targeting tumor vasculature with tnf leads effector t ...control antibodies (syriam hamster igg; bio...

Cancer Therapy: Preclinical

Targeting Tumor Vasculature with TNFLeads Effector T Cells to the Tumor andEnhances Therapeutic Efficacy of ImmuneCheckpoint Blockers in Combination withAdoptive Cell TherapyAngela Rita Elia1,2,3, Matteo Grioni1, Veronica Basso2, Flavio Curnis3,Massimo Freschi4, Angelo Corti3,5, Anna Mondino2, and Matteo Bellone1

Abstract

Purpose: Irregular bloodflowand endothelial cell anergy,whichcharacterize many solid tumors, hinder tumor infiltration bycytotoxic T lymphocytes (CTL). This confers resistance to cancerimmunotherapy with monoclonal antibodies directed againstregulatory pathways in T lymphocytes (i.e., immune checkpointblockade, ICB). We investigated whether NGR-TNF, a TNF deriv-ative capable of targeting the tumor vasculature, and improvingintratumor infiltration by activated CTLs, could sensitize tumors toICB with antibodies specific for the PD-1 and CTLA-4 receptors.

Experimental Design: Transgenic adenocarcinoma of themouse prostate (TRAMP) mice with autochthonous prostatecancer and C57BL/6 mice with orthotopic B16 melanoma weretreated with NGR-TNF, adoptive T-cell therapy (ACT), and ICB,andmonitored for immune surveillance and disease progression.

Results: The combination of ACT, NGR-TNF, and ICB wasthe most effective in delaying disease progression, andin improving overall survival of mice bearing ICB-resistantprostate cancer or melanoma. Mechanistically, the therapeuticeffects were associated with potent tumor infiltration, espe-cially by endogenous but also by adoptively transferred PD-1þ, granzyme Bþ, and interferon-gþ CTLs. The therapeuticeffects were also associated with favorable T-effector/regula-tory T cell ratios.

Conclusions: Targeting the tumor vasculature withlow-dose TNF in association with ACT may represent a novelstrategy for enhancing T-cell infiltration in tumors and over-coming resistance to immune checkpoint blockers. Clin CancerRes; 24(9); 2171–81. �2018 AACR.

IntroductionTumor-specific cytotoxic T lymphocytes (CTL) epitomize the

ultimate effector mechanism of cancer immunotherapy: if fullyactivated and deployed at the cancer forefront, CTLs can selec-tively terminate cancer cells (1). Indeed, a highCTL infiltrate oftenassociateswith goodprognosis and response to immunotherapies(2). One exception to the rule is prostate adenocarcinoma, whichis generally considered anon–T-cell-infiltrated tumor (3, 4) and in

which a relatively highCD8þT-cell infiltrate,mostly located in thetumor stroma (5), correlates with clinical progression (6). How-ever, even a brisk CTL infiltrate, as that frequently found inmelanoma (7), can be outsmarted by cancer cells generating animmunosuppressive microenvironment (1).

A strategy that cancer cells adopt tododge the immune responseis the expression of program death-ligand 1 (PD-L1) on their cellsurface, thus interacting with the immune checkpoint inhibitorPD-1 on activated T cells, and interfering with T-cell receptor(TCR) and CD28 signaling (8, 9). An additional immune check-point involved in blunting the tumor-specific immune response isthe cytotoxic T lymphocyte antigen-4 (CTLA-4), which is inducedby T-cell activation, and, by competing with CD28 for CD80/CD86 binding on antigen-presenting cells, downregulates T cells(10, 11). Also, Foxp3þ regulatory T cells (Treg), which infiltrateboth prostate cancer (12) and melanoma (13), use CTLA-4 todampen T-cell activation (14).

Based on these notions, Allison and colleagues pioneered theuse of immune checkpoint blockade (i.e., monoclonal antibodiesdirected against regulatory pathways in T lymphocytes, ICB) incancer (4). Clinical results with anti–CTLA-4 and anti–PD-1/PD-L1 antibodies have been above expectations, and the FDAhas approved several of these antibodies for the treatment ofvarious cancer histotypes. However, ICB has failed to arrest tumorprogression in many patients affected by these tumors, and vari-ous histological types, among which prostate cancer (15), arepoorly sensitive to ICB. Indeed, treatmentwith either anti–CTLA-4

1Cellular Immunology Unit, Division of Immunology, Transplantation and Infec-tious Diseases, IRCCS San Raffaele Scientific Institute, Milan, Italy. 2LymphocyteActivation Unit, Division of Immunology, Transplantation and Infectious Dis-eases, IRCCS San Raffaele Scientific Institute, Milan, Italy. 3Tumor Biology andVascular Targeting Unit, Division of Experimental Oncology, IRCCS San RaffaeleScientific Institute, Milan, Italy. 4Unit of Pathology, Division of ExperimentalOncology, IRCCS San Raffaele Scientific Institute, Milan, Italy. 5Vita-Salute SanRaffaele University, Milan, Italy.

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

A. Corti, A. Mondino, and M. Bellone contributed equally to this article.

Corresponding Authors: Matteo Bellone, San Raffaele Scientific Institute, ViaOlgettina, 58, Milan 20132, Italy. Phone: 390-2264-34789; Fax: 390-2264-34786;E-mail: [email protected]; Anna Mondino, [email protected]; andAngelo Corti, [email protected]

doi: 10.1158/1078-0432.CCR-17-2210

�2018 American Association for Cancer Research.

ClinicalCancerResearch

www.aacrjournals.org 2171

on April 9, 2021. © 2018 American Association for Cancer Research. clincancerres.aacrjournals.org Downloaded from

Published OnlineFirst February 28, 2018; DOI: 10.1158/1078-0432.CCR-17-2210

http://crossmark.crossref.org/dialog/?doi=10.1158/1078-0432.CCR-17-2210&domain=pdf&date_stamp=2018-4-6

http://clincancerres.aacrjournals.org/

or anti–PD-1/PD-L1 monotherapy has shown limited clinicalbenefits in prostate cancer patients (16, 17). This might be dueto the scarce spontaneous T-cell infiltrate that characterizes mostprostate cancer lesions (6).

In general, the presence of hypoxia in solid tumors can lead tothe formation of new vessels that are more tortuous and leakythan the vessels of normal tissues, thereby promoting irregularblood flow and increasing the tumor interstitial pressure (18).Additionally, leucocyte adhesion molecules are poorly expressedin endothelial cells lining neoangiogenic vessels (a phenomenondefined as "endothelial anergy"; ref. 19). These phenomena canlimit the interaction between endothelial cells and leukocytes,and, consequently, their extravasation. Therefore, crossing theabnormal endothelial barrier and interstitium in solid tumorsremains a major hurdle that either endogenous or adoptivelytransferred tumor-specific CTLs must overcome to get in directcontactwith their targets (20). Thismay also explain the resistanceof some solid tumors to ICB (15, 21).

We have previously demonstrated that targeted delivery ofminute amounts of tumor necrosis factor-a (TNF) to the tumorendotheliumwithCys–Asn–Gly–Arg–Cys–Gly–TNF (NGR-TNF),a fusionprotein capable of targeting the tumor vasculature (22), issufficient to activate endothelial cells and to improve tumorinfiltration by CTLs (23). We have also shown that this approachcan ameliorate the adoptive cell therapy (ACT) with TCR-redir-ected T cells (24).

Based on these notions, we tested the hypothesis that the ther-apeutic efficacy of ICB with antibodies against PD-1 or CTLA-4could be ameliorated by a combination with NGR-TNF and ACT,thus favoring intratumoral T-cell infiltration and enhancing theireffector functions. We tested this triple combination in the auto-chthonous transgenic adenocarcinoma of the mouse prostate(TRAMP) model (25) and in C57BL/6 mice bearing established,orthotopic B16 melanoma expressing the surrogate antigen oval-bumin (OVA; ref. 26). These tumors are poorly sensitive to eachsingle therapeutic agent (27, 28), therefore mimicking humanprostate cancer and ICB-resistant melanoma. We have chosen thecombination of anti–CTLA-4 and anti–PD-1 antibodies becauseanti–CTLA-4 monotherapy induces expression of PD-L1 andother inhibitory immune checkpoints in tumor cells, T cells, and

tumor-associated macrophages (29), as well as expansion ofTregs (30). Additionally, a significantly longer progression-freesurvival has been reported in metastatic melanoma patientstreated with anti–CTLA-4 and anti–PD-1 antibodies than withanti–CTLA-4 alone (31). Disease progression and animal survivalwere analyzed in relation to the distribution and function oftumor-specific T cells. We found that ACT, NGR-TNF, and ICBwere all required to obtain the most effective delay in tumorprogression, to improveoverallmice survival, andpromote tumorinfiltration by PD-1þ, granzyme Bþ, and IFNgþ CTLs.

Materials and MethodsAnimal, cell lines, and reagent

Wild-type C57BL/6J mice (Charles River Breeding Laborato-ries), heterozygous TRAMPmice on a C57BL/6 background (25),C57BL/6-Tg(TcraTcrb)1100Mjb/J mice backcrossed to B6.129S7-Rag1tm1Mom/J mice to generate OT1/Rag1�/� mice (23) werehoused in a pathogen-free animal facility and treated in accor-dance with the European Community guidelines. The EthicalCommittee of the Istituto Scientifico San Raffaele (IACUC 707)approved the in vivo experiments. B16-OVAare B16-F1melanomacells (American Type Culture Collection) expressing a truncatedand nonsecreted form of OVA (26). Cell lines were not authen-ticated. After thawing, cells were cultured for 3 to 4 passages inRPMI 1640 (Lonza, Euroclone) with penicillin–streptomycin(Lonza), 10% heat-inactivated FCS (Capricorn) and hygromycin(100 mg/mL; Roche, Milan, Italy). Before injection, mycoplasmainfection was excluded by PCR. The Tag-IV 404–411 (VVYDFLKC)and OVA257–264 (SIINFEKL) peptides were purchased fromResearch Genetics. NGR-TNF was produced and characterized asdescribed in ref. (22) andwas administrated i.p. (100pgdissolvedin physiologic solution containing 100 mg/mL of human serumalbumin). Anti–CTLA-4 (clone 9H10; Bio X Cell), anti–PD-1(clone RMP1-14; Bio X Cell) monoclonal antibodies, and isotypecontrol antibodies (Syriam Hamster IgG; Bio X Cell) were pur-chased from DBA Italia s.r.l. Isotype control rat IgG were fromSigma-Aldrich. All antibodies were administrated i.p.

Retroviral transduction of T cellspMX-SV40 vector was used to transfect Phoenix-E packaging

cells to generate retrovirus as described previously (32). The pCL-ECO (IMGENEX) packaging plasmid was used to maximizeproduction. Briefly, splenocytes were activated in vitro with con-canavalin A (ConA, 2 mg/mL) in the presence of IL7 (1 ng/mL;PeproTech, tebu-bio) for 48 hours. Cells were harvested, resus-pended in viral supernatant (4 � 106/mL), plated in retronectin-coated 24-well plates (50 mg/mL; Takara), subjected to spintransduction (90 minutes/2,000 rpm), and incubated overnightat 37�C. After transduction, cells were harvested and transductionefficiency was evaluated by FACS, by measuring the percentage ofVb9 (tumor redirected) within CD8þ T cells.

ACT, vaccination, and ICB treatment in the TRAMP modelMale mice received a sublethal dose of total body irradiation

(TBI; 6 Gy) and the following daywere i.v. injected into the lateraltail vein with 0.5 � 106 autologous TAG-IV–redirected CD8þ

T cells according to previous reports (32). Bone marrow–deriveddendritic cells (DC) were prepared and evaluated by FACSas described before (33). LPS-matured DCs were pulsed with2 mg/mL of Tag-IV404–411 peptide for 1 hour at 37�C, washed, andresuspended at 2.5 � 106 cells/mL in PBS (Euroclone). DCs

Translational Relevance

Strategies to sensitize tumors and/or overcome their resis-tance to monoclonal antibodies directed against regulatorypathways in T lymphocytes (i.e., immune checkpoint block-ade, ICB) are urgently needed. Elia et al. show that simulta-neous targeting of the tumor microenvironment with low-dose NGR-TNF, a TNF derivative capable of targeting thetumor vasculature and promoting intratumoral T-cell infiltra-tion, and ICB, supports extravasation, persistence, and effectorfunctions of adoptively transferred T cells. More importantly,the combined therapy has beneficial effects on endogenousimmune surveillance, through depletion of Tregs and expan-sion of a fully functional, polyclonal repertoire of CTLs. Thus,cold tumors and tumors resistant to ICB become sensitive tosuch treatment. As NGR-TNF, adoptive cell therapy and ICBare already in clinical testing, this combined strategy isexpected to be on a fast track for clinical translation.

Elia et al.

Clin Cancer Res; 24(9) May 1, 2018 Clinical Cancer Research2172




were injected intradermally (i.d.; 5 � 105 cells/mouse) in theright flank. Three doses of anti–CTLA-4 (100 mg) plus anti–PD-1(250 mg) antibodies or isotype antibodies (100 mg of SyriamHamster IgG and Rat IgG 250 mg) were administrated once every3 days, followed by anti–PD-1 or Rat IgG one month until theend of experiment. NGR-TNF (100 pg/mouse) was injected intra-peritoneally on day 7 after ACT.

Activation of OT1 cellsOVA-specific and in vitro–activated CD8þ T (OT1) cells from

RAG-1�/� OT1 female mice were obtained as previously de-scribed (23). Briefly, single-cell suspensions (1 � 106 cells/mL)of spleen and lymph nodes cells from OT1/Rag1�/� mice wereseeded into 6-well plates and stimulated with of 100 ng/mLOVA 257–264 in RPMI 1640 supplemented with penicillin/streptomycin, 10 mmol/L HEPES (Lonza), 10 mmol/L sodium-pyruvate (Lonza), 50 mmol/L 2-ME (Life Technologies),10% heat-inactivated FCS, and IL12 (3.5 ng/mL; R&D Sys-tems). On day 3 of culture, cells were harvested and seededinto new 6-well plates together with culture medium and50 U/mL of IL2 (R&D Systems). At day 5, cells were recovered,washed, and suspended in PBS and injected into the tail veinof mice.

Tumor implantation and treatment in the B16-OVA modelC57BL/6J females were challenged s.c. in the left flank with 2�

105 B16-OVA cells, and at day 7 with activated OT1 T cells wereinjected (6� 106/mouse) into the tail vein. The day after, 3 dosesof anti–CTLA-4 (100 mg) plus anti–PD-1 (250 mg) mAbs orisotype antibodies (100 mg of Syriam Hamster IgG and Rat IgG250 mg) were administrated once every 3 days, followed by anti–PD-1 or Rat IgG every 3 days until the end of the experiment.Tumor size was evaluated by measuring two perpendicular dia-meters and depth with a caliper. In survival experiments, animalswere killed when the tumor reached 10 mm of diameter orbecame ulcerated. Tumors were collected when they reachedapproximately 4 mm of diameter, then were disaggregated, dig-ested in collagenases A–B–D (Roche) with DNAse (100 mg/mL,Roche) for 1 hour at 37�C to obtain single-cell suspension.

Flow cytometrySingle-cell suspension from B16-OVA tumor, PDLNs from

TRAMP mice, redirected TAG-IV T cells and activated OT1 cellswere labeled with fluorochrome-conjugated mAbs (either fromBD Bioscience, Biolegend Europe, or eBioscience Inc.) after neu-tralization of unspecific binding with FcR blocker (BD Bios-ciences). Dead cells were labeled with 7AAD, and samples wereacquired on a BD LSR Fortessa. Data were analyzed using theFlowJo software (TreeStar Inc.). Tumor-infiltrating cells were alsoassessed for intracellular Granzyme B expression in the absence ofstimulation or for cytokine production after 4 hours at 37�C ofstimulation with Phorbol Myristate Acetate (PMA)/ionomycin.Brefeldin A (Sigma-Aldrich) was added to the samples during thelast 2 hours of culture. Cells were surface stained, fixed, andpermeabilized, and further stained for intracellular cytokines asdescribed previously (34). Samples were acquired on a BD LSRFortessa, and data were analyzed using the FlowJo software.

Histology and IHCUrogenital apparata were embedded in the Killik Frozen Sec-

tion Medium (OCT, Bio-Optica). Cryostat sections were stained

for hematoxylin and eosin (H&E). Prostate tissue sections (5-mmthick) were stained with anti-CD3mAb (Serotec) according to themanufacturer's instructions, were digitally scanned (ScanScope,Aperio), and then analyzed with Spectrum Plus software.

Statistical analysisStatistical analyses were performed with unpaired Student t

test. Survival curves were compared using the log-rank test; Forall tests, symbols mean �, P < 0.05; ��, P < 0.01; ��, P < 0.001;��, P < 0.0001; ns, not significant.

ResultsTargeting tumor vasculaturewithNGR-TNF leads effector T cellsto the tumor and enhances prostate cancer sensitivity to ICB incombination with ACT

TRAMP mice at 16 to 18 weeks of age have established, andpoorly T-cell–infiltrated prostate cancer (24, 25), are fully tolerantto the products of the SV40 early oncogenes (small and large Tantigens; Tag; ref. 35), and are resistant to anti–CTLA-4 mono-therapy (28). To investigate if resistance to ICB could be due tolimited infiltration of tumor-reactive effector T cells within thetumor, TRAMPmicewere treatedwith a combination of ACT, ICB,and NGR-TNF. A schematic representation of the treatment isdepicted in Fig. 1A. Nonmyeloablative, total body irradiation(TBI) was given to favor antigen-presenting cell activation (36)and to reduce competition for survival or proliferation signals(37). ACT consisted of autologous TRAMP-derived T cells engi-neered by retroviral-mediated transduction to express the Tag-IV–restricted TCR(24, 32). Transduced cell productswere representedby CD3þ T cells, of which 50% were CD8þ (Fig. 1B, first panel).Approximately 40%ofCD8þT cells expressed the transduced Tag-IV–specific Vb9 TCR chain (Fig. 1B, second panel; frequency ofVb9þ cells in untransduced cells was about 3%, data not shown).Transduced Vb9þCD8þ T revealed a central/effector memoryphenotype (CD44hiCD62Lhigh/low, Fig. 1B, third panel), expressedlow levels of PD-1 (Fig. 1B, fourth panel), and produced IFNg andTNFa upon stimulation with the cognate Tag-IV peptide (Fig. 1B,fifth panel). Because vaccination is instrumental to initiate thegraft versus tumor response that results in acute tumor debulking,aswell as in the ameliorationof long-term survival rates in TRAMPmice (34), 1 day after ACT, mice received a single intradermalinjection of dendritic cells pulsed with the immunodominat Tag-IV peptide 404–411 (DC Vax), a vaccination schedule optimizedto generate a long-lasting tumor-specific memory response (38).NGR-TNF was given at the peak of the immune response inducedby the vaccination, i.e., 1week after ACT andDCVax, as this favorsinfiltration of fully activated, donor-derived CD3þ T cells withinthe prostate tumor (23). Finally, ICB (i.e., anti–CTLA-4 and anti–PD-1 antibodies) was administered starting the day after DC Vaxto prevent T-cell exhaustion and immunosuppressive mechan-isms. For the sake of simplicity, TBI, ACT, and DC Vax are referredto as ACT thereafter. Groups of untreated TRAMPmice, or treatedwith ICB, or isotype antibodies (IgG), or ACTþ ICB, ICBþNGR-TNF, or ACTþ IgGþNGR-TNF served as controls. The use of ICBalone did not support better animal survival when comparedwith untreated mice or mice treated with isotype control anti-bodies (Supplementary Fig. S1A), thus suggesting that eventhe combinationof anti–CTLA-4 and anti–PD-1 cannot overcomethe tumor immunosuppressive environment in TRAMP mice(39, 40). Likewise, the combination of NGR-TNF and ICB failed

Overcoming Resistance to Immune ICB

www.aacrjournals.org Clin Cancer Res; 24(9) May 1, 2018 2173




Figure 1.

Targeting tumor vasculature with NGR-TNF leads effector T cells to the tumor and enhances prostate cancer sensitivity to ICB in combination with ACT. A, Schemeand schedule of treatments. Sixteen–18-week-old TRAMP mice received TBI, the following day they were infused with autologous Tag-IV–redirected T cells. Aday later, mice were vaccinated with Tag-IV404-411 pulsed DCs (DC Vax). After 6 additional days, mice were injected with NGR-TNF. Anti–CTLA-4 and anti–PD-1antibodies (ICB) or isotype antibodies were administrated as indicated. B, Phenotype and cytokine production upon challenge with Tag-IV404-411 peptidewere investigated in Tag-IV–redirected T cells by flow cytometry after staining with the indicated antibodies. Each panel is representative of at least threeindependent experiments.C,Kaplan–Meier survival curve of untreatedmice (dotted line; n¼ 14), ormice treatedwith ACTþ ICBþNGR-TNF (red line; n¼ 9), ACTþIgG þ NGR-TNF (blue line; n ¼ 11), or ACT þ ICB (black line; n ¼ 4). Log-rank test: Untreated versus ACTþICB, P < 0.05; untreated versus ACT þ IgG þNGR-TNF, P < 0.0001; untreated versus ACT þ ICB þ NGR-TNF, P < 0.0001; ACT þ ICB versus ACT þ IgG þ NGR-TNF, P > 0.05; ACT þ ICB versus ACT þ ICB þNGR-TNF, P < 0.05; and ACT þ IgG þ NGR-TNF versus ACT þ ICB þ NGR-TNF, P < 0.01. D, Treated mice were killed at day 42, and prostate tissues wereanalyzed by IHC for CD3 expression. Panels depict representative IHC images from the prostate of TRAMP mice treated with ACTþ IgGþ NGR-TNF, ACTþ ICB, orACT þ ICB þ NGR-TNF. E, Quantification of CD3þ cells/mm2 of tissue [white circle, ACT þ IgG þ NGR-TNF (n ¼ 3); black triangles, ACT þ ICB (n ¼ 4); andblack square, ACT þ ICB þ NGR-TNF (n ¼ 5)]. F–I, Single-cell suspensions obtained from the PDLNs of TRAMP mice treated with ACT þ IgG þ NGR-TNF (n ¼ 5),ACT þ ICB (n ¼ 4), or ACT þ ICB þ NGR-TNF (n ¼ 7) were analyzed by flow cytometry. The panels show the frequency of CD8þ cells within CD3þ cells (F)and their absolute numbers� SEM (G), and the frequency of redirected T cells, identified as Vb9þwithin CD8þCD3þ cells (H) and their absolute numbers (I). Studentt test; �, P < 0.05; �� , P < 0.01; �� , P < 0.001; �� , P < 0.0001; ns, not significant. Data were aggregated from two independent experiments.

Elia et al.





to evoke tumor debulking at day 42 (n ¼ 3; SupplementaryFig. S1B), which coincides with complete tumor eradication inmice exposed tomore aggressive treatments (34). The addition ofACT to ICB prolonged TRAMP survival (Fig. 1C), but not longenough to be statistically significant when comparedwith TRAMPmice treated with ICB alone (Supplementary Fig. S1A). Therefore,an exogenous supply of tumor-specific CTLs is not enough toovercome resistance to ICB in this experimental setting. Thecombination of ACT and NGR-TNF significantly prolonged ani-mal overall survival (Fig. 1C). Nevertheless, all TRAMP miceeventually succumbed to the disease by week 82 of age (Fig.1C). In sharp contrast, more than 50% of the TRAMP micereceiving the triple combined treatment were still alive at week82 (Fig. 1C). The survival of the ACTþ ICBþNGR-TNF groupwasdramatically increased when compared with ACT þ IgG þ NGR-TNF (p ¼ 0.0016), or ACT þ ICB (p ¼ 0.0149). No loss of bodyweight was observed inmice treated with ACTþ ICBþNGR-TNF,when compared with the other groups (data not shown), suggest-ing that the triple combination could induce greater therapeuticeffects in the absence of toxicity.

To look for immune correlates of therapeutic effects, at day 42the prostates and prostate-draining lymph nodes (PDLN) wererecovered fromTRAMPmice treated with ACTþ IgGþNGR-TNF,ACT þ ICB, or ACT þ ICB þ NGR-TNF and analyzed by IHC andflow cytometry. Indeed, while anti–PD-1 antibodies primarily actat the tumor site, anti–CTLA-4 antibodies are believed to mainlyimpact on T-cell activation within secondary lymphoid organsand in tertiary lymphoid structures (4). Day 42 was chosen as atthis time disease status well correlates with tumor immuneinfiltration (34). IHC of prostate tissue sections showed in micetreatedwith ACTþ ICBþNGR-TNF a potent infiltration byCD3þ

cells, which outnumbered that induced either by ACT þ ICB orACT þ IgG þ NGR-TNF (Fig. 1D and E).

PDLNs were next analyzed. The frequency of both CD3þCD8þ

T cells (Fig. 1F) and CD3þCD8þVb9þ (i.e., TCR-redirected Tcells; Fig. 1H) was increased in PDLNs of mice treated with theACT þ ICB þ NGR-TNF compared with the ACT þ IgG þ NGR-TNF group. Although an increase was evident in absolute cellnumbers, variability in lymph node dimension among differentexperiments did not allow reaching statistically significant differ-ences (Fig. 1G–I). Altogether, these findings suggest that ICBpromotes T-cell responses both in PDLNs and in the tumor, andthat the beneficial effects of the triple combined therapy are likelyrelated to (i) ACT-mediated refreshment of the pool of T cellscapable of recognizing the tumor; (ii)NGR-TNF–induced increasein tumor immune infiltration; and (iii) ICB-promoted enhance-ment of effector cell activity/persistence in tumors. Thus, the triplecombined therapy can convert a cold prostate tumor into animmune infiltrated neoplasm, thereby sensitizing it to ICB.

The combined therapy synergizes to prolong overall survival ofmelanoma-bearing mice

Approximately 50% of advanced melanoma patients are resis-tant or acquire resistance to the combination of anti–CTLA-4 andanti–PD-1 antibodies (31). To validate the therapeutic potentialof the triple combined therapy in a different model, and also tofurther dissect the mechanisms underlying its efficacy, we usedmice bearing established B16 melanoma. This tumor is poorlysensitive to anti–CTLA-4 and anti–PD-1, evenwhen administeredin combination (27), thus modeling ICB resistance in humanmelanoma. To enumerate tumor-specific T-cell responses, we

took advantage of B16 cells expressing a nonsecreted form ofovalbumin (B16-OVA: ref. 26). AsACT,we exploitedOVA-specificTCR transgenic OT1/Rag-1�/� T cells induced to differentiate intoeffector cells by antigen/IL2-driven cultures (23). We previouslyfound that activatedOT1T cells preferentially infiltratedB16-OVAtumors when NGR-TNF was administered 2 hours before T-cellinfusion (23). Because of this, and of the fact that B16-derivedtumors are fast growing,we adopted this administration schedule.C57BL/6 females bearing a 7-day-old subcutaneous B16-OVAmelanoma were treated as depicted in Fig. 2A and monitored fordisease progression. Treatment with ICB or ACT slightly delayedtumor growth (Fig. 2B) and significantly but modestly prolongedanimal survival when compared with vehicle-treated mice (Fig.2D), thus confirming the poor sensitivity of B16 tumors to ICB(27) or ACT (23) single therapies. When ACT and ICB werecombined, tumor growth (Fig. 2B) and progression were signif-icantly delayed (Fig. 2C), andmice survived longer (Fig. 2D) thancontrols (ICBorACTalone). Thus, ACT improves responses to ICBin mice bearing an established melanoma.

We then evaluated the addition of NGR-TNF, which wepreviously reported to synergize with ACT (23). As expected,administration of NGR-TNF 2 hours before ACT delayed tumorgrowth similarly to the ACT þ ICB (Fig. 2E–G). However, whenNGR-TNF was administered in conjunction with ACT and ICB,a stronger delay in tumor progression was obtained (Fig. 2E),leading to tumors with smaller volumes (Fig. 2F), and alimited, yet significant improvement of mice overall survival(Fig. 2G), with no evidence of body weight loss (data notshown). Although tumors eventually developed in these mice,they showed a substantial delay in progression (Fig. 2H). Thus,as observed with prostate tumor (Fig. 1), the combination ofNGR-TNF and ACT converted melanomas poorly sensitive toICB into ICB-sensitive tumors.

NGR-TNF increases sensitivity to ICB by supporting intratumorCTL infiltration and effector functions

To get mechanistic insights into the therapeutic efficacy ofthe combined therapy, flow cytometry analyses were conductedon cells infiltrating well-established and not yet necrotic mel-anomas. Thus, mice were sacrificed when the tumor reachedapproximately 4 mm in mean diameter (i.e., between days 15and 19) to avoid possible biases due to the impact of differenttumor dimension on the immune infiltrate. Indeed, tumorscollected from the different groups were similar in weight (Fig.3A). ACT þ ICB induced intratumor infiltration by CD8 andCD4 T cells (Fig. 3B and C, respectively). Although NGR-TNFdid not seem to improve tumor infiltration by CD8 T cells whencombined to controls (i.e., ACT only), it significantly did sowhen combined with ACT þ ICB (Fig. 3B and C, respectively).As the effects of NGR-TNF on neoangiogenic vessels are rapidand transient (23), this could be explained by the capability ofICB to induce survival and proliferation of extravasated T cellsin tumor tissues (Fig. 3B).

As an additional consequence of ACT þ ICB, the frequency ofCD25þFoxP3þ Tregs dramatically dropped (Fig. 3D), thus signif-icantly increasing theCD8þ Teff/Treg ratio in the tumors (Fig. 3E).This was also observed in combination with NGR-TNF (Fig. 3E).These immunological parameters have been already associatedwith response to ICB (41).

To better understand the effects of ICB and NGR-TNF ontumor-directed T-cell responses, we stratified tumor-infiltrating






Figure 2.

The triple combined therapy prolongs overall survival of melanoma-bearing mice. A, Scheme and schedule of treatments. Mice bearing a 7-day-old B16-OVAmelanoma were randomly assigned to either one of following treatments: vehicle (n ¼ 22–28), anti–CTLA-4 (100 mg/mouse), and anti–PD-1 (250 mg/mouse)antibodies (ICB n¼ 14–16), OT1 cells (6� 106/mouse; ACT; n¼ 28), ACTþ ICB (n¼ 27), NGR-TNFþ ACTþ ICB (n¼ 9), or NGR-TNF plus ACT and isotype specificantibodies (IgG; n¼ 14–15). Average tumor volume (B and E), area under the curve (AUC) at days 19 (C) and 24 (F), survival (%; D and G), and delay in tumor growth(H) for each experimental group. The delay in tumor growth was calculated by determining the mean time necessary to obtain tumor volumes of 0.5 cm3 for eachexperimental group, comparedwith animals treatedwith vehicle. Each symbol inC,F, andH represents an individualmouse. Bars showmean� SEM. Student t test; � ,P < 0.05; �� , P < 0.01; �� , P < 0.001; ��, P < 0.0001; ns, not significant.

Elia et al.





CD3þCD8þ cells according to TCR expression (being ACT-derived OT1 cells Vb5.1þVa2þ, and endogenous T cells mostlyVb5.1�Va2�) and characterized them for effector phenotype(representative dot plots are depicted in Supplementary Figs. S2and S3). Frequencies and total numbers are reported inSupplementary Fig. S4 and in Fig. 4, respectively. ACT-derivedT cells (OT1) were equally represented in all groups of mice(Fig. 4A) and were composed of CD44þCD62Llowcells (Fig. 4Band Supplementary Fig. S4B), of which approximately 40% to50% expressed PD-1 (Fig. 4C and Supplementary Fig S4C),Granzyme B (Fig. 4D and Supplementary Fig. S4D), and pro-duced IFNg (Fig. 4E and Supplementary Fig. S4E). The frequen-cy of IFNg-producing OT1 cells slightly and yet significantlyincreased in mice treated by ICB (Fig. 4E and SupplementaryFig. S4E), suggesting the antibodies promoted full effectorfunction within the tumor.

Unexpectedly, the triple combined treatment exerted morestriking effects on endogenous T cells. Indeed, treatment withACT þ ICB increased >5-fold the number of infiltrating endog-enous CD8þ T cells when compared with ACT (Fig 4F). Under thepressure of ICB, tumor-infiltrating endogenous CD8 T cells wereenriched in fully differentiated effectors, expressing high levels ofCD44 (Fig. 4G and Supplementary Fig. S4G) and PD-1 (Fig. 4Hand Supplementary Fig. S4H), low levels of CD62L (Fig. 4G andSupplementary Fig. S4G) and showing superior cytolytic (Fig. 4Iand Supplementary Fig. S4I) and IFNg-producing properties (Fig.4J and Supplementary Fig. S4J). NGR-TNF further improvedtumor infiltration by endogenous T cells when compared withthat observed inmice treated with ACT or ACTþ ICB, resulting inoptimal intratumoral representation of fully differentiated effec-tors (Fig. 4).

Altogether, these data demonstrate that NGR-TNF improvesintratumor T-cell responses, and tumor infiltration especially byendogenous CD8þ T cells, whose full effector function is sup-ported by ICB treatment. Thus, the triple combined therapy isneeded to take full advantage of ICB-mediated therapeutic activityagainst poorly ICB-sensitive tumors.

DiscussionOur results indicate that in tumor bearing mice, the combina-

tion of NGR-TNF, ACT, and ICB enhances tumor infiltration,persistence, and effector functions of adoptively transferred Tcells. More importantly, it has a beneficial effect on the endog-enous immune surveillance, through depletion of Tregs andexpansion of a fully functional, polyclonal repertoire of CTLs. Asa consequence of the synergy between these treatments, coldtumors or tumors poorly sensitive to ICB responded to suchtreatment, and mice affected either by autochthonous prostate

Figure 3.NGR-TNF enhances sensitivity to ICB by supporting intratumoral CD8þ andCD4þ cell infiltration. Mice bearing a 7-day-old B16-OVA melanoma wererandomly assigned to the treatments described in Fig. 2. Mice were killed whenthe mean tumor diameter was approximately 4 mm. Single-cell suspensionsfrom each tumorwere analyzed by flow cytometry.A, tumorweight reported asmg of tissue; B, absolute number of CD8þ cells; C, absolute number of CD4þ

T cells; D, percentage of CD4þ T cells expressing CD25 and FoxP3; E, CD8þ

T effector to Treg ratio. Each symbol represents an individual mouse. Bars,mean� SEM. Data are aggregated from two independent experiments. Studentt test: � , P < 0.05; �� , P < 0.01; �� , P < 0.001; �� , P < 0.0001; ns, not significant.






Figure 4.

NGR-TNF enhances sensitivity to ICB by recruitingendogenous T cells into protective antitumor immuneresponses. Tumors from mice treated as described in thelegend to Fig. 3 were processed to single cells and analyzedby flow cytometry. Graphs report absolute numbers (mean�SEM) of OT1 (A–E), and endogenous CD8þ T-cellsubpopulations (F–J) analyzed aswhole population (A andF)or CD44þCD62L� cells (B and G), CD44þPD-1þ cells(C and H), Granzyme Bþ cells (D and I), and IFNgþ cells(E and J). Each symbol represents an individual mouse. Dataare aggregated from two independent experiments.Student t test: � , P < 0.05; �� , P < 0.01; �� , P < 0.001;�� , P < 0.0001; ns, not significant.

Elia et al.





cancer or by orthotopic melanoma experience a substantialincrease in overall survival.

ICB was likely active both at the tumor site and in secondarylymphoid organs, given that the triple therapy improved T cellsresponses at both sites in TRAMPmice. ACT, ICB, or NGR-TNF assingle agents did not impact the disease, and only the combina-tion of ACT þ ICB þ NGR-TNF outperformed ACT þ ICB, ICB þNGR-TNF, or ACT þ NGR-TNF combinations. Thus, the concur-rent action of tumor vessel targeting by NGR-TNF and ICB isneeded to allow sufficient tumor-specificCTLs in the tumor tissue,retain proper effector function, and debulk the established tumor.Although our experiments were not designed to investigate theinduction of long-lasting, tumor-specific immunity (38), morethan 50% of TRAMP mice treated with ACT þ ICB þ NGR-TNFsurvived to the end of the experiment. This suggests that thecombined treatment generated a protective memory response.Indeed, due to hormone-dependent oncogene (25), TRAMPmiceexperiencing treatment-related tumor debulking should contin-uously repopulate the prostate with Tag-expressing cells able toundergo transformation. Thus, our data support the possibilitythat the triple combined treatment generates in TRAMP mice aTag-specific memory response able to efficiently patrol againstarising neoplastic foci.

Our data also indicate that the tumor-specific T-cell responsesdeveloped in the context of the combined therapy protectedTRAMP mice from the development of hormone-independent,undifferentiated prostate tumors. Indeed, while these usuallyappear in approximately 10% to 15% of the untreated TRAMPmice in our colony byweeks 16 to 18 (24), they were not found inmice subjected to the triple combined therapy. Thus, in thecontext of the highly immunosuppressive prostate tumor, ACTþ ICB þ NGR-TNF therapy seems to evoke a potent curative andprotective antitumor immunity.

As stated above, treatment with NGR-TNF did not sensitizeTRAMP mice to ICB unless it was combined to ACT. Also in theMC38 colon cancer model, which is ICB sensitive, the additionof NGR-TNF did not increase the therapeutic index of ICB(Supplementary Fig. S5). We speculate that in TRAMP mice,tumor infiltration and tumor-directed T-cell responses are sopoor that ACT is required to rejuvenate the response and rendera cold tumor sensitive to ICB. In the case of the MC38 model,being tumors more immunogenic and likely infiltrated by Tcells, NGR-TNF alone does not add to the therapeutic potentialof ICB.

Althoughmelanoma is generally considered a T-cell–infiltratedor "hot" tumor (3, 4), a relevant percentage of metastatic mela-noma patients are resistant or acquire resistance to ICB (31). Wefound that the triple combined therapy not only allowed ACTcells to infiltrate the tumor, but also favored the recruitmentof endogenous CTLs in the tumor response, supporting theirfunctionality. Thus, the triple combined therapy might also beeffective in patients with less ICB-sensitive melanomas, and byrecruiting patients' natural immune responses, might promoteepitope spreading, and further improve protective immunesurveillance. It is worth noticing that the triple combinedtherapy dramatically ameliorated overall survival in TRAMPmice, when compared with B16 tumor–bearing mice. A possi-ble explanation is that TRAMP mice totally rely on ACT torejuvenate the immune response to the growing tumor, andthat effects of the combined treatment might be more pro-nounced in slow-progressing tumors.

Considering that original preclinical data measured in the B16melanoma underestimated the therapeutic benefits of ICB laterobserved in melanoma patients (4), we expect that the proposedtreatment will significantly improve response rates both in mel-anoma and prostate cancer patients, and that it could be extendedto other cold ICB-resistant malignancies.

Because both TRAMP-derived tumor cells (data not shown)and B16-OVA cells (23) express the NGR receptor CD13 ontheir surface, NGR-TNF interaction with tumor cells might havefavored immunogenic cell death and by that favor activation ofnewly recruited T cells. Although we cannot exclude this couldbe the case in vivo, in vitro NGR-TNF failed to induce tumor celldeath when used alone (not shown) or in combination withchemotherapy even when used at 10 ng/mL (23). Thus, it ismore likely that NGR-TNF mainly targeted tumor endothelialcells in our models. Still, we cannot exclude an immunogenictumor cell death induced by NGR-TNF–mediated vasculardamage.

Other approaches have been proposed to target neoangiogen-esis, including those aiming at blocking angiogenesis and/ornormalizing the tumor vasculature, some of which have beentested in association with ICB (42–45). At difference with othercompounds, NGR-TNF is an inflammatory vascular targetingagent that, when given at picogram doses, induces transientvascular activation (i.e., upregulation of adhesion molecules onendothelial cells) with no evidence of systemic toxicity (23).

Treatment with NGR-TNF also induces the transient releasein the tumor microenvironment of several cytokines and che-mokines, among which CCL-2 and CCL-7 (23), which attractactivated T cells (46), and participate in the generation ofmemory T cells (47). On this line, other investigators haveprepared a fusion compoundmade by the tumor necrosis factorsuperfamily member LIGHT, a protein involved in generationof tertiary lymphoid structures, and an antibody specific forepidermal growth factor receptor, a membrane proteinexpressed in tumors poorly infiltrated by CTLs (48). Treatmentwith this fusion product induced the production of CCL21 andfavored massive T-cell infiltration, transforming a cold tumorinto a hot one, and overcoming ICB resistance (48). Thus,vascular-targeting with cytokines, such as LIGHT or TNF,appears a valid approach for increasing the therapeutic indexof ICB. Notably, NGR-TNF has been tested in patients and inassociation with immunotherapy (49), proving good tolerabil-ity (50). Thus, NGR-TNF is worth investigating in humans inassociation with ICB. Remarkably, most of the biological effectsinduced by NGR-TNF in murine tumors have been observedwith equivalent doses in patients (50). Thus, as NGR-TNF, ACT,and ICB are already in clinical testing, the combined strategycould be on a fast track for clinical translation.

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

Authors' ContributionsConception and design: F. Curnis, A. Corti, A. Mondino, M. BelloneDevelopment of methodology: A.R. Elia, V. Basso, F. Curnis, A. MondinoAcquisition of data (provided animals, acquired and managed patients,provided facilities, etc.): A.R. Elia, M. Grioni, F. Curnis, A. MondinoAnalysis and interpretation of data (e.g., statistical analysis, biostatistics,computational analysis): A.R. Elia, F. Curnis, M. Freschi, A. Corti, A. Mondino,M. Bellone






Writing, review, and/or revision of the manuscript: A.R. Elia, F. Curnis,A. Corti, A. Mondino, M. BelloneAdministrative, technical, or material support (i.e., reporting or organizingdata, constructing databases): M. Grioni, V. BassoStudy supervision: A. Corti, A. Mondino, M. Bellone

AcknowledgmentsThe work was mostly supported by the Cancer Research Institute, New York

(CRI; Clinical and Laboratory Integration Program award to M. Bellone) andalso in part by the Associazione Italiana per la Ricerca sul Cancro, Milan, IT (IG

2015 Id.16807 toM. Bellone; 5� 1000-9965 and IG-19220 to A. Corti, and IG-15883 to A. Mondino).

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 July 29, 2017; revised January 16, 2018; accepted February 23, 2018;published first February 28, 2018.

References1. Palucka AK, Coussens LM. The basis of oncoimmunology. Cell

2016;164:1233–47.2. Becht E, Giraldo NA, Germain C, de Reynies A, Laurent-Puig P, Zucman-

Rossi J, et al. Immune contexture, immunoscore, and malignant cellmolecular subgroups for prognostic and theranostic classifications ofcancers. Adv Immunol 2016;130:95–190.

3. Gajewski TF, Schreiber H, Fu YX. Innate and adaptive immune cells in thetumor microenvironment. Nat Immunol 2013;14:1014–22.

4. Sharma P, Allison JP. The future of immune checkpoint therapy. Science2015;348:56–61.

5. Karja V, Aaltomaa S, Lipponen P, Isotalo T, Talja M, Mokka R. Tumour-infiltrating lymphocytes: a prognostic factor of PSA-free survival in patientswith local prostate carcinoma treated by radical prostatectomy. AnticancerRes 2005;25:4435–8.

6. Petitprez F, Fossati N, Vano Y, Freschi M, Becht E, Luciano R, et al. PD-L1expression and CD8þ T cell infiltrate are associated with clinical progres-sion in patients with node positive prostate cancer. Eur Urol Focus 2017;doi.org/10.1016/j.euf.2017.05.013.

7. Clemente CG, Mihm MC Jr, Bufalino R, Zurrida S, Collini P, Casci-nelli N. Prognostic value of tumor infiltrating lymphocytes in thevertical growth phase of primary cutaneous melanoma. Cancer 1996;77:1303–10.

8. Hui E, Cheung J, Zhu J, Su X, Taylor MJ, Wallweber HA, et al. T cellcostimulatory receptor CD28 is a primary target for PD-1-mediated inhi-bition. Science 2017;355:1428–33.

9. Greenwald RJ, Freeman GJ, Sharpe AH. The B7 family revisited. Annu RevImmunol 2005;23:515–48.

10. Walunas TL, Lenschow DJ, Bakker CY, Linsley PS, Freeman GJ, Green JM,et al. CTLA-4 can function as a negative regulator of T cell activation.Immunity 1994;1:405–13.

11. Krummel MF, Allison JP. CD28 and CTLA-4 have opposing effects on theresponse of T cells to stimulation. J Exp Med 1995;182:459–65.

12. Miller AM, Lundberg K, Ozenci V, Banham AH, Hellstrom M, Egevad L,et al. CD4þCD25high T cells are enriched in the tumor and peripheralblood of prostate cancer patients. J Immunol 2006;177:7398–405.

13. Miracco C,Mourmouras V, Biagioli M, Rubegni P,Mannucci S,Monciatti I,et al. Utility of tumour-infiltrating CD25þFOXP3þ regulatory T cellevaluation in predicting local recurrence in vertical growth phase cutane-ous melanoma. Oncol Rep 2007;18:1115–22.

14. Sakaguchi S, Vignali DA, Rudensky AY, Niec RE, Waldmann H. Theplasticity and stability of regulatory T cells. Nat Rev Immunol 2013;13:461–7.

15. Sharma P, Hu-Lieskovan S, Wargo JA, Ribas A. Primary, adaptive,and acquired resistance to cancer immunotherapy. Cell 2017;168:707–23.

16. Beer TM, Kwon ED, Drake CG, Fizazi K, Logothetis C, Gravis G, et al.Randomized, double-blind, phase III trial of ipilimumab versus placebo inasymptomatic or minimally symptomatic patients with metastatic che-motherapy-naive castration-resistant prostate cancer. J Clin Oncol 2017;35:40–7.

17. Kwon ED, Drake CG, Scher HI, Fizazi K, Bossi A, van den Eertwegh AJ, et al.Ipilimumab versus placebo after radiotherapy in patients with metastaticcastration-resistant prostate cancer that had progressed after docetaxelchemotherapy (CA184–043): a multicentre, randomised, double-blind,phase 3 trial. Lancet Oncol 2014;15:700–12.

18. Chung AS, Lee J, FerraraN. Targeting the tumour vasculature: insights fromphysiological angiogenesis. Nat Rev Cancer 2010;10:505–14.

19. 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.

20. Bellone M, Calcinotto A. Ways to enhance lymphocyte trafficking intotumors and fitness of tumor infiltrating lymphocytes. Front Oncol2013;3:231.

21. BelloneM, Elia AR. Constitutive and acquired mechanisms of resistance toimmune checkpoint blockade in human cancer. Cytokine Growth FactorRev 2017.doi 10.1016/j.cytogfr.2017.06.002.

22. Curnis F, Sacchi A, Borgna L, Magni F, Gasparri A, Corti A. Enhancement oftumor necrosis factor alpha antitumor immunotherapeutic properties bytargeted delivery to aminopeptidase N (CD13). Nat Biotechnol 2000;18:1185–90.

23. Calcinotto A, Grioni M, Jachetti E, Curnis F, Mondino A, Parmiani G, et al.Targeting TNF-alpha to neoangiogenic vessels enhances lymphocyte infil-tration in tumors and increases the therapeutic potential of immunother-apy. J Immunol 2012;188:2687–94.

24. Manzo T, Sturmheit T, Basso V, Petrozziello E, Hess Michelini R, Riba M,et al. T cells redirected to a minor histocompatibility antigen instructintratumoral TNFalpha expression and empower adoptive cell therapyfor solid tumors. Cancer Res 2017;77:658–71.

25. GreenbergNM,DeMayo F, FinegoldMJ,MedinaD, TilleyWD, Aspinall JO,et al. Prostate cancer in a transgenic mouse. Proc Natl Acad Sci U S A1995;92:3439–43.

26. Bellone M, Cantarella D, Castiglioni P, Crosti MC, Ronchetti A, Moro M,et al. Relevance of the tumor antigen in the validation of three vaccinationstrategies for melanoma. J Immunol 2000;165:2651–6.

27. Curran MA, Montalvo W, Yagita H, Allison JP. PD-1 and CTLA-4 combi-nation blockade expands infiltrating T cells and reduces regulatory T andmyeloid cells within B16 melanoma tumors. Proc Natl Acad Sci U S A2010;107:4275–80.

28. Hurwitz AA, Foster BA, Kwon ED, Truong T, Choi EM, Greenberg NM,et al. Combination immunotherapy of primary prostate cancer in atransgenic mouse model using CTLA-4 blockade. Cancer Res 2000;60:2444–8.

29. Gao J, Ward JF, Pettaway CA, Shi LZ, Subudhi SK, Vence LM, et al.VISTA is an inhibitory immune checkpoint that is increased afteripilimumab therapy in patients with prostate cancer. Nat Med 2017;23:551–5.

30. Kavanagh B, O'Brien S, Lee D, Hou Y, Weinberg V, Rini B, et al. CTLA4blockade expands FoxP3þ regulatory and activated effectorCD4þT cells ina dose-dependent fashion. Blood 2008;112:1175–83.

31. Larkin J, Chiarion-Sileni V, Gonzalez R, Grob JJ, Cowey CL, Lao CD, et al.Combined nivolumab and ipilimumab or monotherapy in untreatedmelanoma. N Engl J Med 2015;373:23–34.

32. de Witte MA, Bendle GM, van den Boom MD, Coccoris M, SchellTD, Tevethia SS, et al. TCR gene therapy of spontaneous prostatecarcinoma requires in vivo T cell activation. J Immunol 2008;181:2563–71.

33. Camporeale A, Boni A, Iezzi G, Degl'Innocenti E, Grioni M, Mondino A,et al. Critical impact of the kinetics of dendritic cells activation on the invivo induction of tumor-specific T lymphocytes. Cancer Res 2003;63:3688–94.

34. HessMichelini R,Manzo T, Sturmheit T, Basso V, RocchiM, FreschiM, et al.Vaccine-instructed intratumoral IFN-gamma enables regression of autoch-thonousmouse prostate cancer in allogeneic T-cell transplantation. CancerRes 2013;73:4641–52.

Elia et al.





35. Degl'Innocenti E, Grioni M, Boni A, Camporeale A, Bertilaccio MT, FreschiM, et al. Peripheral T cell tolerance occurs early during spontaneousprostate cancer development and can be rescued by dendritic cell immu-nization. Eur J Immunol 2005;35:66–75.

36. Paulos CM, Wrzesinski C, Kaiser A, Hinrichs CS, Chieppa M, Cassard L,et al. Microbial translocation augments the function of adoptively trans-ferred self/tumor-specific CD8þ T cells via TLR4 signaling. J Clin Invest2007;117:2197–204.

37. Wrzesinski C, Paulos CM, Gattinoni L, Palmer DC, Kaiser A, Yu Z, et al.Hematopoietic stem cells promote the expansion and function ofadoptively transferred antitumor CD8 T cells. J Clin Invest 2007;117:492–501.

38. Ricupito A, Grioni M, Calcinotto A, Hess Michelini R, Longhi R,Mondino A, et al. Booster vaccinations against cancer are critical inprophylactic but detrimental in therapeutic settings. Cancer Res 2013;73:3545–54.

39. Rigamonti N, Bellone M. Prostate cancer, tumor immunity and a renewedsense of optimism in immunotherapy. Cancer Immunol Immunother2012;61:453–68.

40. Drake CG. Prostate cancer as amodel for tumour immunotherapy. Nat RevImmunol 2010;10:580–93.

41. Twyman-Saint Victor C, Rech AJ,Maity A, Rengan R, Pauken KE, Stelekati E,et al. Radiation and dual checkpoint blockade activate non-redundantimmune mechanisms in cancer. Nature 2015;520:373–7.

42. Allen E, Jabouille A, Rivera LB, Lodewijckx I, Missiaen R, Steri V, et al.Combined antiangiogenic and anti-PD-L1 therapy stimulates tumorimmunity through HEV formation. Sci Transl Med 2017;9doi 10.1126/scitranslmed.aak9679.

43. 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.

44. Nadal R, Amin A, Geynisman DM, Voss MH, Weinstock M, Doyle J, et al.Safety and clinical activity of vascular endothelial growth factor receptor(VEGFR)-tyrosine kinase inhibitors after programmed cell death 1 inhib-itor treatment in patients with metastatic clear cell renal cell carcinoma.Ann Oncol 2016;27:1304–11.

45. Schmittnaegel M, Rigamonti N, Kadioglu E, Cassara A, Wyser Rmili C,Kiialainen A, et al. Dual angiopoietin-2 and VEGFA inhibition elicitsantitumor immunity that is enhanced by PD-1 checkpoint blockade. SciTransl Med 2017;9doi 10.1126/scitranslmed.aak9670.

46. Proost P,Wuyts A, VanDamme J. Humanmonocyte chemotactic proteins-2 and -3: structural and functional comparison with MCP-1. J Leukoc Biol1996;59:67–74.

47. Wang T, Dai H, Wan N, Moore Y, Dai Z. The role for monocyte chemoat-tractant protein-1 in the generation and function ofmemoryCD8þ T cells. JImmunol 2008;180:2886–93.

48. TangH,Wang Y, Chlewicki LK, Zhang Y, Guo J, LiangW, et al. Facilitating Tcell infiltration in tumormicroenvironment overcomes resistance to PD-L1blockade. Cancer Cell 2016;29:285–96.

49. Parmiani G, Pilla L, Corti A, Doglioni C, Cimminiello C, BelloneM, et al. Apilot Phase I study combining peptide-based vaccination and NGR-hTNFvessel targeting therapy in metastatic melanoma. Oncoimmunology2014;3:e963406.

50. Corti A, Curnis F, Rossoni G, Marcucci F, Gregorc V. Peptide-mediatedtargeting of cytokines to tumor vasculature: the NGR-hTNF example.BioDrugs 2013;27:591–603.






2018;24:2171-2181. Published OnlineFirst February 28, 2018.Clin Cancer Res Angela Rita Elia, Matteo Grioni, Veronica Basso, et al. Blockers in Combination with Adoptive Cell TherapyTumor and Enhances Therapeutic Efficacy of Immune Checkpoint Targeting Tumor Vasculature with TNF Leads Effector T Cells to the

Updated version

10.1158/1078-0432.CCR-17-2210doi:

Access the most recent version of this article at:

Material

Supplementary

http://clincancerres.aacrjournals.org/content/suppl/2018/02/28/1078-0432.CCR-17-2210.DC1

Access the most recent supplemental material at:

Cited articles

http://clincancerres.aacrjournals.org/content/24/9/2171.full#ref-list-1

This article cites 46 articles, 19 of which you can access for free at:

Citing articles

http://clincancerres.aacrjournals.org/content/24/9/2171.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/24/9/2171To 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-17-2210

http://clincancerres.aacrjournals.org/content/suppl/2018/02/28/1078-0432.CCR-17-2210.DC1

http://clincancerres.aacrjournals.org/content/24/9/2171.full#ref-list-1

http://clincancerres.aacrjournals.org/content/24/9/2171.full#related-urls

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

mailto:[email protected]

http://clincancerres.aacrjournals.org/content/24/9/2171


targeting tumor vasculature with tnf leads effector t ...control antibodies (syriam hamster igg; bio...

Documents