SG, et al. CXCchemokine receptor 3 expression by activated CD8þ T cells is associatedwith survival in melanoma patients with stage III disease. Cancer Res2004;64:7697–701.

45. Toiyama Y, Fujikawa H, Kawamura M, Matsushita K, Saigusa S, TanakaK, et al. Evaluation of CXCL10 as a novel serum marker for predictingliver metastasis and prognosis in colorectal cancer. Int J Oncol 2012;40:560–6.

46. Khalili JS, Liu S, Rodriguez-Cruz TG,WhittingtonM,Wardell S, Liu C, et al.Oncogenic BRAF(V600E) promotes stromal cell-mediated immunosup-pression via induction of interleukin-1 in melanoma. Clin Cancer Res2012;18:5329–40.

47. Spranger S, Spaapen RM, Zha Y, Williams J, Meng Y, Ha TT, et al. Up-regulation of PD-L1, IDO, and T(regs) in the melanoma tumor microen-vironment is driven by CD8(þ) T cells. Sci Transl Med 2013;5:200ra116.

48. Fuertes MB, Kacha AK, Kline J, Woo SR, Kranz DM, Murphy KM, et al. Hosttype I IFN signals are required for antitumorCD8þT cell responses throughCD8{alpha}þ dendritic cells. J Exp Med 2011;208:2005–16.

49. Rafique I, Kirkwood JM, Tarhini AA. Immune checkpoint blockade andinterferon-alpha in melanoma. Semin Oncol 2015;42:436–47.

50. Kudo-Saito C.Cancer-associated mesenchymal stem cells aggravate tumorprogression. Front Cell Dev Biol 2015;3:23.


Wu et al.




2016;4:858-868. Published OnlineFirst August 22, 2016.Cancer Immunol Res Xinqi Wu, Anita Giobbie-Hurder, Xiaoyun Liao, et al. Infiltration and Humoral Recognition in MelanomaCellular Factors Associated with Enhancing Lymphocyte VEGF Neutralization Plus CTLA-4 Blockade Alters Soluble and

Updated version

10.1158/2326-6066.CIR-16-0084doi:

Access the most recent version of this article at:

Material

Supplementary

http://cancerimmunolres.aacrjournals.org/content/suppl/2016/08/19/2326-6066.CIR-16-0084.DC1

Access the most recent supplemental material at:

Cited articles

http://cancerimmunolres.aacrjournals.org/content/4/10/858.full#ref-list-1

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

Citing articles

http://cancerimmunolres.aacrjournals.org/content/4/10/858.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

Permissions

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

.http://cancerimmunolres.aacrjournals.org/content/4/10/858To request permission to re-use all or part of this article, use this link



http://cancerimmunolres.aacrjournals.org/lookup/doi/10.1158/2326-6066.CIR-16-0084

http://cancerimmunolres.aacrjournals.org/content/suppl/2016/08/19/2326-6066.CIR-16-0084.DC1

http://cancerimmunolres.aacrjournals.org/content/4/10/858.full#ref-list-1

http://cancerimmunolres.aacrjournals.org/content/4/10/858.full#related-urls

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

mailto:[email protected]

http://cancerimmunolres.aacrjournals.org/content/4/10/858


Documents

VEGF Neutralization Plus CTLA-4 BlockadeAlters Soluble and ... · Cancer Immunol Res; 4(10); 858–68. 2016 AACR. Introduction Escape from immune surveillance is a hallmark of metastatic